This study shows that injury-related mortality is high in urban Uganda and that injuries result in 25% of all recorded deaths in Kampala. The odds of dying due to injury are five times higher in Kampala than in San Francisco or the United States. Road traffic injuries accounted for 46% of injury-related deaths in Kampala, and injury to the head and neck was a commonly implicated cause of death.
Our study’s injury mortality rate of 6.7/10,000 is much lower than the 22/10,000 previously reported by community surveys in urban Kampala [20
]. One reason for this difference may be overestimation in the previous study due to recall or interviewer bias. In addition, the previous study was conducted in only one political district in Kampala and may not have been representative of the entire city. Our study included deaths recorded in all of Kampala except any undocumented burials or deaths not reported to the Kamala City Council Mortuary by the private and mission hospitals in the city. Moreover, our study was not affected by recall or interviewer bias. Thus, our results may more accurately describe the overall urban injury mortality in Kampala.
Our findings can be compared with injury-related mortality reports from other countries in Africa. In Kumasi, Ghana, injuries were implicated in 8.6% of all deaths recorded in 1996 [11
]. However, this estimate may have been low as the study included only one mortuary in the city; immediate burials without notification of medical facilities, mortuaries, or the country’s Birth and Death Registry may have occurred. Additionally, our much higher injury mortality rate in Kampala may be due to rapid increases in the burden of injuries and urban population density in Uganda since 1996. Between 1985 and 1995 alone, the number of licensed motor vehicles in Uganda increased by 220%, and the city’s population increased by 4–5% per year [15
]. These rapid changes may be contributing to a much greater role of injury as a cause of death in Kampala. In fact, more recent hospital-based analyses of injury mortality from Nigeria have reported findings similar to ours [19
]. Furthermore, as many as 81% of injury-related deaths in Kumasi have been reported to occur in the prehospital setting [11
], unlike our findings in Kampala, where 53% of all injury deaths took place outside the hospital. This difference may be explained by differences in types of injury, the mechanism or injury severity seen in the two settings, or the on-scene care provided by an informal network of first responders in Kampala [21
Although studies on emergency department visits for injury have been conducted in Kampala [17
] ours is the first study to describe the injury-related mortality patterns in detail using a new, comprehensive death statistics database. It also provides a comparison with the more organized trauma system in San Francisco. Our findings raise several important concerns about the etiology of injury in this setting, the availability of appropriate prevention measures, prehospital and trauma care, and the need for organized data collection. In addition, our study provides further evidence of the tremendously high burden of injury in a developing country.
First, the high prevalence of road traffic crashes and injury to the head and neck in Kampala must be explored further to generate specific policy responses. Improving road safety involves a multisectoral approach with buy-in from law enforcement agencies, road users, and city planners. Traffic safety enforcement has been suggested as a cost-effective method of addressing the injury epidemic in Kampala [23
]. The Kampala Police Department and the Injury Control Center, Uganda have led ongoing advocacy for helmet use by motorcycle drivers and passengers. Increased rates of helmet use may be a valuable component when addressing the high injury-related mortality in Kampala, particularly the high proportion of traumatic deaths that include injury to the head and neck.
Second, the lack of a formal emergency medical system in Kampala is an important issue. Immediate prehospital care by trained first responders has been shown to save lives in other settings [22
]. Adequate airway protection and hemorrhage control alone could potentially salvage many of those with head injury who may otherwise die in the prehospital setting. Mechanisms to formalize prehospital care in Kampala, including training lay first responders in basic prehospital trauma care, are being explored and may be one approach to addressing injury-related mortality [21
]. Furthermore, improving hospital-based trauma care by implementing context-specific trauma training programs for nurses and doctors may ensure that injury victims who reach the hospital receive fast and appropriate care. Such programs have reduced injury mortality in other settings [24
]. Because 47% of injury-related deaths in Kampala (n
= 347) occurred at the hospital level, compared to only 21% in San Francisco (n
= 74), optimizing in-hospital trauma care may be a key area for policy change to reduce the burden of injury in Kampala. Additionally, engaging in such capacity-building interventions for trauma and emergency surgical services across the spectrum of care in Kampala may improve medical services for other emergency conditions. Thus, these programs could strengthen the health system broadly compared to other vertical, disease-specific programs.
Third, another critical issue raised by our study is the need for accurate injury-related vital statistics. One to two million deaths occurring annually in low- and middle-income countries are potentially avertable with better trauma systems [9
]. It is unclear how many injury-related deaths in Kampala are avertable today; a prospective vital statistics database would help track injury deaths and determine the deaths that are avoidable. Resource-rich countries have begun to understand the importance of investing in injury prevention, relative to other causes; and there is increasing evidence that such investments are even more necessary in low- and middle-income countries such as Uganda. A vital statistics database could also help evaluate essential interventions for injury prevention and trauma care. For example, the WHO and the World Bank have recommended a lay first-responder system in settings where formal emergency services do not exist. Preliminary results of a pilot first-responder program in Kampala have been reported elsewhere [21
]. However, the impact of such a program on injury-related mortality can only be evaluated with adequate infrastructure to capture vital statistics.
Our study has several limitations. First, we may not have captured all injury deaths in urban Kampala because although Mulago Hospital serves as the sole 24-hour public emergency department in the city and sees 75% of all injury victims in Kampala [17
] deaths occurring either at home or en route to any of the city’s other hospitals may not have been recorded accurately. Although we assumed that most of these deaths would have been reported to the Kampala City Council Mortuary, some of these deaths may have been missed.
Second, the three data sources in Kampala collect data based on the occurrence of death within the city and not specifically the residence of the decedent. Residents who died outside the city would not be captured by the sources. Likewise, nonresidents of Kampala may be incorrectly assigned a local address. Thus, the records may have overestimated or underestimated deaths of Kampala residents. We assumed that these figures are either negligible or balance each other out. However there can be considerable movement of people between Kampala and the rest of country at the end of the year, when families travel for the holiday season, leading to an increase in the incidence of injuries, although no seasonal variations in injury deaths were apparent in our data. The estimates of the population at risk may also be inaccurate, although this issue exists in all registries. Finally, data quality may be an issue. For example, the records of deaths from medical causes may underestimate the contributions of infectious disease because of diagnostic challenges in the postmortem setting.
Third, the validity of comparing Kampala and San Francisco depends on intrinsic comparability of both cities. Although we accounted for differences in demographics through direct standardization, each city varies from the other in regard to cultural norms, road infrastructure, and traffic safety regulations. These issues may influence injury-related mortality rates in different ways in each setting, making this comparison less valid. For example, prehospital injury-related deaths occur at a higher rate in San Francisco than in Kampala possibly due to the high incidence of assault, particularly homicide, which may be resulting in severe nonsurvivable penetrating injury in the U.S. setting. Therefore, controlling for injury severity may improve the accuracy of our findings. Nevertheless, the two cities may have comparable injury patterns based on key similarities such as high pedestrian traffic, reliance on public transportation, and provision of definitive trauma care at a single city hospital.
Despite these limitations, our study contributes to a better understanding of the burden of injuries in a resource-constrained setting. Many other issues related to injury still have to be addressed. The disability associated with injury is a substantial burden on society [2
] and needs to be further explored in this setting. An economic evaluation of the high injury-related mortality rate in Uganda would also add to a greater understanding of the true societal cost of this epidemic and better inform resource allocation.