The picture that emerged from this pre-program evaluation was encouraging, at least from the perspective of those nurses who took the opportunity to share their views. Both rural and urban nurses held relatively positive attitudes toward advising patients to quit, and despite a general lack of intervention training and little agreement that brief advice to quit is effective (as has been found in other studies ), the majority spent at least some time intervening. The most frequently performed activities tended to take the least amount of time, while the more complex activities (e.g., teaching coping skills and pharmacotherapy education) were seldom performed. Patient-related factors (quitting benefits and motivation) encouraged nurses to intervene and workloads and lack of time discouraged them, the latter of which is consistent with the literature . Unexpected were the significant differences between rural and urban nurses. Rural nurses were more likely to perceive intervening as part of their role, had higher confidence for intervening, more frequently assisted patients with quitting and arranged follow-up, reported having more systems in place to support cessation, and were less likely to report workload as being a barrier to intervening.
Although the response rate was relatively low, interpretation of the findings was enhanced by using previously published survey items with good internal reliability and comparing the outcomes to the study on which the current study was based and which had a high response rate --the outcomes of the 5A protocol steps, especially for rural nurses, were remarkably similar between the two studies. Compared to other studies, the levels of "ever intervening" with tobacco in the current study were substantially higher but "frequently intervening" was lower [11, 12]. However, it is difficult to compare with other studies due to inconsistencies of measurement scaling and reporting across studies . Some researchers treat interval scales as categorical and report proportions by scale option for each survey item (e.g., % reporting frequent, occasional, seldom, and never) without presenting an overall percentage for each of the global 5A steps , while others report the proportion who have ever done a given activity, without reference to frequency . Some studies use "unbalanced" rating scales (i.e., an unequal number of favourable responses at one end of the scale, such as "almost always" and "frequently"), which tends to result in more positive responses ; these studies then dichotomize the reporting of "frequently" intervening using the top two positive responses [11, 12], whereas "frequently" in the current study represented only the top anchor choice. These methodological inconsistencies suggest the need for more research in this area to develop standardized measures and methods of reporting.
The rural-urban differences were important findings, and given there were no differences in cessation training, might be reasonably attributed, at least in part, to the uniqueness of the rural context . Positive role perception for tobacco interventions, for example, might derive from the fact that rural nurses often have an expanded role of practice  and from our experience, they take on a number of different roles within a single nursing position such as staff educator, manager of chronic diseases, and telehealth coordinator, and within any one of their roles, there might be a clear indication for tobacco cessation treatment. Increased confidence might be due to the need for a wide range of knowledge and skills in rural nursing, which is influenced by the community demographics and the need to work with different populations in different areas of the hospital , often all in the same day, which in turn, represents a level of cross-training not common in larger centres. The higher rural smoking rates could have influenced intervention frequency as nurses are faced with various consequences of smoking across the units they work on, and they tend to be responsive to their community's needs because, as other studies have shown, "we see those people in our churches and in our grocery stores...in the end, we are the ones who see these people outside of our work life too." (pg 22) . In the USA, nurses in states with higher smoking rates vs. lower rate also intervene more often .
In relation to the systems differences between urban and rural nurses, rural nurses might be more aware of systems such as patient materials due to the physical environment layout. Nursing stations in the participating rural hospitals are commonly shared among many units and serve as central storage areas for staff and patient education materials. Most of the rural hospitals also have patient self-help materials on wall-racks located in the main foyer, which serves as the reception/waiting area, and through which staff pass on their way to/from work. Relative to documenting tobacco use/counselling, the rural hospitals used paper charting so every form is physically in the nurses' hands and easy to see. In contrast, each form in the urban hospital's electronic charting system has a separate tab that needs to be pulled down, and the cessation counselling tab in the urban hospital was optional, relatively new, and not well-promoted.
Overall, the findings provide insight into what is likely realistic in the context of staff nurses' daily practice. Although training would be expected to increase the frequency of intervening , the activities (those that could be done in a short period of time) and time (< 10 minutes) that the nurses reported in relation to intervening, and the most frequently noted barriers to intervening (time and workload) indicate that what nurses are willing and able to do maps onto the brief versus intensive interventions recommended by the Guidelines . In turn, brief interventions seem reasonable given the context of today's faster paced healthcare environment in which patient contact hours have been reduced, units are downsizing, and patient acuity and turnover have increased . Even brief interventions, however, are not necessarily easy to maintain--training must be ongoing and can be complex in larger hospitals; there is often staff resistance and a lack of accountability when the responsibility is diffused over a large number of staff; it is difficult to measure outcomes because to do so requires standardized charting and doing chart reviews; and brief interventions require some form of central organization and funds from the operating budget, even if only for training . The activities that the nurses reported doing less frequently are more in line with more intensive interventions that require more time, more training, and often a dedicated staff position.
Implications for practice
It was encouraging that the majority of nurses would likely be agreeable to providing cessation interventions as part of standard practice, and that they would not have to be convinced about the importance and desirability of intervening as much as it would be important to develop an intervention that would fit with the realities of their busy practice. Based on this needs assessment, we developed a brief intervention consistent with the Guidelines [4, 8] with the assist and arrange steps consisting primarily of distributing printed materials and avoiding the more complex types of counselling. For training, we developed a 30-minute in-service being mindful of nurses' heavy workloads and preferences for training, and provided information where nurses could engage in more in-depth training if interested . To highlight the patient benefits of quitting, which was the primary encouraging factor for intervening, we included smoking rates for each community in the training, as well as additional risks to hospitalized smokers and immediate benefits of quitting. We also included information from the Guidelines [4, 8] to increase nurses' knowledge of what was being recommended and to highlight the effectiveness of brief interventions. To enhance nurses' confidence to provide brief interventions, we demonstrated how to perform the 5A's in 1-3 minutes.
One limitation of this study is that the data were self-reported and not validated by medical charts. Other studies suggest a tendency to over-report (vs. under-report) desirable cessation-related activities . Another limitation is the relatively low response rate. Although the response rate was within the expected range  and the outcomes were similar to the study  on which the current study was based, the nurses who responded to the survey might not be representative of the target population and thus the outcomes are not necessarily generalizable to all acute-care nurses working in the region or in other areas of the province or country. Obtaining a high survey response rate, especially in the fast-paced environment of acute care, can be difficult to achieve and is one of the disadvantages of survey research .