INTRODUCTION
Tobacco smoking continues to pose a major public health challenge globally, leading to millions of preventable deaths annually and playing a significant role in cardiovascular diseases, cancers, and chronic respiratory conditions. Although there have been worldwide efforts to control tobacco use, smoking remains a primary cause of illness and death, especially in low- and middle-income nations where prevention and cessation initiatives are frequently inadequate1.
HCWs are essential in the fight against tobacco use, as they educate patients about health, demonstrate healthy behaviors, and offer support for quitting smoking. However, when healthcare professionals smoke themselves, it can damage their credibility and weaken their efforts to help others quit. Research indicates that smoking rates among HCWs remain alarmingly high in various countries, emphasizing the need for ongoing focus on this important issue2.
Sociodemographic factors like gender, age, and professional specialty are consistently recognized as important influences on smoking behavior. Research indicates that male HCWs and those in certain roles, such as nursing and administrative positions, tend to have higher smoking rates3. The high-stress nature of healthcare, marked by long hours, emotional fatigue, and burnout, also plays a significant role, with many individuals turning to tobacco as a way to cope4. Additionally, the social environment, particularly the presence of family members or friends who smoke, is a well-known factor that predicts both the initiation and continuation of smoking5. A notable issue among HCWs is the gap between knowledge and action. Although they generally have a clear understanding of the serious health risks associated with smoking, this awareness often is not enough to prevent them from smoking or to help them quit2,6.
A systematic review involving 63 countries revealed that smoking among HCWs is still a significant problem, particularly among men and younger professionals7. In Cyprus, similar trends were noted, with nearly one-third of HCWs identified as smokers. The study highlighted that factors such as being male, younger, single, and having family members who smoke were strongly linked to tobacco use8. In Jordan, another study found that around 25% of HCWs smoked, despite the presence of smoke-free policies in healthcare environments. Most participants reported that these policies were not effectively enforced and that many lacked proper training in providing smoking cessation support9.
In Iraq, especially in northern regions like Kirkuk, there is limited information on smoking habits among HCWs. It is crucial to understand the prevalence of smoking and the factors related to it in this group in order to develop effective intervention strategies. HCWs in Iraq face unique challenges, such as job-related stress, heavy workloads, and the influence of community smoking norms, all of which can affect their smoking behaviors and motivation to quit.
To address this gap, the current study was conducted to assess the smoking prevalence among HCWs in Kirkuk and to investigate the factors associated with smoking behavior, including sociodemographic and occupational aspects. The research also examined motivational factors related to quitting smoking through multivariable regression and factor analysis.
METHODS
Study design and setting
A cross-sectional study was carried out among HCWs at public hospitals and primary healthcare centers in Kirkuk Governorate, located in northern Iraq. Data collection took place from January to August 2025. Kirkuk was chosen for this study because it reflects a diverse and rapidly evolving region, encompassing both urban and semi-urban populations, which offers a representative snapshot of healthcare personnel in northern Iraq.
Study population and sample size
The study aimed to include all types of HCWs, such as physicians, dentists, pharmacists, nurses, laboratory staff, medical technologists, and administrative or support personnel. Participants were eligible if they were currently employed at a healthcare facility in Kirkuk, were aged ≥20 years, and were willing to take part in the study. Those who were temporary or non-medical staff, as well as individuals who chose not to participate, were excluded.
The sample size was calculated a priori using a single population proportion formula based on an expected smoking prevalence of 20%, derived from previous studies among HCWs10,11, with a 95% confidence level and a margin of error of 3%. After adjusting for a design effect of 1.5 and an anticipated non-response rate, the minimum required sample size was estimated to be 1067 participants. To enhance statistical power and precision, a total of 1618 HCWs were ultimately included in the study. The detailed sample size calculation is provided in the Supplementary file.
Sampling technique
A stratified multistage cluster sampling method was used for the study. In the first stage, healthcare institutions in Kirkuk were categorized by type into hospitals and primary healthcare centers. In the second stage, a random selection of facilities was made from each category. In the third stage, HCWs within each chosen facility were sampled in proportion to their professional roles, including physicians, nurses, pharmacists, technicians, and administrative staff. This approach ensured that all types of facilities and healthcare professions were adequately represented throughout the governorate.
Data collection tool
A structured, self-administered questionnaire was utilized, adapted from previously validated tools used in similar studies involving healthcare professionals8-11. The questionnaire consisted of four main sections. The first section collected sociodemographic and occupational information, including age, gender, education level, profession, and length of service. The second section assessed smoking behavior, covering smoking status, types of tobacco products used, duration and frequency of smoking, and previous quit attempts. The third section focused on perceptions and attitudes toward smoking in the workplace and awareness of smoking-related health risks. The fourth section explored motivational and support factors related to smoking and smoking cessation, including reasons for smoking or quitting and preferred cessation support.
The questionnaire was pretested with 30 HCWs who were not part of the final sample to evaluate its clarity, consistency, and the time needed for completion. Minor adjustments were made based on this feedback. The internal reliability of the questionnaire was deemed acceptable, with Cronbach’s alpha coefficients of 0.81 for attitude items and 0.84 for motivation items.
Data collection procedure
Data collection was conducted by trained public health researchers during official working hours. Participants were informed about the study’s objectives, assured of confidentiality, and made aware that their participation was voluntary. Completed questionnaires were reviewed for completeness before data entry.
Ethical considerations
Ethical approval was granted by the Scientific Research Ethics Committee of the Kirkuk Health Directorate (IRB 2024). All participants provided written informed consent. Participation was entirely voluntary, and data were anonymized to ensure confidentiality throughout the study.
Operational definitions
Smoking status was classified based on the type of tobacco product used. Combustible tobacco products included cigarettes and hookah (shisha), while non-combustible or novel tobacco products included electronic cigarettes (vaping devices). A current smoker was defined as a participant who reported using any combustible or non-combustible tobacco product within the past 30 days. A former smoker was defined as an individual who had previously used tobacco products but had abstained from all forms of tobacco use for more than one month. A non-smoker was defined as a participant who had never used any tobacco product.
Reasons for smoking were assessed using predefined questionnaire items. Participants were asked to indicate their perceived reasons for smoking, including stress and psychological pressure, social influence (friends or family), and addiction. The term ‘addiction’ referred to self-perceived dependence on tobacco, as reported by the participants, and was not based on a clinical diagnosis or a validated nicotine dependence scale. Participants were allowed to select more than one reason, and overlap between categories was possible.
Statistical analysis
Data were analyzed using the Statistical Package for the Social Sciences (SPSS) version 26 (IBM Corp., Armonk, NY, USA)12. The distribution of continuous variables was assessed for normality using visual inspection of histograms and the Shapiro–Wilk test. As the data showed no substantial deviation from normality, continuous variables are presented as mean ± standard deviation (SD). Descriptive statistics were employed to summarize the variables, presenting means ± standard deviations for continuous data, and frequencies and percentages for categorical data.
Bivariate associations between smoking status and independent variables were assessed using the chi-squared (χ²) test. Variables with p<0.25 in the bivariate analysis were entered into a multivariable binary logistic regression model using a forward stepwise (likelihood ratio) method to identify independent predictors of current smoking. The final adjusted model retained gender, age group, duration of work, family member or close friend who smokes, and perceptions of smoking in the workplace, and adjusted odds ratios (AORs) with 95% confidence intervals (CIs) are reported.
An ordinal logistic regression model was used to explore predictors of smoking status categorized as non-smoker, former smoker, or current smoker. The model’s fit was evaluated using Nagelkerke R², the Hosmer–Lemeshow test, and the test of parallel lines to check for proportional odds assumptions.
Principal component analysis (PCA) with Varimax rotation was conducted on items related to motivations for quitting smoking to extract underlying factors. Components with eigenvalues >1.0 were retained, and factor loadings ≥0.40 were deemed significant.
RESULTS
A total of 1618 HCWs took part in the study, representing a range of professional categories from hospitals and primary healthcare centers in Kirkuk Governorate. The mean age of participants was 35.7 ± 9.7 years, indicating that most were in the early to mid-career stage.
Table 1 illustrates the demographic and occupational characteristics of the participants. Among the total participants, 54.8% were female, and 45.2% were male. The average age was around 35.7 years, with nearly half of the participants (47.6%) falling into the 20–30 age group, followed by 27.6% in the age of 31–40 years range and 15.6% in the age group of 41–50 years. Nurses made up the largest professional group at 25.0%, followed by assistant medical staff at 20.5% and laboratory staff at 15.8%, while physicians represented only 3.0%. Most participants (48.9%) held a technical diploma, with 38.7% having a Bachelor’s degree and 4.2% possessing postgraduate education. Almost two-thirds (66.1%) had <10 years of work experience. In terms of smoking behavior, the majority of participants were non-smokers (83.6%), while 7.0% were current cigarette smokers, 4.2% smoked hookah (shisha), 2.3% used electronic cigarettes, and 2.9% were former smokers.
Table 1
Sociodemographic characteristics of HCWs in Kirkuk, Iraq, January–August 2025 (N=1618)
Table 2 presents the smoking-related characteristics of the participants. Current smokers made up 13.5% of the study population, while former smokers accounted for 2.9%. Among those who currently smoke, the duration of smoking varied: 32.0% had smoked for 1–5 years, 26.5% for 5–10 years, and 28.8% for >10 years. In terms of smoking intensity, half of the smokers (50%) reported smoking 10–20 cigarettes per day, while 28.1% smoked >20 cigarettes daily.
Table 2
Smoking patterns, behaviors, and perceptions among HCWs in Kirkuk, Iraq, January–August 2025 (N=1618)
The primary reasons for smoking included stress and psychological pressure (48.4%), followed by social influences (34.2%) and self-perceived addiction (17.4%). Despite these challenges, a considerable number of participants (69.2%) had tried to quit smoking at least once. Among those who attempted to quit, 74.5% were unsuccessful, while only 25.5% managed to quit successfully. The most common reasons for wanting to quit were health concerns (66.0%), family influence (21.3%), and financial considerations (12.8%). Most former smokers (59.6%) had quit within the past one to five years. Additionally, a majority (62.4%) reported having family members or close friends who smoke, which may normalize smoking behavior in their social circles.
While awareness of the health risks associated with smoking was high (92.9%), this knowledge did not prevent some individuals from smoking. Attitudes towards smoking in the workplace were largely negative, with 88.6% opposing or strongly opposing smoking at work. In terms of support for cessation, 56.9% preferred health education and awareness campaigns, while 23.4% favored pharmacological aids, and 19.7% sought medical consultation.
Table 3 illustrates the bivariate associations between smoking status and various characteristics. Gender was found to have a strong correlation with smoking (p<0.001), with 27.9% of males identified as smokers compared to just 1.7% of females. Although the age group did not show a significant relationship to smoking (p=0.073), there was a slightly higher prevalence among younger workers. Occupation demonstrated a significant association (p=0.002), with administrative staff and dentists exhibiting higher smoking rates than other professions. Education level also had a significant impact (p=0.032), as participants with lower education level reported higher smoking rates. The duration of work did not reveal a significant relationship (p=0.071).
Table 3
Association between smoking status and sociodemographic characteristics among HCWs in Kirkuk, Iraq, January–August 2025 (N=1618)
A strong link was identified between smoking and having family members or close friends who smoke (p<0.001). Awareness of the health effects of smoking did not significantly influence smoking behavior (p=0.331). Lastly, attitudes towards smoking in the workplace were significantly associated with smoking status (p<0.001); those who opposed smoking at work were notably less likely to smoke themselves.
Table 4 displays the findings from the multivariable binary logistic regression model. After adjustment in the multivariable model, gender, years of service, family smoking history, and perceptions of smoking in the workplace were identified as significant predictors. Male HCWs were found to be significantly more likely to smoke than their female counterparts (AOR=28.1; 95% CI: 15.9–49.6, p<0.001). Participants with family members who smoke had more than three times the odds of being smokers themselves (AOR=3.22, p<0.001). In contrast, those with longer tenures (>20 years) were less likely to smoke (AOR=0.27, p=0.004), suggesting that smoking prevalence was lower among HCWs with longer work duration. Workers who strongly opposed smoking in their workplace were significantly less likely to smoke (AOR=0.10, p<0.001). The model showed strong predictive capability (Nagelkerke R²=0.457) and demonstrated a good overall fit (Hosmer–Lemeshow p=0.930).
Table 4
Binary logistic regression analysis of factors associated with smoking status among HCWs in Kirkuk, Iraq, January–August 2025 (N=1618)
| Predictor variable | Smoking status | ||||
|---|---|---|---|---|---|
| n | OR (95% CI) | p | AOR (95% CI)* | p | |
| Gender | |||||
| Female (ref.) | 886 | 1 | 1 | ||
| Male | 732 | 22.45 (13.13–38.32) | 0.001 | 28.10 (15.91–49.60) | 0.001 |
| Age (years) | |||||
| 20–30 (ref.) | 770 | 1 | 1 | ||
| 31–40 | 447 | 1.15 (0.83–1.61) | 0.387 | 1.97 (1.24–3.12) | 0.004 |
| 41–50 | 252 | 0.98 (0.65–1.49) | 0.954 | 2.02 (0.98–4.33) | 0.057 |
| >50 | 149 | 0.51 (0.26–0.97) | 0.039 | 0.80 (0.29–2.21) | 0.679 |
| Occupation | |||||
| Administrative staff (ref.) | 196 | 1 | |||
| Dentist | 125 | 1.49 (0.85–2.61) | 0.161 | - | - |
| Assistant medical | 331 | 0.66 (0.39–1.09) | 0.104 | ||
| Laboratory staff | 255 | 0.58 (0.34–1.01) | 0.055 | ||
| Medical technologist | 160 | 0.75 (0.41–1.35) | 0.333 | ||
| Nursing staff | 405 | 0.62 (0.38–1.01) | 0.052 | ||
| Pharmacist | 98 | 1.35 (0.73–2.48) | 0.339 | ||
| Physician | 48 | 0.45 (0.15–1.34) | 0.150 | ||
| Education level | |||||
| Secondary/preparatory school (ref.) | 133 | 1 | - | - | |
| Technical diploma | 791 | 0.64 (0.38–1.06) | 0.085 | ||
| Bachelor's | 626 | 0.97 (0.58–1.61) | 0.908 | ||
| Master's or PhD | |||||
| Duration of work (years) | 68 | 0.58 (0.23–1.43) | 0.237 | ||
| 1–10 (ref.) | 1070 | 1 | 1 | ||
| 11–20 | 330 | 0.82 (0.57–1.18) | 0.287 | 0.59 (0.32–1.07) | 0.085 |
| >20 | 218 | 0.58 (0.36–0.95) | 0.031 | 0.27 (0.12–0.65) | 0.004 |
| Family member or close friend who smokes | |||||
| No (ref.) | 609 | 1 | 1 | ||
| Yes | 1009 | 2.87 (2.01–4.09) | 0.001 | 3.22 (2.13–4.86) | 0.001 |
| Aware of the adverse health effects of smoking | |||||
| No (ref.) | 115 | 1 | |||
| Yes | 1503 | 0.78 (0.46–1.29) | 0.333 | - | - |
| Perceptions of smoking in healthy workplaces | |||||
| Strongly support smoking in the workplace (ref.) | 30 | 1 | 1 | ||
| I support smoking in the workplace | 19 | 1.18 (0.37–3.73) | 0.782 | 1.49 (0.31–7.23) | 0.619 |
| Neutral | 136 | 1.03 (3.93–12.39) | 0.938 | 0.76 (0.28–2.09) | 0.596 |
| I oppose smoking in the workplace | 43 | 0.29 (0.14–0.62) | 0.001 | 0.30 (0.11–0.76) | 0.013 |
| I strongly oppose smoking in the workplace | 1030 | 0.09 (0.04–0.19) | 0.001 | 0.10 (0.03–0.25) | 0.001 |
| Constant | |||||
| Model χ2 (df) | 15.225 (8) | ||||
| Nagelkerke R² | 0.457 | ||||
| Hosmer–Lemeshow p | 0.930 | ||||
The results of the rotated component matrix obtained from the principal component analysis (PCA) are presented in Supplementary file Table S1. Two primary motivational factors for smoking cessation were identified. The first, termed health-driven motivation, included items such as health concerns (loading=0.77), financial reasons (0.79), and influence from family or friends (0.84). This factor reflects intrinsic motivations for quitting. The second factor, support-oriented motivation, comprised items like health education (0.81), medical consultation (0.74), and pharmacological aids (0.69), representing external supports that aid in the quitting process. All factor loadings were >0.40, indicating strong construct validity.
Two motivational components related to smoking cessation were identified through principal component analysis. Together, these components accounted for 69.5% of the total variance. The first factor, health-driven motivation, explained 40.2% of the variance and reflected concerns related to health and financial consequences. The second factor, support-oriented motivation, accounted for 29.3% of the variance and encompassed external support mechanisms such as health education, medical consultation, and pharmacological assistance. A summary of the extracted motivational factors and the variance explained is provided in Supplementary file Table S2.
DISCUSSION
This study investigated the prevalence of smoking, behaviors, and determinants among HCWs in the Kirkuk Governorate of Iraq, while also exploring motivational factors for quitting. The findings indicate a significant smoking prevalence among healthcare professionals, with strong connections identified between gender, social exposure to smokers, workplace attitudes, and smoking behavior.
The smoking prevalence we observed among HCWs in Kirkuk is consistent with global systematic review data, which indicate an overall prevalence of around 21% among healthcare professionals, with higher rates in males (approximately 31%) compared to females (around 17%)7. This finding highlights the ongoing issue of tobacco use, even among health professionals who are expected to serve as role models. With regard to novel tobacco products, the prevalence of electronic cigarette use in the present study was lower than that of traditional combustible tobacco products, a pattern that has also been reported in several Middle Eastern settings, whereas higher e-cigarette use has been observed among healthcare workers in some European and North American studies13-15.
In regional studies, such as one conducted in northern Jordan, the prevalence of smoking among primary healthcare professionals was found to be around 25%, which coincided with weak enforcement of smoke-free policies9. Our findings highlight the potential relevance of workplace policies and attitudes, with opposition to smoking in the work environment being associated with a lower likelihood of smoking among HCWs. This association suggests that organizational culture and workplace policies may influence smoking behaviors, although causality cannot be established.
Multivariable and ordinal logistic regression analyses identified male gender, the presence of smoking family members or friends, shorter professional experience, and negative perceptions of workplace smoking policies as independent predictors of current smoking. The finding that male gender is a strong predictor aligns with numerous studies, as smoking prevalence among male physicians globally remains higher than among their female colleagues3. The influence of social exposure, where family and friends smoke, supports the concept of social modeling, indicating that HCWs are not immune to peer or familial influences, which correlate with increased smoking rates16. Shorter professional experience may point to a younger age, early-career stress, or weaker adherence to institutional norms17.
The protective association of opposing workplace smoking suggests that a supportive institutional culture and clear norm-setting can effectively reduce smoking among HCWs18,19. Therefore, focusing solely on individual-level factors is insufficient; organizational and policy-level interventions are essential to promote smoking cessation.
The principal component analysis in this study identified two key motivational constructs: 1) Health-driven motivation, which includes concerns about personal and family health as well as financial reasons; and 2) Support-oriented motivation, which encompasses education, medical counseling, and pharmacotherapy. The variance explained by these constructs highlights that both intrinsic and extrinsic factors significantly influence smoking cessation among HCWs. This dual-mechanism finding aligns with existing literature on cessation, which suggests that the motivation to quit is multifaceted; health concerns alone are often insufficient without supportive structures in place. For instance, a study conducted in Cyprus revealed that while approximately 72% of HCWs expressed a strong desire to quit smoking, only about 5–7% utilized cessation aids, illustrating a significant gap between intention and action8.
In addition, patterns such as dual or multiple tobacco product use deserve further attention, as some healthcare workers may concurrently use cigarettes, hookah, or electronic cigarettes, potentially increasing nicotine exposure and dependence13,20. The workplace environment may also contribute to nicotine dependence, particularly in settings characterized by high occupational stress, shift work, or social acceptance of smoking21. Furthermore, initiation of smoking during early employment in healthcare settings may be influenced by peer behaviors, workplace norms, and job-related stressors2. These considerations underscore the importance of preventive measures tailored to healthcare professionals, including early workplace-based education, stress management interventions, strict enforcement of smoke-free policies, and accessible cessation support services22.
Strengths and limitations
The strengths of this study include its large sample size, representation from various healthcare professions, and the combined use of bivariate, multivariable, and factor analysis methods, which together offer a comprehensive understanding of the determinants and motivations related to smoking. Conducting the study in Kirkuk is particularly valuable, as there are limited data available from this region. However, there are several limitations to consider. The cross-sectional design prevents the establishment of causal relationships; while associations can be identified, we cannot confirm the direction of these relationships. Additionally, smoking status was self-reported, which may lead to an underestimation of prevalence due to social desirability bias, especially among HCWs. Lastly, although motivational factors were examined through principal component analysis, qualitative data could provide deeper insights into the underlying psychosocial mechanisms. Furthermore, as the study was conducted in a single governorate, the findings may not be fully generalizable to healthcare workers in other regions of Iraq or in different national and healthcare system contexts. In addition, the reliance on questionnaire-based self-reported data introduces the possibility of information bias, including recall bias and social desirability bias, particularly given the professional role of healthcare workers. Moreover, smoking status was not validated using biochemical measures such as cotinine testing; therefore, some degree of misclassification of tobacco exposure cannot be excluded. Nevertheless, self-administered questionnaires remain a widely accepted and practical method for assessing smoking behavior in large epidemiological studies.
CONCLUSIONS
Smoking continues to be a significant issue among healthcare workers in Kirkuk, Iraq. The main factors associated with smoking include male gender, social exposure to smokers, shorter professional experience, and permissive workplace attitudes toward smoking. Motivations for quitting appear to be driven by concerns about personal health and financial costs, as well as the availability of external support. Addressing smoking among HCWs may benefit from a comprehensive approach that integrates individual-level cessation support, such as counseling and pharmacotherapy, alongside the reinforcement of smoke-free workplace policies. Targeted interventions focusing on younger healthcare professionals and those more vulnerable to social influences could help inform efforts to strengthen tobacco control within healthcare settings.
