Assessing the Effect of an Inpatient Smoking Cessation Brief Intervention Tool on Long-Term Smoking Cessation Rates and Patient Interest for Nicotine Replacement Therapy: A Retrospective Analysis

Introduction: Smoking cessation among hospital inpatients is essential to reduce risk of surgical complications and all-cause mortality. In the Australian state of Queensland, the Smoking Cessation Clinical Pathway (SCCP), a brief intervention tool, has been used by clinical staff in public hospitals to uniformly support patients to quit smoking since 2015. This study aims to assess the effect of the SCCP on long-term smoking cessation rates recorded in subsequent readmissions, and whether the SCCP as an intervention affects inpatients’ interest in nicotine replacement therapy (NRT) during admission and after discharge. Methods: We retrospectively analysed data provided by the Princess Alexandra Hospital (PAH) on patients who self-identified as a current smoker on admission to any ward and were admitted to the PAH between 1st January 2018 and 31st December 2019. Smoking cessation rates and patient interest in NRT by SCCP completion were analysed using χ2 tests and a multinomial logistic regression. Results: Of 1,902 included patients, NRT was offered to 1,397 patients (73.4%) and accepted by 332 patients (17.5%). Patients who had completed a SCCP were more likely to be offered NRT than those who had not (p < 0.0001). Of the 452 patients with multiple readmissions, 100 (22%) ceased smoking at any point in the 2-year study period. At the end of the 2-year study period, 75 (75%) patients remained abstinent and only 25 (25%) relapsed to smoking as per their final smoking status at the end of the 2-year study period. Patients with a completed SCCP were 1.8 times (RRR: 1.825, p = 0.030) more likely to quit smoking at any point in the 2-year study period, and twice as likely to have quit at the end of the 2-year study period (RRR: 2.064, p = 0.044). Discussion: The SCCP may be effective at increasing smoking cessation rates among hospital inpatients. Future policies promoting long-term smoking cessation should consider implementation of post-discharge follow-up appointments.

cessation rates and patient interest in NRT by SCCP completion were analysed using χ 2 tests and a multinomial logistic regression. Results: Of 1,902 included patients, NRT was offered to 1,397 patients (73.4%) and accepted by 332 patients (17.5%). Patients who had completed a SCCP were more likely to be offered NRT than those who had not (p < 0.0001). Of the 452 patients with multiple readmissions, 100 (22%) ceased smoking at any point in the 2-year study period. At the end of the 2-year study period, 75 (75%) patients remained abstinent and only 25 (25%) relapsed to smoking as per their final smoking status at the end of the 2-year study period. Patients with a completed SCCP were 1.8 times (RRR: 1.825, p = 0.030) more likely to quit smoking at any point in the 2-year study period, and twice as likely to have quit at the end of the 2-year study period (RRR: 2.064, p = 0.044). Discussion: The SCCP may be effective at increasing smoking cessation rates among hospital inpatients. Future policies promoting long-term smoking cessation should consider implementation of post-discharge follow-up appointments.

Introduction
Tobacco smoking is a leading cause of preventable mortality and morbidity in Australia; for example, in 2015, it was responsible for approximately 21,000 deaths and hospitalisations in Australia [1,2]. Smoking cessation is vital for hospital inpatients as it reduces risk of allcause mortality and surgical complications, as well as mitigating comorbidity severity [3]. In recognition of the importance of promoting smoking cessation in this setting, the Australian Government has legislated the implementation of smoke-free policies in public hospitals [4,5]. As patients admitted to these hospitals are thus forced to quit smoking for the duration of their hospitalisation, Australian health authorities offer evidence-based smoking cessation support to people who smoke upon admission [6][7][8]. In Queensland, this support is in the form of the Smoking Cessation Clinical Pathway (SCCP).
The SCCP, a brief intervention tool, has been used by clinical staff in public hospitals to uniformly support patients to quit smoking since 2015 [9]. The SCCP is intended to provide cessation support to patients who smoke throughout the duration of their hospitalisation, and ideally to encourage long-term smoking abstinence upon discharge [9]. The tool, provided as online supplementary Figure 1 (for all online suppl. material, see www. karger.com/doi/10.1159/000528864), is based on the Ottawa Model's 5A (Ask, Advise, Assess, Assist, and Arrange) approach [10]. First, the tool guides clinicians to Ask patients if they have smoked tobacco in the past 30 days, and then, among those who indicate they have, to Assess the nicotine dependence, desire to quit and current use of nicotine replacement therapy (NRT) or prescription medications. The tool then guides clinicians to Advise patients to stop smoking using clear and non-confrontational language, and to then Assist them to stop smoking through the provision of NRT patches or gum, and by advising them and referring them to Quitline telephone counselling. The final step in the form encourages clinicians to Arrange for the provision of a script for NRT upon discharge.
While several Australian studies have investigated the effects of hospital-based smoking cessation interventions [11][12][13], there has been limited research describing the efficacy and utility of the SCCP tool in promoting smoking cessation. A prior study found that the SCCP prompted clinicians to offer NRT to patients, leading to increased NRT prescribing and use. However, this study had a relatively limited timeframe of 3 months and data on patients' sociodemographic characteristics were unavailable [14].
A further study investigated the potential impact of the SCCP on smoking cessation among hospital inpatients, finding that smoking cessation rates were similar regardless of SCCP completion. However, the 12-month timeframe and lack of readmissions limited the conclusions drawn [15].
To build on these findings, this study aims to assess the effect of SCCP completion on long-term smoking cessation rates by examining patients' smoking status on subsequent readmissions, and whether the SCCP as an intervention affects inpatients' interest in NRT during admission and after discharge over a 2-year period. These findings will be useful in understanding the effectiveness of the SCCP in promoting smoking cessation among hospital inpatients.

Materials and Methods
This study involved a retrospective review of routinely collected data from patients' SCCP responses and the integrated electronic medical record system from the Princess Alexandra Hospital (PAH) in Queensland, Australia. Ethics approval for this study was received from Metro South Human Research Ethics Committee, approval number LNR/2019/QMS/59322.

Data Source
Data were extracted in June 2020 for a randomly selected list of inpatients who met the following inclusion criteria: self-identified as a current smoker on admission to any ward, and admitted to the PAH in Queensland between 1st January 2018 and 31st December 2019. Participants were excluded if they self-identified as having had no history of smoking, or if their admission date was outside the timeframe of the 1st January 2018 and 31st December 2019.

Measures
Data relating to participants' sociodemographic characteristics including age, gender, health conditions, and smoking status were extracted from the integrated electronic medical record. Readmission data were determined from admission dates and reoccurring unique patient identification numbers. Data relating to smoking in the last 30 days, nicotine dependence, number of cigarettes per day, wanting to quit, and NRT-related data were extracted from patients' SCCPs. Each participant's medical record was reviewed to determine if a SCCP was completed during their initial admission; whether they were taking NRT prior to admission; whether NRT was offered to the patient; whether the NRT was accepted or declined when offered; and whether NRT was prescribed on discharge (interest). For a small number of patients, the SCCP is not completed -for these patients, the dataset shows "unable to complete pathway," but no further information is provided. No formal follow-up was conducted, and thus, "follow-up" was based on readmission information available from the provided data. Smoking status at any readmission (that was not the final readmission) was recorded as one of two options: (1) quit (for those who self-reported as ex-smokers at a subsequent readmission) or (2) did not quit (for those who maintained their smoker status for all subsequent admissions). Smoking status at final readmission was recorded as one of three options: (1) quit (for those who selfreported as ex-smokers at a subsequent readmission); (2) relapsed (for those who had multiple admissions and self-reported being an ex-smoker and then a smoker at a later admission); and (3) unchanged (for those who maintained their smoker status for all subsequent admissions).

Analysis
After confirming that all statistical assumptions were met, χ 2 tests were used to compare inpatient NRT interest between patients who had or had not completed a SCCP. Next, a multinomial logistic regression was used to understand the relationship between the outcomes of smoking status at any previous admission and smoking status at final readmission, as predicted by completion of a SCCP at initial admission. For this part of the analysis, participants were excluded if they had no subsequent readmissions. Of the remaining eligible patients, the number of patients who had not completed a SCCP at initial admission was determined to be used as the control comparator. Equal-sized groups were preferable to maximise statistical power. Hence, a list of patients who had completed a SCCP was generated using a computerised list generator and patients were randomly selected for cohort matching. Given the lack of other sociodemographic information provided in the dataset and the proportion of gender and distribution of age, we opted for a randomised block design (blocking by SCCP completion). The study was designed to have 90% power with α = 0.05 to compare outcomes between two groups, resulting in a target sample size of at least 207 participants using a priori power analysis. A preliminary simple regression was conducted to check for any differences with sex, age, and intention to quit between those who did and did not quit smoking but found no statistically significant differences for our sample.
Patients were clustered by patient ID to account for standard error associated with repeated measures. We adjusted the analysis for age, sex, length of stay, and number of readmissions. Statistical significance was set at p < 0.05. Analyses were conducted in Stata version 16 [16].

Results
Out of 49,785 eligible patients admitted to the hospital during the time period of interest and who were recorded as a person who smokes, 951 did not have a complete SCCP at their initial admission. This means that 98.1% of patients who identify as a person who smokes on admission to the hospital had an SCCP completed during 2018-2019. For consistency between both analyses, the 951 patients that did not have a complete SCCP at their initial admission were matched with 951 other randomly selected patients that did have a complete SCCP, resulting in a final sample of 1,902, exceeding the target of 207 to achieve sufficient power. Hence, 47,883 patients were excluded from analyses. Of the 1,902 included patients, 342 (18%) indicated an intention to quit smoking. Participants' mean age was 51.4 years (SD: 19.2), and 57% were male.

NRT Interest
NRT was offered to 1,397 of the 1,902 patients (73.4%) and was subsequently accepted by 332 patients (17.5%). Patients who had completed a SCCP were more likely to be offered NRT than those who had not (77% vs. 69.8%, p < 0.0001; Table 1). The use of NRT prior to admission and NRT prescription on discharge was comparable across both groups (p = 0.381 and p = 0.642, respectively). Similarly, acceptance of offered NRT (as indicated by prescribing data) was also comparable regardless of SCCP completion (p = 0.108).

Smoking Cessation
The records of 951 patients that did not complete a SCCP at initial admission were then further analysed to determine whether SCCP completion influenced smoking status at readmission, with 725 patients further excluded as they had no recorded readmission up until June  (Table 3).

Discussion/Conclusion
Our study found that completing the SCCP brief intervention tool during hospital admission appears to promote smoking cessation when examining subsequent hospital readmissions over a 2-year period. That is, patients who completed the SCCP tool were more likely to cease smoking at any point in the 2-year study period and more likely to be abstinent from smoking as per their final Status unchanged, n = 352, 68% (SCCP at any previous admission, n = 169) *Remaining patients were randomly matched with patients who had readmissions and a completed SCCP, increasing the sample size to n = 452 Not readmitted (excluded from analysis), n = 725 The reference category for smoking status at any previous admission is "Did not quit." The reference category for smoking status at final readmission is "Unchanged status." The reference category for completion of SCCP at any previous admission is "Not completed." smoking status at the end of the 2-year study period. This indicates that the SCCP is likely an effective tool for providing a brief smoking cessation intervention to hospital inpatients. While we are unaware of other recent studies investigating the effectiveness of the SCCP in promoting smoking cessation, our findings reflect those of other recent studies showing that similar brief interventions are effective at promoting smoking cessation in hospital inpatient settings in Brazil [17], Iran [18], and Canada [19,20].
Our findings also showed that SCCP completion was associated with increased likelihood of NRT being offered to patients (as indicated on the SCCP). This finding is unsurprising given that offering NRT is one of the SCCP's steps. However, actual prescribing and patient use, which is dependent on patients accepting the offered NRT, was comparable between those that did complete a SCCP and those that did not. This suggests that most patients were offered NRT regardless of SCCP completion, potentially due to the hospital's smoke-free policy. Future research would benefit from exploring why the SCCP is not completed and why NRT is not offered in some cases. However, the data suggest that patients were more likely to accept NRT if they already intended to quit. Alternatively, the NRT offered upon discharge (nicotine patches) may not be their preferred form of NRT or cessation support, which may influence their likelihood to accept. For example, previous research shows low acceptability of nicotine patches in some populations due to experiences of adverse side effects such as nightmares [21,22].
These data also showed a fifty-fold increase in the proportion of eligible inpatients who had a completed SCCP compared to previous studies conducted since the introduction of the SCCP in 2015 [14]. This demonstrates that clinical staff are now routinely administering the intervention as part of the patients' admission, indicating that clinical staff may have increasing competency and confidence in administering the tool and may be more mindful of providing a formal intervention to optimise smoking cessation support. A systematic review by Rigotti et al. [23] investigating the efficacy of interventions for smoking cessation in hospitalised patients found that high-intensity behavioural interventions initiated at admission that included at least 1 month of supportive contact after discharge best promoted smoking cessation. Although intensive hospital interventions were investigated instead of a brief intervention tool as in our study, Rigotti et al. [23] highlight the benefit of post-discharge contact to sustain smoking cessation support for the patient outside of a hospital setting. They also highlight that the addition of NRT prescribing can further increase likelihood of smoking cessation post-discharge. These findings are compatible with an earlier systematic review which found that completion of hospital-based smoking cessation interventions was successful in increasing quit rates, whether or not NRT was used or prescribed [24]. Several studies indicate that hospital-based smoking cessation interventions are effective in increasing cessation rates when postdischarge follow-up is implemented or NRT use is encouraged [25][26][27]. This may be applicable to brief interventions like the SCCP, such that its efficacy can be improved.

Limitations
While the longer timeframe of this study may have better captured the ongoing smoking status of readmitted inpatients, allowing their behavioural change over time to be observed from contemplating quitting to actioning and maintaining quitting [28], the cross-sectional nature of the study design limits our ability to make temporal associations and conclusions. Moreover, this study had small cohort-to-cohort matching due to the small number of patients who did not receive the SCCP. Further research, using larger sample sizes and longitudinal study designs, is desirable to clarify and better elucidate the impact of SCCP as an intervention on supporting long-term smoking cessation. Replication of this study in other hospitals, both in Australia and globally, would also be desirable to confirm the generalisability of our findings. The reduction in smoking behaviours attributed to the SCCP in this study may also be explained by other factors (such as policies, differences between hospitals and patients, or disease scares). Further research delving into these factors may also be beneficial. Despite these limitations, this study provides new information that can guide and inform hospital-level smoking cessation advocates in developing sustained smoking cessation support for inpatients post-discharge.

Concluding Statement
This study, involving a retrospective review of 2 years of hospital-collected patient data, demonstrated that administering the evidence-based SCCP intervention tool may promote smoking cessation rates among hospital patients when subsequent readmissions are examined, and seemingly has negligible effect on patients' interest in NRT use during or after admission. Further research is required to assess the effect of the SCCP tool on longterm smoking cessation support. Our results have provided further evidence that hospital-based interventions can have a positive impact on sustained smoking cessation in patients who smoked at the time of admission and should therefore be used as part of routine care.