Background
Chronic Obstructive Pulmonary Disease (COPD) is a chronic, life-threatening lung disease and the third leading cause of death worldwide. Active smoking is the main risk factor, while other factors such as occupational factors, infections and the role of air pollution are becoming more influential [
1]. In the Netherlands approximately 325,000 people are formally diagnosed with COPD and more than 6,000 people die of COPD annually [
2]. Clinical practice guidelines have been developed to assist healthcare practitioners in clinical decision making, by providing recommendations about appropriate healthcare for specific circumstances [
3]. In the Netherlands an evidence-based guideline for physical therapy diagnosis and treatment in patients with COPD was issued by the Royal Dutch Society for Physical Therapy (KNGF) and developed according to a rigorous procedure [
4]. An important component of the guideline is the recommended use of measurement instruments to establish parameters for treatment and evaluation [
5].
Publication of guidelines does not automatically lead to their uptake and change of physical therapists’ behavior based on guideline recommendations shows room for improvement [
6]. Several studies reported multiple barriers for adherence of physical therapists to clinical guidelines and measurement instruments [
7‐
9]. Multifaceted implementation strategies are more likely to result in change of professional behavior compared to educational activities [
10‐
12], and a comprehensive implementation strategy is essential in promoting the uptake of clinical guidelines. Such an implementation strategy should be tailored to specific barriers and facilitators [
13]. Therefore, assessment of specific barriers is important for implementation planning.
The focus of many interventions to enhance uptake is aimed at health care practitioners (knowledge, skills, attitude), although social factors (influence of patients, colleagues, stakeholders) and external factors (financial, organizational, regulatory influences, guideline characteristics) are also important to address when designing implementation strategies [
14].
Several theories and models exist to understand behavioral change of healthcare practitioners [
15‐
18]. These models suggest that behavior is determined by three important factors: attitude, defined as positive and negative beliefs associated with a particular behavior; social norms, defined as beliefs of reference persons about a particular behavior; and self-efficacy expectations, defined as abilities to perform a particular behavior. Such models have been used to develop frameworks for implementation in order to identify barriers to change at different healthcare levels [
14]. Each type of barrier may require a specific intervention, and the analysis of facilitators and barriers is a crucial step in a systematic approach for designing implementation interventions. Methods such as intervention mapping [
16] or the behavior change wheel approach [
19] can be used to design and select interventions.
A recently updated systematic review showed that interventions tailored to prospectively identified barriers are more likely to improve professional practice compared to no intervention or dissemination of guidelines only. Barriers are typically identified using interviews and questionnaires, while the most effective ways to identify barriers are unknown [
20]. There is a need for valid methods to identify specific barriers for the topic of interest, integrated in comprehensive approaches for designing implementation strategies.
In the Netherlands a questionnaire was developed and tested among general practitioners to identify barriers for the implementation of clinical practice guidelines [
21]. The questionnaire allows for tailoring to specific context and target groups, and has been used in follow-up studies to conduct surveys among Dutch general practitioners and physical therapists [
22,
23]. To identify barriers towards the use of measurement instruments, a general questionnaire was developed in assessing the attitude of physical therapists by asking their beliefs about the use of measurement instruments and reasons for (not) using measurement instruments [
24]. We were interested whether the combined use of these two available questionnaires may be useful to identify barriers to implementing the Dutch physical therapy COPD guideline.
The specific objective of our study was to construct a questionnaire for reliable and valid assessment of barriers and facilitators among Dutch physical therapists for implementing the COPD guideline and its recommended measurement instruments.
Discussion
This study showed that the construct of the developed questionnaire revealed relevant underlying domains for the identification of barriers and facilitators for implementing the COPD guideline and its recommended measurement instruments within the target group of physical therapists. A practical approach was useful in constructing an overall questionnaire based on previously developed questionnaires. The factor analysis allowed for exploring the content and meaning of the underlying domains, as well as the relevance of the questionnaire to develop implementation strategies.
Construct of the questionnaire
The characteristics of the identified factors in the 5-factor solution reflect the domains guideline (applicability of the COPD guideline), provider (knowledge & skills; attitude), context (required investment of time & money), and patient (patient characteristics) from the Peters questionnaire [
21]. However, the 5-factor solution resulted in only two variables loading on the fifth factor (patients), while it is recommended that at least three variables should be represented in each common factor [
28]. In the 4-factor solution the domains time & money and patients were combined with four variables, although this factor showed an insufficient alpha for internal consistency [
30]. The small sample size of our study does not allow for clear conclusions about the best factor solution, but we considered the 5-factor solution most relevant. Patient characteristics are considered important in addressing barriers to implementation [
13,
14]. More data are needed to determine the robustness of the identified factors.
Our analysis showed little overlap of items of the two parts of the questionnaires across the identified domains, supporting the presentation of the questionnaire in two parts. However, we advocate an integrated approach in the further use of the questionnaire. Measurement instruments are core elements in guideline recommendations, and specific barriers towards the use of these instruments are important to address.
Relevance of the questionnaire
The identified underlying domains in the questionnaire are considered important for implementation of clinical guidelines, and reflect existing frameworks for barriers to adhering to guideline recommendations [
10,
14]. However, the question remains whether the content of the questionnaire is actually valid for designing an implementation strategy for the COPD guideline. The main barriers for implementation were required time investment and financial constraints, which have also been reported as barriers in comparable studies in other professions [
31,
32]. These barriers may be considered generic and the questionnaire does not provide specific information for addressing these barriers for the specific purpose of implementing the COPD guideline.
The main facilitators show that participating physical therapists were very positive about using the COPD guideline and its recommended measurement instruments and reported to have enough knowledge and skills. Outcomes of studies that assessed barriers and facilitators for using guidelines and measurement instruments in the field of physical therapy vary, and negative [
8] as well as positive attitudes [
23,
24] have been reported. One study reported lack of knowledge as important barrier among general physical therapists [
24], as opposed to another study in which specialized physical therapists reported to have sufficient knowledge [
23]. The participation of physical therapists in specialized COPD networks in our study may have contributed to the positive attitude and focus on the COPD guideline to support clinical decision-making.
Despite the positive attitude of the participating physical therapists towards using the COPD guideline, we also noted barriers to working with the guideline especially related to external factors (resources, collaboration with general practitioners). Several factors may contribute to these barriers. One aspect is that organizational and external prerequisites may be lacking, thus limiting the feasibility for adhering to certain guideline recommendations. But these barriers may also be related to lack of specific competencies and routine of physical therapists. This requires more in-depth analysis based on the results of the questionnaire and can consequently be addressed in implementation planning. The complexity of COPD care requires a multidisciplinary approach, for which quality improvement collaboratives may be of added value to address shared barriers with other disciplines [
33].
Although the results of the questionnaire provide a good start for further implementing the guideline, more specific information is required for designing an implementation strategy. The lack of specific and targeted information from generic questionnaires has also been reported in other professions. A commonly used questionnaire in nursing to assess barriers to utilization of research is the BARRIERS scale. Although the scale is reliable in its use among nurses, the validity of the scale has been questioned due to its nonspecific nature and lack of evidence in being useful for designing implementation interventions [
34]. Lugtenberg [
22] conducted a survey among Dutch general practitioners to identify barriers for adhering to four different guidelines. The survey was derived from the results of a qualitative focus group study and specifically tailored to key recommendations in guidelines, rather than on barriers to guidelines as a whole. Their study resulted in a large variety of perceived barriers across recommendations, revealing more detailed information on potential interventions needed to improve guideline adherence [
22,
35].
Methodological considerations
While the full development of new measurement scales requires a comprehensive approach [
25,
26], we were able to use a practical approach in constructing the questionnaire and consequently assess the underlying factor structure and relevance of the questionnaire. The main rationale was the availability of the comprehensively constructed questionnaire of Peters among several groups of health providers in the Netherlands [
21], and the availability of a questionnaire that was directly applicable to the target group of physical therapists [
24].
In conducting the exploratory factor analysis, several decisions were related to choosing the appropriate method for extracting factors. We chose the method of principal axis factoring with oblique rotation. Principal axis factoring allowed for comparing results with parallel analysis for the number of actors to retain, and oblique rotation was used based on the assumption that barriers for implementation in the different domains were correlated. To estimate consistency of the results, we repeated the analyses using the alternative methods of maximum likelihood and orthogonal rotation [
36]. The results of these alternative analyses showed similar factor solution with some small differences in the loading of items on the different factors. The outcomes of the alternative analyses are available via the corresponding author.
Main limitation of this study for interpreting the results of the factor analysis is the relatively low sample size of 139 participants, which may have compromised the robustness of the analysis. It is recommended to use a subject to item ratio of at least 5:1 while a 10:1 ratio is considered as rule of thumb for determining a priori sample size [
36,
37], even though a survey among 303 published studies showed that more than 40% used a ratio of <5:1 [
37]. The ratio in our study was 3:1 and as a result we need to be cautious when interpreting the outcomes, although alternative analysis showed stable results. Additional analysis of the questionnaire with a larger sample would benefit confirming the underlying factor structure. The response rate of 57% contributed to the small sample size. We did not conduct a non-response analysis and therefore have no specific information about the non-responders. Several factors may have contributed to non-responses. We used only one method for administration of the survey and people may not have felt comfortable with the online form management system. Physical therapists with a positive attitude towards guidelines may have been more inclined to fill out the survey, which could have biased our findings, in overestimating the positive attitude towards the COPD guideline [
25,
26].
Implications for further strategy
The importance of tailoring interventions to prospectively identified barriers provides a clear rationale for the further development of methods to identify barriers [
20]. However, the actual contribution of barrier analysis for developing implementation strategies is still considered a black box [
38]. Quantitative assessment of barriers via surveys and qualitative methods via in-depth interviews are often combined and allow for in-depth analysis. Specific tailoring of surveys to key recommendations in guidelines may result in better insight in specific barriers at different levels and reveal more detailed information to develop implementation interventions [
22,
35]. These barriers should then be integrated in methods for designing interventions such as intervention mapping [
16] or the behavior change wheel approach [
19].
The construct of the developed questionnaire revealed relevant underlying domains for the identification of barriers and facilitators for implementing the COPD guideline and its recommended measurement instruments within the target group of physical therapists. The domains need to be confirmed in a follow-up study with a large sample size. While the positive attitude of physical therapists towards using the COPD clinical guideline is promising for further implementation, the results of the questionnaire alone do not provide sufficient information to inform the development of an implementation strategy.
Focus on the further development of methods for prospective identification of barriers and consequent tailoring of implementation interventions is required. The questionnaire allows for tailoring to the target group and may be used across health care professionals as basis for in-depth analysis of barriers to specific recommendations in guidelines. Efficiency and effectiveness can be enhanced by identifying specific barriers during the guideline development process. The infrastructure set up for developing the guideline can then be used for addressing key barriers by the guideline development group, using the questionnaire as well as in-depth analysis e.g. via focus group interviews. This would result in a (draft) strategy upon completion of the guideline allowing for swift progress after publication and dissemination. In addition, the identification of specific barriers during the development process will inform the guideline development group to modify the content of guidelines and to increase their usability [
39].
Acknowledgement
This study was conducted with financial support of the Royal Dutch Society for Physical Therapy (KNGF). The sponsor did not have a role in study design, the collection, analysis, and interpretation of data; in the writing of the manuscript; or in the decision to submit the manuscript for publication.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
PW carried out the planning, data analysis and writing of the manuscript. CZ contributed to the planning and data analysis of the manuscript. SD participated in the data analysis and writing of the manuscript. EH contributed to statistical analysis of the manuscript. MN participated in writing of the manuscript. RB contributed to planning and writing of the manuscript. All authors read and approved the final manuscript.