Abstract
Objective(s)To compare agreement between self-reported height, weight and blood pressure measurements submitted to an online contraceptive service with researcher-measured values and document strategies used for self-reporting.DesignAn observational study.SettingAn online sexual health service which provided the combined oral contraceptive pill, free of charge, to users in Southeast London, England.ParticipantsBetween August 2017 and August 2019, 365 participants were recruited.Primary and secondary outcome measuresThe primary outcome, for which the study was powered, was the agreement between self-reported and researcher-measured body mass index (BMI) and blood pressure measurements, compared using kappa coefficients. Secondary measures of agreement included sensitivity, specificity and Bland Altman plots. The study also describes strategies used for self-reporting and classifies their clinical appropriateness.Results327 participants fully described their process of blood pressure measurement with 296 (90.5%) classified as clinically appropriate. Agreement between self-reported and researcher-measured BMI was substantial (0.72 (95% CI 0.42 to 1.0)), but poor for blood pressure (0.06 (95% CI −0.11 to 0.23)). Self-reported height and weight readings identified 80.0% (95% CI 28.4 to 99.5) of individuals with a researcher-measured high BMI (≥than 35 kg/m2) and 9.1% (95% CI 0.23 to 41.3) of participants with a researcher-measured high blood pressure (≥140/90 mm Hg).ConclusionIn this study, while self-reported BMI was found to have substantial agreement with researcher-measured BMI, self-reported blood pressure was shown to have poor agreement with researcher-measured blood pressure. This may be due to the inherent variability of blood pressure, overdiagnosis of hypertension by researchers due to ‘white coat syndrome’ or inaccurate self-reporting. Strategies to improve self-reporting of blood pressure for remote prescription of the combined pill are needed.
Funder
National Institute for Health Research