Can the short index of problems (SIP) be improved? Validity and reliability of the three-month SIP in an emergency department sample.

Abstract

Although the Short Index of Problems (SIP) is often used, little is known about the psychometric properties of the SIP in special populations. The present study seeks to determine the following: (1) whether it is possible to substitute items to enhance the psychometric properties of the SIP and (2) whether the SIP, or improved scale, is as sensitive as the Drinker Inventory of Consequences (DrInC) to assess intervention effectiveness. The sample consisted of 404 injured patients who were treated in the Emergency Department (ED) of a major teaching hospital that serves southern New England. Three approaches were used to guide development of the 3-month SIP-R, the potential alternative to the SIP. Cronbach's alpha assessed intrascale reliability; hierarchical multiple regression assessed construct validity; performance of the scales assessing intervention change were compared to the total 3-month DrInC as a function of intervention using analysis of covariance (ANCOVA). There was no evidence that changing the current SIP items will significantly improve performance. The 3-month SIP performed as well as the 3-month DrInC-2R in predicting 12-month DrInC scores and in determining intervention change at 12 months. Of the 45 DrInC items, 31 also predicted a difference across intervention groups. These results suggest there is no advantage to changing the current SIP items. The 3-month SIP is a psychometrically sound measure for assessing consequences of alcohol consumption in an ED sample and is almost as sensitive to intervention change as the full DrInC.