Statistical analysis
We used descriptive statistics like frequencies, proportions, and mean (standard deviation, SD) to summarise the sample characteristics. We dichotomised the outcome variable BPCR as optimal (followed at least four of the six steps of BPCR) and suboptimal (followed fewer than four steps of BPCR). Using binary logistic regression analysis, we computed the Odds and adjusted Odds ratios (OR and Adj OR) with a 95% confidence interval (95% CI). The ORs for age (continuous variable) was related to a one-year increase in age. We considered those predictor variables with p<0.2 in the unadjusted analysis for adjusted analysis. A two-sided p<0.05 was considered statistically significant.