A cross-sectional facility-based survey was conducted at two public health clinics of Lusaka, Zambia in July 2017 about alcohol consumption before and during pregnancy by administering, validated screening questionnaires to identify problem drinking in a pregnant population (see Additional file 1).17,24,25
The study was located at two public primary care antenatal clinics within peri-urban and urban communities in Lusaka province Zambia where the majority face high levels of socioeconomic adversity and drug abuse. For example, it has recently been estimated that about 16.8% of the female population in Kalingalinga consume alcohol.26 In 2017, Mtendere clinic’s catchment population was 114,064, with an annual antenatal target of 6,159 while Kalingalinga clinic’s catchment population was 94,649, with an annual antenatal target of 5,111.
A sample of pregnant women were recruited from two public health clinics of Kalingalinga urban clinic and Mtendere peri-urban clinic, which provide free prenatal care to pregnant women in these parts of Lusaka province in Zambia. At Kalingalinga urban clinic, 90 women were approached and 79 (87.78%) completed the tool vs. 109 of 120 women approached at Mtendere clinic (90.83%). In order to be eligible to complete a T-ACE (Tolerance (T); Annoyed (A); Cut down and eye-opener (E)) screening tool, women had to be pregnant, aged 18 or over and visiting one of the clinics at the time of the study. Excluded from the study were eligible women who refused to answer the T-ACE tool (n=8 at Kalingalinga clinic and n=6 at Mtendere clinic) and adolescents aged below 16.
Every third woman attending their first prenatal visit was invited to participate in the study through systematic sampling as previously used in similar studies.27 A major advantage of this approach is that it optimized inclusion of women who were willing to participate, maximized the effectiveness of data collection efforts, and because the consideration of subjects representative of the entire population was not an objective of this study.28
Recruitment occurred from July 19 to 31, 2017. Women were recruited with the help of the district health manager. First, selected for inclusion were two health clinics (peri-and urban) with the highest HIV prevalence rates in the district. Secondly, researchers approached pregnant women while they were waiting for their routine prenatal care, provided at set times each day of the week, and to reach the maximum number of women, recruitment occurred during these times. Those who opted to participate in the study received verbal and written information about the study and reassurance about confidentiality. If they agreed to participate, they completed the alcohol-screening questionnaire right away in English or in a preferred participant local language of choice (Bemba and Nyanja). The questionnaire were completed in private and non-judgmental settings (e.g., conference room).
All women who provided written consent to participate were screened in two stages. First, administered was a brief sociodemographic questionnaire. Demographic information included age, gender, and socioeconomic status as approximated by parental education. Second was verbal and self- administration of two commonly used alcohol-screening questionnaires in a face-to-face interview conducted by the researcher and a trained research assistant.
A validated T-ACE alcohol-screening questionnaire was used based on its widespread use in detecting tolerance to alcohol and lifetime alcohol abuse or addiction issues. The tool was translated in Bemba and Nyanja, two languages commonly spoken in Lusaka by the author (native speaker), and validated in these languages. The T-ACE has been found to have high levels of sensitivity (69–88%) and specificity (71–89%) for identifying risk drinking and problem drinking among pregnant women,13 and has been validated for use in wide range of settings19,29,30 including Sub-Saharan Africa31, and for verbal administration by an interviewer in both English and local language. A score of two or more on the “T-ACE” indicates at-risk drinking.24 Four question comprise the T-ACE: (1) How many drinks does it take to make you feel high? (2) Have people annoyed you by criticizing you’re drinking? (3) Have you ever felt ou ought to cut down on your drinking? (4) Have you ever had a drink first thing in the morning to steady your nerves or get rid of a hangover?
To identify overall severity of dependence or periconceptional period alcohol use, a five-item questions were administered with the T-ACE screening tool: (1) During the time you were pregnant, but didn’t know you were pregnant, how many alcohol drinks did you usually have at one time? (2) During the time you were pregnant, but didn’t know you were pregnant, how often did you drink beer, wine or other alcoholic beverages? (3) How often did you have four or more drinks in one day in the past 30 days? (4) How many drinks did you have on a typical day when you were drinking alcohol in the past 30 days? (5) During the past 30 days, on how many days did you drink one or more drinks of an alcoholic beverage?
IBM SPSS Statistics, version 24.0 (Armonk, NY: IBM Corp) was used to perform descriptive statistics. To compare demographic factors and alcohol use patterns among participants by clinic and among those who scored two or more points on the T-ACE questionnaire, bivariate analyses were performed using the χ2 test for dichotomous variables and the t-test for continuous variables. Because alcohol use data were not normally distributed, I used medians and first (Q1) and third (Q3) quartiles to describe the data distribution on any alcohol use and binge drinking episode in the past 30 days (number of drinks and drank ≥4 alcoholic drinks on at least one day), and any at-risk drinking (scored >2 on the T-ACE questionnaire). A mixed-effects linear model (188 pregnant women in 10 wards) was used to evaluate the effect of the log-transformed outcome variables (number of drinks consumed ≥4 alcoholic drinks on at least one day in the past 30 days and scoring >2 on the T-ACE questionnaire), after adjusting for patient-level variables. Patient-level variables included age, marital status, education and prenatal care regular attendance. The mixed-effects analysis were used to account for dependence resulting from pregnant women being nested within administrative wards. Statistical significance of fixed effects were evaluated at p < .05.