Assessment of substance use among injured persons seeking emergency care in Nairobi, Kenya.

Introduction: Trauma is a leading cause of morbidity and mortality in Kenya. In many countries, substance use is common among patients presenting with injuries to an emergency center (EC). Objective: To describe the epidemiology of self-reported substance use among adult injured patients seeking ED care in Nairobi, Kenya.
Methods: This prospective cross-sectional study, assessed patients presenting with injuries to the Kenyatta National Hospital ED in Nairobi, Kenya from March through June of 2021. Data on substance use, injury characteristics and ED disposition were collected. Substances of interest were alcohol, stimulants, marijuana, and opiates. The Alcohol Use Disorders Identification Test-Concise (AUDIT-C) tool was used to characterize hazardous alcohol use.
Results: A total of 1,282 patients were screened for participation, of which 646 were enrolled. Among participants, 322 (49.8%) reported substance use in the past month (AUDIT-C positive, stimulants, opiates, and/or marijuana). Hazardous alcohol use was reported by 271 (42.0%) patients who screened positive with AUDIT-C. Polysubstance use, (≥2 substances) was reported by 87 participants in the past month. Median time from injury to ED arrival was 13.1 h for all enrolees, and this number was significantly higher among substance users (median 15.4 h, IQR 5.5 - 25.5; p = 0.029). Conclusions: In the population studied, reported substance use was common with a substantial proportion of injured persons screening positive for hazardous alcohol use. Those with substance use had later presentations for injury care. These data suggest that ED programming for substance use disorder screening and care linkage could be impactful in the study setting.

African relevance

Alcohol use, as well as other substance use, is associate with injury, and subsequently seeking acute care. There is currently limited data from injured acute care patients in Africa and Kenya on hazardous alcohol use or polysubstance use.

There is an urgent need for research into how to best develop and implement substance use assistance interventions in the emergency care setting in sub-Saharan Africa.

Current data demonstrate that a history of alcohol use was common among injured patients in Nairobi, Kenya. These data, along with the burdens of use for other dangerous substances, suggest that the ED venue may be an impactful environment to access and provide care for persons with substance use disorders.

Introduction

Trauma is a leading cause of morbidity and mortality around the world, with particularly high impacts across Africa [1]. In Kenya, which has a developing emergency care system, one in eight adults are injured on an annual basis and injuries are among the most common causes for seeking acute medical attention [2], [3], [4], [5], [6]. At Kenya's largest hospital emergency center (Kenyatta National Hospital, KNH), prior work has shown that 25% of encounters are related to injuries [7].

Alcohol use is a known associated risk factor for injury, and use or abuse has been identified among Kenyan patients who suffer injuries [8], [9], [10], [11]. More broadly, substance use beyond alcohol, such as stimulants (including khat, cocaine, and methamphetamines), marijuana, and/or opiates, have been seen to be common among injured emergency center (EC) patients in multiple settings [12], [13], [14], [15], [16], [17]. There is limited data from injured EC patients in Kenya on hazardous alcohol use, or the use of two or more substances, polysubstance use. Prior work has not evaluated the role of self-reported substance use (as opposed to acute intoxication) and its potential role in screening for substance use and implementing interventions for change in EC settings in Kenya. This analysis evaluates self-reported substance use among traumatically injured patients in a national referral hospital emergency center in Nairobi, Kenya.

Methods

This study is a cross-sectional analysis of patients presenting with injuries to the KNH EC. This is an a priori secondary analysis of prospective data collected from March to June 2021 from persons with injuries seeking EC care [18]. KNH is the largest hospital in Kenya, the main public hospital serving the nation's capital, and it is also an important receiving center for referrals from across the country. The EC functions 24-hours a day and has continually available diagnostic and treatment resources including surgical specialists for injured patients. All participants enrolled in the study received standard care for injuries at the discretion of their treating providers in the EC, who were not members of the research team. The study was approved by the KNH Ethics and Research Committee and the Institutional Review Board of Rhode Island Hospital.

Patients with injuries at the KNH EC were screened 24-hours a day by research staff as close to time of arrival as possible during the enrolment period. Designation of a patient as a trauma case occurred at time of triage by the non-study affiliated clinical personnel at the EC who use the South African Triage Scale, which includes a variable indicating if the patient is seeking care for trauma [19]. All participants provided informed consent in either Kiswahili or English. Patients were eligible for inclusion if ≥18 years of age, seeking EC care for a physical injury, and both able and willing to provide informed consent. Exclusion criteria included: age <18 years, seeking care for non-injury reasons, known to be pregnant, a legal prisoner of the state, and unable or unwilling to provide informed consent. Patients with altered mental status or clinical instability on arrival who did not have sufficient awareness to make decisions were excluded given their inability to provide informed consent, however these patients were reassessed throughout their EC course for screening and enrolment if their clinical state allowed. Participants were compensated for study participation.

Data on demographics, medical history, and substance use were collected at enrollment. Patients were then followed throughout the EC course by study staff. Data were again collected at the time of completion of EC care, regarding disposition and identified injuries. Data were collected using password protected electronic tablets and study data were collected and managed using REDCap electronic data capture tools hosted at Brown University [20,21]. Data sources in EC included direct participant questioning and EC medical record review to identify details of identified injuries.

Substance use was based on self-report of any predefined substance of interest used at least one time within 30-days of presentation for injury care. The preceding thirty day timeframe is frequently used in research evaluating individual substance use, including in the U.S.A., Africa, and in Kenya in validated tools such as the Global School-Based Student Health Survey [22], [23], [24], [25], [26], [27]. Substance of interest were alcohol, stimulants (khat, cocaine, or methamphetamines), marijuana, and opiates. To characterize alcohol use, the 3-question Alcohol Use Disorders Identification Test-Concise (AUDIT-C) tool was used; AUDIT-C has been validated as an appropriate screening tool for hazardous drinking in a variety of settings, including in the emergency center among trauma patients, as well as in Kenya [15,[28], [29], [30], [31]]. An AUDIT-C score is considered positive for hazardous alcohol use in males with a score ≥4, a score ≥3 in females [32].

Data analysis was completed using STATA Statistical Software Release 16.0 (College Station, Texas, USA). Descriptive analysis was performed for the population stratified based on substance use categories. Comparisons were made between patients reporting no substance use and independently compared to those reporting one or more substance(s) and those reporting two or more substances (polysubstance use) using chi-squared, Fisher's exact nonparametric testing or Wilcoxon rank sum as appropriate based on observation frequency and conformity with normality of distribution. Patient's alcohol use was assessed; only those patients with identified hazardous drinking using the AUDIT-C tool were considered as alcohol substance users.

Results

During the study enrolment period, 1282 patients presented for injury care to the KNH EC who were screened for participation. Among those, 563 (43.9%) did not meet inclusion criteria and 73 (5.7%) declined participation. A total 646 participants were enrolled (Fig. 1).

Fig. 1

Patient enrolment.

Of the 646 enrolled patients, 567 were male (87.8%) and the median age was 29 years (IQR: 25 - 37 years). Characteristics of patients reporting no substance use in the past month against those with reported ≥1 substance use (Table 1), and those with no substance use and those with reported ≥2 substances (Table 2) are outlined below. Substance users were statistically more likely to be male (≥1 substance p≤0.001, ≥2 substance p = 0.047). Age was similar across all groups. Both categories of substance users were less likely to have professional jobs and were also less likely to have full time employment. There was a statistical difference in the vocation and education levels between those with no and ≥1 substance use (vocation p = 0.041, education p = 0.002) and those with no and ≥2 substance use (vocation p = 0.005, education p≤0.001). Less than a quarter of all enrolled patients had attained post-secondary education.

Table 1

Characteristics, no substance use vs ≥1 substance use.

	Non-Substance	≥1 Substance	p =
Number	324	322
Demographicsact
Age, years (Median, IQR)	28, 24–37	30, 25–37	0.121
Male (No,%)	268 (82.7%)	299 (92.9%)	<0.001
Vocation (No,%)			0.041
Full Time Professional	37 (11.4)	24 (7.6)
Full Time Laborer	103 (31.8)	99 (31.1)
Part Time Professional	12 (3.7)	5 (1.6)
Part Time Laborer	46 (14.2)	75 (23.6)
Self Employed	68 (21.0)	61 (19.2)
Unemployed	57 (17.6)	53 (16.7)
Missing	1	5
Highest Education Level (No,%)			0.002
Post Secondary	70 (21.6%)	61 (18.9%)
Secondary	149 (46%)	114 (35.4%)
Primary	105 (32.4%)	147 (45.7%)
Injury and Referral
Transferred from Outside Facility (No,%)	146 (45.3%)	126 (39.1%)	0.111
Hours, Injury to Arrival (Median, IQR)	10.2 (4.5, 24.0)	15.4 (5.5, 25.5)	0.029
Disposition from KNH ED (No,%)			0.250
Discharged	177 (54.6%)	184 (57.3%)
Admitted	139 (42.9%)	131 (40.8%)
Transferred from KNH	5 (1.5)	1 (0.3)
Eloped	2 (0.6)	5 (1.6)
Deceased in ED	1 (0.3)	0 (0)
Missing	0	1

Table 2

Characteristics, no substance use vs ≥2 substance (polysubstance) use.

	Non-Substance	≥2 Substance	p =
Number	324	87
Demographics
Age, years (Median, IQR)]	28, 24–37	28, 24–35	0.048
Male (No,%)	268 (82.7%)	82 (94.3)	0.047
Vocation (No,%)			0.005
Full Time Professional	37 (11.4)	5 (5.9%)
Full Time Laborer	103 (31.8)	21 (24.7%)
Part Time Professional	12 (3.7)	0 (0)
Part Time Laborer	46 (14.2)	23 (27.1%)
Self Employed	68 (21.0)	12 (14.1%)
Unemployed	57 (17.6)	23 (27.1%)
Missing	1	3
Highest Education Level (No,%)			<0.001
Post Secondary	70 (21.6%)	13 (14.9%)
Secondary	149 (46%)	23 (26.4%)
Primary	105 (32.4%)	51 (58.6%)
Injury and Referral
Transferred from Outside Facility (No,%)	146 (45.3%)	31 (35.6%)	0.18
Hours, Injury to Arrival (Median, IQR)	10.2 (4.5, 24.0)	16.7 (6.0, 27.0)	0.293
Disposition from KNH ED (No,%)			0.51
Discharged	177 (54.6%)	51 (59.3%)
Admitted	139 (42.9%)	33 (38.4%)
Transferred from KNH	5 (1.5)	0 (0)
Eloped	2 (0.6)	2 (2.3)
Deceased in ED	1 (0.3)	0 (0)
Missing	0	1

History of arriving directly to the KNH EC versus being transferred from an outside facility was not significant between substance users and non-users (p = 0.111). The time from injury occurrence to KNH EC arrival was 13.1 h for all enrolees, and significantly longer for ≥1 substance users (median 15.4 h, IQR 5.5 - 25.5) than non-substance users (median 10.2 h, IQR 4.5 - 24.0; p = 0.029). The disposition destination from the EC was not significantly different between non-substance users, ≥1 substance users, and polysubstance users (Tables 1 and 2; p = 0.250 and p = 0.51, respectively).

Among enrolees, 322 patients (49.8%) reported some type of substance use in the past month. Patients also reported using stimulants (n = 79), marijuana (n = 71), and opiates (n = 6). Among those reporting substance use, 87 participants reported polysubstance use. The most common reported polysubstance combination was stimulant use reported by individuals who were AUDIT-C positive for hazardous alcohol use (n = 57).

Among all enrolled patients, 271 patients (42.0% of enrolees, 248 males and 23 females) reported alcohol use characteristics that yielded AUDIT-C positive outcomes for hazardous alcohol use (Table 3). There was not a significant difference between AUDIT-C positive males and females (p = 0.055). Of note, 10.5% of all enrolees reported having ≥6 alcohol drinks on one occasion daily, while a larger proportion (34.4%) reported ≥6 alcoholic drinks on one occasion at least weekly.

Table 3

Alcohol use disorders identification test-concise parameters (n = 314).

	No. (%)
How often do you drink alcohol?
Once a month	74 (23.6)
2–4 times a month	103 (32.8)
2–3 times a week	94 (29.9)
4 or more times a week	41 (13.1)
Missing	2 (0.6)
How much alcohol (# drinks) do you have on a day when you drink?
1–2	178 (56.7)
3–4	96 (30.6)
5–6	26 (8.3)
7–9	8 (2.6)
≥10	4 (1.2)
Missing	2 (0.6)
How often do you drink ≥6 drinks on one occasion?
Daily	33 (10.5)
Once per week	108 (34.4)
Once per month	71 (22.6)
Less than monthly	15 (4.8)
Never	86 (27.4)
Missing	1 (0.3)
AUDIT-C Positive
Male (score >=4)	248 (91.5)
Female (score >=3)	23 (8.5)

Discussion

Among the studied injured persons in the emergency care setting in Kenya, nearly half reported substance use in the past month, with over one in eight reporting polysubstance use during that time frame. This observed burden in the injured population suggests that the EC venue may be an impactful environment to access and provide care for substance use disorders in Kenya and other similar settings. Furthermore as the EC treatment period has been documented as a setting in which health promotion and readiness to change is present, interventions for substance use may be both feasible and efficacious [33], [34], [35]. However, there is an urgent need for research into how to best develop and implement substance use assistance interventions in the EC setting in sub-Saharan Africa.

Alcohol use was the most common substance reported by the enrolled injured persons seeking care. There is limited prior research evaluating alcohol or drug use among injured patients in Kenya. Work by Odero et al. in 1995 found that 16.5% of their patients reported alcohol use within 12 h of injury and he found in 2007 that 32.8% of EC injury patients were suspected of being under the influence of alcohol preceding the injury based on the detection of “the smell of alcohol on the subjects’ breath” [9,36]. A survey among KNH road traffic injured patients in 2005 found an alcohol use prevalence 26.3% [37]. The current data found 42.0% of all enrolees were AUDIT-C positive, indicating a high percent of all trauma patients at the KNH EC have behavior consistent with hazardous drinking and/or active alcohol use disorder. Further, there were very high levels of reported binge drinking, with 44.9% of those with alcohol use in the past month reporting ≥6 alcoholic drinks either daily or weekly. There no universal definition of binge drinking, though most commonly it is defined by the number of alcoholic drinks consumed in one sitting: ≥5 for men and ≥4 for women [38,39]. Alternatively, the AUDIT screening tool uses a definition of ≥6 drinks in one sitting for men and women [39]. Our study's high frequency of binge drinking suggests that the injured population is at high-risk and injured EC patients in Kenya may benefit from interventions to support reduction in alcohol misuse.

While the prior work by Odero and colleagues in Kenyan ECs was around alcohol intoxication at the time of injury, substance use immediately preceding injury only captures a portion of the risk of future injury associated with substance use. In the current data the frequency of substance use over the past month was used to better understand the burden of substance use among EC patients. This wider scope of understanding of substance use, presenting in the EC at one point in time, provides the opportunity to intervene and aim to reduce harm and prevent future injury, even if and when substance use may not have contributed to the index care presentation. As a consequence, every emergency care encounter presents an opportunity for both brief interventions and linkages to care (including efforts towards harm reduction and cessation) [40], [41], [42], [43]. Given the high substance use burdens observed among the population studied, there is a need to develop targeted interventions and referral pathways, both within the Kenyan EC context and in other low-and middle-income countries).

Previous work among trauma patients in Kenya has shown significant delays in presenting to the EC for care. Delays to definitive care can be exacerbated when patients initially present to lower-level health facilities without the capacity to manage significant injuries. Generally, wide ranges in prior studies indicate significant heterogeneity in EC arrival time from injury [3,44,45]. In this study the time of injury to EC arrival was 13.1 h for all patients, with significantly longer delays among those reporting substance use. It is unknown what specifically led to the longer delays in presenting to KNH EC for trauma care among those with substance use, however it is possible that the substance users were less readily identified as needing transfer, which could impact the time to care access of trauma care in an already high-risk sub-population. Additionally, it is possible that there was a delay in the time to initial presentation for injury care due to patient intoxication. Although KNH is a national referral hospital, delays so much longer than previously reported in this and other large hospitals in and around Nairobi are likely multifactorial, and possibly complicated by the SARS CoV-2 pandemic which was ongoing during the time of data collection, but not easily or clearly identified. Future research in which prior and current substance use data is obtained would be beneficial to better understand delays in care as it related to substance use in the injured population.

Study inclusion required individual participants to provide their own informed consent. As such, patients with persistent altered mental status were excluded from enrollment (with particular concern for patients who may have been acutely intoxicated and/or critically injured). Therefore this study was not able to capture the entire population of substance users in the KNH EC during the study period and may have underestimated the burden of use. In several of the study questions, substance use was reported on the interval of “in the past month” and even self-reported substance use in the hours leading up to a patient's injury was not evaluated. Our findings may suffer from an element of either recall and/or social desirability bias. Given the nature of the data of interest pertaining to substance use if this type of bias was present it would most likely bias toward the null and the estimates may be an underestimation of the true burden of substance use in the population. Lastly, this study was performed during the pandemic caused by SARS CoV-2, and the findings may be less generalizable to other non-pandemic conditions. However as there has been evidence of increased alcohol and substance use during the SARS CoV-2 pandemic (as well as in prior epidemics) the results demonstrating high use in the population studied are consistent with the existing literature [46], [47], [48].

Conclusion

There is a gap in substance use research from high-risk emergency care settings in Africa such as Kenya. The current data demonstrate that a history of alcohol use was common and that there was a high prevalence of hazardous alcohol use based on the validated AUDIT-C tool among the injured population seeking care. These data along with the burdens of use for other dangerous substances suggest that the EC venue may be an impactful environment to access and provide care for persons with substance use disorders. Further study on substance misuse in the EC setting among injured persons are needed to inform development and implementation of substance use assistance interventions in Kenya.

CRediT authorship contribution statement

J． Austin Lee: Conceptualization, Formal analysis, Writing – original draft. Eric O. Ochola: Data curation, Software, Methodology. Janet Sugut: Investigation, Project administration, Writing – review & editing. Beatrice Ngila: Investigation, Project administration. Daniel K. Ojuka: Investigation, Writing – review & editing. Michael J. Mello: Supervision, Writing – review & editing. Adam R. Aluisio: Funding acquisition, Conceptualization, Supervision, Writing – review & editing.

Declaration of Competing Interests

The authors declare no conflicts of interest.