Assessment of blood pressure measurement quality in the baseline surveys of the WHO MONICA project.

Because of the general inavailability of reference standards, there exist no common procedures to assess the quality of blood pressure measurements in epidemiological population surveys. To approach this problem within the collaborative international WHO MONICA Project, a standardized assessment of BP measurement quality was developed and applied to the forty-seven baseline surveys of that project. The entire assessments were carried out in retrospect, that is, only after each population survey had been completed. The assessment was focused on the procedures of quality assurance and control as reportedly applied in each survey, and on quality indicators which were derived from the recorded blood pressure values of each survey. The definitions of specific quality assessment items were based on the MONICA project protocol and on sources in the pertinent literature. The available information on quality assurance and control procedures depended solely on self-reports by local survey organizers and on site visits, and was occasionally found to be at variance with the actual blood pressure recordings. Therefore, quality indicators derived from actual blood pressure recordings were far more informative and comparable between surveys. Each survey was rated as optimal, satisfactory, or unsatisfactory with regard to single quality items and these single scores were used jointly to compute a summary score of blood pressure measurement quality for each survey. This summary score indicated that 39 out of 47 MONICA baseline surveys showed optimal or satisfactory BP measurement quality. Limitations and potentials for improvement of quality assessments became apparent. We conclude that a standardized assessment of BP measurement quality in epidemiological population surveys seems feasible and propose that quality assessment methods similar to the ones suggested here become a routine part of future epidemiological analyses of blood pressure values and hypertension in populations. This should facilitate valid study comparisons.