Does Therapist Experience Improve Therapy Outcomes?

Tracking Client Outcomes as
Therapists Mature from Trainee to
What does the therapist contribute to outcome?
 The therapist contribution
Personal qualities
 As practitioners, we would like to believe that we not only
influence outcomes, but that the outcomes of our clients
improve with experience or training
 Evidence has been hard to come by
Mixed results in the literature with a few studies showing a positive
relationship between experience and outcome but most metaanalyses suggesting little support for the belief that experienced
clinicians have better outcomes
Smith & Glass, 1977; Shapiro & Shapiro, 1982; Stein & Lambert, 1995
Difficulties inherent in studying the therapist
 Methodological issues:
 Few studies have been designed to explicitly investigate
therapist experience. This is looked at post-hoc in studies
designed to identify if therapy works.
 Designs are cross-sectional, comparing different groups of
therapists. In such instances, it is hard to disentangle
therapist experience from individual differences between
therapists on a variety of attributes (e.g. personality,
theoretical orientation)
 Experience level is typically classified by degree, which may
ignore variance attributed to years in practice.
Masters level vs PhD level therapists
How this study improves on past studies
 Experience was classified by years trained rather than
comparing different degrees
 All therapists trained in Ph.D. programs in psychology
Experience levels could be compared more easily
Similar training time line/pattern of skill acquisition over time
Training was broken down into four different levels or status
 The same therapists were tracked over time, providing a
longitudinal/within-subjects design.
Decrease in error variance
 Therapists had many years experience compared to most
studies (minimum of 8, maximum of 11)
 Therapists in the study had seen many clients (minimum
of 72)
 Participants
 14 Therapists
All therapists provided services at the CCC at some point during their graduate
training as well as post licensure.
5 females, 9 males
All Ph.D.s: 5 Clinical Psych, 9 Counseling Psych
Range of years of total experience = 8-11 years
Minimum number of clients tracked was 72, maximum was 1015, mean was
383.29, (SD=269.29); median=348.5
Client participants included all students/patients who agreed to participate in
archival outcome research and were seen by one of the participating therapists
during the years 1996-2011
 Database
 Since 1996, 304 different professionals and trainees have provided treatment.
The CCC's database includes approximately 273,000 appointments, with
progress notes; over 181,000 individual therapy appointments; over 184,000
OQ-45s have been gathered since 1996.
Measure of Outcome: Outcome Questionnaire-45
 45 items across 4 domains
 Yields a total score reflecting amount of distress a
person has been experiencing during the last 7 days
 Scores range from 0 to 180
 Higher scores signify more distress
 64 represents the cutoff between dysfunctional &
functional populations
 There is a significant effect of training status on OQ
change at the p<.05 level. F(3,5362)=3.73, p=0.011
 Post-hoc Tukey HSD tests showed that licensed
professionals OQ change scores are 2.18 points smaller
than practicum students OQ change scores at the p<.05
level. No other comparisons were significant.
 To investigate this data more thoroughly, we need to
account for clustering of clients within therapists. To
account for the clustering, doing a repeated measures
ANOVA or HLM would be preferable.
 There was no significant difference in pretest OQ scores
between training status.
Mean Change, Across Experience Levels, For
Clients Seen Only in Individual Therapy
Individual Therapist Patterns
Individual Therapist Patterns
Individual Therapist Patterns
Individual Therapist Patterns
Individual Therapist Patterns
Interesting Findings
 Therapist Experience was NOT associated with better client
As years of therapist experience increased, mean change in client
outcome scores trended downward.
Licensed professionals change scores were significantly lower than
practicum students.
 There was ambiguous support for small differences in OQ
scores at intake across training status, with licensed
professional averaging higher intake scores (seeing more
distressed clients).
 What’s going on???
Patterns in treatment in our center:
Most experienced therapists have increased client load, fewer opportunities
for consultation, and they see clients less frequently
The mean level of client distress was higher for the Professional therapists
 Difficulty measuring/operationalizing experience
Training status may not effectively measure ‘experience’
Experience may be better captured by looking at total number of sessions provided,
familiarity or specialty with a specific type of intervention or patient problem
 Difficult to eliminate factors, other than experience, that influence client
treatment (e.g. referrals to group, medication, biofeedback, frequency of
visits, therapist consultation)
 Problems inherent in ‘real world’ research
This study investigated patterns of outcome in a counseling center.
In contrast to controlled studies, the therapy followed many varied courses.
For instance, in this study, different therapists provided different numbers of sessions at each
level of training.
Length of therapy and gaps in sessions varied
Patients changed providers from time to time
 There were only 14 therapists
 Need to account for clustering of the clients within therapists, perhaps by
using hierarchical linear modeling.

similar documents