Source of data

This study used secondary data from the Zambia WHO-STEPS Survey 2017 for non-communicable diseases (NCDs) prevalence and risk factors, mental and oral health. The Zambia STEPS survey is a cross-sectional national survey designed to obtain data representative of the adult population, aged 18–69 years, living in Zambia. The STEPS survey uses a multi-stage cluster sampling technique and utilizes the household listing from the Zambia Population-Based HIV Impact Assessment (ZAMPHIA). The first stage of sampling was a selection of Standard Enumeration Areas (SEAs) from each of the 10 provinces using probability proportional to size (PPS). In the second stage, 15 households in the rural SEAs and 20 households in urban SEAs were selected systematically using appropriate sampling intervals based on the number of households in that SEA. In the third stage, one member of the household who was eligible was selected for an interview. The questionnaire administered to survey participants included demographic data, as well as physical and biochemical measurements. Overall, 4301 participants were included in the survey, with a response rate of 77.7%. Details of the WHO-STEPS methodology can be found elsewhere²⁵.

Eligibility criteria

All participants aged 18–69 years who reported ‘yes’ to the question ‘Do you currently smoke any tobacco products, such as cigarettes, shisha, cigars, or pipes?’ during the WHO-STEPS Survey were eligible for the study.

Variables and measurements

Primary outcomes

There were two primary outcome variables of interest: 1) receipt of advise to quit smoking among current smokers who visited a doctor or other healthcare worker in the past 12 months (yes/no response to the question ‘During any visit to a doctor or other health worker in the past 12 months, were you advised to quit smoking tobacco?); and 2) past year quit attempt status (yes/no response to the question ‘During the past 12 months, have you tried to stop smoking?’).

Sociodemographic variables

Age, gender (male/female), residence (rural/urban), education level (no or less than primary, primary, secondary school, or higher education), marital status (never married or married/cohabiting), and employment status (self-employed/formally employed, unemployed/retired, or nonpaid/student/homemaker).

Behavioral and psychological variables

Current smokers who responded ‘yes’ to the question ‘Have you consumed any alcohol within the past 30 days?’ were classified as alcohol consumers, whereas those who reported ‘no’ were classified as non-consumers. Current smokers were asked: ‘During the past 12 months, have you attempted suicide?’. Those who answered ‘yes’ were classified as recent suicide attempt whereas those who answered ‘no’ were classified as no recent suicide attempt.

Clinical variables

Two specific questions were asked to assess history of cardiovascular disease (CVD): 1) ‘Have you ever had a heart attack?’ (yes/no); and 2) ‘Have you ever been told by a doctor or other health worker that you have raised blood pressure or hypertension?’ (yes/no). History of metabolic disease (diabetes mellitus) was assessed by asking: ‘Have you ever been told by a doctor or other health worker that you have raised blood sugar or diabetes?’ (yes/no).

Statistical analysis

The first analysis included 475 current smokers and the second analysis included 387 current smokers who visited a doctor or other healthcare workers in the past 12 months (Figure 1). We calculated the proportion of smokers who received advise to quit smoking and attempted to quit smoking, overall and stratified by sociodemographic, behavioral and psychological, and clinical variables. Categorical variables are reported as percentages. Chi-squared tests of association were used to determine associations between quit attempts and selected predictor variables. Logistic regression models producing unadjusted (ORs) and adjusted odds ratios (AORs) and the corresponding 95% confidence intervals (CIs), were used to assess the association between advise to quit and quit attempt. Covariates were determined a priori and included in the full model regardless of the statistical significance in the univariate regression model (age, gender, recent suicide attempt, heart attack, hypertension, and diabetes mellitus). Statistical significance in the adjusted model was set at a two-sided 0.05 level. All analyses were conducted using STATA version 15.

Figure 1

Flow diagram of criteria for inclusion in analysis of smokers and those who received quit advise from the healthcare provider (WHO-STEPS Survey 2017)

Ethical approval

The University of Zambia Research Ethics Committee (UNZABREC) approved the STEPS Survey and written informed consent was obtained from all the participants²⁶.

Characteristics of overall current smokers included in the analysis

Among the 475 smokers included in the analysis, more than half were aged 18–44 years (63.8%) (Table 1). The majority of the smokers were men (87.2%), of rural residence (70.9%), with no education or less than primary (42.6%), married/cohabiting (68.8%), and self-employed/formally employed (60.9%). In terms of comorbidities, 2.9% had a history of heart attack, 16.9% had hypertension, and 4.2% had diabetes mellitus. Recent suicide attempt and alcohol consumption was reported in 11.9% and 75.6% of smokers, respectively. Less than half (47.6%) of the current smokers included in the analysis attempted to quit smoking.

Characteristics of current smokers who visited a health provider

Among the 387 current smokers who visited a health provider in the past 12 months for any other medical reason, the majority were aged 18–44 years (63.3%) (Table 1). Most were men (87.1%), of rural residence (70.3%), no education or less than primary (42.1%), married/cohabiting (67.5%), and self-employed/formally employed (59.2%). In terms of comorbidities, 3.4% had a history of heart attack, 7.5% had hypertension, and 7.2% had diabetes mellitus. Recent suicide attempt and alcohol consumption was reported in 12.9% and 76.5% of smokers who visited a healthcare provider, respectively. Only 18.6% of current smokers who visited a health provider in the past 12 months for any other medical reason, reported receiving smoking advise (Figure 1).

Factors associated with attempting to quit smoking

In univariate analysis, women and unemployed/retired occupational status were significantly associated with decreased odds of attempting to quit smoking (Table 2). In contrast, primary and secondary education level and smoking cessation advise were significantly associated with increased odds of attempting to quit smoking. In adjusted analysis, being advised to quit was associated with increased odds of attempting to quit (OR=2.84; 95% CI: 1.11–7.25).

Limitations

Our study is not without limitations. First, we were unable to assess changes in variables over time. This being a cross-sectional design, we were only able to identify factors associated with quitting smoking and were unable to address causality. Second, although we adjusted for known confounders in the multivariable model, the potential for residual confounders inherent in observational studies remains and might have affected study outcomes. Third, the data are based on self-reported responses, which might have resulted in information bias related to misclassification of tobacco smoking and comorbidities. Fourth, we were unable to capture nuance (or spectrum) of some variables, such as alcohol use and recent suicide attempts. In addition, the dataset had limited variables to establish the correlates of the healthcare provider’s advise. Finally, the current study may have limited generalizability to other national or international settings. Nevertheless, our study, however, had the advantage of standardized questionnaires that enabled evaluation of NCD risk factors, including tobacco smoking, in Zambia.