Premium
Testing Predictive Efficiency of Patient Classifications for Methadone Maintenance Clients
Author(s) -
Babst Dean V.,
Ellis Rosalind,
Schmeidler James
Publication year - 1976
Publication title -
british journal of addiction to alcohol and other drugs
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.424
H-Index - 193
eISSN - 1360-0443
pISSN - 0007-0890
DOI - 10.1111/j.1360-0443.1976.tb00107.x
Subject(s) - typology , methadone maintenance , methadone , predictive validity , consistency (knowledge bases) , classification scheme , test (biology) , psychology , medicine , clinical psychology , psychiatry , computer science , machine learning , artificial intelligence , paleontology , archaeology , biology , history
Abstract This study tests the predictive accuracy of a typology for male methadone maintenance patients. The test consists of ascertaining how well the patient classification developed earlier predicts program retention for a new group. Each client was followed up after exactly two years to determine whether he was still in the program. The typology was developed using data based on the first patients admitted to the Dole‐Nyswander methadone maintenance program from fiscal 1965 through 1968. The predictive efficiency of the typology was tested with subjects admitted in fiscal 1969. The classification scheme had some predictive ability, but in general it performed poorer than most other prediction tests of this type. Within this study, the patient classifications with the most consistency were those based on two main items, number of prior convictions and job status at admission.