By Aaron Abajian, MD
Every rehabilitation discipline runs on the same quiet bargain: the clinician provides assessment and a program, and the patient or family executes most of that program somewhere the clinician will never see. Physical therapy, occupational therapy, and speech-language pathology all call it something slightly different, but the home exercise program, the HEP, is where the dose actually lives. And the dose, very often, is not being taken.
What the adherence literature actually says
Precise numbers vary by discipline, population, and how adherence is measured, which is itself part of the problem. But studies consistently estimate home program adherence in the broad range of 30 to 50 percent, and self-reported adherence is widely understood to overstate the truth, since patients report to the person they do not want to disappoint. Whatever the exact figure for a given caseload, two findings recur across the literature: adherence declines over time from an initially motivated baseline, and clinicians systematically overestimate it.
It is worth being honest about the measurement fog rather than quoting a false-precision statistic. If roughly half of prescribed home practice does not happen, then for many patients the clinic visit is the minority of their intended treatment, and the majority is running unsupervised at an unknown and probably low rate.
Silence is the missing signal
Here is the structural flaw: in almost every practice, non-adherence produces no signal at all. A patient who stops their exercises generates no alert, no data point, no flag in the chart. The information surfaces days or weeks later, at the next visit, filtered through embarrassment, and by then the response options have narrowed to re-teaching and starting over.
Compare that with how clinicians treat every other meaningful variable. We would not accept a vital sign that reported only at monthly visits, only by patient recollection, and only when the news was good. Yet that is precisely the telemetry we accept for the single variable that most determines whether the program works: whether it is being done.
The fix is not surveillance and it is not more motivational conversation at the visit. It is a lightweight, recurring question, asked between visits, with an answer that is effortless to give and honest by design.
The churn link: families that stop practicing quit therapy
There is a second cost hiding behind the clinical one. A family that stops practicing does not usually announce it. They drift: practice slips, progress plateaus because the dose fell, the plateau reads as "therapy is not working," and discouragement does the rest. Cancellations get softer, then more frequent, and eventually the family is gone. The plateau gets blamed, but the quiet weeks that caused it were visible weeks earlier, to anyone who had a way to see them.
That makes adherence visibility a practice-health issue as well as a clinical one. The moment to save that family was the second quiet week, with a smaller program and a kinder conversation, not the intake call from the practice down the street six months later.
A loop that surfaces drift early
The mechanics of the fix are small enough to feel anticlimactic. With Cronote, it is one weekly check-in per family, one honest one-tap answer, and one alert when the answers stop.
- The weekly check-in: written once, repeating forever, delivered as a normal text or email. A caseload imports from a CSV. Families install nothing and create no account: one tap in the browser logs the answer.
- An honest answer set: "Done or We struggled." The struggling option is deliberately first-class, because an answer set that only accepts success teaches families to stop answering exactly when the signal matters. We struggled, logged three weeks running, is a clean indication to shrink or reshape the program.
- Accumulation, never streaks: the detail page shows which weeks each family logged, counting up. Nothing resets on a bad week, so a bad week costs nothing to admit.
- The silence alarm: "No answer? Notify me" alerts the clinician when a family has not answered for 10 days. Ten quiet days is drift; ten quiet weeks is churn. The alert arrives while it is still the former.
Practice week. Did you get your exercises in most days? Done or We struggled.
Message content stays free of clinical detail by design: the day, the activity, and the one-tap answer. No diagnosis, no goals, nothing about the patient's condition travels in the text. For clinics, a Business Associate Agreement is available.
What this does and does not claim
A check-in loop does not make anyone exercise, and no reminder system should claim otherwise. What it changes is what the clinician knows and when: which families are practicing, which are struggling, and which have gone quiet, surfaced weekly instead of discovered at the next visit. Clinicians consistently report that the earlier that information arrives, the more options they have, and families that stay engaged and honest are families you can keep adjusting for. The dose conversation gets to happen while it can still change the outcome of the plan of care.
The economics are deliberately boring: accounts are free, email and app delivery are unlimited at no cost, and texts run on one-time credit packs, $5 for 100, $15 for 500, $40 for 2,000, which never expire. A weekly text is about 4 to 5 texts per family per month, so a 40-family caseload runs about a year on the $40 pack. No subscription, no per-seat pricing, no procurement meeting.