Falls documentation in nursing homes: Agreement between the minimum data set and chart abstractions of medical and nursing documentation

OBJECTIVES: To assess the agreement between falls as recorded in the Minimum Data Set (MDS) and fall events abstracted from chart documentation of elderly nursing home (NH) residents. DESIGN: Secondary analysis of data from a longitudinal panel study. SETTING: Fifty-six randomly selected NHs in Maryland stratified by facility size and geographic region. PARTICIPANTS: Four hundred sixty-two NH residents, aged 65 and older, in NHs for 1 year. MEASUREMENTS: Falls were abstracted from resident charts and compared with MDS fall variables. Fall events data obtained from other sources of chart documentation were matched for the corresponding periods of 30 and 180 days before the 1-year MDS assessment date. RESULTS: For a 30-day period, concordance between the MDS and chart abstractions of falls occurred in 65% of cases, with a kappa coefficient of 0.29 (P < .001), indicating fair agreement. Concordance occurred between the sources for 75% of cases for a 180-day period, with a kappa of 0.50 (P < .001), indicating moderate agreement. During the 180-day period, chart abstractions showed that 49% of the sample fell, whereas the MDS revealed that only 28% fell. An analysis of residents whose falls the MDS missed indicated that these residents had significantly more activity of daily living impairment and significantly less unsteady gait and cane/walker use. CONCLUSION: The MDS underreported falls. Nurses completing MDS assessments must carefully review residents' medical records for falls documentation. Future studies should use caution when employing MDS data as the only indicator of falls.