Inconsistency in prevalence of hypertension based on self-reports and use of standard tests: Implications for large scale surveys.

Abstract

Objective: Biomarkers are increasingly integrated into population-based surveys to provide reliable estimates of the prevalence of specific diseases. The Demographic and Health Surveys have recently incorporated blood pressure measurements; however, little is known about the extent of agreement between measured and reported levels of hypertension in India. The objective of this study was to examine the extent of agreement between self-reported hypertension and the results of standard blood pressure measurements, as well as to explore the risk groups and factors associated with inconsistencies in self-reported and biomedically measured hypertension.

Methods: Reliability measures such as sensitivity, specificity, and kappa statistics were used to examine inconsistencies in self-reported and biomedically measured hypertension in the National Family Health Survey-4 data. Multilevel logistic models were adopted to analyse the respondent characteristics related to both false-positive and false-negative responses in the survey.

Results: Compared to biomedically measured hypertension, self-reported hypertension was inconsistent and disproportionate at disaggregated levels in India. While self-reports severely underestimated hypertension among men aged 15-54 years and women aged 35-49 years, it overestimated hypertension among women below the age of 35 years. The inconsistency in self-reported and biomedically examined hypertension had deviations from a sex standpoint. Women aged <35 years reported a false-positive prevalence of hypertension. False-negative responses were elucidated among women aged ≥35 years and men aged 15-54 years. The likelihood of false-positive responses was higher among pregnant and obese respondents, and those who consumed alcohol.

Conclusion: The significant deviance of self-reporting of hypertension from the prevalence derived based on standard tests further indicates the need for adopting standard tests in all emerging future large-scale surveys. A back-check survey is recommended to understand and differentiate the excessive false-positive reporting of hypertension among women aged 15-35 years.