Assessing readmission rate of alcoholism treatment: Developing patient-treatment matching by using Chi-square automatic interaction detection (CHAID).

摘要

The high health care costs of treating alcoholics, especially alcoholics with multiple readmissions for alcohol treatment, have focused increased attention on the effectiveness of alcoholism treatment. Because several recent studies have revealed the heterogeneous nature of alcoholics, researchers and clinicians now generally believe that there is no single treatment that is effective for all alcoholics. Therefore, the focus of research on effectiveness has evolved toward patient-treatment matching. While some health services research has been conducted to match patients with specific treatment intervention, those studies have so far failed to overcome a variety of conceptual and methodological problems. Methodological problems include selecting effective matching variables on a comprehensive basis and analyzing higher order interaction effects among the categorized variables.; Drawing upon the SARS data set of the Iowa Department of Public Health which contains information on 6,912 alcoholics, this study incorporates four categories of patient variables including demographic, psychiatric, alcohol-related, and social and personal. These variables are analyzed by means of the Chi-square automatic interaction detection (CHAID) to first discover the determinants of the outcome of alcoholism treatment (readmission) and then to develop a patient-treatment model.; The findings of this study indicate that times of prior arrest, insurance status, current frequency of alcohol use, gender, source of referral, multiple addictions, and employment status significantly predict the readmission of alcoholism treatment. However, the findings also reveal that interactions among patient's variables exist. This suggests that further study directed at matching alcoholics to treatments should take such interactions into account.; Although it is generally assumed that inpatient detoxification is a necessary step in the treatment of severe alcoholics and that inpatient treatment results inevitably produce a better outcome, the findings of this study do not support either of the above general assumptions. However, more intensity of treatment such as residential treatment, intensive outpatient, or double doses help to improve the outcome of alcoholism treatment depending on the characteristics of the subgroup. Several important implications for policy and administration are discussed at the conclusion of this study.