Abstract
Background:
Smoking is a known trigger of asthma exacerbation. Yet some patients with asthma continue to smoke. Little is known about their smoking literacy, which is defined as an individual’s capacity to comprehend and use healthcare information relating to smoking. This study aimed to determine the level of smoking literacy amongst patients with asthma, who were smokers (S) and non-smokers (NS).
Methods:
Standardised questionnaires were administered by trained research assistants to adult multi-ethnic Asian patients with asthma, who were managed in three public primary care clinics (polyclinics) in Singapore. Demographic data and information on a subject’s understanding and effects of smoking on health and complications were collected. The latter were computed into knowledge scores and the mean scores between S and NS were compared using the independent two-sample t-test with adjustment for potential confounders via multiple linear regression.
Results:
A total of 174 S and 183 NS were recruited with a higher proportion of younger male S. Overall, the knowledge score for S was significantly lower than NS (absolute mean difference (AMD) = −0.95, 95% confidence interval (CI) −1.58 to −0.33, p = 0.003). This was particularly for smoking associated specific respiratory disease (AMD = −0.23, 95% CI −0.44 to −0.02, p =0.034) and asthma symptoms (AMD = −0.73, 95% CI −0.98 to −0.47, p < 0.001). The differences persisted after adjusting for other important factors such as gender and education level.
Conclusions:
Compared with NS, S generally reported lower scores in their overall knowledge, and the knowledge in the association between smoking and specific respiratory diseases and asthma symptoms. Smoker patients may require measures to raise their smoking literacy during the counselling to optimise their smoking cessation success rates.
Keywords
Introduction
Health literacy is defined as ‘the degree to which individuals have the capacity to obtain, process and understand basic health information and services needed to make appropriate health decisions’. 1 Rosas-Salazar et al., 2 who have succinctly summarised numerous studies, showed that low levels of education and limited health literacy were associated with poor health and poor healthcare utilisation, increased barriers to care and early death. Health literacy tends to be broad-based and covers a wide spectrum of health related activities and behaviour. Health literacy can be scoped down to focus the attention on the level of awareness and practice of specific diseases or habits such as smoking. Suka et al. reported that individuals with higher health literacy were less likely to have risky habits of smoking. 3 However, they did not specify if higher level of literacy on the specific effects of smoking would be associated with lower smoking rates.
Hence we define ‘smoking literacy’ as an individual’s capacity to comprehend and use healthcare information relating to smoking. A survey of 1600 Indian subjects from rural and urban populations in Punjab had shown that while 96% were aware that smoking was harmful to health, 25% of them continued to smoke. 4 This finding seems to reflect that smoking literacy varied between subgroups in a population.
One subset of the population who will be more susceptible to the ill effects of smoking comprises patients with bronchial asthma. Tobacco smoke is a known trigger of asthma exacerbation, which can be potentially life-threatening. The cigarette constituents in the smoke settling on the airways will trigger inflammation and result in bronchospasm. 5 Smoking literacy becomes particularly important for patients with asthma. These patients need to appreciate the adverse effects of smoking on their airways and take appropriate measure to mitigate their asthma exacerbations, including smoking cessation. Hence, assessing their smoking literacy is a strategic step to be incorporated in a quit-smoking programme.
As a proportion of patients with asthma continued to smoke, we postulated that the smoking literacy level differed between the smokers and non-smokers. This study thus aimed to determine the level of smoking literacy amongst adult patients with asthma, who were smokers (S) as compared to non-smokers (NS). The findings from this study will provide healthcare professionals with an insight into the smoking literacy level amongst patients with asthma. The results will be used to design targeted education programme to enable them to understand and to facilitate their smoking cessation efforts.
Method
Setting
This study was conducted in SingHealth Polyclinics, a public primary care institution which managed 1.7 million patient attendances in Singapore in 2014. Forty-five per cent of the multi-ethnic Asian population in Singapore with chronic diseases, including asthma, are managed in public primary care clinics (polyclinics). 6 Three such polyclinics, located at Pasir Ris, Bedok and Geylang, respectively, on the eastern and southern regions of the island-state where the physician and the nurse clinicians in the investigator team are stationed, served as the study recruitment sites. The study was approved by the SingHealth Centralised Institutional Review Board (CIRB no: SHP 2006/01/A001).
Study population
The target participants in this study were adult patients with asthma who visited the primary care clinics for their respiratory condition at least twice annually. They included Singaporean or permanent residents, aged 21–50 years whose diagnosis was based on disease coding of ‘asthma’ using ICD-9 classification in their electronic health records. The age limit was set at 50 years to reduce the probability of recruiting patients with chronic obstructive pulmonary disease (COPD) or asthma–COPD overlap syndrome (ACOS), as no spirometry was carried out to exclude these respiratory conditions. Patients with cognitive and hearing impairment detected via direct interaction during the recruitment process and supported by existing diagnosis from electronic medical records or those who refused consent were excluded.
Based on the hypothesis that smoking literacy level differed between patients who smoked and those who did not, a ‘case-control’ approach was adopted for this study. The ‘cases’ were ‘smokers’ (S) whereas the comparator group comprised of ‘non-smokers’ (NS) based on their self-reporting. Smokers were asthmatic patients who had smoked at least one cigarette within the preceding 6 months of the survey or asthmatic patients who had recently quitted smoking for less than 6 months. In the local setting, most of those who self-reported to have quit smoking recently or have often lit cigarettes periodically, will remained classified as ‘S’. Non-smokers were asthmatic patients who had never smoked before.
Sample size calculation was based on the mean overall literacy score which was postulated to be 14 for the NS group and 13 for the S group, with a common standard deviation of 3. This postulation was estimated with reference to a study by William et al., 7 in which the knowledge score ranged from 11.9 to 15.1 (with SD ranging from 2.5 to 2.8 amongst respondents with asthma). Under the assumption of 5% significance level, 85% power and equal number of participants in NS and S, 163 patients were needed in each group or about 326 subjects in total.
Study instrument
The investigators are primary healthcare physicians and nurses who manage patients with asthma in their daily clinical practices. They developed the questionnaire as there was no precedent or existing instrument for assessing smoking literacy. The questionnaire collected data on demography, smoking-related literacy, including its effects on asthma control and associated diseases. A pilot study was carried out on 10 patients in one study site to establish face validity of the instrument. Minor amendments to the questionnaire were made prior to commencement of the main study.
Smoking literacy was assessed by computing the scores of correct responses to the questions in the respective domains of the questionnaire. The 20 items in the questionnaire were grouped into four domains: (1) general understanding of cigarette smoking; (2) effects of smoking on general diseases; (3) effects of smoking on specific respiratory diseases; and (4) effects of smoking on asthma symptoms. For each question, a participant was given a score of ‘1’ if he/she answered the question correctly, and ‘0’ otherwise. The total score for each domain was the aggregate score of the number of ‘correct’ answers. It could range from 0 to n, according to the number of questions within each domain. The higher the aggregated scores for the entire questionnaire, the higher the level of the smoking literacy.
The survey was conducted in English as it is the main language of communication in Singapore. Each study site has an approved list of primary healthcare professionals who are multi-lingual. For participants who were less fluent in English, translation of the questionnaire was conducted with the assistance of the enlisted translators.
Subject recruitment
Nurse counsellors screened potential patients with asthma during their counselling session and recruited those who satisfied the inclusion and exclusion criteria on a consecutive case encounter basis at the study sites. After obtaining written informed consent, participants were interviewed face-to-face by trained investigators using standardised structured questionnaires. The recruitment period lasted from March 2007 to December 2009.
Statistical analysis
Wilcoxon rank sum test and
Results
Three hundred and seventy-two participants were approached at the three designated study sites, of which 357 completed the questionnaire, generating a response rate of 96%. The study population comprised 174 S and 183 NS. The demographic profile of the participants is presented in Table 1. More S were males, singles, and younger as compared with NS.
Demographic characteristics of 357 study participants.
Table 2 summarises the level of knowledge of the participants according to the four domains.
Literacy comparison between smokers (S) and non-smokers (NS) based on their correct answer to questions.
‘n (%)’ refers to the number (percentage) of subjects who answered ‘Yes’ to the question.
General understanding of effects of cigarette smoking
More S understood the varying nicotine content of the different brands of cigarettes (44% vs. 28%, p = 0.003).
Effects of smoking on major diseases
There was a significantly higher proportion of S (95% vs. 87%, p= 0.010) who were aware of the association between acute myocardial infarction and smoking. More S perceived that smoking would lead to early death (33% vs. 25%, p= 0.068). No difference was noted between the S and NS on their recognition of other complications.
Effects of smoking on specific respiratory diseases
A lower proportion of S understood the association between smoking and specific respiratory diseases as compared to NS, although the differences did not reach statistical significance.
Effect of smoking on asthma symptoms
S generally scored lower than NS in recognising the association between smoking and asthma symptoms. Fewer S were aware that passive smoking increased the risk of asthma amongst children (87% vs. 95%, p = 0.006) and that cigarette smoke irritated the throat (71% vs. 95%, p < 0.001) or increased phlegm production (79% vs. 90%, p = 0.007). Similarly, fewer S reckoned the association between smoking and asthma attacks (28% vs. 18%, p = 0.022). More S perceived that smoking cessation would lead to coughing (46% vs. 32%, p=0.008).
Table 3 shows the mean scores in each domain and the overall knowledge scores of S and NS. The comparison revealed a significantly lower overall knowledge score of S as compared to NS (absolute mean difference (AMD) = −0.95, 95% confidence interval (CI) −1.58 to −0.33, p = 0.003). More specifically, S demonstrated lower knowledge score in the association between smoking and specific respiratory diseases (AMD = −0.23, 95% CI −0.44 to −0.02, p = 0.034) and asthma symptoms (AMD = −0.73, 95% CI −0.98 to −0.47, p < 0.001).
Comparison of mean (SD) knowledge scores between smokers and non-smokers.
The Cronbach’s alpha indices for domains 1, 2, 3 and 4 were 0.17, 0.51. 0.45 and 0.48, respectively.
The multiple linear regressions suggested that education level and gender were important demographic factors that contributed to the difference in knowledge scores between S and NS (Table 4).
Effect of smoking on knowledge scores adjusted for important factors by multiple linear regression.
Discussion
Overall, the majority of S and NS in this study population had adequate smoking literacy pertaining to the effects of cigarette smoking on health, especially on the much publicised complications such as heart attack, stroke, hypertension and miscarriage. Overall, smoking literacy among the S was shown to be significantly lower than those in the non-smoker group. It was found that with low smoking literacy, the odds of smoking increase among asthmatics. 8
The results revealed that education seemed to be the most significant predictor of participants’ knowledge score. Participants with higher education level (secondary school and above) generally demonstrated significantly higher overall knowledge. For example, participants with secondary education level on average scored 2.57 higher than those with none or primary education, while participants with junior college/diploma/polytechnic and tertiary education scored 3.35 and 3.16 higher, respectively. Similar trends were observed between education level and each of the four specific knowledge domains. Female had lower score in general understanding of cigarette smoking as compared to male (coefficient = −0.20, 95% CI −0.39 to −0.01, p = 0.037). However, the difference was not observed in other knowledge domains and overall knowledge.
Low health literacy has been shown to be an independent risk factor for poor smoking cessation outcomes among ethnically diverse smokers belonging to lower socioeconomic strata. 9 Nonetheless, inadequate health literacy can be surmounted by using tailored education on the target participants. 10 Little is known about specific smoking literacy and its relationship with smoking cessation outcome. There is a need to first raise the level of smoking literacy amongst the patients with asthma by addressing the current gaps shown in this study. Aside from smoking-related complications, smokers should be informed by their asthma educators of the effect of smoking on aggravating their nocturnal asthma symptoms and disturbing their sleep, and ultimately impairing their quality of life. 11 It is thus critical to enhance the smoking literacy amongst smokers with asthma to avert further deterioration of their asthma control status. Nevertheless, the modality to raise the smoking literacy needs to be evaluated for its effectiveness with further research.
Limitations
The study is a pioneering attempt to evaluate smoking literacy of local adult patients with asthma using knowledge scores. Simple lay language was used in the questionnaire to frame the items in the questionnaire. Although no validation study was conducted for this questionnaire, a pilot study was conducted to assess its utility, understanding and acceptance by the multi-ethnic Asian participants, resulting in amendments and a revised version prior to its application in the main study. Nonetheless, the Cronbach alpha indices were low; suggesting limited internal consistency of each domain of the questionnaire. The instrument would require further validation study to enhance its content for future utility.
Each interviewer conducted the survey at one of the two study sites. Both were not blinded to the smoking status of the subjects due to the selection of the cases and controls. Interviewer bias is another potential study limitation, as inter-rater reliability assessment could not be carried out.
A noteworthy consideration is the classification of the participants who self-reported to quit smoking in the past 6 months prior to study enrolment as ‘smokers’. It is known to the investigators that S continued to smoke occasionally in the early stage of their smoking cessation endeavour. The investigators decided to classify them as ‘smokers’ to be distinct from the other participants (‘non-smokers’), who had never smoked in their life time.
The study did not compute the number of non-respondents and response rates of patients who agreed to be referred to the counsellors, as suitable patients were approached at multiple nurse counsellors’ stations at the study sites, which constituted a limitation.
While the study sites were typical public polyclinics, smoking literacy of patients with asthma was not assessed at private GP clinics. Generalisation of the results to the general population should be restricted.
Conclusion
Smokers achieved lower scores in their overall knowledge, including the association between smoking and specific respiratory diseases and asthma symptoms. Measures to promote successful smoking cessation for smoker-patients would likely require interventions to raise the level of their smoking literacy.
Footnotes
Acknowledgements
The authors thank the polyclinic nurses at the study sites located at Pasir Ris, Bedok and Geylang, who had rendered their assistance to screen for suitable subjects before referring to the investigators on-site.
Declaration of Conflicting Interests
The authors declare that there are no conflicts of interest.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.