Epidemiology of traumatic brain injury based on clinical symptoms amongst inhabitants of the Southwest Region of Cameroon: a community-based study

ABSTRACT

Background

The majority of studies investigating the epidemiology of traumatic brain injury (TBI) in sub-Saharan Africa are primarily hospital-based, missing fatal, mild, and other cases of TBI that do not present to formal care settings. This study aims to bridge this gap in data by describing the epidemiology of TBI in the Southwest Region of Cameroon.

Methods

This was a cross-sectional community-based study. Using a three-stage cluster sampling, local research assistants surveyed households with a pre-tested questionnaire to identify individuals with symptoms of TBI in nine health districts in the Southwest Region of Cameroon from 2016 to 2017.

Results

Data gathered on 8,065 individuals revealed 78 cases of suspected TBI. Road traffic injury (RTI) comprised 55% of subjects’ mechanism of injury. Formal medical care was sought by 82.1% of subjects; three subjects died at the time of injury. Following injury, 59% of subjects reported difficulty affording basic necessities and 87.2% of subjects were unable to perform activities of their primary occupation.

Conclusions

This study postulates an incidence of TBI in Southwest Cameroon of 975.57 per 100,000 individuals, significantly greater than prior findings. A large proportion of TBI is secondary to RTI.

KEYWORDS:

• Traumatic brain injury
• Cameroon
• sub-Saharan Africa
• TBI

Introduction

Traumatic brain injury (TBI) is a leading cause of trauma-related death and disability worldwide, with an estimated 64–74 million new cases occurring worldwide each year (1). TBIs are associated with high mortality rates, especially at the site of the injury event. In addition to morbidity resulting from injury, there is a highly demonstrated socioeconomic burden on affected individuals and their families (2). TBI has been termed the “silent epidemic” of modern society because problems resulting from TBI are usually not immediately visible nor often reported by patients (3,4). Thus, the incidence of TBI is often underestimated and the societal impact is largely unknown.

The burden of TBIs is disproportionately felt by low- and middle-income countries (LMICs), likely because of rapid urbanization and motorization without adequate corresponding traffic control measures. This includes poor road networks, inadequate road safety measures, and poorly enforced road traffic laws (5). The World Health Organization (WHO) estimated that in Africa there are approximately two million cases of TBI requiring neurosurgical care annually – but that existing capacity can only manage around 100,000 neurosurgical cases (6). Furthermore, there is also a dearth of medical capacity necessary to manage the health outcomes associated with TBI (6,7).

While TBI is thought to be a significant health issue in LMICs, the majority of multicentre trials and research on TBI has predominantly focused on populations in high-income countries (8). The epidemiology of TBI in sub-Saharan Africa remains unclear. Available studies are primarily hospital-based with vaguely defined catchment areas; thus, it is perceivable that both fatal and mild cases of TBI, which do not present to formal care, are not captured in the literature (9–14). Prior work in Ghana demonstrated that there is low utilization of formal medical services by injured individuals, particularly in rural areas (15). There are also significant financial and cultural barriers, along with a lack of transportation, which prevent formal care seeking in this setting (16). Lack of reliable estimates of the epidemiology of TBI in LMICs hinders efforts toward appropriate funding allocation, injury prevention efforts, and development of effective treatments.

This study aims to bridge this gap in data by describing the epidemiology of TBI in the Southwest Region of Cameroon through a cross-sectional community-based survey, with the goal of capturing cases on both ends of the severity spectrum. The primary aim of this study is to estimate the incidence of TBI in Southwest Cameroon. Our secondary aim is to measure the disability and socioeconomic consequences associated with TBI experienced by injured people and their families.

Methods

Study design, setting, and sampling: We conducted a cross-sectional community-based study on the clinical symptoms of TBI. This study was designed as a sub-analysis of a study powered to identify the incidence of injury in the Southwest Region of Cameroon from 2016 to 2017 (17). The Southwest Region is one of the 10 regions of Cameroon. The region has a mixed urban/rural distribution with a combined population of 1,575,224, with 18 health districts containing 36 health areas (18,19). We used a three-stage cluster sampling framework to select enumeration areas with selection probability proportionate to size. This methodology has been previously described (17). Of the 18 health districts, nine were selected (two districts, Akwaya and Bakassi, were excluded from selection due to feasibility and safety constraints). Within each of these nine health districts, four health areas were then selected. Households were randomly selected using geolocation mapping, a grid, and a random number generator.

Sample size calculation: Sample size for the estimation of injury was derived using a 78% power to detect a 6% yearly injury incidence (estimated from prior studies) with a precision of ±5% and a 95% confidence interval. This was adjusted for 20% predicted non-response rate and a design effect of two for the study’s multi-cluster sampling framework to yield a sample size of 4,680. This calculated sample size was deliberately exceeded during the collection by approximately 50% at each site to account for multiple sub-analyses of relatively rare events, such as TBI. A minimum sample size of 806 was calculated to verify that this sample was large enough to conduct a TBI sub-analyses, using a TBI prevalence of 30% from a previous study, with a precision of ±5% and a 95% confidence interval adjusting for 20% non-response and a design effect of two (7).

Surveys and interviews: Data were obtained between January 3, 2017, and March 3, 2017, using a pre-tested questionnaire which assessed the following variables: history of TBI, basic demographic data, mechanism of injury, care-seeking behavior, treatment, health outcomes, disability, household costs, and familial impact. A team of trained local research assistants administered questionnaires in the subject’s preferred language. Within each household, one family member greater than 18 years old acted as the family representative and provided injury data for all other family members. We excluded subjects not living in the Southwest Region, those who failed to provide informed consent, and households that did not have anyone aged 18 years or older to provide consent after two attempts. We followed reporting guidelines from the American Association for Public Opinion Research (AAPOR) to calculate survey outcome rates (20). A standard, semi-structured interview was designed and employed as well, based on the access to care framework developed by Peters et al., which explored perceptions of care and determinants of post-injury care decision-making (21).

Disability was measured using standard disability categories from the Surgeons OverSeas Assessment of Surgical Need (SOSAS) tool and the World Health Organization’s (WHO) Guidelines for Conducting Community Surveys on Injuries and Violence (22,23). Familial impact was assessed by measuring the extent to which caretakers shifted their daily activities, lost wages, and had new financial difficulties. TBI was defined as by the American Congress of Rehabilitation Medicine: traumatically induced physiological disruption of brain function as manifested by at least one of the following: any period of loss of consciousness, any loss of memory for events immediately before or after the accident, any alteration in mental state at the time of the accident, and/or focal neurological deficit(s) that may or may not be transient (24). As a standard method of assessing transient neurologic deficits in the community setting was not available, patients were surveyed for loss of consciousness, confusion, or memory loss following head injury.

The International Monetary Fund’s 2016 conversion rate of 593 Franc of the Francophone Community of Africa (CFA franc) to 1 United States Dollar (USD) was utilized when reporting financial impact on wages (25).

Data management and analysis

Data collected on paper surveys were entered into REDCap and were analyzed using Stata 14; svy commands were used as appropriate to account for sampling weights and cluster survey design (26–28). The incidence of subjects with suspected TBI based on clinical symptoms of loss of consciousness (LOC), amnesia, or confusion was calculated. Groups were characterized using descriptive statistics, including means and standard errors (SE) for parametric continuous variables and medians with interquartile ranges (IQR) for non-parametric variables, proportions for binary variables, and percentage frequencies for categorical variables. The difference between groups was evaluated using Chi-square test and the adjusted Wald test. A p-value of less than 0.05 was considered statistically significant.

Ethical approval

The study was carried out after approval by the University of Buea Institutional Ethical Review Board, the University of Douala Ethical Review Board, and the University of California, San Francisco’s Institutional Review Board. All individuals involved in data collection, management, and analysis received an online course on Human Subjects Research from the Collaborative Institutional Training Initiative Program (29).

Results

General population characteristics

During this study 1,551 households were approached and surveyed. Of these, 151 were not eligible and 113 did not consent. Data were collected on 1,287 households, which provided information on 8,065 individuals. Based on reported symptoms of loss of consciousness (LOC), confusion, or amnesia after an injury, 77 subjects were suspected of having 78 events of TBI within the last 12 months (Figure 1). Only 10 recognized that they had a TBI. Males comprised 50 subjects with suspected TBI, accounting for 55% of the total TBI cohort when adjusted for the multi-cluster survey sampling method (Table 1). The median age was 24 (IQR: 7–34) years for participants with suspected TBI, 24 (IQR: 14–38) years for those with injuries other than TBI, and 20 (IQR: 10–34) years for uninjured participants.

Table 1. Comparison of demographic and socioeconomic variables between individuals with suspected TBI, non-TBI injury, and the non-injured population (n = 8065^a).^b

Variables	Suspected TBI (n = 77)	Non-TB injury (n = 394)	No injury (n = 7594)	p-Value
Age Mean (SE) Median (IQR)	25.4 (3.75) 24 (7–34)()	27.7 (0.91) 24 (14–38)	23.9 (0.27) 20 (10–34)	0.711
Sex, n = 7995 (99%) Male Female	50 (55%) 27 (45%)	236 (56%) 153 (44%)	3560 (46%) 3969 (54%)	0.067
Urban residence, n = 7953 (98.6%)	20 (31%)	110 (40%)	2203 (39%)	0.542
Own agricultural land, n = 7949 (98.6%)	50 (61%)	271 (64%)	4844 (56%)	0.152
Cooking fuel, n = 8014 (99.4%) Wood Charcoal LPG Kerosene	70 (83%) 4 (5%) 30 (56%) 10 (15%)	361 (90%) 73 (25%) 185 (54%) 61 (16%)	6975 (89%) 1201 (22%) 3233 (53%) 1215 (20%)	0.496 0.020 0.750 0.323
Level of education, n = 7910 (98.1%) None Primary Secondary Tertiary	0 (0%) 27 (35%) 28 (32%) 21 (33%)	9 (2%) 80 (16%) 139 (36%) 156 (46%)	148 (1%) 1540 (17%) 2788 (36%) 2956 (45%)	0.481
Household possess cellphone, n = 7871 (97.6%)	74 (99%)	366 (95%)	7075 (96%)	0.064

^aVariables had missing data; hence, the total n differs for each variable. Missing data for each variable were excluded from the analysis.

^bPercentages adjusted for the multi-cluster survey sampling method.

Figure 1. Subject inclusion in the study.

Epidemiology of suspected TBI

Of the 78 suspected TBI events, most of the subjects were farmers (31, 40%) and students (18, 23%). The median education level was 9 years (IQR: 7–11 years), and the median weekly income was 4.22 USD (IQR: 0–25.30 USD).

Mechanism of injury

Out of the 78 suspected TBI events, 43 (55.1%) occurred via road traffic injury (RTI). Of these, 39 (90.7%) subjects were riding a motorbike at the time of injury. Helmets were not worn in 85.3% of motorcycle injuries. Other major causes of TBI were falls (16, 20.5%), blunt force trauma from objects (8, 10.3%), and assault (5, 6.41%). Within the 78 cases with TBI, confusion after injury was reported by 47 subjects, LOC by 44 subjects, and amnesia by 9.

Treatment

Medical care was sought following 64 (82.1%) cases of TBI. Of those who presented at the hospital, 32 (50%) stayed until discharged, 8 (12.5%) went home against medical advice, and 20 (31.3%) were evaluated but never admitted. The median stay in the hospital was two nights (IQR 0–7 nights). Additionally, 17 (21.8%) subjects with TBI were later readmitted or needed additional treatment.

There were 16 (20.5%) subjects who presented to traditional healers or bone setters. This decision was made by seven subjects due to personal preference, five because formal care was too expensive, and three because they did not believe that their injury was serious. The final subject did not provide a specific reason.

Health outcomes

Of the 8,065 participants, there were three fatal injuries due to suspected TBI. Following TBI, 68 (87.1%) subjects were unable to perform activities of their primary occupation, with a median of 30 disability days (IQR: 7–60 days). In regard to basic activities of daily living, 39 (50.0%) had trouble dressing themselves, eating, or going to the bathroom individually, while 38 (48.7%) had difficulty with functional mobility. Furthermore, there were 20 (25.6%) who had problems going to school, 15 (19.2%) who had trouble engaging with friends and family, 12 (15.4%) who had difficulty speaking or communicating, and seven (9.0%) who reported depression or shame (Table 2).

Table 2. Disabilities identified in subjects with suspected TBI events (n = 78)

Disability* (frequency, proportion)	n (%)	Severity of disability (frequency)
		Mild	Moderate	Severe
Depression or shame	7 (9.1)	1	2	4
Difficulty picking things up or using their hand	25 (33)	2	9	14
Have trouble going to school	20 (26)	1	6	13
Difficulty standing or walking	38 (49)	9	14	15
Have trouble leaving the home, traveling, or shopping	35 (45)	2	13	20
Need assistance dressing, eating, or going to the bathroom	39 (51)	5	14	20
Difficulty speaking or communicating	12 (16)	1	3	8
Unable to engage with friends and family	15 (19)	1	6	8
Weakness, shortness of breath, and fatigue	10 (13)	1	6	3

^aCategories of disability synthesized using the SOSAS tool and the WHO’s Guidelines for Conducting Community Surveys on Injuries and Violence.^20,21 Severity classified based on respondent’s personal perception of their disability.

Economic impact

The median total cost of treatment for individuals with suspected TBI was $59.00 (IQR: $16.86–$122.23). In 22 (28.2%) cases of TBI, subjects lost their jobs or stopped going to school as a result of the injury. The median time lost at work or school by caretakers was 14 days (IQR: 7–30 days).

Increased difficulty affording food or rent following the injury was reported by 42 (58%) families (Figure 2). In 23 (29%) cases, a family member was forced to spend savings, 6 (8%) needed to sell their assets, and 27 (35%) were forced to borrow money.

Figure 2. Socioeconomic consequences experienced by injured individuals and/or their families following TBI.

Discussion

This study found 78 suspected TBI events in 8,065 individuals living in Southwest Cameroon over a 12-month span. While our study was powered to estimate the overall injury incidence, the number of TBI events identified would correspond to an incidence of 975.57 per 100,000 (95% CI of 781.41−1,218.0) individuals in this region. Road traffic injuries, primarily in motorcyclists not wearing helmets, were responsible for the majority of cases.

Our findings in Southwest Cameroon suggest a higher estimated incidence of TBI compared to rates previously published in South Africa (Figure 3). The Global Burden of Disease Study reported an incidence of 326 per 100,000 people in all of sub-Saharan Africa, specifically 320 per 100,000 in Cameroon, and Congo, the neighboring country with the highest incidence, at 356 per 100,000 (30). The majority of the studies contributing to estimates of TBI incidence in LMICs rely on hospital-based analyses and are unlikely to include those who never seek medical attention, as well as those who die before reaching the hospital.

Figure 3. Comparison of the estimated incidence of suspected TBI in Cameroon compared with other countries according to the Global Burden of Disease study 2016.

Comparatively, the incidence of TBI in the United States (US) is 333 per 100,000 and 302 per 100,000 in Canada (30). In the US, the leading cause of TBI in 2014 was from unintentional falls, with individuals >75 years old experiencing the highest rate of incidence and mortality. Intentional self-harm was the leading cause of TBI-related death (31). RTI seems to comprise a greater proportion of TBI in Africa compared to the US/Canada (1). Contributing factors may include limited traffic enforcement, safety education, vehicle safety technology, and road environment organization commonly experienced in LMICs (32). However, it is worth noting that two health districts were excluded due to safety concerns; these areas may have had a higher proportion of violence-related TBI that was not captured by this study.

The majority of studies exploring the costs associated with TBI have been conducted in the US or European countries, and existing studies in Africa focus primarily on hospital-associated costs (2,33). This study found notable economic burden on patients and their families following TBI, with significant household time redistributed away from work and school as a result of the injury. After paying for treatment, 58% of patients had difficulty affording basic necessities. In low-income settings, where much of the population may live in poverty, even mild additional economic hardship can lead to significant consequences, such as food insecurity and long-term educational losses. Furthermore, students and age groups making up the bulk of the workforce were disproportionately affected by TBI, which may translate to a loss in country's economic productivity over time (34,35).

There are several ongoing efforts to address TBI on a global scale, ranging from partnerships between academic neurosurgery departments in high-income countries (Duke University, University of Alabama, University of Miami, and others) to internationally funded organizations such as the NIHR Global Health Research Group on Neurotrauma. Preventative measures are likely to play the greatest role in reducing the impact of TBI in LMICs. Increased legislation and education on road traffic are necessary. Helmet laws have had clear cost–benefit ratios in the USA and notable decreases in head injuries and death in LMICs when properly enforced (36,37). A paucity of helmet use was reflected in our study; 85.3% of subjects injured while on a motorcycle were not wearing a helmet. RTI without helmet usage has been found to be associated with a 6.90-fold increase in odds of presenting with a moderate-to-severe Glasgow Coma Scale score and 3.76-fold increased odds of needing neurosurgery (38). Persisting neurologic disabilities have been reported in as high as 54% of patients in this setting (39). Multifaceted strategies that include enforcement, training, and other elements are necessary; simple passage of legislation has not been found to be sufficient (40,41). Successful injury prevention efforts are often multifactorial, involving effective implementation in several sectors, including environmental engineering, urban planning, legislation, education, and enforcement, among others.

There have been other previously studied interventions found to be effective in reducing rates of injury in LMICs. In Kampala, Uganda, increased traffic patrol deployment was associated with a 17% drop in road deaths and was also found to be cost-effective (42). Educational interventions with a focus on cyclist injury rates in Brazil and Thailand have resulted in reduced overall injuries and increased helmet use (43). Simple infrastructure changes can have notable effects as well; implementation of speed bumps in residential areas in two neighborhoods in South Africa was associated with a drop in pedestrian–vehicle collisions of 22–23% and a decrease in fatal collisions by 50–68% (44).

Despite aggressive prevention measures, improvement of neurosurgical capacity in LMICs is also necessary to fully address the burden of TBI. In Africa, the neurosurgical workforce must increase by greater than 1700% to meet the demand of regional cases (6). Improvement of surgical infrastructure and government awareness of the need for accessible neurosurgical care is vital to improving the outcomes of patients with TBI in this setting (45).

Study limitations

This study had several limitations. Given the necessity of using a community-based design to capture events that typically are not included in hospital data sources, reports of TBI were based on clinical symptoms reported by patients and/or family rather than medical records or examination; this may result in underreporting of TBI. Also, since information was gathered through familial report, there may be recall bias resulting in further underreporting of seemingly minor injuries and inaccuracy in disability details. Also, while we queried respondents about deaths secondary to injury, deaths may not have been recognized as secondary to TBI in the community setting. Our study was also limited to certain health districts due to safety concerns; thus, regions with greater rates of TBI may have been missed. Finally, the study power was originally calculated for injury rather than TBI specifically.

Conclusions

This study found 78 suspected cases of TBI in 8,065 individuals, suggesting an estimated incidence of TBI in the Southwest Region of Cameroon of 975.57 per 100,000 individuals, notably higher than reported in prior studies. The majority of TBI in our cohort occurred due to RTI, with patients and their families suffering from significant economic burden following TBI. Our findings underscore the need for increased investment in injury prevention efforts designed for LMIC contexts and in regional neurosurgical capacity to reduce the burden of TBI in these settings.

Disclosure statement

The authors have nothing to declare. S. Ariane Christie was supported by the Association for Academic Surgery (AAS) Global Surgery Research Fellowship Award.

Additional information

Funding

S. Ariane Christie was supported by the Association for Academic Surgery (AAS) Global Surgery Research Fellowship Award.