How many sessions do you need? Evidence of dose and frequency in therapy

"How long is this going to last?" is probably the most frequent question of a first interview and one of the worst we answer. The honest answer ("depends") is true and at the same time useless for someone who needs to organize their time and budget.

It turns out that the research has enough to say about it. Not a magic figure, but replicated patterns about how the improvement is distributed throughout a treatment and, above all, a pretty strong finding about something that we hardly ever discussed: frequency.

TL;DR

• The ratio between number of sessions and improvement is curvilinear: much progress is made at the beginning and each additional session contributes less and less.
• A systematic review of 26 studies in real services placed the optimal range between 4-26 sessions, according to severity and context: between 4 and 6 in mild to moderate malaise with low intensity interventions, and up to 26 in high discomfort.
• The frequency seems to weigh as much as the total amount: a randomized trial with 200 patients with depression showed that two sessions per week exceeded one, with d = 0.55 at 6 months.
• In the same study, the higher frequency reduced abandonment by half (16 abandoned versus 32).
• There are documented exceptions to the typical curve, including eating disorders and severe psychiatric populations.

Why it Matters

In Argentina there is a strong tradition of long treatments, often without explicit review of objectives. And there is an economic reality in which a large part of patients sustain therapy with effort and decide to follow or not depending on whether they perceive progress.

Understanding the dose-response curve serves two very concrete decisions: when to seriously review a treatment that does not progress, and how to pose the initial framing without promising more or less.

What the evidence says

The Curve

Robinson, Delgadillo and Kellett (2020) systematically reviewed 26 dose-response studies in psychological therapies delivered in real services (not in controlled trials), and published the results in Psychotherapy Research.

The central finding, consistently replicated, is that the relationship between treatment duration and outcome is curvilinear (log-linear or cubic): the improvement is concentrated in the first sessions and then the rhythm is flattened. It is not to stop improving, it is that more and more treatment is needed to achieve the same increase.

On optimal doses, the review found ranges that vary according to context, population, and measurement:

The reported global range ranges from 4 to 26 sessions. and there is an additional fact that often goes unnoticed: weekly therapy accelerates the rate of improvement compared to less frequent schemes.

Frequency

Bruijniks et al. (2020) published in British Journal of Psychiatry a randomized multicenter trial with 200 adults with depression in nine specialized centers in the Netherlands, with factorial design 2 × 2: CCT or interpersonal therapy, and one or two sessions per week, during 16 to 24 weeks, with a maximum of 20 sessions.

That is: same total number of meetings, different frequency.

The results consistently favoured the two weekly sessions:

• Average difference of 3.85 points in BDI-II at 6 months, with d = 0.55.
• Higher response rate (hazard ratio 1,48; CI 95% [1.00; 2,18]).
• Less abandonment: 16 abandoned in the group of two weekly sessions versus 32 in that of one.

That last point connects directly with what we know about why patients leave therapy: the slower the perceived progress, the more likely the treatment will fall.

What does it mean in your clinical practice

• Plan the frame in ranges, not in indefining. "We will work and in about eight sessions we review together how we come" is more honest and more sustainable than "that cannot be known." It does not compromise you to a discharge and gives the patient a horizon.
• I put an explicit review point. If there is no movement at 8 or 10 sessions, that is clinical information, no lack of time. It may indicate that objectives, formulation, framing or referral need to be reviewed.
• Consider increasing frequency in depressive pictures, especially at the beginning. Evidence suggests that two weekly sessions during the first weeks accelerate the improvement without increasing the total. It is a distributional adjustment, not total cost.
• Eye with fortnightly session by default. Much is used as an economic solution, and the evidence suggests that it may slow the process to the point of compromising it. If it is the only viable option for the patient, it is worth knowing that you are working against the curve.
• The curve flattens, does not fall. That the rhythm drops does not mean that sustaining a long treatment is useless: it means that from a certain point it is appropriate to explain what is being sought at that stage. There a written treatment plan prevents the therapy from becoming inertia.

Killings and limitations

• Correlation in naturalistic data. In real service studies, patients who need more sessions are different from those who need less. Part of the curve reflects who stays, not just how much it serves to stay.
• The "optimal dose" depends on the criterion of success. It is not the same to define it by reliable change, by remission or by satisfaction.
• Documented exceptions. The review found that the curvilinear pattern does not sustain the same in eating disorders or in severe psychiatric populations.
• Frequency trial is on depression, with CCT and PTI, in a specialized health system. It is not automatically extrapolated to other tables or other models.
• None of this replaces individual evaluation. They are population averages useful to guide decisions, not to take them for you.

In summary

The evidence does not say how many sessions your patient needs. It says three useful things: that most of the improvement is concentrated at first, that the range where it is usually obtained goes from a few to about two dozen sessions according to severity, and that how you distribute those sessions over time matters as much as how many sessions are.

To be able to see this evolution requires registration. An orderly medical records, with consistent session-to-session notes, is what turns an impression ("I think it's better") into a clinical reading. Brauni takes care of the draft of each note so that reviewing the evolution of a treatment does not depend on your memory or loose papers.

References

• Robinson, L., Delgadillo, J., & Kellett, S. (2020). The dose-response effect in routinely delivered psychological therapies: A systematic review. Psychotherapy Research, 30(1), 79-96. doi.org/10.1080/10503307.2019.1566676
• Bruijniks, S. J. E., Lemmens, L. H. J. M., Hollon, S. D., Peeters, F. P. M. L., Cuijpers, P., Arntz, A., ... Huibers, M. J. H. (2020). The effects of once- versus twice-weekly sessions on psychotherapy outcomes in depressed patients. British Journal of Psychiatry, 216(4), 222-230. doi.org/10.1192/bjp.2019.265