O05 TERMINAL DIGIT PATTERN IN COMMUNITY-BASED HYPERTENSION SCREENING PROGRAMS: 5-YEAR ANALYSIS FROM MAY MEASURE MONTH NEPAL

Background and Objectives: Terminal digit preference leads to misclassification of hypertension. Despite strong promotion of community-based hypertension screening, the extent of terminal digit preference in such programs is not well understood. We aimed to identify the proportion of blood pressure (BP) ending in zero, and factors associated with the most repeated digit. Methods: A cross-sectional study was conducted using data from May Measure Month (MMM) Nepal 2017, 2018, 2020, 2021, and 2022. Participants with available data on the type of BP monitors were included. In MMM Nepal, adults≥18 years of age were eligible and recruited via convenience sampling. Data enumerators are community health workers or college students trained for BP measurement through short trainings. Without bias, the proportion of BP ending in zero is expected to be 10% and 20% for automated and manual monitors, respectively. Results: A total of 48,442 participant were included with 29.2% age≥50 years and 51.8% females. BP appeared to be normally distributed with “spikes” at terminal digits of zero (Figure 1). 90.9% of participants were measured their BP with automated monitors and 8.1% with manual monitors. The proportion of BP ending in zero was about 1.7 and 4.8 times greater than the expected proportion using automated and manual monitors, respectively (automated: 17.0% [SBP], 16.7% [DBP], manual: 95.3% [SBP], 93.8% [DBP]). Logistic regression showed age≥50 (OR 1.13 [95%CI 1.06, 1.20]) and SBP≥140mmHg (1.63 [1.49, 1.78]) were associated with SBP ending in zero, and age≥50 (1.14 [1.08, 1.22]) and DBP≥90 mmHg (2.11 [1.96, 2.28]) were associated with DBP ending in zero, both after controlling for sex, smoking, alcohol use, diabetes, cardiovascular diseases, antihypertensive medicine use, and the type of BP monitors used. Conclusion: The proportion of BP ending in zero was higher than expected with any type of BP monitors used. More studies are needed to investigate whether these distributions were by chance or due to errors.