Underreporting of recognized adverse drug reactions by primary care physicians: an exploratory study

Pharmacoepidemiol Drug Saf. 2011 Dec;20(12):1287-94. doi: 10.1002/pds.2172. Epub 2011 Jun 23.


Purpose: This study evaluated the magnitude of underreporting of adverse drug reactions (ADRs) and investigated possible reporting patterns according to patient characteristics and the type of reaction based on the integration of information obtained from primary care electronic medical records (EMRs) and the Spanish Pharmacovigilance System.

Methods: This investigation was a descriptive retrospective study analysing ADRs recorded in 2005 in the EMRs from six health centers in Zaragoza (Aragon, Spain) with a covered population of 126,838 subjects. The associations between the probability of reporting and the reaction and drug type were studied using logistic regression models adjusted by age and sex.

Results: The total number of ADRs recorded in the EMRs was 543, of which 65.7% were reported to the Spanish Pharmacovigilance System. Positive associations were found between the probability of reporting an ADR and advanced age of patients (OR for ≥76 years = 2.0; 95%CI 1.1-3.6), involvement of the reproductive system (OR = 7.9; 95%CI 1.02-60.2) and involvement of psychiatric disorders (OR = 4.0; 95%CI 1.4-11.6). Negative associations were found between reporting an ADR and early age of patients (OR for 0-14 years = 0.2; 95%CI 0.1-0.6) and the use of antimicrobial drugs (OR = 0.6; 95%CI 0.4-0.9).

Conclusions: This study tackles an important public health problem directly related to patients' safety and highlights the utility of EMRs for investigating the current significance of ADR underreporting. It also makes us think that primary care physicians seem to have selective reporting patterns based on their familiarity with the reaction type and the drug causing the reaction as well as on the age of patients.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Adverse Drug Reaction Reporting Systems / statistics & numerical data*
  • Age Factors
  • Aged
  • Child
  • Child, Preschool
  • Electronic Health Records / statistics & numerical data*
  • Female
  • Humans
  • Infant
  • Logistic Models
  • Male
  • Middle Aged
  • Pharmacovigilance
  • Physicians, Primary Care / statistics & numerical data*
  • Retrospective Studies
  • Spain
  • Young Adult