The Utility of Alcohol Saliva Test Strips Compared to the Breathalyzer in Trauma Patients in a Resource-Limited Setting

Abstract

Introduction:

The correlation between alcohol consumption and injury is undeniable. However, past research relying on self-reporting alcohol use likely resulted in underreporting, emphasizing the need to increase alcohol testing, especially in resource-limited settings where the burden of injuries is highest.

Methods:

This is a prospective analysis of injured patients presenting to the trauma center at Kamuzu Central Hospital in Lilongwe, Malawi. We collected information including patient age, sex, admission date, mechanism of injury, breathalyzer test and Rapid Response^™ Alcohol Saliva Test Strips (AST) result, and survival.

Results:

A total of 805 trauma patients were included. The overall prevalence of alcohol consumption in this trauma cohort is 18.3%. There was a 95.5% agreement between the AST and breathalyzer test with a Kappa coefficient of 0.83. The sensitivity and specificity of the AST were determined to be 78.5% (CI 75.7 to 81.2) and 99.3% (CI 98.7 to 99.9), respectively. ROC analyses showed the AST to have excellent discrimination with an area under the curve of 0.88 (95% CI 0.85–0.92).

Conclusion:

The prevalence of alcohol-related injury is high in Malawi, and the use of the Alcohol Saliva Test Strips is feasible and correlated with results derived from the breathalyzer. Routine alcohol testing for trauma patients presenting to a resource-limited setting is imperative and should be implemented.

Introduction

The global mortality burden attributable to trauma is more significant than any other single cause in the most economically productive ages of 15–45 years global working population and more than all contagious diseases added together, including COVID-19.[1] There are 40 million people permanently injured per year and up to 100 million temporarily injured. The number of people injured, temporarily or permanently, is greater than any other medical condition. Between 80 and 90% of all traumatic injuries occur in low- and middle-income countries (LMICs). Given that the majority of the population in most Low-Income Countries (LICs) is under the age of 35, this affects the working population disproportionately in those countries. [2,3]

Alcohol is associated with almost all types and mechanisms of injury. The correlation between alcohol consumption and injury is undeniable, as it adversely impacts cognitive and motor function, escalating the risk of intoxicated individuals participating in acts of violence.[4] However, past research relying on self-reporting alcohol use likely resulted in underreporting, emphasizing the need to increase testing overall. [5,6]

In low-income countries, the surge in road traffic injuries is expected to rise due to increasing economic growth, motorization, and heightened alcohol use, resulting in a potential rise in drunk driving incidents.[7] Compounding this challenge are large proportions of vulnerable road users such as pedestrians and cyclists, poor road infrastructure, insufficient law enforcement, and inadequate technical standards on vehicles, collectively increasing the traffic crash risk for those under the influence of alcohol in these settings.

In sub-Saharan Africa, particularly Malawi, there is a limited understanding of alcohol use among those sustaining injuries by any mechanism. Despite Malawi being regarded as a country with relatively low alcohol consumption, it has been noted to have a very high rate of road traffic deaths at 31/100,000 inhabitants. [3] Notably, blood alcohol screening is not routinely measured in the emergency room, presenting a critical gap in knowledge during medical management. Non-invasive methods for quantitatively estimating BAC have primarily used breath testing. Although a breath analyzer provides a rapid result, it requires regular calibration and patient cooperation, which may be difficult for combative or unconscious patients, and they are not currently available in most hospitals in sub–Saharan Africa. To address this gap, implementing alcohol screening for trauma victims in Malawi becomes imperative, with the potential integration of affordable saliva testing strips offering a pragmatic solution in low-income settings where traditional methods such as blood alcohol levels or breathalyzers are unavailable. We, therefore, aimed to determine the sensitivity and specificity of alcohol saliva strip testing using the breathalyzer as the pragmatic gold standard in a resource-limited setting.

Methods

This study is based at Kamuzu Central Hospital (KCH), a 1000-bed tertiary referral center serving approximately 9 million people in central Malawi. A team of attendings, residents, and clinical officers provides surgical care at KCH in the emergency department, wards, and operating rooms. We performed a prospective analysis of patients presenting to the trauma center at Kamuzu Central Hospital in Lilongwe, Malawi, between March 2023 and October 2023. We collected information including patient age, sex, date of admission, mechanism of injury, time to presentation to hospital, breathalyzer test result, alcohol saliva test result, comorbidities, date of discharge, length of hospital stay, and survival.

We utilized the track S80 breathalyzer to determine Blood Alcohol Concentration (BAC). A mouthpiece and two chambers filled with liquid are connected to a meter that assesses color changes. The patient is asked to exhale into the mouthpiece. The exhaled air enters a chamber filled with red-orange potassium dichromate and other chemicals. The alcohol in the exhaled breath reacts with the potassium dichromate, changing the color from red-orange to green. The degree of color change is related to the alcohol level in your exhaled breath. A photocell then compares the color difference. This color difference generates an electrical current, which is then converted to a numerical value for the BAC.

We also utilized the Rapid Response^™ Alcohol Saliva Test Strips BTNX^inc, identifying alcohol levels at 0%, 0.02%, 0.04%, 0.08%, and 0.3% for this study. The alcohol saliva strip contains Tetramethylbenzidine (TMB) 0.12mg, Alcohol Oxidase 0.5 IU, Peroxidase 0.35 IU, and Proteins 0.15mg. The AST strip is based on the high specificity of alcohol oxidase (ALOx) for ethyl alcohol in the presence of peroxidase and enzyme substrate such as tetramethylbenzidine (TMB), as shown in the following:

$$\text{EtOH}+\text{TMB ALOx}/\text{Peroxidase CH}3\text{CHO}+\text{Colored TMB}$$

NEGATIVE: No color change appears on the reaction pad. The color should match the color block on the pouch corresponding to a negative (‐) result. This indicates that alcohol has not been detected. POSITIVE: A color change appears on the reaction pad. The BAC ranges from 0.02%, 0.04%, 0.08% to 0.30%, with the color on the reaction pad varying from a light blue to a dark blue, falling on or between the corresponding color blocks on the pouch

The Rapid Response^™ Alcohol Test Strips were qualitatively verified using a test solution prepared by adding 5 drops of 80-proof distilled spirits to 30 mL of water. This solution should produce a color change on the reaction pad corresponding to a 0.02% or greater BAC. The color reaction with alcohol in saliva is somewhat slower and less intense than with alcohol in an aqueous solution. [8]

All trauma patients ≥18 years who sustained injuries following motor vehicular collision or interpersonal violence were invited to participate in the study. We limited our study to this trauma population as we have previously shown the correlation between MVC and Interpersonal violence with self-reported alcohol use in Malawi. [9,10] Every consented patient received both a breathalyzer and saliva tests. A descriptive analysis of the sociodemographic and clinical characteristics of the population was performed. Likewise, the level of concordance (Cohen’s Kappa coefficient) between Rapid Response^™ Alcohol Saliva Test Strips BTNXinc and BAC track S80 breathalyzer (gold standard blood alcohol level) was calculated for clinical validations, as well as sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), the area under the curve (AUC), and accuracy. A negative test was determined if the saliva test was 0%, and a positive test was determined if the saliva test was 0.02% or above. Receiver operating characteristic (ROC) analyses were performed to evaluate the performance of the Rapid Response^™ Alcohol Test Strips.

We utilized StataCorp v18.0, College Station, Texas, for all statistical analyses. The Malawi National Health Science Research Committee and the University of North Carolina Institutional Review Board approved this study.

Results

In our analysis, a total of 805 trauma patients were included. The median age was 32 years (IQR 31.2–32.7), with a male preponderance of 753 (87%). Of the enrolled patients, 768 (89%) received treatment in the emergency department and were discharged, while 88 (10.2%) were admitted to an inpatient ward. Only five patients were admitted to the Surgical High Dependency Unit (0.58%), and two reported fatalities in the emergency department (0.23%). (Table 1). The overall prevalence of alcohol-related injury in this cohort is 18.3%. (Table 2)

Table 1.

Demographic and patient characteristics

	Patients N=864
Age, mean in years ± standard deviation	31.9 ± 11
Sex
Male, no. (%)	753 (87.4%)
Female, no. (%)	109 (12.65%)
Mechanism
Assault, no. (%)	255 (29.6%)
Motor Vehicle Accident, no. (%)	608 (70.5%)
Outcome
Outpatient Department, no. (%)	768 (88.9%)
Admitted to the ward, no. (%)	88 (10.2%)
Admitted to the Surgical HDU*, no. (%)	5 (0.58%)
Died in casualty, no. (%)	2 (0.23%)

^*HDU High Dependency Unit

Table 2.

Alcohol Saliva test and breathalyzer test results of the participants

Test	Positive No. (%)	Negative No. (%)
Saliva	129 (14.9%)	735 (85.1%)
Breathalyzer	158 (18.3%)	706 (81.7%)

The majority of the saliva tests had negative results at 733 (85%) in comparison to 707 (82%) of patients testing negative via breathalyzer. Positive results for a 0.02% alcohol level were observed in 87 patients (10%) through saliva tests, whereas only 10 patients (1.1%) showed a positive result on the breathalyzer. Similarly, 22 patients (2.55%) had a positive result for a 0.04% alcohol level in saliva tests, while 26 (3%) tested positive on the breathalyzer. For a 0.08% alcohol level, 20 patients (2.3%) had a positive result in saliva tests, as opposed to 13 patients (1.5%) with a positive result on the breathalyzer. There was a 95.5% agreement between the AST and Breathalyzer test with a Kappa coefficient of 0.83.

The sensitivity of saliva alcohol test strips was 78.5% (CI 75.7 to 81.2), with a positive predictive value of 96.1% (CI 94.8 to 97.4). On the other hand, the specificity of saliva alcohol test strips was 99.3% (CI 98.7 to 99.9), with a negative predictive value of 95.4% (CI 93.97 to 96.77%). (Table 3). ROC analyses showed the Rapid Response^™ Alcohol Saliva Test Strips to have excellent discrimination with an area under the curve (AUC) of 0.88 (95% CI 0.85–0.92). (Figure 1). The overall accuracy of the Alcohol Saliva Test Strips is 95.4%.

Table 3.

Sensitivity and specificity of using Alcohol Saliva Test

	Percentage (Confidence Interval)
Sensitivity	78.48% (75.74% to 81.22%)
Specificity	99.29% (98.73% to 99.88%)
PPV*	96.12% (94.84% to 97.41%)
NPV*	95.37% (93.97% to 96.77%)

^*Prevalence of 18.29% (15.71% to 20.86%)

^*Positive Predictive Value

^*Negative Predictive Value

Figure 1.

ROC curve for Saliva Alcohol diagnostic test.

Discussion

Alcohol production and use have been part of the social and cultural fabric of all societies. With misuse and abuse, alcohol consumption has become an important contributor to the global burden of disease. Indeed, unintentional and intentional injuries are the cause of almost half (46%) of the global burden of alcohol-related mortality. This study shows the high utility and accuracy of the saliva alcohol test strips in a resource-limited setting with a homogenous Black population with a sensitivity and specificity of 78.5 and 99.3%. This study shows that the Rapid Response^™ Alcohol Saliva Strips Test (AST) can be used to determine the alcohol status of trauma patients. Our results complement previous reports that this test is accurate compared to serum BAC and breath BAC measurements. [11–13] Although the precision of the AST is not equivalent to serum determination, it can provide an accurate estimate of BAC in most trauma patients, particularly in resource-limited settings where alcohol breathalyzers may not be available.

In a national household survey in Malawi, the prevalence of alcohol consumption among people aged 24–64 stands at 30.1% for males and 4.1% for females. This survey also found that 19.2 % and 2.3% of males and females, respectively, are considered to be heavy drinkers. [14] We found in this study that the prevalence of positive alcohol intake before trauma was 18.3%. This finding is in contrast to the alcohol prevalence in trauma patients presenting to a Level 1 trauma center in the United States at 8.9%. [15] Our results are similar to a study performed in Malawi, which showed a prevalence of 20% in adult trauma patients [16] but was significantly lower than a study conducted in Tanzania, which had a prevalence rate of 30%.[17]

The burden of alcohol consumption is not distributed equally between countries and regions. Three main determinants affect it: the proportion of people of Muslim faith, socioeconomic status, and the prevalence of HIV/AIDS. While religion can contribute to lower levels of alcohol-related harm, both economic wealth and the prevalence of HIV/AIDS contribute to increases in alcohol-related harm. Malawi has one of the highest HIV prevalence in the world. Studies show key patterns of interaction between alcohol use and sexual behavior, therefore posing greater risks of sexually transmitted infections (STIs), including HIV infection.[18]

The advantages of AST include 1) results are not influenced by the presence of blood in the oral cavity, 2) the non-invasive nature of AST minimizes the risk of needle stick injuries for staff and multiple needle punctures for patients, 3) AST provides a determination of the BAC within 5 minutes, 4) it could also be used in determining postmortem saliva ethanol levels and 5) it can be utilized in obtunded patients. Lastly, because of the relatively low cost of the AST, the saliva test is a cost-effective alternative in public health settings where mildly to moderately intoxicated persons are encountered. [19] In addition, the AST is a one-time strip that does not require equipment cleaning and maintenance over time.

Alcohol saliva strip test (AST) is used to determine a BAC of 0.02% or higher via saliva, which is unique in providing on-the-spot, quantitative results. However, it has some limitations, such as a positive result indicating only the presence of alcohol and does not indicate or measure intoxication. Technical or procedural errors are possible, and other intra-oral substances may interfere with the test, leading to erroneous results.

Nevertheless, the AST still has good reliability, validity, and accuracy in the non-invasive, quantitative estimation of BAC. The AST could serve as an inexpensive and practical tool in the clinical setting to capture patients who repeatedly present with alcohol-related injuries for further interventions. Furthermore, the AST could also be a valuable tool in the enforcement of traffic laws to reduce alcohol-related harms in LMICs where the burden of injuries is highest. Additional data generated from the implementation of AST as a routine alcohol screen could contribute to the development of a national alcohol policy for Malawi.

Conclusion

The prevalence of alcohol-related injury is high in Malawi, and the use of the Alcohol Saliva Test Strips is feasible, effective, and highly correlated with results derived from the breathalyzer. Routine alcohol testing for trauma patients presenting to a resource-limited setting is imperative and should be implemented. This study reinforces the need for a national alcohol policy for reducing alcohol-attributable harms.