Treatment effectiveness score as an outcome measure in clinical trials.

A variety of measures are used for evaluating patients’ responses to substance abuse treatments. These range from physical measures (such as samples of urine, breath, hair, or blood), self-reports of drug use (such as the Addiction Severity Index (ASI) or the Time Line Follow-Back), self-reports of psychological or physiological functioning (such as symptom checklists or craving or mood ratings), and collateral reports. Physical indices of recent drug use, such as urine toxicology screens, are preferable to self-report or collateral reports for evaluating patients’ responses to drug abuse treatments because of their objectivity. In order to optimize the likelihood of both detecting individual episodes of problem drug use and correctly inferring drug abstinence based on urine toxicology results, guidelines have been suggested for collection procedures and timing for collection of urine specimens (Blaine et al. 1994; Cone and Dickerson 1992; Jain 1992). However, the difficult task of aggregating urine toxicology results remains, whether when interpreting the response of a single patient to a specific treatment or when evaluating a treatment’s effectiveness based on a group of patients’ responses in a clinical trial. Difficulties in aggregating urine toxicology results include, but certainly would not be limited to, such problems as the frequency and sensitivity of toxicology screens, early termination of some patients from treatment (or, conversely, the continued participation of some patients who respond poorly to treatment), and problems of analyzing a data matrix that contains a large number of missing datapoints. This chapter reviews the objective indices of treatment response that have traditionally been used and suggests three composite methods for evaluating these data: the Treatment Effectiveness Score (TES), the Joint Probability score (JP), and the Clinical Stabilization Score (CSS).