Utility of routine data reporting injuries requiring hospitalisation in Nepal: a secondary data analysis

Abstract

Background Injuries are an important public health issue in Nepal, contributing significantly to both morbidity and mortality. There is no injury surveillance system available, however healthcare service use is routinely reported to central government using the Health Management Information System (HMIS). The study was conducted as part of a wider programme of research to explore the burden of injuries in Nepal, funded by the United Kingdom National Institute for Health and Care Research. Objectives To explore the utility of Health Management Information System data to understand the burden of injuries in Nepal, inequalities by age and sex, and changes over time. Design Secondary analysis of published data. Setting Nepal. Data sources We used published national HMIS data on hospitalised injuries in Nepal, between 2009/10 and 2016/17, classified using International Classification of Disease codes. We grouped codes to report data by injury type, using the Global Burden of Disease injury classification framework where possible. We calculated crude rates of total unintentional injuries and self-harm, and crude rates by year for each type of unintentional injury, with correlation coefficients to describe any trends over time. Results The trend in crude unintentional injury rate increased over time. Road traffic injury admissions increased from 4.28/100,000 (95% confidence interval 4.03 to 4.52) of the population in 2009/10 to 10.55/100,000 (95% confidence interval 10.17 to 10.92) in 2016/17 (r = 0.93), while admissions following poisoning almost halved over the same period, from 7.52/100,000 (95% confidence interval 7.19 to 7.84) to 3.62/100,000 (95% confidence interval 3.40 to 3.84) (r = –0.87). Inequalities by age and gender were noted: during the period 2014/15–2016/17, admissions following road traffic injury most commonly affected adults of working age (13.82/100,000; 95% confidence interval 13.50 to 14.14) and were 1.78 times more common in men (13.63/100,000; 95% confidence interval 13.27 to 13.99) than women (7.77/100,000; 95% confidence interval 7.49 to 8.05). Limitations The coding and completeness of the injury data currently limit the utility of their use for monitoring and decision-making. Conclusions The cause of injury admissions between 2009/10 and 2016/17 appears to have shifted over time, with trends varying by injury type. In the absence of an injury surveillance system, routine inpatient data collected through the HMIS has the potential to inform policy and practice. Future work Support to enhance the completeness of data collection, and accuracy and consistency of data coding has the potential to enhance the utility of this existing data system. Funding This article presents independent research funded by the National Institute for Health and Care Research (NIHR) Global Health Research programme as award number 16/137/49.