Validity of self-reported diagnoses leading to hospitalization: a comparison of self-reports with hospital records in a prospective study of American adults

Am J Epidemiol. 1998 May 15;147(10):969-77. doi: 10.1093/oxfordjournals.aje.a009387.


The authors compared interview reports with hospitalization records of participants in a nationally representative survey to determine the accuracy of self-reports of ischemic heart disease, stroke, gallbladder disease, ulcers, cataract, hip fracture, colon polyps, and cancers of the colon, breast, prostate, and lung. The study cohort consisted of 10,523 participants from the First National Health and Nutrition Examination Survey in 1971-1975 who were aged 25-74 years at the baseline examination and who completed a follow-up interview in 1982-1984. Self-reports of hospitalization for breast cancer were confirmed as accurate for 100% of cases where a hospital record was available. Self-report accuracy was also high for ischemic heart disease (84%), cataract (83%), and hip fracture (81%); it was moderate for lung cancer (78%), prostate cancer (75%), gallbladder disease (74%), colon cancer (71%), and stroke (67%); but it was low for ulcers (54%) and colon polyps (32%). Some of the self-reports of ulcers (20%), hip fracture (9%), ischemic heart disease (7%), and stroke (7%) were found to reflect diagnoses of other conditions of anatomic proximity. Accuracy of self-reports improved with higher levels of education, but was not generally related to age, gender, race, alcohol use, or smoking. The results suggest that self-reports of some diseases can be taken as accurate, but self-reports of other conditions might require medical record verification in epidemiologic follow-up studies.

Publication types

  • Comparative Study
  • Research Support, Non-U.S. Gov't
  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Adult
  • Aged
  • Bias
  • Cohort Studies
  • Data Collection / standards*
  • Diagnosis*
  • Diagnosis-Related Groups / statistics & numerical data*
  • Educational Status
  • Follow-Up Studies
  • Hospitalization / statistics & numerical data*
  • Humans
  • Length of Stay / statistics & numerical data
  • Middle Aged
  • Patient Participation / statistics & numerical data*
  • Population Surveillance
  • Predictive Value of Tests
  • Prospective Studies
  • Truth Disclosure
  • United States