Objective: Functional status measures are potent independent predictors of hospital outcomes and mortality. The study objective was to compare medical record with interview data for functional status.
Subjects and methods: Subjects were 525 medical patients, aged 70 years or older, hospitalized at an academic medical center. Patient interviews determined status for 7 basic activities of daily living (BADLs) and 7 instrumental activities of daily living (IADLs). Medical records were reviewed to assess documentation of BADLs and IADLs.
Results: Most medical records contained no documentation of individual BADLs and IADLs (61% to 98% of records lacking documentation), with the exception of walking (24% of medical records lacking documentation). Impairment prevalence was lower in medical records than at interview for all BADLs and IADLs, and agreement between interview and medical record was poor (kappa < 0.40 for individual BADLs and IADLs). Sensitivity of the medical record for BADL and IADL impairment was poor (range 95% to 44%), using the interview as a reference standard. Sensitivity and specificity of the medical record for detection of BADL and IADL impairment changed substantially when records with nondocumentation of functional status were excluded or were assumed to be equivalent to independence.
Conclusions: The results suggest that the medical record is a poor source of data on many functional status measures, and that assuming that nondocumentation of functional status is equivalent to independence may be unwarranted. Given the prognostic importance of functional status measures, the results highlight the importance of developing reliable and efficient means of obtaining functional status information on hospitalized older patients.