What do physical therapists think about evidence-based practice? A systematic review

Tatiane Mota da Silva, Lucíola da Cunha Menezes Costa , Alessandra Narciso Garcia, Leonardo Oliveira Pena Costa

Evidence-Based Practice (EBP) has been increasingly recognised and used by physiotherapists as a result of the growing volume and accessibility of high-quality research (Maher et al., 2004). Currently, the term Evidence-Based Physiotherapy can be applied to physiotherapy based on high-quality clinical research (Herbert et al., 2011). However decision-making must also take into account the wishes, expectations, and values of the patient as well as the therapist’s experience and knowledge (Herbert et al., 2005). Although the concepts of EBP are well-defined, EBP faces a variety of challenges with regards to its implementation (Haynes and Haines, 1998). These challenges are closely related to current health policies, the complexity of physiotherapy practice, access to the studies and continued education programs (Haynes and Haines, 1998). Previous studies on different health professions have identified multiple barriers such as lack of time (McColl et al., 1998; Metcalfe et al., 2001; Jette et al., 2003; Majid et al., 2011), lack of access to full-text articles (Maher et al., 2004; Bennett et al., 2007) and/or lack of skills in finding and understanding the studies (Jette et al., 2003; Bennett et al., 2007; Caldwell et al., 2007; Ahmadi et al., 2012). The inability of health professionals to comprehend and select high-quality studies has been attributed to poor training in EBP due to great variability of teaching methods during university training (Ahmadi et al., 2012). Other barriers include the questionable quality of the studies (Petrisor and Bhandari, 2006; Caldwell et al., 2007; Spallek et al., 2010) and the conflicting results of different studies investigating the same topic (Metcalfe et al., 2001; Spallek et al., 2010). Some studies have different characteristics which do not represent the real clinical practice which can make the clinical application more difficult (Flores et al., 2000; Cranney et al., 2001; Jette et al., 2003). Another factor that can interfere in the application of EBP is the language of publication. Most studies are published in English (Shiwa et al., 2013), which can hinder their use by non-English proficient readers (Maher et al., 2004). EBP has also been progressively used in different areas of health, such as medicine, nursing, physiotherapy, occupational therapy, nutrition, dentistry, and in the area of healthcare management and economics (Evidence-Based Medicine Working Group, 1992). There are several studies that investigated different aspects required in EBP in a specific population of physiotherapists (Jette et al., 2003; Iles and Davidson, 2006; Bauer et al., 2007; Grimmer-Somers et al., 2007; Salbach et al., 2007, 2009; Buchard, 2009; Nilsagård Ylva, 2010; Gorgon et al., 2013; Scholten-Peeters et al., 2013). Another review (Scurlock-Evans et al., 2014) has showed that physiotherapists tend to present favourable opinions toward EBP and the mainly barriers faced by them are usually related with lack of time and skills, and also misperceptions of EBP. Besides, this review identified some interventions that seem to be promising to a better implementation of EBP. Although this review has assessed some important characteristics about EBP in physiotherapy, the results were presented as a textual synthesis, without presenting the frequencies of the analysed characteristics, and included only studies published in English. Although there are studies about this topic, our systematic review can potentially contributes with more detailed information about this topic from retrieving all available evidence. Then, we considered important a new systematic review to better inform professionals about these characteristics and to identify the most important difficulties faced by physiotherapists with regards to EBP. Therefore we aimed to systematically review the evidence on EBP knowledge, skills, behaviour, opinions, and barriers faced by physiotherapists.

2. Methods

2.1. Search strategy

Systematic searches were conducted on the following electronic databases MEDLINE, EMBASE, CINAHL, PSYCINFO, LILACS, and SciELO including publications since the inception of these databases until 05th September 2014. These searches were adjusted to each of the databases used. Detailed search strategies used in each database are described in Appendix.

2.2. Inclusion and exclusion criteria

Studies were considered for inclusion if they met the following criteria: (1) quantitative cross-sectional studies, (2) studies that investigated EBP characteristics by physiotherapy graduates (regardless of degree) who are working in the field and (3) studies that investigated EBP knowledge, skills, behaviour, opinions, and barriers in physiotherapy.

2.3. Study selection

The study selection process included: (1) analysis and selection by screening the titles; (2) analysis and selection by reading the abstracts, and (3) analysis and selection by reading the full texts. Potentially eligible studies were also searched by reading the references lists of eligible article. If an eligible study was published in a language different than English, Portuguese and Spanish, all possible attempts were taken to translate it. The data were extracted by two independent reviewers and, in case of disagreement; consensus was reached by discussion between the reviewers or by arbitration by a third reviewer.

2.4. Data extraction

A data-extraction form based on similar studies (Dijk et al., 2010; Ubbink et al., 2013) was designed for recording information on: (1) year of publication, (2) country of origin, (3) source of sample, (4) number of participants, (5) the aspects of EBP that were analysed, including knowledge, skills, behaviour, opinions, and barriers. The data were extracted by two independent reviewers and, in case of disagreement; consensus was reached by discussion between the reviewers or by arbitration by a third reviewer. The authors were contacted by email in order to obtain any additional information that might not be reported in the original articles.

2.5. Risk of bias

Risk of bias was assessed using the criteria developed by Ferreira (Ferreira et al., 2010) and Leboeuf-Yde and Lauritsen (1995). These criteria described the representativeness of the sample (measured by 2 items), the response rate, the accuracy of the data, evidence of power calculation and the instrument used. Some of these criteria have been adapted for our study and are described in Appendix 2. Each study received a score as risk of their bias, expressing the number of criteria met on a 6-point scale, higher scores being representing low risk of bias (Leboeuf-Yde and Lauritsen, 1995; Ferreira et al., 2010). The risk of bias was rated by two independent reviewers and, in case of disagreement; consensus was reached by discussion or by arbitration by a third reviewer.

2.6. Data analysis

We considered as outcomes: EBP knowledge, skills, behaviour, opinions, and barriers in physiotherapy. We defined the outcomes of our review as follows: Knowledge: “facts, information, and skills acquired through experience or education; the theoretical or practical understanding of a subject”; (University, 2013) Skill: “the ability to do something well; expertise, a particular ability”; (University, 2013) Behaviour: “the way in which one acts or conducts oneself, especially towards others, behaviour patterns”; (University, 2013) Opinion: “a view or judgement formed about something, not necessarily based on fact or knowledge, the beliefs or views of a group or majority of people”; (University, 2013) Barrier: “a circumstance or obstacle that keeps people or things apart or prevents communication or progress” (University, 2013).

It was not possible to perform a meta-analysis of the studies included in the review due to the large heterogeneity among the studies. This heterogeneity is largely related to different data collection methods, different measurement instruments used, and different response options used on each instrument. Therefore, our results were reported descriptively. Although our data precluded the use of a meta-analysis approach, we aggregated the response options by presenting the range of responses (minimum and maximum) for each characteristic analysed.

3. Results

3.1. Study inclusion

The search yielded 12,392 potentially eligible studies, of which 12,313 were excluded after the titles and abstracts were read. Moreover, a total of 20 studies were excluded after we have read the full-text. One abstract was considered potentially eligible but was not included in the analysis because a full text could not be obtained (Hamzat and Amusat, 2002). No additional studies were found in the references of the selected studies. The entire process of eligibility assessment is described in Fig. 1.

3.2. Characteristics of the studies

Twelve studies were included in the analysis (total sample ¼ 6411 participants; mean ¼ 534 participants per study). The studies were conducted in 9 different countries from 5 different continents. The participants were recruited from different sources, such as registration boards (Grimmer-Somers et al., 2007; Salbach et al., 2007, 2009; Nilsagård Ylva, 2010), associations (Jette et al., 2003; Buchard, 2009), hospitals (Iles and Davidson, 2006; Gorgon et al., 2013; Queiroz and Santos, 2013; Bernhardsson et al., 2014) or clinics (Queiroz and Santos, 2013; Bernhardsson et al., 2014) and university (Scholten-Peeters et al., 2013). Table 1 summarises the characteristics of the studies. The results of two articles (Salbach et al., 2007, 2009) were reported together as they came from the same study. We contact eight authors by email to ask additional information during the extraction process (Jette et al., 2003; Iles and Davidson, 2006; Bauer et al., 2007; Buchard, 2009; Salbach et al., 2009; Nilsagård Ylva, 2010; Queiroz and Santos, 2013; Bernhardsson et al., 2014), and got information from 5 of them (Grimmer-Somers et al., 2007; Salbach et al., 2007; Nilsagård Ylva, 2010; Gorgon et al., 2013; Bernhardsson et al., 2014). The aspects of EBP described by each study are found in the Appendix 3. A total of six studies (Jette et al., 2003; Iles and Davidson, 2006; Salbach et al., 2007, 2009; Nilsagård Ylva, 2010; Gorgon et al., 2013) were based on the questionnaire developed by McColl (McColl et al., 1998), which measures the EBP knowledge, behaviour, opinions, and barriers of general practitioners. Data collection was conducted in different ways, including: Emails (Jette et al., 2003; Salbach et al., 2007, 2009; Nilsagård Ylva, 2010; Gorgon et al., 2013; Queiroz and Santos, 2013; Scholten-Peeters et al., 2013; Bernhardsson et al., 2014); Telephone calls (Iles and Davidson, 2006; Gorgon et al., 2013); Letters delivered in person (Iles and Davidson, 2006; Gorgon et al., 2013); Letters delivered by post services (Iles and Davidson, 2006; Grimmer-Somers et al., 2007; Buchard, 2009; Gorgon et al., 2013); and internet (Grimmer-Somers et al., 2007). The studies obtained responses from 39 (Bauer et al., 2007) to 833 (Nilsagård Ylva, 2010) participants. The response rates ranged from 20% (Scholten-Peeters et al., 2013) to 81% (Salbach et al., 2007, 2009; Gorgon et al., 2013).

3.3. Risk of bias of included studies

The total score ranged from 2 to 5 points (on a scale ranging from 0 to 6, median score ¼ 3 points; mean score ¼ 3.4 points; Standard deviation (SD) ¼ 0.64), reflecting moderate risk of bias. The items most frequently satisfied were the description of the sample, response rate, and data primarily collected to describe knowledge, skills, behaviour, opinions and barriers in EBP. The items less frequently satisfied were about the non-responders, failure to validate the questionnaire used and description of power calculation. The results of the risk of bias are shown in Table 2.

3.4. EBP knowledge

Ten studies (Jette et al., 2003; Iles and Davidson, 2006; Bauer et al., 2007; Salbach et al., 2007; Buchard, 2009; Nilsagård Ylva, 2010; Gorgon et al., 2013; Queiroz and Santos, 2013; ScholtenPeeters et al., 2013; Bernhardsson et al., 2014) reported information on EBP knowledge. In three studies between 21% (Nilsagård Ylva, 2010) and 82% (Jette et al., 2003) of the respondents reported having previous knowledge on EBP. In 2 studies, 47% (Salbach et al., 2007) and 70% (Jette et al., 2003) of the sample reported having learned how to perform a search in databases, and 59% (Iles and Davidson, 2006) and 70% (Nilsagård Ylva, 2010) reported they knew how to formulate a clinical question. In three studies, between 27.3% (Buchard, 2009) and 84% (Gorgon et al., 2013) reported they performed searches in databases, and 48% (Gorgon et al., 2013), 67% (Jette et al., 2003) and 70% (Nilsagård Ylva, 2010) of the sample reported having critical appraisal skills. Table 3 illustrates the data most frequently reported with regards to aspects of EBP, and Appendix 2 presents all of the aspects included in each study.

3.5. EBP skills and behaviour

Ten studies analysed skills or behaviour (Jette et al., 2003; Iles and Davidson, 2006; Bauer et al., 2007; Buchard, 2009; Salbach et al., 2009; Nilsagård Ylva, 2010; Gorgon et al., 2013; Queiroz and Santos, 2013; Scholten-Peeters et al., 2013; Bernhardsson et al., 2014). In 4 studies, 8% (Nilsagård Ylva, 2010), 16% (Gorgon et al., 2013), and 32.8% (Salbach et al., 2009) of the participants reported that they use databases sometimes in a typical week and 20% (Nilsagård Ylva, 2010), 62.3% (Salbach et al., 2009), and 65% (Jette et al., 2003) use monthly. In two studies, nearly half of the sample had used Cochrane, CINAHL or PEDro e 47% (Gorgon et al., 2013) and 10.6% had used PEDro, 15.3% Cochrane and 26.6% CINAHL (Iles and Davidson, 2006). In 4 studies, 15% (Nilsagård Ylva, 2010), 56% (Salbach et al., 2009), and 66% (Jette et al., 2003) of the respondents reported reading scientific articles sometimes in a typical week and 10% (Gorgon et al., 2013), 17% (Jette et al., 2003), 26.7% (Salbach et al., 2009), and 40% (Nilsagård Ylva, 2010) monthly. Two studies reported that 25.8% (Iles and Davidson, 2006) and 86% (Nilsagård Ylva, 2010) use guidelines in clinical practice. In other four studies, 23.9% (Queiroz and Santos, 2013), 48% (Nilsagård Ylva, 2010), 62% (Bernhardsson et al., 2014) and 67% (Jette et al., 2003) claimed to have support from work to use EBP.

3.6. EBP opinions

Seven studies reported information on EBP opinions (Jette et al., 2003; Bauer et al., 2007; Grimmer-Somers et al., 2007; Salbach et al., 2007; Buchard, 2009; Queiroz and Santos, 2013; Bernhardsson et al., 2014). In 6 studies, 48% (Queiroz and Santos, 2013), 69.2% (Bauer et al., 2007; Queiroz and Santos, 2013; Bernhardsson et al., 2014), 80% (Buchard, 2009), 90% (Jette et al., 2003; Bernhardsson et al., 2014), and 93.2% (Salbach et al., 2007) of the sample considered EBP necessary or important and 78.1% (Salbach et al., 2007), 82% (Jette et al., 2003), and 90.3% (Grimmer-Somers et al., 2007) believed that scientific literature is important to clinical practice. Additionally, 51.4% (Buchard, 2009), 79% (Jette et al., 2003), and 84% (Salbach et al., 2007) believed that EBP improves the quality of care, and between 66% (Queiroz and Santos, 2013) and 83% (Bernhardsson et al., 2014) of the sample from four studies believed that scientific evidence helps in making clinical decisions. In 3 studies, 70% (Buchard, 2009), 85% (Jette et al., 2003), and 90.2% (Salbach et al., 2007) of the respondents reported interest in learning more about EBP.

3.7. EBP barriers

The most frequently reported barriers were: lack of time varying from 31.2% (Grimmer-Somers et al., 2007) to 93.8% (Buchard, 2009) of respondents, inability to understand statistical data 30.4% (Salbach et al., 2007) to 54.3% (Grimmer-Somers et al., 2007), lack of support from employer e 6.7% (Grimmer-Somers et al., 2007) to 56% (Gorgon et al., 2013), lack of resources (computer and internet access) e 15.6% (Salbach et al., 2007) to 53% (Gorgon et al., 2013), colleagues not favourable to EBP implementation e 3.7% (Salbach et al., 2007) to 42% (Nilsagård Ylva, 2010), lack of interest e 3.3% (Salbach et al., 2007) to 36% (Nilsagård Ylva, 2010) and lack of generalisation of results e 20% (Grimmer-Somers et al., 2007) to 33.7% (Salbach et al., 2007).

4. Discussion

Evidence-Based Physiotherapy has been increasingly discussed and disseminated over the last few years. This review shows that only 21% (Nilsagård Ylva, 2010) and 82%, (Jette et al., 2003) of professionals had received a formal training on EBP. The data related to skills and behaviour show a rate of use sometimes in a typical week of databases ranging between 8% (Nilsagård Ylva, 2010) and 32.8% (Salbach et al., 2009) and a rate of monthly use between 20% (Nilsagård Ylva, 2010) and 65% (Jette et al., 2003). Nearly half of the professionals had used databases to aid in clinical decision-making (Iles and Davidson, 2006; Gorgon et al., 2013). The data related to opinions about EBP shows that part of the professionals seems to be favourable with the implementation of EBP, considering that the majority of them believe that scientific evidence is an important component to be considered in clinical practice. The barriers reported by the studies were often related to lack of time, inability to understand statistical data, lack of support from employer, lack of resources, colleagues who are not favourable to EBP implementation, lack of interest, and lack of generalisation of results. Comparing the characteristics among different countries, it was not possible to observe a clear pattern of knowledge or skills. In a study conducted in the USA (Jette et al., 2003) 82% of physiotherapists received some information or prior training in EBP, 67% believed to have skills to perform a search strategy for articles, and 66% read 2e5 articles in a typical month. In a study conducted in Australia (Iles and Davidson, 2006), 80% assumed to have a lack of knowledge in EBP, however, 59% stated that they were able to formulate clinical questions often and 69% read articles regularly. Paradoxically, in a German study (Bauer et al., 2007), only 30.8% of physiotherapists reported having received prior information about EBP and 41% claimed to use scientific evidence in their daily practice. Also, a French study (Buchard, 2009), found that 74.6% of physiotherapists reported to have no knowledge in EBP, 71.6% have difficulty applying the information of scientific articles in clinical practice, and 79.1% considered to have little ability to access articles. These findings from studies conducted in Germany and France can be explained because physiotherapists from these countries did not receive their training at a university level. We observed some inconsistencies about EBP over time. In 2003 (Jette et al., 2003) 82% of respondents agreed that they had received information on the principles about EBP, 65% said they had research skills and 55% felt confident regarding these skills. In 2006 (Iles and Davidson, 2006), 80% of respondents agreed that there were gaps in EBP knowledge, in 2007 (Bauer et al., 2007) 48.7% did not know the concept of EBP and 41% do not work based on evidence. In 2009 (Buchard, 2009) 74.6% believed they did not have full knowledge of the main sources of information in research, but 58.6% based their clinical reasoning on EBP. In 2010 (Nilsagård Ylva, 2010) nearly 70% reported good ability to formulate a clinical question and critically appraise scientific literature, but only 21% reported that they have received education in EBP. In 2011 (Scholten-Peeters et al., 2013), 46.3% of physiotherapists rated their EBP knowledge as average, and in 2013 (Queiroz and Santos, 2013) 59.7% of the participants were confident to critically appraise scientific articles, approximately 75% confirmed having knowledge about EBP and 50% agreed that they have learned the foundations of EBP. Finally in 2014 (Bernhardsson et al., 2014) 62% of the respondents agreed or strongly agreed that they felt confident in their ability to find relevant research for their clinical questions. In conclusion, there is no clear pattern of improvement knowledge about EBP over time. In addition, although the lack of time was considered the most important barrier in the studies included, there are some considerable barriers reported related to the workplace like no access to scientific literature at work (Gorgon et al., 2013), no policies at work to stimulate the use of evidence (Gorgon et al., 2013), lack of training at work to use the evidence (Gorgon et al., 2013), inability to make changes in the workplace (Iles and Davidson, 2006), and employers’ lack of interest (Nilsagård Ylva, 2010). There were some barriers related to specific skills such as lack of skills in searching for evidence (Jette et al., 2003; Iles and Davidson, 2006; Gorgon et al., 2013), lack of skills to critically appraise an article (Grimmer-Somers et al., 2007; Salbach et al., 2007; Gorgon et al., 2013), and lack of skills for statistical interpretation (Jette et al., 2003; Grimmer-Somers et al., 2007; Salbach et al., 2007; Gorgon et al., 2013). Besides, there are some barriers related to lack of resources to perform a search strategy (Gorgon et al., 2013), lack of access to computers (Iles and Davidson, 2006), lack of technological equipment (Nilsagård Ylva, 2010), and other related to lack of generalisation of the data for the patient (Jette et al., 2003; Salbach et al., 2007; Queiroz and Santos, 2013). Most of these barriers can be overtakenwith training as well as by promoting workplace changes with regard to a better use of EBP. Some studies in other health professions point out similar difficulties about knowledge (Hadley et al., 2008; Ubbink et al., 2013) and skills (Ubbink et al., 2013), report that the professionals agreed that EBP is important to optimise treatment (Hadley et al., 2008; Ubbink et al., 2013) and that EBP is essential to clinical practice (Hadley et al., 2008). Other studies (Dijk et al., 2010; Zwolsman et al., 2012) with medical students and general practitioners corroborate some of our findings about barriers that also identified lack of time as the main barrier (Dijk et al., 2010; Zwolsman et al., 2012), in addition to factors related to knowledge and skills (Dijk et al., 2010; Zwolsman et al., 2012), factors associated with the profession (Dijk et al., 2010), and attitudes such as lack of motivation or lack of support from colleagues (Dijk et al., 2010), problems with conflicting evidence (Zwolsman et al., 2012), lack of access (Zwolsman et al., 2012) and patient-related barriers (Zwolsman et al., 2012). 4.1. Limitations This systematic review included all studies about the main characteristics that might influence the use of evidence-based physiotherapy. For this review, a sensitive search was performed in the main databases over a wide period of time, and all efforts were made to read and extract the data of all potentially eligible studies and for translation the studies that were reported in another language. Also, all data were presented using frequency analysis, to show clearly the points of view of the professionals. However, it was not possible to find one potentially eligible article in full (Hamzat and Amusat, 2002), which may have biased our estimates. Besides, we consider that some of the response options to the questions used in the included papers were quite vague (e.g. having critical appraisal skills e different people will interpret what this means differently and have different levels of skill, but these differences cannot be ascertained from the data). Similarly, the question responses around using databases and reading scientific articles ‘sometimes in a typical week’ or ‘monthly’ e there is likely to be large variation in these. 4.2. Implications for future studies The study shows many difficulties about the implementation of evidence-based physiotherapy mostly related with lack of knowledge and skills. Lack of support from the employer or the colleges, and limitations related with the profession as lack of time, or lack of autonomy were cited as major barriers for implementing EBP in real life. Nevertheless, our study can be used for rethinking about academic, work, councils, or association initiatives to encourage the access, understanding and use of research in the practice. Although evidence-based physiotherapy have been increasingly recognised and used, the profession does not seem to follow such development. Our review shows many difficulties that professionals have about knowledge, skills, behaviour and barriers the implementation of EBP. However, the original studies do not explain likely explanations for these limitations and do not suggest any interventions for improving these problems. It is clear that more information is needed for a better understanding about EBP.

5. Conclusion

Physiotherapists need to improve their knowledge, skills and behaviour towards EBP, however, they have a positive opinion of EBP. Additionally, the main barriers that hinder the implementation of EBP in physiotherapy are lack of time, inability to understand statistical data, lack of support from employer, lack of resources, lack of support from colleagues for EBP implementation, lack of interest, and lack of generalisation of results.

 

Veja o artigo completo: https://www.researchgate.net/publication/269183360_What_do_physical_therapists_think_about_evidence-based_practice_A_systematic_review