Article Text

Lifejacket wear and the associated factors among boaters involved in occupational boating activities on Lake Albert, Uganda: a cross-sectional survey
1. Frederick Oporia1,
2. Fred Nuwaha1,
3. Simon P S Kibira2,
4. Olive Kobusingye1,3,
5. Fredrick Edward Makumbi4,
6. Mary Nakafeero1,
7. Ronald Ssenyonga4,
8. John Bosco Isunju1,
9. Jagnoor Jagnoor5
1. 1Department of Disease Control and Environmental Health, Makerere University College of Health Sciences, Kampala, Uganda
2. 2Department of Community Health and Behavioural Sciences, Makerere University College of Health Sciences, Kampala, Uganda
3. 3The George Institute for Global Health, Camperdown, New South Wales, Australia
4. 4Department of Epidemiology and Biostatistics, Makerere University College of Health Sciences, Kampala, Uganda
5. 5George Institute for Global Health, Camperdown, New South Wales, Australia
1. Correspondence to Frederick Oporia, Department of Disease Control and Environmental Health, Makerere University College of Health Sciences, Kampala, Kampala, Uganda; foporia{at}musph.ac.ug

## Abstract

Background Drowning death rates in lakeside fishing communities in Uganda are the highest recorded globally. Over 95% of people who drowned from a boating activity in Uganda were not wearing a lifejacket. This study describes the prevalence of lifejacket wear and associated factors among boaters involved in occupational boating activities on Lake Albert, Uganda.

Methods We conducted a cross-sectional survey, grounded on etic epistemology and a positivist ontological paradigm. We interviewed 1343 boaters across 18 landing sites on Lake Albert, Uganda. Lifejacket wear was assessed through observation as boaters disembarked from their boats and self-reported wear for those who ‘always wore a life jacket while on the lake’. We used a mixed-effects multilevel Poisson regression, with landing site-specific random intercepts to elicit associations with lifejacket wear. We report adjusted prevalence ratios (PRs) at 95% confidence intervals.

Results The majority of respondents were male, 99.6% (1338/1343), and the largest proportion, 38.4% (516/1343) was aged 20–29 years. Observed lifejacket wear was 0.7% (10/1343). However, self-reported wear was 31.9% (428/1343). Tertiary-level education (adjusted PR 1.57, 95% CI 1.29- 1.91), boat occupancy of at least four people (adjusted PR 2.12, 95% CI 1.28 - 3.52), big boat size (adjusted PR 1.55, 95% CI 1.13 - 2.12) and attending a lifejacket-use training session (adjusted PR 1.25, 95% CI 1.01 - 1.56) were associated with higher prevalence of self-reported lifejacket wear. Self-reported wear was lower among the 30–39 year-olds compared to those who were aged less than 20 years (adjusted PR 0.66, 95% CI 0.45 - 0.99).

Conclusion Lifejacket wear was low. Training on lifejacket use may improve wear among boaters involved in occupational boating activities on Lake Albert.

• Drowning
• Fisher-folk
• Water transport
• Low-Middle Income Country
• Sub-Saharan Africa

## Data availability statement

Data are available upon reasonable request. Due to confidentiality, data are publicly unavailable. However, data may be availed upon reasonable request to the corresponding author on foporia@musph.ac.ug.

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

## 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.

#### WHAT IS ALREADY KNOWN ON THIS TOPIC

• Factors associated with lifejacket wear among leisure boaters in high-income countries are known. There is no evidence as to whether these factors are generalisable to boaters involved in occupational boating activities in rural low-income settings.

• Lakeside fishing communities in Uganda are among the most affected globally. The majority of those who drown from boating activities are not wearing lifejackets.

• This study estimates the prevalence of lifejacket wear and the associated factors among the boaters involved in occupational boating activities on one of the major lakes in Uganda.

• This study identifies potential interventions that may improve lifejacket wear among boaters involved in occupational boating activities in Uganda.

#### HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE AND/OR POLICY

• This study does not have a direct influence on research, practice and policy in Uganda. However, it provides an understanding of the current state of lifejacket wear and the associated factors among the communities known to be the most at risk of drowning. This information may be used to inform government efforts to improve the safety of water transport as a contribution toward achieving the country’s Vision 2040. Indeed, Lake Albert is one of the major lakes in Uganda that is known for frequent drowning incidents. The findings of this study may be used as a basis to develop, pilot and scale up interventions aimed at improving the safety of water transport on Lake Albert in Western Uganda.

## Background

Drowning accounts for 7% of the global burden of injury deaths, and people with frequent exposure to water such as boating have an increased risk.1 Over the last decade, the estimated number of unintentional drowning deaths has slowly decreased from 372 000 in 20122 to 236 000 in 2020.3 Low and middle-income countries (LMICs) suffer the world’s highest drowning death rates and continue to bear over 90% of the burden.2 4 Although Africa has the least data on drowning, the WHO estimates that the region is among the most affected, with death rates at 8 per 100 000 population.2 3 Unfortunately, these global estimates exclude drownings from water transportation and flood disasters which are frequent in Africa and many other LMICs. Drowning death rates in lakeside fishing communities in Uganda are the highest recorded globally, estimated at 502 per 100 000 population.5

Risk factors for drowning are categorised as modifiable (can be changed) and non-modifiable. Modifiable risk factors include inconsistent lifejacket wear, frequent exposure to water and the seaworthiness of watercraft, while the non-modifiable risk factors include age, sex and weather.4 6 Lifejackets are above 80% effective in preventing drowning deaths.7–9 Despite this high effectiveness, lifejacket wear is chronically low, both in high-income countries (HICs) and LMICs. Over 80% of the people who drown from leisure boating activities do not wear life jackets.5 10 11 In Uganda, 95% of people who drowned from boating activities were not wearing a lifejacket. In Lake Victoria fishing communities, lifejacket wear ranges from 2% on the Tanzanian side to 26% in Uganda.12 13 Little is known about the communities around Lake Albert in Western Uganda that have different sociocultural characteristics.

In HICs, low lifejacket wear is driven by a perceived low risk of drowning, perceived strong swimming ability and discomfort. Factors associated with increased lifejacket wear include female gender, boat type (non-motorised), boat size (small) and role modelling.4 14 In Uganda, the majority of people who frequently access water are young adults involved in occupational boating activities of fishing, transportation and other economic activities.12 15 16 Safety practices are left to individual decisions due to limited national legislation and enforcement. Moreover, findings from leisure boating activities in HICs cannot be generalised to occupational boating activities in rural low-resource settings. This study estimates the prevalence of lifejacket wear and the associated factors among boaters involved in occupational boating activities on Lake Albert, Western Uganda, as part of the preliminary studies to inform the development of appropriate interventions.

## Methodology

### Study design and setting

We conducted a cross-sectional survey along the shorelines (landing sites) of Lake Albert in Western Uganda. To objectively measure the outcome of interest, we grounded our study on etic epistemology and a positivist ontological paradigm.17 Lake Albert is Africa’s seventh largest freshwater body, located at the border between Uganda and the Democratic Republic of Congo.18 According to the National Fisheries Resources Research Institute and Uganda Police Marines, there are over 70 landing sites on the Ugandan side but only about half are gazetted. However, due to the rising lake water levels, many had been flooded and vacated, leaving only 18 accessible and occupied at the time of this study. The landing sites, used for embarking and disembarking, are spread across five districts: Kikuube, Hoima, Ntoroko, Buliisa and Pakwach. The inhabitants of the landing sites mainly speak Alur, Runyooro, Rutooro and Lugungu. Apart from the Alur speakers who are from the Luo ethnic group, the other languages belong to the same ethnic group, locally known as Banyakitara, and have similar sociocultural characteristics. The lake supports the local livelihoods of about 4 million people on the Ugandan side who mainly depend on fishing and water transportation businesses.19

### Study participants

The study was conducted among boaters involved in occupational boating activities on Lake Albert, Uganda. We defined occupational boaters to include fishermen and transporters who use boats or canoes, seafarers, coxswains and boat crew (collectively referred to as boaters in this study). From the leadership of Lake Albert Boat Owners’ Association, the number of boaters on the landing sites ranges from 150 to 370. We included boaters who had worked for at least 1 month at the time of the interview. We chose this period because we believed that it was a long enough experience with water to identify associated risks. We excluded boat passengers because of the different exposure risks which are not comparable to those of the occupational boaters who are on the water daily. The occupational boaters spend an average of 12 hours on the lake daily.

### Sample size determination

We determined the sample size using the Leslie Kish formula.20 We considered the following assumptions: a standard normal deviate Zα=1.96; estimated prevalence of lifejacket use, 26% from a previous study in Uganda12 and a 3% precision. We inflated our sample with assumptions of 10% non-response and a design effect (DE) of 1.5 to cater for clustering. This DE has been used and recommended in multiple indicator cluster surveys (MICS).21 These assumptions yielded a final sample size of 1355 boaters.

### Sampling procedures

We included the 18 gazetted landing sites (figure 1) that were functional at the time of this study. From each landing site, an estimated population of boaters was obtained from the landing site leadership. Because the landing sites had different population sizes, we employed proportionate-to-population size sampling to obtain the required number of boaters per site. Study participants disembarking from their boats were interviewed consecutively from 07:00 to 13:00 daily until the required number was obtained from each landing site.

Figure 1

Map of Uganda showing the location of project landing sites.

### Data collection

We collected data using a structured questionnaire programmed in open data kit (ODK) software installed on tablets. The questionnaire development was informed by a synthesis of literature on factors influencing lifejacket use in LMICs, tools from previous studies12 13 15 16 and local knowledge about the community. The questionnaire was pretested at Ggaba landing site on Lake Victoria, and the necessary revisions were made. A Cronbach’s alpha of 0.62 was calculated, indicating an acceptable level of internal consistency reliability22 of the items in the questionnaire. Data collection focused on the domains of socio-demographics, use of life jackets, experience with water, boat ownership and other variables in the behaviour change of the capability, opportunities and motivation for behavior change (COM-B) model of the behavior-change wheel.23

We categorised boat sizes as small (<3 meters long) and big (≥3 meters long) as defined by the Uganda Inland Water Transport Act 2021.24 Lifejacket wear was measured through observation as the boaters disembarked from their boats, as well as self-reported wear. Self-reported lifejacket wear was measured as a binary variable; boaters who ‘always wore a life jacket while on the lake’ were considered to have self-reported lifejacket wear and were assigned to the ‘yes’ category and the rest to the ‘no’ category.

### Data management and analysis

We ensured data quality using plausible ranges pre-programmed into the electronic questionnaire. We imported the dataset into Stata V.15 software for further cleaning and analysis. Exploratory analyses were conducted to assess for outliers and suspect entries. Observed and self-reported lifejacket wear are presented as counts and percentages. Due to the small number of observed lifejacket wear, we were unable to perform analysis beyond descriptive. Therefore, we used self-reported lifejacket wear for further analyses. We used a mixed-effects Poisson regression with landing site-specific random intercepts to elicit associations between self-reported lifejacket wear and the independent variables. The mixed-effects Poisson regression that contains both fixed and random effects was considered appropriate because it allows for modelling intra-cluster correlation. We report prevalence ratios (PRs) as opposed to odds ratios (ORs) to reduce statistical noise because the prevalence of self-reported lifejacket wear was above 10%.25 We assessed multicollinearity among the independent variables using variance inflation factors (VIFs); none of the VIFs was greater than 5, suggesting absence of multicollinearity.26

Before performing multilevel analyses, we ran a null model to calculate the intra-cluster correlation coefficient (ICC), which reflects the proportion of total variance in lifejacket wear outcome explained by landing site. An ICC value of 0.129 was obtained, suggesting that a large amount of variation was accounted for by the landing site,27 28 thus justifying the use of multilevel analysis. We employed the logical model building procedure where all variables that met the 0.2 level of significance at bivariable analysis, as well as those that were not statistically significant but important in literature as known/potential confounders, were included in the multiple regression model. The goodness of fit (GOF) of the model was assessed using the Hosmer-Lemeshow (HL) test. The HL GOF was chosen because it has an asymptotic χ2 distribution for many generalised linear models in the exponential dispersion family, to which the Poisson model belongs.29 We report crude and adjusted PRs at 95% CIs. A level of 5% with two-tailed test was used to signify statistical significance.

## Results

The response rate was high, 99% (1343/1355), with the majority of the respondents being male, 99.6% (1338/1343) and the largest proportion, 38.4% (516/1343) aged 20–29 years. Fishermen constituted the majority, 89.9% (1207/1343) of the sample. The majority, 70.2% (943/1343) had attained secondary school education, while a few. 1.7% (23/1343) had tertiary-level education. A summary of the sociodemographic characteristics is given in table 1.

Table 1

Sociodemographic characteristics of boaters on Lake Albert, Uganda

### Lifejacket wear among the boaters on lake Albert Uganda

Out of the 1343 boaters who participated in this study, only 10 (0.7%) were observed wearing lifejackets. However, 42.4% (570/1343) reported having a lifejacket and only 31.9% (428/1343) reported always wearing a lifejacket while on the lake (table 2).

Table 2

Prevalence of self-reported lifejacket wear by landing site on Lake Albert, Uganda

### Factors associated with reported lifejacket wear among the boaters on Lake Albert, Uganda

Several factors were significantly associated with self-reported lifejacket wear at bivariable analysis. The prevalence of lifejacket wear was higher among boaters who had a household weekly expenditure of above 200 000/− (approx. US$57) compared with those who spent less than 50 000/− (approx. US$14) (unadjusted PR=1.86, 95% CI 1.31 - 2.65). Boat occupancy of at least four people had over a twofold prevalence of lifejacket wear compared with that of two people (unadjusted PR=2.64, 95%-CI 1.43 - 4.89). Lifejacket wear was higher if a big boat was used (unadjusted PR 1.46, 95% CI 1.07 - 2.00). Furthermore, people whose boats had provisions for storage of lifejackets (unadjusted PR=1.52, 95% CI 1.25 - 1.85) and those who had ever attended a training session on lifejacket use (unadjusted PR 1.37, 95% CI 1.12 - 1.68) had a higher prevalence of self-reported lifejacket wear.

At multivariable analysis, the HL GOF test gave a Pearson χ2 p value of 0.98, indicating that the final model fit the data reasonably well. Tertiary-level education (adjusted PR 1.57, 95% CI 1.29 - 1.91) and attending a training session on lifejacket use (adjusted PR 1.25, 95% CI 1.01 - 1.56) were associated with a higher prevalence of lifejacket wear. The prevalence of lifejacket wear was higher when boat occupancy was four or more people (adjusted PR 2.12, 95% CI 1.28 - 3.52), when boat size was more than 3 meters in length (in this study referred to as big boat) (adjusted PR 1.55, 95% CI 1.13 - 2.12), and when the participant owned a boat (adjusted PR 1.44, 95% CI 1.14 to 1.83). However, relative to the boaters aged less than 20 years, the prevalence of lifejacket wear was significantly lower among people who were aged 30–39 years (adjusted PR 0.66, 95% CI 0.45 - 0.99). Taking intoxicating substances such as alcohol was associated with less lifejacket wear, although this was not significant. Self-reported lifejacket wear was lower among the boaters in the two districts at the extreme ends of Lake Albert (see figure 1): Ntoroko (adjusted PR 0.59, 95% CI 0.40 - 0.85) and Pakwach (adjusted PR 0.47, 95% CI 0.23 - 0.96) (table 3).

Table 3

Factors associated with lifejacket wear among occupational boaters on Lake Albert, Western Uganda

## Discussion

This study describes the prevalence of lifejacket wear and the associated factors among boaters involved in occupational boating activities on Lake Albert, Western Uganda. Males constituted the majority in our sample, which may indicate that fishing and transportation are largely dominated by men. This finding can be related to many studies that show the burden of drowning to be particularly high among men.15 16 30 The results show that the largest proportion of boaters were young adults aged 20–39 years. Previous studies demonstrated that the majority of people who drown are under the age of 40 years.12 13 15

Observed lifejacket wear was less than 1%. However, self-reported wear was much higher. In a similar study among fishing communities on Lake Victoria, about two-thirds (67%) of the participants reported using a lifejacket at some point.12 We cannot rule out social desirability bias from the self-reported lifejacket wear because the respondents may have mentioned what they thought the study team wanted to hear. Furthermore, the boaters may take off their lifejackets as they approach the landing, and therefore, our observation as they disembarked from their boats may have missed it. It should be noted that drowning can occur at any point on the lake regardless of the distance from shore. Therefore, lifejackets should be worn at all times while still on water.1 While lifejacket ownership was high, this study shows that it does not necessarily translate into use. There could be other influencers. Previous studies found that people do not wear lifejackets because of distrust in the quality of the lifejackets available at the landing sites, while others perceived a low risk of drowning especially when the waters are calm.12 16 31 In another similar setting like Tanzanian Lake Victoria fishing communities, lifejacket wear was low at 2%13 but substantially higher at 26% among the Ugandan counterparts on the same lake.12

Boat occupancy was associated with self-reported lifejacket wear. This is plausible because the boat occupants may remind each other to wear lifejackets, a behaviour that may be relatable to peer influence. A systematic review of factors associated with lifejacket use found that role modelling was a predictor of increased lifejacket wear among adolescents and indigenous communities.32 This study found that self-reported lifejacket wear was significantly higher among people who used bigger boats than those who used smaller boats. Bigger boats may have more space for storage of lifejackets compared with the smaller ones. Indeed, as evidenced in this study, the provision of storage space in the boats was also associated with higher lifejacket wear. Our results are consistent with another study that also found that people in bigger boats were more likely to wear lifejackets11 but contrary to a study which found that smaller boats were associated with increased lifejacket wear.32

Education/training was the only modifiable factor associated with higher lifejacket wear. These results suggest that training sessions on lifejacket wear may yield positive results as demonstrated in some other studies elsewhere.33 34 Moreover, about 88% of the study participants had never received any form of training or sensitisation on lifejacket use, suggesting that training of boaters could have a high population attributable fraction (PAF) in increasing lifejacket use.35 In addition, the results show that people who had attained a tertiary-level education were more likely users of lifejackets compared with those who had not gone to school. It is possible that those who had tertiary-level education understood the risks associated with non-use of lifejackets. Our findings are, however, different from a systematic review that reported education as one of the factors associated with inconsistent lifejacket use.32

A rowboat (non-motorised) or a fishing/transport boat (motorised) did not show a significant association with lifejacket wear. A case–control study of boat-related injuries in Washington State, USA, showed that people in non-motorised boats were at risk of drowning.36 However, a systematic review of factors associated with lifejacket wear, and another study found that moving in a motorised boat was a predictor of increased lifejacket wear.32 37 Noteworthy in these studies, the boats were categorised as small if they were ≤6 meters long,32 different from our categorisation of <3 meters for a small boat. Furthermore, this study shows that lifejacket wear was significantly lower among people aged 30–39 years. In a systematic review on personal, social and environmental factors associated with lifejacket wear, younger age, especially children, was associated with increased lifejacket wear, but this started reducing as age increased.32 From this study, lifejacket wear among fishermen and transporters was not statistically different, contrary to a study that showed that being a fisherman was positively associated with lifejacket wear.32 However, it should be noted that in this context, it is not important to distinguish between a fisherman or transporter because they have a similar duration of exposure to water, and hence similar risk of drowning. According to the WHO, frequent access to water and non-use or inconsistent use of lifejackets are among the risk factors for drowning.1

There was no significant difference between people who perceived themselves as strong swimmers and those who did not know how to swim. It would be expected that people who know how to swim may be aware of the dangers associated with water, based on their experiences. Overestimation and higher confidence alongside perceived low risk of drowning tagged on swimming expertise is a major risk factor for drowning, as reported in other studies.38 39 Elsewhere, perceived swimming expertise was associated with low lifejacket wear.11 32 40 Consumption of intoxicating substances such as alcohol was also associated with reduced lifejacket wear. The consumption of such substances leads to poor judgement of the water and weather conditions, resulting in overconfidence and therefore reduced lifejacket wear. Although we expected higher lifejacket wear among the people who had ever been in a boat that capsized, our results show the contrary. There was less lifejacket wear among the boaters who had ever experienced a boat capsize. A previous study found that the boaters usually hang on their boats in the event of a boat capsize, citing it as one of the substitutes for lifejacket wear.31

This study is limited by the fact that, first, we relied on self-reported lifejacket wear. Self-reported practices are liable to information bias, especially social desirability bias, which might have made the participants report what they felt was acceptable. Second, our interviews always started at 07:00; it is possible that we missed out on the boaters who returned earlier than that time. In addition, due to the diverse cultural orientations, this study cannot be generalised to all boaters in Uganda because it was conducted among boaters on one lake out of the many in the country. However, we believe that our sample was powered enough to represent the boaters in the districts that neighbour Lake Albert on the Ugandan side.

## Conclusion

This study shows that the observed lifejacket wear among the boaters involved in occupational boating activities on Lake Albert in Uganda is low, while self-reported lifejacket wear was substantially higher. Targeted, contextually relevant training on lifejacket use has the potential to improve lifejacket wear practices among the boaters involved in occupational boating activities on Lake Albert, Uganda.

## Data availability statement

Data are available upon reasonable request. Due to confidentiality, data are publicly unavailable. However, data may be availed upon reasonable request to the corresponding author on foporia@musph.ac.ug.

## Ethics statements

### Ethics approval

This study involves human participants and was approved by the Makerere University School of Public Health Higher Degrees Research and Ethics Committee and registered with Uganda National Council for Science and Technology (registration #SS992ES). Administrative clearance from Uganda Police Marines and the leadership of the landing sites was obtained before the study. The participants gave informed consent to participate in the study before taking part.

## Acknowledgments

We are grateful to the management of the landing sites where the study was conducted. We thank our experienced research assistants (Monica Aweko, Jeff Agenonga, Lawrence Magara, Morris C Jabero, Grace Kabasinguzi and Edgar Ayesiga Wandigali). We are especially grateful to Dr Tessa Clemens from the Centers for Disease Control and Prevention, USA, for her technical advice throughout the development of the protocol and conduct of this study. We thank Otto Businge, Bonny Enock Balugaba, Timothy Mbaziira, Dr Arthur Bagonza, Dr Milton Mutto, Dr Esther Buregyeya and Nishimwe Aurore for their support during this study. We also thank Professor Kjell Torén for his mentorship.

## Footnotes

• Contributors All authors made significant contributions to merit co-authorship. Frederick Oporia conceptualised the study, led the writing of the proposal and obtained ethical clearance, supervised the data collection process, oversaw the analysis, led the writing of the manuscript and played a supervisory role in the entire process. FEM provided technical guidance on the best statistical analysis approach for the study and the interpretation of the results. FN and SPSK participated in reviewing tools and advised on the data collection process and participated in the review and interpretation of findings. JBI participated in the review of the manuscript to ensure intellectual integrity. MN and RS supported the data analysis process, while JJ and OK provided expert advice based on their vast experience and knowledge in the field of drowning prevention, and reviewed the data collection tools to ensure the required data were collected. All authors reviewed and approved the final manuscript. However, FO takes full responsibility for the conduct of the study and final manuscript as the guarantor; he had full access to the data and controlled the decision to publish.

• Funding This study was partly supported by Bloomberg Philanthropies (51606) through the CDC Foundation and the Consortium for Advanced Research Training in Africa (CARTA). CARTA is jointly led by the African Population and Health Research Centre and the University of the Witwatersrand and funded by the Carnegie Corporation of New York (grant number G-19–57145), Sida (grant number 54100113), Uppsala Monitoring Centre and the DELTAS Africa Initiative (grant number 107768/Z/15/Z). The DELTAS Africa Initiative is an independent funding scheme of the African Academy of Sciences’s Alliance for Accelerating Excellence in Science in Africa and supported by the New Partnership for Africa’s Development Planning and Coordinating Agency with funding from the Wellcome Trust (UK) and the UK government.

• Competing interests None declared.

• Patient and public involvement Patients and/or the public were involved in the design, conduct, reporting or dissemination plans of this research. Refer to the Methodology section for further details.

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