Quality of Life Disparities in Chronic Health Conditions: A Social Model Analysis - Prompt

Below is the prompt I use to generate the report and supplementary documents. I used Claude Opus 4.1, on 2025-08-28, via claude.ai.

At the bottom of this page are some of my thoughts on how the parameters of the prompt may have influenced the outcome.

I am interested in comparing how patients with different chronic health conditions rate their quality of life. For instance, does someone living with multiple sclerosis generally have a better quality of life than someone living with arthritis? Gather data from appropriate and high quality sources and ensure it is of good trustworthiness. I am not looking for medical information or advice here, I am researching the social aspects of disability and patient's ability to manage them within the society they live in, as part of the social model of disability - a foundational core of the approach to this analysis.

I will use the terms "chronic health conditions" and "disabilities" or "disability" interchangeably in this prompt. You should use "chronic health condition" or similar when talking about the medical condition and symptoms, but "disability" when talking about the impact the medical condition and symptoms have on a patients ability to live an independent life to the fullest.

*Chronic health conditions to investigate for relative quality of life
I want you to research the following medical conditions:

And then additionally choose from a list of the most common chronic health conditions to make the total number of conditions in your analysis 20.

Exploring the 'Why'

I don't want this report to just be a simple bar chart with some sad numbers on it. I want to generate some actionable insights and some try to investigate why certain conditions have lower quality of life scores than others, and if they have anything in common that might be the reason behind that. Therefore, I want you to analyse in more depth what you have discovered so far and then research disability rights movements, patient advocacy groups, and other sources where patients talk about their experiences and their struggles. Why do chronic health condition sufferers feel the way they do? What could be done to improve their lives? Is there a potentially linked causal relationship? Is there any other key information do you think is important to highlight the difficulties chronic medical condition sufferers experience? Go above and beyond here, really impress me with your insights and your presentation -I want ineractive visualisations, tooltips, a range of chart types appropriate for the data type, and clear citations to the underlying data sources.

Outcomes, not outputs
I want you to create a few outputs for me, but your focus should always be on outcomes rather than outputs - what outcome are we trying to drive here? That outcome is to highlight disparities in quality of life for different chronic health conditions, and to try to investigate why those exist and if we can do anything to improve quality of life for those suffering the most from their disabilities.

I want you to create the following for me, all of which should contain the same data and analysis:

Style and format

Your writing style should be casual but not chatty. Be objective and avoid adding additional emotional weight to the report. Use British English and metric measurements.

Your visual style should be minimal and clean and be based on Google's Material Design 3, documentation here: https://m3.material.io/. Use teal as the primary colour, and amber as the accent colour.

Prompt biases

There are a few things in the prompt that may have influenced the output in a way that makes it less trustworthy.

  1. I specifically asked for analysis of a few of my own medical conditions. This potentially means I am biased in terms of what conditions are ranked where in the final output.
    1. I tried to mitigate this by asking for conditions I do not have (e.g. diabetes) and also asking it to analyse twenty conditions in total, adding in the most common chronic health conditions not mentioned.
      1. I think this adds enough "known uninvolveds" (e.g. diabetes) and "unknown uninvolveds" (i.e. the extra conditions Claude chose to investigate itself, which I only gave a vague "popular" conditions condition)
    2. What inspired me to investigate this matter was actually seeing someone say that ME/CFS has the lowest reported quality of life for chronic health conditions, so the fact that it also scored lowest in Claude's report should not be too surprising.
    3. The numbers are backed up by various data sources. I have not thoroughly investigated them yet, however. While I did ask Claude to investigate specific conditions, I did not limit how it should investigate them, or what sources it should use.

Overall, I think that this is not overly likely to have caused the output to be significantly biased in terms of underreporting the quality of life, or overly focussing on, medical conditions I myself have

  1. Geographically, there is a focus on the UK.
    1. After sending my initial prompt, Claude asked me if I wanted a specific geographical focus.
    2. I asked for a UK focus (where I live) but pulling in global data where it is of good quality.

I don't think this poses a problem for the impartiality of the output, given the intended focus is in the UK.

  1. Claude has a focus on the social model of disability, rather than the medical.
    1. This is pure bias from me. I fully believe that the social model provides a much more realistic, pragmatic, and humane approach to improving the lives of disabled people.
      1. If you disagree, well, come fite me >:3
    2. While I did ask Claude to perform the analysis with that focus, it did bring in various terms and concepts I did not mention (deliberately, to allow it to freewheel), such as invisible disabilities or the impact in income and employment.
    3. So, while I do think my prompt steering it very firmly towards concepts related to the social model, various other disablity concepts were pulled in too, and it created a compelling and enthusiastic case for the veracity and strength of the social model
      1. Like I said… come fite me, but this time with Claude too it seems 🤖🤝🐗

On a more serious note, I will admit the model was deliberately steered in that direction, though it did go much more enthusiastic than I wanted it too, despite me specifying this in the prompt:
Be objective and avoid adding additional emotional weight to the report.
I felt a little bit of sycophancy coming through, like Claude was trying to say how amazing I was for finally discovering the last insight to this long lost mystery - why are some disabled people sad? It is because society hates us :'( According to Claude, no one seems to have thought of that before.