From: Elliott,R.,Watson,J.,Timulak,L., &Sharbanee, J. (in press). Research on Humanistic-Experiential Psychotherapies.In M. Barkham, W. Lutz, & L Castonguay (eds.). Garfield & Bergin’s Handbook of Psychotherapy & Behavior Change (7th ed.). New York: Wiley.
Current mental health politics urgently require continuing collection, integration, and dissemination of information about the rapidly expanding body of outcome evidence, to help deal with challenges to HEPs in many countries. HEP outcome research has grown rapidly, with a fifty per cent increase in the past 10 years. This has allowed us to pursue increasingly sophisticated analytic strategies and to break down the evidence by client subpopulation and type of HEP. We believe that these analyses go a long way toward meeting the demands of the various national guideline development groups (e.g., APA Division 12 Task Force on Empirically Supported Treatments in the United States; National Institute for Clinical Excellence [NICE] in the United Kingdom).
Looking at our current data set of 91 recent studies, together with our previous collection of nearly 200 outcome studies, we see that evidence for the effectiveness of HEPs comes from five separate lines of evidence and supports the following conclusions:
First, overall, HEPs are associated with large pre-postclient change. These client changes are maintained over the early posttherapy period (< 12 months), although not enough recent studies have addressed late (a year or more) outcomes.
Second, in controlled studies, clients in HEPs generally show large gains relative to clients who receive no therapy, regardless of whether studies are randomized or not. This allows the causal inference that HEP, in general, causes client change; or rather, speaking from the client’s perspective, we can say that clients use HEP to cause themselves to change.
Third, in comparative outcomestudies, HEPs overall are statistically and clinically equivalent in effectiveness to other therapies (especially non-CBT therapies), regardless of whether studies are randomized or not.
Fourth, in the current dataset, CBT appears to have a small advantage over HEPs. However, this effect seems to be due in part to non–bona fide treatments usually labeled by researchers as supportive(or sometimes nondirective), which are generally less effective than CBT. These therapies are typically delivered when there is a negative researcher allegiance and in non–bona fide versions, and appear to be the mediator for the substantial researcher allegiance effect that we have found repeatedly. In our previous review, when the supportive treatments were removed from the sample, or when researcher allegiance was controlled for statistically, HEPs appeared to be equivalent to CBT in their effectiveness. However, levels of researcher allegiance in the current sample were so high that it proved difficult to control for them statistically, leading to an equivocal finding in favor of CBTover person-centered therapy, in contrast to the clear equivalence finding we reported in our previous review.
Fifth, in terms of type of HEP, EFT continues to fare the best, although the number of recent controlled and comparative studies is too small towarrant a strong conclusion. There are plenty of studies of supportive-nondirective therapy, a weaker form of HEP, and these continue to do most poorly against CBT. However, it is not clear how much this is due to negative researcher allegiance effects and how much is due to this approach being less effective. In terms of effectiveness, Person-centered therapy falls in between supportive-nondirective therapies and EFT, a consistent finding across both our previous and current meta-analyses.
Going beyond these general conclusions, we have argued that there is now enough research to warrant varying positive valuations of HEP in six important client populations: depression, relationship/interpersonal problems, anxiety, coping with chronic medical conditions, psychosis, and self-damaging activities.
For depression, HEPs continue to be extensively researched, with large pre-post effects and medium controlled effects; for this dataset we found an equivocally negative comparative effect size, characterized by overwhelmingly negative researcher allegiance. This result is suspect because it contradicts our previous clear equivalence finding that supported the use of HEPs for depression generally, but particularly for EFT for mild to moderate depression (e.g., Goldman et al., 2006; Watson et al., 2003), and PCT for perinatal depression (e.g., Cooper et al., 2003; Holden et al., 1989).However, recent large-scale balanced allegiance studies (Barkham & Saxon, 2018; Barkham et al., 2020) suggest that there are some specific situations in which CBT may do better that PCT: for example, with more severely distressed clients seen for more sessions or tracked a year later. This in turn points to the possibility that when applied to depression PCT may need to bolstered with more powerful methods, e.g., EFT chair work.
For relationship and interpersonal problems HEPs clearly meet criteria as an efficacious treatment, based on pre-post and controlled effects. Our previous review contained a substantial numberof studiesof couples therapy, with large pre-post, controlled and even comparative effects. However, the current review was dominated by individual therapy and included more studies of social anxiety and PTSD, which resulted in smaller comparative effects in the equivalent or trivially less effective range.
For helping clients cope psychologically with chronic medical conditions, we found a large body of studies and replicated our previous finding of reasonably large pre-post effects, clear superiority to no treatment control conditions, and equivalence to other treatments including CBT. This client population continues to be a promising one for further exploration of the value HEPs including supportive-nondirective and PCT.
For habitual self-damaging activities, including eating difficulties, our analysis points to the effectiveness of HEPs in general (primarily supportive and other HEPs). Although time and resources precluded including Motivational Interviewing (MI) here, our results are again comparable to those commonly reported for MI (Lundahl et al., 2010).
For anxiety difficulties overall, the recent evidence is as in the previous review mixed, but sufficient to warrant a general continuing verdict of possibly efficacious: We found large pre-post and very large controlled effects, but general superiority of CBT to supportive-nondirective treatments in negative researcher allegiance studies. There are at least two major comparative treatment studies (e.g., Timulak et al. 2018) pitting EFT against CBT currently in progress, but none of these have yet reported results.
For psychotic conditions such as schizophrenia, we replicated our previous finding of promising pre-post and comparative effects supporting the use of HEPS for this challenging client population. This directly contradicts the UK guideline contraindicating humanistic counseling for clients with this condition (National Collaborating Centre for Mental Health, 2010). In fact, the comparative evidence we have reviewed points to the possibility that HEPs may in some cases be moreeffective than the other therapies to which they have been compared. Clearly, this is an area that warrants further investigation and treatment development.