Treatment mode
A per-patient care approach for appetite taken by medical treatment rather than by an eating disorder. Chemotherapy is the usual case. It strips out everything that motivates or scores eating, adds one-tap symptom tags, and goes silent about food around treatment days.
Some patients are not avoiding food. They cannot eat because treatment has taken their appetite, damaged their mouth, or turned their sense of taste inside out. Chemotherapy is the case this was built for; the same picture holds for radiotherapy and transplant.
Almost everything Pippa does for an eating disorder is wrong for them. Coins and streaks turn eating into a performance, the post-meal check-in asks about urges that do not exist, and the celebration copy congratulates someone for something they had no control over. Treatment mode removes all of it and leaves a log, a symptom tag, and your dashboard.
Off by default, and off means the app has no treatment-mode surface at all. Turn it on per patient in Care approach. It clears ARFID mode and sets structured, because a child on chemotherapy given eating-disorder machinery gets the wrong advice.
Why the app stops cheering
The instinct is to encourage. The evidence says do not.
Families in paediatric oncology already reach for pressure: bribes, bargaining, the threat of a feeding tube. Mealtime conflict is common and it makes intake worse, not better. The standing guidance is to reduce demanding feeding behaviour and treat the side effects instead. Appetite loss here is outside the patient's control, and trying harder is not a lever they have.
So in this mode nothing pleasant is earned by eating. The pet stays, because a familiar face is company rather than pressure, but every reward that was triggered by a meal is switched off: the participation coin, the streak coin, the new-food bonus, and the water reward. Streaks and the coin balance disappear from the patient's home screen entirely.
What you get instead
Symptom tags
The patient can tag why eating was hard, in one tap, from a fixed list: feeling sick, mouth hurts, tasted wrong, smell was too much, full really fast, too tired, just not hungry.
Crucially these attach to a skipped meal as well as an eaten one. This is the part worth paying attention to:
A missed meal is a gap in a chart. A missed meal tagged "mouth hurts" is a mucositis report you can treat.
Better control of nausea and mucositis measurably reduces taste disturbance, so the symptom is usually the actionable end of the problem, not the intake number.
Going down lately
A rolling fourteen-day list of foods that actually went down, newest first. Taste change affects around 60% of children on chemotherapy, and the direction varies per child and shifts over time — some swing from sweet to savoury mid-treatment. A static preference list would be wrong within a week, so this is deliberately a live document.
It answers the question every parent asks at five o'clock: what can he actually eat right now? Build from what is on it today rather than from what worked last month.
Foods logged with a tastes-wrong, smell, or mouth-hurts tag are excluded. A list headed "going down" that included them would cost the list its only job.
Eats per day
A simple count per day for the last two weeks, with treatment days greyed out and marked tx.
There is no risk styling anywhere on this card and no alerting. A patient on chemotherapy not eating is not news to their oncologist, and an alert that fires most days is an alert nobody reads.
Treatment days, and why the app goes quiet
Add infusion dates in Care approach as you schedule them. On those days, and the evening before, the app says nothing about food. No reminders, no prompts, no suggestions. Logging still works normally; Pippa simply stops speaking.
This is the piece with the strongest evidence behind it and the least obvious mechanism, so it is worth understanding rather than just switching on.
Food eaten close to chemotherapy can pick up a lasting aversion from a single pairing. This is conditioned taste aversion, and it is why the "scapegoat" technique works: give a strongly flavoured novel sweet between the meal and the infusion and the novel flavour absorbs the aversion instead of the meal. Children given a scapegoat were twice as likely to eat the test meal later.
The consequence for an app is uncomfortable. A well-meant nudge toward something the patient will actually eat, landing in that window, can destroy that food for them for years. A missed reminder costs one meal.
Past dates stay in the list, greyed rather than deleted, because they are the reason last week looks the way it does on the card.
Known limit: the silence window is applied when the app rebuilds its notification schedule, which happens on app open. A phone left closed across a treatment day can still fire a reminder that was armed earlier.
What stays on
Logging in every form, the day view, water, the meal plan, and your whole dashboard. Gentle meal reminders stay too: in this group the failure is forgetting rather than avoiding, and the reminders never mention goals, numbers, or streaks. They are silenced inside the treatment window.
What to tell the family
Two things are worth saying out loud at the start:
- Nothing in the app is keeping score. There is no streak to break and no target to miss.
- Tagging a skipped meal is the most useful thing they can do. It is not an admission of failure; it is the message that reaches you.
Interactions
| Setting | Behaviour |
|---|---|
| ARFID mode | Cleared automatically. The two are mutually exclusive. |
| Intuitive eating | Cleared automatically; treatment mode sets structured. |
| Finished-plate photo | Allowed, off by default. Here it is straightforwardly useful, since intake is the medical question. |
| Post-meal check-in | Hidden. Symptom tags replace it. |
Android currently syncs the mode flag and applies the reward changes; the symptom chips and the stripped home screen are on iOS first.