Identifying and recruiting smokers for preoperative smoking cessation—a systematic review of methods reported in published studies

Background Smoking cessation before surgery reduces postoperative complications, and the benefit is positively associated with the duration of being abstinent before a surgical procedure. A key issue in providing preoperative smoking cessation support is to identify people who smoke as early as possible before elective surgery. This review aims to summarise methods used to identify and recruit smokers awaiting elective surgery. Methods We searched MEDLINE, EMBASE, CINAHL, and PsycINFO, and references of relevant reviews (up to May 2014) to identify prospective studies that evaluated preoperative smoking cessation programmes. One reviewer extracted and a second reviewer checked data from the included studies. Data extracted from included studies were presented in tables and narratively described. Results We included 32 relevant studies, including 18 randomised controlled trials (RCTs) and 14 non-randomised studies (NRS). Smokers were recruited at preoperative clinics (n = 18), from surgery waiting lists (n = 6), or by general practitioners (n = 1), and the recruitment methods were not explicitly described in seven studies. Time points of preoperative recruitment of smokers was unclear in four studies, less than 4 weeks before surgery in 17 studies, and at least 4 weeks before surgery in only 11 studies. The recruitment rate tended to be lower in RCTs (median 58.2 %, range 9.1 to 90.9 %) than that in NRS (median 99.1 %, range 12.3 to 100 %) and lower in preoperative clinic-based RCTs (median 54.4 %, range 9.1 to 82.4 %) than that in waiting list-based RCTs (median 70.1 %, range 36.8 to 85.0 %). Smokers were recruited at least 4 weeks before surgery in four of the six waiting list-based studies and in only three of the 18 preoperative clinic-based studies. Conclusions Published studies often inadequately described the methods for recruiting smokers into preoperative smoking cessation programmes. Although smoking cessation at any time is beneficial, many programmes recruited smokers at times very close to scheduled surgery so that the benefit of preoperative smoking cessation may have not been fully effected. Optimal delivery of preoperative smoking cessation remains challenging, and further research is required to develop effective preoperative cessation programmes for smokers awaiting elective operations. Electronic supplementary material The online version of this article (doi:10.1186/s13643-015-0152-x) contains supplementary material, which is available to authorized users.


Background
Tobacco use remains a serious global public health problem [1]. It has been suggested that surgery may offer a powerful "teachable moment" for patients to quit smoking [2,3]. The risk of postoperative complications is much higher in smokers compared to nonsmokers [4]. Smoking cessation before surgery reduces postoperative complications, and the benefit is positively associated with the duration of being abstinent before a surgical procedure [5]. Interventions for preoperative smoking cessation may include brief advice, educational booklets, pharmacotherapy, and referral to telephone quitlines and professional smoking cessation clinics [6]. According to findings from a Cochrane systematic review, intensive smoking cessation interventions provided 4-8 weeks before surgery may be more effective than less intensive and later interventions in terms of postoperative complications and long-term smoking abstinence [7].
A key issue in providing preoperative smoking cessation support is to identify people who smoke as early as possible before elective surgery. A Cochrane systematic review of methods for recruiting smokers into cessation programmes did not include any study recruiting smokers specifically for preoperative smoking cessation [8]. Cost-effective strategies for recruiting smokers awaiting elective surgery are likely to be different from strategies for identifying and recruiting smokers in the general population. The National Institute for Health and Clinical Excellence (NICE) in the United Kingdom (UK) recommended that all smokers awaiting elective surgery should be identified, and that the smoking cessation interventions could be provided "at the point of referral in primary care, during secondary care consultations and/or at preoperative assessment" [9]. Methods for identifying and recruiting smokers for preoperative smoking cessation may be defined at least according to where (e.g. primary care or preoperative evaluation clinics), when (time before surgery), and by whom (e.g. general practitioners, nurses, or surgeons). However, it is still unclear which methods are most appropriate and feasible in practice to recruit smokers into preoperative smoking cessation programmes. This systematic review aims to summarise methods that could be used to identify and recruit smokers awaiting elective surgery.

Methods
This systematic review of methods for identifying and recruiting smokers into preoperative smoking cessation programmes was conducted according to an outline protocol (Additional file 1 and Additional file 5).

Inclusion and exclusion criteria
We included prospective studies that (1) compared different methods of recruiting smokers for preoperative smoking cessation programmes or (2) evaluated preoperative smoking cessation programmes.
We excluded the following studies: Studies in which no preoperative smoking cessation interventions were provided Studies with a retrospective design Studies not published in full or published in languages other than English

Literature search and study selection
We searched MEDLINE, EMBASE, CINAHL, and PsycINFO, and references of relevant reviews (see Additional file 1 for search strategies used) to identify relevant studies. Literature searches were conducted in May 2014, and no date limitation was applied. Two reviewers independently screened titles and abstracts of all references identified from searching bibliographic databases. Full texts were obtained and assessed according to the inclusion and exclusion criteria by two independent reviewers. Any disagreements were resolved through discussion or the involvement of a third reviewer if required.

Data extraction and evidence synthesis
One reviewer extracted and a second reviewer checked data extracted from the included studies. The following data were extracted from the included studies: the study design; the type of surgical procedure; country and setting (hospital or general practice); the methods for identifying and recruiting smokers (when, where, how, and by whom); the type of preoperative smoking cessation interventions provided; the outcome measures; the target quit duration before surgery; the number of recruited preoperative smokers; and the recruitment rate (defined as the proportion of recruited smokers in all eligible smokers identified). We contacted authors of some studies by emails to request additional information on missing or unclear data when available. Data and information extracted from included studies were presented in tables and narratively described.

Ethics and consent
This study is a literature-based systematic review, without directly involving human subjects. Therefore, the approval of an ethics committee is not required and informed consent for participation is not applicable.

Results
After screening 1612 references initially located by searching literature databases, we examined 101 full text articles and included 32 relevant studies ( Fig. 1 and Additional file 4). Table 1 shows the main characteristics and methods used in the included studies to identify and recruit smokers for preoperative smoking cessation. We obtained additional information from authors of ten studies (Table 1). More details on study characteristics of the included studies, including preoperative smoking cessation interventions, outcome measures used, and additional publications of the same studies, are provided in Additional file 2. A list of full text articles excluded and reasons for exclusion are available in Additional file 3.
The methods and timing of identifying and recruiting smokers before surgery are presented in Table 1. In general, recruitment information from the included studies was limited. Studies were generally focused on the effectiveness of the interventions on smoking abstinence or on reducing postoperative complications rather than on reporting specific information on methods for identifying and recruiting smokers.
Most studies (n = 18) identified and enrolled smokers at preoperative evaluation or pre-admission clinics, including anaesthetic [12,19,[23][24][25]36] and surgical clinics [10,15,18,[26][27][28][29][30][31][32][33]35]. Eligible smokers were usually identified and recruited by dedicated research personnel. Smokers were identified from elective surgery waiting lists (or medical records) in six studies [11,14,16,17,39,40]. There was only one study in which the identification of smokers for preoperative smoking cessation was started at the point of referral in primary care [38]. The methods for identification and recruitment were not explicitly described for the remaining seven studies. The number of smokers for preoperative smoking cessation ranged from 46 to 391 (median 149) in the included RCTs and ranged from 15 to 714 (median 59) in the included non-randomised studies. There were a total of 5137 patients included in the 32 studies (2990 in RCTs and 2147 in non-randomised studies).
According to 24 studies that provided sufficient data, the rate of recruitment of eligible smokers ranged from 9.1 to 90.9 % (median 58.2 %) in 15 RCTs and ranged from 12.3 to 100 % (median 99.1 %) in nine nonrandomised studies (Table 1). Therefore, RCTs tended to report relatively lower recruitment rate, compared with non-randomised studies. In addition, preoperative clinicbased approach tended to be associated with a lower recruitment rate than waiting list-based approach (Fig. 2  Smokers about to undergo surgery or diagnostic procedure were identified using electronic medical records system and received a personalised introductory letter about the study from the chief of preventive medicine (with an "opt-out" postcard). 1-2 weeks prior to procedure date, trained telephone interviewers contacted smoking patients who did not decline via postcard.   GP (199) were invited to identify daily smoking when referring them to surgery and to refer high-risk patients to a preoperative risk reduction programme. However, only 2 patients were referred a few months after starting the programme, and additional efforts slightly increased referral to the programme (7/72). High-risk patients for elective surgery were also identified at the department of orthopaedic surgery and surgical gastroenterology (at their first contact to hospital a ).

days
Recruitment rate was calculated using the number of smokers recruited as numerator and the number of invited eligible smokers as denominator. Target quit period before surgery was based on recommended quit date, not necessarily reflect the actual quit period preoperatively NA not available, SD standard deviation, POE preoperative evaluation, GP general practitioner, CVD cardiovascular diseases, IQR interquartile range, USA United States of America a Authors of the studies responded to request for additional information. We were able to contact authors of 14 studies by emails and received response from authors of ten studies with some additional information on missing or unclear data 70.1 % (range 36.8 to 85.0 %) in three waiting list-based RCTs. Because of wide ranges of the reported recruitment rates and great heterogeneity in study characteristics (such as diverse patients and interventions), the observed recruitment rates should be interpreted with great caution. For example, a recruitment rate of 9.1 % was reported in a study that recruited smokers at preoperative evaluation clinics [24], whereas a much higher recruitment rate of 68.5 % was reported by the same author in a more recent study that used the same recruitment approach [23].
Smokers were enrolled into smoking cessation programmes at various time points before surgery ( Table 1). The duration of preoperative recruitment of smokers was unclear in four studies, less than 4 weeks before surgery in 17 studies, and at least 4 weeks before surgery in 11 studies [10, 13-16, 20, 34, 35, 38-40].
Smokers were recruited at least 4 weeks before surgery in four of the six waiting list-based studies (67 %) and in only three of the 18 preoperative clinic-based studies (17 %) ( Table 1). In one study of patients awaiting coronary surgery, smokers were identified and recruited from elective surgery waiting lists many months (mean 8 months, standard deviation (SD) 2.7) before surgery [14]. In a recent study [40], smoking cessation materials (leaflets and quitline referral forms) were posted to all patients (smokers and non-smokers) at the time of waiting list placement (>4 weeks before surgery). Investigators of a study of elective forefoot surgery recruited participants approximately 6 months before surgery [39]. In another study that identified smokers from elective surgery waiting lists, smokers were enrolled into the smoking cessation programme 8 to 14 weeks before surgery [16].
The target quit dates before surgery generally corresponded to the time when smokers were recruited (Table 1). It was unclear in six studies. In four studies conducted by the same research team, smokers were identified at preoperative evaluation clinics a few days before surgery and asked to be abstinent from smoking on the day of surgery [19,[23][24][25]. The target quit date was at least 4 weeks before surgery in nine studies [13,15,16,20,34,35,[38][39][40].
Several of the included studies explicitly reported difficulties in identifying and recruiting smokers into preoperative smoking cessation programmes. Myles and colleagues identified smokers at the time of hospital admission for surgery but reported a low rate of recruitment and a high dropout rate [16]. Authors of two other studies reported failure to achieve the recruitment target, but the methods for recruiting smokers were not explicitly described [13,21]. Only one study evaluated the involvement of general practitioners (GPs) in recruiting smokers into preoperative programmes, in which GPs referred only seven of the 72 high-risk patients for a preoperative programme in a period of 9 months, after some intensive efforts to encourage such GP referrals [38].

Discussion
In this systematic review, we found that information on the recruitment of smokers was often inadequately reported in studies that evaluated preoperative smoking cessation programmes. Although smoking cessation at any time is beneficial to patients, the recruitment of smokers was often close (less than 4 weeks) to scheduled surgery and the benefit of preoperative smoking cessation for preventing postoperative complications may not have been fully effected.
Surgical treatment may be a "teachable moment" for patients to quit smoking [2]. Intensive behavioural support and pharmacological interventions are known to be effective strategies for smoking cessation [6]. However, many smokers, as well as surgeons, may not be fully aware of the risks of active smokers and the benefits of smoking cessation before surgery [3,42]. One of the key clinical issues in providing an effective smoking cessation programme for patients having elective surgery is to identify all patients who smoke and ensure that appropriate referral pathways for cessation support are in place [9]. This review has shown that published studies of preoperative smoking cessation programmes provide inadequate information on the methods used to identify and recruit smokers. Without this information, it is impossible for other researchers to replicate and verify the effectiveness of a preoperative smoking cessation programme. A wide range of different strategies have been used to recruit smokers into general (not preoperative specifically) cessation programmes, including mass media, telephone-or internet-based approaches, and face-to-face recruitment by health practitioners or others [8]. The recruitment of patients into preoperative smoking cessation programmes may require alternative strategies, although evidence of the effectiveness of different strategies is lacking in this area. The methods reported in the included studies were mainly (1) recruitment of smokers from preoperative outpatient clinics or (2) recruitment of smokers from waiting lists or medical notes (Table 1). However, difficulties in identifying and recruiting smokers into preoperative smoking cessation programmes have been explicitly reported in several of the included studies.
Available evidence indicated that waiting list-based approaches may be associated with a higher recruitment rate and earlier identification of smokers before surgery, compared with preoperative clinic-based approaches. However, this finding is tentative due to great heterogeneity across studies with similar approaches, and waiting list-based recruitment of smokers may not be feasible in some settings. Further research and empirical evidence are required to compare the different strategies to recruit smokers into preoperative smoking cessation programmes.
Many studies employed research personnel to identify and recruit smokers, to obtain consent, and collect data from study participants. Without the involvement of dedicated research staff, it may be uncertain whether preoperative smoking cessation interventions evaluated in studies could be implemented in normal practice. We found that the recruitment rates of eligible smokers for preoperative smoking cessation tended to be lower in the included RCTs than that in the non-randomised studies. It is assumed that preoperative smoking cessation interventions may be more acceptable to patients if such interventions are standard of care, so that the recruitment rate would be higher in clinical practice than that in research studies. However, this assumption needs to be confirmed with empirical evidence. In addition, further qualitative research may be required to understand reasons for difficulties in recruiting smokers into preoperative smoking cessation programmes, as reported in some included studies.

Limitations
Data on smoking cessation interventions used and outcome measures in the included studies were extracted (see Additional file 2), but these data were not further considered in this systematic review. To highlight lacking of relevant research and inadequate reporting in published studies, we focused on methods for recruiting smokers into preoperative smoking cessation programmes. Because we were not aware of a suitable quality assessment tool, quality of the included studies was not systematically assessed. Identification and recruitment of smokers awaiting elective surgery may be considered as an essential component in any preoperative smoking cessation interventions. However, reporting of methods for recruiting eligible patients has not been explicitly included in the TIDieR (template for interventions description and replication) checklist for better reporting of interventions [43]. Findings of this systematic review indicate that existing checklists for reporting of interventions need to be improved.
We did not identify any studies that directly compared different methods for recruiting smokers into preoperative smoking cessation programmes. The differences in recruitment rates and target quit dates before surgery between waiting list-based and preoperative clinic-based recruitment methods were emerged from indirect comparisons across different studies, which should be interpreted with great caution. In addition, the number of relevant studies was small, and there was considerable heterogeneity in the reported recruitment rates and study characteristics. Due to resource and time restrictions, we excluded studies that were not fully published and studies that were published in languages other than English. However, methods for recruiting preoperative smokers may be more likely inadequately reported in studies that were not fully published.

Conclusions
Published studies often inadequately described the methods for recruiting smokers into preoperative smoking cessation programmes and utilised study designs that may not be applicable to normal clinical practice. Although smoking cessation at any time is beneficial to patients, many programmes recruited smokers at times very close to scheduled procedure so the benefit of preoperative smoking cessation may not have been fully effected. Consequently, optimal delivery of preoperative smoking cessation remains challenging, and further research is urgently required to develop effective preoperative cessation programmes for smokers awaiting elective operations.