Critical analysis

Under the trial chart on the Progress page, Critical analysis reads the chart back to you and answers the four questions people actually have when they look at it. The chart shows you where you are; this tells you what that means.

The four questions

• Your titration pace: how your climb up the ladder compares with the trial protocol’s, stated as timing and dose: “you reached 7.5 mg at week 14; that protocol was there by week 8.”
• Response for the drug you’ve taken: at your total milligrams so far, where was each cohort? This is the fairest comparison you have, and it’s the one people mean when they ask “is it working for me?”
• Have you maxed out?: two kinds of room. Runway: what the cohort went on to do after your week: being at −10% in week 20 isn’t a ceiling if that cohort was only two-thirds of the way to their own final result at the same point. Ladder: how many dose rungs are above you, and how long those cohorts had already been at their top dose.
• What could mislead you: the one thing most likely to trip you up in your picture, rather than a generic list of caveats.

Encouragement, but only the honest kind

The chart freezes each cohort at the week you’re standing on, which quietly makes it look like their story ended there. It didn’t. In SURMOUNT-1 the 10 mg cohort was at −13.1% in week 20 and finished at −19.5%: most of the rest of their loss came later. That’s a real, published fact, and it’s the clearest answer to “have I maxed out?”

But it’s what happened to them. “That cohort went on to lose another 6.4 points after this week” is reporting; “you have another 6.4 points coming” would be a prediction, and you won’t see one here. Room existing isn’t room promised.

Pace is a measurement, not a verdict

It will tell you that you climbed to your dose six weeks later than a protocol did, because that’s a fact about two schedules. It will not tell you that you’re titrating too slowly, or that you should move: that’s a dose opinion, and it isn’t ours to hold. Same with headroom: “you have three rungs left” is arithmetic; “use them” is a prescribing decision.

Will a slower climb pay off later?

This is the question everyone wants a yes to, and the honest answer is that nothing here can settle it. The trials never randomised titration speed, every arm followed one fixed schedule, so there’s simply no comparison in the data between climbing slowly and climbing fast. Reaching a given loss on less drug is a real observation about what has happened so far. It isn’t evidence that the slower climb caused it, and it isn’t a promise of more to come. The analysis is meant to name that hope and then be straight with you about it.

What it will never do

It will not tell you what dose to take. Not directly, not as a hint, not as “most people would”. Dose decisions belong to you and the person who prescribes for you.

• No dose or titration recommendations, and no advice about starting or stopping.
• No predictions about your final result: an early position is weak evidence about a later one.
• No claim that your choices caused your results. One person without a control group can’t show that.
• No numbers of its own: it can only use the figures already on your panel, so it can’t invent a statistic or recall one from a study.

How it works

It’s AI-generated, and it works only from the figures already measured on your panel: it explains them, it never recalculates them, so it cannot arrive at a number of its own. It is rewritten when those figures change (a new weigh-in, a dose change, another shot) rather than every time you open the page.

Those figures are the same ones it reads, and you can see them yourself under “The figures behind this” at the bottom of the panel. Nothing is hidden: the reading is just easier to follow than the bare findings, so the findings sit one click away instead of on top. If the reading can’t be produced, they take over the panel so you still get your numbers.

AI interpretation of your own tracking data. Not medical advice.