2012 National Meeting

1105 — Measuring Veteran Recovery at Mental Health Clinics: The Mental Health Recovery Measure (MHRM)

Armstrong NP, Greater Los Angeles VA; Cohen AN, Greater Los Angeles VA & UCLA; Hellemann G, UCLA; Niv N, Long Beach VA; Hamilton AB, and Young AS, Greater Los Angeles VA & UCLA;

Objectives:
The Uniform Mental Health Services Package requires a transformation to recovery-oriented care in VA mental health. To evaluate this organizational change, VA would benefit from measures of recovery in patients receiving specialty mental health care. Although promising instruments exist, most are burdensome and have multiple domains that have not been established as separate constructs. This paper evaluates the psychometric properties of the Mental Health Recovery Measure (MHRM) in a large sample of VA patients and suggests improvements.

Methods:
As the largest quality improvement study in VA specialty mental health, “Enhancing QUality of care In Psychosis (EQUIP)” provided an opportunity to collect recovery data. EQUIP is a VA HSR&D/QUERI trial that enrolled 801 Veterans with schizophrenia across 8 sites and 4 VISNs. At baseline, Veterans completed the MHRM and measures of functioning, quality of life, and healthcare satisfaction. Reliability and validity of the MHRM were examined through internal consistency, factor analysis, and Pearson correlation procedures.

Results:
Despite eight conceptual domains suggested by MHRM authors, factor analysis revealed that a single factor model best fit the data. Examination of inter-item correlations and item correlations with Lehman Quality of Life and MIRECC-version GAF (occupational and social functioning) scores were used to reduce the number of MHRM items from 30 to 10 (MHRM-10). Total mean score for the MHRM-10 was 27.6 (SD = 6.7; range = 0-40). MHRM-10 internal consistency was comparable to that of the 30-item MHRM (alpha = 0.9). MHRM-10 total scores significantly correlated with measures of occupational functioning (r = 0.16, p <0.01), quality of life (r = 0.36, p <0.01), and satisfaction with care (Client Satisfaction Questionnaire: r = 0.14, p <0.01).

Implications:
The MHRM is a reliable and valid measure that is best interpreted as a total score. A 10-item version of the MHRM maintained comparable psychometric properties, reducing measurement burden. Replication of findings in a different sample and establishing sensitivity to change over time is suggested.

Impacts:
By evaluating recovery outcomes in patients, VA can set benchmarks for and measure progress towards recovery transformation.