Skip to main content
NCBI home page
Journal of Studies on Alcohol and Drugs logoLink to Journal of Studies on Alcohol and Drugs
. 2022 Sep 21;83(5):760–767. doi: 10.15288/jsad.21-00141

Utility of Family Reports in Predicting Emergency Department Patient Alcohol Use in Tanzania

Arthi S Kozhumam a, Carter Lovvorn a, Paige O’Leary a, Linda Minja b, Judith Boshe c,,d, João Ricardo Nickenig Vissoci a,,e, Blandina T Mmbaga b,,c,,d, Catherine A Staton a,,e,,*
PMCID: PMC9523754  PMID: 36136447

Abstract

Objective:

Myriad reasons, including stigma, may prevent patients from self-reporting harmful use of alcohol in Tanzania. Family members may be more forthright but might not know the extent of the patient's alcohol use or suffer alcohol-related stigma as well. Our study aims to compare the reporting of patient alcohol use by emergency department patients themselves and their family members in Tanzania in order to describe the potential use of family reports as a proxy for patient self-reports.

Method:

We conducted a secondary descriptive analysis of a prospective cohort of adult patients seeking treatment for injury and their family members. We evaluated alcohol use behavior, alcohol-related consequences, and alcohol-related stigma reported by 231 patients and 231 family members (both majority male, ages 25–45 years), measured by the Alcohol Use Disorders Identification Test (AUDIT), Perceived Alcohol Stigma (PAS) scale, and the Drinker Inventory of Consequences (DrInC). Alcohol use behavior concordance/discordance between patients and families was established, and alcohol use and perceived stigma were analyzed.

Results:

More than 72% of patient–family pairs showed alcohol use (AUDIT) concordance. Receiver operating characteristic curve and regression analysis suggests family reports to be clinically relevant, significant, and potentially accurate markers of patient alcohol use (sensitivity: 95.10%, specificity: 69.77%). Findings support the existence of stigma toward alcohol in this context, with similar stigma levels of patients and family members.

Conclusions:

Family-reported patient alcohol use may be an accurate proxy for patient self-reporting. Further research is needed into stigma toward alcohol that is culturally appropriate and adopted.

Harmful use of alcohol is responsible for 5.1% of the global burden of disease—7.1% for males and 2.2% for females, respectively—and 3 million deaths per year (World Health Organization [WHO], 2018). Although an estimated 1.4% of the global population has an alcohol use disorder (AUD) (Ritchie & Roser, 2018), this burden is not equitably distributed worldwide. Individuals in African countries consume about 13% more alcohol than the global average per capita (Milanzi & Ndasauka, 2019), with heavy drinking contributing to preventable alcohol-related injuries (Hingson et al., 2005, 2009).

Despite the growing burden alcohol places on sub-Saharan Africa, little is known about the true proportion of patients who have consumed alcohol, as there are few surveillance systems in emergency departments (EDs; Staton et al., 2018) and limited self-reporting by patients. On questionnaires such as the Alcohol Use Disorders Identification Test (AUDIT) implemented in northern Tanzania, the prevalence of alcohol use (estimated 47%–70% had ever used alcohol) was found to be higher than WHO and systematic review estimates (Francis et al., 2015). In both 2010 and 2016, 3.2 of 9.7 liters of alcohol consumed per capita by those 15 years or older in Moshi, Tanzania, were unrecorded (World Health Organization, 2018). Significant stigma and negative attitudes toward substance and alcohol use exist in Tanzania (El-Gabri et al., 2020b; Sorsdahl et al., 2012). Although these may prevent ED patients from self-reporting their use, family members may be more honest if they are aware of their family member's drinking habits. It is also possible, however, that multi-informant reports may be discordant because of incorrect family member eyewitness reports or even innate bias (Haeny et al., 2018).

Harmful use of alcohol has been described as a “family illness,” often leading to long-standing interpersonal trauma and psychological harm (Kühn & Slabbert, 2017). Patient and family reporting systems and involvement have been implemented at ambulatory visits (Bourgeois et al., 2019) and error identification (Etchegaray et al., 2014), showing instances of inaccuracies and differences between patient and family reports. Although work by De Los Reyes and others has explored how multi-informant reports impact anxiety and family relations, significant discrepancies found between reports hold the potential for clinical meaning (De Los Reyes & Ohannessian, 2016; De Los Reyes et al., 2013; Makol et al., 2019). At the same time, studies involving multiple reports of one subject's alcohol use are limited.

Given the potential inconsistency in accuracy and forthrightness of alcohol reporting between multi-informant reports, as well as the large impact of alcohol use and potential stigma on family health and life, our study aimed to understand if questioning of alcohol use should include information from family members in order to identify patients with a high risk of alcohol use and evaluate perceived stigma toward alcohol in Tanzania. We aimed to assess the concordance between patient self-reporting and family reporting for alcohol use (AUDIT score) and identify whether variables such as perceived alcohol stigma are associated with either patient or family responses in order to assess the use of family-reported patient AUDIT as a proxy for patient self-report.

Method

Ethics

This study was approved by the Duke University Medical Center Institutional Review Board, Kilimanjaro Christian Medical Center Ethics Committee, and the Tanzanian National Institute of Medical Research.

Study design

Study setting. The study was conducted within Moshi, Tanzania, located in northern Tanzania with a population of more than 180,000 (Tanzania National Bureau of Statistics, 2012). A total of 27.5% of college-age younger men in the northern Tanzania region have an AUDIT score at or above 8, often where the line for an AUD is drawn (Francis et al., 2015), with these patients considered “hazardous drinkers” and/or having AUD (Staton et al., 2018), although cutoff scores for hazardous drinking vary depending on the surveyed location's environment and drinking patterns (Babor & Robaina, 2016). Located in the Kilimanjaro Region, Moshi is home to Kilimanjaro Christian Medical Center (KCMC), a referral hospital for more than 15 million residents. We selected this hospital to assess alcohol reporting because of its central location and service to both rural and urban populations of Moshi.

Study design and population. This is a secondary descriptive analysis of a prospectively collected cohort of patients seeking treatment for an injury at the KCMC Emergency Department and members of their families, following the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines (Vandenbroucke et al., 2007). Family members reported by patients as accompanying them were predominantly biological (n = 220, 95.24%), including mothers (n = 12, 5.45%), fathers (n = 16, 7.27%), aunts (n = 2, 0.91%), sisters (n = 36, 16.36%), brothers (n = 98, 44.55%), husbands/wives (n = 34, 15.45%), sons (n = 14, 6.36%), daughters (n = 4, 1.82%), grandchildren (n = 1, 0.45%), and grandparents (n = 2, 0.91%), with one brother-in-law (n = 1, 0.45%); n = 11 (4.76%) patients selected other individuals including friends and coworkers. All injury patients at KCMC between July 2016 and May 2017 were selected for the study, regardless of alcohol use behavior. Self-reported gender was collected for all study participants. Specific inclusion criteria included ability to provide full informed consent for participation, being clinically sober at the time of enrollment, ability to speak Swahili, and medical stability. Patients younger than 18 or with non-injury complaints were excluded, as were patients who were medically unstable or in deteriorating or critical condition. Injury patients who were illiterate to any degree were given the option to verbally provide consent and were not excluded from participation. Enrolled patients were informed that their accompanying family members would be questioned about the patient's health; however, they were not told that family members would receive the same questionnaires or that their results would be recorded.

Measures and data collection

Measures. The three main constructs evaluated in this project were alcohol use behavior, alcohol-related consequences, and alcohol-related stigma reported by patients or family members. Patients and family members were both given the AUDIT to answer regarding the patient's alcohol use and were administered the Perceived Alcohol Stigma (PAS) scale. In addition, the Drinker Inventory of Consequences (DrInC) scale was administered to patients to evaluate self-perceived consequences of alcohol use.

The AUDIT is an instrument that examines one's harmful use of and dependence on alcohol (Saunders et al., 1993). The AUDIT is a 10-item self-reported scale with scores ranging from 0 to 40. This measure has been validated in multiple countries (Dybek et al., 2006; Moretti-Pires & Corradi-Webster, 2011; Pal et al., 2004; Pradhan et al., 2012). In addition, the Swahili translated version of the AUDIT questionnaire was found to have excellent language clarity and domain coherence, with AUDIT scores significantly lower in nondrinkers than in drinkers (Nickenig Vissoci et al., 2018). AUDIT scores were categorized as AUDIT “+” (ascore greater than 7) and AUDIT “−” (scores between 0 and 7).

The PAS scale, which is an adapted version of the Perceived Devaluation-Discrimination scale, in this adaptation for alcohol use, evaluates perceived alcohol stigma toward alcohol drinkers (Glass et al., 2013b). The PAS has been shown to have strong psychometric properties as well as satisfactory content and face validity (Glass et al., 2013a), with higher PAS scores correlated with a lower likelihood of treatment-seeking behavior (Glass et al., 2013b). This questionnaire consists of 12 questions, with possible scores ranging from 0 to 72, in which the person surveyed responds on a 6-point Likert scale from strongly agree to strongly disagree, with an average score of 3 or greater showing a degree of stigma toward alcohol users (Glass et al., 2013b). Scores were scaled from 0 to 100 for this analysis, with 0 representing no stigma and 100 the highest possible stigma.

The DrInC is composed of 50 yes-or-no questions designed to understand negative consequences of drinking that often correlate with one's well-being and psychosocial functioning (Lee et al., 2010). It is composed of five subscores—interpersonal, physical, social, impulsive, and intrapersonal—to quantify specific consequences. Each sub-score is summed to produce “lifetime” and “past-3-month” measures of consequences, which can be combined for an overall sum (Miller et al., 1995). The DrInC has been found to be significantly correlated with AUDIT scores and alcohol consumption quality when validated in Tanzania (Zhao et al., 2018).

Data collection process. Data were collected by two nurses natively fluent in Swahili who have more than 10 years of experience conducting research on sensitive and often stigmatized topics, including alcohol, sexual activity, and HIV. In preparation for survey administration, these nurses were trained in medical ethics, background information in Swahili on prior use and validation of the questionnaires, and research protocol. Upon the patients’ arrival at the ED and after stabilization, they underwent a screening process including asking for informed consent, determining fitness for inclusion, and administering a preliminary breath alcohol analysis test for sobriety identification. Surveys for the questionnaires were conducted in a private, secluded setting and recorded on paper. They were then recorded in REDCap, an online database (Harris et al., 2009), and double checked for completeness.

Data analysis

We included only patients who self-reported AUDIT scores and had a family member who reported the patient's AUDIT as well. Of 341 patients and 234 family members who were part of our original injury data set, 231 patient–family member pairs of reports were used for this study.

AUDIT concordance was established if categorical values of the AUDIT score matched (i.e., both AUDIT+ or both AUDIT−) between patient and family reports. Discordance was when categorical values of the AUDIT did not match. Given that some participants were not active drinkers, we also created AUDIT categories with “0” for lifetime alcohol abstainers and “1” for those who had consumed an alcoholic drink in their lives.

Descriptive statistics were used to represent the data using medians and interquartile ranges for numeric variables and frequencies for categorical variables. Age and gender were used as confounding variables, and AUDIT+ or AUDIT− categories were defined as the outcome variables of interest. Family-reported AUDIT scores, DrInC scores, and PAS scores were included as predictors. Descriptive data of all variables were depicted with its comparison between concordance groups for all participants and for non-abstainer participants. Differences in the predictors between AUDIT concordance groups were compared using the Mann–Whitney test, assuming a 5% significance level.

Multivariable binomial logistic regression models were used to identify predictors of AUDIT concordance for all participants and for the non-abstainers, controlling for the confounding of age and gender. The same approach was used to evaluate the association of the predictors with patient self-reported categorical AUDIT scores. A correlation matrix of the predictors included in the model is reported, using Pearson coefficients.

To evaluate the accuracy of family AUDIT reports in identifying a patient AUDIT+ self-report, we calculated sensitivity, specificity, positive predictive value, negative predictive value, and area under the receiver operating characteristics (ROC) curve. To account for variations in this predictive effect by age and gender, we stratified by patient and family gender (male and female) and age (<35 and ≥35 years) in this analysis.

Analysis was conducted in R via the psych and corrplot packages for correlation diagrams, the kruskal.test function was used for Kruskal–Wallis rank sum tests, the chisq.test function was used for chi-square analysis, the glm function in R was used for binomial regression, and the ROC and Epi packages in R were used for area under the curve analysis (RStudio, 2018).

Results

Demographics

The majority of patients in the sample were male (81.40%) and young to middle aged (Mdn = 33; IQR = 25–45), exhibiting AUDIT scores within the lowest quartile of possible scores (Mdn = 6; IQR = 0–16) and DrInC (Mdn = 5; IQR = 2–18) scores within the lowest third of the scale. PAS scores for patients are generally at the top half of possible scores (M = 76.04; IQR = 54.91–87.83). Family members are also majority male (62.17%), of a similar age distribution (Mdn = 35; IQR = 26–45), and with PAS scores similar to patients (M = 76.04; IQR = 60.99–87.83).

Concordance

AUDIT concordance between patient self-report and family proxy report was seen for 72% of patient–family pairs in the sample (n = 167, 72.29%). Patient self and family AUDIT reports exhibit similar ranges overall (n = 231 pairs), for concordant pairs (n = 167), and for discordant pairs (n = 64). Distributions of alcohol use frequency differed significantly for pairs that were concordant and those discordant, χ2(2) = 41.66, p < .001.

For AUDIT-concordant patients, 40.12% reported abstain ing from alcohol, 15.57% had drunk but not in the past year, and 44.31% had drunk in the past year. Because of the high proportion of AUDIT-concordant patients abstaining from alcohol, we include here an overall analysis including all patient–family pairs (Table 1, columns 1–3) and a separate analysis excluding abstainers (Table 1, columns 4–6). The vast majority of non-abstainers were also AUDIT concordant (n = 119, 73.45%).

Table 1.

Self- and family-report questionnaire results overall and by Alcohol Use Disorders Identification Test (AUDIT) concordance level between patient and family report

graphic file with name jsad.21-00141tbl1.jpg

Demographics/ questionnaire Total, family–patient pairs Non-abstainers, family–patient pairs
(n = 231) AUDIT concordant (n = 167, 72%) AUDIT discordant (n = 64, 28%) (n = 162) AUDIT concordant (n = 119, 73%) AUDIT discordant (n = 43, 27%)
Demographics
 Patient age median (IQR) 33 (25–45) 32 (25–45) 35 (26–50) 36.5 (27–47) 37 (27–48) 35 (26–41)
 Patient gender (% male) 81% 81% 81% 81% 81% 81%
 Family age median (IQR) 35 (26–45) 34 (26–43) 39 (26.5–50) 35 (27–45) 35 (25–50) 35 (28–43.75) 35 (25–50)
 Family gender (% male) 62% 64%a 58% 60% 60% 60%
AUDIT
 Patient median (IQR) 6 (0–16) 0 (0–12) 13 (8–20) 6 (2–10.75) 4 (2–10.5) 8 (5.5–10.5)
 Family median (IQR) 2 (0–7) 0 (0–8) 3 (2–6) 3.5 (1–10) 3 (1–9.5) 6 (3–10)
 P–F comparison χ2(27) = χ2(26) = χ2(14) = χ2(26) = χ2(26) = χ2(15) =
163.04, p < .01 143.09, p < .01 29.74, p < .05 84.76,p < .01 87.86, p < .01 34.35, p < .01
PAS
 Patient median (IQR) 76 (55–88) 76 (55–88) 76 (56–88) 78 (59–88) 81 (62–88) 72 (58–88)
 Family median (IQR) 76 (61–87) 76 (61–88) 71 (62–88) 76.04 (62–88) 76 (63–88) 71 (62–88)
 P–F comparison χ2(126) = χ2(97) = χ2(39) = χ2(95) = χ2(75) = χ2(28) =
136.05, p = .26 103.84, p = .30 45.17, p = .23 107.33, p = .18 77.14, p = .41 32.94, p = .24
DrInC
 Patient median (IQR) 5 (2–18) 6 (2.50–18) 5 (2–13.75) 4 (2–11.75) 3 (1–9.5) 8 (3–21.5)
Open in a new tab

Notes: IQR = interquartile range; PAS = Perceived Alcohol Stigma scale; DrInC = Drinker Inventory of Consequences.

a

One missing value.

Predictability

Regression analysis shows a family-reported categorical AUDIT score of greater than 7 to be a significant predictor of patient categorical AUDIT as well as patient–family AUDIT discordance (Table 2). As the family-reported categorical AUDIT was shown to be a significant predictor of a patient's self-reported categorical alcohol use, we conducted ROC analysis to determine the sensitivity, specificity, and positive and negative predictive values for the use of family reporting as a continuous variable as a proxy for patient self-report AUDIT category (considered to be a gold standard). Analysis shows lower continuous family-reported AUDIT scores have the greatest power overall to predict patient self-reported categorical AUDIT score, with high overall sensitivity (95.10%) and negative predictive value (94.74%), but lower specificity (69.77%) and positive predictive value (71.32%) (Table 3).

Table 2.

Regression analyses overall and by Alcohol Use Disorders Identification Test (AUDIT) concordance level between patient and family report

graphic file with name jsad.21-00141tbl2.jpg

Demographics/questionnaire Total, family–patient pairs Non-abstainers, family–patient pairs
AUDIT discordance (n = 231), OR [95% CI], p Patient categorical AUDIT (n = 231), OR [95% CI], p AUDIT discordance (n = 231), OR [95% CI], p Patient categorical AUDIT (n = 231), OR [95% CI], p
Demographics
 Patient age 1.00 [0.97, 1.03], p = .95 1.00 [0.97, 1.03], p = .91 1.00 [0.97, 1.03], p = .87 1.00 [0.97, 1.02], p = .84
 Patient gender 1.06 [0.40, 2.61], p = .73 (female) 0.62 [0.21, 1.77], p = .36 (female) 1.16 [0.42, 3.03],p = .76 (female) 0.58 [0.18, 1.67], p = .33
 Family age 1.00 [0.97, 1.03], p = .10 1.01 [0.98, 1.04], p = .69 1.00 [0.97, 1.03], p = .97 1.01 [0.98, 1.06], p = .39
 Family gender 1.28 [0.61, 2.65], p = .51 (female) 1.09 [0.51, 2.30], p = .82 (female) 1.03 [0.47, 2.21], p = .94 (female) 0.91 [0.40, 2.03], p = .82
AUDIT
 Patient N.A. N.A. N.A. N.A.
 Family 3.13 [1.52, 6.48], p < .05 (categorical, Yes) 9.64 [4.78, 20.12], p < .001 (categorical, Yes) 1.24 [0.55, 2.75], p = .60 (categorical, Yes) 4.53 [2.09, 10.08],p < .001 (categorical, Yes)
PAS
 Patient 1.01 [0.99, 1.03], p = .26 1.01 [0.99, 1.02], p = .42 1.00 [0.98, 1.02], p = .87 1.00 [0.98, 1.02], p = .91
 Family 1.00 [0.98, 1.02], p = .95 1.00 [0.98, 1.01], p = .63 1.00 [0.98, 1.02], p = .84 0.99 [0.97, 1.01], p = .45
DrInC
 Patient 1.01 [0.98, 1.04], p = .53 1.03 [1.00, 1.07], p = .08 1.04 [1.00, 1.08], p = .05 1.07 [1.03, 1.12], p < .001
Open in a new tab

Notes: OR = odds ratio: CI = confidence interval; N.A. = not applicable; PAS = Perceived Alcohol Stigma scale; DrInC = Drinker Inventory of Consequences; bold indicates statistical significance (p < .05).

Table 3.

Receiver operating characteristics data for model of family Alcohol Use Disorders Identification Test (AUDIT) score predicting patient self-reported categorical AUDIT, stratifying for demographic variables

graphic file with name jsad.21-00141tbl3.jpg

Variable Sensitivity % Specificity % PPV % NPV % AUC estimate [95% CI]
Overall 90.16 72.35 53.92 95.35 0.74 [0.82, 0.92]
Patient male 88.89 71.64 55.81 94.11 0.87 [0.81, 0.92]
Patient female 100.00 75.00 43.75 100 0.90 [0.81, 0.90]
Family member male 90.00 70.87 54.55 94.81 0.862 [0.797, 0.927]
Family member female 90.48 74.24 52.78 96.08 0.89 [0.81, 0.96]
Patient age < 35 96.55 70.97 50.91 98.51 0.88 [0.82, 0.95]
Patient age ≥ 35 78.13 83.33 67.57 89.55 0.87 [0.79, 0.94]
Family age < 35 92.86 70.24 50.98 96.72 0.89 [0.82, 0.95]
Family age ≥ 35 87.88 74.11 56.86 94.03 0.85 [0.78, 0.93]
Open in a new tab

Notes: PPV = positive predictive value; NPV = negative predictive value; AUC = area under the curve; CI = confidence interval.

Correlation tables of the variables used in regressions, including all 231 patient–family pairs, are shown in Table 4. Results suggest patient and family gender, family AUDIT and family PAS, family AUDIT and patient DrInC, and patient PAS and family PAS to be relatively correlated.

Table 4.

Correlation tables for predictor variables included in regression analyses

graphic file with name jsad.21-00141tbl4.jpg

Variable Patient AUDIT r (p) Family AUDIT r (p) Patient PAS r (p) Family PAS r (p) Patient DrInC r (p)
Patient – .694 .113 .098 .261
 AUDIT (.001) (.086) (.1375) (.001)
Family .694 – .061 .108 .245
 AUDIT (.001) (.360) (.099) (.001)
Patient .113 .061 – .208 -.033
 PAS (.086) (.360) (.002) (.616)
Family .098 .108 .208 – .113
 PAS (.1375) (.099) (.002) (.086)
Patient .261 .245  .033 .113 –
 DrInC (.001) (.001) (.616) (.086)
Open in a new tab

Notes: AUDIT = Alcohol Use Disorders Identification Test; PAS = Perceived Alcohol Stigma scale; DrInC = Drinker Inventory of Consequences.

Discussion

This is the first study, to our knowledge, that attempts to understand the predictive ability of family self-reports on patient alcohol use for injury patients presenting to an ED of a low-income country. Similarly, we tried to understand how perceived alcohol stigma affected this predictive ability to add to the current literature about the presence of perceived alcohol stigma, associated risk behaviors, and reporting by injury patients (El-Gabri et al., 2020a, 2020b; Friedman et al., 2020; Meier et al., 2020; Staton et al., 2018, 2020; Zhao et al., 2020). We found that patient and family reports are concordant 73% of the time, suggesting family reports as having predictive ability. Both receiver operating curve analysis and high patient–family AUDIT concordance verified this, suggesting that the use of family reporting of alcohol as a proxy for patient self-reported alcohol use has acceptable sensitivity and specificity and is a significant predictor.

In our clinical setting, given a long prehospital time, we have patients who suffer delirium or alcohol withdrawal early in their hospital stay, causing significant morbidity or death. As such, these data can support the use of family AUDIT reporting to help identify those patients who should be monitored/treated for alcohol withdrawal. As the AUDIT scale has shown in other usages, it is a very sensitive tool; our result here that the family member AUDIT score is highly sensitive to self-reported patient AUDIT implies that a family-reported point of view of patient alcohol use in this setting is sufficient enough to find patients with problematic alcohol use. Although the relatively lower specificity could lead to false positives, and thus unnecessary clinical evaluations or alcohol-related treatment, we believe the high sensitivity to be clinically important in identifying patients to target. Further research is therefore warranted into causes of discordance in alcohol use reporting, as well as using family reporting of AUDIT clinically in other settings because of cultural, stigma-related, reporting behavior among others.

The correlation between self-reported AUDIT and DrInC was lower than expected and previously reported (Zhao et al., 2018). When we evaluated this correlation for the whole patient sample (n = 341), we found a higher correlation, which could be explained by a social desirability of the dyads in reporting either DrInC or AUDIT scores.

Even when patients are identified by family reports, they still must decide to actively change their alcohol behaviors. Although total scores in perceived alcohol stigma do not appear to significantly correlate with AUDIT alcohol use scores, our results show a correlation between self- and family-reported PAS, a finding that may indicate increased perceived patient stigma if an individual thinks or is aware of family members knowing their drinking habits. Literature has discussed a positive role of the family environment and beliefs in substance use and risky behaviors; however, it has focused on parental relationships of adolescents and children (Nash et al., 2005) and the use of family-focused interventions (Spoth et al., 2002) and skills training programs (Mejía et al., 2020) specific to AUD in low- and middle-income countries. The implication that there may be alcohol-related stigma underlying the family dynamic (Haverfield & Theiss, 2016) even between adults supports the need for more information about family members’ reports and further research to better inform how we should ask questions of family members. Family-reported stigma and information about where this stigma comes from can be used to incorporate both patients and families into stigma-reduction interventions (Keyes et al., 2010).

Along with the need for further research on family-reported substance use such as alcohol as well as their perceptions of stigma to better inform the need and benefits of family enrollment in treatment and conducting patient and family member stigma-reduction programs, further research is needed into defining stigma and assessing it using a scale that is more appropriately culturally adapted. In addition, previous studies on the PAS scale have provided impetus for looking at specific stigmas rather than an overall scale (Griffin et al., 2020; Staton et al., 2018). Our results showed contradicting opinions of stigma through correlational analysis between PAS questions, suggesting that the scale itself (which was developed in a high-income country and translated but not culturally adapted for Tanzania, unlike the AUDIT; Nickenig Vissoci et al., 2018) may have inherent issues (Glass et al., 2013a).

This study has possible limits in generalizability inherent to the gendered nature of alcohol use in Tanzania (Francis et al., 2015), age of patients coming into the KCMC ED (majority 25–45 years), sample size (231 patient–family pairs), distribution of family members brought in to respond (4.76% including friends/coworkers and the majority of biological family members being parents), and selection bias, as we could only include patients in the analysis for whom family members were available. Although this secondary analysis used data from a single ED referral center, distance of family members from the center could affect their availability to be present; however, we believe this population is still an accurate representation of the Moshi population because of its serving more than 15 million rural and urban residents of the Kilimanjaro Region. We found a small sample (n = 64) of AUDIT-discordant patient–family pairs but conducted the nonparametric Kruskal–Wallis rank sum test to account for a nonnormal distribution of data. In addition, analysis of perceived alcohol stigma is limited by the use of the PAS scale, which our results suggest may not provide a detailed or specific enough understanding of individual alcohol stigma. This suggests that a de novo establishment of a scale for alcohol stigma may be necessary for better contextual representation of alcohol-related stigma.

Conclusions

We found a high concordance between patient self-reported and family-reported AUDIT scores, suggesting that patients may be realistic about their alcohol use and their family members tend to be aware of alcohol use behavior and that that information can be used for early detection and recognition of patient alcohol use or misuse. Patient–family alcohol use concordance coupled with ROC analysis suggests family reports to be clinically relevant and accurate markers of patient alcohol use. Given the discrepancy in correlations for total PAS scores and individual PAS questions with the categorical AUDIT, we recommend further research into how, when, and from whom stigma toward alcohol is developed and which aspects are most important or commonly held. Further research is needed to understand how factors beyond stigma may affect patient and family reporting differently.

Acknowledgments

The authors acknowledge the Kilimanjaro Christian Medical Center Emergency Department research team for their dedication to improving the lives of their patients.

Footnotes

This research was supported by the Fogarty International Center of the National Institutes of Health underAward Number K01TW010000 (principal investigator, Catherine A. Staton). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Conceptualization of this article was performed by J.R.N.V., B.T.M., and C.A.S.; methodology: A.S.K., C.L., J.R.N.V., and C.A.S.; formal analysis and investigation: A.S.K., C.L., P.O., J.R.N.V., and C.A.S.; writing of the original draft: A.S.K., C.L., and P.O.; writing (review and editing): A.S.K., P.O., L.M., J.B., J.R.N.V., B.T.M., and C.A.S.; funding acquisition: C.A.S. and B.T.M.; and supervision: J.R.N.V., B.T.M., and C.A.S.

References

  1. Babor T. F., Robaina K. The Alcohol Use Disorders Identification Test (AUDIT): A review of graded severity algorithms and national adaptations. International Journal of Alcohol and Drug Research. 2016;5:17–24. doi:10.7895/ijadr.v5i2.222. [Google Scholar]
  2. Bourgeois F. C., Fossa A., Gerard M., Davis M. E., Taylor Y. J., Connor C. D., Bell S. K. A patient and family reporting system for perceived ambulatory note mistakes: Experience at 3 U.S. healthcare centers. Journal of the American Medical Informatics Association. 2019;26:1566–1573. doi: 10.1093/jamia/ocz142. doi:10.1093/jamia/ocz142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. De Los Reyes A., Bunnell B. E., Beidel D. C.2013Informant discrepancies in adult social anxiety disorder assessments: Links with contextual variations in observed behavior Journal of Abnormal Psychology 122376–386.doi:10.1037/a0031150 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. De Los Reyes A., Ohannessian C. M. Introduction to the special issue: Discrepancies in adolescent-parent perceptions of the family and adolescent adjustment. Journal of Youth and Adolescence. 2016;45:1957–1972. doi: 10.1007/s10964-016-0533-z. doi:10.1007/s10964-016-0533-z. [DOI] [PubMed] [Google Scholar]
  5. Dybek I., Bischof G., Grothues J., Reinhardt S., Meyer C., Hapke U.,. . . Rumpf H.-J.2006The reliability and validity of the Alcohol Use Disorders Identification Test (AUDIT) in a German general practice population sample Journal of Studies on Alcohol 67473–481.doi:10.15288/jsa.2006.67.473 [DOI] [PubMed] [Google Scholar]
  6. El-Gabri D., Barcenas L. K., Meier B., Mvungi M., Haglund M., Gerardo C. J., Staton C. A. Injury patients' perceptions of drink-driving: A qualitative assessment of drink-driving behavior in Moshi, Tanzania. PLoS One. 2020a;15:e0230662. doi: 10.1371/journal.pone.0230662. doi:10.1371/journal.pone.0230662. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. El-Gabri D., Nickenig Vissoci J. R., Meier B. J., Mvungi M., Haglund M., Swahn M., Staton C. A.2020bAlcohol stigma as it relates to drinking behaviors and perceptions of drink drivers: A mixed method study in Moshi, Tanzania Alcohol 8873–81.doi:10.1016/j.alcohol.2020.04.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Etchegaray J. M., Ottosen M. J., Burress L., Sage W. M., Bell S. K., Gallagher T. H., Thomas E. J. Structuring patient and family involvement in medical error event disclosure and analysis. Health Affairs (Project Hope) 2014;33:46–52. doi: 10.1377/hlthaff.2013.0831. doi:10.1377/hlthaff.2013.0831. [DOI] [PubMed] [Google Scholar]
  9. Francis J. M., Weiss H. A., Mshana G., Baisley K., Grosskurth H., Kapiga S. H. The epidemiology of alcohol use and alcohol use disorders among young people in Northern Tanzania. Friedman. PLoS One. 2015;10:e0140041. doi: 10.1371/journal.pone.0140041. doi:10.1371/journal.pone.0140041. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. K., Raymond J., Minnig M. C., Sakita F., Vissoci J. R. N., Staton C.2020Association between alcohol use and violence-related injury in emergency department patients in Moshi, Tanzania: Analysis of a prospective trauma registry The Lancet. Global Health 8S18.doi:10.1016/S2214-109X(20)30159-5 [Google Scholar]
  11. Glass J. E., Kristjansson S. D., Bucholz K. K.2013aPerceived alcohol stigma: Factor structure and construct validation Alcoholism: Clinical and Experimental Research 37Supplement 1E237–E246doi:10.1111/j.1530-0277.2012.01887.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Glass J. E., Mowbray O. P., Link B. G., Kristjansson S. D., Bucholz K. K. Alcohol stigma and persistence of alcohol and other psychiatric disorders: A modified labeling theory approach. Drug and Alcohol Dependence. 2013b;133:685–692. doi: 10.1016/j.drugalcdep.2013.08.016. doi:10.1016/j.drugalcdep.2013.08.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Griffin S. M., Karia F. P., Zimmerman A., Minnig M. C. C., Swahn M., Makelarski J., Staton C. A. Mixed-methods study: Sex differences in experiences of stigma associated with alcoholism and alcohol use disorders among injury patients in Tanzania. Alcoholism: Clinical and Experimental Research. 2020;44:1700–1707. doi: 10.1111/acer.14402. doi:10.1111/acer.14402. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Haeny A. M., Littlefield A. K., Wood P. K., Sher K. J. Method effects of the relation between family history of alcoholism and parent reports of offspring impulsive behavior. Addictive Behaviors. 2018;87:251–259. doi: 10.1016/j.addbeh.2018.07.022. doi:10.1016/j.addbeh.2018.07.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Harris P. A., Taylor R., Thielke R., Payne J., Gonzalez N., Conde J. G. Research electronic data capture (REDCap)–a metadata-driven methodology and workflow process for providing translational research informatics support. Journal of Biomedical Informatics. 2009;42:377–381. doi: 10.1016/j.jbi.2008.08.010. doi:10.1016/j.jbi.2008.08.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Haverfield M. C., Theiss J. A. Parent's alcoholism severity and family topic avoidance about alcohol as predictors of perceived stigma among adult children of alcoholics: Implications for emotional and psychological resilience. Health Communication. 2016;31:606–616. doi: 10.1080/10410236.2014.981665. doi:10. 1080/10410236.2014.981665. [DOI] [PubMed] [Google Scholar]
  17. Hingson R., Heeren T., Winter M., Wechsler H. Magnitude of alcohol-related mortality and morbidity among U.S. college students ages 18-24: Changes from 1998 to 2001. Annual Review of Public Health. 2005;26:259–279. doi: 10.1146/annurev.publhealth.26.021304.144652. doi:10.1146/annurev.publhealth.26.021304.144652. [DOI] [PubMed] [Google Scholar]
  18. Hingson R. W., Zha W., Weitzman E. R. Magnitude of and trends in alcohol-related mortality and morbidity among U.S. college students ages 18-24, 1998-2005. Journal of Studies on Alcohol and Drugs. 2009;Supplement 16:12–20. doi: 10.15288/jsads.2009.s16.12. doi:10.15288/jsads.2009.s16.12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Keyes K. M., Hatzenbuehler M. L., McLaughlin K. A., Link B., Olfson M., Grant B. F., Hasin D. Stigma and treatment for alcohol disorders in the United States. American Journal of Epidemiology. 2010;172:1364–1372. doi: 10.1093/aje/kwq304. doi:10.1093/aje/kwq304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kühn J., Slabbert I. The effects of a father's alcohol misuse on the wellbeing of his family: Views of social workers. Social Work. 2017;53(3) doi:10.15270/53-3-578. [Google Scholar]
  21. Lee C. S., Baird J., Longabaugh R., Nirenberg T. D., Mello M. J., Woolard R. Change plan as an active ingredient of brief motivational interventions for reducing negative consequences of drinking in hazardous drinking emergency-department patients. Journal of Studies on Alcohol and Drugs. 2010;71:726–733. doi: 10.15288/jsad.2010.71.726. doi:10.15288/jsad.2010.71.726. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Makol B. A., De Los Reyes A., Ostrander R. S., Reynolds E. K. Parent-youth divergence (and convergence) in reports of youth internalizing problems in psychiatric inpatient care. Journal of Abnormal Child Psychology. 2019;47:1677–1689. doi: 10.1007/s10802-019-00540-7. doi:10.1007/s10802-019-00540-7. [DOI] [PubMed] [Google Scholar]
  23. Meier B. J., El-Gabri D., Friedman K., Mvungi M., Mmbaga B. T., Nickenig Vissoci J. R., Staton C. A. Perceptions of alcohol use among injury patients and their family members in Tanzanian society. Alcohol. 2020;83:9–15. doi: 10.1016/j.alcohol.2019.06.001. doi:10.1016/j.alcohol.2019.06.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Mejía A., Bertello L., Gil J., Griffith J., López A. I., Moreno M., Calam R. Evaluation of family skills training programs to prevent alcohol and drug use: A critical review of the field in Latin America. International Journal of Mental Health and Addiction. 2020;18:482–489. doi:10.1007/s11469-019-00060-x. [Google Scholar]
  25. Milanzi E. B., Ndasauka Y.2019Prevalence of alcohol addiction in Africa Ndasauka Y., Kayange G. M. (Eds.), Addiction in South and East Africa: Interdisciplinary approaches (pp. 215–228)Cham, Switzerland: Springer International Publishing; doi:10.1007/978-3-030-13593-5_13 [Google Scholar]
  26. Miller W. R., Tonigan J. S., Longabaugh R.1995The Drinker Inventory of Consequences (DrInC): An instrument for assessing adverse consequences of alcohol abuse Test manual (NIAAA Project MATCH Monograph Series4NIH Publication No. 95-3911). Bethesda, MD: National Institute on Alcohol Abuse and Alcoholism [Google Scholar]
  27. Moretti-Pires R. O., Corradi-Webster C. M. [Adaptation and validation of the Alcohol Use Disorders Identification Test (AUDIT) for a river population in the Brazilian Amazon] Cadernos de Saude Publica. 2011;27:497–509. doi: 10.1590/s0102-311x2011000300010. doi:10.1590/S0102-311X2011000300010. [DOI] [PubMed] [Google Scholar]
  28. Nash S. G., McQueen A., Bray J. H. Pathways to adolescent alcohol use: Family environment, peer influence, and parental expectations. Journal of Adolescent Health. 2005;37:19–28. doi: 10.1016/j.jadohealth.2004.06.004. doi:10.1016/j.jadohealth.2004.06.004. [DOI] [PubMed] [Google Scholar]
  29. Nickenig Vissoci J. R., Hertz J., El-Gabri D., Andrade Do Nascimento J. R., Pestillo De Oliveira L., Mmbaga B. T., Staton C. A. Cross-cultural adaptation and psychometric properties of the AUDIT and CAGE questionnaires in Tanzanian Swahili for a traumatic brain injury population. Alcohol and Alcoholism. 2018;53:112–120. doi: 10.1093/alcalc/agx058. doi:10.1093/alcalc/agx058. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Pal H. R., Jena R., Yadav D. Validation of the Alcohol Use Disorders Identification Test (AUDIT) in urban community outreach and de-addiction center samples in north India. Journal of Studies on Alcohol. 2004;65:794–800. doi: 10.15288/jsa.2004.65.794. doi:10.15288/jsa.2004.65.794. [DOI] [PubMed] [Google Scholar]
  31. Pradhan B., Chappuis F., Baral D., Karki P., Rijal S., Hadengue A., Gache P. The Alcohol Use Disorders Identification Test (AUDIT): Validation of a Nepali version for the detection of alcohol use disorders and hazardous drinking in medical settings. Substance Abuse Treatment, Prevention, and Policy. 2012;7:42. doi: 10.1186/1747-597X-7-42. doi:10.1186/1747-597X-7-42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Ritchie H., Roser M. 2018. Alcohol consumption. Our World in Data. Retrieved from https://ourworldindata.org/alcohol-consumption. [Google Scholar]
  33. RStudio. Boston, MA: Author; 2018. RStudio: Integrated Development Environment for R (IDE) (Version 1.1456) Retrieved from https://www.rstudio.com. [Google Scholar]
  34. Saunders J. B., Aasland O. G., Babor T. F., de la Fuente J. R., Grant M. Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO Collaborative Project on Early Detection of Persons with Harmful Alcohol Consumption-II. Addiction. 1993;88:791–804. doi: 10.1111/j.1360-0443.1993.tb02093.x. doi:10.1111/j.1360-0443.1993.tb02093.x. [DOI] [PubMed] [Google Scholar]
  35. Sorsdahl K., Stein D. J., Myers B. Negative attributions towards people with substance use disorders in South Africa: Variation across substances and by gender. BMC Psychiatry. 2012;12:101. doi: 10.1186/1471-244X-12-101. doi:10.1186/1471-244X-12-101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Spoth R. L., Guyll M., Day S. X. Universal family-focused interventions in alcohol-use disorder prevention: Cost-effectiveness and cost-benefit analyses of two interventions. Journal of Studies on Alcohol. 2002;63:219–228. doi: 10.15288/jsa.2002.63.219. doi:10.15288/jsa.2002.63.219. [DOI] [PubMed] [Google Scholar]
  37. Staton C. A., Vissoci J. R. N., Toomey N., Abdelgadir J., Chou P., Haglund M., Swahn M. The impact of alcohol among injury patients in Moshi, Tanzania: A nested case-crossover study. BMC Public Health. 2018;18:275. doi: 10.1186/s12889-018-5144-z. doi:10.1186/s12889-018-5144-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Staton C. A., Zhao D., Ginalis E. E., Hirshon J. M., Sakita F., Swahn M. H., Nickenig Vissoci J. R. Alcohol availability, cost, age of first drink, and its association with at-risk alcohol use in Moshi, Tanzania. Alcoholism: Clinical and Experimental Research. 2020;44:2266–2274. doi: 10.1111/acer.14457. doi:10.1111/acer.14457. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Tanzania National Bureau of Statistics. Population and housing census: Population distribution by administrative areas. Ministry of Finance, Dar es Salaam. 2012. Retrieved from https://www.nbs.go.tz/index.php/en/census-surveys/population-and-housing-census/162-2012-phc-population-distribution-by-administrative-areas.
  40. Vandenbroucke J. P., von Elm E., Altman D. G., Gøtzsche P. C., Mulrow C.D., Pocock S. J., Egger M. S. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): Explanation and elaboration. PLoS Med. 2007;4(10):297. doi: 10.1371/journal.pmed.0040297. doi:10.1371/journal.pmed.0040297. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. World Health Organization. Global status report on alcohol and health 2018. 2018 Retrieved from https://www.who.int/publications/i/item/9789241565639. [Google Scholar]
  42. Zhao D., Nickenig Vissoci J. R., Mmbaga B. T., Abdullah A. S., Staton C. A. Alcohol-related risky behavior patterns and their association with alcohol use and perceived alcohol stigma in Moshi, Tanzania. Journal of Studies on Alcohol and Drugs. 2020;81:614–623. doi: 10.15288/jsad.2020.81.614. doi:10.15288/jsad.2020.81.614. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Zhao D., Staton C. A., He Q., Mmbaga B. T., Nickenig Vissoci J. R. Cross-culture adaptation and psychometric properties of the DrInC Questionnaire in Tanzanian Swahili. Frontiers in Public Health. 2018;6:330. doi: 10.3389/fpubh.2018.00330. doi:10.3389/fpubh.2018.00330. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Journal of Studies on Alcohol and Drugs are provided here courtesy of Rutgers University. Center of Alcohol Studies

RESOURCES