You are here

Article Text

Article menu
Original research
Estimates of road traffic deaths in Tanzania
  1. Leah Watetu Mbugua1,
  2. Sudeshna Mitra1,
  3. Kazuyuki Neki1,
  4. Hialy Gutierrez2,
  5. Ramshankar Balasubramaniyan2,
  6. Mercer Winer2,
  7. Jaeda Roberts2,
  8. Theo Vos3,
  9. Erin Hamilton3,
  10. Mohsen Naghavi3,
  11. James E Harrison4,
  12. Soames Job1,
  13. Kavi Bhalla2
  1. 1 Global Road Safety Facility, World Bank, Washington, DC, USA
  2. 2 Public Health Sciences, University of Chicago, Chicago, Illinois, USA
  3. 3 Institute for Health Metrics and Evaluation, University of Washington, Seattle, Washington, USA
  4. 4 Flinders University, Bedford Park, South Australia, Australia
  1. Correspondence to Dr Kavi Bhalla, Public Health Sciences, University of Chicago, Chicago, Illinois, USA; kavibhalla{at}gmail.com

Abstract

Introduction There is considerable uncertainty in estimates of traffic deaths in many sub-Saharan African countries, with the Global Burden of Disease (GBD) and the Global Status Report on Road Safety (GSRRS) reporting widely differing estimates. As a case study, we reviewed and compared estimates for Tanzania.

Methods We estimated the incidence of traffic deaths and vehicle ownership in Tanzania from nationally representative surveys. We compared findings with GBD and GSRRS estimates.

Results Traffic death estimates based on the 2012 census (9382 deaths; 95% CI: 7565 to 11 199) and the 2011–2014 Sample Vital Registration with Verbal Autopsy (8778; 95% CI: 7631 to 9925) were consistent with each other and were about halfway between GBD (5 608; 95% UI: 4506 to 7014) and WHO (16 252; 95% CI: 13 130 to 19 374) estimates and more than twice official statistics (3885 deaths in 2013). Surveys and vehicle registrations data show that motorcycles have increased rapidly since 2007 and now comprise 66% of vehicles. However, these trends are not reflected in GBD estimates of motorcycles in the country, likely resulting in an underestimation of motorcyclist deaths.

Conclusion Reducing discrepancies between GBD and GSRRS estimates and demonstrating consistency with local epidemiological data will increase the legitimacy of such estimates among national stakeholders. GBD, which is the only project that models the road-user distribution of traffic deaths in all countries, likely severely underestimates motorcycle deaths in countries where there has been a recent increase in motorcycles. Addressing police under-reporting and strengthening surveillance capacity in Tanzania will allow a better understanding of the road safety problem and better targeting of interventions.

  • Motor vehicle - Non traffic
  • Motor vehicle - Occupant
  • Motorcycle

Data availability statement

The study analysed secondary data collected by other parties described in the manuscript. These data are not publicly available but may be obtained by contacting these other parties.

http://dx.doi.org/10.1136/injuryprev-2022-044555

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Despite extensive global advocacy, the first UN Decade of Action on Road Safety ended in 2020 without a marked increase in evidence-based road safety programmes in most low-income and middle-income countries (LMICs), especially in the sub-Saharan African region.1 2 One important reason for the low priority of traffic injuries in the national policy agenda of LMICs is that the scale of the problem is underestimated by country governments.3 Therefore, in recent years, researchers and international agencies have sought to highlight under-reporting in official statistics of the incidence of road traffic injuries.3–5 This has primarily been done by comparing official country reports with estimates based on available health-sector data produced by the Institute for Health Metrics and Evaluation’s (IHME) Global Burden of Disease (GBD) Study and the WHO’s Global Health Estimates (GHE), whose road traffic injury estimates are published in the Global Status Report on Road Safety (GSRRS).6 7 However, in this dialogue, the issue of the reliability of the GBD and GSRRS estimates has received little attention.

    For instance, consider the wide discrepancy in estimates for countries in GBD’s Eastern sub-Saharan Africa region. The GSRRS estimate for road traffic deaths in the region (26.9 per 100 000) is more than twice GBD-2019 estimates (11.6 per 100 000 population).6 7 The discrepancy is even more pronounced in country-level estimates. In Ethiopia, Kenya and Tanzania—the three most populous countries in the region—WHO estimates are 3.1, 3.0 and 2.8 times GBD estimates, respectively.6 7 Although GBD and GSRRS estimates are invariably higher than official statistics, such discrepancies make it easy for national stakeholders to dismiss GHE as unreliable, sidestepping a public debate about under-reporting in official statistics and the extent of the road traffic injury problem. In the ideal scenario, reliable statistical estimates of road traffic mortality would provide benchmarks for strengthening traffic injury surveillance, which is essential for understanding risk factors and developing effective solutions.

    We have recently shown that it is increasingly common for national health surveys and censuses in many LMICs to include questions related to the incidence of road traffic injuries.3 Our analysis of traffic deaths measured in two censuses and two intercensal household surveys in Cambodia showed internally consistent results. Verbal autopsy surveys are a validated tool for estimating causes of death that performs particularly well for estimating deaths from injuries.8–10 Therefore, nationally representative verbal autopsy surveys are used in many countries (including India,11 China12 and Thailand13) for estimating cause-specific mortality. When these data sources are not included as inputs in GBD and GSRRS analyses,3 they provide an opportunity for external validation of estimates. In addition, comparisons with information from epidemiological data sources collected by local research institutions are more likely to create a productive dialogue within the country about the nature of under-reporting and how it should be addressed.

    Therefore, as a case study, we sought to systematically review all sources of epidemiological information on road traffic deaths and injuries in Tanzania and compare estimates from various sources. We discuss the implications for estimating the true death toll in the country and elsewhere and efforts to improve national statistics. This study parallels a similar case study for Cambodia.14

    Methods

    We searched for nationally representative data sources from Tanzania, focusing on household surveys and population censuses that included questions that allow estimating the incidence of road traffic deaths and non-fatal injuries. Additionally, we searched for sources that allow estimating household ownership of bicycles, motorcycles, cars and other vehicles, because vehicle ownership is an important covariate of traffic injuries and are used in modelling studies like GBD to improve estimates of deaths for different types of road users in countries where primary data sources on road traffic injuries are sparse or unavailable.

    In particular, we conducted the following searches: We searched PubMed and Google Scholar for research publications using the following search strategy: (((traffic injuries)) AND (Tanzania)) AND ((“1990”[Date - Publication]: “3000”[Date - Publication])) and reviewed articles where a title and abstract review suggested the use of nationally representative data sources. We then conducted snowball searches based on the articles retrieved. We searched for surveys and censuses from Tanzania in the Integrated Public Use Microdata Series database,15 the International Household Survey Network16 and Monitoring and Evaluation to Assess and Use Results Demographic and Health Surveys.17 We reviewed the questionnaires of surveys and censuses to identify questions on road traffic injuries and vehicle ownership. We searched government websites for road safety statistics, including reports from the Tanzania Police Force,18 and information reported by Tanzania’s national road traffic injury surveillance system, Road Accident Information System (RAIS).19 Finally, we reviewed IHME’s Global Health Data Exchange and the GBD-2019 data sources used for Tanzania.20

    We acquired microdata of the sources identified wherever possible and estimated the incidence of road traffic injuries and household ownership of bicycles and motor vehicles. Where microdata were not available, estimates reported in secondary sources (eg, published tabulations) were extracted.

    We compared our estimates with those produced by various revisions of GBD and GSRRS. Publications from these projects21–23 provide details of estimation methods. Note that each revision of GBD involves a fresh analysis of all historic data from all countries and results in new time series of national estimates. GBD estimates between revisions are different because the historic pool of global data grows as new data sources are identified and included and because of changes to estimation methods. While the most recent revision (GBD-2019) provides the best current GBD estimate, we have included prior revisions because the evolving estimates provide an additional indication of the level of uncertainty in estimates. Only GBD-2017 and GBD-2019 estimates are currently available online. For earlier revisions, we used data that we had previously downloaded. For GBD-2010 and GBD-2015, we only had access to the estimate for the most recent year.

    Patient and public involvement

    Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

    Results

    Data sources identified

    The literature review yielded 25 publications with potentially relevant road injury data (online supplemental table A1). Based on a review of the title, abstract and full text, we only identified one study, Levira et al,24 that provided a nationally representative data estimate of causes of death in Tanzania using verbal autopsy.

    We identified 27 nationally representative censuses and surveys that asked questions potentially relevant for estimating the incidence of road traffic mortality in Tanzania (3 sources) and the household ownership of vehicles (24 sources). We did not find any nationally representative data sources that could be used to estimate the incidence of non-fatal traffic injuries. Online supplemental table A2 provides details of all sources identified.

    Among the sources that collected information on road traffic deaths (table 1) was the 2012 national Population and Housing Census, where 30% of the enumeration areas received a long questionnaire that included a section on deaths in the household. This module asked about the age, sex and cause (including an option for ‘Road Accident’) for all deaths in the household. We obtained microdata for the census and computed age-specific and sex-specific proportions of all-cause deaths that were due to road traffic injuries after weighting for sampling probability. We applied these proportions to the age-specific and sex-specific estimates of all-cause deaths in Tanzania in 2012 estimated by GBD.

    View this table:
    Table 1

    Nationally representative surveys and census with questions pertaining to the incidence of road traffic injuries in Tanzania

    The Sample Vital Registration with Verbal Autopsy (SAVVY) Study collected information about mortality from a nationally representative sample of 155 000 households between 2011 and 2014. Causes of death were determined based on verbal autopsy interviews with next of kin or other caregivers. Microdata from the study were not available to us but estimates of road traffic deaths from SAVVY were reported by Levira et al 24 who estimated that 2.43% (95% CI: 2.01 to 2.84) of all-cause deaths were due to road traffic injuries in 2016. We applied these proportions to GBD estimate of all-cause deaths.

    Finally, one source, the 2014 Global Student Health Survey, only collected information on injuries among students aged 13–17 years and was excluded from further analysis. Our review of data sources used by GBD-2019 and GSRRS showed that neither uses the data sources we have identified for estimating mortality.

    From among the 24 data sources with questions on household ownership of vehicles, we were able to access and extract estimates from 12 sources. Among these, we excluded the Southern and Eastern African Consortium for Monitoring Educational Quality surveys because they collected data at schools and were not representative of the national population.

    Comparisons of estimates of road traffic deaths

    Figure 1 shows a large discrepancy between GBD and GSRRS estimates for traffic deaths in Tanzania. For the year 2016, GBD-2019 estimated 5608 deaths (95% UI: 4506 to 7014) but GSRRS-2018 estimates were almost three times higher (16 252 deaths; 95% CI: 13 130 to 19 374). This discrepancy has existed in all previous revisions, with GSRRS estimates always being substantially higher than GBD. Our estimates based on analysis of the 2012 census (9382 deaths; 95% CI: 7565 to 11 199) and the 2011–2014 SAVVY verbal autopsy (8778 deaths; 95% CI: 7631 to 9925) were consistent with each other and were about halfway between GBD and WHO estimates. Notably, the estimates based on the census and the SAVVY verbal autopsy are more than twice the deaths reported in official statistics (3885 deaths in 2013).

    Figure 1
    Figure 1

    Comparison of various sources of estimates of road traffic deaths in Tanzania. GBD, Global Burden of Disease; GSRRS, Global Status Report on Road Safety; SAVVY, Sample Vital Registration with Verbal Autopsy.

    Figure 2 describes trends in household ownership of bicycles and motor vehicles in the country. Ownership levels of all vehicles have increased in time. Notably, the number of households that own a motorcycle has increased by almost 10 times since around 2007, such that household motorcycle ownership now substantially exceeds car ownership. The dramatic increase in motorcycle ownership is also evident in motor vehicle registration data shown in figure 3. However, notably, GBD’s estimate of the motorcycle ownership trend in the country does not track the large increase in registered motorcycles (figure 3) and household ownership of motorcycles (figure 2). The number of registered motorcycles in Tanzania at present is 13 times GBD’s estimate.

    Figure 2
    Figure 2

    Percentage of households that own a bicycle, motorised two wheeler or private car. AIS, AIDS Indicator Survey; DHS, demographic and health surveys; GAS, Pew Global Attitudes Survey; MICS, Multiple Indicator Cluster Survey.

    Figure 3
    Figure 3

    Comparison of GBD-2019 estimates of registered motor vehicles and official motor vehicle registration statistics. GBD: Global Burden of Disease.

    Figure 4 compares GBD estimates of the proportions of traffic fatalities by their mode of transport with official statistics. The underestimation of motorcycle ownership by GBD is reflected in a much lower estimate of motorcyclist deaths in the country. According to GBD-2019, motorcyclists account for only 7% of road traffic deaths as opposed to 23% in official statistics.

    Figure 4
    Figure 4

    GBD-2019 estimates of types of road users killed in 2016 compared with official statistics as reported in the 2018 GSRRS report. GBD: Global Burden of Disease.

    Discussion

    Uncertainty in global health models of road traffic deaths

    There is a large discrepancy in the estimates of road traffic deaths in Tanzania produced by global health modelling efforts. WHO (GHE/GSRRS) estimates are almost three times those reported by IHME (GBD-2019). Such discrepancies are common in many LMICs, especially for countries where the two projects have not included local epidemiological sources of information on traffic deaths. In this regard, the case of Tanzania is instructive. In general, both GBD and GHE/GSRRS estimate road traffic deaths based on covariates that describe the country’s income level, motorisation levels, health system, demographic characteristics and transport system. However, the modelling details differ in important ways. WHO’s models are calibrated only on countries with high-quality national vital registration data but GBD incorporates many more sources (such as health surveys, verbal autopsies and data representative at the subnational level). Such differences often result in fairly large discrepancies between country-level estimates. In fact, changes in the modelling strategies between the various revisions of GBD alone result in large changes in estimates.14

    In the case of Tanzania, our review identified two nationally representative studies (population census and a national verbal autopsy) that provide estimates of road traffic deaths at the national level. The two sources provide estimates that are fairly consistent with each other but inconsistent with both GBD and WHO estimates. Notably, neither of these data sources was used in GSRRS or GBD models.7 20 GBD-2019 included results from five verbal autopsy studies from the country.20 Only one of these, the Adult Morbidity and Mortality Project (AMMP, 1992–2003), included both urban and rural populations.25 Notably, publications from AMMP26 estimated transport injury death rates at approximately twice those estimated by GBD-2019. The remaining GBD-2019 data sources from Tanzania are from rural settings (Muleba, Rufiji, Korogwe and Ifakara), where population exposure to road traffic is lower and road traffic death rates are likely to be much lower than in urban settings. Thus, it is likely that GBD underestimates traffic deaths in Tanzania. The inclusion of data from the 2012 census and the SAVVY verbal autopsy in future GBD revisions may help improve estimates. In contrast to GBD, WHO’s mortality models are only able to incorporate data from relatively complete vital registers which are not available in most African countries. It is important that future GSRRS estimates include the types of data sources (eg, household health surveys, verbal autopsies, census mortality modules) that are often available in the region.3

    Between the two global health estimation projects, only GBD estimates traffic deaths disaggregated by victims’ mode of transport. While GSRRS does report regional and global distributions of the types of victims killed, these results rely on the proportions reported in official statistics and are based on an unstated assumption that official statistics provide an unbiased estimate. However, this is a problematic assumption. There is substantial evidence that police under-reporting is affected by such factors as the location of the event (eg, urban/rural) and the income, age and gender of victims, all of which are associated with the victim’s mode of transport.27–29 Especially in countries with large under-reporting (including Tanzania and most LMICs), we should expect that reported crashes are severely biased. Therefore, we recommend that the GSRRS models explicitly model the road user distribution rather than rely on official statistics.

    Unlike GSRRS, however, GBD includes separate statistical models for estimating deaths for each road-user type. However, our analysis suggests that GBD likely severely underestimates the proportion of motorcycle deaths in Tanzania. National motor-vehicle registration data show a rapid and large increase in the number of motorcycles in the country starting around 2007, which is also reflected in estimates of the proportions of households that own a motorcycle, such that by 2019 motorcycles account for 60% of all motor vehicles in the country. Therefore, GBD’s estimate that only 7% of traffic deaths in the country are motorcyclists appears to be implausibly low. Significantly, GBD-2019’s data sources for Tanzania did not include any that disaggregated road traffic deaths by road-user type. Therefore, GBD’s estimates for motorcyclist deaths are primarily determined by covariates, among which the number of motorcycles per capita in the country is likely the most important variable. Our results show that GBD’s estimate of the trend in the number of motorcycles in Tanzania does not track the large increases evident in vehicle registration data (figure 3). This is likely because GBD-2019’s estimates of vehicles were based on the International Road Federation 2009 World Road Statistics database, which did not include data beyond 2007. Therefore, GBD’s estimates of vehicles for the period beyond 2007 are extrapolated from older data that predates the rapid growth in the motorcycle fleet in Tanzania.

    Although the true road-user distribution of traffic deaths in Tanzania is unknown, hospital-based studies suggest that the proportion that is motorcyclists is likely much higher than that reported by GBD (7%) and also higher than official statistics (23%). In seven studies that reported road-user distributions of hospital visits in Mwanza city, Dar es Salaam and Morogo regions, the proportion of motorcyclists ranged from 37% to 87% (see Supplementary Appendix Table A3 for details).30–36

    A similarly large and recent increase in motorcycles has happened in many African countries, and it is likely that GBD underestimates motorcyclist deaths in other countries as well. This is particularly concerning because motorcycles impose a much higher injury risk on users than other vehicles37 and represent a growing threat to safety in Africa that is not yet fully appreciated.

    Implications for road traffic injury surveillance in Tanzania

    Our study focuses on the need to improve statistical estimates of road traffic injuries in Tanzania by using reliable local epidemiological data sources. Such work is important because it can help country stakeholders acknowledge the true magnitude of the problem and prioritise road safety appropriately in the policy agenda. Our analysis makes clear that road traffic deaths are severely under-reported in official statistics in the country. The national census and the SAVVY survey provide a consistent estimate of the true road death toll (approximately 9000 deaths), which is 2.4 times the officially reported figure. This higher estimate would place road traffic injuries among the top-10 causes in GBD-2019’s cause-of-death ranking for Tanzania. The official statistic would place road traffic deaths much lower at rank 22.

    However, it is important to note that such survey-based measurements and statistical estimates (eg, GBD, GSRRS) cannot provide all the information needed for running effective road safety programmes. The Safe System approach recommended by the WHO and the World Bank requires reliable, timely and detailed data on crash circumstances and risk factors. The approach involves developing a road safety strategy based on assessing population-level risks and allocating resources to the most cost-effective interventions. Surveillance data are needed for setting meaningful targets for final outcomes (ie, road traffic deaths and injuries), intermediate outcomes (eg, helmet and seatbelt use) and the institutional outputs (eg, enforcement levels) needed to achieve these outcomes. Finally, data are needed for monitoring the performance of the safety programme and to allow recalibration of efforts on an ongoing basis.

    Fulfilling these information needs will require Tanzania to make a major investment in developing the capacity of its police to do routine surveillance of traffic injuries and key risk factors. Some efforts to modernise the reporting infrastructure are already underway, notably including the development of the RAIS through World Bank support. RAIS aims to establish infrastructure for recording and reporting crash information using standardised forms designed for the purpose of managing road safety. Periodic survey-based measurements of traffic deaths and injuries can provide an external benchmark to assess the completeness of statistics produced by RAIS and monitor progress towards building institutional capacity for traffic injury surveillance.

    What is already known on the subject?

    • Global health estimates from the Global Burden of Disease (GBD) and the Global Status Report on Road Safety (GSRRS) report widely differing estimates for many countries in sub-Saharan Africa.

    • Some countries have conducted surveys and censuses that allow estimating the incidence of road traffic injuries but are not included as input sources in GBD and GSRRS models.

    What this study adds?

    • In Tanzania, estimates from the national census and a verbal autopsy survey suggest that there are approximately 9000 road traffic deaths annually, halfway between GBD and GSRRS estimates but twice the number reported in official statistics. Including local epidemiological data sources can help reduce uncertainty and discrepancies in estimates of traffic injuries produced by GBD and GSRRS.

    • GBD is the only global health study that estimates the road-user distribution of traffic deaths but likely underestimates motorcyclist deaths in Tanzania and other countries that have had a recent increase in motorcycle use.

    • Strengthening the capacity of national police to do surveillance of traffic injuries is vital for running evidence-based road safety programmes.

    Data availability statement

    The study analysed secondary data collected by other parties described in the manuscript. These data are not publicly available but may be obtained by contacting these other parties.

    Ethics statements

    Patient consent for publication

    Ethics approval

    Not applicable.

    References

    1. ↵
      1. Third Global Ministerial Conference on Road Safety
      . Stockholm declaration: achieving global goals 2030, 2020. Available: https://www.roadsafetysweden.com/about-the-conference/stockholm-declaration/
    2. ↵
      2. Bonnet E ,
      3. Nikiéma A ,
      4. Adoléhoume A , et al
      . Better data for better action: rethinking road injury data in francophone West Africa. BMJ Glob Health 2020;5:e002521.doi:10.1136/bmjgh-2020-002521 pmid:http://www.ncbi.nlm.nih.gov/pubmed/32371569
      OpenUrlFREE Full Text
    3. ↵
      2. Mitra S ,
      3. Neki K ,
      4. Mbugua LW , et al
      . Availability of population-level data sources for tracking the incidence of deaths and injuries from road traffic crashes in low-income and middle-income countries. BMJ Glob Health 2021;6:e007296.doi:10.1136/bmjgh-2021-007296 pmid:http://www.ncbi.nlm.nih.gov/pubmed/34782357
      OpenUrlAbstract/FREE Full Text
    4. ↵
      2. Bhalla K ,
      3. Shotten M ,
      4. Cohen A
      . Transport for health: the global burden of disease due to injuries and air pollution from motorized road transport. Washington, DC: World Bank Global Road Safety Facility, and Institute for Health Metrics and Evaluation, 2014. http://www.healthmetricsandevaluation.org/gbd/publications/policy-report/transport-health-global-burden-disease-motorized-road-transport
    5. ↵
      2. RFS J ,
      3. Wambulwa WM
      . Features of low-income and middle-income countries making road safety more challenging. J Road Saf 2020;31:79–84.
      OpenUrl
    6. ↵
      2. Vos T ,
      3. Lim SS ,
      4. Abbafati C , et al
      . Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the global burden of disease study 2019. The Lancet 2020;396:1204–22.doi:10.1016/S0140-6736(20)30925-9
      OpenUrlCrossRefPubMed
    7. ↵
      1. World Health Organization
      . Global status report on road safety 2018. Geneva: World Health Organization, 2018.
    8. ↵
      2. Setel PW ,
      3. Whiting DR ,
      4. Hemed Y , et al
      . Validity of verbal autopsy procedures for determining cause of death in Tanzania. Trop Med Int Health 2006;11:681–96.doi:10.1111/j.1365-3156.2006.01603.x pmid:http://www.ncbi.nlm.nih.gov/pubmed/16640621
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      2. Yang G ,
      3. Rao C ,
      4. Ma J , et al
      . Validation of verbal autopsy procedures for adult deaths in China. Int J Epidemiol 2006;35:741–8.doi:10.1093/ije/dyi181 pmid:http://www.ncbi.nlm.nih.gov/pubmed/16144861
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      2. Misganaw A ,
      3. Mariam DH ,
      4. Araya T , et al
      . Validity of verbal autopsy method to determine causes of death among adults in the urban setting of Ethiopia. BMC Med Res Methodol 2012;12:130.doi:10.1186/1471-2288-12-130 pmid:http://www.ncbi.nlm.nih.gov/pubmed/22928712
      OpenUrlPubMed
    11. ↵
      2. Gomes M ,
      3. Begum R ,
      4. Sati P , et al
      . Nationwide mortality studies to quantify causes of death: relevant lessons from India's million death study. Health Aff 2017;36:1887–95.doi:10.1377/hlthaff.2017.0635 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29137507
      OpenUrlPubMed
    12. ↵
      2. Yang G ,
      3. Hu J ,
      4. Rao KQ , et al
      . Mortality registration and surveillance in China: history, current situation and challenges. Popul Health Metr 2005;3:3: 3.doi:10.1186/1478-7954-3-3 pmid:http://www.ncbi.nlm.nih.gov/pubmed/15769298
      OpenUrlPubMed
    13. ↵
      2. Ditsuwan V ,
      3. Veerman LJ ,
      4. Barendregt JJ , et al
      . The National burden of road traffic injuries in Thailand. Popul Health Metr 2011;9:2.doi:10.1186/1478-7954-9-2 pmid:http://www.ncbi.nlm.nih.gov/pubmed/21244666
      OpenUrlCrossRefPubMed
    14. ↵
      2. Gutierrez H ,
      3. Mitra S ,
      4. Neki K , et al
      . Comparing estimates of road traffic deaths and non-fatal road traffic injuries in Cambodia. Inj Prev 2022;28:340–6.doi:10.1136/injuryprev-2021-044504 pmid:http://www.ncbi.nlm.nih.gov/pubmed/35149595
      OpenUrlAbstract/FREE Full Text
    15. ↵
      1. Minnesota Population Center
      . Integrated Public Use Microdata Series: Version 7.2 [dataset]., 2019. Available: https://www.ipums.org/
    16. ↵
      1. International Household Survey Network
      . Available: http://www.ihsn.org/
    17. ↵
      Demographic and health surveys (DHS) program. Available: https://dhsprogram.com/
    18. ↵
      1. Tanzania Police Force
      . Crime and traffic incidents statistics report: January-December 2015. Tanzania, Dar es Salaam: National Bureau of Statistics, 2016.
    19. ↵
      2. Walugembe F ,
      3. Levira F ,
      4. Ganesh B , et al
      . A retrospective study on the epidemiology and trends of road traffic accidents, fatalities and injuries in three municipalities of Dar es Salaam region, Tanzania between 2014-2018. Pan Afr Med J 2020;36:1–10.doi:10.11604/pamj.2020.36.24.21754
      OpenUrlCrossRef
    20. ↵
      Global burden of disease study 2019 (GBD 2019) data input sources tool, 2019. Available: http://ghdx.healthdata.org/gbd-2019/data-input-sources
    21. ↵
      1. India State-Level Disease Burden Initiative Road Injury Collaborators
      . Mortality due to road injuries in the states of India: the global burden of disease study 1990-2017. Lancet Public Health 2020;5:e86–98.doi:10.1016/S2468-2667(19)30246-4 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31879251
      OpenUrlPubMed
    22. ↵
      1. World Health Organization
      . WHO methods and data sources for country-level causes of death 2000-2019. Geneva: World Health Organization, 2020.
    23. ↵
      1. World Health Organization
      . Understanding and bridging the differences between country-reported and WHO-estimated road traffic fatality data. Geneva, Switzerland: World Health Organization, 2019.
    24. ↵
      2. Levira F ,
      3. Newton CR ,
      4. Masanja H , et al
      . Mortality of neurological disorders in Tanzania: analysis of baseline data from sample vital registration with verbal autopsy (SAVVY). Glob Health Action 2019;12:1596378.doi:10.1080/16549716.2019.1596378 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31144608
      OpenUrlPubMed
    25. ↵
      2. Kitange HM ,
      3. Machibya H ,
      4. Black J , et al
      . Outlook for survivors of childhood in sub-Saharan Africa: adult mortality in Tanzania. adult morbidity and mortality project. BMJ 1996;312:216–20.doi:10.1136/bmj.312.7025.216 pmid:http://www.ncbi.nlm.nih.gov/pubmed/8563587
      OpenUrlAbstract/FREE Full Text
    26. ↵
      2. Moshiro C ,
      3. Mswia R ,
      4. Alberti KG , et al
      . The importance of injury as a cause of death in sub-Saharan Africa: results of a community-based study in Tanzania. Public Health 2001;115:96–102.doi:10.1016/S0033-3506(01)00426-7 pmid:http://www.ncbi.nlm.nih.gov/pubmed/11406773
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      1. World Bank
      . Guide for road safety opportunities and challenges: low- and middle- income country profiles. Washington DC, USA: World Bank, 2020. https://www.roadsafetyfacility.org/publications/guide-road-safety-opportunities-and-challenges-low-and-middle-income-country-profiles
    28. ↵
      2. Dandona R ,
      3. Kumar GA ,
      4. Ameer MA , et al
      . Under-reporting of road traffic injuries to the police: results from two data sources in urban India. Inj Prev 2008;14:360–5.doi:10.1136/ip.2008.019638 pmid:http://www.ncbi.nlm.nih.gov/pubmed/19074240
      OpenUrlAbstract/FREE Full Text
    29. ↵
      2. Aeron-Thomas A
      . Under-reporting of road traffic casualties in low income countries. UK: Transport Research Laboratory, Berkshire, 2000.
    30. ↵
      2. Boniface R ,
      3. Museru L ,
      4. Kiloloma O , et al
      . Factors associated with road traffic injuries in Tanzania. Pan Afr Med J 2016;23:46.doi:10.11604/pamj.2016.23.46.7487 pmid:http://www.ncbi.nlm.nih.gov/pubmed/27217872
      OpenUrlPubMed
    31. ↵
      2. Chalya PL ,
      3. Mabula JB ,
      4. Ngayomela IH , et al
      . Motorcycle injuries as an emerging public health problem in Mwanza City, north-western Tanzania. Tanzan J Health Res 2010;12:214–21.pmid:http://www.ncbi.nlm.nih.gov/pubmed/24409627
      OpenUrlPubMed
    32. ↵
      2. Chalya PL ,
      3. Mabula JB ,
      4. Dass RM , et al
      . Injury characteristics and outcome of road traffic crash victims at Bugando medical centre in Northwestern Tanzania. J Trauma Manag Outcomes 2012;6:1.doi:10.1186/1752-2897-6-1
      OpenUrlPubMed
    33. ↵
      2. Chalya PL ,
      3. Dass RM ,
      4. Mchembe MD , et al
      . Citywide trauma experience in Mwanza, Tanzania: a need for urgent intervention. J Trauma Manag Outcomes 2013;7:9.doi:10.1186/1752-2897-7-9
      OpenUrlPubMed
    34. ↵
      2. Chalya PL ,
      3. Ngayomela IH ,
      4. Mabula JB , et al
      . Injury outcome among helmeted and non-helmeted motorcycle riders and passengers at a tertiary care hospital in north-western Tanzania. Tanzan J Health Res 2014;16:280–8.doi:10.4314/thrb.v16i4.4 pmid:http://www.ncbi.nlm.nih.gov/pubmed/26891517
      OpenUrlPubMed
    35. ↵
      2. Ngallaba SE ,
      3. Makerere DJ ,
      4. Kapesa A , et al
      . A retrospective study on the unseen epidemic of road traffic injuries and deaths due to accidents in Mwanza City—Tanzania. Open J Prev Med 2014;04:222–8.doi:10.4236/ojpm.2014.44028
      OpenUrl
    36. ↵
      2. Simon R ,
      3. Gilyoma JM ,
      4. Dass RM , et al
      . Paediatric injuries at Bugando medical centre in northwestern Tanzania: a prospective review of 150 cases. J Trauma Manag Outcomes 2013;7:10.doi:10.1186/1752-2897-7-10
      OpenUrl
    37. ↵
      1. World Health Organization
      . Helmets: a road safety manual for decision makers and practitioners. Geneva: World Health Organization, 2006.

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • Contributors SM and KB contributed to the study design and jointly led all aspects of the study. RFSJ and SM initiated the research into the issue. KN, LWM, HG, MW, RB and JR searched for data sources, reviewed questionnaires and conducted data analysis. LWM and KB wrote the first draft of the article. All authors contributed to the discussion and interpretation of the results, writing of the manuscript and have read and approved the final manuscript. KB is the guarantor of the manuscript.

    • Funding This work was supported by the World Bank Global Road Safety Facility, University of Chicago (no award/grant number) and UKAID (Award No. 7197082).

    • Competing interests None declared.

    • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.