A comparison of patient and clinician assessments of functional ability in predicting number of hospitalizations for older patients with left ventricular dysfunction
To compare the relative ability of patient self-assessments and clinician assessments of functional limitations for predicting hospitalizations for older adults with left ventricular dysfunction.
Latent growth model analyses used secondary data from a clinical drug study, the Studies of Left Ventricular Dysfunction. Self-report and clinician-report data of patients' functional limitations were collected at baseline, and 6 weeks, 12 months, and 24 months postbaseline to assess the effects of an angiotensin-converting enzyme inhibitor versus placebo. Latent growth model analyses were used to assess the relative effect of baseline ratings and changes in these 2 ratings of patient functional limitations over the initial 2 years of the study (ie, the drug study component) as potential predictors of the number of hospitalizations occurring over 3 subsequent years.
Data from 1099 patients aged 65+ years were analyzed. Both baseline patient assessments and changes in patient assessments of functional limitations were significant predictors of the number of subsequent hospitalizations (? = 0.119 and 0.273, respectively). This was not the case for clinician assessments. Moreover, baseline patient and clinician assessments of the patient's functional limitations were correlated at 0.42, indicating only a modest relationship between the 2 assessments.
These results demonstrate the greater utility of patient self-report measures for predicting hospitalizations. Incorporating patient-reported outcomes into clinical practice will expand understanding of the effect of diseases. Assessing changes in patient reports yields important insights into those patients more likely to have multiple admissions for heart failure.