A health service accounts for itself in numbers, and the numbers are few and old: average length of stay, occupancy rate, response times, output — appointments, surgeries, scans. In Portugal, guaranteed maximum response times and production targets structure much of what a hospital has to report.
None of these indicators sees whether someone was frightened.
This is not a complaint about management
Worth saying, because the easy criticism is unfair. These indicators exist for a good reason: they are auditable. They are counted the same way everywhere, do not depend on who records them, and allow comparison. A public system spending everyone's money needs exactly that.
The problem is not that the indicators are bad. It is that they are incomplete in a biased way: what is easy to count gets in, what is hard to count stays out — and then it starts to look as though what stayed out does not exist.
What stays out
- Whether the patient was frightened during the scan.
- Whether the hour before surgery was spent in panic or in calm.
- Whether the person with dementia spent the afternoon agitated or settled.
- Whether the family was present or left outside.
- Whether someone was treated with dignity on the worst day of their life.
None of this is an indicator. So none of it competes for budget, appears in a contract, or justifies an investment decision.
The practical consequence
This is where it stops being philosophy.
An intervention that improves only what is invisible has no way to justify itself inside the system, even when it works. It is not rejected — it is incomparable, which in practice is worse, because it never even reaches the list where things are decided.
It is the same family of problem as budget silos, described in whoever pays is not whoever saves: there the benefit exists but lands in the wrong pocket; here it exists and lands nowhere, because there is no line to put it on.
And there is a nastier side effect: the people working on the invisible side are left without language. A palliative team, a care home, a service that cares well has no numbers that say so. It ends up defending its work with adjectives, against departments that defend theirs with tables.
What can actually be counted
We have no solution to the big problem, and be wary of anyone who does. We have three useful observations.
1. Counting what was done is possible, and it is more than existed. Every session of any non-pharmacological intervention can be logged: when, with whom, how long, whether it was stopped. After months that is a time series. It does not prove effect. It allows a conversation with data instead of impressions.
2. Patient-experience instruments exist and are underused. PREMs — patient reported experience measures — are the only family of indicators that looks at this side, and most institutions do not apply them. That is an institutional decision, not a technological limitation, and it is probably the most effective thing a management team can do here.
3. Always distinguish activity from outcome. "We ran 340 sessions" is activity. It is not a clinical outcome, and presenting it as one is the mistake that most discredits people working on this side. We develop that in the ROI that cannot be proven.
Where we stop
Being specific, because this is where the industry behaves worst:
- RVer does not measure emotional state. There is no wellbeing score, no comfort index, no mood inferred from anything.
- A supplier selling you a "wellbeing index" computed by the equipment is inventing an indicator — and inventing it precisely because the market needs one. An invented number is worse than none, because it displaces the conversation and cannot be audited.
- We cannot turn comfort into an indicator. Nobody can, for now.
What exists is the record of what the team did and observed, in their words.
In short
- The system measures what is auditable, and has good reasons to.
- Comfort, fear and dignity are not indicators, so they do not compete.
- An intervention purely on the invisible side is not rejected — it is incomparable.
- Counting sessions is not measuring outcome, and confusing the two discredits whoever does it.
- Distrust any wellbeing score produced by a piece of equipment.