Problems of HRQL assessment: how much is too much?

Few investigations have explored the extent to which a number of less serious problems contribute to health-related quality of life (HRQL) impairment beyond a smaller number of more serious problems. Our study examined the relative impact of the number and severity of patient problems on HRQL. This study analyzed results from 100 patients with irritable bowel syndrome (IBS) and 100 women with polycystic ovary syndrome (PCOS). We compared an aggregation method that generated scores using all items that could affect HRQL as the denominator (all items) to a second method that considered only items that patients experienced as problems (problem items). For each IBS and PCOS questionnaire domain, we used regression analysis to examine the relative contribution of scores from the all-items and problem-items approaches to prediction of scores on the other instruments. Of 57 correlations between IBS domains and scores on the other questionnaires, the all-items method explained a statistically significant additional proportion of the variance beyond the problem items in 29 cases and the problem items a significant additional proportion of the variance between the all items in 7 (p < 0.001). Of the 28 correlations between the PCOS and the other questionnaires, the all-items approach explained a significant additional proportion of the variance in 21 cases and the problem items approach in 5 (p < 0.001). With IBS and PCOS, including all potential items in calculating a domain score provided a more accurate portrayal of HRQL than a selected approach focusing on problem items