In this study, we were only able to show a small increase, from zero to ten percent, in the frequency of patients directly referred by GPs to the surgical departments and the hospital's preoperative smoking and alcohol cessation programme. Though a ten percent increase in some situations is of importance, this low referral rate would not be acceptable in case of surgery from a patient safety view. This is because ninety percent of the high-risk patients that could benefit from a timely risk reduction are overlooked.
It was surprising that overcoming the barriers known to be related to GPs' assessment and counselling for tobacco cessation [23
], did not facilitate the referral of high-risk patients in our study.
We discovered that in spite of a positive attitude among the GPs, as well as approval and support from their organisation, the referral rate was very low.
It might be speculated that this poor referral rate was due to patients refusing to be referred when offered this option by their GPs. However we discovered that this was not the case in the present study. We learnt in interviews with the GPs, that in daily practice, they forgot to play their role in risk reduction before surgery. In view of this the GPS asked for more information and reminders for both the patients and themselves. However, meeting these local wishes and suggestions did not result in an increased number of referred patients.
Bias and limitations
Some of the elements in the integrated preoperative programme that had been identified in interviews or from previous studies as possible barriers have not yet been subjects for specific intervention research. They may therefore, theoretically, be of minor significance, which would reduce the effect of overcoming them. Another theoretical possibility could be that some of the efforts to overcome the barriers might have a negative effect and thereby neutralise a positive effect from other elements in the integrated programme. Furthermore, some barriers may be of greater importance in different countries; e.g. the simple referral model and easy access to smoking cessation experts has been shown to be effective among GPs in England [30
], but not in the Netherlands [31
]. The factors given in table originated from Europe, North America and Australia; however, it is unclear whether the individual factors are specific to concrete contexts or health-care systems. The extra reimbursement for the GP reflected the "pay per service" or "quality-based reimbursement principle", which seemed promising in some studies, but not in others [38
]. One other explanation for the poor referral rate could be that it is unusual tradition to have such integrated collaboration regarding risk reduction by changing lifestyle for surgical patients.
The tight time schedule imposed by the requirements of the integrated preoperative programme prior to surgery, could pose a significant challenge for GP's. In other situations the "window of intervention" for smoking and hazardous drinking may either be open longer or there is an opportunity to repeat the intervention programmes, for example in patients with COPD, diabetes, liver diseases etc.
Furthermore, the legal responsibility for the surgical pathway may play a role. According to the law in most countries the surgeon is the person responsible for having informed the patient sufficiently before surgery regarding the benefits and harms of that surgery. This is necessary to ensure that the patient has the correct basis for giving their informed consent to the operation.
The clinical perspective
From a clinical point of view, integration of a preoperative smoking and alcohol cessation intervention in the surgical pathway is highly desirable. The effective preoperative lifestyle intervention programmes published hitherto either took place in the hospital out-patient clinics or in units in close relation to the surgical pathway and were given by experts [5
In the long-term it would be relevant to establish a better tradition for collaboration between the primary care providers and the surgical departments. In the short-term it is, however, important to evaluate other strategies for risk reduction to offer the programmes in due time before surgery. Otherwise the consequences would be tremendous for the patient and for the healthcare system in general. In Denmark the annual costs related to the increased complications rate and prolonged hospital stay among harmful drinkers without preoperative lifestyle intervention programmes are approximately 30-50 Euros per capita [39
]. The corresponding amount for smokers may be similar. The preoperative smoking intervention has already proven cost-effective in the immediate postoperative period [11
The implementation of this programme may also arise from a political level, which may decide that all high-risk patients on waiting lists for elective operations should be offered a lifestyle intervention programme prior to surgery [40
The patient perspective
From a patient perspective, patients, relatives and patient organisations need improved information regarding the high-risk of surgery and the effectiveness of risk reduction programmes prior to surgery. Though smokers and hazardous drinkers undergoing surgery may never develop into a strong and demanding patient organisation, they may complain about misinformation or lost opportunities to improve their surgical outcomes in the future.
The research perspective
This study gives rise to a hypothesis that overcoming identified barriers (table ) are insufficient when implementing new integrated guidelines. It might be necessary to add new elements that can stimulate GPs to remember to follow the guidelines in their daily practice.
In addition, new research could change the emphasis from projects investigating barriers to exploration of fully implemented and integrated procedures in primary and secondary care. This may help to generate new strategies for implementation among surgical drinking and smoking patients.