Is suboptimal prescribing a risk factor for poor health outcomes in community-dwelling elders? The ICARe Dicomano study

Pharmacoepidemiol Drug Saf. 2010 Sep;19(9):954-60. doi: 10.1002/pds.1997.


Purpose: Mostly because of comorbidity and drugs consumption, older persons are often exposed to an increased risk of sub-optimal prescribing (SP). At present, few studies investigated the association between SP and long-term health outcomes. We examined the relation between SP and the risk of mortality and hospitalization in Italian older community-dwellers.

Methods: Older (65+ years) community-dwelling residents of a small town in Tuscany were enrolled in a longitudinal study. SP was defined as polypharmacy (use of 5+ drugs), prescription of inappropriate drugs (ID) according to Beers' criteria, and of potentially interacting drugs (PID), evaluated in 1995 and 1999. These three forms of SP were entered as time-dependent exposures into multivariable Cox regression analysis models, whose outcomes were mortality and hospitalizations through 2003.

Results: Of 1022 participants (mean age 73.0 +/- 6.8, 57% women), 220 were evaluated in 1995, 234 in 1999 and 568 in both waves. In univariate analysis, mortality was two-fold higher in participants with polypharmacy (73.4/1000 person/years, 95% CI 58.2-92.4 vs. 34.1, 95% CI 29.7-39.2; p < 0.001) or PID (72.7/1000 person/years, 95% CI 46.3-113.9 vs. 38.0, 95% CI 33.5-43.1; p < 0.001), whereas it was unrelated to the presence of ID. Hospitalization rates were independent of any form of SP. In multivariable models, polypharmacy, ID, and PID were no longer associated with an increased risk of death, and ID predicted a slightly increased risk of hospitalizations (HR 1.03, 95% CI 1.0-1.06, p = 0.048).

Conclusions: In this cohort, SP was not associated with an excess risk of poor health outcomes.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Aged
  • Aged, 80 and over
  • Databases, Factual
  • Drug Interactions
  • Female
  • Hospitalization / statistics & numerical data
  • Humans
  • Italy
  • Longitudinal Studies
  • Male
  • Multivariate Analysis
  • Outcome Assessment, Health Care*
  • Polypharmacy*
  • Practice Patterns, Physicians' / standards*
  • Proportional Hazards Models
  • Risk Factors
  • Time Factors