Abstract
Background
SafeBoda is a motorcycle taxi company that provides road safety training and helmets to its drivers in Kampala, Uganda. We sought to determine whether SafeBoda drivers are more likely to engage in safe riding behaviours than regular drivers (motorcycle taxi drivers not part of SafeBoda).
Methods
We measured riding behaviours in SafeBoda and regular drivers through: a) computer-assisted personal interview (CAPI), where 400 drivers were asked about their riding behaviours (e.g., helmet and mobile phone use) and b) roadside observation, where riding behaviours were observed in 3000 boda-boda drivers and their passengers along major roads in Kampala.
Results
Across the two cross-sectional studies, a higher proportion of SafeBoda drivers than regular drivers engaged in safe riding behaviours. For instance, helmet use among SafeBoda compared to regular drivers was 21 percent points higher (95% CI: 0.15–0.27; p<0.001) based on the CAPI and 45 percent points higher (95% CI: 0.43–0.47; p<0.001) based on roadside observation. Furthermore, compared to regular drivers, SafeBoda drivers were more likely to report having a driver’s license (66.3% vs 33.5%; p<0.001) and a reflective jacket (99.5% vs 50.5%; p<0.001) and were less likely to report driving towards oncoming traffic (4% vs 45.7%; p<0.001) in the past 30 days.
Conclusion
The SafeBoda program is associated with increased safe riding behaviours among motorcycle taxi drivers in Kampala. Therefore, the promotion and expansion of such programs may lead to a reduction in morbidity and mortality due to road injuries.
INTRODUCTION
In Uganda, motorcycle taxis known locally as boda-boda,a are a major form of transportation because they are relatively affordable, convenient for short distances, and efficient for navigating heavy vehicular traffic.1–5 In the city of Kampala alone, there are an estimated 50,000 to 80,000 boda-boda drivers.2 Despite their advantages, motorcycle taxis are involved in a substantial number of road traffic crashes (RTCs) and associated injuries in Uganda.6–10 Boda-boda drivers are the second largest road user category involved in RTCs in Uganda. In 2015, they were involved in 28.2% of RTCs and represented 23.5% of road traffic injuries in Uganda.6 According to the Road Safety Performance Review report, road fatalities involving boda-boda drivers doubled between 2011 and 2015 (from 570 deaths to 1170 deaths).10 Moreover, Mulago, the main national public hospital in Uganda, receives between 10 and 20 cases of boda-boda related RTCs daily, and spends more than 60% of its annual surgery budget on trauma cases from these crashes.2,9,11 Poor riding behaviours such as speeding, unsafe overtaking, and low use of protective gear have been found to be associated with increased crash and injury risk in this road user group.1,5,8–10 One study conducted among boda-boda drivers presenting at the emergency unit at Mulago Hospital in Kampala, found 57% lower odds of head injury in drivers wearing a helmet during the crash compared to those not wearing a helmet.9 Yet, despite this evidence, helmet use among boda-boda drivers and their passengers in Kampala is low. For instance, a 2011 roadside observation study of 12189 boda-boda drivers and passengers found helmet use of 30.8% and <1% respectively.1
In recent years, private companies such as SafeBoda and Taxify have cropped up to formalize the boda-boda sector and to address the issues around poor riding behaviors through training and provision of protective equipment.12,13 SafeBoda is a road safety-focused private company founded in 2015. It currently provides motorcycle taxi services in Kampala using a trained community of drivers. The company provides multiphase road safety training, two helmets (for driver and passenger), and vehicle maintenance to its drivers in an effort to increase safe riding behaviours and reduce crashes and injuries. The company also provides hairnets to encourage helmet use in passengers concerned about contracting skin diseases from a shared helmet.12 SafeBoda currently has over 5000 drivers in Kampala.
There is contradictory evidence on whether programs such as SafeBoda that provide protective gear and training to motorcycle drivers in low-and middle-income countries are associated with increased safe riding behaviours and reduced incidents of RTC and injuries.14–18 Therefore, we sought to determine whether drivers enrolled in the SafeBoda program were more likely to engage in safe riding behaviours than regular drivers in Kampala, Uganda. In this study and throughout this paper, regular drivers refer to boda-boda drivers who were not identified as part of the SafeBoda program. At the time this study was conducted, SafeBoda was the only company of its kind operating in Kampala. Since that time, additional companies (i.e., UberBoda, Taxify, and Dial Jack) have started to provide similar services in the city.
METHODS
We conducted two cross-sectional studies to measure safe riding behaviours (e.g., helmet use) in SafeBoda and regular drivers in Kampala.
Computer-assisted personal interview (CAPI) study:
We recruited 400 drivers (200 SafeBoda and 200 regular drivers) between October and December 2017 from boda-boda stages across Kampala. A stage is a location where a boda-boda driver is generally stationed when not out driving a passenger or looking for one.
Study size:
We used helmet use as the primary outcome to calculate the study sample size. Previous studies had reported helmet use between 18% and 30% among boda-boda drivers in Uganda.1,9 We assumed that 30% of regular drivers wore a helmet in Kampala. With a type I error rate of 0.05 using a χ2 test, we needed a sample of approximately 400 boda-boda drivers to have 80% power to detect a prevalence ratio (PR) of 1.45 in helmet use comparing SafeBoda and regular drivers.19 We used a purposive and convenience sampling method (i.e., targeting areas with boda-boda stages with at least one SafeBoda driver) to recruit participants for the study. This method had previously been used to recruit boda-boda drivers in Kampala.4 We recruited a maximum of 4 drivers from each stage by enrolling the first 4 to give consent to be in the study. We limited the number recruited from each stage to avoid overrepresentation of any one stage in the study. Drivers were eligible to participate if they were 18 years or older, able to communicate in English or Luganda (a widely spoken language in Uganda), and had been working as a boda-boda driver for at least six consecutive months at the time of recruitment. The choice of six months was to help us to collect covariate information such as past crash history.
Data collection:
Consenting drivers were administered a 15-minute questionnaire in a reasonably quiet location near the recruitment stage. The questionnaire contained questions on demographic and personal characteristics (e.g., age, education, income, years of boda-boda experience, and hours worked per day) and on safe riding behaviours (e.g., helmet use, possession of a driver’s license and reflective vest, and alcohol and mobile phone use). The exposure (SafeBoda status) was ascertained through self-report and/or by observing presence or absence of a SafeBoda branded jacket at the time of the interview. The primary outcome (frequency of helmet use in the past 30 days) was measured through self-report and was categorized in the analysis as “yes” for always, “no” for other responses).. All variables were measured through self-report or direct observation.
The CAPI questions were derived from previous studies and discussion with injury experts in the study team.4,14,20–24 The questionnaire was available in English and Luganda.
Spotter observation study:
In order to validate the self-reported helmet use from the CAPI study, we observed actual helmet use on leaving the boda-boda stage in 20% of the drivers in the CAPI study. Every 2nd CAPI study participant to be interviewed at each stage was systematically selected and observed for helmet use as they left the stage following the interview. Recruitment was continued until the sample size of 80 drivers was reached. Participants did not know they would be observed.
Roadside observation study:
We collected data on SafeBoda status and helmet use and other riding behaviours (e.g., mobile phone use while driving and passenger helmet use) from 3000 boda-boda drivers along purposively selected streets in Kampala. Streets were selected based on traffic volume and to be representative of the five administrative divisions of Kampala (i.e., Rubaga, Kawempe, Kampala Central, Nakawa, and Makindye). SafeBoda status was straightforward to ascertain as SafeBoda drivers wear brightly colored orange reflective jackets inscribed with the company’s logo. Regular boda-boda drivers were distinguished from non-commercial motorcyclists by the size of their motorcycle (engine size of 100–125 cc) and their demeanor (e.g., looking around for customers). We collected data on the first boda-boda driver to pass through the lane nearest to the study’s observation point. This was done at two-minute intervals. If no boda-boda driver drove past the observation point during the first 30 seconds of the two-minute interval, no observation was made until the next interval. For each street, we observed in one travel direction, which was chosen by the flip of a coin. This was done to minimize the risk of double counting. To capture both peak and non-peak hours, we collected data at each observation point for two peak hours (between 7 am and 9 am or 4 pm and 7 pm) and two non-peak hours (between 10 am and 3 pm) in one day.
Analysis:
All analyses were conducted in the statistical package R (version 3.5.1).25 We used Pearson’s chi-square with continuity correction (for categorical variables) and Student’s t-tests (for continuous variables) to compare outcomes (e.g., helmet ownership and past crash history) between SafeBoda and regular drivers. We then compared the proportion of helmet use by SafeBoda and regular drivers from the two studies. We used the EpiStats package in R to assess differences in proportions of self-reported helmet use, observed helmet use, and road traffic crashes between SafeBoda and regular drivers.26 Other riding behaviours (e.g., alcohol use before driving and driving on pedestrian walkways) were analyzed as proportions.
Lastly, we ran a Poisson regression model to estimate the PR to assess whether being in the SafeBoda program was associated with increased self-reported helmet use while adjusting for a minimum set of potential confounders determined a priori. The minimum set of confounders (age, education, and driver’s license) came from a directed acyclic graph of hypothesized variable relationships (based on theory and existing injury literature). We opted to use a Poisson model instead of a logistic model due to helmet use not being a rare outcome in the boda-boda population. We used robust standard errors in the Poisson model to account for any violation of the distribution assumption that the variance equals the mean.
Sensitivity analysis:
To examine the adequacy of the the covariate adjustment model, we used another approach (propensity score matching) to control for confounding of the SafeBoda-helmet use association. The matched exposure sets were created using the MatchIt package in R.27 Due to limitations in the MatchIt package, this analysis used only observations with complete data. We included in the matching the exposure and variables believed to be associated with the exposure and outcome (i.e., age, education, and driver’s license). A Poisson regression model adjusting for these confounders was then used to compare matched pairs of SafeBoda and regular drivers.
We conducted a second sensitivity analysis to assess how robust our self-reported helmet use data were against potential outcome misclassification due to social desirability bias. We imputed observed helmet use for drivers with self-reported helmet use but who were not observed in the spotter observation study (internal validation group). We took advantage of the relationships between observed helmet use, self-reported helmet use, SafeBoda status, and relevant covariates (e.g., age) in the validation group to impute the missing observed helmet use for drivers not in the validation group.28,29 The mice package in R was used for the multiple imputation process.30 We used a logistic regression model for monotone missing data to impute the missing observed helmet use.28,31 Upon completion of the imputation process, we generated a single PR comparing imputed observed helmet use in SafeBoda and regular drivers by pooling the PRs from 10 imputation rounds (using a Poisson model). We compared this PR to the PR from the naïve (self-reported) and spotter observation analyses. All models adjusted for the same set of covariates.
Ethical approval and consent:
Approval for the study was granted by University of Washington, Makerere University, and the Uganda National Council for Science and Technology. Participants in the CAPI study provided verbal consent to be interviewed and were compensated for their time.
RESULTS
Among 200 SafeBoda and 200 regular drivers who completed the CAPI, compared to regular drivers, SafeBoda drivers had more boda-boda experience, were more educated, and made more money per week (Table 1).
Table 1.
Comparison of baseline characteristics of SafeBoda and regular boda-boda drivers in Kampala (CAPI study)
| Characteristic |
SafeBoda drivers N = 200 |
Regular drivers N = 200 |
P value |
|---|---|---|---|
| n (%) or mean (SD) | n (%) or mean (SD) | ||
| Age (years) | 33.5 (7.2) | 32.1 (7.0) | 0.05 |
| Education (years) | 8.4 (3.7) | 8.0 (3.6) | 0.25 |
| Weekly net income (in USD) | 25.05 (14.19) | 23.16 (12.33) | 0.16 |
| Number of trips per day | 18.3 (7.9) | 18.6 (7.7) | 0.65 |
| Hours worked as boda-boda per day | 12.6 (2.2) | 12.2 (2.0) | 0.03 |
| Has a driving license | 132 (66.3%) | 67 (33.5%) | <0.001 |
| Has other job | 64 (32%) | 76 (38.0%) | 0.25 |
| Used a phone while driving in past 30 days | 39 (19.5%) | 92 (46%) | <0.001 |
| Alcohol use in past 2 hours before driving in past 30 days | 4 (12.1%) | 26 (41.3%) | 0.007 |
| Had a road traffic crash in past 6 months | 43 (21.5%) | 67 (33.5%) | 0.01 |
| Own a helmet | 198 (99%) | 185 (92.5%) | 0.003 |
| Helmet cost (in USD) | 27.47 (9.95) | 10.02 (7.06) | <0.001 |
| Has a reflective jacket | 199 (99.5%) | 101 (50.5%) | <0.001 |
| Has ever received a road safety training | 198 (100%) | 146 (74.1%) | <0.001 |
| Has driven on a pedestrian sidewalk in past 30 days | 14 (7.0%) | 91 (45.5%) | <0.001 |
| Has driven towards oncoming traffic in past 30 days | 8 (4.0%) | 91 (45.7%) | <0.001 |
| Carried more than one passenger in past 30 days | 18 (9.1%) | 156 (78.4%) | <0.001 |
| Years working as a boda-boda driver | 7.0 (4.2) | 6.3 (4.1) | 0.10 |
Note: Continuous variables are reported as mean with standard deviation (in parenthesis), while categorical variables are reported as count with percent (in parenthesis). Some observations have missing data.
In the roadside observation study of 3000 boda-boda drivers in Kampala, 49 (1.6%) were identified as SafeBoda drivers and 2951 (98.4%) were classified as regular drivers (Table 2). Only 57% of the drivers were observed to carry a passenger, of which 9% carried more than one passenger (none among SafeBoda and 9.5% among regular drivers). About two-thirds of the passengers carried were male and this was similar in both SafeBoda and regular drivers. A small percentage of drivers (1.2%) were observed to engage in distracted driving behaviour by using their phone while driving (6.1% among SafeBoda vs 1.2% among regular drivers).
Table 2.
Characteristics of boda-boda drivers in Kampala (roadside observation study)
| Characteristic |
SafeBoda drivers N = 49 |
Regular drivers N = 2951 |
|---|---|---|
| n (%) | n (%) | |
| Division of Kampala | ||
| Kampala Central | 23 (46.9%) | 794 (26.9%) |
| Nakawa | 12 (24.5%) | 805 (27.3%) |
| Makindye | 10 (20.4%) | 569 (19.3%) |
| Kawempe | 3 (6.1%) | 313 (10.6%) |
| Rubaga | 1 (2.0%) | 470 (15.9%) |
| Carried a passenger during observation | 35 (71.4%) | 1,689 (57.2%) |
| Carried more than one passenger | 0 (0.0%) | 161 (9.5%) |
| Driver observed to be using their mobile phone | 3 (6.1%) | 34 (1.2%) |
| Sex of first boda-boda passenger | ||
| Female | 13 (37.1%) | 622 (36.8%) |
| Male | 22 (62.9%) | 1,067 (63.2%) |
| First passenger wore a helmet | 8 (22.9%) | 11 (0.7%) |
| First passenger helmet was correctly strapped | 6 (75%) | 8 (72.7%) |
Note:
Categorical variables are reported as count with percent (in parenthesis). Data exclude missing observations.
Adjusting for division did not significantly change observed difference in helmet use in SafeBoda and regular drivers (PR: 1.81 vs 1.76)
Helmet use:
We compared the proportion of helmet use in boda-boda drivers, as measured by self-report and direct observation. Of the 400 drivers who reported their frequency of helmet use in the 30 days before the study, 88% (95% CI: 84.3%–90.9%) reported always using a helmet while driving, compared to 69.6% (95% CI: 58.1%–79.2%) observed to use a helmet during the spotter observation of 80 participants in the CAPI. In the roadside observation study of 3000 boda-boda drivers, we observed helmet use by 56.0% (95% CI: 54.2%–57.8%) of drivers. Observed helmet use among passengers was very low at 1.1% (95% CI: 0.69%–1.8%).
Across the two studies, a higher proportion of SafeBoda drivers used a helmet while driving than did regular drivers (Table 3). The difference in the proportion of helmet use between SafeBoda and regular drivers was 0.21 (95% CI: 0.15–0.27; p<0.001) in the CAPI study and 0.45 (95% CI: 0.43–0.47; p<0.001) in the roadside observation study. From the roadside observation, the difference in the proportion of helmet use between passengers carried by SafeBoda drivers and passengers carried by regular drivers was 0.22 (95% CI: 0.08–0.36; p<0.001).
Table 3.
Proportion of helmet use for SafeBoda and regular drivers: results from two approaches
| Approach | SafeBoda driver | SafeBoda passenger | Regular driver | Regular passenger |
|---|---|---|---|---|
| Self-reported (CAPI) | 0.99 (0.95, 1.00) | - | 0.78 (0.71, 0.83) | - |
| Spotter observation (CAPI) | 0.96 (0.78, 1.00) | - | 0.57 (0.43, 0.70) | - |
| Roadside observation | 1.00 (0.91, 1.00) | 0.23 (0.11, 0.41) | 0.55 (0.53, 0.57) | 0.007 (0.003, 0.012) |
Upon covariate adjustment for potential confounders (Figure S1), SafeBoda drivers were 1.22 times more likely to use a helmet while driving than regular drivers (95% CI: 1.14–1.31; p<0.001) in the CAPI study (Table 4). Adjustment for confounding with propensity score matching gave similar results (PR: 1.21, 95% CI: 1.11–1.32; p<0.001). There were 129 matched exposure pairs from a dataset of 394 participants (Figures S2–S4).
Table 4.
Adjusted* prevalence ratios for self-reported helmet use comparing SafeBoda and regular drivers
| PR (covariate adjustment) | PR (propensity score matching) | |
|---|---|---|
| Regular boda-boda driver | 1.00 | 1.00 |
| SafeBoda driver | 1.22 (1.14–1.31) | 1.21 (1.11–1.32) |
Note: Adjusted for age, education, and possession of a driver’s license
Other riding behaviours:
Compared to regular drivers, SafeBoda drivers were more likely to self-report other safe riding behaviours in the CAPI study. They were more likely to report having a driver’s license (66.3% vs 33.5%; p<0.001) and a reflective jacket (99.5% vs 50.5%; p<0.001). Moreover, they were less likely to report talking on the phone while driving (19.5% vs 46%; p<0.001), driving towards oncoming traffic (4% vs 45.7%; p<0.001), and carrying more than one passenger (9.1% vs 78.4%; p<0.001) in the past 30 days (Table 1). From the roadside observation, SafeBoda drivers were more likely to be carrying a passenger (71% vs 57%; p = 0.07), less likely to be carrying more than 1 passenger (0% vs 9.5%; p = 0.10), and more likely to be observed using a mobile phone while driving (6.1% vs 1.2%; p = 0.01) (Table 2).
Road traffic crashes:
Compared to regular drivers, SafeBoda drivers were less likely to report having been involved in a road traffic crash in the six months prior to the CAPI study (RR: 0.64, 95% CI: 0.46–0.89; p = 0.007). The majority of the 110 reported crashes involved either a collision with another motorcycle (27.3%) or a car (53.6%). Speeding (38.2%), making an illegal turn (30.9%), and driving towards oncoming traffic (17.3%) were the most frequently mentioned causes of the reported crashes. Among those who reported a crash, 89 (80.9%) were injuredb and 81% of these injuries required a visit to a health facility.
Sensitivity analysis:
Compared to the effect size based on self-reported helmet use (PR = 1.22), we found higher PRs comparing SafeBoda and regular drivers based on observed helmet use. Using the spotter observation data, SafeBoda drivers were 61% more likely to use a helmet while driving than regular drivers (PR: 1.61, 95% CI: 1.26–2.04). Results from the multiple imputation were similar (PR: 1.60, 95% CI: 1.53–1.67) but with narrower confidence intervals.
DISCUSSION
Our findings suggest that drivers in the SafeBoda program were not only more likely to wear a helmet while driving but they were also more likely to engage in other safe riding behaviours such as wearing a reflective vest, and less likely to engage in risky behaviours such as driving towards oncoming traffic. However, observed mobile phone use while driving was higher in SafeBoda compared to regular drivers, although the observed number was quite small.
There are multiple mechanisms that may explain the observed higher rate of safe driving behaviours and helmet use among SafeBoda drivers. For instance, it is possible that the SafeBoda program may be recruiting more safety conscious drivers. However, other plausible mechanisms for the observed findings in SafeBoda drivers include the provision of road safety training and protective equipment and the sense of community among SafeBoda drivers. Not only are the drivers provided with two helmets and a reflective vest, they are also trained to view themselves as a community. In a way, SafeBoda drivers look out for each other as they drive. They follow traffic regulations unlike regular boda-boda drivers because they were trained to do so but also because other SafeBoda drivers are watching them. SafeBoda passengers were found to be more likely to wear helmets. This may be because helmets are provided, whereas regular boda-boda passengers would need to purchase their own. It could also be because SafeBoda passengers are more safety conscious. In addition, messages urging helmet use appear on passenger phone screen when ordering for a SafeBoda via their app.
Our sensitivity analyses seem to confirm the possibility of outcome misclassification from social desirability bias. We found a higher PR for observed helmet use in the sensitivity analyses than in the primary analysis. However, a model with an interaction term for the exposure (SafeBoda) and misclassified outcome (self-reported helmet use) was not statistically significant. In addition, the similar PRs for the spotter observation and multiply imputed data likely point to missingness at random in our data. This discrepancy in self-reported and observed helmet use has been previously reported in other settings.24,32,33
Our findings corroborate those of previous investigators that found generally low helmet use and frequent high-risk driving behaviours among boda-boda drivers in Kampala.1,4,9 Similar to a 2011 roadside observation study by Roehler and colleagues in Kampala and a 2017 study by Bachani et al. in Kenya, we found low helmet use (1.1%) in boda-boda passengers.1,24 However, unlike Roehler et al., we found a much higher observed helmet use in boda-boda drivers in our roadside observation study (56% versus 31%).1 This discrepancy could be due to differences in our sampling approaches or to an increasing trend of helmet use over time (especially with the advent of programs such as SafeBoda).
This study has some limitations. First, we relied on self-report for the majority of the variables in the analysis. This may have introduced bias in our study estimates. We attempted to control for this potential bias, in both the design and analytic phases, by including direct observation studies and conducting sensitivity analyses. Second, the numbers of SafeBoda drivers and passengers observed during the roadside observation study were quite small. This has implications on the precision and generalizability of our findings. In addition, since we did not use probability sampling for the CAPI study, the generalizability of our findings to the whole boda-boda population in Kampala might be limited. However, it was not possible to use probability sampling due to the lack of sampling frame for boda-boda stages in Kampala and the study’s focus on stages with at least one SafeBoda driver. Third, there is a possibility for selection bias if boda-boda drivers recruited from stages are systematically different from those riding on the streets. However, we found similar results (i.e., more road safety consciousness in SafeBoda drivers) in both the CAPI and roadside observation studies. Lastly, there was potential to misclassify non-commercial motorcyclists as regular boda-boda drivers during the roadside observation study. This could lead to an attenuated association given that non-commercial motorcyclists (i.e., courier and delivery drivers and drivers of personal motorcycles) generally have higher rates of helmet use than regular drivers. However, we believe there was little potential for misclassification of non-commercial motorcyclists as regular boda-boda drivers because regular boda-boda drivers in Kampala have a unique profile, including the types of the motorcycle they use and their behaviour on the road. Another related misclassification limitation was the difficulty in identifying whether or not a driver’s or a passenger’s helmet was correctly strapped during the roadside observation study. We used two observers during the study to minimize this limitation.
CONCLUSION
Our findings suggest that the SafeBoda program is associated with increased usage of helmets for both drivers and passengers and also with other safe behaviours such as not riding with more than one passenger. Given these findings, we believe SafeBoda and similar programs that recently joined the boda-boda industry (e.g., Taxify and UberBoda) may be worth scaling up in order to accelerate adoption of safe riding behaviours among boda-boda drivers.
Supplementary Material
What’s already known on the subject
Helmets reduce risk of head injury during a crash.
Helmet use is low among boda-boda drivers in Uganda.
What this study adds
Helmet use estimates after introduction of the SafeBoda program.
Helmet use estimates from three measurement approaches.
Safe riding behaviours disaggregated by type of boda-boda driver.
Acknowledgments:
The authors are grateful to Grace Magambo, Joseph Mugisha, and Simon Okurut for their support throughout the study.
Funding:
Study was supported by NIH Research Training Grant #D43 TW009345 awarded to the Northern Pacific Global Health Fellows Program by the Fogarty International Center.
Footnotes
Competing interests: None. SafeBoda was informed about the study but the company had no influence on the study.
Ethics approval: The study was approved by University of Washington IRB, Makerere University School of Public Health IRB, and Uganda National Council for Science and Technology.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement: The data that support the findings of this study are available from the corresponding author, [KM], upon reasonable request.
This term originally referred to bicycle taxis that operated at the Uganda-Kenya border. However, it is now a generic term for bicycle and motorcycle taxis in East Africa. In present use, boda-boda increasingly refers to motorcycle taxis.
Injury was defined as any wound or bruise from a road traffic crash regardless of whether treatment was sought.
REFERENCES
- 1.Roehler DR, Naumann RB, Mutatina B, et al. Using baseline and formative evaluation data to inform the Uganda Helmet Vaccine Initiative. Glob Health Promot. 2013;20(4 Suppl):37–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Sebaggala R, Matovu F, Ayebale D, Kisenyi V, Katusiimeh M. The cost of motorcycle Accidents in Uganda. 2015.
- 3.Goodfellow T Politics on two wheels: motorcycle taxis and divergent development in Rwanda and Uganda. http://africanarguments.org/2015/02/17/politics-on-two-wheels-motorcycle-taxis-and-divergent-development-rwanda-and-uganda-by-tom-goodfellow/. Published 2015. Accessed.
- 4.Tumwesigye NM, Atuyambe LM, Kobusingye OK. Factors Associated with Injuries among Commercial Motorcyclists: Evidence from a Matched Case Control Study in Kampala City, Uganda. PLoS One. 2016;11(2):e0148511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Raynor B Informal transportation in Uganda: a case study of the Boda Boda. Kenyon College;2014. [Google Scholar]
- 6.UBOS. Statistical Abstract. Uganda Bureau of Statistics;2015. [Google Scholar]
- 7.Kitara D Boda Boda Injuries in Gulu Regional Hospital, Northern Uganda. East and Central African Journal of Surgery. 2011;16(2). [Google Scholar]
- 8.Galukande M, Jombwe J, Fualal J, Gakwaya A. Boda-boda injuries a health problem and a burden of disease in Uganda: A tertiary hospital survey. East and Central African Journal of Surgery. 2009;14(2). [Google Scholar]
- 9.Kamulegeya LH, Kizito M, Nassali R, Bagayana S, Elobu AE. The scourge of head injury among commercial motorcycle riders in Kampala; a preventable clinical and public health menace. Afr Health Sci. 2015;15(3):1016–1022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Naimanye Andrew, Nganwa Racheal, Kwamusi P, Tomusange M. Uganda Road Safety Performance Review. Kampala: Secretariat for the Secretary General’s Special Envoy for Road Safety; 2018. [Google Scholar]
- 11.UHVI. Putting helmets on heads – Making roads safe. 2013. [Google Scholar]
- 12.SafeBoda. Safety | SafeBoda. http://www.safeboda.com/home/safety/. Published 2016. Accessed 11/9/2016, 2016. [Google Scholar]
- 13.Mpairwe H Taxify Go is the new low cost hailing option in Kampala. https://www.techjaja.com/taxify-go-is-the-new-low-cost-hailing-option-in-kampala/. Published 2018. Accessed 8/20, 2018.
- 14.Luchidio M, Kahuthia-Gathu R, Gatebe E. Impact of training boda boda operators and safety status in Kakamega county, Kenya. International Journal of Advance Research. 2013;1(9). [Google Scholar]
- 15.Kardamanidis K, Martiniuk A, Ivers RQ, Stevenson MR, Thistlethwaite K. Motorcycle rider training for the prevention of road traffic crashes. Cochrane Database Syst Rev. 2010(10):Cd005240. [DOI] [PubMed] [Google Scholar]
- 16.Woratanarat P, Ingsathit A, Chatchaipan P, Suriyawongpaisal P. Safety riding program and motorcycle-related injuries in Thailand. Accid Anal Prev. 2013;58:115–121. [DOI] [PubMed] [Google Scholar]
- 17.Ivers RQ, Sakashita C, Senserrick T, et al. Does an on-road motorcycle coaching program reduce crashes in novice riders? A randomised control trial. Accid Anal Prev. 2016;86:40–46. [DOI] [PubMed] [Google Scholar]
- 18.Swaddiwudhipong W, Boonmak C, Nguntra P, Mahasakpan P. Effect of motorcycle rider education on changes in risk behaviours and motorcycle-related injuries in rural Thailand. Trop Med Int Health. 1998;3(10):767–770. [DOI] [PubMed] [Google Scholar]
- 19.Dean A, Sullivan K, Soe M. Open Source Epidemiologic Statistics for Public Health. In:2013.
- 20.Johnson O. Prevalence and Pattern of Road Traffic Accidents among Commercial Motorcyclists in a city in Southern Nigeria. (PDF Download Available). Education Research. 2012;3(6):537–542. [Google Scholar]
- 21.PO A, Illika A, Asuzu A. Predictors of road traffic accident, road traffic injury and death among commercial motorcyclists in an urban area of Nigeria. - PubMed - NCBI. 2017. [DOI] [PubMed] [Google Scholar]
- 22.Oginni FO, Ugboko VI, Adewole RA. Knowledge, attitude, and practice of Nigerian commercial motorcyclists in the use of crash helmet and other safety measures. Traffic Inj Prev. 2007;8(2):137–141. [DOI] [PubMed] [Google Scholar]
- 23.Amoran OE, Eme O, Giwa OA, Gbolahan OB. Road safety practices among commercial motorcyclists in a rural town in Nigeria: implications for health education. Int Q Community Health Educ. 2005;24(1):55–64. [DOI] [PubMed] [Google Scholar]
- 24.Bachani AM, Hung YW, Mogere S, Akunga D, Nyamari J, Hyder AA. Helmet wearing in Kenya: prevalence, knowledge, attitude, practice and implications. Public Health. 2017;144s:S23–s31. [DOI] [PubMed] [Google Scholar]
- 25.R Core Team. R: A language and environment for statistical computing. In. Vienna, Austria: R Foundation for Statistical Computing; 2018. [Google Scholar]
- 26.Decorps JP. EpiStats: Tools for Epidemiologists. R package version 1.0. In:2017. [Google Scholar]
- 27.Ho D, Imai K, King G, Stuart E. MatchIt: Nonparametric preprocessing for parametric causal inference Journal of Statistical Software. 2011;42(8):1–28. [Google Scholar]
- 28.Edwards JK, Cole SR, Troester MA, Richardson DB. Accounting for Misclassified Outcomes in Binary Regression Models Using Multiple Imputation With Internal Validation Data. Am J Epidemiol. 2013;177(9):904–912. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Cole SR, Chu H, Greenland S. Multiple-imputation for measurement-error correction. Int J Epidemiol. 2006;35(4):1074–1081. [DOI] [PubMed] [Google Scholar]
- 30.van Buuren S, Groothuis-Oudshoor K. mice: Multivariate Imputation by Chained Equations in R Journal of Statistical Software. 2011;45(3):1–67. [Google Scholar]
- 31.Rubin DB. Multiple Imputation for Nonresponse in Surveys. John Wiley & Sons, Inc.; 1987. [Google Scholar]
- 32.Wadhwaniya S, Gupta S, Mitra S, et al. A comparison of observed and self-reported helmet use and associated factors among motorcyclists in Hyderabad city, India. Public Health. 2017;144s:S62–s69. [DOI] [PubMed] [Google Scholar]
- 33.Bachani AM, Tran NT, Sann S, et al. Helmet use among motorcyclists in Cambodia: a survey of use, knowledge, attitudes, and practices. Traffic Inj Prev. 2012;13 Suppl 1:31–36. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.