1 Walton M, et al. Inj Prev 2017;0:1–8. doi:10.1136/injuryprev-2017-042341 ABSTRACT Background Vietnam has prioritised the establishment of a civil registration system for deaths but as yet is unable to report accurate national statistics for the population of 93.5 million people due to inadequate mortality data. Verbal autopsy data suggest that injury is a third leading cause of death (by International Statistical Classiication of Diseases and Related Health Problems 10th Revision chapters) in Vietnam, with road trafic injuries in particular a signiicant and increasing challenge. The study aims to present a more accurate proile of the number and probable causes of these deaths based on data collected hospitals using a version of the WHO death certiicate modiied for the Vietnamese context. Methods Death data collected from Viet Duc Surgical and Trauma Hospital in Vietnam between 1 March 2013 to 31 March 2015 was analysed to explore the number and probable causes of deaths for deaths resulting from an injury. Results A total of 1616 deaths were recorded for Viet Duc Hospital, of which 73% (1181/1616) were associated with an injury. Most (83%; n=871/1049) injury-related deaths for which immediate cause of death was documented were as a result of head/brain injuries. Injury-related deaths were most commonly caused by from trafic accidents (72%, 853/1181). The majority of patients suffering injury-related deaths were discharged home to die (93%, 1097/1181). Conclusion The study conirms some indings from previous studies about deaths from injuries, while disagreeing with others, highlighting the challenge for Vietnam in collecting these data. Gathering detailed death data provides essential evidence on which to base decisions about allocation of government funding and policy for injury prevention and treatment. INTRODUCTION Vietnam has prioritised the establishment of a civil registration system for deaths but as yet is unable to report accurate national statistics for the population of 93.5 million people due to inadequate mortality data. The Vietnam Ministry of Justice is responsible for collecting mortality data and storing it within the Vital Registration and Primary Vital record systems. 1 Mortality data collection Verbal autopsy (VA) is widely used in many low-in- come countries to record probable causes of death. The process involves interviewing family members and witnesses. The majority of VA reports are not completed by doctors or medically trained people resulting in concerns about the accuracy of VA data. 2–4 The General Statistics Office of Vietnam currently publishes VA reports via the A6 Mortality Reporting System to identify the number of deaths and their causes. 5 Limited data are also available from three mortality databases for children. 6 Basic death data are routinely collected in Vietnamese hospitals using the same A6 form (called the Death Notification Form in the hospitals), but this form rarely includes clinical information or an accu- rate description of the primary cause of death or underlying conditions. 7 Given that most deaths in Vietnam occur outside medical facilities, VA remains the best available approach for assessing deaths in the home. 8 In 2001, the Ministry of Health requested medical practitioners to use ICD-10 codes (International Statistical Classification of Diseases and Related Health Problems 10thRevision) when recording deaths. 9 However, a 2015 study of hospital deaths 7 and an earlier 2009 study of the quality of hospital medical records demonstrated doctors experienced significant challenges recording accurate death data. Both studies identified reasons provided for poor reporting of ICD-10 data included: the unavailability of a translated ICD-10 index into Vietnamese, referral to only the summary sheet in the medical record when completing the coding, the need for training in ICD-10 coding and data abstraction. 7 10 The 2009 study found 90% of the records of injury-related deaths recorded little information about the extent of the injury, 85% recorded no indication of the activity that led to the injury and 67% lacked information about where the injury geographically occurred. 10 Although, the 2015 pilot study in two national hospitals found that basic patient identification and death informa- tion were fully and consistently completed in the piloted death report form. Even so, doctors found the complex identification of underlying causes of death and ICD-10 coding challenging. 7 Injury-related deaths Injury is a third leading cause of death (by ICD-10 chapters) in Vietnam, with road traffic injuries in particular a significant and increasing challenge. A 2012 study of 5273 deaths in Vietnam recorded by the A6 Mortality Reporting System (reliant on VA reports) found 14% of deaths were caused by inju- ries (n=763/5273), with an overall injury mortality rate of 55.3 persons per 100 000 person-years. 11 A study of 2800 households in seven districts in Hanoi Original article Improving hospital death certiication in Vietnam: results of a pilot study of injury-related fatalities Merrilyn Walton, 1 Reema Harrison, 2 Anna Chevalier, 3 Esmond Esguerra, 4 Nguyen Duc Chinh, 5 Haphan Haian, 6 Dang Van Duong, 6 Huong Giang 6 To cite: Walton M, Harrison R, Chevalier A, et al.Inj Prev Published Online First: [please include Day Month Year]. doi:10.1136/ injuryprev-2017-042341 1 School of Public Health, University of Sydney, Sydney, New South Wales, Australia 2 School of Public Health and Community Medicine, University of New South Wales, Sydney, New South Wales, Australia 3 Safer Roads Consulting, Thirroul, New South Wales, Australia 4 University of Sydney, Ofice for Global Health, Sydney, New South Wales, Australia 5 Benh Vien Viet Duc, Hanoi, Vietnam 6 Bach Mai Hospital, Hanoi, Vietnam Correspondence to Dr Reema Harrison, School of Public Health and Community Medicine, University of New South Wales, Sydney, New South Wales 2052, Australia; reema. harrison@unsw.edu.au Received 23 January 2017 Revised 26 April 2017 Accepted 12 June 2017 IP Online First, published on August 1, 2017 as 10.1136/injuryprev-2017-042341 Copyright Article author (or their employer) 2017. Produced by BMJ Publishing Group Ltd under licence. group.bmj.com on August 1, 2017 - Published by http://injuryprevention.bmj.com/ Downloaded from