Pacing with ME/CFS: Why I built a small tool that promises nothing

The phone rings. Someone I had not spoken to in five years. I pick up. She is crying.

“You do something with AI, right? Can you help us?”

Someone in her family has ME/CFS. Severely affected: bedridden around the clock, needing help with eating and drinking. The family is trying to manage their presence, but without structure, by feel. And that instinct almost always misleads, because the response to exertion comes days later.

I am not a doctor. I am a person who happens to be able to build AI tools right now. So I built one. Free, no account. I am sharing it here because I believe other families are in exactly this situation.

One thing at a time: first the illness, then the tool.

What ME/CFS is

ME/CFS stands for Myalgic Encephalomyelitis/Chronic Fatigue Syndrome. It is not fatigue in the everyday sense, not burnout under another name. It is a serious chronic systemic illness with no known causal therapy.

The core symptom is Post-Exertional Malaise (PEM): a worsening after exertion that typically appears 12 to 48 hours later. Whoever steers caregiving by current mood is steering by a signal already a day old. Pacing means staying below the exertion threshold before exhaustion becomes noticeable.

What Prof. Simon explains, and what it means for caregivers

The two-part video interview with Prof. Dr. Dr. Simon (sports medicine, University of Mainz) and Dr. Th. Weiss, embedded on me-cfs.net and available as a video on YouTube, is one of the clearest explanatory pieces I found. Fasynation.de adds the practical daily-life implications.

Prof. Simon's starting point: the organs themselves appear healthy. Heart, lungs, muscles, all technically fine. The problem lies in microcirculation. Blood flows past the muscle cells without releasing oxygen. The cascade then runs as follows: local oxygen deficiency in a small muscle area, from cutting bread, washing hair, climbing stairs, is reported to the brain. The brain raises breathing rate and heart rate. That does not fix the underlying problem, but causes other areas of the body to be flooded with too much oxygen. Oxidative stress develops, the immune system reacts. The result: severe exhaustion, typically two days after the exertion.

The 30-second threshold follows from this: up to that point, the body can still compensate for the insufficient oxygen supply via anaerobic metabolism. Beyond it, hypoxic damage to the tissue develops. This applies primarily to the physical self-exertion of the affected person.

What caught my attention in the video: Prof. Simon explicitly addresses cognitive and multisensory load. When someone simultaneously processes acoustic signals, takes in visual input, and thinks, small areas of the brain can no longer be adequately supplied with oxygen, the same pattern as in the muscle. The brain issues the same breathing signal, the same cascade begins. This explains why speech, presence, light and sounds from outside can have a measurable effect on severely affected individuals. It is not direct evidence that the pacing tool works physiologically in the same way as Simon's approach for self-exertion, but it is the mechanism that supports the logic behind it.

The rest in the tool therefore has no upper limit and no automatic return to green. Not because that was an arbitrary design choice, but because the physiology suggests it: when enough rest has passed, only the person being cared for can signal that.

What the tool can do

The result of this request is a simple browser app that requires no installation, no account and shares no data with third parties. Available at degeler.consulting/pacing. Free of charge. In German and English.

The screen shows two states. Green: interaction is running. Red: rest. When the set time runs out, green switches automatically to red. That actively ends the interaction without anyone needing to decide in the moment. Red never switches back to green automatically, because rests should have no upper limit.

Adjustable: the maximum interaction time (default 30 seconds), the recommended minimum rest (default 60 seconds), screen brightness for darkened rooms, a sound signal as a fallback and vibration on Android devices. Spacebar and Enter work as keyboard shortcuts. The tool works offline.

The tool records per day: how many cycles took place and how much interaction time in total. The following day, it can be marked whether the person was better or worse. This marker is the core of learning over time.

The group code: several caregivers, one shared history

In most families, several people take turns accompanying. The optional group code solves this: whoever enters the same code on different devices automatically shares the same history. No login, no account.

The code is hashed with SHA-256 in the browser before it reaches the server. The server only sees a cryptographic string, never the code itself. All data is stored on Degeler Consulting's server in Germany; no third-party provider is involved.

The analysis: what the data shows over time

After at least three marked days, the tool shows an analysis: how much interaction time was there on average on days before a good following day, and how much before a worse following day. From this, an estimate of the personal threshold emerges.

Not a medical value. Just a reference point from your own data that becomes more meaningful over time. The article inside the tool also explains how to account for the delay when reading the data: today's marker describes the response to yesterday, not to today.

Not a medical device

This tool is not based on clinically validated evidence for this use case. Anyone with questions about the diagnosis, progression or treatment of ME/CFS should consult a doctor.