TEST-RETEST RELIABILITY OF THE ADDICTION SEVERITY INDEX COMPOSITE SCORES AMONG CLIENTS IN RESIDENTIAL-TREATMENT

The authors investigated the test-retest reliability of the Addiction Severity Index (ASI) composite scores and the use of ''free-living vs. institutionalized'' reporting periods among 112 patients admitted to either of two residential, drug-free treatment programs in New England . Alcohol and drug scores were substantially lower for institutionaliz ed vs. free-living reporting periods. Intraclass correlation coefficie nts were high (> 0.80) for the alcohol, drug, employment and legal sco res, but lower (0.50-0.59) for the psychiatric and medical scores and did not differ substantially by, type of reporting period or by lag pe riod between end of the reporting period and the interview. The use of free-living reporting periods yields more valid ASI scores for drugs and alcohol, in particular, and yields comparable reliability.