Original articles
Validity of a Self-reported History of Doctor-diagnosed Angina

https://doi.org/10.1016/S0895-4356(98)00146-2Get rights and content

Abstract

The objective of this study was to assess the validity of a self-reported history of doctor-diagnosed angina in population-based studies in men. Subjects were 5789 men from the British Regional Heart Study who reported being without an angina diagnosis at entry (1978–1980) and were alive at the end of 1992, aged 52 to 75 years. In 1992, subjects were asked in a self-administered questionnaire if they recalled ever having had a doctor diagnosis of angina. Self-report of diagnosed angina was compared with general practice (GP) record of angina obtained from reviews of medical records from study entry to the end of 1992. Men were followed for a further 3 years from 1992 for major ischemic heart disease events. The prevalence of diagnosed angina in 1992 was 10.1% according to self-reported history and 8.9% according to GP record review. There was substantial agreement between the two sources of information: 80% of men with a GP record of angina reported their diagnosis, and 70% of men who reported an angina diagnosis had confirmation of this from the record review. When all ischemic heart disease (angina or myocardial infarction) was considered, agreement was higher. Genuine angina was likely in many of the 177 men who had self-reported angina not confirmed by the GP record review: 78 had an ischemic heart disease history (myocardial infarction or coronary revascularization) identified by the review, and 31 had a GP record of angina after 1992. Angina symptoms, nitrate use, cardiological investigation, and surgical intervention for angina compared between agreement groups showed a very consistent pattern. All these indicators of angina were most common in men with both self-report and GP record of angina, least common in men with neither self-report nor GP record of angina, but had a substantially higher prevalence in men with self-reported angina only than in those with GP-recorded angina only. After 3 years follow-up from 1992, 9.5% of men with both self-report and GP record of angina, and 11.3% of men with self-reported angina only had experienced a new major ischemic heart disease event; compared to 5.7% of men with a GP record of angina only and 2.7% of those without angina by either criteria. This pattern of risk remained similar after adjustment for age and previous myocardial infarction. These results suggest that self-reported history of a doctor diagnosis of angina is a valid measure of diagnosed angina in population-based studies in men.

Introduction

Although epidemiological studies and surveys often rely on subjects’ self-reported medical history to ascertain the presence of a variety of diagnoses and conditions, several investigators have drawn attention to the lack of information regarding the validity of these measures 1, 2, 3. In the field of cardiovascular epidemiology, there are now a number of studies which have addressed this issue by comparing self-reports with medical records for a range of cardiovascular conditions 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14. The agreement between these two sources of information has varied considerably depending on the condition in question. Agreement has generally been found to be good or reasonable for relatively well defined diagnoses such as diabetes 3, 4, 5, 6, 7, 8, 9, 10 and myocardial infarction 3, 5, 6, 10, 11, 12, 13, 14, but of concern is the low level of agreement found for conditions with less clear diagnostic criteria 9, 10. Very few studies have assessed the degree of concordance between self-report and medical record specifically for angina 6, 10, 13.

When assessing the validity of a measurement against a reference or standard, sensitivity, specificity, and predictive value are the appropriate analytic measures [15], and these have been used widely in validation studies comparing self-reports with medical records. However, while medical records may provide the best available comparison for self-reported medical history, many authors have acknowledged that they are not an ideal gold standard 3, 4, 5, 8, 12, and some have suggested that it cannot be assumed that they necessarily provide the more accurate source of information 6, 9. Studies investigating the quality and completeness of medical records indicate that errors and omissions are not uncommon 16, 17, 18, in addition, any process used to review the records is likely to introduce a further source of error. It therefore seems inappropriate to judge the validity of self-reported medical history entirely on the basis of comparison with medical records.

This study aimed to investigate the validity of self-reported history of a doctor diagnosis of angina obtained from a self-administered questionnaire in men in the British Regional Heart Study. We examined the agreement between self-report of diagnosed angina and angina ascertained from reviews of medical records. In addition, we also assessed validity by examining the associations of self-reported angina and of recorded angina with other indicators of angina (angina symptoms, cardiological investigations, and treatment), and with the risk of a subsequent new major ischemic heart disease event.

Section snippets

Methods

The British Regional Heart Study (BRHS) is a prospective study of cardiovascular disease in which 7735 middle-aged men, randomly selected from one general practice in each of 24 towns in Britain, have been followed up since 1978–80. Details of selection of towns, general practices, and subjects have been reported [19]. The men were aged 40–59 years at the start of the study, and the baseline examination included a questionnaire administered by research nurses, from which pre-existing

Subjects Included in Validation Study

The 1992 questionnaire (Q92) was mailed to 6582 men who were then aged 52–75 years. Of the original 7735 men, 1088 had died and 65 had emigrated by November 1992. There was a response to the questionnaire from 5934 men (90%), 145 of whom were excluded from the validation study as they had either reported doctor-diagnosed angina at baseline (n = 140) or had not responded to the question on doctor-diagnosed angina at baseline (n = 5).

Prevalence of Diagnosed Angina

Of the remaining 5789 men, a doctor diagnosis of angina was

Discussion

For surviving men in the British Regional Heart Study, the level of agreement between self-report and medical record was substantial for a doctor diagnosis of angina, and very high for a doctor diagnosis of ischemic heart disease (angina or myocardial infarction). However, there was evidence that medical record reviews may not provide an adequate gold standard for diagnosed angina: the prevalence of other indicators of angina and the occurrence of subsequent major ischemic heart disease events

Conclusion

Although self-reported history and medical record reviews used together provide the most complete information on diagnosed angina in epidemiological studies, self-reported history alone appears to have a high degree of validity as a measure of doctor-diagnosed angina in population-based studies in middle-aged and elderly men. The use of self-reported history is further justified when all ischemic heart disease is considered.

Acknowledgements

The authors thank Professor A.G. Shaper and Dr. S.G. Wannamethee for their comments. The British Regional Heart Study is a British Heart Foundation Research Group and also receives support from the Department of Health. Fiona Lampe is supported by the Department of Health.

References (30)

  • R. Kehoe et al.

    Comparing self-reported and physician-reported medical history

    Am J Epidemiol

    (1994)
  • N. Haapanen

    Miilunpalo S, Pasanen M, Oja P, Vuori I. Agreement between questionnaire data and medical records of chronic diseases in middle-aged and elderly Finnish men and women

    Am J Epidemiol

    (1997)
  • G.A. Colditz et al.

    Validation of questionnaire information on risk factors and disease outcomes in a prospective cohort study of women

    Am J Epidemiol

    (1986)
  • A. Paganini-Hill et al.

    Accuracy of recall of hip fracture, heart attack and cancerA comparison of postal survey data and medical records

    Am J Epidemiol

    (1993)
  • A.M. Lagaay et al.

    Validation of medical history taking as part of a population based survey in subjects aged 85 and over

    Br Med J

    (1992)
  • Cited by (0)

    View full text