Ask the patient for feedback: what the evidence about FIT says

There is a practice that sounds so sensible that it is difficult to discuss it: ask the patient, session-to-session, how he is doing and how he is feeling the work with you. Two scales of four items, two minutes, and a brief conversation about what the scale shows. That is, in essence, the feedback-reported treatment (FIT) and its most widespread system, the PCOMS (Partners for Change Outcome Management System), with its two instruments: the ORS (Outcome Rating Scale, how next week) and the SRS (Session Rating Scale, how the session comes).

The promise is attractive: better results, less abandonment and less deterioration, without changing the theoretical model. The honest question is how much evidence that promise holds. And the answer is more nuanced than is usually told in training courses.

TL;DR

• The most complete meta-analysis of the PCOMS found a small effect on general symptoms (g = 0.27), but that effect disappears in psychiatric contexts (g = 0,10, not significant).
• Another independent meta-analysis, with outcome measures outside the system itself, found no effect on well-being (g = 0,03) or on the number of sessions (g = 0,13).
• Much of the positive effect appears when the outcome measure is the same scale of the intervention (the ORS) and when the authors have a link to the system.
• The evidence about the feedback of progress in general (not only PCOMS) is more solid than that of PCOMS in particular: there are consistent improvements in results and abandonment.
• Practical conclusion: asking feedback is still a good clinic, but it should be done because of what it contributes to the conversation, not because it relies on a large effect that the evidence does not support.

Why it Matters

In Argentina the FIT circulates mainly through formations and workshops, almost always presented as a practice "based on evidence" without further clarification. And there is something true there: the underlying logic (measure, return, adjust) is the same of the monitoring of results, which does have support.

But "logic has support" is not the same as "this specific system has support." And the difference matters when you're deciding if you're going to spend five minutes of each session, with each patient, for years.

What the evidence says

The favorable meta-analysis (and its asterisks)

Østergård, Randa and Hougaard (2020) reviewed 18 studies (14 randomised) with 2.910 participants. They found a small but significant effect of the PCOMS on general symptoms: g = 0.27.

The detail is in the moderators when they separated by context:

• On counseling: g = 0,45 devices (moderate effect).
• On psychiatric: devices g = 0,10, without statistical significance.

The authors themselves point out why the large number should be read carefully: almost all studies of the favorable group had the researcher's alliance with the system (the authors were promoters of PCOMS) and used the ORS as the only result measure. That is, the scale that is part of the intervention was also the rod with which success was measured.

Unfavorable meta-analysis

That same year, Pejtersen, Viinholt and Hansen (2020) published in Journal of Counseling Psychology a systematic review of 14 randomized trials, designed specifically to look at independent measures.

The results:

• Quantity of sessions: g = 0,13 (IC 95% [0.001, 0.26]), which amounts to a difference of less than one session.
• Patient welfare, measured with scales outside the PCOMS: g = 0,03 (IC 95% [-0.18, 0.23]), i.e. nothing.

The authors' conclusion is direct: they found no evidence that PCOMS would improve the well-being of patients or the number of sessions they attend.

The replica

Duncan and Sparks (2023), central figures of the development of the PCOMS, responded publicly questioning the inclusion of studies they consider confused and defending the validity of the ORS against independent measures. The discussion remains open and worth knowing: it is a legitimate methodological debate, not a closed case in either direction.

The broader context

Here comes the nuance that changes the reading. The meta-level analysis of Jong et al. (2021), on overall progress feedback (several systems, not just PCOMS), did find consistent effects: better results, less abandonment and less deterioration, with the greatest benefit in cases that were bad ("not-on-track").

That is: the practice of measuring and returning has backup; the specific brand is not what makes the difference.

What does it mean in your clinical practice

• I asked for feedback, but for the right reason. The main value of asking "how did you feel this session?, was there anything that didn't close you?" isn't statistical, it's relational. It opens a conversation about the link that many patients don't start alone, and that connects directly with what is well established about the therapeutic alliance.
• Don't delegate clinical judgment to a number. An ORS from 26 doesn't tell you what to do. It tells you it's worth asking.
• If you are going to measure result, consider an independent measure. A validated scale of the picture you are dealing with (PHQ-9, GAD-7, BDI) gives you less self-referential information than a scale that is part of the feedback ritual itself.
• Pay special attention to cases that do not improve. It is where the feedback shows its highest performance and where the treatment abandonment is most played.
• Eye with Social Desire. Many patients score high on SRS to avoid discomfort. If all your SRSs are close to maximum, the instrument is probably not capturing anything: there the useful data is that uniformity, not the score.

Killings and limitations

• Small effects are not null effects. A g of 0,27 at population level may be clinically relevant if the cost of the intervention is low, and here the cost is a few minutes per session.
• The quality of the studies is uneven. Much of the favorable evidence comes from team with an interest in the system, and much of the unfavorable evidence comes from studies in contexts very different from Argentina's private clinical practice.
• How it is implemented matters more than if it is implemented. Applying the scale without talking about the result makes feedback an administrative procedure. In studies where it works, return is a real clinical conversation.
• Almost none of this evidence is Latin American. The cultural transfer of short self-reporting instruments is not automatic.

In summary

The informed feedback is not the revolution that is sometimes sold, but it is not smoke either. The honest reading of the evidence is this: to systematically ask the patient how it is doing and how we are going is a reasonable, low cost and probably beneficial practice, especially when things are not working. What is not supported is the idea that a one-time system, with two scales of four items, improves outcomes substantially by itself.

None of this replaces your clinical reading: the scale opens the door, the conversation is yours. And if what you're looking for is to be able to look at a patient's evolution without relying on memory or loose sheets, having the clinical record sorted and queryable is the basis on which any feedback system works. That's exactly what it does Brauni with your session notes.

References

• Østergård, O. K., Randa, H., & Hougaard, E. (2020). The effect of using the Partners for Change Outcome Management System as feedback tool in psychotherapy: A systematic review and meta-analysis. Psychotherapy Research, 30(2), 195-212. doi.org/10.1080/10503307.2018.1517949
• Pejtersen, J. H., Viinholt, B. C. A., & Hansen, H. (2020). Feedback-informed treatment: A systematic review and meta-analysis of the Partners for Change Outcome Management System. Journal of Counseling Psychology, 67(6), 723-735. doi.org/10.1037/cou0000420
• Duncan, B. L., & Sparks, J. A. (2023). When meta-analysis continues to mislead: A reply to Østergård and Hougaard (2020). Psychological Services, 20(Suppl 2), 238-240. doi.org/10.1037/ser0000666
• de Jong, K., Conijn, J. M., Gallagher, R. A. V., Reshetnikova, A. S., Heij, M., & Lutz, M. C. (2021). Using progress feedback to improve outcomes and reduce drop-out, treatment duration, and deterioration: A multilevel meta-analysis. Clinical Psychology Review, 85, 102002. doi.org/10.1016/j.cpr.2021.102002