Comparing the Effectiveness of Combined NRT With Single NRT in Primary Care Clinics in Hong Kong

Background Globally, smoking causes six million deaths a year1. By 2030, if current trends continue, the number of deaths will rise to eight million. According to Hong Kong Thematic Household Survey Reports, although the prevalence of current smokers among aged 15 and over has dropped from 15.3% in 2006 to 10.8% in 2017 in Hong Kong, smoking still kills 5700 persons per year and contributes to 14% of all deaths from non-communicable diseases. Thus, enhancement in smoking cessation would be crucial in improving health all smokers.

Among all pharmacological treatment for smoking cessation, evidences showed that oral treatment such as Varenicline is the most clinically effective smoking cessation medication9. However, there have been concerns about its adverse effect on neuropsychiatric and cardiovascular aspects. Besides, many smokers are reluctant to use it due to fear of its high rate of side effects and being medicalized for smoking cessation. The use of nicotine replacement therapy (NRT) has been largely studied, and, over the last decade, many studies had been carried out to compare the effect of monotherapy with combined nicotine replacement therapy. While the effect of monotherapy had been found to be small in heavy smokers due to significant withdrawal symptoms, combined NRT is believed to provide a stable baseline nicotine level by means of nicotine patch plus intermittent usage of short acting NRT e.g. gums, lozenges or inhalers for withdrawal symptoms. Several studies have shown that combined NRT is associated with lower withdrawal scores and higher 6-month quit rates (26.9 to 36.9%) when compared with monotherapies (19-23%)18-21. Combined NRT has also been shown to be safe as trials of combining various NRTs did not report that combination treatment produced increased adverse events.

In Hong Kong, Hospital Authority is one of the major service providers for smoking cessation. The target recipients of our smoking cessation service are primarily patients attending public general out-patient clinics (GOPCs) for management of chronic illnesses such as hypertension and diabetes mellitus, as well as patients with episodic complaints. Through effective interventions provided in the smoking cessation service, it is aimed that their medical conditions could be further improved. However, as majority of these smokers have moderate or above nicotine dependence, how effective is combined NRT compared to single NRT in this Chinese population is unknown.

Hypothesis:

Combined NRT gives a higher abstinence rate than single NRT.

Objective:

This study aims to compare the effectiveness of combined NRT with single NRT in Hong Kong primary care clinics.

Assessment Patients were seen at baseline for assessment, and then at 4 weeks, 12 weeks, 26 weeks and 52 weeks. Study medication was given at baseline and at week 4. In baseline assessment, smoking history including daily cigarette consumption and past quitting method, past medical health, drug history and allergy would be obtained. In follow up visits patients were assessed on nicotine withdrawal symptoms, carbon monoxide level, side effects from treatment and medication compliance. Counselling would be given in all follow up visits.

Pharmacological interventions Patients were randomized to either intervention or usual care for smoking cessation. Nicotine replacement therapy (NRT) was given for 8 weeks in both arms. Intervention consisted of counseling and combined NRT of nicotine patch and gum. Usual care involved counseling and single NRT of nicotine patch. NRT patch regimen used in usual care was the same as that in intervention group.

Data analysis Baseline characteristics were reported and compared by treatment groups, two-sample t test was conducted for continuous variables and Chi-square test for categorical variables. At each visit, crude odds ratio (OR) (combined NRT vs. single NRT) was reported, simple logistic regression was utilized without adjustment first, and then adjusted for potential confounders, age, sex, baseline CO level and cluster site of the subject recruitment. The overall treatment effect (combined NRT vs. single NRT) over the study period was estimated by Generalized Estimating Equation (GEE). In the GEE model, time (repeated measures at 4, 12, 26, and 52 weeks) was included as a continuous covariate, adjusted odds ratio (AOR) was reported and treatment-time interaction was tested. Potential confounders were also adjusted for in the GEE model. As GEE only uses non-missing records, missing at complete random was assumed to avoid bias. To support the random missing assumption, supplementary analysis was conducted for missingness pattern Furthermore, imputation was conducted under a conservative scenario of 'all missing records as "failure", i.e., not quit'. Statistical significance level was set at two-sided p<0.05 for all tests. Analysis was conducted by R version 3.2.235. Side effects from NRT were also recorded from both groups. Chi-squared test was used to compare the difference.