Acute confusion indicators: risk factors and prevalence using MDS data

OBJECTIVE: The purpose of this study was to use Minimum Data Set (MDS) data from LTC (long-term care) in one Midwestern state to test whether risk variables derived from a conceptual model developed from findings in acute care predicted acute confusion in long-term care residents. DESIGN: Cross-sectional. SETTING: MDS nursing home records. POPULATION: The sample was composed predominantly of women (n = 1,775). INTERVENTIONS: The 1995 MIS annual reviews of nursing home residents from a Midwestern state provided the data for analysis; however missing data, on the average about 10% for the acute confusion/delirium indicators, reduced the sample to 2,318. Based on the conceptual model and items available in the MDS, precipitating factors selected for analysis included: dehydration, hypoxia, infections, and medications. Individual vulnerability factors selected for analysis included: age, diagnosis of dementia, various chronic medical diagnoses/conditions, indicators of frailty such as falls; multiple medications; and sensory impairments. MAIN OUTCOME MEASURE(S): The prevalence of acute confusion found in these data (13.98%) is similar to other studies of acute confusion prevalence in abstracted data files. Frequencies of the indicators suggest that cognitive ability variation and periods of motor restlessness/lethargy are the most readily recognized symptoms of acute confusion by nursing home staff. Of the variables that were significant in the univariate analysis, the only variables that contributed to the explanation of acute confusion in the logistic regression analysis were inadequate fluid intake, dementia status, and a fall in the past 30 days. The strongest contributing factor to acute confusion in this population was inadequate fluid intake. Although medications are the most frequent cause of acute confusion in older hospital patients, possibly because of the age-related sensitivity of the brain to the effects of drugs, this was not the case in these LTC residents. RESULTS/CONCLUSIONS: Although these results support the hypothesized relationship of dehydration and acute confusion, limitations of this cross-sectional analysis need to be addressed. The items used for fluid intake lacked the accuracy of quantitative measurement in a more precise clinical investigation. There is a concern about the reliability of the delirium indicators. Hawes et al. conclude that nursing home staff do not consistently detect and document these indicators. Also, the amount of missing data is of concern. Future research endeavors should focus on refining the assessment of acute confusion in LTC. Other research efforts should focus on using the MDS data to identify areas for clinical intervention. The MDS is one of the few mandated clinical databases that captures nursing care. Nurses involved in health services research can use the data to affect nursing care issues in long-term care. [CINAHL abstract]