Factors influencing the implementation of guidelines
The most frequently investigated guideline characteristic concerns "complexity". Several of the systematic reviews included in our meta-review indicated that when a guideline can be relatively easily understood and tried out, the chance is greater that the guideline will be used.
It is important therefore for guideline developers to take into account the complexity of the guidelines. Particularly for developers of multi-disciplinary guidelines directed at several target groups with varying educational levels and backgrounds (e.g. physicians, nurses, patients), it is a challenge to describe recommendations that are understandable and usable for all target groups.
The finding in the Sachs review [
51], that involving the targeted professionals already in the development phase enhances the chance of successful implementation, may be relevant for guideline developers as well. Prominent groups like the WHO Advisory Committee on Health Research [
27] and the AGREE Collaboration [
59]also recommend that groups that develop guidelines should be broadly composed and include all relevant health professionals. In addition, involvement of the target group may imply that the guideline is first being tested in practice before large-scale implementation takes place [
59].
Still, Davies et al. [
58] assert that findings are not always unanimous with regard to whether guidelines that are developed by end users (amongst others) are more often used. Future research will have to provide more insight into this issue.
Our meta-review also describes specific influential characteristics of professionals. Implementers of guidelines and policy makers in particular should take into account that implementation may be hampered by the simple fact that professionals are often unaware that the guidelines exist, or are not familiar with their content [
49]. Clearly, it is not sufficient to merely disseminate a guideline. Targeted implementation interventions – in which professionals themselves are preferably directly and actively involved – should take place to create awareness. Examples of such targeted interventions may be combinations of (web-based, written or face-to-face) practical recommendations, educational material, and educational meetings (see for instance the Sachs article [
51]).
Characteristics of patients, too, appear to exert influence: for instance, co-morbidity in a patient appears to reduce the chance that guidelines are followed [
57]. Professionals, presumably, assume that guidelines are based on a general clinical picture and are insufficiently tailored to the often complex care needs of patients with co-morbidity. For instance, Tinetti et al. [
60] and Durso [
61] therefore argue for greater attention among guideline developers to the specific needs of patients with co-morbidity. To improve guideline implementation, these authors recommend that guidelines should also provide guidance for interventions in patients with multiple conditions as well as information on risks of specific interventions in these patients.
In addition, environmental characteristics influence the implementation of guidelines. For example, support by peers or superiors in following the guidelines, and sufficient staff and time appear to be important for guideline implementation [
49].
However, with respect to environmental characteristics, and also regarding patients' and professionals' characteristics, existing systematic reviews lack methodological rigour, and underlying primary research often focuses on rather heterogeneous guideline subjects and target groups. This hampers evidence-based conclusions. Future sound methodological research regarding these kinds of characteristics is therefore recommended.
More research is already being performed regarding the category "characteristics of the implementation strategies". Almost all included reviews investigated the influence of certain implementation strategies (see Additional file
1), which provides relevant information, particularly for guideline implementers. Most of the reviews indicate that effective implementation strategies often have multiple components and that the use of one single strategy, such as reminders only or one educational intervention, is less effective than a combination of strategies. However, interpretation is hampered by the fact that the high-quality Grimshaw review published in 2004 and 2006 [
41,
42] does not show a correlation between the number of components in implementation strategies and their effectiveness.
Explanations for contradictory results
The contradictory results found for "characteristics of the implementation strategies" may be partially explained by limited overlap in target groups and guideline subjects. The Grimshaw review included a large number of primary studies (n = 235) on guidelines with varying target groups and topics, while other relevant reviews involved fewer studies and had a narrower focus, e.g. on guidelines for mental health care [
14], guidelines for pressure ulcers [
55] or guidelines for nursing practice [
51].
Another (partial) explanation may be that Grimshaw et al. [
41,
42] were more rigorous in their analysis of primary studies than the other reviewers. Although Grimshaw and colleagues did not conduct a formal meta-analysis (because of the large heterogeneity of studies), they did take effect sizes into account.
However, a methodologically strong review by Wensing et al. [
23] also contradicts the conclusion of Grimshaw et al. [
41,
42]. Wensing and colleagues reviewed studies on a rather broad range of professional, educational and quality interventions (which was why the Wensing review did not completely match our inclusion criteria, and was not discussed in previous sections). Wensing et al. concluded that combined implementation strategies with many different aspects are, in general, the most effective. Furthermore, findings of two other methodologically sound meta-reviews [
20,
62] are relevant in this regard. These previous meta-reviews were likewise excluded from our meta-review since they focussed on a broad range of professional or educational interventions, and not specifically on clinical guidelines. Yet these meta-reviews equally concluded that successful implementation strategies are often multi-faceted.