Determinants of Willingness to Undergo Lung Cancer Screening among High-Risk Current and Ex-smokers in Sabah, Malaysia: A Cross-Sectional Pilot Study

1. Study design and participants

This was a cross-sectional, interviewer-administered questionnaire study of current or former smokers presenting as inpatients to our Medical and Respiratory Units of the respective study sites. Potential respondents were opportunistically identified by the investigators and invited to participate in the study if they fulfilled the inclusion criteria. The questionnaire was made available in both English and Malay languages.

The study was conducted in two public hospitals and one private hospital located in Kota Kinabalu, Sabah, Malaysia. Recruitment and data collection were carried out between May 2022 to November 2022. We included individuals aged 55 to 80 years, who were not known to have lung cancer, and were current smokers with a history of smoking minimum 20 pack-years or ex-smokers with a history of smoking minimum 20 pack-years and had stopped smoking within the last 15 years. The minimum age of 55 years was determined, as the objective risk calculator, PLCO_m2012norace model, only accounted for individuals who were aged 55 years and above [7]. Meanwhile, the smoking criteria were adapted from the United States Preventive Services Task Force (USPSTF) criteria [8]. We excluded individuals with a known lung cancer diagnosis, and those who did not fulfill the age or smoking criterion described above.

2. Bilingual questionnaire design and validation

The questionnaire was adapted from two validated questionnaires, with permission from their respective authors. The two validated questionnaires consisted of a study comparing the absolute and perceived risk [9], and another study on the psychological determinants of lung screening uptake [10]. Malay translation was performed as follows: the content was qualitatively validated in terms of relevance, importance, and coverage by three authors who were subject-matter experts. Bilingual comprehensibility and accuracy were ascertained by forward-backward translation. Forward translation from English to Malay was separately and independently conducted by a linguistic expert and subject-matter expert, followed by backward translation from Malay to English by a linguistic expert and subject-matter expert. Then, three authors (Huan NC, Rosli T, and Ramarmuty HYD) reviewed the separate versions of the questionnaire and finalized the pre-final translated version in Malay. Face validation was then conducted with five participants, and two authors (Nyanti LE and Rosli T) reviewed and finalized the final translated version in Malay.

3. Questionnaire content

The questionnaire was divided into three sections, with a total of 34 questions. Section one consisted of 13 questions covering the demographics, smoking history, perceived lung cancer risk, and objective lung cancer risk, and it was adapted with permission from a validated survey from an Australian study [9]. Minor changes were made to the question on race and ethnicity to reflect Malaysian racial demographics. Questions on objective lung cancer risk were derived using the Prostate Lung Colon Ovarian Cancer 2012 risk prediction model excluding the race and ethnicity predictor (PLCO_m2012norace) [8]. A 1.5% risk threshold was used, as it was the threshold at which the LDCT mortality reduction benefits over a chest radiograph began [11].

Section two presents an overview of LDCT screening, including benefits and harms to the respondent, and it was also adapted with permission from the Australian study [9]. Section three consisted of 21 questions adapted from the Self-Regulatory Questionnaire for Lung Cancer Screening (SQR-LCS) questionnaire [10], which is a validated measure of psychological factors that affect lung cancer screening uptake. The scope of the questions encompassed several psychological constructs, such as consequences (question 14), personal control (question 15), treatment control (question 16), illness coherence (question 18), emotional representation (question 19), behavioral response and appraisal (questions 20, 21, and 22), risk perception (question 32), response efficacy for smoking cessation (question 33), treatment intention (question 24), perceived stigma (question 23), and lung cancer survival (question 17). Additional questions not covered in either prior studies (questions 10 and 11) were integrated; these were questions related to the local Malaysian setting, such as costs incurred for screening and treatment of lung cancer (questions 25 and 26), perceived time burden to undergo screening (question 27), perceived logistical burden to undergo screening (questions 28 and 29), perceived community stigma towards lung cancer (question 30), and cultural beliefs related to cancer treatment (question 31). Willingness of the respondent to undergo lung cancer screening was indicated in the final question (question 34).

4. Statistical analysis

The sample size was calculated according to the prevalence formula [12], which was derived from a smoker prevalence of 13% in Sabah state, as reported in the National Health and Morbidity Survey (NHMS) 2019 [13]. We used a z value of 1.96, corresponding to a confidence coefficient of 0.95; and a precision, or margin of error of 7.5%. The minimum sample size needed to achieve a 7.5% precision in estimating the prevalence was 77 subjects.

Demographic characteristics were presented using descriptive analyses, namely frequencies, percentages, means and standard deviations, or medians and interquartile ranges. Group comparisons between patients who preferred and did not prefer screening were performed using the chi-squared test, t-test, or Wilcoxon Signed Rank test, as appropriate. A p-value <0.05 was considered significant. Missing data were not imputed. Analysis was performed using SPSS version 28 (IBM Co., Armonk, NY, USA).

5. Ethics approval

This study received ethics approval from the Medical Research & Ethics Committee, Ministry of Health Malaysia (NMRR ID-22-00669-QWZ). Written informed consent was obtained from all patients.

1. Demographic and smoking characteristics

We interviewed a total of 95 respondents with a mean age of 63.3 years, and they were predominantly of male gender (n=92) and Sabahan native ethnicity (n=44). Overall, our respondents had a low to moderate level of education, with nearly two-thirds not having completed secondary school and nearly one-third who completed secondary school but did not progress further. Onefourth of the respondents had chronic obstructive pul-monary disease (COPD), while the prevalence of past tuberculosis infection was 16.8% (n=16). More than half of the respondents were overweight. There were only two respondents from the private hospital. The demographic details are presented in Table 1.

Table 1.

Demographic characteristics, smoking history, and perceived and absolute risk of the study participants

The earliest age of smoking commencement was 5 years, while the median age of smoking commencement and total duration of smoking was 17.5 and 41 years, respectively. One-third (n=32) of the respondents were current smokers, while the remaining twothirds were ex-smokers; with a median daily cigarette consumption of 20 cigarettes, and median exposure pack-years of 42 years. Among the ex-smokers, median years since quitting smoking was 6 years, but seven individuals (12.5%) conceded that they might start smoking again. A vast majority of current smokers expressed their intention to quit (n=27, 84.3%), but they varied in terms of the actions taken. Meanwhile, five current smokers had no intention of smoking cessation. Further data on the smoking history and perceptions are detailed in Table 1.

2. Absolute and perceived risk

Among our respondents, the median 6-year risk of developing lung cancer was 2.6% (Table 1). Interestingly, 18 respondents (18.9%) had a risk of less than 1.5%, which made them ineligible for screening. The risk-reducing characteristics of these 18 respondents, included having smoked less in terms of the duration and quantity of cigarettes, a higher body mass index, absence of COPD, absence of personal or family history of other cancers, being a former smoker, and a higher level of education.

In terms of perceived risk, the majority of the respondents perceived themselves to be at an equal or lower risk of developing lung cancer compared to non-smoking individuals of the same age and gender. A greater proportion of these respondents were eligible for screening compared to those who perceived themselves to be at higher risk (Figure 1). Only one-third (n=29) of the respondents perceived themselves to be at higher risk of developing lung cancer; out of these, one-third had a PLCO_m2012norace 6-year risk of less than 1.5%, indicating that they may not actually require screening.

Fig. 1.

Eligibility for screening based on the absolute risk (perceived lung cancer risk and objective risk [PLCO_m2012norace] >1.5%) in the higher, same, and lower perceived risk groups.

Five respondents expressed their unwillingness to undergo LDCT screening. One of these five respondents demonstrated concurrence between the absolute risk (PLCO_m2012norace 6-year risk: 1.0%) and perceived risk; having accurately perceived himself to be at a lower risk than a non-smoking individual of the same age and gender. The remaining four respondents were eligible for screening, with a PLCO_m2012norace 6-year risk of 9.0%, 2.2%, 4.7%, and 1.9%, respectively. However, only one respondent perceived himself to be at a higher risk. The respondent with the highest absolute risk of 9.0% perceived himself to be at an equal risk of developing lung cancer compared to a non-smoking individual of the same age and gender.

3. Attitudes towards lung cancer screening, diagnosis, and management

Nearly two-thirds of the respondents indicated that they were unsure about the symptoms of lung cancer; thus, demonstrating a visible knowledge gap (Figure 2, Question 2). More than three-quarters of the respondents indicated that they would experience fear and anxiety if they were diagnosed with lung cancer (Figure 2, Question 1), but preferred to know about the lung cancer diagnosis at an early-stage and agreed that LDCT was beneficial in detecting early-stage lung cancer (Figure 3, Questions 5 and 6). More than three-quarters of the respondents felt that a diagnosis of lung cancer would burden their loved ones (Figure 2, Question 6). However, most of the respondents would not blame themselves if they were diagnosed with lung cancer (Figure 3, Question 3), and one-third of the respondents feared external stigma from others (Figure 4, Question 3).

Fig. 2.

Psychosocial attitudes towards lung cancer, screening, and treatment (part one of three).

Fig. 3.

Psychosocial attitudes towards lung cancer, screening, and treatment (part two of three). CT: computed tomography.

Fig. 4.

Psychosocial attitudes towards lung cancer, screening, and treatment (part three of three). CT: computed tomography.

Most of the respondents believed that early-stage lung cancer is curable, and that treatment is potentially curative (Figure 2, Questions 3 and 4), and they would be willing to undergo surgical resection if they were diagnosed with early-stage lung cancer (Figure 3, Question 2). However, one-third of the respondents would opt for traditional or alternative treatment options above standard medical care (Figure 4, Question 2). Nearly all respondents indicated increased willingness to undergo screening if the LDCT facility was located close to their homes. One-third of the respondents did not believe that smoking cessation would result in a reduction in cancer risk (Figure 4, Question 1).

A majority of the respondents were concerned about the financial implications of screening or treatment (Figure 3, Question 1; and Figure 4, Question 7). Most of the respondents disagreed that they had time or logistic constraints to undergo LDCT screening (Figure 4, Questions 5 and 6), but nearly all respondents demonstrated increasing willingness to undergo LDCT screening if the screening facility was located close to their homes.

4. Determinants of willingness to undergo screening

We used univariable binary logistic regression to examine the demographic or psychosocial factors affecting the willingness to undergo LDCT screening (Table 2). Desire to know about a lung cancer diagnosis early, having perceived time constraints to undergo screening, and having LDCT facilities closer to their place of residence were associated with significantly higher odds of willingness to undergo LDCT screening. Negative attitudes towards lung cancer treatment or ability of LDCT to detect early lung cancer, preference for traditional treatments above conventional cancer therapy, perceived burden towards loved ones, and higher education level demonstrated reduced odds to undergo screening, but they were not statistically significant.

Table 2.

Univariable analysis of the demographic characteristics and psychosocial factors contributing to willingness to undergo LDCT screening