Problems with the Sickness Impact Profile: a theoretically based analysis and a proposal for a new method of implementation and scoring

The Sickness Impact Profile (SIP) is one of the most widely used health status measures, but there are problems with the measure that lead to inconsistent and illogical scores. There are many desirable features to the SIP development methodology in that it is based on a good range of items and the item weightings are valuable. The current method of scoring the SIP is the use of a summated total and was selected based on limited empirical evidence. However, in this paper we argue that there are problems with the SIP because the current empirically derived method of scoring is incompatible with both the underlying theoretical scaling framework (Thurstone scaling) and the nature of the items in the SIP. In addition, the items do not have properties consistent with the scaling methodology. We suggest that it is crucial to take both a theoretical and empirical approach to selecting a scoring method. To examine problems associated with the SIP we explored the underlying scaling methodology and identified the types of items in the SIP. A new method of scoring is proposed that is consistent with the items and scaling in the SIP, namely using the maximum individual weighting from the items that were checked as a category score. This new method of scoring resolves many of the previously observed problems in the SIP. The new method of scoring also presented the opportunity for a new implementation of the SIP that reduces the number of items that most respondents, especially those with severe limitations, would be asked. Without taking a theoretical approach to scoring we suggest that subsequent, empirically based, scale amendments are unlikely to solve the problems. It is proposed that this new method of scoring justifies a thorough empirical investigation.