The five-star ghost

A clinician I spoke with once pulled up the national satisfaction figures and went quiet. On the dimension closest to her own heart, the one the instrument calls bejegening en serieus nemen, how people feel treated and taken seriously, the score was 4.8 out of 5. Begrijpelijke uitleg, clear explanation, sat at 4.7. Near the ceiling. By those numbers, Dutch mental health care treats people as people, and treats them well. She wasn't proud of the number. She was confused by it. Because she kept thinking of the ones who weren't in it. The man who stopped coming after the fourth session and never said why. The woman who, somewhere around the intake, had quietly decided she was a case being processed, not a person being met. Where were they in the 4.8? The dropouts? The never-admitted? The ones who left without a word? Nowhere.

So which is true? The 4.8, or the patients she couldn't stop picturing? Here is the thing about that score, and it is the whole story. The CQi, the national client-experience index, is filled in after the treatment trajectory ends. Akwa, who maintain it, say so plainly: it is the questionnaire a client completes na afloop van het behandeltraject. Read that timing slowly. The people the survey can reach are, by design, the people who reached the end. The finishers.

The survey was never built to find them

And the others? The man who left after the fourth session is not in the sample. The woman who was never admitted, who is still on a waiting list somewhere, is not in the sample. The person who felt like a number and quietly stopped showing up did not fail to answer the survey. The survey was never built to reach her. So when you hold "bejegening 4.8" next to "niet gehoord, niet gezien", not heard, not seen, you are not looking at a contradiction to be resolved. You are looking at two instruments pointed at two different people.

This is not a Dutch flaw. It is what satisfaction surveys do everywhere, and we can put a number on it. In a study of one large patient-experience survey, only 16.5% of patients responded, and the ones who did differed systematically from the ones who stayed silent, in age, in circumstance, in how their care had gone (Tyser et al., 2016). The satisfied answer. The disappointed drift off. So the average drifts up, and the people most likely to have felt processed rather than known are exactly the people least likely to be counted. A high score is not a lie. It is a portrait of whoever stayed in the frame.

Where the data goes quiet, the wound speaks

You might ask, fairly, what about the angry reviews, the forums, the panels where "niet gehoord" is the loudest theme? Doesn't that data exist? It does, and it has to be read as what it is: a tail, not a survey. The Dutch review layer is in open crisis of trust. Only 9.2% of institutions on the largest review platform clear its own reliability threshold, and two of the country's medical bodies withdrew from its board over how hard the entries are to verify (Skipr, 2025). So you cannot quote the tail as a representative rate. But you can read its shape. And the shape keeps pointing at the same wound: being handled as a symptom-cluster rather than known as a person.

There is one cleaner number that maps straight onto it. When MIND asked more than 750 clients whether their care provider had made clear agreements with them about diagnosis and choice of treatment, more than 66% answered negatively (MIND, 2023). Two in three felt they had no real say in the naming of their own problem. That is not a complaint about waiting rooms or paperwork. That is the experience of being a case that decisions happen to, rather than a person decisions are made with. And the population that never reaches adequate care, the one the post-treatment survey can never see, is not shrinking. Across the NEMESIS-3 measurement window, the share of people with moderate or severe disorders receiving intensive treatment fell from 67.2% to 53.2%, while twelve-month prevalence among adults rose from 17% to 26% (Trimbos, 2023). More people unwell, a smaller slice reaching the intensive help, and none of that absence shows up in a survey of the ones who finished.

So take the patients on your own list this month, and ask the quiet version of the question. For how many of them would a satisfaction score, collected at the very end, honestly capture how seen they felt at the start? The 4.8 belongs to the people who made it to the end of the road. "Niet gehoord" belongs to the ones who turned off it. The wound is real precisely where the data is silent, because the silence and the satisfaction are measuring two different people.

This is the part where Freudche does something narrow. It holds the thread of who a person is across the whole arc of treatment, so that when you open the next session you find the through-line waiting for you: not a fresh symptom list to re-derive, but the same human being you sat with last time, picked up where the two of you left off. Each hour builds on the last instead of starting the inventory over. It does not diagnose, and the clinical judgement, what this person needs and where the work goes next, stays entirely yours. What it lifts is the quiet pressure that turns a person into a cluster: the having to rebuild someone from scratch under the clock, the moment that makes even a careful clinician sound, for one session, like she is tracking a framework and not a face.

The patients in that 4.8 felt treated as people because, for them, the road held together long enough to be remembered well. The ghost in the score is the one for whom it didn't, the one who became a number somewhere in the gap between one session and the next, and left before anyone could see it happen. The point was never to push the average higher. It was to keep the person in view long enough that fewer of them ever turn into the silence the survey can't hear.

Researched by Freudche.