The multimorbidity interaction severity index (MISI): A proof of concept study.

Therapeutic decision-making for patients with multimorbidity (MM) is challenging. Clinical practice guidelines inadequately address harmful interactions and resulting therapeutic conflicts within and among diseases. A patient-specific measure of MM severity that takes account of this conflict is needed.As a proof of concept, we evaluated whether the new Multimorbidity Interaction Severity Index (MISI) could be used to reliably differentiate patients in terms of lower versus higher potential for harmful interactions.Two hypothetical patient cases were generated, each with 6 concurrent morbidities. One case had a low (i.e., low conflict case) and the other a high (i.e., high conflict case) potential for harmful interactions. All possible interactions between conditions and treatments were extracted from each case's record into a multimorbidity interaction matrix. Experienced general internists (N = 18) judged each interaction in the matrix in terms of likely resource utilization needed to manage the interaction. Based on these judgements, a composite index of MM interaction severity, that is, the MISI, was generated for each physician and case.The difference between each physician's MISI score for the 2 cases (MISIdiff) was computed. Based on MISIdiff, the high conflict case was judged to be of significantly greater MM severity than was the low conflict case. The positive values of the inter-quartile range, a measure of variation (or disagreement) between the 2 cases, indicated general consistency of individual physicians in judging MM severity.The data indicate that the MISI can be used to reliably differentiate hypothetical multimorbid patients in terms of lesser versus greater severity of potentially harmful interactive effects. On this basis, the MISI will be further developed for use in patient-specific assessment and management of MM. The clinical relevance of the MISI as an alternative approach to defining MM severity is discussed.