Drug Burden Index and Hospitalization among Community-Dwelling Older People

Background: Medications with anticholinergic and sedative effects carry significant risks in older people. Adverse events arising from the use of these medications may also lead to hospitalization and contribute to length of stay. The Drug Burden Index (DBI) is a tool that measures a person’s total exposure to medications with anticholinergic and sedative properties, using the principles of dose response and maximal effect. Cumulative anticholinergic and sedative drug burden measured using the DBI has been associated with clinically important outcomes in older people. The association between the DBI and hospitalization still remains relatively unknown.

Objective: The main aim of this study was to evaluate the relationship between DBI and hospitalization in a population-based sample of community-dwelling older Finns over a 1-year period.

Methods: The health status and medication use of 339 community-dwelling ≥75-year-old Finns were assessed in 2004. Data on hospitalizations over the following year were obtained from the national discharge register. Two different measures were used to assess hospitalizations in the study sample: (i) the proportion of hospitalized participants; and (ii) the number of hospital days per person-year. Estimates for the number of hospital days per person-year and rate ratios (RRs) with 95% confidence intervals (CIs) were calculated using Poisson or negative binomial regression analysis.

Results: A total of 127 participants (38%) were exposed to DBI medications; 27% had a low DBI (>0 to <1), and 11% had a high DBI (≥1). The number of hospital days per person-year was 7.9 (95% CI 7.6, 8.3) for the unexposed participants (DBI = 0) and 13.4 (95% CI 12.8, 14.1) for the exposed participants (DBI >1); the age, gender and co-morbidity adjusted RR of hospital days per person-year between the exposed and unexposed participants was 1.26 (95% CI 1.18, 1.35). Between the low and high DBI groups, the difference in the number of hospital days per person-year was insignificant (p = 0.42). In multivariate analyses, the number of regularly used medications (RR= 1.12 [95% CI 1.00, 1.26] per additional medication) and the measure of basic activities of daily living Barthel Index (RR = 0.94 [95% CI 0.88, 0.99] per increase) were independently associated with the use of hospital days.

Conclusion: Exposure to DBI medications was associated with a greater use of hospital days, but a cumulative dose-response relationship between DBI and hospitalization was not observed. The number of regularly used medications and functioning in the basic activities of daily living predicted hospital care utilization.