Completeness of Police Records for Injury Surveillance: A Systematic Review

Background: Iinjuries are an emerging global health epidemic. They claim more than 16,000 lives across the world daily and account for almost 50% of deaths in people aged 10 to 24 years. The burden of injuries is likely to grow in the coming years. Reliable data that would help in the analysis of injury problems are lacking in many countries. Police records are a potentially important source of data on injuries. Objectives: to summarise and present the worldwide literature on the proportions of injuries ascertained by police records; and the extent to which police records are utilised in injury surveillance systems. Methods: Peer reviewed and grey literature published between January 2000 to July 2020 in English language on ascertainment of unintentional injuries by police records, or on use of police records for injury surveillance was included. Databases searched were Medline, EMBASE, PubMed, Google Scholar, Cochrane database and the Database of Abstracts of Reviews of Effectiveness (DARE). Reference lists of eligible studies were also scanned. The review was not registered and a review protocol was not published. Results: Out of 754 unique studies identified, 712 were excluded after screening of the abstract and 10 were excluded after reading the full text. Four studies were included after screening the reference lists of eligible studies. Ascertainment of fatal injuries by police records was reported between 35% to 96.6% in High-Income-Countries (HICs) 46.4% to 56.2% in Upper-Middle-Income-Countries (UMICs), 4.2% to 77.8% in Lower-Middle-Income- Countries (LMICs) and 46.3% to 58% in Low-Income-Countries (LICs). Ascertainment of nonfatal injuries was reported between 16% to 82% in HICs, 4.7% to 51.5% in UMICs, 6.7% to 24.7% in LMICs and 17% in LICs. Police records are used for injury surveillance either alone or in combination with other data sources. Conclusion: Police records are a potentially useful source of information on unintentional injuries. Ascertainment was found to be higher for fatal injuries and in HICs as compared to LICs and LMICs. However, the use of police records as the basis of unintentional injury surveillance systems is presently at a nascent stage. Better enforcement of legal provisions regarding reporting of injuries to the police and increasing awareness is likely to improve the proportions of ascertainment.