We know that people with mental health issues can benefit from psychotherapy when the content of sessions draws from empirical research. However, recent evidence suggests that how frequently therapists and clients meet, and how long their sessions last may be just as important to outcomes as what happens during these encounters.

New findings cast doubt on long held traditions about session frequency and duration, and suggest that not all clients are the same when it comes to what works – and what doesn’t. It may be a good idea to consider new ways that therapy can be delivered not just because it can save money, but because it can be more effective.

Like most therapists across North America, I struggle with the fact that there is simply not enough time in the day to fit in all the sessions with clients that are required. No matter how I adjust my schedule, and despite repeated efforts with my colleagues to streamline our intake process and adhere to proven methods, we just can’t seem to get our wait lists down to reasonable levels. Early in my career, this was not a problem, since I was concerned with filling my schedule, not opening it up. But most good private practitioners and just about anyone in the public service will eventually reach a point where demand exceeds supply. Instead of talking about the methods we are using to treat people, team meetings gradually get consumed by concerns about how to deal with urgent referrals, how to fit more sessions into each week of work, what to do about pressure from benefit providers and employers to have us offer cost-effective solutions, and how to decide when treatment gains are good enough.

Part of the problem stems from some long held assumptions that therapists don’t even think about. For example, most of us adhere to the “50 minute hour” model that offers clients a solid chunk of time to talk about important issues and leaves several minutes for the therapist to document what occurred before the next session begins. Ideally, these sessions are held at regular intervals. A weekly or twice monthly frequency is most common. For clients who can afford the luxury, or perhaps in the event of a crisis, twice weekly sessions are not uncommon. At the other end of the spectrum, when therapists are busy and schedules get blocked in, the frequency of sessions can slow to once every three or four weeks, and sometimes even longer. Behind all of this lie a key set of beliefs that can be summarized as follows: In an ideal world, most therapists would have an hour for every session, meet with each client every week or two, and continue with sessions for as long as it took to get symptoms down to “normal” levels.

During my career as a consulting psychologist, I’ve worked in several different settings. These have included a busy private practice, several probation and parole offices, and a number of family health teams. I have also been active in a networking group where we take turns presenting on a variety of mental health topics. Not long ago I got into a discussion with some colleagues about how we all tended to follow the assumptions about session frequency and duration that I’ve mentioned above. Everyone seemed to be struggling with lengthy wait lists, busy schedules, and a growing sense that we needed to look more closely at our methods. Was it possible that the data in our client records might clarify what was going on?

To answer this question, we performed a quality assurance review of close to 120 treatment records divided among three clinicians. The available data included the mental health diagnosis, whether the client was taking any medication, how frequently they were being seen, and how long treatment lasted. Along with this information, we looked at session-by-session records of symptom severity for a total of over 800 treatment encounters.

Not Your Typical Drug Study

The results of this quality review proved to be surprising. For example, unlike the type of client in research studies that typically suffer from just one discrete problem, such as major depressive disorder, nearly all of the clients described a mixture of both anxiety and depression, and many had additional diagnoses, like opiate addiction or alcohol dependence. We also discovered that most were already taking some form of antidepressant medication to treat their symptoms when they were referred. This meant that the therapists were seeing clients with more complex issues than those who populate most treatment research studies, not to mention that they were trying to treat people who were probably not getting an ideal response from medication. In other words, the problems facing the typical client they see, and that probably thousands of other therapists work with across the nation, are much more complex than the ones that research studies base their findings on when looking at things like treatment effectiveness.

Initial Symptom Level Matters

Based on this quality review we were pleased to learn that despite how serious their problems were, on the whole, most of the clients were improving significantly with respect to their life satisfaction and key symptoms of depression and anxiety. Even when we controlled for the passage of time, which in itself can bring about healing, therapy proved to be having some beneficial effects. However, there were a number of surprises. For example, we discovered that the initial symptom levels that clients described at the start of therapy were very relevant to outcomes. People who started therapy with more severe symptoms and lower ratings of life satisfaction in comparison to other clients tended to end treatment with lower life satisfaction and higher symptom distress. But they also showed the same degree of improvement. In other words, people who start off worse finish worse, but improve just as much.

Different People Show Different Rates of Change

We also learned that regardless of how severe a person’s symptoms were at the start of therapy, different people experienced different rates of change. Some of the clients improved quite dramatically from one session to the next, whereas others seemed to improve more slowly. Furthermore, just as preliminary studies looking at the Myndplan assessment inventories had revealed, some very specific client characteristics predicted this rate of change. When taken together with our finding of symptom severity, this means that with the right tools it should be possible to identify not only how quickly a client is likely to respond to therapy, but also how many sessions they will require in order to achieve the same relative improvement.

Some Clients are More Variable Than Others

When we looked at clients’ mood monitor scores more closely, we also found some interesting things. Most importantly, while many clients demonstrated slower or faster rates of improvement over time, a subset of people showed unusual variability in their scores across sessions. It seems that some people are far more changeable with respect to their day-to-day mood states, perhaps because they are in more stressful or more desperate circumstances, or perhaps because they are more sensitive to the effects of everyday stress than the average person. Regardless, it seems that more of this variability, particularly in the case of anxiety, corresponds to both the severity of symptoms at the outset of treatment and the rate of response over time. It seems likely that this type of client will be far more likely to do better with more frequent sessions than if they are left with large blocks of time between sessions.

Fewer is Better

A final insight proved remarkably counter-intuitive. This involved the results when we looked more closely at the number of sessions clients were attending over the course of therapy. Contrary to our expectation that more is better, we were surprised to discover that clients who had more sessions showed less improvement, and in some cases, their condition even deteriorated! While you might be tempted to point out that this was probably a function of more severe symptoms or reflected a slower response to therapy, neither of these explanations seemed to account for the difference. For example, when we looked at first session symptom monitor scores they were unrelated to the duration of therapy. We also found no significant relationship between the passage of time and outcomes. Most of the difference in outcomes seems to be based on the actual number of sessions. This suggests that grouping fewer therapy sessions close together in time may be just as good, or even better, than more sessions spread out over time.

What Does Current Research Say About This?

These findings are consistent with some of the more recently published research on the topic of session frequency and duration. For example, Erekson et al 1 looked at the relationship between session frequency and psychotherapy outcomes in over 21,000 clients who were treated at university counseling centers and found clinically significant gains were achieved faster for weekly sessions than for twice monthly sessions. Cuijpers et al looked at 70 randomized treatment trials that included data on session frequency and duration and found that neither the total number of sessions nor the number of weeks in therapy had a significant impact on outcomes. Furthermore, the length of sessions was found to have no impact. What did make a difference? It turns out that the only significant effect that emerged in this study was for session frequency 2.

Another study by Storch et al 3 revealed that daily treatment sessions for 14 days were as effective as weekly treatment sessions for 14 weeks, with no difference observed on follow-up. These findings and the results of our own quality assurance work are consistent with Bohni et al’s proposal that exposure is an essential piece of any effective psychotherapy method 4. Whether we are trying to get someone with phobic anxiety or a panic disorder to face a feared situation or object, convince someone with depression to ignore negative thoughts or get out socially, or persuade someone with anger issues to keep calm and carry on, most of our work encourages people to practice behavior change every day that puts them in uncomfortable situations and encourages them to ride it out. As most undergraduates learn in their first introductory psychology course, more frequent exposure is almost always better than the exposure that is spread out over time when a client is trying to change their behavior, learn new behaviors, or maintain treatment gains.

Why is this important?

Psychotherapy is meant to help alleviate the suffering associated with mental health problems by reducing symptom severity, increasing adaptive behaviors, and improving quality of life. In addition to using the right methods, it appears that something as simple as the frequency and duration of treatment sessions can significantly impact these goals. Severity scores at the outset of treatment may be more useful for determining realistic goals than for estimating how long a person will be in therapy. Session length may be less important than most therapists – and clients – would like to believe. The results may seem counterintuitive, but in the end, the best outcomes may come from fewer sessions, shorter sessions, and more frequent sessions. This may defy tradition, but the evidence seems to be mounting that how we deliver treatment, not just what we deliver, may reduce suffering in addition to saving time and money.

Equally important here is the fact that if therapists and clients are not measuring symptom levels or keeping track of progress, it will be more difficult for them to judge how difficult the situation is in the first place, if progress is within normal limits, whether there are any characteristics that suggest a change in tactics is required, and when realistic goals have been achieved. As we start to phase in active symptom monitoring for clients using the Myndplan Beta prototype, we are hoping to address these issues and many more that are relevant to both providers and consumers. Stay tuned!