Worldwide, tobacco kills over eight million people yearly (1). Even though 80% of smokers is from low to middle income countries, 20.2% of Dutch adults smoked cigarettes in 2020 (2). The Dutch national institute of public health estimates that, yearly, 19.000 people in the Netherlands die due to the direct consequences of tobacco (3). This excess mortality is mainly caused by lung diseases, heart conditions and cancers.

Smoking cessation reduces the relative risk of morbidity and mortality (4, 5). Of the 10 years of life expectancy lost as a smoker, one regains 10, 9, and 6 years by stopping at 30, 40, or 50 years of age, respectively (4). Unfortunately, smoking cessation is not easy. Fewer than one in ten quit attempts from Dutch smokers is successful (6). However, studies show that a brief advice to quit smoking on medical grounds can increase long-term smoking abstinence by 47% (7). Because of this, the World Health Organization (WHO) calls on physicians and politicians to help fight the tobacco epidemic (8).

The WHO encourages physicians to assess smoking status in each patient, warn people about the dangers of tobacco and offer smoking cessation counseling (SCC) to help people to quit smoking (8). An easy and effective method of assessing smoking status, warning and offering SCC is the Ask-Advice-Connect (AAC) method. Using AAC, the physician asks about smoking, advises to quit and offers to connect the patient to an external smoking cessation program. The AAC method has a higher treatment enrolment rate than Ask-Advice-Refer (AAR) that requires patients to call the SCC provider on their own accord (9). With AAC, the contact details of the smoker are passed to the SCC provider who then contacts the smoker within 72 h (10). Despite the availability of the efficient AAC method, the majority of medical professionals still fail to connect patients to quit-smoking interventions (11). A questionnaire-based study among medical professionals of numerous departments in an academic hospital in India showed that 98% of professionals thought it was part of their role to help patients quit smoking (12). Furthermore, 95% of professionals indicated that they assessed smoking status, 94% expressed that they advised patients to quit when necessary but only 50% assisted and 28% arranged follow-up. A similar study found that cardiologists reported to Ask 73% of patients, Advise 62% of identified smokers and Refer 50% of identified smokers (13). Yet, when the patient is asked, only around one third confirms that they have been advised to quit (14). Clearly, there is a gap between practitioner reported behavior and patient reported advise rates. Thus, in order to increase smoking cessation rates we need a more thorough implementation of AAC and an objective evaluation of smoking assessment.

Known barriers to AAC implementation are the accountability of smokers, the responsibility of HPs, the role of HPs and the perceived lack of time of HPs (15, 16). Multiple methods exist to increase the efficacy of smoking cessation practices. Recent studies show that sending care providers two text messages a week can increase knowledge about the AAC model and support implementation of smoking cessation guidelines (17, 18). Whether text messages can improve implementation of the AAC method is yet unknown.

The main aim of this study was to evaluate the effect of a blended intervention aimed at HPs of the cardiology department in the Leiden University Medical Center to increase frequency of (i) smoking status assessment and (ii) connection to an external smoking cessation program. Furthermore, this study aimed to qualitatively explore the attitudes of HPs with regard to AAC and facilitators and barriers for inquiring about smoking status.

The purpose of this study was to evaluate the effect of a blended intervention aimed at HPs to increase the number of patients asked about smoking status, and the percentage of identified smokers connected with a smoking cessation program. The main conclusions of this study are that an implementation intervention for the AAC method in a university hospital did not increase the number of patients with a known smoking status or the number of identified smokers connected to a smoking cessation program. The results further support the hypothesis that this is most likely caused by priority issues and time pressure on the HPs involved.

This pattern of results is inconsistent with the previous literature, in which text messages have been reported to lead to a better implementation of smoking cessation strategies by HPs (17, 18). In our view, muliptle factors may explain the contrasting outcome of our study. First of all, focusing on the percentage of known smoking statuses, the lack of improvement can be partially explained by the ceiling effect. Whereas, past researchers have started at 21.2% of known smoking statuses, our study started at 74% (17). The significantly higher baseline in our study could make further improvement more challenging. A high baseline has been reported in other studies, where the percentages of patients asked about smoking were reported to be as high as 73–95% (12, 13). However, in these studies, HPs were asked how often they ask about smoking using surveys which potentially are subject to bias. This could imply that the 95% outcomes might not be attainable when the percentage of registered smoking statuses is objectively obtained from the EMR. Second, our setting is a University hospital with second opinions and complicated emergencies. It seems plausible that, with extremely rare and complex genetic diseases or in acute life-threatening situations, asking about smoking status is seen as a lesser priority. Third, with our intervention we tried to change behavior by providing knowledge about smoking cessation and the relevance of AAC to HPs. Studies show that changing people's behavior is not easily done by solely providing knowledge and relevance of a topic (19–21). For this, you also need more intensive guidance such as motivational interviewing, or environmental changes using choice architecture (22, 23). All in all, our intervention did not significantly increase the number of patients asked about smoking.

Similar explanation are relevant when focusing on the number of identified smokers connected to an external smoking cessation program. The present research showed that the implementation of a method of quick connection (one-click) accompanied by an instructive email in June 2018 led to connection rates of 9.3% of identified smokers. Which means that 90.7% of the identified smokers in the hospital is not connected to a smoking cessation program. Naturally, not every smoker is ready to quit at the time of presentation. However, taking into account that 68–80% of smokers wants to quit, there appears to be a lot of room for improvement (2, 24). On the other hand, past researchers reported similar rates (between 7.8 and 11.8%) when implementing AAC methodology in family or community clinics (9, 10). In previous literature, there are studies reporting higher rates of setting up follow-up for identified smokers (26–50%) but these were again in studies with self-reported outcome measures where HPs were asked to fill out surveys about SCC, potentially introducing self-reporting bias. In addition, these studies use the AAR method where one Refers (gives a referral note to the patient) instead of Connects (gives the patients' contact to an SCC provider) (12, 13, 25). AAR might be more easily conducted, but the percentage of connected patients enrolling into a smoking cessation program is significantly lower than with AAC (10).

The results of this qualitative assessment showed that almost all HPs who were interviewed (selected from the HPs with the lowest “ask” and “connect” rates across all professional roles) considered assessment of smoking status an essential part of their job description. These results are consistent with previous research reporting a high positive attitude toward smoking cessation strategies from HPs (26). This is an encouraging finding, as a positive HP attitude toward SCC is an essential starting point for every smoking cessation program. However, it is interesting that these were the HPs that asked about smoking the least and only half of the interviewed HPs reported having ever connected a patient to the external smoking cessation program. These findings indicate a huge gap between intention and practice of SCC. Notably, reporting bias due to social desirability may play a significant role in these interviews. More than half of HPs spontaneously mentioned that different ways should be implemented to obtain patients' smoking status and assist in smoking cessation, as their time with the patient during counseling is limited and does not allow for additional activities including SCC. Even if SCC consists of the brief AAC, HPs do not experience they have the time needed to ask about smoking. This conflicting attitude between HP's SCC ambition and implementation willingness has been reported previously, along with large differences between specialists and persisting unclarity about which HP is responsible for SCC (15).

Taken together, the results imply that the HPs agree that AAC should be integrated in hospital policies but that patients should be Asked and Connected by other means or professionals.

One limitation of this study is the pre-post study design which can lead to historical bias where events unrelated to the intervention can have an influence on the outcome. Another can be the fact that only the Ask and Connect phases of AAC were objectively measured in this study. Asking after smoking status can be immediately followed by Connecting an identified smoker to a smoking cessation program, forgoing the Advice phase. Especially since HPs can find it patronizing to Advice smokers to quit, it is plausible that the Advice phase is sometimes skipped (27). We do not expect this to influence the results of this study but it should be taken into account as Advice is an integral part of AAC. As for strengths, first, the large number of patient visits included provides sufficient power to compare the percentage of known smoking statuses between the three groups. Second, all 67 presently active HPs working at the cardiology department in the Leiden University Medical Center provided informed consent for this study. This indicates that smoking cessation is considered an important aspect of care by HP's. On the same line, 52 out of 67 (75%) HPs agreed to receive messages during our study, suggesting good reach of the messages. However, no information is available about whether the messages were read and engaged with. This makes it possible that busy HPs quickly disregarded the messages. Third, the endpoints used were objective and easily obtainable leading to a low amount of missing data. Last, the qualitative assessment focused on the HPs with the lowest implementation rates, providing essential information to improve SCC effectiveness. However, it should be acknowledged that, in the future, important information could also be identified through interview with HPs with higher rates of Asking and Connecting.

In short, this mixed-methods study on 48,321 patient-consultations evaluated the effect of a blended intervention aimed at HPs. Unexpectedly, our intervention did not increase the numbers of patients asked by HPs about smoking or the number of identified smokers connected to an external smoking cessation program, most likely due to inadequate time available to HPs during patient consultations for ACC in addition to their prevailing clinical tasks. However, we found that AAC is an accepted smoking cessation strategy by HPs in a university hospital setting. Therefore, it is recommended to facilitate in-hospital AAC through an optimized logistical process involving rapid one-click connection to evidence-based smoking cessation therapy and using procedures that minimally impact the time restricted consultation with the HP without squandering the effect of a brief advice of HPs.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Medisch-Ethische Toetsingscommissie Leiden Den Haag Delft. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

DF performed the interviews, data collection, and analyses. VJ, EM, and DA critically reviewed the manuscript in multiple stages. All authors participated in planning the study and read and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.