Foundations for Polaris-MH Development: Dosage, Phase and ETR Models

Polaris-MH assesses (a) a patient’s subjective well-being, (b) the severity of patient symptoms associated with the most common disorders treated in outpatient settings, (c) the impact of the patient’s psychological problems upon the patient’s life (functional disability), and (d) therapeutic bond/satisfaction with treatment. These domains are grounded in extensive research on mental health outcomes. They are readily accepted by clinicians, operating from all major therapeutic models, as being central to clinical decisions and outcomes assessment. They provide the framework for constructive dialogue between clinicians and utilization review staff.

Dosage Model

On the basis of a meta-analysis, a dosage model of psychotherapeutic effectiveness was described that demonstrated a positive relationship between the log of the number of sessions (dose) and the normalized probability of patient improvement (effect) (4). Subsequent dose-effect work has provided evidence for the differential, but lawful, responsiveness to psychotherapy of various symptoms (7), interpersonal problems (3; 9), and diagnoses (4; 11). Basically, the dosage model describes a pattern of relatively rapid early improvement, with more and more sessions needed to achieve incremental improvement later in treatment (a pattern of diminishing returns).

Phase Model

The phase model (5) extended and interpreted the dosage model. The phase model proposes three progressive sequential phases of the psychotherapeutic recovery process: (a) remoralization, the enhancement of well-being; (b) remediation, the achievement of symptomatic relief; and (c) rehabilitation, the reduction of troublesome, maladaptive behaviors that interfere with life functioning (e.g. functioning in areas such as family relationships and work). The phase model suggests that the decelerating curve of improvement for a patient can be attributed to the increasing difficulty of treatment goals as they change over the course of treatment. Both the dosage and phase models rely on group data to provide outcome information for an average patient. However, research has shown that patterns of improvement for individuals vary around this general trend (1; 6; 8; 10).

Expected Treatment Response (ETR) Model

The Expected Treatment Response (ETR) model assumes an underlying log-linear course of recovery in treatment for each patient, as described in the dosage model. ETR model utilizes a hierarchical linear modeling strategy (2) to depict a patient’s behavioral health status over treatment as a log-linear function of session number; then it uses pretreatment clinical characteristics (e.g., severity, chronicity, previous treatment, treatment expectation) to predict the patient's expected response over the course of his or her treatment. Using the results of such an individualized growth curve analysis for a large sample of outpatients in psychotherapy, a single patient's course of treatment can be predicted as soon as his/her intake information is available. Furthermore, ongoing therapeutic effectiveness can be assessed for a single patient by tracking the patient’s actual progress in comparison to his or her expected progress.

This new technique, with its focus on the clinical management of an individual patient makes new kinds of measurement systems necessary. The instruments must be consumer friendly (e.g., relevant to the patient, not too long, not too complicated), easy to use in daily practice (several times over the course of treatment), and have enough information to support clinical decision-making in an ongoing treatment. Polaris-MH was developed to fulfill these new kinds of criteria for an outcome assessment system.