The Evidence Base for Psychoanalytic Therapies: It Just Might Be Better Than You Think


Sunset 2019In recent days there’s been a bit of a kerfuffle on Twitter regarding the relative efficacy of psychoanalytic and cognitive-behavioral therapies (CBT). Of course, the standard mantra in the media and among many mental health professionals is that the science shows that CBT is superior and the treatment of choice for many, if not most, mental and emotional problems. Well, as is often the case in life and psychotherapy, reality is much less clear.

This post isn’t about fake news or alternative facts. Instead, I hope it’s about a balanced perspective. As a psychotherapist-counselor-professor-clinical psychologist, I like to think I don’t have an allegiance to any single therapy approach. Although I know I can’t claim perfect objectivity, I do have a broad view. One factor that has helped me have a broad view is that I read lots of professional journal articles in order to be able to write my theories of counseling and psychotherapy textbook.

Below, I’ve inserted an excerpt from the end of the psychoanalytic chapter of our textbook. Whether you’re a CBT or psychoanalytic fan, or perhaps a fan of a different approach, I hope you find this short review of psychoanalytic treatment efficacy interesting. The bottom line for me is captured by an old quotation from Freud (who wasn’t known for his flexible thinking). Purportedly, he said, “There are many ways and means of conducting psychotherapy. All that lead to recovery are good.” I might add the following to Freud’s comment: There are many different clients with many different problems and many different individual and cultural perspectives. I’m convinced that most clients are best served if therapists tweak their approaches to fit the client, rather than expecting the client to fit into narrow clinical procedures based on pure (or rigid) theoretical perspectives.

Here’s the excerpt . . .

***********************************

Conducting rigorous research on longer-term treatments, such as psychoanalytic therapy, is challenging and cost prohibitive. Psychoanalytic approaches are often less symptom- or diagnosis-focused, seeking instead to facilitate client insight and improve interpersonal relationships. Because empirically supported treatments focus on whether a specific psychological procedure reduces symptoms associated with a medical diagnosis, “proving” the efficacy of complex therapy approaches is difficult—especially when compared to the lesser challenges inherent in evaluating symptom-focused treatments. Partly because of these complexities, some reviewers contend that psychoanalytic psychotherapies are less efficacious than cognitive and behavioral therapies (Busch, 2015; Tolin, 2010).

The good news for psychoanalytic therapy fans is that evidence is accumulating to support treatment efficacy. The less good news is that some of the research support remains methodologically weak and the wide variety of psychoanalytic approaches makes it difficult to come to clear conclusions. Nevertheless, the most recent meta-analytic studies, literature reviews, and individual randomized controlled studies support the efficacy of psychoanalytically oriented therapies for the treatment of a variety of mental disorders. According to Leichsenring, Klein, and Salzer (2014), there is empirical support for the efficacy of psychoanalytic psychotherapies in treating:

  • Depressive disorders.
  • Anxiety disorders.
  • Somatic symptom disorders.
  • Eating disorders.
  • Substance-related disorders.
  • Borderline personality disorder.

The evidence for the efficacy of psychodynamic approaches for depressive disorders is strong. In a recent meta-analysis, Driessen and colleagues (2015) evaluated 54 studies, including 3,946 patients. They reported that short-term psychodynamic psychotherapy (STPP) was associated with improvements in general psychopathology and quality of life measures (d = 0.49–0.69) and all outcome measures (d = 0.57–1.18); they also noted that patients continued to improve at follow-up (d = 0.20–1.04). Further, no differences were found between STPP and other psychotherapies. On anxiety measures, STPP appeared significantly superior to other psychotherapies at post-treatment (d = 0.35) and follow-up (d = 0.76).

In a previous meta-analytic review, Shedler (2010) also concluded that psychodynamic therapies were equivalent to “. . . other treatments that have been actively promoted as ‘empirically supported’ and ‘evidence based’” (p. 107). He also reported that psychodynamic therapies had more robust long-term effects.

Table 2.2 provides a sampling of meta-analytic evidence supporting psychodynamic therapies. For comparison purposes, the original meta-analyses conducted by Smith and colleagues are included (Smith & Glass, 1977; Smith et al., 1980). Notably, Smith, Glass, and Miller reported that psychodynamic approaches were significantly more efficacious than no treatment and approximately equivalent to other therapy approaches.

Table 2.2 also includes the average effect size (ES or d; see Chapter 1) for antidepressant medications (ES = 0.31 for serotonin-specific reuptake inhibitors or SSRIs). This comparison data shows that psychodynamic psychotherapy is more effective than SSRI treatment for depression. Additionally, the benefits of psychoanalytic therapy tend to increase over time (Driessen et al., 2015; Shedler, 2010). This implies that psychoanalytic psychotherapy clients develop insights and acquire skills that continue to improve their functioning into the future—which is clearly not the case for antidepressant medication treatment (Whitaker, 2010). One of the ways psychotherapists explain this difference in longer term efficacy is with the statement: “A pill is not a skill.”

Table 2.2 A Sampling of Psychodynamic Psychotherapy Meta-analyses

Authors Outcome focus Number of studies ES or d
Abbass et al. (2009) General psychiatric symptoms 8 0.6
Anderson & Lambert (1995) Various 9 0.85
de Maat et al. (2009) Long-term treatment 10 0.78
Driessen et al. (2015) Depression 54 0.57–1.18
Comparison research
Turner et al. (2008) Meds for Major depression 74 0.31
Smith et al. (1977) Different therapies 375 0.68
  Many problems    
Smith et al. (1980) Different therapies 475 0.75
  Many problems    

Note: This is a sampling of meta-analytic psychoanalytic psychotherapy reviews. We’ve omitted several reviews with very high effect sizes partly because of criticisms related to their statistical methodology (see Driessen et al., 2015, and Shedler, 2010, for more complete reviews). This table is not comprehensive; it’s only a reasonable representation of psychoanalytic psychotherapy meta-analyses.

We recommend you take the preceding research findings (and Table 2.2) with a grain of salt. Conducting systematic research on something as subjective as human mental and emotional problems always includes error. One source of error is the allegiance effect (Luborsky et al., 1999). The allegiance effect is the empirically supported tendency for the researcher’s therapy preference or allegiance to significantly predict outcome study results.  Luborsky and colleagues (1999) analyzed results from 29 different adult psychotherapy studies and reported that about two thirds of the variation in outcome was accounted for by the researcher’s theoretical orientation (e.g., psychoanalytic researchers reported more positive outcomes for psychoanalytic therapy and behavior therapists discovered that behavior therapy was more effective).

The implications of the allegiance effect help explain why, shortly after Shedler’s (2010) publication extolling the virtues of psychodynamic psychotherapy, several critiques and rebuttals were published (Anestis, Anestis, & Lilienfeld, 2011; McKay, 2011). The critics claimed that Shedler’s review was biased and accused him of overlooking weaknesses within the meta-analyses he reviewed (e.g., poor outcome measures, pooling the effects of small samples with little power and poor designs, lack of treatment integrity effects). Although Shedler’s critics raised important points, the critics themselves had their own biases. The problem is that all researchers (and writers) have an allegiance of one sort of another.

One of our favorite ways of understanding the allegiance effect is articulated in a story about the great New York Yankee baseball player, Yogi Berra. One day, when a player on Yogi’s team was called out on a close play at second base, Yogi went charging on to the field to protest. The umpire explained that he, unlike Yogi, was an objective observer and that he, unlike Yogi, had been only about 5 feet from the play, while Yogi had been over 100 feet away, in the dugout. When Yogi heard the umpire’s logic, he became even angrier and snapped back, “Listen ump, I wouldn’t have seen it, if I hadn’t believed it” (adapted from Leber, 1991).

The “I saw it because I believed it” phenomenon is also called confirmation bias (Masnick & Zimmerman, 2009; Nickerson, 1998). Confirmation bias involves seeking, interpreting, and valuing evidence that supports pre-existing beliefs, while ignoring and devaluing evidence contrary to preexisting beliefs. Consequently, psychoanalytically oriented individuals see support for their perspective and behavior therapists see support for theirs. However, despite these caveats, based on accumulating research, psychodynamic approaches have a reasonably good record of efficacy.

***********************************

Although this particular review has many limitations, I’m convinced that most of us, most of the time, are better off following the advice of Marvin Goldrfried (and others) and focusing on the common therapeutic factors, or, as Norcross calls a subset of common factors, empirically-supported relationships.

For more information, check out Goldfried’s recent article on obtaining consensus in psychotherapy:  https://www.stonybrook.edu/commcms/psychology/_pdfs/clinical/Goldfried%20AP%20Consensus%20AP.pdf

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s