The use of a Multi-morbidity Index (MMI) in Patients Receiving Home Parenteral Nutrition (HPN)

Abstract

Introduction: Although HPN patients often have concomitant medical conditions, little has been done to assess the role of multi-morbidity in relation to HPN care (1-3). A validated method for determining multi-morbidity in HPN patients could be beneficial in determining the needs and risks for each HPN patient. We previously studied the effect of a multidisciplinary nutritional support team (MNST) on quality improvement HPN care (QIP-PN) (4). In a secondary analysis, we explored the options for measuring multi-morbidity of patients in the QIP-PN study. In this report, we describe the results of a pilot study on measuring multi-morbidity in HPN patients to determine its value in HPN care. Objective: Determine the role of multi-morbidity scoring in HPN care. Methods: 60 HPN patients (30 study patients and 30 case-matched controls) were reviewed for demographics, CIRS Score, number of PN formula changes, hospitalizations and hospital length of stay. Histograms were produced to visually examine the data and cutoffs were selected for CIRS scores based on this information. A matrix plot was drawn showing Spearman correlations among the three variables of interest. The data was then subjected to Kruskal-Wallis and Wilcoxon Rank Sum Tests and Negative Binomial Regression Models to determine if the groups differed significantly. Results: 28 patients had CIRS scores <17, while 32 patients had CIRS scores ≥17. The results of negative binomial regression indicate that the number of hospitalizations was a significant predictor of total hospital LOS at the 5% alpha level and that the total hospital LOS of patients with a CIRS Score <17 was significantly different from that of patients with a CIRS≥17 at 5% alpha level. Conclusion: We found the CIRS method was well suited to use in HPN patients. It was easy to learn and perform. It provided a single variable which could be used a summary statement of patient multimorbidity. When the CIRS data on 60 long term HPN patients was analyzed, a trend emerged wherein a higher CIRS score was associated with greater hospital LOS. This approach requires greater study and validation.