Severity measurement using a generic instrument - A feasibility study in ambulatory care involving patients with diabetes or asthma

The object of the study was to assess the performance of a generic severity measure [the Duke severity of illness (DUSOI) scale) in ambulatory care in the UK, This was done in 2 settings. First, all 181 patients with non-insulin-requiring diabetes mellitus attending a single general practice had 4 retrospective ratings of severity of illness, 2 before and 2 after training of the rater, Secondly, a convenience sample of 232 patients with non-insulin-requiring diabetes mellitus and 376 patients with asthma from 57 general practices and 7 hospitals in the north of England had a single rating of severity of illness and completed a questionnaire on functional health status, From the first setting we estimated the test-retest reliability of the rating of severity: the intraclass correlation coefficient for the 2 ratings before training was 0.72 and after it was 0.85, Training increased diagnosis identification by 0.91 (95% CI: 0.72-1.1) diagnoses per patient, From the second setting we estimated the concurrent validity of the DUSOI scale as assessed by correlation with the functional health status measures, For 42% of patients with diabetes their diabetes was rated as their most severe diagnosis; for 85% of patients with asthma their asthma was rated as their most severe diagnosis, For patients with asthma the DUSOI scores correlated with physical health domains. For patients with diabetes similar correlations were much weaker, We conclude that the DUSOI scale is capable of reliable use and has concurrent validity, However, it requires further evaluation to fully establish its utility in UK ambulatory care.