Diabetes

Pain and translation and using AI to improve healthcare at an individual level

by Dana Lewis

I think differently from most people. Sometimes, this is a strength; and sometimes this is a challenge. This is noticeable when I approach healthcare encounters in particular: the way I perceive signals from my body is different from a typical person. I didn’t know this for the longest time, but it’s something I have been becoming more aware of over the years.

The most noticeable incident that brought me to this realization involved when I pitched head first off a mountain trail in New Zealand over five years ago. I remember yelling – in flight – help, I broke my ankle, help. When I had arrested my fall, clung on, and then the human daisy chain was pulling me back up onto the trail, I yelped and stopped because I could not use my right ankle to help me climb up the trail. I had to reposition my knee to help move me up. When we got up to the trail and had me sitting on a rock, resting, I felt a wave of nausea crest over me. People suggested that it was dehydration and I should drink. I didn’t feel dehydrated, but ok. Then because I was able to gently rest my foot on the ground at a normal perpendicular angle, the trail guides hypothesized that it was not broken, just sprained. It wasn’t swollen enough to look like a fracture, either. I felt like it hurt really bad, worse than I’d ever hurt an ankle before and it didn’t feel like a sprain, but I had never broken a bone before so maybe it was the trauma of the incident contributing to how I was feeling. We taped it and I tried walking. Nope. Too-strong pain. We made a new goal of having me use poles as crutches to get me to a nearby stream a half mile a way, to try to ice my ankle. Nope, could not use poles as crutches, even partial weight bearing was undoable. I ended up doing a mix of hopping, holding on to Scott and one of the guides. That got exhausting on my other leg pretty quickly, so I also got down on all fours (with my right knee on the ground but lifting my foot and ankle in the air behind me) to crawl some. Eventually, we realized I wasn’t going to be able to make it to the stream and the trail guides decided to call for a helicopter evacuation. The medics, too, when they arrived via helicopter thought it likely wasn’t broken. I got flown to an ER and taken to X-Ray. When the technician came out, I asked her if she saw anything obvious and whether it looked broken or not. She laughed and said oh yes, there’s a break. When the ER doc came in to talk to me he said “you must have a really high pain tolerance” and I said “oh really? So it’s definitely broken?” and he looked at me like I was crazy, saying “it’s broken in 3 different places”. (And then he gave me extra pain meds before setting my ankle and putting the cast on to compensate for the fact that I have high pain tolerance and/or don’t communicate pain levels in quite the typical way.)

A week later, when I was trying not to fall on my broken ankle and broke my toe, I knew instantly that I had broken my toe, both by the pain and the nausea that followed. Years later when I smashed another toe on another chair, I again knew that my toe was broken because of the pain + following wave of nausea. Nausea, for me, is apparently a response to very high level pain. And this is something I’ve carried forward to help me identify and communicate when my pain levels are significant, because otherwise my pain tolerance is such that I don’t feel like I’m taken seriously because my pain scale is so different from other people’s pain scales.

Flash forward to the last few weeks. I have an autoimmune disease causing issues with multiple areas of my body. I have some progressive slight muscle weakness that began to concern me, especially as it spread to multiple limbs and areas of my body. This was followed with pain in different parts of my spine which has escalated. Last weekend, riding in the car, I started to get nauseous from the pain and had to take anti-nausea medicine (which thankfully helped) as well as pain medicine (OTC, and thankfully it also helped lower it down to manageable levels). This has happened several other times.

Some of the symptoms are concerning to my healthcare provider and she agreed I should probably have a MRI and a consult from neurology. Sadly, the first available new patient appointment with the neurologist I was assigned to was in late September. Gulp. I was admittedly nervous about my symptom progression, my pain levels (intermittent as they are), and how bad things might get if we are not able to take any action between now and September. I also, admittedly, was not quite sure how I would cope with the level of pain I have been experiencing at those peak moments that cause nausea.

I had last spoken to my provider a week prior, before the spine pain started. I reached out to give her an update, confirm that my specialist appointment was not until September, and express my concern about the progression and timeline. She too was concerned and I ended up going in for imaging sooner.

Over the last week, because I’ve been having these progressive symptoms, I used Katie McCurdy’s free templates from Pictal Health to help visualize and show the progression of symptoms over time. I wasn’t planning on sending my visuals to my doctor, but it helped me concretely articulate my symptoms and confirm that I was including everything that I thought was meaningful for my healthcare providers to know. I also shared them with Scott to confirm he didn’t think I had missed anything. The icons in some cases were helpful but in other cases didn’t quite match how I was experiencing pain and I modified them somewhat to better match how I saw the pain I was experiencing.

(PS – check out Katie’s templates here, you can make a copy in Google Drive and use them yourself!)

As I spoke with the nurse who was recording my information at intake for imaging, she asked me to characterize the pain. I did and explained that it was probably usually a 7/10 then but periodically gets stronger to the point of causing nausea, which for me is a broken bone pain-level response. She asked me to characterize the pain – was it burning, tingling…? None of the words she said matched how it feels. It’s strong pain; it sometimes gets worse. But it’s not any of the words she mentioned.

When the nurse asked if it was “sharp”, Scott spoke up and explained the icon that I had used on my visual, saying maybe it was “sharp” pain. I thought about it and agreed that it was probably the closest word (at least, it wasn’t a hard no like the words burning, tingling, etc. were), and the nurse wrote it down. That became the word I was able to use as the closest approximation to how the pain felt, but again with the emphasis of it periodically reaching nausea-inducing levels equivalent to broken bone pain, because I felt saying “sharp” pain alone did not characterize it fully.

This, then, is one of the areas where I feel that artificial intelligence (AI) gives me a huge helping hand. I often will start working with an LLM (a large language model) and describing symptoms. Sometimes I give it a persona to respond as (different healthcare provider roles); sometimes I clarify my role as a patient or sometimes as a similar provider expert role. I use different words and phrases in different questions and follow ups; I then study the language it uses in response.

If you’re not familiar with LLMs, you should know it is not human intelligence; there is no brain that “knows things”. It’s not an encyclopedia. It’s a tool that’s been trained on a bajillion words, and it learns patterns of words as a result, and records “weights” that are basically cues about how those patterns of words relate to each other. When you ask it a question, it’s basically autocompleting the next word based on the likelihood of it being the next word in a similar pattern. It can therefore be wildly wrong; it can also still be wildly useful in a lot of ways, including this context.

What I often do in these situations is not looking for factual information. Again, it’s not an encyclopedia. But I myself am observing the LLM in using a pattern of words so that I am in turn building my own set of “weights” – meaning, building an understanding of the patterns of words it uses – to figure out a general outline of what is commonly known by doctors and medical literature; the common terminology that is being used likely by doctors to intake and output recommendations; and basically build a list of things that do and do not match my scenario or symptoms or words, or whatever it is I am seeking to learn about.

I can then learn (from the LLM as well as in person clinical encounters) that doctors and other providers typically ask about burning, tingling, etc and can make it clear that none of those words match at all. I can then accept from them (or Scott, or use a word I learned from an LLM) an alternative suggestion where I’m not quite sure if it’s a perfect match, but it’s not absolutely wrong and therefore is ok to use to describe somewhat of the sensation I am experiencing.

The LLM and AI, basically, have become a translator for me. Again, notice that I’m not asking it to describe my pain for me; it would make up words based on patterns that have nothing to do with me. But when I observe the words it uses I can then use my own experience to rule things in/out and decide what best fits and whether and when to use any of those, if they are appropriate.

Often, I can do this in advance of a live healthcare encounter. And that’s really helpful because it makes me a better historian (to use clinical terms, meaning I’m able to report the symptoms and chronology and characterization more succinctly without them having to play 20 questions to draw it out of me); and it saves me and the clinicians time for being able to move on to other things.

At this imaging appointment, this was incredibly helpful. I had the necessary imaging and had the results at my fingertips and was able to begin exploring and discussing the raw data with my LLM. When I then spoke with the clinician, I was able to better characterize my symptoms in context of the imaging results and ask questions that I felt were more aligned with what I was experiencing, and it was useful for a more efficient but effective conversation with the clinician about what our working hypothesis was; what next short-term and long-term pathways looked like; etc.

This is often how I use LLMs overall. If you ask an LLM if it knows who Dana Lewis is, it “does” know. It’ll tell you things about me that are mostly correct. If you ask it to write a bio about me, it will solidly make up ⅓ of it that is fully inaccurate. Again, remember it is not an encyclopedia and does not “know things”. When you remember that the LLM is autocompleting words based on the likelihood that they match the previous words – and think about how much information is on the internet and how many weights (patterns of words) it’s been able to build about a topic – you can then get a better spidey-sense about when things are slightly more or less accurate at a general level. I have actually used part of a LLM-written bio, but not by asking it to write a bio. That doesn’t work because of made up facts. I have instead asked it to describe my work, and it does a pretty decent job. This is due to the number of articles I have written and authored; the number of articles describing my work; and the number of bios I’ve actually written and posted online for conferences and such. So it has a lot of “weights” probably tied to the types of things I work on, and having it describe the type of work I do or am known for gets pretty accurate results, because it’s writing in a general high level without enough detail to get anything “wrong” like a fact about an award, etc.

This is how I recommend others use LLMs, too, especially those of us as patients or working in healthcare. LLMs pattern match on words in their training; and they output likely patterns of words. We in turn as humans can observe and learn from the patterns, while recognizing these are PATTERNS of connected words that can in fact be wrong. Systemic bias is baked into human behavior and medical literature, and this then has been pattern-matched by the LLM. (Note I didn’t say “learned”; but they’ve created weights based on the patterns they observe over and over again). You can’t necessarily course-correct the LLM (it’ll pretend to apologize and maybe for a short while adjust it’s word patterns but in a new chat it’s prone to make the same mistakes because the training has not been updated based on your feedback, so it reverts to using the ‘weights’ (patterns) it was trained on); instead, we need to create more of the correct/right information and have it voluminously available for LLMs to train on in the future. At an individual level then, we can let go of the obvious not-right things it’s saying and focus on what we can benefit from in the patterns of words it gives us.

And for people like me, with a high (or different type of) pain tolerance and a different vocabulary for what my body is feeling like, this has become a critical tool in my toolbox for optimizing my healthcare encounters. Do I have to do this to get adequate care? No. But I’m an optimizer, and I want to give the best inputs to the healthcare system (providers and my medical records) in order to increase my chances of getting the best possible outputs from the healthcare system to help me maintain and improve and save my health when these things are needed.

TLDR: LLMs can be powerful tools in the hands of patients, including for real-time or ahead-of-time translation and creating shared, understandable language for improving communication between patients and providers. Just as you shouldn’t tell a patient not to use Dr. Google, you should similarly avoid falling into the trap of telling a patient not to use LLMs because they’re “wrong”. Being wrong in some cases and some ways does not mean LLMs are useless or should not be used by patients. Each of these tools has limitations but a lot of upside and benefits; restricting patients or trying to limit use of tools is like limiting the use of other accessibility tools. I spotted a quote from Dr. Wes Ely that is relevant: “Maleficence can be created with beneficent intent”. In simple words, he is pointing out that harm can happen even with good intent.

Don’t do harm by restricting or recommending avoiding tools like LLMs.

Being a raccoon and living with chronic disease

by Dana Lewis

Being a raccoon loading a dishwasher is a really useful analogy for figuring out: what you want to spend a lot of effort and precision on, where you can lower your effort and precision and still obtain reasonable outcomes, or where you can allow someone else to step in and help you when you don’t care how as long as the job gets done.

Huh? Raccoons?!

A few years ago Scott and I spotted a meme/joke going around that in every relationship there is a person who loads the dishwasher precisely (usually “stacks the dishwasher like a Scandinavian architect”) and one who loads the dishwasher like a “raccoon on meth” or a “rabid raccoon” or similar.

Our relationship and personality with dishwasher loading isn’t as opposite on the spectrum as that analogy suggests. However, Scott has a strong preference for how the dishwasher should be loaded, along with a high level of precision in achieving it. I have a high level of precision, but very low preference for how it gets done. Thus, we have evolved our strategy where I put things in and he re-arranges them. If I put things in with a high amount of effort and a high level of precision? He would rearrange them ANYWAY. So there is no point in me also spending high levels of effort to apply a first style of precision when that work gets undone. It is more efficient for me to put things in, and then he re-organizes as he sees fit.

Thus, I’ve embraced being the ‘raccoon’ that loads the dishwasher in this house. (Not quite as dramatic as some!)

A ChatGPT-created illustration of a cute raccoon happily loading the dishwasher, which looks fine but not precisely loaded.This came to mind because he went on a work trip, and I stuck things in the dishwasher for 2 days, and jokingly texted him to “come home and do the dishes that the raccoon left”. He came home well after dinner that night, and the next day texted when he opened the dishwasher for the first time that he “opened the raccoon cage for the first time”. (LOL).

Over the years, we’ve found other household tasks and chores where one of us has strong preferences about the way things should be done and the other person has less strong preferences. Similarly, there are some things that feel high-effort (and not worth it) for one of us but not the other. Over time, we’ve sorted tasks so things that feel high-effort can be done by the person for which it doesn’t feel high-effort, and depending on the preference level determines ‘how’ it gets done. But usually, the person who does it (because it’s low-effort) gets to apply their preferences, unless it’s a really weak preference and the preference of the non-doer doesn’t require additional effort.

Here are some examples of tasks and how our effort/preference works out. You can look at this and see that Scott ends up doing the dishwasher organization (after I load it like a raccoon) before starting the dishwasher and also has stronger preferences about laundry than I do. On the flip side, I seem to find it easier with routines for staying on top of household supply management including buying/re-ordering and acquiring and putting those away where we have them ready to go, because they’re not on a clear scheduled cadence. Ditto for managing the cats’ health via flea/tick medication schedules, scheduling and taking them to the vet, signing them up for cat camp when we travel, coordinating with the human involved in their beloved cat camp, etc. We end up doing a mix of overall work, split between the two of us.

A four-quadrant grid. Across the top it says "Effort" with low effort on the left and high effort on the right. Along the side it says "preference" with weak preference on the bottom and strong preference at the top. The implication is you can have a mix of preference and how much work certain chores are. Usually the person in the top left quadrant for a particular chore - representing easy or lower effort anad stronger preference - ends up doing that chore. For me that's household supply ordering; managing cat vet appts, etc. where due to Scott's much stronger preference his include the dishwasher, laundry, etc.
Could each of us do those tasks? Sure, and sometimes we do. But we don’t have to each do all of them, all the time, and we generally have a split list of who does which type of things as the primary doer.

Raccoons and burnout with chronic diseases

I have now lived with type 1 diabetes for almost 22 years. When I met Scott, I had been living with diabetes for 11 years. When he asked on one of our early dates what he could do to help, my answer was: “…nothing?” I’m an adult, and I’ve successfully managed my diabetes solo for decades.

Obviously, we ended up finding various ways for him to help, starting with iterating together on technological solutions for remote monitoring (DIYPS) to eventually closing the loop with an automated insulin delivery system (OpenAPS). But for the longest time, I still did all the physical tasks of ordering supplies, physically moving them around, opening them, managing them, etc. both at a 3-month-supply order level and also every 3 days with refilling reservoirs and changing pump sites and sensors.

Most of the time, these decades-long routines are literally routines and I do them without thought, the same way I put on my shoes before I leave the house. Yet when burnout is approaching – often from a combination of having five autoimmune diseases or having a lot of life going on while also juggling the ‘routine’ tasks that are voluminous – these can start to feel harder than they should.

Should, being the key word here.

Scott would offer to do something for me and I would say no, because I felt like I “should” do it because I normally can/am able to with minimal effort. However, the activation energy required (because of burnout or volume of other tasks) sometimes changes, and these minimal, low-effort tasks suddenly feel high-effort. Thus, it’s a good time to examine whether someone can – even in the short term and as a one-off – help.

It’s hard, though, to eradicate the “should”. I “should” be able to do X, I “should” be able to handle Y. But honestly? I should NOT have to deal with all the stuff and management of living with 5 autoimmune diseases and juggling them day in and day out. But I do have to deal with these and therefore do these things to stay in optimal health. “Should” is something that I catch myself thinking and now use that as a verbal flag to say “hey, just because I CAN do this usually doesn’t mean I HAVE to do it right now, and maybe it’s ok to take a break from always doing X and let Scott do X or help me with Y.”

Some of these “I should do it” tasks have actually become tasks that I’ve handed off long-term to Scott, because they’re super low effort for him but they’re mildly annoying for me because I have roughly 247 other tasks to deal with (no, I didn’t count them: that would make it 248).

For example, one time I asked him to open my shipment with 3 months worth of pump supplies, and unbox them so I could put them away. He also carried them into the room where we store supplies and put them where they belonged. Tiny, but huge! Only 246 tasks left on my list. Now, I order supplies, and he unboxes and puts them away and manages the inventory rotation: putting the oldest boxes on top (that I draw from first) and newer ones on bottom. This goes for pump supplies, CGM supplies, and anything else mail order like that.

A similar four quadrant chart with the same axes as the other graph, with effort on top (low left; high right) and execution preference (weak bottom, strong top). Similar to chores, we look at how our preferences and how much work it feels like, relative to each other, to decide if there are any tasks I can ask Scott to take on related to chronic disease management (like opening boxes and rotating stock of supplies being lower effort for him than me, due to my overall volume of tasks being higher)

This isn’t always as straightforward, but there are a lot of things I have been doing for 20+ years and thus find very low effort once the supplies are in my hand, like changing my pump site and CGM. So I do those. (If I was incapacitated, I have no doubt Scott could do those if needed.) But there’s other stuff that’s low effort and low preference like the opening of boxes and arranging of supplies that I don’t have to do and Scott is happy to take on to lower my task list of 247 things so that I only have about 240 things left to do for routine management.

Can I do them? Sure. Should I do them? Well, again, I can but that doesn’t mean I have to if there’s someone who is volunteering to help.

And sometimes that help is really useful in breaking down tasks that are USUALLY low effort – like changing a pump site – but become high effort for psychological reasons. Sometimes I’ll say out loud that I need to change my pump site, but I don’t want to. Some of that might be burnout, some of that is the mental energy it takes to figure out where to put the next pump site (and remembering the last couple of placements from previous sites, so I rotate them), combined with the physical activation energy to get up from wherever I am and go pull out the supplies to do it. In these cases, divide and conquer works! Scott often is more than happy to go and pull out a pump site and reservoir and place it where it’s convenient for me when I do get up to go do something else. For me, I often do pump site changes (putting a new one on, but I keep the old one on for a few extra hours in case the new one works) after my shower, so he’ll grab a pump site and reservoir and set it on the bathroom counter. Barrier removed. Then I don’t have to get up now and do it, but I also won’t forget to do it because it’s there in flow with my other tasks to do after my shower.

A gif showing a similar four quadrant graph (effort across top, execution preference along the side), showing a task going from the top left (low effort usually, strong preference for how it is done) moving to the right (high effort and still high preference), then showing it being split into two halves, one of which becomes a Scott task because it's lower effort sub-tasks and the remaining part is still high preference for me but has lowered the effort it takes.

There are a lot of chronic disease-related tasks like this that when I’m starting to feel burnout from the sheer number of tasks, I can look for (or sometimes Scott can spot) opportunities) to break a task into multiple steps and do them at different times, or to have someone do the task portions they can do, like getting out supplies. That then lowers the overall effort required to do that task, or lowers the activation energy depending on the task. A lot of these are simple-ish tasks, like opening something, getting something out and moving it across the house to a key action spot (like the bathroom counter for after a shower), or putting things away when they no longer need to be out. The latter is the raccoon-style approach. A lot of times I’ll have the activation energy to start and do a task, but not complete (like breaking down supply boxes for recycling). I’ll set them aside to do later, or Scott will spot this ‘raccoon’ stash of tasks and tackle it when he has time/energy, usually faster than me getting around to do it because he’s not burdened with 240 other tasks like that. (He does of course have a larger pile of tasks than without this, but the magnitude of his task list is a lot smaller, because 5 autoimmune diseases vs 0.)

Be a friend to your friend who needs to be a raccoon some of the time

I am VERY lucky to have met & fallen in love & married someone who is so incredibly able and willing to help. I recognize not everyone is in this situation. But there may be some ways our friends and family who don’t live with us can help, too. I had a really fantastic example of this lately where someone who isn’t Scott stepped up and made my raccoon-life instantly better before I even got to the stage of being a raccoon about it.

I have a bunch of things going on currently, and my doctor recommended that I have an MRI done. I haven’t had an MRI in years and the last one was pre-pandemic. Nowadays, I am still masking in any indoor spaces including healthcare appointments, and I plan to mask for my MRI. But my go-to n95 mask has metal in the nose bridge, which means I need to find a safe alternative for my upcoming MRI.

I was busy trying to schedule appointments and hadn’t gotten to the stage of figuring out what I would wear as an alternative for my MRI. But I mentioned to a friend that I was going to have an MRI and she asked what mask I was going to wear, because she knows that I mask for healthcare appointments. I told her I hadn’t figured it out yet but needed to eventually figure it out.

She instantly sprang into action. She looked up options for MRI-safe masks and asked a local friend who uses a CAN99 mask without wire whether the friend had a spare for me to try. She also ordered a sample pack of another n95 mask style that uses adhesive to stick to the face (and thus doesn’t have a metal nose bridge piece). She ordered these, collected the CAN99 from the local friend, and then told me when they’d be here, which was well over a week before I would need it for the MRI and offered to bring them by my house so I had them as soon as possible.

Meanwhile, I was gobsmacked with relief and appreciation because I would have been a hot mess of a raccoon trying to get around to sorting that out days or a week after she had sorted a variety of options for me to try. Instead, she predicted my raccoon-ness or otherwise was being a really amazing friend and stepping up to take something off my plate so I had one less thing to deal with.

Yay for helpers. In this case, she knew exactly what was needed. But a lot of times, we have friends or family who want to help but don’t know how to or aren’t equipped with the knowledge of what would be helpful. Thus, it’s useful – when you have energy – to think through how you could break apart tasks and what you could offer up or ask as a task for someone else to do that would lower the burden for you.

That might be virtual tasks or physical tasks:

  • It might be coming over and taking a bunch of supplies out of boxes (or medicine) and splitting them up and helping put them in piles or all the places those things need to go
  • It could be researching safe places for you to eat, if you have food allergies or restrictions or things like celiac
  • It could be helping divvy up food into individual portions or whatever re-sizing you need for whatever purpose
  • It could be researching and brainstorming and identifying some safe options for group activities, e.g. finding places with outdoor dining or cool places to walk and hike that suits everyone’s abilities and interests

Sometimes it’s the physical burden that it’s helpful to lift; sometimes it’s the mental energy burden that is helpful to lift; sometimes a temporary relief in all the things we feel like we have to do ourselves is more important than the task itself.

If you have a chronic disease, it’s ok to be a raccoon. There is no SHOULD.

Part of the reason I really like the raccoon analogy is because now instead of being annoyed at throwing things in the dishwasher, because whether I exert energy or not Scott is going to re-load it his way anyway, I put the dishes in without much precision and giggle about being a raccoon.

The same goes for chronic disease related-tasks. Even for tasks where Scott is not involved, but I’m starting to feel annoyed at something I need to do, I find ways to raccoon it a little bit. I change my pump site but leave the supplies on the counter because I don’t HAVE to put those away at the same moment. I usually do, but I don’t HAVE to. And so I raccoon it a bit and put the supplies away later, because it doesn’t hurt anyone or anything (including me) for those not to get put away at the same time. And that provides a little bit of comedic relief to me and lightens the task of changing my pump site.

It also helps me move away from the SHOULD weighing heavily in my brain. I should be able to get all my pump site stuff out, change it, and throw away and put away the supplies when done. It shouldn’t be hard. No one else has this challenge occasionally (or so my brain tells me).

But the burden isn’t about that task alone. It’s one task in the list of 247 things I’m doing every day to take care of myself. And sometimes, my list GROWS. January 18, my list was about 212 things I needed to do. Beginning January 19, my list jumped up to 230. Last week, it grew again. I have noticed this pattern that when my list of things to do grows, some of the existing “easy” tasks that I’ve done for 20 years suddenly feel hard. Because it is hard to split my energy across more tasks and more things to focus on; it takes time to adapt. And so being a raccoon for some of those tasks, for some of the time, provides a helpful steam-valve to output some of the challenges I’m juggling of dealing with all the tasks, because those tasks 100% don’t have to be done in the same way as I might do during “calm” static times where my task list hasn’t expanded suddenly.

And it doesn’t matter what anyone else does or what they care about. Thus, remove the should. You should be able to do this, sure – if you weren’t juggling 246 other things. But you do have 246 things and that blows apart the “should”.

Free yourself of the “should” wherever possible, and be a raccoon wherever it helps.

How to Exercise When Exercise Is Harder Than Your Normal

by Dana Lewis

I’ve been spending a lot of time thinking lately about how to optimize exercise and physical activity when your body doesn’t do what it’s supposed to do (or what you want it to do). We don’t always have control over our bodies; whereas we do, sometimes, have control over our actions and what we can try to do and how we do physical activity. A lot of my strategies for optimizing exercise and physical activity have actually been updating my mental models, and I think they might be useful to other people, too.

But first, let me outline a couple of scenarios and how they differ so we have a shared framework for discussing some of the mental strategies for incorporating activity and exercise into life with chronic diseases like autoimmune diseases.

Let’s imagine you’re running and you come to a cliff.

  • In scenario A, there’s a bridge across to the other side at the same level. It’s no big deal to continue running across and continue on your way.
  • In scenario B, there’s no bridge, and you tumble off the cliff, but then you are able to (eventually) work your way back up to the other side at the same level as the person who could just stroll across the bridge.
  • In scenario C, there’s no bridge but the cliff isn’t as steep of a drop off; instead, it’s like a 2% railroad grade trail sloping away and down. You continue down it, but you end up well below the other side where a bridge would’ve connected, and there’s no way up to that level. The longer you go, the farther you are from that level.
  • In scenario D, there is a cliff that you fall off of, and you pick yourself up and keep going but it’s on that 2% railroad grade sloping away and down. Like scenario C, you end up well below – and even farther below – where you would have been if the bridge had been there.

Illustration of a runner crossing a bridge; running up a slope after the trail drops first then returns to the same height (B); running down a slope that takes them below the target height (C); and a combination of a sharp drop then slope down (D), as explained in more words throughout the blog post.

This is basically illustrative of the different types of situations you can find yourself in related to health status.

  • If all is well, you’re in scenario A: no bumps in the road, you just carry on.
  • Scenario B is like when you have a short-term injury or accident (like breaking your ankle or a toe) where you have a sudden drop in ability but you are able to build back up to the level you were at before. It may take longer and feel like a hard slog, but usually you can get there.
  • Scenario C is when you have a chronic disease (or are experiencing aging over time) where there’s small changes in the situation or in your ability. Because of these factors, you end up below where you maybe would like to be.
  • Scenario D is when there’s an acute situation that triggers or results in a significant, sudden drop followed by a chronic state that mimics the downward 2% small change slope that adds up significantly over time, meaning you are well below compared to where you would like to be.

My personal experiences and living in Scenario D

I have dealt with scenario B via a broken ankle and a broken toe in past years. Those stink. They’re hard. But they’re a different kind of hard than scenario C and scenario D, where I’ve found myself in the last few years and more acutely, I now am clearly operating in scenario D: I have had an acute drop-off in lung function and have autoimmune diseases that are affecting my ability to exercise, especially as compared to “before”. In fact, I keep having cycles of scenario D where my VO2 max drops off a cliff (losing a full point or more) within 2-3 days, then plateaus at the low level during the length of that round of symptoms, before maybe responding to my efforts to bring it back up. And it doesn’t always go back up or respond to exercise the way it used to do, “before”, because well, my lungs don’t work like they used to.

It’s been pretty frustrating. I want to keep building on the hard work I’ve put into my last 2-3 years of ultrarunning. Last year around this time, I ran a personal best 100k (62 miles) and beat my brother-in-law’s 100k time. I’m pretty proud of that because I’m pretty slow; but in ultras if you pace well and fuel well, you can beat faster runners. (As opposed to much shorter distances where speed matters more!).

This year, however, I can barely trek out – on the best day – for a 4 mile run. I had originally envisioned that, due to my fitness level and cumulative mileage build up, I would be able to train for and run a fast marathon (26.2 miles / ~42k) this spring, and that was supposed to be what I was training for. (Fast being “fast for me”.) But instead of running ~30-40 miles a week, I have been running 8-16 miles per week and have only clocked in half of the total mileage I had done by this point last year. Argh. I didn’t expect to do as much overall, but 210 instead of 420 miles by the beginning of April shows how different it’s been and how limited I have been. I’ve dropped the scheduled plan for marathon training – or any hopes of ultra training this year, unless something changes drastically in a positive way that I’m not expecting.

I finally realized that comparing my abilities to “before” is the crux of a lot of my angst. It is a little hard when you realize over time (scenario C) that you can’t do something that you think you should be able to. For example, me trying to run fast: it just has never worked the way training to run fast seems to work for other people. Eventually, in “before times”, I had settled into a strategy of running far, but doing so more slowly, and that’s turned out to be way more fun for me. But when you have an acute adjustment in ability that isn’t like scenario B (e.g. you can expect to regain strength/function/ability over time), it’s really hard to wrap your brain around. And comparisons to ‘before’ feel inevitable. They’re probably part of the grieving process in recognizing that things have changed. But at some point, it’s helpful to recognize and picture that you ARE in scenario D. This includes grappling with and accepting the fact that something has changed; and you likely do not have control over it.

I have updated my mental model with some strategies, to help me frame and recognize that on bad days, I don’t have to push myself (even if deep down I want to, because I want to rebuild/gain fitness to where I “should” be) – and that I should save that strategy for “good” days.

Here’s what I’ve landed on, for general strategy approach, which applies to whatever activity that I ultimately choose for the day:

Overlapping circles of good days and bad days, showing that regardless of which day it is, I still go out every day. Strategies for 'bad' days include lowering expectations; changing activities; pacing slower; taking breaks; turning around; and not comparing to 'before'. Good/better days can involve a slow start but speed up or add distance if it feels good, as long as I pace/do it in a way that doesn't overdo it such that I can't be active as desired any following day.
The other thing, in addition to comparing distance, time and pacing to “before” abilities, that I have struggled with, is not having a training plan or schedule. Because my ‘good’ days (where my lungs do not seem to limit my activity) are unpredictable, I can’t build a training schedule and build up mileage/ability the way I used to. Ultimately, I have had to land on a strategy that I don’t like but accept is the most feasible one for now (suggested by Scott): have a “checklist” of activities for my ‘good days’, and have a checklist of activities for my ‘bad days’. This has helped me separate my before-desire for running being my primary activity (and thinking about my running ‘schedule’ that I wish I could go back to), and instead be more realistic on the day-of about what activities are ideal for the type of day I’m actually dealing with.

For example, on my worst days, I cannot run for 30 seconds without gasping for breath and any type of intensive activity (anything more than a really slow meandering walk or a few seconds of a really slow run) feels terrible. Walking feels yuck too but it’s tolerable when I go slow enough, even though my lungs still feel physically uncomfortable inside my rib cage. On medium bad days, I maybe can do a slow, easy, short run with 20 seconds run intervals; a walk; an easy super slow hike with lots of stopping; or an e-bike ride; or easy pace cross-country skiing (when it was winter). On good days? I can do anything! Which means I can hike more elevation at clippier paces (and I can actually push myself on pace) or run with some modicum of effort above a snail’s pace or run a snail’s pace that doesn’t hurt for 30 second intervals. Those are my favorite activities, so those are high on my list (depending on whether it’s the weekday or weekend) to try to do when I’m feeling good. On the bad days or less good days, I take whatever activity is available to me however I can get it.

Activity choice check list for really bad days (e.g. walk or easy e-bike) vs less bad days (slow, easy short run or very slow hike or easy ski) versus the better days where I can run, hike longer/faster, and ski any distance I want.
There are tons of activities so if you’re reading this, know that I’m making this list based on MY favorite types of activities (and the climate I live in). You should make your list of activities and sort them if it’s helpful, to know which ones bring joy even on the worst days and those are what you should prioritize figuring out how to do more of, as the days permit.

Some of this stuff maybe seems “duh” and super intuitive to a lot of people, especially if you’re not living in Scenario D. Hello to everyone in Scenario A! But, when you’ve been thrust off a metaphorical cliff into Scenario D, and there’s no way to do what you did “before”, figuring out how to pace and push yourself to regain what fitness you can OR preserve basic health functionality as long as you can…it’s all an experiment of balancing what amount of activity pushes you in a positive way and builds strength, fitness and health and balancing against going over the point where it causes short-term harm (to the point where it impedes your activity the following days) and/or long-term harm (e.g. further hurts your lungs or other body parts in a way that is either irreversible or hard to recover from).

The pep talk I wish I got that I’m giving to you now

Before I lived in Scenario D (lung stuff), I lived a lot in Scenario C: running with type 1 diabetes AND celiac AND Grave’s AND exocrine pancreatic insufficiency (which means I have to juggle glucose management while only eating gluten free and calculating and eating enzymes for any of that gluten free food I eat as fuel while running) was a lot to juggle, in of itself. I often thought about how much I was juggling while running along, while recognizing that a lot of that juggling was invisible to the outside. Which made me think and observe that even though I feel like every other runner was flying by me and not dealing with the exact same set of balls to juggle; some of those runners WERE probably juggling their own health stuff and limitations, too (or are parents juggling jobs and kid schedules and running, etc). Everyone’s got baggage that they’re carrying, of some sort, or are juggling things in a different way. So, juggling is hard. Kudos to everyone for getting out there for juggling with what they’ve got.

But especially now in Scenario D, it’s even more important to me that it’s not about being out there and running certain paces or hiking certain distances: it’s getting out there AT ALL which is the entire point. And I’ve made it my mission to try to compliment people for getting out there, when it feels like it’s appropriate to do so.

Last week, I was handed the perfect opportunity, and it turned out to be the best conversation I’ve had in a long time. A woman was coming uphill and commented that I had not forgotten my hiking poles like she had. I said yeah, they make a difference going downhill as well as up! She said something about huffing and puffing because she has asthma. DING DING: opportunity to celebrate her for being out there hiking uphill, even with asthma. (I pretty much said that and complimented her). She and Scott were trading comments about it being the beginning of hiking season and how they had forgotten their hiking poles and we told them we were making a list throughout the hike of everything else we had forgotten. They mentioned that they were 70 (wow!) and 75 (wow!) and so they didn’t think they needed walkie talkies because they would not separate on the trail (one of the things that we forgot to bring in case Scott mountain-goated-ahead of me on the trail at any point). We gave them our sincere compliments for being out there hiking (because, goals! I am aiming hard and working hard to get to the age of 70 and be able to hike like that!). She talked about it being hard because she has asthma and was struggling to breathe at first before she remembered to take her albuterol…and I pointed out that even if she was struggling and had to stop every few minutes, it didn’t matter: she was out there, she was hiking, and it doesn’t matter how long it takes! She thought that was the best thing to hear, but it was really what I try to tell myself because I love to hear it, too, which is celebrating going and not worrying about pace/slow/whatever. I told her I had a lung condition too (she’s the first stranger I’ve ever told) and she asked if I was stopping every 2 minutes and whether I had taken an inhaler. I explained most of my lung condition doesn’t respond to an inhaler but that yes, I too had to stop and catch my breath. But it was an awesome, gorgeous day and worth hiking in and that I was glad I had gone up. Ultimately, she said a lot of things that made it seem like my shared experience helped her – but in turn, seeing her and talking to her helped ME just as much, because it reminded me that yes, everyone else is juggling things while hiking too. And it’s really not about speed/pace/time; it’s absolutely about being out there and enjoying it.

So that’s what I’m trying to do: I’m trying to move beyond the comparison from what I did before, and simply compare to “am I going out at all and trying”. Trying = winning; going = winning, and that’s the new mental model that has been working really well for me as I spend more time in Scenario D.

—

PS – if you read this and are in a similar situation of Scenario B, C, or D and want a virtual high five and to feel “seen” for what you’re working through – feel free to comment here or email any time. I see you going out and trying; which means you’re winning! And I’m happy to give a virtual comment the way I am trying to give comments out on the trails and compliment folks for the process of being out moving through the world in all the ways that we can, however we can. 

Meet me in the gray area: beyond prevention, before progression

by Dana Lewis

Two things can simultaneously be true:

  • Doctors may wish they had more opportunities to help patients prevent having worse/later stage outcomes of a disease.
  • Doctors may struggle when a patient seeks health care at an early stage, asking for strategies and intervention support against developing worse/later stage outcomes of a disease.

The struggle may be for a few reasons. There’s often a lack of systemic infrastructure to support patients who show up earlier rather than later in a disease progression, especially when the frequency/timeline of care is much quicker than the system is currently resourced for. There’s often a lack of research for these earlier stages and what effective strategies are for preventing progression and treating earlier stage outcomes.

When a clinician struggles with this, it’s not a moral failing of the clinician if they don’t feel equipped to tackle those challenges. However, I do wish clinicians would more often clearly communicate to patients about these struggles. The patient might have a choice: do they pursue another clinician who might have different resources (including time/energy) or expertise in navigating the unknown? Or do they work with the existing clinician to navigate the murky waters together, figuring it out as they go? But patients only have a choice if they realize it themselves and are equipped to pursue alternative paths – or are told that this is a fork in the path.

The challenge is this is a gray area for all of us – patients and clinicians alike. But the reality is, the gray area (for a patient) betwixt and between prevention and progression is our life. The black and white that may emerge after the gray space can be as significant, literally, as life and death. We as patients are highly motivated to navigate the gray area and reduce suffering and possibly try different or new strategies that have shown early promise (although maybe haven’t yet been tested to RCT or the ideal standard, or in the specific disease or stage of the disease in question). We as patients may not have time to wait for the evidence to evolve further.

Clinicians may be aware of the gray space that the patient has landed in. The reality that many clinicians may not know or forget – or have slipped out of their mind – is that the gray space is even more daunting to face alone. If the instinct is to simply shoot down every patient idea with “that’s not approved for use in this disease” without forthrightly contextualizing against a recognition “there’s nothing tested or proven for this part/stage of the disease”, it can begin to put cracks in a relationship. What clinicians might not realize is that a patient may not have time to be in the gray space with a clinician who simply says no to trying anything, because no one has ever studied it before and when little study is being done at all about the gray area the patient is within. Or maybe clinicians do realize it, and sometimes rely on the power of the broken systemic infrastructure that keeps a patient from finding a clinician who does feel equipped to walk through the gray area with them.

What I wish is for clinicians to be equipped to identify this situation, standing on the edge of the gray area with a patient. And to say up front, then and there, if they don’t feel comfortable pursuing off-label strategies when there are zero documented on-label strategies beyond waiting for the worse outcomes to progress. I don’t like that (because why wait for permanent damage to do something, when permanent damage is not inevitable if action is taken sooner), but I very much highly respect and appreciate a clinician who is forthright and willing to say they don’t have time/energy/feel equipped to do so.

Why? Because if I’m already in the gray space, past prevention and before serious progression, it gives me a better opportunity to find someone else who can partner there. It might take a try or two, but it keeps me from wasting time and energy and trying to invest in developing a relationship with a clinician who has already decided they can’t help me until I cross over into black and white worse outcomes.

When we talk about prevention, it’s often about preventing a disease. In the world I live in, and the body I live in – now inhibited by five autoimmune diseases, I don’t have a choice about disease prevention for the most part. My body is clearly equipped with a superstar hyperactive immune system. While I’ve seen some research working on addressing autoimmune stuff, it’s likely decades away from any cure of any one condition that I have (let alone all of them) or fixing the hyperactivity of my immune system and preventing additional autoimmune diseases. Sure, I can work to prevent other diseases that aren’t autoimmune (exercising, staying in as best health overall that I can, etc.), but my focus right now is keeping each of my five autoimmune conditions from being bigger headaches than they already are.

(As a side note, I recently read this paper looking at rates of autoimmune conditions after T1D, based on a registry analysis in Sweden of people with T1D. It’s interesting that the risk of “one more” condition following T1D is 17%, two more is essentially the square of that, etc. etc. all the way down…so the risk is typically about 17% and it’s not additive; having two does not mean you’re more likely to get three, it means you have about the same 17% chance of something else. That’s a useful mental model to me, understanding that I got unlucky 4 more times…and that combination of luck is rare among people with T1D. They went all the way to the category of “three or more” autoimmune conditions after T1D, calculating that 0.3% of people with T1D have 3 or more autoimmune conditions after T1D. They stopped there, but you can extrapolate by multiplying by 17% again and estimate it’s 0.08% for four or more…which is where I’m at. This shows me that I’m not alone in dealing with so many things, but it puts me at about 1 in 1,250 of people with T1D or around one in a million – heh – in the general population if you extrapolate based on global population estimates and assume similar rates/risks of autoimmune conditions in the general public.)

Four of the five are easy enough (although, the fourth took about a year and a half to get to ‘easy enough’, overlapping with the third taking two to three years). The fifth, though, is the gray area that I currently inhabit. Possibly because I am in tune with my body because of the experience with these other autoimmune conditions, I have been presenting to the healthcare system to address this fifth autoimmune condition earlier than most people. Like many autoimmune conditions, it takes years to decades for some people to get diagnosed. Many are diagnosed after systemic manifestations have fully kicked in, e.g. these later stage worse outcomes I referred to above. I’m in a gray area, at the edge of seeing systemic activity, and able to identify it as a red flag, but before – I hope – permanent irreversible damage has been done. The question remains, however, for me to figure out how to navigate this gray area and with which clinicians, in order to achieve care that will possibly prevent or delay or reduce the severity of the outcomes that I will end up with.

I speak from personal experience with this gray area. It’s not fun to navigate, even if you do have a really great clinician partner. But it’s infinitely more challenging to stand there in the gray, unsure of the ability or willingness of a clinician to partner with you.

Meet me in the gray: beyond prevention, before progression - a blog written by Dana M. Lewis on DIYPS.org

A Slackbot for using Slack to access and use a chat-based LLM in public

by Dana Lewis

I’ve been thinking a lot about how to help my family, friends, and colleagues use LLMs to power their work. (As I’ve written about here, and more recently here with lots of tips on prompting and effectively using LLMs for different kinds of projects). 

Scott has been on the same page, especially thinking about how to help colleagues use LLMs effectively, but taking a slightly different approach: he built a Slackbot (a bot for Slack) which uses GPT-3.5 and GPT-4 to answer questions. This uses the API of GPT but presents it to the user in Slack instead of having to use ChatGPT as the chat interface. So, it’s a LLM chatbot, different than ChatGPT (because it’s a different chat interface), but uses the same AI (GPT-3.5 and GPT-4 from OpenAI). You could implement the same idea (a chat-based bot in Slack) using different AIs/LLMs, of course.

Using a slack-based bot for an LLM achieves a couple of things:

  1. More people can try GPT-4 and compare it to GPT-3.5 to get a taste for prompting and responses, without having to pay $20/month for a ChatGPT Pro account to get access to GPT-4.
  2. If you spend a lot of time in Slack for work, you don’t have to switch windows to access GPT.
  3. If your employer doesn’t want you to use the ChatGPT consumer product for work due to security concerns, but is more comfortable with OpenAI’s confidentiality guarantees for their API, a Slack bot interface provides a more secure alternative.
  4. You can see “in public” how other people in your Slack workspace are using it, and learn from their prompts and interactions new tricks and ideas (in addition to reading my long post here with lots of tips and suggestions).
  5. You can privately (in DM by yourself) talk to the bot and ask it your “dumb” questions (which aren’t dumb), and you can reason through things at your own speed or learn how to do things that you haven’t done before, and do so in private just like if you had your own ChatGPT account. It doesn’t burn social capital or take other people’s time, which you can save for more nuanced questions.

Scott has implemented it for work, but also came up with another fun use for it: we created a family-focused Slack channel for kids in our lives (we have ten (!) nieces and nephews – you can see how I used GPT to help with special custom books for some of them recently) so their parents and they can access GPT in a safe environment, too. ChatGPT (and OpenAI’s GPT API) have some reasonable safeguards in it already to protect kids, but more usefully we want to be able to provide custom instructions to keep it always at a kid-focused level. Plus this way we can share the experience with them, too.

I asked the bot to summarize itself for a lay audience not familiar with the project, and this was the output:

### Summary of the SlackAskBot

The SlackAskBot is a smart assistant designed to operate within Slack, a popular communication platform used by teams for messaging and collaboration. Imagine having a helpful robot in your group chats or direct messages that can understand your questions and provide informative responses. That’s what the SlackAskBot does. It’s like adding a knowledgeable team member to your Slack workspace who is always ready to assist.

#### How It Works:

  1. **Integration with Slack:**

– The bot is integrated into Slack, meaning it can read messages in channels it’s added to or when directly messaged. It’s like inviting a friend into your group chat who can listen and respond to conversations.

  1. **Understanding Messages:**

– When you type a message that includes the bot (either by mentioning it with an @ symbol in a channel or by sending it a direct message), the bot pays attention. It’s designed to understand that you’re talking to it and that you’re expecting some form of response.

  1. **Processing Your Requests:**

– Once the bot receives a message, it doesn’t just look at the latest message you’ve sent. It considers the context by looking at the conversation history. This way, it can provide responses that make sense within the ongoing discussion. It’s akin to a human joining a conversation, where understanding the previous discussion points is crucial for making relevant comments.

  1. **Generating Responses:**

– The bot uses a powerful language model (think of it as a very advanced AI that understands and generates human-like text) to come up with responses to your messages. It’s as if the bot thinks about what you’ve asked and then crafts a reply that aims to be helpful and informative.

  1. **Interacting Back:**

– After thinking through your message and generating a response, the bot then sends this response back into the Slack channel or direct message. You’ll see its reply just like you would from any other team member, providing the information or assistance you requested.

  1. **Special Features:**

– The bot is designed with specific settings for different channels or direct messages. This means it can tailor its responses based on where it’s talking. For example, it might respond differently in a casual chat compared to a more formal project discussion channel.

  1. **Respecting Privacy and Permissions:**

– The bot is programmed to respect Slack’s rules and the privacy of your conversations. It only listens and responds to messages where it’s explicitly mentioned or directly messaged, ensuring that it doesn’t intrude on private discussions.

#### In Summary:

The SlackAskBot acts as a virtual assistant within your Slack workspace, ready to provide information and answers to your questions. It’s designed to understand the context of conversations, generate helpful responses, and interact with users in a way that enhances team communication and productivity. Whether you’re asking for quick facts, seeking advice on a project, or just interacting for fun, the SlackAskBot is there to make your Slack experience more informative and engaging.

—
The SlackAskBot is open source – you can check out the code and instructions for how to implement it in your own Slack environment in the repository here.

We are using Slack as a chat interface for LLMs like GPT, making the user experience similar to ChatGPT

Effective Pair Programming and Coding and Prompt Engineering and Writing with LLMs like ChatGPT and other AI tools

by Dana Lewis

I’ve been puzzled when I see people online say that LLM’s “don’t write good code”. In my experience, they do. But given that most of these LLMs are used in chatbot mode – meaning you chat and give it instructions to generate the code – that might be where the disconnect lies. To get good code, you need effective prompting and to do so, you need clear thinking and ideas on what you are trying to achieve and how.

My recipe and understanding is:

Clear thinking + clear communication of ideas/request = effective prompting => effective code and other outputs

It also involves understanding what these systems can and can’t do. For example, as I’ve written about before, they can’t “know” things (although they can increasingly look things up) and they can’t do “mental” math. But, they can generally repeat patterns of words to help you see what is known about a topic and they can write code that you can execute (or it can execute, depending on settings) to solve a math problem.

What the system does well is help code small chunks, walk you through processes to link these sections of code up, and help you implement them (if you ask for it). The smaller the task (ask), the more effective it is. Or also – the easier it is for you to see when it completes the task and when it hasn’t been able to finish due to limitations like response length limits, information falling out of the context window (what it knows that you’ve told it); unclear prompting; and/or because you’re asking it to do things for which it doesn’t have expertise. Some of the last part – lack of expertise – can be improved with specific prompting techniques –  and that’s also true for right-sizing the task it’s focusing on.

Right-size the task by giving a clear ask

If I were to ask an LLM to write me code for an iOS app to do XYZ, it could write me some code, but it certainly wouldn’t (at this point in history, written in February 2024), write all code and give me a downloadable file that includes it all and the ability to simply run it. What it can do is start writing chunks and snippets of code for bits and pieces of files that I can take and place and build upon.

How do I know this? Because I made that mistake when trying to build my first iOS apps in April and May 2023 (last year). It can’t do that (and still can’t today; I repeated the experiment). I had zero ideas how to build an iOS app; I had a sense that it involved XCode and pushing to the Apple iOS App Store, and that I needed “Swift” as the programming language. Luckily, though, I had a much stronger sense of how I wanted to structure the app user experience and what the app needed to do.

I followed the following steps:

  1. First, I initiated chat as a complete novice app builder. I told it I was new to building iOS apps and wanted to use XCode. I had XCode downloaded, but that was it. I told it to give me step by step instructions for opening XCode and setting up a project. Success! That was effective.
  2. I opened a different chat window after that, to start a new chat. I told it that it was an expert in iOS programming using Swift and XCode. Then I described the app that I wanted to build, said where I was in the process (e.g. had opened and started a project in XCode but had no code yet), and asked it for code to put on the home screen so I could build and open the app and it would have content on the home screen. Success!
  3. From there, I was able to stay in the same chat window and ask it for pieces at a time. I wanted to have a new user complete an onboarding flow the very first time they opened the app. I explained the number of screens and content I wanted on those screens; the chat was able to generate code, tell me how to create that in a file, and how to write code that would trigger this only for new users. Success!
  4. I was able to then add buttons to the home screen; have those buttons open new screens of the app; add navigation back to the home; etc. Success!
  5. (Rinse and repeat, continuing until all of the functionality was built out a step at a time).

To someone with familiarity building and programming things, this probably follows a logical process of how you might build apps. If you’ve built iOS apps before and are an expert in Swift programming, you’re either not reading this blog post or are thinking I (the human) am dumb and inexperienced.

Inexperienced, yes, I was (in April 2023). But what I am trying to show here is for someone new to a process and language, this is how we need to break down steps and work with LLMs to give it small tasks to help us understand and implement the code it produces before moving forward with a new task (ask). It takes these small building block tasks in order to build up to a complete app with all the functionality that we want. Nowadays, even though I can now whip up a prototype project and iOS app and deploy it to my phone within an hour (by working with an LLM as described above, but skipping some of the introductory set-up steps now that I have experience in those), I still follow the same general process to give the LLM the big picture and efficiently ask it to code pieces of the puzzle I want to create.

As the human, you need to be able to keep the big picture – full app purpose and functionality – in mind while subcontracting with the LLM to generate code for specific chunks of code to help achieve new functionality in our project.

In my experience, this is very much like pair programming with a human. In fact, this is exactly what we did when we built DIYPS over ten years ago (wow) and then OpenAPS within the following year. I’ve talked endlessly about how Scott and I would discuss an idea and agree on the big picture task; then I would direct sub-tasks and asks that he, then also Ben and others would be coding on (at first, because I didn’t have as much experience coding and this was 10 years ago without LLMs; I gradually took on more of those coding steps and roles as well). I was in charge of the big picture project and process and end goal; it didn’t matter who wrote which code or how; we worked together to achieve the intended end result. (And it worked amazingly well; here I am 10 years later still using DIYPS and OpenAPS; and tens of thousands of people globally are all using open source AID systems spun off of the algorithm we built through this process!)

Two purple boxes. The one on the left says "big picture project idea" and has a bunch of smaller size boxes within labeled LLM, attempting to show how an LLM can do small-size tasks within the scope of a bigger project that you direct it to do. On the right, the box simply says "finished project". Today, I would say the same is true. It doesn’t matter – for my types of projects – if a human or an LLM “wrote” the code. What matters is: does it work as intended? Does it achieve the goal? Does it contribute to the goal of the project?

Coding can be done – often by anyone (human with relevant coding expertise) or anything (LLM with effective prompting) – for any purpose. The critical key is knowing what the purpose is of the project and keeping the coding heading in the direction of serving that purpose.

Tips for right-sizing the ask

  1. Consider using different chat windows for different purposes, rather than trying to do it all in one. Yes, context windows are getting bigger, but you’ll still likely benefit from giving different prompts in different windows (more on effective prompting below).Start with one window for getting started with setting up a project (e.g. how to get XCode on a Mac and start a project; what file structure to use for an app/project that will do XYZ; how to start a Jupyter notebook for doing data science with python; etc); brainstorming ideas to scope your project; then separately for starting a series of coding sub-tasks (e.g. write code for the home page screen for your app; add a button that allows voice entry functionality; add in HealthKit permission functionality; etc.) that serves the big picture goal.
  2. Make a list for yourself of the steps needed to build a new piece of functionality for your project. If you know what the steps are, you can specifically ask the LLM for that.Again, use a separate window if you need to. For example, if you want to add in the ability to save data to HealthKit from your app, you may start a new chat window that asks the LLM generally how does one add HealthKit functionality for an app? It’ll describe the process of certain settings that need to be done in XCode for the project; adding code that prompts the user with correct permissions; and then code that actually does the saving/revising to HealthKit.

    Make your list (by yourself or with help), then you can go ask the LLM to do those things in your coding/task window for your specific project. You can go set the settings in XCode yourself, and skip to asking it for the task you need it to do, e.g. “write code to prompt the user with HealthKit permissions when button X is clicked”.

    (Sure, you can do the ask for help in outlining steps in the same window that you’ve been prompting for coding sub-tasks, just be aware that the more you do this, the more quickly you’ll burn through your context window. Sometimes that’s ok, and you’ll get a feel for when to do a separate window with the more experience you get.)

  • Pay attention as you go and see how much code it can generate and when it falls short of an ask. This will help you improve the rate at which you successfully ask and it fully completes a task for future asks. I observe that when I don’t know – due to my lack of expertise – the right size of a task, it’s more prone to give me ½-⅔ of the code and solution but need additional prompting after that. Sometimes I ask it to continue where it cut off; other times I start implementing/working with the bits of code (the first ⅔) it gave me, and have a mental or written note that this did not completely generate all steps/code for the functionality and to come back.Part of why sometimes it is effective to get started with ⅔ of the code is because you’ll likely need to debug/test the first bit of code, anyway. Sometimes when you paste in code it’s using methods that don’t match the version you’re targeting (e.g. functionality that is outdated as of iOS 15, for example, when you’re targeting iOS 17 and newer) and it’ll flag a warning or block it from working until you fix it.

    Once you’ve debugged/tested as much as you can of the original ⅔ of code it gave you, you can prompt it to say “Ok, I’ve done X and Y. We were trying to (repeat initial instructions/prompt) – what are the remaining next steps? Please code that.” to go back and finish the remaining pieces of that functionality.

    (Note that saying “please code that” isn’t necessarily good prompt technique, see below).

    Again, much of this is paying attention to how the sub-task is getting done in service of the overall big picture goal of your project; or the chunk that you’ve been working on if you’re building new functionality. Keeping track with whatever method you prefer – in your head, a physical written list, a checklist digitally, or notes showing what you’ve done/not done – is helpful.

Most of the above I used for coding examples, but I follow the same general process when writing research papers, blog posts, research protocols, etc. My point is that this works for all types of projects that you’d work on with an LLM, whether the output generation intended is code or human-focused language that you’d write or speak.

But, coding or writing language, the other thing that makes a difference in addition to right-sizing the task is effective prompting. I’ve intuitively noticed that has made the biggest difference in my projects for getting the output matching my expertise. Conversely, I have actually peer reviewed papers for medical journals that do a horrifying job with prompting. You’ll hear people talk about “prompt engineering” and this is what it is referring to: how do you engineer (write) a prompt to get the ideal response from the LLM?

Tips for effective prompting with an LLM

    1. Personas and roles can make a difference, both for you and for the LLM. What do I mean by this? Start your prompt by telling the LLM what perspective you want it to take. Without it, you’re going to make it guess what information and style of response you’re looking for. Here’s an example: if you asked it what caused cancer, it’s going to default to safety and give you a general public answer about causes of cancer in very plain, lay language. Which may be fine. But if you’re looking to generate a better understanding of the causal mechanism of cancer; what is known; and what is not known, you will get better results if you prompt it with “You are an experienced medical oncologist” so it speaks from the generated perspective of that role. Similarly, you can tell it your role. Follow it with “Please describe the causal mechanisms of cancer and what is known and not known” and/or “I am also an experienced medical researcher, although not an oncologist” to help contextualize that you want a deeper, technical approach to the answer and not high level plain language in the response.

      Compare and contrast when you prompt the following:

      A. “What causes cancer?”

      B. “You are an experienced medical oncologist. What causes cancer? How would you explain this differently in lay language to a patient, and how would you explain this to another doctor who is not an oncologist?”

      C. “You are an experienced medical oncologist. Please describe the causal mechanisms of cancer and what is known and not known. I am also an experienced medical researcher, although not an oncologist.”

      You’ll likely get different types of answers, with some overlap between A and the first part of answer B. Ditto for a tiny bit of overlap between the latter half of answer B and for C.

      I do the same kind of prompting with technical projects where I want code. Often, I will say “You are an expert data scientist with experience writing code in Python for a Jupyter Notebook” or “You are an AI programming assistant with expertise in building iOS apps using XCode and SwiftUI”. Those will then be followed with a brief description of my project (more on why this is brief below) and the first task I’m giving it.

      The same also goes for writing-related tasks; the persona I give it and/or the role I reference for myself makes a sizable difference in getting the quality of the output to match the style and quality I was seeking in a response.

  • Be specific. Saying “please code that” or “please write that” might work, sometimes, but more often or not will get a less effective output than if you provide a more specific prompt.I am a literal person, so this is something I think about a lot because I’m always parsing and mentally reviewing what people say to me because my instinct is to take their words literally and I have to think through the likelihood that those words were intended literally or if there is context that should be used to filter those words to be less literal. Sometimes, you’ll be thinking about something and start talking to someone about something, and they have no idea what on earth you’re talking about because the last part of your out-loud conversation with them was about a completely different topic!

    LLMs are the same as the confused conversational partner who doesn’t know what you’re thinking about. LLMs only know what you’ve last/recently told it (and more quickly than humans will ‘forget’ what you told it about a project). Remember the above tips about brainstorming and making a list of tasks for a project? Providing a description of the task along with the ask (e.g. we are doing X related to the purpose of achieving Y, please code X) will get you better output more closely matching what you wanted than saying “please code that” where the LLM might code something else to achieve Y if you didn’t tell it you wanted to focus on X.

    I find this even more necessary with writing related projects. I often find I need to give it the persona “You are an expert medical researcher”, the project “we are writing a research paper for a medical journal”, the task “we need to write the methods section of the paper”, and a clear ask “please review the code and analyses and make an outline of the steps that we have completed in this process, with sufficient detail that we could later write a methods section of a research paper”. A follow up ask is then “please take this list and draft it into the methods section”. That process with all of that specific context gives better results than “write a methods section” or “write the methods” etc.

  • Be willing to start over with a new window/chat. Sometimes the LLM can get itself lost in solving a sub-task and lose sight (via lost context window) of the big picture of a project, and you’ll find yourself having to repeat over and over again what you’re asking it to do. Don’t be afraid to cut your losses and start a new chat for a sub-task that you’ve been stuck on. You may be able to eventually come back to the same window as before, or the new window might become your new ‘home’ for the project…or sometimes a third, fourth, or fifth window will.
  • Try, try again.
    I may hold the record for the longest running bug that I (and the LLM) could. Not. solve. This was so, so annoying. No users apparently noticed it but I knew about it and it bugged me for months and months. Every few weeks I would go to an old window and also start a new window, describe the problem, paste the code in, and ask for help to solve it. I asked it to identify problems with the code; I asked it to explain the code and unexpected/unintended functionality from it; I asked it what types of general things would be likely to cause that type of bug. It couldn’t find the problem. I couldn’t find the problem. Finally, one day, I did all of the above, but then also started pasting every single file from my project and asking if it was likely to include code that could be related to the problem. By forcing myself to review all my code files with this problem in mind, even though the files weren’t related at all to the file/bug….I finally spotted the problem myself. I pasted the code in, asked if it was a possibility that it was related to the problem, the LLM said yes, I tried a change and…voila! Bug solved on January 16 after plaguing me since November 8. (And probably existed before then but I didn’t have functionality built until November 8 where I realized it was a problem). I was beating myself up about it and posted to Twitter about finally solving the bug (but very much with the mindset of feeling very stupid about it). Someone replied and said “congrats! sounds like it was a tough one!”. Which I realized was a very kind framing and one that I liked, because it was a tough one; and also I am doing a tough thing that no one else is doing and I would not have been willing to try to do without an LLM to support.

    Similarly, just this last week on Tuesday I spent about 3 hours working on a sub-task for a new project. It took 3 hours to do something that on a previous project took me about 40 minutes, so I was hyper aware of the time mismatch and perceiving that 3 hours was a long time to spend on the task. I vented to Scott quite a bit on Tuesday night, and he reminded me that sure it took “3 hours” but I did something in 3 hours that would take 3 years otherwise because no one else would do (or is doing) the project that I’m working on. Then on Wednesday, I spent an hour doing another part of the project and Thursday whipped through another hour and a half of doing huge chunks of work that ended up being highly efficient and much faster than they would have been, in part because the “three hours” it took on Tuesday wasn’t just about the code but about organizing my thinking, scoping the project and research protocol, etc. and doing a huge portion of other work to organize my thinking to be able to effectively prompt the LLM to do the sub-task (that probably did actually take closer to the ~40 minutes, similar to the prior project).

    All this to say: LLMs have become pair programmers and collaborators and writers that are helping me achieve tasks and projects that no one else in the world is working on yet. (It reminds me very much of my early work with DIYPS and OpenAPS where we did the work, quietly, and people eventually took notice and paid attention, albeit slower than we wished but years faster than had we not done that work. I’m doing the same thing in a new field/project space now.) Sometimes, the first attempt to delegate a sub-task doesn’t work. It may be because I haven’t organized my thinking enough, and the lack of ideal output shows that I have not prompted effectively yet. Sometimes I can quickly fix the prompt to be effective; but sometimes it highlights that my thinking is not yet clear; my ability to communicate the project/task/big picture is not yet sufficient; and the process of achieving the clarity of thinking and translating to the LLM takes time (e.g. “that took 3 hours when it should have taken 40 minutes”) but ultimately still moves me forward to solving the problem or achieving the tasks and sub-tasks that I wanted to do. Remember what I said at the beginning:

    Clear thinking + clear communication of ideas/request = effective prompting => effective code and other outputs

 

  • Try it anyway.
    I am trying to get out of the habit of saying “I can’t do X”, like “I can’t code/program an iOS app”…because now I can. I’ve in fact built and shipped/launched/made available multiple iOS apps (check out Carb Pilot if you’re interested in macronutrient estimates for any reason; you can customize so you only see the one(s) you care about; or if you have EPI, check out PERT Pilot, which is the world’s first and only app for tracking pancreatic enzyme replacement therapy and has the same AI feature for generating macronutrient estimates to aid in adjusting enzyme dosing for EPI.) I’ve also made really cool, 100% custom-to-me niche apps to serve a personal purpose that save me tons of time and energy. I can do those things, because I tried. I flopped a bunch along the way – it took me several hours to solve a simple iOS programming error related to home screen navigation in my first few apps – but in the process I learned how to do those things and now I can build apps. I’ve coded and developed for OpenAPS and other open source projects, including a tool for data conversion that no one else in the world had built. Yet, my brain still tries to tell me I can’t code/program/etc (and to be fair, humans try to tell me that sometimes, too).

    I bring that up to contextualize that I’m working on – and I wish others would work on to – trying to address the reflexive thoughts of what we can and can’t do, based on prior knowledge. The world is different now and tools like LLMs make it possible to learn new things and build new projects that maybe we didn’t have time/energy to do before (not that we couldn’t). The bar to entry and the bar to starting and trying is so much lower than it was even a year ago. It really comes down to willingness to try and see, which I recognize is hard: I have those thought patterns too of “I can’t do X”, but I’m trying to notice when I have those patterns; shift my thinking to “I used to not be able to do X; I wonder if it is possible to work with an LLM to do part of X or learn how to do Y so that I could try to do X”.

    A recent real example for me is power calculations and sample size estimates for future clinical trials. That’s something I can’t do; it requires a statistician and specialized software and expertise.

    Or…does it?

    I asked my LLM how power calculations are done. It explained. I asked if it was possible to do it using Python code in a Jupyter notebook. I asked what information would be needed to do so. It walked me through the decisions I needed to make about power and significance, and highlighted variables I needed to define/collect to put into the calculation. I had generated the data from a previous study so I had all the pieces (variables) I needed. I asked it to write code for me to run in a Jupyter notebook, and it did. I tweaked the code, input my variables, ran it..and got the result. I had run a power calculation! (Shocked face here). But then I got imposter syndrome again, reached out to a statistician who I had previously worked with on a research project. I shared my code and asked if that was the correct or an acceptable approach and if I was interpreting it correctly. His response? It was correct, and “I couldn’t have done it better myself”.

    (I’m still shocked about this).

    He also kindly took my variables and put it in the specialized software he uses and confirmed that the results output matched what my code did, then pointed out something that taught me something for future projects that might be different (where the data is/isn’t normally distributed) although it didn’t influence the output of my calculation for this project.

    What I learned from this was a) this statistician is amazing (which I already knew from working with him in the past) and kind to support my learning like this; b) I can do pieces of projects that I previously thought were far beyond my expertise; c) the blocker is truly in my head, and the more we break out of or identify the patterns stopping us from trying, the farther we will get.

    “Try it anyway” also refers to trying things over time. The LLMs are improving every few months and often have new capabilities that didn’t before. Much of my work is done with GPT-4 and the more nuanced, advanced technical tasks are way more efficient than when using GPT-3.5. That being said, some tasks can absolutely be done with GPT-3.5-level AI. Doing something now and not quite figuring it out could be something that you sort out in a few weeks/months (see above about my 3 month bug); it could be something that is easier to do once you advance your thinking ; or it could be more efficiently done with the next model of the LLM you’re working with.

  • Test whether custom instructions help. Be aware though that sometimes too many instructions can conflict and also take up some of your context window. Plus if you forget what instructions you gave it, you might get seemingly unexpected responses in future chats. (You can always change the custom instructions and/or turn it on and off.)

I’m hoping this helps give people confidence or context to try things with LLMs that they were not willing to try before; or to help get in the habit of remembering to try things with LLMs; and to get the best possible output for the project that they’re working on.

Remember:

  • Right-size the task by making a clear ask.
  • You can use different chat windows for different levels of the same project.
  • Use a list to help you, the human, keep track of all the pieces that contribute to the bigger picture of the project.
  • Try giving the LLM a persona for an ask; and test whether you also need to assign yourself a persona or not for a particular type of request.
  • Be specific, think of the LLM as a conversational partner that can’t read your mind.
  • Don’t be afraid to start over with a new context window/chat.
  • Things that were hard a year ago might be easier with an LLM; you should try again.
  • You can do more, partnering with an LLM, than you can on your own, and likely can do things you didn’t realize were possible for you to do!

Clear thinking + clear communication of ideas/request = effective prompting => effective code and other outputs

—

Have any tips to help others get more effective output from LLMs? I’d love to hear them, please comment below and share your tips as well!

Tips for prompting LLMs like ChatGPT, written by Dana M. Lewis and available from DIYPS.org

Personalized Story Prompts for Kids Books and Early Reader Books

by Dana Lewis

For the holidays this year, I decided to try my hand at creating another set of custom, illustrated stories for my nieces and nephews (and bonus nieces and nephews). I have a few that are very advanced readers and/or too old for this, but I ended up with a list of 8 kids in my life from not-yet-reading to beginning reading to early 2nd grade reading level. I wanted to write stories that would appeal to each kid, include them as the main character, be appropriate for their reading (or read-to) level, and also include some of their interests.

Their interests were varied which made it quite a challenge! Here’s the list I worked from:

  • 2nd grade reading level, Minecraft
  • early 2nd grade reading level: soccer, stunt biking, parkour, ninja, Minecraft
  • beginning reading level: soccer, stunt biking, ninja, Spiderman
  • beginning reading level: Peppa Pig, moko jumbies
  • (read to younger child): Minnie Mouse, Peppa Pig, Bluey, and tea parties
  • (read to younger child): Bluey, Olaf, Elsa, & Anna
  • (read to younger child): cars/vehicles

I enlisted ChatGPT, an LLM, and ended up creating stories for each kid, matching their grade levels and interests, then illustrating them.

But illustrating them was actually a challenge (still), trying to create images with similar characters that would be on every page of the story and similar enough throughout that they were the “same” character.

Illustration challenges and how I got successful prompts:

My first pass on images wasn’t very good. I could get basic details to repeat, but often had images that looked like this – slightly different style and character throughout:

8 different illustrations in slightly different styles and almost different characters of a girl with blonde, shoulder length hair and a purple dress in an enchanted forest

Different styles throughout and that makes it look like a different character, even though it’s the same character in the whole story. This was a book to read to a <3 year old, though, and I thought she wouldn’t mind the different styles and left it as is. I also battled with adding, for personal use, the characters that most interested her: Peppa Pig and Minnie Mouse.

Interestingly, if I described with a prompt to illustrate a scene including a character “inspired by, but distinct from, Peppa Pig”…it essentially drew Peppa Pig or a character from it. No problems.

But if you gave the same prompt “inspired by, but distinct from, Minnie Mouse”? No go. No image at all: ChatGPT would block it for copyright reasons and wouldn’t draw any of the image. I riffed a bunch of times and finally was able to prompt a good enough mouse with round ears and a red dress with white polka dots. I had to ultimately illustrate the mouse character alone with the human character, because if I tried to get a Peppa-inspired character and then separately a mouse character, it wanted to draw the mouse with a pig-style face in the correct dress! I could never work around that effectively for the time I had available (and all the other books I was trying to illustrate!) so I stopped with what I had.

This was true for other characters, too, with copyright issues. It won’t draw anything from or like Bluey – or Frozen, when prompted. But I could get it to draw “an ethereal but warm, tall female adult with icy blonde hair, blue eyes, in an icy blue dress”, which you can see in the fourth image on the top row here:

Another series of illustrations with slightly different characters but closer in style throughout. there's one image showing a Frozen-inspired female character that I got by not prompting with Frozen.

I also managed to get slightly closer matching characters throughout this, but still quite a bit of variability. Again, for a young being-read-to-child, it was good enough for my purposes. (I never could get it to draw a Bluey-like character, even when I stopped referencing Bluey by name and described the shape and character, so I gave up on that.)

I tried a variety of prompts and series of prompts for each book. Sometimes, I would give it the story and prompt it with each page’s text, asking for an illustration and to keep it in the same style and the same character as the previous image. That didn’t work well, even when I told it in every prompt to use the same style and character plus the actual image prompt. I then tried to create a “custom” GPT, with the GPT’s instructions to use the same style throughout. That started to give me slightly better results, but I still had to remind it constantly to use the same style.

I also played around with taking an image that I liked, starting a new chat, and asking it to describe that image. Then I’d use that prompt to create a new prompt, describing the character in the same way. That started to get me slightly better results, especially when I did so using the custom GPT I had designed (you can try using this GPT here). I started to get better, more consistent characters:

A series of images of a young cartoon-drawn boy with wavy blonde hair riding a bike through an enchanted forest.

 

A series of drawings of a cartoon-like character with spiky blonde hair, blue eyes, and various outfits including a ninja costume

Those two had some variability, but a lot improved beyond the first several books. They are for the beginning and second-grade reading levels, too, so they are older kids with more attention to detail so it was worth the extra effort to try to get theirs to be more consistent.

The last one with the ninja and ninja outfits is another one that ran into copyright issues. I tried to have it illustrate a character inspired by, but distinct from, Spiderman – nope, no illustration at all. I asked it to illustrate the first picture in the soccer park with a spider strand looping in the corner of the image, like Spiderman had swung by but was out of sight and not picture – NOPE. You can’t even get an image that has Spiderman in the prompt at all, even if Spiderman isn’t in the picture! (I gave up and moved on without illustrating spiderwebs, even though Spiderman is described in the story).

My other favorite and pretty consistent one was two more of the early reader ones:

A series of images showing a young cartoon boy with wavy brown hair at a car fair

The hard part from that book was actually trying to do the cars consistently, rather than the human character. The human character was fairly consistent (although in different outfits, despite clear outfit prompts – argh) throughout, because I had learned from the previous images and prompt processes and used the Custom GPT, but the cars varied more. But, for a younger reader, hopefully that doesn’t matter.

The other, more-consistent character one for an early reader had some variations in style but did a better job matching the character throughout even when the style changed.

Another example with a mostly consistent young cartoon drawn girl with whispy blonde pigtails and big blue eyes, plus moko jumbies and peppa pig

How I wrote each story:

I also found some processes for building better stories. Again, see the above list of very, varied interests for each kid. Some prompts were straight forward (Minecraft) and other were about really different characters or activities (moko jumbies and Peppa Pig? Minnie Mouse and Peppa Pig? soccer ninja and Minecraft?).

What I ended up doing for each:

  1. In a new ChatGPT window (not the custom GPT for illustrating): Describe the reading level; the name of the character(s); and the interests. Ask it to brainstorm story ideas based on these interests.
  2. It usually gave 3 story ideas in a few sentences each, including a title. Sometimes, I would pick one and move on. Other times, I would take one of the ideas and tweak it a bit and ask for more ideas based on that. Or, I’d have it try again generally, asking for 3 more ideas.
  3. Once I had an idea that I liked, I would ask it to outline the story, based on the chosen story idea and the grade level we were targeting. Sometimes I would tweak the title and other times I would take the title as-is.
  4. Once it had the outline, I could have it then write the entire story (especially for the younger, beginner reader or read-to levels that are so short), but for the “chapter” books of early 2nd and 2nd grade reading level, I had it give me a chapter at a time, based on the outline. As each chapter was generated, I edited and tweaked it and took the text to where I would build the book. Sometimes, I would re-write the whole chapter myself, then give it back the chapter text and ask it to write the next one. If you didn’t give it back, it wouldn’t know what the chapter ended up as, so this is an important step to do when you’re making more than minor sentence construction changes.
  5. Because I know my audience(s) well, I tweaked it heavily as I went, incorporating their interests. For example, in the second images I showed above, there’s a dancing dog. It’s their actual dog, with the dog named in the story along with them as characters. Or in the chapter book for the character with the bike, it described running up a big mountain on a quest and being tired. I tossed in an Aunt-Dana reference including reminding the character about run-walking as a way to keep moving forward without stopping and cover the distance that needs to be covered. I also tweaked the stories to include character traits (like kindness) that each child has, and/or behaviors that their family prioritizes.

I described the images processes first, then the story writing, in this blog post, but I actually did the opposite for each book. I would write (brainstorm, outline, write, edit, write) the entire book, then I would go start a new chat window (eventually solely using my custom GPT) and ask for illustrations. Sometimes, I would give it the page of the story’s text and ask it to illustrate it. That’s helpful when you don’t know what to illustrate, and it did fairly well for some of the images (especially the Minecraft-inspired ones!). Ultimately, though, I would often get an image, ask what the prompt was for the image, tweak the prompt, and give it back to better match the story or what I wanted to illustrate. Once I was regularly asking for the image prompts, I had realized that giving the character details repeatedly for every image helped with consistency. Then I would use the ad-nauseam details myself for a longer prompt, which resulted in better images throughout, so I spent more energy deciding myself what to illustrate to best match the story.

All in all, I made 7 custom books (and 8 copies, one of the Minecraft books I copied and converted to a different named character for a friend’s child!). Between writing and editing, and illustrating, I probably spent an average of one hour per book! That’s a lot of time, but it did get more efficient as I went, and in some cases the hour included completely starting over and re-working the images in the book for consistency compared to the version I had before. The next books I create will probably take less time, both because I figured out the above processes but also because hopefully DALL*E and other illustration tools will get better about being able to illustrate the same character consistently across multiple prompts to illustrate a story.

How other people can use this to create stories – and why:

I have been so excited about this project. I love, love, love to read and I love reading with my nieces and nephews (and bonus kids in my life) and finding books that match their interest and help spark or maintain their love of reading. That’s why I did this project, and I have been bursting for WEEKS waiting to be able to give everyone their books! I wanted it to be a surprise for their parents, too, which meant that I couldn’t tell 2/3 of my closest circles about my cool project.

One of my friends without young kids that I finally told about my project loved the idea: she works as staff at an elementary school, supporting some students who are working on their reading skills who are nonverbal. She thought it would be cool to make a book for one student in particular, and described some of her interests: violins, drums, raspberries, and unicorns. I was in the car when she told me this, and I was able to follow the same process as above in the mobile ChatGPT app and list the interests, ask for a brainstorm of story ideas for a beginning reading level style book that had some repetitive text using the interests to aid in reading. It created a story about a unicorn who gathers other animals in the forest to play in an orchestra (with drums and violins) and eat raspberries. I had it illustrate the story, and it did so (with slightly different unicorns throughout). I only had to have it re-draw one image, because it put text in one of the last images that didn’t need to be there.

Illsutrations from a quick story about a unicorn, drums, violin, and an orchestra, plus raspberries

It was quick and easy, and my friend and her student LOVED it, and the other teachers and staff at the school are now working on personalized books for a lot of other students to help them with reading skills!

It really is an efficient and relatively easy way to generate personalized content; it can do so at different reading levels (especially when a teacher or someone who knows the student can tweak it to better match the reading level or sounds and words they are working on next); and you can generate pretty good matching illustrations too.

The hardest part is consistent characters; but when you don’t need consistency throughout a whole book, the time it takes drops to ~5 or so minutes to write, tweak, and illustrate an entire story.

—

Illustrations require a paid ChatGPT account, but if you have one and want to try out the custom GPT I built for (slightly more consistent) illustrations of stories, you can check it out here.

Custom stories: prompting and effective illustrating with ChatGPT, a blog post by Dana M. Lewis from DIYPS.org

Running a Multi-Day Ultramarathon (Aiming for 200 Miles)

by Dana Lewis

I used to make a lot of statements about things I thought I couldn’t do. I thought I couldn’t run overnight, so I couldn’t attempt to run 100 miles. I could never run 200 mile races the way other people did. Etc. Yet last year I found myself training for and attempting 100 miles (I chose to stop at 82, but successfully ran overnight and for 25 hours) and this year I found myself working through the excessive mental logistics and puzzle of determining that I could train for and attempt to run 200 miles, or as many miles as I could across 3-4 days.

Like my 100 mile attempt, I found some useful blog recaps and race reports of people’s official races they did for 200-ish mile races. However, like the 100 attempts, I found myself wanting more information for the mental training and logistical preparation people put into it. While my 200 mile training and prep anchored heavily on what I did before, this post describes more detail on how my training, prep, and ‘race’ experience for a multi-day or 200 mile ultra attempt.

DIY-ing a 200

For context, I have a previous post describing the myriad reasons of why I often choose to run DIY ultras, meaning I’m not signing up for an official race. Most of those reasons hold true for why I chose to DIY my 200. Like my 100 (82) miles, I mapped a route that was based on my home paved trail that takes me out and around the trails I’m familiar with. It has its downsides, but also the upsides: really good trail bathrooms and I feel safe running them. Plus, it’s easy and convenient for my husband to crew me. Since I expected this adventure to take 3-4 days (more on that below), that’s a heavy ask of my husband’s time and energy, so sticking with the easy routes that work for him is optimal, too. So while I also sought to run 200 miles just like any other 200-mile ultra runner, my course happens to have minimal elevation. Not all 200 mile ultramarathon races have a ton of elevation – some like the Cowboy 200 are pretty flat – so my experience is closer to that than the experience of those running mountain based ultras with 30,000 feet (or more) of elevation gain. And I’m ok with that!

Sleep

One of the puzzles I had to figure out to decide I could even attempt a 200 miler is sleep. With a 100 mile race, most people don’t sleep at all (nor did I) and we just run through the night. With 200 miles, that’s impossible, because it takes 3, 4, 5 days to finish and biologically you need sleep. Plus, I need more sleep than the average person. I’m a champion sleeper; I typically sleep much longer than everyone else; and I know I couldn’t function with an hour here or there like many people do at traditional races. So I actually designed my 200 mile ultra with this in mind: how could I cover 200 miles AND get sleep? Because I’m running to/from home, I have access to my kitchen, shower, and bed, so I decided that I would set up my run to run each day and come home and eat dinner, shower, and sleep each night for a short night in my bed.

I then decided that instead of winging it and running until I dropped before eating, showering, and sleeping, I would aim for running 50 miles each day. Then I’d come in, eat, shower, and sleep and get up the next morning and go again. 4 days, 3 nights, 50 miles each day: that would have me finishing around 87-90ish hours total (with the clock running from my initial start), including ~25 hours or more of total downtime between the eating/showering/sleeping/getting ready. That breakdown of 3.67 days is well within the typical finish times of many 200 mile ultras (yes, comparing to those with elevation gain), so it felt like it was both a stretch for me but also doable and in a sensible way that works for me and my needs. I mapped it all out in my spreadsheet, with the number of laps and my routes and pacing to finish 50 miles per day; the two times per day I would need my husband to come out and crew me at ‘aid station stops’ in between laps, and what time I would finish each night. I then factored in time to eat and shower and get ready for bed, sleep, and time to get up in the morning. Given the fact that I expected to run slower each day, the sleep windows go from 8 hours down to less than 6 hours by night 3. That being said, if I managed to sleep 5 hours per night and 15 hours total, that’s probably almost twice as much as most people get during traditional races!

Like sleep, I was also very cognizant of the fact that a 200 probably comes down to mental fortitude and will power to keep going; meticulous fueling; and excellent foot care. Plus reasonable training, of course.

Meticulous fueling

I have previously written about building and using a spreadsheet to track my fuel intake during ultras. This method works really well for me because after each training run I can see how much I consumed and any trends. I started to spot that as I got tired, I would tend to choose certain snacks that happened to be slightly lower calorie. Not by much, but the snack selections went from those that are 150-180 calories to 120-140 calories, in part because I perceived them to be both ‘smaller’ (less volume) and ‘easier to swallow’ when I was tired. Doubled up in the same hour, this meant that I started to have hours of 240 calories instead of more than 250. That doesn’t sound like much, but I need every calorie I can get.

I mapped out my estimated energy expenditure based on the 50 miles per day, and even consuming 250 calories per hour, I would end up with several thousand calories of deficit each day! I spent a lot of time testing food that I think I can eat for dinner on the 3 nights to ensure that I get a good 1000 calories or more in before going to bed, to help address and reduce the growing energy deficit. But I also ended up optimizing my race fuel, too. Because I ran so many long runs in training where I fueled every 30 minutes, and because I had been mapping out my snack list for each lap for 50 miles a day for 4 days, I’ve been aware for months that I would probably get food fatigue if I didn’t expand my fuel list. I worked really hard to test a bunch of new snacks and add them to the rotation. That really helped even in training, across all 12 laps (3 laps a day to get 50 miles, times 4 days), I carefully made sure I wouldn’t have too many repeats and get sick of one food or one group of things I planned to eat. I also recently realized that some of the smaller items (e.g. 120 calorie servings) could be increased. I’m already portioning out servings from a big bag into small baggies; in some cases adding one more pretzel or one more piece of candy (or more) would drive up the calories by 10-20 per serving. Those small tweaks I made to 5 of my ~18 possible snacks means that I added about 200 calories on top of what was already represented in those snacks. If I happen to choose those 5 snacks as part of my list for any one lap, that means I have a bonus 200 calories I’ve convinced myself to consume without it being a big deal, because it’s simply one more pretzel or one more piece of candy in the snack that I’m already use to consuming. (Again, because I’m DIYing my race and have specific needs relative to running with celiac, diabetes, and exocrine pancreatic insufficiency, for me, pre-planning my fuel and having it laid out in advance for every run, or in the race every single lap, is what works for me personally.)

Here’s a view of how I laid out my fuel. I had worked on a list of what I wanted for each lap, checking against repeats across the same day and making sure I wasn’t too heavily relying on any one snack throughout all the days. I then bagged up all snacks individually, then followed my list to lay them out by each lap and day accordingly. I also have a bag per day each for enzymes and electrolytes, which you’ll see on the left. Previously, I’ve done one bag per lap, but to reduce the number of things I’m pulling in and out of my vest each time, I decided I could do one big bag each per day (and that did end up working out well).

Two pictures side by side, with papers on the floor showing left to right laps 1-3 on the top and along the left side days 1-4, to create a grid to lay out my snacks. On the left picture, I have my enzymes, electrolytes per day and then a pile of snacks grouped for each lap. On the right, all the snacks and enzymes and electrolytes have been put into gallon bags, one for each lap.

Contingency planning

Like I did for my 100, I was (clearly) planning for as many possibilities as I could. I knew that during the run – and each evening after the run – I would have limited excess mental capacity for new ideas and brainstorming solutions when problems come up. The more I prepared for things that I knew were likely to happen – fatigue, sore body, blisters, chafing, dropping things, getting tired of eating, etc – the more likely that they would be small things and not big things that can contribute to ending a race attempt. This includes learning from my past 100 attempt and how I dealt with the rain. First of all, I planned to move my race if it looks like we’ll get 6 months of rain in a single 24 hour period! But also, I scheduled my race so that if I do have a few hours of really hard rain, I could choose to take a break and come in and eat/shower/change/rest and go back out later, or extend and finish a lap on the last day or the day after that. I was not running a race that would yank me from the course, but I did have a hard limit after day 5 based on a pre-planned doctor’s appointment that would be a hassle to reschedule, so I needed to finish by the night after day 5. But this gave me the flexibility to take breaks (that I wasn’t really planning to take but was prepared to if I needed to due to weather conditions).

Training for a 200 mile ultramarathon

Like training plans for marathons and 100 milers, the training plans I’ve read about for 200 mile ultramarathons intimidate me. So much mileage! So much time for a slow run/walker like me. I did try to look at sample 200 mile ultra plans and get a sense for what they’re trying to achieve – e.g. when do they peak their mileage before the race, how many back to back runs of what general length in terms of time etc – and then loosely keep that in mind.

But basically, I trained for this 200 mile ultra just like I trained for my marathon, 50k, 100k, and 82 miler. I like to end up doing long runs (which for me are run/walks of 30 seconds run, 60 seconds walk, just like I do shorter runs) of up to around 50k distance. This time, I did two total training runs that were each around 29 miles, just based on the length of the trail I had to run. I could have run longer, but mentally had the confidence that another ~45 minutes per run wasn’t going to change my ability to attempt 50 miles a day for 4 days. If I didn’t have 3 years of this training style under my personal belt, I might feel different about it. That’s longer than many people run, but I find the experience of 7-8 hours of time on my feet fueling, run/walking, and problem solving (including building up my willpower to spend that much time moving) to be what works for me.

The main difference for my 200 is probably also that it’s my 3rd year of ultrarunning. I was able to increase my long runs a little bit more of a time, when historically I used to add 2 miles a time to a long run. I jumped up 4 miles at a time – again, run/walking so very easy on my legs – when building up my long runs, so I was able to end up with 2 different 29 mile runs, two weeks apart, even though I really kicked off training specifically for this 8 weeks prior (10 weeks including taper) to the run. In between I also did a weekend of back to back to back runs (meaning 3 days in a row) where I ran 16 miles, another 16 miles, and 13 miles to practice getting up and running on tired legs. In past cycles I had done a lot more back to back (2-day) with a long and a medium run, but this time I did less of the 2-day and did the one big 3-day since I was targeting a 4-day experience. In future, if I were to do this again, given how well my body held up with all this training, I might have done more back to back, but I took things very cautiously and wanted to not overtrain and cause injury from ramping up too quickly.

As part of that (trying not to over do it), instead of doing several little runs throughout the week I focused on more medium-long runs with my vest and fueling, so I would do something like a long run (starting at 10 miles building up to 29 miles), a medium-long run (8 miles up to 13 miles or 16 miles) and another medium-ish run (usually 8 miles). Three runs a week, and that was it. Earlier in the 8 weeks, I was still doing a lot of hiking off the season, so I had plenty of other time-on-feet experiences. Later in the season I sometimes squeezed in a 4th short run of the week if we wouldn’t be hiking, and ran without my vest and tried to do some ‘speed work’ (aka run a little faster than my easy long run pace). Nothing fancy. Again, this is based on my slow running style (that’s actually a fixed interval of short run and short walk, usually 30 seconds run and 60 seconds walk), my schedule, my personality, and more. If you read this, don’t think my mileage or training style is the answer. But I did want to share what I did and that it generally worked for me.

I did struggle with wondering if I was training “enough”. But I never train “enough” compared to others’ marathon, 50k, 100k, 100 mile plans, either. I’m a low mileage-ish trainer overall, even though I do throw in a few longer runs than most people do. My peak training for marathon, 50k, and 100k is usually around low 50s (miles per week). Surprisingly, this 200 cycle did get me to some mid 60 mile weeks! One thing that also helped me mentally was adding in a rolling 7 day calculation of the miles, not just looking at miles per calendar week. That helped when I shifted some runs around due to scheduling, because I could see that I was still keeping a reasonable 55-low60s mileage over 7 days even though the calendar week total dropped to low 40s because of the way the runs happened to land in the calendar weeks.

Generally, though, looking back at how my training was more than I had accomplished for previous races; I feel better than ever (good fueling really helps!); I didn’t have any accidents or overtraining injuries or niggles; I decided a few weeks before peak that I was training enough and it was the right amount for me.

Another factor that was slightly different was how much hiking I had done this year. I ran my 100k in March then took some time off, promising my husband that we would hike “more” this year. That also coincided with me not really bouncing back from my 100k recovery period: I didn’t feel like doing much running, so we kept planning hiking adventures. Eventually I realized (because I was diagnosed with Graves’ disease last year, I’m having my thyroid and antibody and other related blood work done every 3 months while we work on getting everything into range) that this coincided with my TSH going too high for my body’s happiness; and my disinterest in long runs was actually a symptom (for me) of slightly too-high TSH. I changed my thyroid medication and within two weeks felt HUGELY more interested in long running, which is what coincided with reinvigorating my interest in a fall ultra, training, and ultimately deciding to go for the 200. But in the meantime, we kept hiking a lot – to the tune of over 225 miles hiked and over 53,000 feet of elevation gain! I never tracked elevation gain for hiking before (last year, not sure I retrospectively tracked it all but it was closer to 100 miles – so definitely likely 2x increase), but I can imagine this is definitely >2x above what I’ve done on my previous biggest hiking year, just given the sheer number of hikes that we went out on. So overall, the strengthening of my muscles from hiking helped, as did the time on feet. Before I kicked off my 8 week cycle, we were easily spending 3-4 hours a hike and usually at least two hikes a weekend, so I had a lot of time on feet almost every hike equivalent to 12 or more miles of running at that point. That really helped when I reintroduced long runs and aided my ability to jump my long run in distance by 4 miles at a time instead of more gently progressing it by 2 miles a week as I had done in the past.

How my 200 mile attempt actually went

Spoiler alert: I DNF (did not finish) 200 miles. Instead, I stopped – happily – at 100 miles. But it wasn’t for a lack of training.

Day 1 – 51 miles – All as planned

I set out on lap 1 on Day 1 as planned and on time, starting in the dark with a waist lamp at 6am. It was dark and just faintly cool, but warm enough (51F) that I didn’t bother with long sleeves because I knew I would warm up. (Instead, for all days, I was happy in shorts and a short sleeve shirt when the temps would range from 49F to 76F and back down again.) I only had to run for about an hour in the dark and the sky gradually brightened. It ended up being a cloudy, overcast and nice weather day so it didn’t get super bright first thing, but because it wasn’t wet and cold, it wasn’t annoying at all. I tried to start and stay at an easy pace, and was running slow enough (about ~30s/mile slower than my training paces) that I didn’t have to alter my planned intervals to slow me down any more. All was fairly well and as planned in the first lap. I stopped to use the bathroom at mile 3.5 and as planned at my 8 mile turnaround point, and also stopped to stuff a little more wool in a spot in my shoe a mile later. That added 2 minutes of time, but I didn’t let it bother me and still managed to finish lap 1 at about a 15:08 min/mi average pace, which was definitely faster than I had predicted. I used the bathroom again at the turnaround while my husband re-filled my hydration pack, then I stuffed the next round of snacks in my vest and took off. The bathroom and re-fueling “aid station” stop only took 5 minutes. Not bad! And on I went.

A background-less shot of me in my ultrarunning gear. I'm wearing a grey moisture-wicking visor; sunglasses; a purple ultrarunning vest packed with snacks in front and the blue tube of my hydration pack looped in front; a bright flourescent pink short sleeve shirt; grey shorts with pockets bulging on the side with my phone (left pocket) and skittles and headphones and keys (right pocket), and in this lap I was wearing bright pink shoes. Lap 2 was also pretty reasonable, although I was surprised by how often I wanted a bathroom. My period had started that morning (fun timing), and while I didn’t have a lot of flow, the signals my abdomen was giving my brain was telling me that I needed to go to the bathroom more often than I would have otherwise. That started to stress me out slightly, because I found myself wishing for a bathroom in the longest stretch without trail bathrooms and in a very populated area, the duration of which was about 5.5 miles long. I tried to drink less but was also aware of trying not to under hydrate or imbalance my electrolytes. I always get a little dehydrated during my period; and I was running a multi-day ultra where I needed a lot of hydration and more sodium than usual; this situation didn’t add up well! But I made it without any embarrassing moments on the trail. The second aid station again only took 5 minutes. (It really makes a world of difference to not have to dry off my feet, Desitin them up, and re-do socks and shoes every single aid station like I did last year!) I could have moved faster, but I was trying to not let small minutes of time frazzle me, and I was succeeding with being efficient but not rushed and continuing on my way. I had slowed down some during lap 2, however – dropping from a 15:08 to 15:20ish min/mi pace. Not much, but noticeable.

At sunset, with light blue sky fading to yellow at the horizon behind the row of tall, skinny bush like trees with gaps and a hot air balloon a hundred or so feet off the ground seen between the trees.Lap 3 I did feel more tired. I talked my husband into bringing me my headlamp toward the end of the last lap, instead of me having to carry it for 4+ hours before the sun went down. (Originally, I thought I would need it 2-3 hours into this last lap, but because I was moving so well it was now looking like 4 hours, and it would be a 2-3 mile e-bike ride for him to bring me the lamp when I wanted it. That was a mental win to not have to run with the lamp when I wasn’t using it!) I was still run/walking the same duration of intervals, but slowed down to about 16:01 pace for this lap. Overall, I would be at 15:40 average for the whole day, but the fatigue and my tired feet started to kick in on the third lap between miles 34-51. Plus, I stopped to take a LOT more pictures, because there was a hot air balloon growing in the distance as it was flying right toward me – and then by me next to the trail! It ended up landing next to the soccer fields a mile behind me after it passed me in this picture. I actually made it home right as the sun set and didn’t have to wear my lamp at all that evening.

Day 1 recovery was better and worse than I expected. I sat down and used my foot massager on my still-socked feet, which felt very good. I took a shower after I peeled my socks off and took a look at my feet for the first time. I had one blister that I didn’t know was growing at all pop about an hour before I finished, but it was under some of my pre-taped area. I decided to leave the tape and see how it looked and felt in the morning. I had 2-3 other tiny, not a big deal blisters that I would tape in the morning but didn’t need any attention that night.

I had planned to eat a reasonably sized dinner – preferably around 1000 calories – each night, to help me address my calorie deficit. And I had a big deficit: I had burned 5,447 calories and consumed 3,051 calories in my 13 hours and 13 minutes of running. But I could only eat ¼ of the pizza I planned for dinner, and that took a lot of work to force myself to eat. So I gave up, and went to bed with a 3,846 calorie deficit, which was bigger than I wanted.

And going to bed hurt. I was stiff, which I could deal with, but my feet that didn’t hurt much while running started SCREAMING at me. All over. They hurt so bad. Not blisters, just intense aches. Ouch! I started to doubt my ability to run the next day, but this is where my pre-planning kicked in (aided by my husband who had agreed to the rules we had decided upon): no matter what, I would get up in the morning, get dressed, and go out and start my first lap. If I decided to quit, I could, but I could not quit at night in bed or in the morning in the bed or in the house. I had to get up and go. So I went to sleep, less optimistic about my ability to finish 50 miles again on day 2, but willing to see what would happen.

Day 2: 34 instead of 50 miles, and walking my first ever lap

I actually woke up before my alarm went off on day 2. Because I had finished so efficiently the day before, I was able to again get a good night’s sleep, even with the early alarm and waking up again at 4:30am with plans to be going by 6am. The extra time was helpful, because I didn’t feel rushed as I got ready to go. I spent some extra time taping my new blisters. Because they hadn’t popped, I put small torn pieces of Kleenex against them and used cut strips of kinesio tape to protect the area. (Read “Fixing Your Feet” for other great ultra-related foot care tips; I learned about Kleenex from that book.) I also use lambs’ wool for areas that rub or might be getting hot spots, so I put wool back in my usual places (between big and second toes, and on the side of the foot) plus another toe that was rubbing but not blistered and could use some cushion. I also this year have been trying Tom’s blister powder in my socks, which seems to help since my feet are extra sweat prone, and I had pre-powdered a stack of socks so I could simply slip them on and get going once I had done the Kleenex/tape and wool setup. The one blister that had popped under my tape wasn’t hurting when I pressed on it, so I left it alone and just added loose wool for a little padding.

A pretty view of the trail with bright blue sky after the sun rose with green bushes (and the river out of sight) to the left, with the trail parallel to a high concrete wall of a road with cheery red and yellow leaved trees leaning over the trail.And off I went. I managed to run/walk from the start, and faster than I had projected on my spreadsheets originally and definitely faster than I thought was possible the night before or even before I started that morning. Sure, I was slower than the day before, but 15:40 min/mi pace was nothing to sneeze at, and I was feeling good. I was really surprised that my legs, hips and body did not hurt at all! My multi-day or back-to-back training seemed to pay off here. All was well for most of the first lap (17 miles again), but then the last 2 or so miles, my pace started dipping unexpectedly so I was doing 16+ min/mi without changing my easy effort. I was disappointed, and tired, when I came into my aid station turnaround. I again didn’t need foot care and spent less than 5 minutes here, but I told Scott as I left that I was going to walk for a while, because my feet had been hurting and they were getting worse. Not blisters: but the balls of my feet were feeling excruciating.

A close up of a yellow shelled snail against the paved trail that I saw while walking the world's slowest 17-mile lap on day 2.I headed out, and within a few minutes he had re-packed up and biked up to ride alongside me for a few minutes and chat. I told him I was probably going to need to walk this entire lap. We agreed this was fine and to be expected, and was in fact built into my schedule that I would slow down. I’ve never walked a full lap in an ultra before, so this would be novel to me. But then my feet got louder and louder and I told him I didn’t think I could even walk the full lap. We decided that I should take some Tylenol, because I wasn’t limping and this wouldn’t mask any pain that would be important cues for my body that I would be overriding, but simply muting the “ow this is a lot” screams that the bones in the balls of my feet were feeling. He biked home, grabbed some, and came back out. I took the Tylenol and sent him home again, walking on. Luckily, the Tylenol did kick in and it went from almost unbearable to manageable super-discomfort, so I continued walking. And walking. And walking. It took FOREVER, it felt like, having gone from 15-16 min/mi pace with 30 seconds of running, 60 seconds of walking, to doing 19-20 minute miles of pure walking. It was boring. I had podcasts, music, audiobooks galore, and I was still bored and uncomfortable and not loving this experience. I also was thinking about it on the way back about how I did not want to do a 3rd lap that day (to get me to my planned 50 miles) walking again.

Scott biked out early to meet me and bring me extra ice, because it was getting hot and I was an hour slower than the day before and risking running out of water that lap if he didn’t. After he refilled my hydration pack and brought it back to me while I walked on, I told him I wanted to be done for the day. He pointed out that when I finished this lap, I would be at 34 miles for the day, and combined with the day before (51), that put me at 85 miles, which would be a new distance PR for me since last year I had stopped at 82. That was true, and that would be a nice place to stop for the day. He reminded me of our ‘rules’ that I could go out the next day and do another lap to get me to 100, and decide during that lap what else I wanted to do. I was pretty sure I didn’t want to do more, but agreed I would decide the next day. So I walked home, completing lap 2 and 34 miles for the day, bringing me to 85 miles overall across 2 days.

Day 2 recovery went a little better, in part because I didn’t do 51 miles (only 34) and I had walked rather than ran the second lap, and also stopped earlier in the day (4pm instead of 7pm). I had more time to shower and bring myself to finally eat an entire 1000 calories before going to bed, again with my feet screaming at me. I had more blisters this time, mostly again on my right foot, but the balls of my feet and the bones of my feet ached in a way they never had before. This time, though, instead of setting my alarm to get up and go by 6am, I decided to sleep for longer, and go out a little later to start my first lap. This was a deviation from my plan, but another deviation I felt was the right one: I needed the sleep to help my body recover to be able to even attempt another lap.

Day 3: Only 16 miles, but hitting 100 for the first time ever

Instead of 6am, I set out on Day 3 around 8:30am. I would have taken even longer to go, but the forecast was for a warm day (we ended up hitting 81F) and I wanted to be done with the lap before the worst of the heat. I thought there was a 10% chance I’d keep going after this lap, but it was a pretty small chance. However, I set out for the planned 16 mile lap and was pleasantly surprised that I was run/walking at about a 15:40 pace! Again, better than I had projected (although yes, I had deviated from my mileage plan the day before), and it felt like a good affirmation that stopping the day before instead of slogging out another walking lap was the right thing to do.

After a first few miles, I toyed with the idea of continuing on. But I knew with the heat I probably wouldn’t stand more than one more lap, which would get me to 116. Even if I went out again the fourth day, and did 1-2 laps, that would MAYBE get me to 150, but I doubted I could do that without starting to cause some serious damage. And it honestly wasn’t feeling fun. I had enjoyed the first day, running in the dark, the fog, the daylight, and the twilight, seeing changing fall leaves and running through piles of them. The second day was also fun for the first lap, but the second lap walking was probably what a lot of ultra marathoners call the “death march” and just not fun. I didn’t want to keep going if it wasn’t fun, and I didn’t want to run myself into the ground (meaning to be so worn down that it would take weeks to months to recover) or into injury, especially when the specific milestones didn’t really mean anything. Sure, I wanted to be a 200 mile ultramarathoner, something that only a few thousand people have ever done – but I didn’t want to do it at the expense of my well-being. I spent a lot of time thinking about it, especially miles 4-8, and was thinking about the fact that the day before I had started, I had gone to a doctor’s appointment and had an official diagnosis confirming my fifth autoimmune disease, then proceeded to run (was running) 100 miles. Despite all the fun challenges of running with autoimmune conditions, I’m in really good health and fitness. My training this year went so well and I really enjoyed it. Most of this ultra had gone so well physically, and my legs and body weren’t hurting at all: the weakness was my feet. I didn’t think I could have trained any differently to address that, nor do I think I could change it moving forward. It’s honestly just hard to run that many hours or that many miles, as most ultramarathoners know, and your feet take a beating. Given that I was running on pavement for all of those hours, it can be even harder – or a different kind of hard – than kicking roots and rocks on a dirt trail. I figured I would metaphorically kick myself if I tried for 116 or 134 and injured myself in a way that would take 6-8 weeks to recover, whereas I felt pretty confident that if I stopped after this lap (at 100), I would have a relatively short and easy recovery, no major issues, and bounce back better than I ever have, despite it being my longest ever ultramarathon. Yes, I was doing it as a multi-day with sleep in between, but both in time on feet and in mileage, it was still the most I’d ever done in 2 or 3 days.

And, I was tired of eating. I was fueling SO well. Per my plans, I set out to do >500 mg of sodium per hour and >250 calories per hour. I had been nailing it every lap and every day! Day 1 I averaged 809 mg of sodium per hour and 290 calories per hour. Day 2 was even increased from that, averaging 934 mg of sodium per hour and 303 calories per hour! Given the decreased caloric burn of day 2 because I walked the second lap, my caloric deficit for day 2 was a mere ~882 calories (given that I also managed to eat a full dinner that night), even though I skipped the last hour as I finished the walking lap. Day 3 I was also fueling above my goals, but I was tired of it. Sooooo tired of it. Remember, I have to take a pill every time I eat, because I have exocrine pancreatic insufficiency (EPI or PEI). I was eating every 30 minutes as I ran or walked, so that meant swallowing at least one pill every 30 minutes. I had swallowed 57 pills on Day 1 and 48 pills on Day 2, between my enzymes and electrolyte pills. SO MANY PILLS. The idea of continuing to eat constantly every 30 minutes for another lap of ~5 or more hours was also not appealing. I knew if I didn’t eat, I couldn’t continue.

A chart with an hourly break down of sodium, calories, and carbs consumed per hour, plus totals of caloric consumption, burn, and calculated deficit across ~27 hours of move time to accomplish 100 miles run.

And so, I decided to stop after one more lap on day 3, even though I was holding up a respectable 15:41 min/mi pace throughout. I hit 100 miles and finished the lap at home, happy with my decision.

Two pictures of me leaning over after my run holding a sign (one reading 50 miles, one reading 100 miles) for each of my cats to sniff.(You can see from these two pictures that I smelled VERY interesting, sweaty and salty and exhausted at the end of day 1 and day 3, when I hit 50 miles and 100 miles, respectively. We have two twin kittens (now 3 years old) and one came out to sniff me first on the first day, and the other came out as I came home on the third day!)

Because I had only run one final lap (16 miles) on day 3, and had so many bonus hours in the rest of the day afterward when I was done and home, I was able to eat more and end up with only a 803 calorie deficit for the day. So overall, day 1 had the biggest deficit and probably influenced my fatigue and perception of pain on day 2, but because I had shortened day 2 and then day 3, my very high calorie intake every hour did a pretty good job matching my calorie expenditure, which is probably why I felt very little muscle fatigue in my body and had no significant sore areas other than the bottoms of my feet. I ended up averaging 821 mg/hr of sodium and 279 calories per hour (taking into account the fact that I skipped two final snacks at the end of day 2 when I was walking it out; ignoring that completely skipped hour would mean the average caloric intake on hours I ate anything at all was closer to 290 calories/hr!)

In total, I ended up consuming 124 pills in approximately 27 hours of move time across my 100 miles. (This doesn’t include enzyme pills for my breakfast or dinners each of those days, either – just the electrolyte and enzyme pills consumed while running!)

AFTERMATH

Recovery after day 3 was pretty similar to day 2, with me being able to eat more and limit my calorie deficit. I’ve had long ~30 mile training runs where I wasn’t very hungry afterward, but it surprised me that even two days after my ultra, I still haven’t really regained my appetite. I would have figured my almost 4000 calorie deficit from day 1 would drive a lot of hunger, so this surprised me.

So too has my physical state: 48 hours following the completion of my 100 miles, I am in *fantastic* shape compared to other multi-day back to back series of runs I’ve done, ultramarathons or not. The few blisters I got, mainly on my right foot, have already flattened themselves up and mostly vanished. I think I get more blisters on my right foot because of breaking my toe last year: my right foot now splays wider in my shoe, so it tends to get more blisters and cause more trouble than my left foot. I got only one blister on my left foot, which is still fluid filled but not painful and starting to visibly deflate now that I’m not rubbing it onto a shoe constantly any more. And my legs don’t feel like I ran at all, let alone running 51+34+16 miles!

I am tired, though. I don’t have brain fog, probably because of my excellent fueling, but I am fatigued in terms of overall energy and lack of motivation to get a lot done yesterday and today (other than writing this blog post!). So that’s probably pretty on par with my effort expended and matches what I expected, but it’s nice to be able to move around without hurting (other than my feet).

My feet in terms of general aches and ows are what came out the worst from my run. Day 2, what hurt was the bottom of the balls of my feet. Starting each night though, I was getting aches all over in all of the bones of my feet. After day 3, that night the foot aches were particularly strong, and I took some Tylenol to help with that. Yesterday evening and today though, the ache has settled down to very minor and only occasionally noticeable. The tendon from the top of my left foot up my ankle is sore and gets cranky when I wear my sneakers (although it didn’t bother me at all while running any of the days), so after tying and re-tying my shoelaces 18 times yesterday to try to find the perfect fit for my left foot, today I went on my recovery walk in flip flops and was much happier.

What I’m taking away from this 200 mile attempt that was only 100 miles:

I feel a little disappointed that I didn’t get anywhere near 200 miles, but obviously, I was not willing to hurt long enough or hard enough to get there. My husband called it a stretch goal. Rationally, I am very happy with my choices to stop at 100 and end up in the fantastic physical shape that I am in, and I recognize that I made a very rational choice and tradeoff between ending in good shape (and health) and the mainly ego-driven benefits of possibly achieving 200 miles (for me).

Would I do anything different? I can’t think of anything. If I somehow had an alternate do-over, I can’t think of anything I would think to change. I’d like to reduce my risk of blisters but I’m already doing all I can there, and dealing with changes in my right foot shape post-broken toe that I have no control over. And I’m not sure how to train more/better for reducing the bottom ball of foot pain that I got: I already trained multiple days, back to back, long hours of feet on pavement. It’s possible that having my doctor’s appointment the day before I started influenced my mental calculation of my future risk/benefit tradeoff of continuing more miles, and so not having had that then may have changed my calculations to do another lap or two, or go out on the 4th day (which I did not). But, I don’t have a do over, and I’ll never know, and I’m not too upset about that because I was able to control what I could control and am again pretty happy with the outcomes. 100 or 150 miles felt about the same to me, psychologically, in terms of satisfaction.

What I would tell other people about attempting multiple day ultramarathons or 200 mile ultramarathons:

Training back to back days is one option, as is long spurts of time on feet walking/hiking/running. I don’t think “just running” has to be the only way to train for these things. I’m also a big proponent of short intervals: If you hear people recommend taking walk breaks, it doesn’t have to be 1 minute every 10 minutes or every mile. It can be as short as every 30 seconds of running, take a walk break! There’s no wrong way to do it, whatever makes your body and brain happy. I get bored running longer (and don’t like it); other people get bored running the short intervals that I do – so find what works for you and what you’re actually willing to do.

Having plans for how you’ll rest X hours and go out and try to make it another lap or to the next aid station works really well, especially if you have crew/pacers/support (for me, my husband) who will stick to those rules and help you get back out there to try the next lap/section. Speaking of sleep/rest, laying down for a while helps as much as sleeping, so even if you can’t sleep, committing to the rest of X hours is also good for resting your feet and everything. I found that the hour laying down before I fell asleep helped my body process the noise of the “ouch” from my feet and it was a lot easier to sleep after that. Plan that you’ll have some down/up time before and after your sleep/rest time, and figure that into your time plans accordingly.

The cheesy “know your why” and “know what you want” recommendations do help. I didn’t want 200 miles badly enough to hurt more for longer and risk months of recovery (or the inability to recover). Maybe you’d be lucky enough to achieve 200 without hurting that bad, that long, or risking injury – or maybe you’ll have to make that choice, and you might make it differently than I did. (Maybe you’re lucky enough to not have 5 autoimmune things to juggle! I hope you don’t have to!) I kind of knew going in that I was only going to hit 200 if all went perfect.

Diabetes and this 200 mile ultramarathon that was a 100 mile ultra:

I just realized that I managed to write an ENTIRE race report without talking about diabetes and glucose management…because I had zero diabetes-related thoughts or issues during these several days of my run! Sweet! (Pun fully intended.)

Remember, I have type 1 diabetes and use an open source automated insulin delivery (AID) system (in my case, still using OpenAPS after alllllll these years), and I’ve talked previously about how I fuel while ultrarunning and juggling blood glucose management. Unlike previous ultras, I had zero pump site malfunctions (phew) and my glucose stayed nicely in range throughout. I think I had one small drift above range for 2 hours due to an hour of higher carb activity right when I shifted to walking the second lap on day 2, but otherwise was nicely in range all days and all nights without any extra thought or energy expended. I didn’t have to take a single “low carb”/hypoglycemia treatment! I think there was one snack I took a few minutes early when I saw I was drifting down slightly, but that was mostly a convenience thing and I probably would not have gone low (below target) even if I had waited for my planned fuel interval. But out of 46 snacks, only one 5-10 minutes early is impressive to me.

I had no issues after each day’s run, either: OpenAPS seamlessly adjusted to the increasing insulin sensitivity (using “autosensitivity” or “autosens”) so I didn’t have to do manual profile shifts or overrides or any manual interference. I did decide each night whether I wanted to let it SMB (supermicrobolus) as usual or stick to temp basal only to reduce the risk of hypoglycemia, but I had no post-dinner or overnight lows at all.

The most “work” I had to do was deciding to wear a second CGM sensor (staggered, 5 days after my other one started) so that I had a CGM sensor session going with good quality data that I could fall back to if my other sensor started to get jumpy, because the sensor session was supposed to end the night of day 4 of my planned run. I obviously didn’t run day 4, but even so I was glad to have another sensor going (worth the cost of overlapping my sensors) in order to have the reassurance of constant data if the first one died or fell out and I could seamlessly switch to an already-warmed up sensor with good data. I didn’t need it, but I was glad to have done that in prep.

(Because I didn’t talk about diabetes a lot in this post, because it was not very relevant to my experiences here, you might want to check out my previous race recaps and posts about utlrarunning like this one where I talk in more detail about balancing fueling, insulin, and glucose management while running for zillions of hours.)

—

TLDR: I ran 100 miles, and I did it my DIY way: my own course, my own (slow pace), with sleep breaks, a lot of fueling, and a lot of satisfaction of setting big goals and attempting to achieve them. I think for me, the process goals of figuring out how to even safely attempt ultramarathons are even more rewarding than the mileage milestones of ultrarunning.

Running a multi-day ultramarathon by Dana M. Lewis from DIYPS.org

New Systematic Review of Exocrine Pancreatic Insufficiency (EPI) In Type 1 Diabetes and Type 2 Diabetes – Focusing on Prevalence and Treatment

by Dana Lewis

I’m thrilled that the research I did evaluating the prevalence and treatment of EPI in both Type 1 diabetes and Type 2 diabetes (also presented as a poster at #ADA2023 – read a summary of the poster here) has now been published as a full systematic review in Diabetes Technology and Therapeutics.

Here is a pre-edited submitted version of my article that you can access if you don’t have journal access; and as a reminder, copies of ALL of my research articles are available on this page: DIYPS.org/research!

And if you don’t want to read the full paper, this is what I think you should take away from it as a person with diabetes or as a healthcare provider:

    1. What is EPI? 

      Exocrine pancreatic insufficiency (known as EPI in some places, and PEI or PI in other places) occurs when the pancreas no longer produces enough enzymes to digest food. People with EPI take pancreatic enzyme replacement therapy (PERT) whenever they eat (or drink anything with fat/protein).

    2. If I have diabetes, or treat people with diabetes, why should I be reading the rest of this about EPI?EPI often occurs in people with cystic fibrosis, pancreatitis, and pancreatic cancer. However, since these diseases are rare (think <0.1% of the general population even when these groups are added up all together), the total number of people with EPI from these causes is quite low. On the other hand, EPI is also common in people with diabetes, but this is less well-studied and understood. The research on other co-conditions is more frequent and often people confuse the prevalence WITHIN those groups with the % of those conditions occurring overall in the EPI community.This paper reviews every paper that includes data on EPI and people with type 1 diabetes or type 2 diabetes to help us better understand what % of people with diabetes are likely to face EPI in their lifetime.
    3. How many people with type 1 diabetes or type 2 diabetes (or diabetes overall) get EPI?TLDR of the paper: EPI prevalence in diabetes varies widely, reported between 5.4% and 77% when the type of diabetes isn’t specified. For Type 1 diabetes, the median EPI prevalence is 33% (range 14-77.5%), and for Type 2 diabetes, the median is 29% (range 16.8-49.2%). In contrast, in non-diabetes control groups, the EPI prevalence ranges from 4.4% to 18% (median 13%). The differences in ranges might be due to geographic variability and different exclusion criteria across studies.Diabetes itself is prevalent in about 10% of the general population. As such, I hypothesize that people with diabetes likely constitute one of the largest sub-groups of individuals with EPI, in contrast to what I described above might be more commonly assumed.
    4. Is pancreatic enzyme replacement therapy (PERT) safe for people with diabetes? 

      Yes. There have been safety and efficacy studies in people with diabetes with EPI, and PERT is effective just like in any other group of people with EPI.

    5. What is the effect of pancreatic enzyme replacement therapy (PERT) on glucose levels in people with diabetes?
      PERT itself does not affect glucose levels, but PERT *d0es* impact the digestion of food, which then changes glucose levels! So, most PERT labels warn to watch for hypoglycemia or hyperglycemia but the medicine itself doesn’t directly cause changes in glucose levels. You can read a previous study I did here using CGM data to show the effect of PERT actually causing improved glucose after meals in someone with Type 1 diabetes. But, in the systematic review, I found only 4 articles that even made note of glucose levels, and only 1 (the paper I linked above) actually included CGM data. Most of the studies are old, so there are no definitive conclusions on whether hypoglycemia or hyperglycemia is more common when a person with diabetes and EPI starts taking PERT. Instead, it’s likely very individual depending on what they’re eating, insulin dosing patterns before, and whether they’re taking enough PERT to match what they’re eating.TLDR here: more studies are needed because there’s no clear single directional effect on glucose levels from PERT in people with diabetes.Note: based on the n=1 study above, and subsequent conversations with other people with diabetes, I hypothesize that high variability and non-optimal post-meal glucose outcomes may be an early ‘symptom’ of EPI in people with diabetes. I’m hoping to eventually generate some studies to evaluate whether we could use this type of data as an input to help increase screening of EPI in people with diabetes.
    6. How common is EPI (PEI / PI) compared to celiac and gastroparesis in Type 1 diabetes and Type 2 diabetes? 

      As a person with (in my case, Type 1) diabetes, I feel like I hear celiac and gastroparesis talked about often in the diabetes community. I had NEVER heard of EPI prior to realizing I had it. Yet, EPI prevalence in Type 1 and Type 2 diabetes is much higher than that of celiac or gastroparesis!The prevalence of EPI is much higher in T1 and T2 than the prevalence of celiac and gastroparesis.Celiac disease is more common in people with diabetes (~5%) than in the general public (0.5-1%). Gastroparesis, when gastric emptying is delayed, is also more common in people with diabetes (5% in PWD).However, the  prevalence of EPI is 14-77.5% (median 33%) in Type 1 diabetes and 16.8-49.2% (median 29%) in Type 2 diabetes (and 5.4-77% prevalence when type of diabetes was not specified). This again is higher than general population prevalence of EPI.

      This data emphasizes that endocrinologists and other diabetes care providers should be more prone to initiate screening (using the non-invasive fecal elastase test) for individuals presenting with gastrointestinal symptoms, as the rates of EPI in diabetes are much higher in both Type 1 and Type 2 diabetes than the rates of celiac and gastroparesis.

    7. What should I do if I think I have EPI?
      Record your symptoms and talk to your doctor and ask for a fecal elastase (FE-1) screening test for EPI. It’s non-invasive. If your results are less than or equal to 200 (μg/g), this means you have EPI and should start on PERT. If you or your doctor feel that your sample may have influenced the results of your test, you can always re-do the test. But if you’re dealing with diarrhea, going on PERT may resolve or improve the diarrhea and improve the quality of the sample for the next test result. PERT doesn’t influence the test result, so you can start PERT and re-run the test any time.Symptoms of EPI can vary. Some people experience diarrhea, while others experience constipation. Steatorrhea or smelly, messy stools that stick to the side of the toilet are also common EPI symptoms, as is bloating, abdominal pain, and generally not feeling well after you eat.

      If you’ve been diagnosed with EPI, you may also want to check out some of my other posts (DIYPS.org/EPI) about my personal experiences with EPI and also this post about the amount of enzymes needed by most people with EPI. You may also want to check out PERT Pilot, a free iOS app, for recording and evaluating your PERT dosing.

—

If you want to read the full article, you can find copies of all of my research articles at DIYPS.org/research

If you’d like to cite this specific article in your future research, here’s an example citation:

Lewis, D. A Systematic Review of Exocrine Pancreatic Insufficiency Prevalence and Treatment in Type 1 and Type 2 Diabetes. Diabetes Technology & Therapeutics. http://doi.org/10.1089/dia.2023.0157

Why DIY AID in 2023? #ADA2023 Debate

by Dana Lewis

I was asked to participate in a ‘debate’ about AID at #ADA2023 (ADA Scientific Sessions), representing the perspective that DIY systems should be an option for people living with diabetes.

I present this perspective as a person with type 1 diabetes who has been using DIY AID for almost a decade (and as a developer/contributor to the open source AID systems used in DIY) – please note my constant reminder that I am not a medical doctor.

Dr. Gregory P. Forlenza, an Associate Professor from Barbara Davis Center, presented a viewpoint as a medical doctor practicing in the US.

FYI: here are my disclosures and Dr. Forlenza’s disclosures:

On the left is my slide (Dana M. Lewis) showing I have no commercial support or conflicts of interest. My research in the last 3 years has previously been funded by the New Zealand Health Research Council (for the CREATE Trial); JDRF; and DiabetesMine. Dr. Forlenza lists research support from NIH, JDRF, NSF, Helmsley Charitable Trust, Medtronic, Dexcom, Abbott, Insulet, Tandem, Beta Bionics, and Lilly. He also lists Consulting/Speaking/AdBoard: Medtronic, Dexcom, Abbott, Insulet, Tandem, Beta Bionics, and Lilly.

I opened the debate with my initial presentation. I talk about the history of DIY in diabetes going back to the 1970s, when people with diabetes had to “DIY” with blood glucose meters because initially healthcare providers did not want people to fingerstick at home because they might do something with the information. Similarly, even insulin pumps and CGMs have been used in different “DIY” ways over the years – notably, people with diabetes began dosing insulin using CGM data for years prior to them being approved for that purpose. It’s therefore less of a surprise in that context to think about DIY being done for AID. (If you’re reading this you probably also know that DIY AID was done years before commercial AID was even available; and that there are multiple DIY systems with multiple pump and CGM options, algorithms, and phone options).

And, for people with diabetes, using DIY is very similar to how a lot of doctors recommend or prescribe doing things off label. Diabetes has a LOT of these types of recommendations, whether it’s different types of insulins used in pumps that weren’t approved for that type of insulin; medications for Type 2 being used for Type 1 (and vice versa); and other things that aren’t regulatory approved at all but often recommended anyway. For example, GLP-1’s that are approved for weight management and not glycemic control, but are often prescribed for glycemic control reasons. Or things like Vitamin D, which are widely prescribed or recommended as a supplement even though it is not regulatory-approved as a pharmaceutical agent.

I always like to emphasize that although open source AID is not necessarily regulated (but can be: one open source system has received regulatory clearance recently), that’s not a synonym for ‘no evidence’. There’s plenty of high quality scientific evidence on DIY use and non-DIY use of open source AID. There’s even a recent RCT in the New England Journal of Medicine, not to mention several other RCTs (see here and here, plus another pending publication forthcoming). In addition to those gold-standard RCTs, there are also reviews of large-scale big data datasets from people with diabetes using AID, such as this one where we reviewed 122 people’s glucose data representing 46,070 days’ worth of data; or another forthcoming publication where we analyzed the n=75 unique (distinct from the previous dataset) DIY AID users with 36,827 days’ of data (average of 491 days per participant) and also found above goal TIR outcomes (e.g. mean TIR 70-180 mg/dL of 82.08%).

Yet, people often choose to DIY with AID not just for the glucose outcomes. Yes, commercial AID systems (especially now second-generation) can similarly reach the goal of 70+% TIR on average. DIY helps provide more choices about the type and amount of work that people with diabetes have to put IN to these systems in order to get these above-goal OUTcomes. They can choose, overall or situationally, whether to bolus, count carbs precisely, announce meals at all, or only announce relative meal size while still achieving >80% TIR, no or little hypoglycemia, and less hyperglycemia. Many people using DIY AID for years have been doing no-bolus and/or no meal announcements at all, bringing this closer to a full closed loop, or at least, an AID system with very, very little user input required on a daily basis if they so choose. I presented data back in 2018(!) showing how this was being done in DIY AID, and it was recently confirmed in a randomized control trial (hello, gold standard!) showing that between traditional use (with meal announcements and meal boluses); meal announcement only (no boluses); and no announcement nor bolusing, that they all got similar outcomes in terms of TIR (all above-goal). There was also no difference in those modes of total daily insulin dose (TDD) or amount of carb intake. There was a small difference in time below range being slightly higher in the first mode (where people were counting carbs and bolusing) as compared to the other two modes – which suggests that MORE user input may actually be limiting the capabilities of the system!

The TLDR here is that people with diabetes can do less work/provide less input into AID and still achieve the same level of ideal, above-goal outcomes – and ongoing studies are showing the increased QOL and other patient-reported outcomes that also improve as a result.

Again, people may be predisposed to think that the main difference between commercial and DIY is whether or not it is regulatory approved (and therefore prescribable by doctors and able to be supported by a company under warranty); the bigger differences are instead around interoperability across devices, data access, and transparency of how the system works.

There’s even an international consensus statement on open source AID, created by an international group of 48 medical and legal experts, endorsed by 9 national and international diabetes organizations, supporting that open source AID used in DIY AID is a safe and effective treatment option, confirming that the scientific evidence exists and it has the potential to help people with diabetes and reduce the burden of diabetes. They emphasize that doctors should support patient (and caregiver) autonomy and choice of DIY AID, and state that doctors have a responsibility to learn about all options that exist including DIY. The consensus statement is focused on open source AID but also, in my opinion, applies to all AID: they say that AID systems should fully disclose how they operate to enable informed decisions and that all users should have real-time and open access to their own data. Yes, please! (This is true of DIY but not true of all commercial systems.)

The elephant in the room that I always bring up is cost, insurance coverage, and therefore access and accessibility of AID. Many places have government or insurance that won’t cover AID. For example, the proposed NICE guidelines in the UK wouldn’t provide AID to everyone who wants one. In other places, some people can get their pump covered but not CGM, or vice versa, and must pay out of pocket. Therefore in some cases, DIY has out of pocket costs (because it’s not covered by insurance), but is still cheaper than AID with insurance coverage (if it’s even covered).

I also want to remind everyone that choosing to DIY – or not – is not a once-in-a-lifetime decision. People who use DIY choose every day to use it and continue to use it; at any time, they could and some do choose to switch to a commercial system. Others try commercial, switch back to DIY, and switch back and forth over time for various reasons. It’s not a single or permanent decision to DIY!

The key point is: DIY AID provides safety and efficacy *and* user choice for people with diabetes.

—

Dr. Forlenza followed my presentation, talking about commercial AID systems and how they’ve moved through development more quickly recently. He points to the RCTs for each approved commercial system that exist, saying commercial AID systems work, and describing different feature sets and variety across commercial systems. He shared his thoughts on advantages of commercial systems including integration between components by the companies; regulatory approval meaning these systems can be prescribed by healthcare providers; company-provided warranties; and company provided training and support of healthcare providers and patients.

He makes a big point about a perceived reporting bias in social media, which is a valid point, and talks about people who cherry pick (my words) data to share online about their TIR.

He puts an observational study and the CREATE Trial RCT data up next to the commercial AID systems RCT data, showing how the second generation commercial AID reach similar TIR outcomes.

He then says “what are you #notwaiting for?”, pointing out in the US that there are 4 commercial systems FDA approved for type 1 diabetes. He says “Data from the DIY trials themselves demonstrate that DIY users, even with extreme selection bias, do not achieve better glycemic control than is seen with commercial systems.” He concludes that commercial AID has a wide variety of options; commercial systems achieve target-level outcomes; a perception that both glucose outcomes and QOL are being addressed by the commercial market, and that “we do not need Unapproved DIY solutions in this space”.

—

After Dr. Forlenza’s presentation, I began my rebuttal, starting with pointing out that he is incorrectly conflating perceived biases/self-reporting of social media posts with gold-standard, rigorously performed scientific trials evaluating DIY. Data from DIY AID trials do not suffer from ‘selection bias’ any more than commercial AID trials do. (In fact, all clinical trials have their own aspects of selection bias, although that isn’t the point here.) I reminded the audience of the not one but multiple RCTs available as well as dozens of other prospective and retrospective clinical trials. Plus, we have 82,000+ data points analyzed showing above-goal outcomes, and many studies that evaluate this data and adjust for starting outcomes still show that people with diabetes who use DIY AID benefit from doing so, regardless of their starting A1c/TIR or demographics. This isn’t cherry-picked social media anecdata.

When studies are done rigorously, as they have been done in DIY, we agree that now second-generation commercial AID systems reach (or exceed, depending on the system) ADA standard of care outcomes. For example, Dr. Forlenza cited the OP5 study with 73.9% TIR which is similar to the CREATE Trial 74.5% TIR.

My point is not that commercial systems don’t work; my point is that DIY systems *do* work and that the fact that commercial systems work doesn’t then override the fact that DIY systems have been shown to work, also! It’s a “yes, and”! Yes, commercial AID systems work; and yes, DIY AID systems work.

The bigger point, which Dr. Forlenza does not address, is that the person with diabetes should get to CHOOSE what is best for them, which is not ONLY about glucose outcomes. Yes, a commercial system- like DIY AID – may help someone get to goal TIR (or above goal), but DIY provides more choice in terms of the input behaviors required to achieve those outcomes! There’s also possible choice of systems with different pumps or CGMs, different (often lower) cost, increased data access and interoperability of data displays, different mobile device options, and more.

Also, supporting user choice of DIY is in fact A STANDARD OF CARE!

It’s in the ADA’s Standards of Care, in fact, as I wrote about here when observing that it’s in the 2023 Standards of Care…as well as in 2022, 2021, 2020, and 2019!

I wouldn’t be surprised if there are people attending the debate who think they don’t have any – or many – patients using DIY AID. For those who think that (or are reading this thinking the same), I ask a question: how many patients have you asked if they are using DIY AID?

There’s a bunch of reasons why it may not come up, if you haven’t asked:

  • They may use the same consumables (sites, reservoirs) with a different or previous pump in a DIY AID system.
  • Their prescribed pump (particularly in Europe and non-US places that have Bluetooth-enabled pumps) may be usable in a DIY AID.
  • They may not be getting their supplies through insurance, so their prescription doesn’t match what they are currently using.
  • Or, they have more urgent priorities to discuss at appointments, so it doesn’t come up.
  • Or, it’s also possible that it hasn’t come up because they don’t need any assistance or support from their healthcare provider.

Speaking of learning and support, it’s worth noting that in DIY AID, because it is open source and the documentation is freely available, users typically begin learning more about the system prior to initiating their start of closed loop (automated insulin delivery). As a result, the process of understanding and developing trust in the system begins prior to closed loop start as well. In contrast, much of the time there is limited available education prior to receiving the prescription for a commercial AID; it often aligns more closely with the timeline of starting the device. Additionally, because it is a “black box” with fewer available details about exactly how it works (and why), the process of developing trust can be a slower process that occurs only after a user begins to use a commercial device.

With DIY AID, because it is open source and the documentation is freely available, users typically begin learning more about the system prior to initiating their start of closed loop (automated insulin delivery). As a result, the process of understanding and developing trust in the system begins prior to closed loop start as well. In contrast, much of the time there is limited available education prior to receiving the prescription for a commercial AID; it often aligns more closely with the timeline of starting the device. Additionally, because it is a black box with less available details about exactly how it works (and why), the process of developing trust can be a slower process that occurs only after a user begins to use a commercial device. The learning & trust in AID timelines is something that needs more attention in commercial AID moving forward.

I closed my rebuttal section by asking a few questions out loud:

I wonder how healthcare providers feel when patients learn something before they do – which is often what happens with DIY AID. Does it make you uncomfortable, excited, curious, or some other feeling? Why?

I encouraged healthcare providers to consider when they are comfortable with off-label prescriptions (or recommending things that aren’t approved, such as Vitamin D), and reflect on how that differs from understanding patients’ choices to DIY.

I also prompted everyone to consider whether they’ve actually evaluated (all of) the safety and efficacy data, of which many studies exist. And to consider who benefits from each type of system, not only commercial/DIY but individual systems within those buckets. And to consider who gets offered/prescribed AID systems (of any sort) and whether subconscious biases around tech literacy, previous glucose outcomes, and other factors (race, gender, other demographic variables) result in particular groups of people being excluded from accessing AID. I also remind everyone to think about what financial incentives influence access and available of AID education, and where the education comes from.

—

Although Dr. Forlenza’s  rebuttal followed mine, I’ll summarize it here before finishing a recap of my rebuttal: he talks about individual selection bias/cherry picked data, acknowledging it can occur in anecdotes with commercial systems as well; talks about the distinction of regulatory approval vs. off label and unapproved; legal concerns for healthcare providers; and closes pointing out that many PWD see primary care providers, he doesn’t believe it is reasonable to expect PCPs to become familiar with DIY since there are no paid device representatives to support their learning, and that growth of AID requires industry support.

—

People probably wanted to walk out of this debate with a black and white, clear answer on what is the ‘right’ type of AID system: DIY or commercial. The answer to that question isn’t straightforward, because it depends.

It depends on whether a system is even AVAILABLE. Not all countries have regulatory-approved systems available, meaning commercial AID is not available everywhere. Some places and people are also limited by ACCESSIBILITY, because their healthcare providers won’t prescribe an AID system to them; or insurance won’t cover it. AFFORDABILITY, even with insurance coverage, also plays a role: commercial AID systems (and even pump and CGM components without AID) are expensive and not everyone can afford them. Finally, ADAPTABILITY matters for some people, and not all systems work well for everyone.

When these factors align – they are available, accessible, affordable, and adaptable – that means for some people in some places in some situations, there are commercial systems that meet those needs. But for other people in other places in other situations, DIY systems instead or also can meet that need.

The point is, though, that we need a bigger overlap of these criteria! We need MORE AID systems to be available, accessible, affordable, and adaptable. Those can either be commercial or DIY AID systems.

The point that Dr. Forlenza and I readily agree on is that we need MORE AID – not less.

This is why I support user choice for people with diabetes and for people who want – for any variety of reasons – to use a DIY system to be able to do so.

People probably want a black and white, clear answer on what is the ‘right’ type of AID system: DIY or commercial. It depends on whether a system is even AVAILABLE. Not all countries have regulatory-approved systems available, meaning commercial AID is not available everywhere. Some places and people are also limited by ACCESSIBILITY, because their healthcare providers won’t prescribe an AID system to them; or insurance won’t cover it. AFFORDABILITY, even if insurance coverage, also plays a role: commercial AID systems (and even pump and CGM components without AID) are expensive and not everyone can afford them. Finally, ADAPTABILITY matters for some people, and not all systems work well for everyone. The point is that we need a bigger overlap of these criteria! We need more alignment of these factors - more AID (DIY and commercial) available, accessible, affordable, and adaptable for people with diabetes. I support user choice for people with diabetes, which includes DIY AID systems

—

PS – I also presented a poster at #ADA2023 about the high prevalence rates of exocrine pancreatic insufficiency (EPI / PEI / PI) in Type 1 and Type 2 diabetes – you can find the poster and a summary of it here.