The four themes suggested in the theoretical framework did serve to cover all the experiences expressed by staff at the PHC units, which shows that the framework is suitable for evaluation of implementation in PHC.
At the adopting unit the explicit implementation strategy had been used, a strategy that has been shown to predict a better implementation outcome if combined with a creative climate at the unit [22
]. The Explicit NA group differed from the adopters regarding a number of factors, indicating that these factors may be more strongly linked to adoption than implementation strategy. One of these factors was expectations. Earlier studies on the implementation of clinical guidelines among GPs in PHC state that dissemination must target the perceived needs [23
]. Expectations seemed to be more important than staff involvement in the decision process, known to be fundamental in organization development [24
]. The intention of the explicit implementation strategy was to involve staff in the decision process, but choice of strategy seems not to have influenced the perception of involvement.
Other important factors were those concerning the innovation characteristics. It was obvious that Explicit NA found the concept incompatible with their way of providing health services, and they saw hardly any relative advantage in using it. The staff seemed to lack a sense of urgency to change, which is consistent with earlier findings in PHC [25
]. According to Rogers [4
], innovation characteristics explain 49-87% of the variance in rate of adoption of innovations. In a study on the implementation of information systems, Yetton et al
] found that the contribution of innovation characteristics to implementation success was higher than that of the implementation process. The innovation implemented in this study, the computer-based lifestyle test, was supposed to be perceived as easy to use, possible to try, observable and compatible with existing routines. Computerized solutions in PHC have been evaluated previously and found to be feasible [27
]. In this study, the staff in the different groups seem to have diverging perceptions of the lifestyle test characteristics, probably due to factors regarding context or adopter characteristics.
An opinion leader facilitating the implementation was mentioned only by the adopters. In most social systems there are key figures - respected and well-informed professionals who personify the group norms and group culture, and who filter new information and pass it forward [7
]. The presence of opinion leaders could have great importance in the implementation process [6
]. However, it has been shown that opinion leaders are valuable to support implementation particularly in highly specialized staff groups [28
]. The nature of PHC is to address a wide range of health care issues, which could be an explanation for the lack of opinion leaders in this particular setting.
The contextual factors studied were all about the actual inner context - the coinciding working situation. Since all units in the Implicit NA group had had a period of heavy work load, organizational change and staff shortage, it is most likely that this affected their ability to adopt the innovation [29
]. The two Explicit NA units both had mangers absent during the study period because of sick-leave or job vacancy, which may have had an impact on adoption because leadership is of great importance for organizational innovation [30
Staff characteristics investigated in this study included opinions about addressing lifestyle issues in primary health care and innovativeness on a group level. The overwhelming awareness about the importance of the issue was not reflected in adoption of the computer-based test. Perceiving it important might lead to an intention to address lifestyle issues, but not necessarily to doing it, since there is a well-documented intention-behavior gap [32
]. Positive opinions about change and innovations expressed by the Explicit A probably helped in forming a receptive context, one of the factors that determine innovativeness in health service organizations [16
The findings from this study might be helpful for policy-makers and for managers and staff in the local setting, aiming to introduce new methods into PHC. A factor not analysed in this study is the contributions of the different professional groups to the results. This will be analyzed further and presented in the future. Another important question that should be assessed in future research is the low overall rates of adoption, and what could be done to encourage the use of the computer-based lifestyle test.
Limitations and strengths
As in all qualitative research this study is limited regarding its generalizability and relevance to other settings. However, we believe that the knowledge gained from this study could be of importance in planning the implementation of new methods under similar circumstances. The factors assessed in this study could be categorized according to the predefined themes in the theoretical framework. The framework was based on an extensive literature review, and was also consistent with other implementation models described in the literature, which should be considered a strength. Theory, however, contains a number of additional factors that our study was not designed to assess, for example adopter characteristics on an individual level, outer context and networks. However, we believe that the factors assessed do influence implementation outcome in an important way. The fact that size differed substantially between units, and that some of the units experienced organizational changes coinciding with the implementation process might have influenced the results. These are factors that should be considered if the study is repeated.
The main difference between the two strategies was the testing period combined with a decision session provided in the explicit strategy. A more extensive implementation effort might have resulted in higher levels of adoption, but would also have required more financial input. A strength in this study was that a better implementation outcome was obtained despite limited financial resources.
Another strength in this study was that the interviews were conducted with the different staff groups separately, which allowed the individuals to reveal their thoughts without fearing the reactions of staff members in other categories. It is well known that PHC is a hierarchic organization, and a mixed group might have hindered an open discussion.