1 PREVENTION OF TREATMENT FAILURE: THE IMPORTANCE OF PROGRESS FEEDBACK & THE THERAPEUTIC ALLIANCE MIDST ALL THE CURRENT EMPHASIS ON TREATMENT GUIDELINES michael J. Lambert, Ph.D. Brigham Young University

The Problem  10-14% of SMI adults and 14 to 25% of child clients deteriorate in psychotherapy  What shall we do about it???? 2

Identifying Cases for Review Little or No Need (50%) Moderate Need (43%) Great Need (7%)

4 Major Issues 1. Development of change sensitive brief measures. 2. Development of expected treatment response and method of predicting treatment failure. 3. Automated method of providing instantaneous feedback to clinicians and patients. 4. Development of Problem-solving tools for failing cases 5. Clinical trails to test effects

Outcome Is:  Symptom Distress—internal pain e.g., I feel hopeless about the future e.g., I feel hopeless about the future  Interpersonal Problems e. g., I feel lonely e. g., I feel lonely  Social Role Functioning e.g., I feel angry enough at work to do something I may regret e.g., I feel angry enough at work to do something I may regret  Well being 5

Substance Abuse Items  After heavy drinking I need a drink the next morning to get going.  I feel annoyed by people who criticize my drinking or drug use.  I have trouble at work/school because of drinking/drug use. 6

Substance Abuse Items (continued)  How many days in the PAST WEEK did you drink any alcohol(beer, wine, or liquor)? 0, 1, 2-3, 4-5, 6-7  How many days in the PAST WEEK did you use any drugs (Marijuana, cocaine, heroin, speed, others)? 0, 1, 2-3, 4-5, 6-7) 7

Measured With  45-item self-report (parent-report) scale taken prior to each treatment session And delivered to clinician in real time—within 2 seconds.  The test provides a mental health “Vital Sign” or “Lab Test” that calibrates current functioning in relation to functioning prior to treatment AND expected treatment response of similar clients WITH ALERTS. 8

ALERTS are Essential Because Clinicians are Overly Optimistic 9

 Final Outcome was predicted for 550 Clients  3 were predicted to have a negative outcome  40 had a negative outcome  Staff identified only one case  Algorithms predicted 85% of those who had a negative outcome but false alarm signals were given at a 2:1 ratio. How Well do Practitioners Predict Treatment Failure?

Hatfield (2010)  Examined case notes of patients who deteriorated to see if therapists noted worsening at the session it occurred.  If the patient got 14 points worse was there any recognition? 21%  If the patient got 30 points worse was there recognition? 32%

14 PDA Administration

15 Clinician Report

Assessment for Signal Cases  My therapist seems glad to see me  At times the tone of my therapist’s voice seems critical  I could count on friendships when something went wrong  I had thoughts of quitting therapy

17

 SIX CLINICAL TRIALS IN WHICH WE ATTEMPTED TO REDUCE DETERIORATION RATES BY PROVIDING PROGRESS FEEDBACK TO PSYCHOTHERAPISTS Lambert, et al Lambert, et al Whipple, et al Hawkins, et al Harmon, et al Slade, et al Crits Christoph et al 2011 Simon, et al. Inpatient Eating Disorders, 2012  Random assignment of patients to experimental condition blocked on therapist (every therapist had patients for whom they received feedback and were denied feedback) N = 4,000 Research Program

Results (Outcome) Recovered or Improved No ChangeDeteriorated NOT-NFb (n = 286) 60 (21%)165 (58%)61 (21%) NOT-Fb (n = 298) 104 (35%)154 (52%)40 (13%) NOT-Fb+CST (n = 239) 121 (51%) 102 (43%) 16 (6%)

Substance Abuse Outcomes Crits-Christoph, et al 2011 (Journal of Substance Abuse Treatment)  Multi-site study—New York, Philadelphia, Salt Lake City  304 patients assigned to feedback or no feedback within therapists  Followed across 12 treatment sessions. 21

22

23

BECOMING AN EMPIRICALLY- VALIDATED PSYCHOTHERAPIST: The value of applying patient- focused research in psychotherapy practice

 Therapist Effects Can be Dramatic  The individual therapist plays a key role in outcome and this role is independent of school-based offerings. Variability of Client Outcome as a Function of the Therapist and What To Do About It.

26 Provider Profile Provider Profile  Providers can profile real time change metrics for themselves—percent of cases recovered, improved, no change & deteriorated—to determine strengths & areas for improvement  Providers can view aggregate cases by diagnosis, gender, etc. for differential effectiveness in case load  Providers contrast differential effectiveness (e.g., diagnosis) with peers

Therapist Variability in Outcome  Since 1996, 270 different professionals and trainees have provided treatment.  The CCC’s database includes nearly 27,009 individual, couple, family, biofeedback, and group sessions.  179,000 OQ-45s have been gathered since  average improvement scores for the 26 current professionals, compared with an average improvement score computed for all previous professional therapists and all past and present trainees.  minimum of 186 clients per therapist to a maximum of 1,054 clients per therapist;

 On average, every therapist’s clients improved.  The average of improvement scores for all past professionals and all trainees is depicted in black.  Average improvement scores for 7 therapists (in blue), are significantly better than this overall average.

 4 therapists (in red), had scores significantly worse than this overall average.  15 therapists (in green), had scores within this average range.  Therapist 1's average improvement score is particularly interesting, as it is significantly better than 24 of the 25 other average improvement scores. Nielsen & Okiishi, 2010

32 In sum…  Ongoing monitoring & feedback: Increases overall outcomes Increases overall outcomes Reduces treatment failures Reduces treatment failures Improves outcomes for substance abuse clients Improves outcomes for substance abuse clients Increases service access by reallocation of staff time Increases service access by reallocation of staff time Identifies best practice groups/clinicians and those in need of peer-supervision Identifies best practice groups/clinicians and those in need of peer-supervision Saves support staff time when using a fully automated system. Saves support staff time when using a fully automated system.