|Home | About | Journals | Submit | Contact Us | Français|
To assess awareness and use of evidence-based medicine (EBM) databases and The Cochrane Library among physicians in Croatia.
A cross-sectional study with a telephone survey was performed among 573 physicians (88.6% response rate from 647 contacted physicians) from family practice and 4 major university hospital centers in Croatia. The main outcome measures were physicians' awareness of The Cochrane Collaboration, awareness and use of The Cochrane Library, access to EBM databases, and access to internet at work.
Overall, 54% of respondents said they had access to EBM databases, but when asked which databases they used, they named mostly non-EBM databases. The question on the highest level of evidence in EBM was correctly answered by 53% respondents, 30% heard of The Cochrane Collaboration, and 34% heard about The Cochrane Library. They obtained information about The Cochrane Library mostly from colleagues and research articles, whereas the information about EBM was gained mainly during continuous medical education. There were more respondents who thought The Cochrane Library could help them in practice (58%) than those who heard about The Cochrane Library (30%). Only 20% of the respondents heard about the initiative for the establishment of the Croatian branch of The Cochrane Collaboration. Family physicians had significantly lower level of awareness, knowledge, and use of EBM and The Cochrane Library than physicians from university hospitals.
There is low awareness about EBM and The Cochrane Library among physicians in Croatia, which creates a need for educational interventions about EBM for the benefit of health care in Croatia.
Evidence-based medicine (EBM) has emerged as a new paradigm for medical practice, integrating individual clinical expertise with the best available external clinical evidence and compassionate use of individual patients’ rights and preferences in making clinical decisions about their care (1). Providing evidence-based care to patients involves turning a clinical problem into an answerable question, systematically searching for the best evidence relevant to the question, critically appraising that evidence, and using the evidence as the basis for clinical decisions to solve the problem (2).
The concept of EBM should persuade clinicians and consumers to pay attention to the best evidence from research that is both valid and clinically applicable, and there still remains much to be done to reach this goal (3). While the overload of medical information today presents a demanding challenge to physicians to sort and identify relevant and valid evidence, it is important to translate that evidence into clinically useful terms (2).
There are two sources of evidence – primary and secondary. Primary evidence is found in original bibliographic citations. Secondary sources include pre-appraised literature, such as systematic reviews written by experts in critical appraisal of primary citations. The use of secondary sources speeds up the process for busy physicians and it helps them to make sense of great amount of available research results by bringing together separately conducted studies, synthesizing their results, and providing evidence for clinical practice (4). EBM databases are resources that provide synopses of evidence-based information in an accessible form, and integration of answers to physicians' patient-related questions into clinical practice in the shortest, most specific way possible. Systematic reviews produced by The Cochrane Library are currently considered to be the “gold standard” in EBM (5).
The Cochrane Library contains high-quality, independent evidence to inform health care decision-making. The Cochrane Library contains several databases, including the Cochrane Database of Systematic Reviews (for Cochrane Reviews), Database of Abstracts of Reviews of Effects or DARE (structured abstracts of systematic reviews), and the Cochrane Central Register of Controlled trials or CENTRAL (which includes records of trials). Cochrane reviews are produced to the highest methodological standards, they search for and synthesize the results of available studies, and are recognized as the gold standard in evidence-based health care (6). The Cochrane Collaboration is an international non-profit organization that was established in 1993 to help various stakeholders to prepare, update, and promote accessibility of Cochrane reviews (7).
Croatia is a South-Eastern European country with a population of 4.4 million (8). Physicians in Croatia are educated in a 6-year medical curriculum with a mandatory course on the principles of research in medicine on the second year, when students are exposed to EBM (9). Further structured education about EBM may be obtained in postgraduate master or doctoral programs or during mandatory continuing medical education (10). In 2004, a non-profit association Evidence-Based Medicine Society Croatia was founded with a purpose to promote EBM in Croatia (11). In 2008, the Croatian Branch of Italian Cochrane Center (CBICC) was founded at the School of Medicine in Split, Croatia (12). The principal goal of the CBICC is the promotion of EBM and systematic reviews through knowledge translation activities (12). Croatian health care institutions and medical schools also have access to a number of EBM databases, including The Cochrane Library. EBM databases are a part of the Center for Online Databases, funded by the Croatian Ministry of Science, Education, and Sports (13).
Since there are no previous studies about EBM awareness and knowledge among physicians in Croatia, there is no evidence to guide planning and providing educational activities of the CBICC or other EBM-related organizations or associations in Croatia. Our study aimed to assess the level of awareness and usage of EBM databases and The Cochrane Library among Croatian physicians working in family practice or clinical hospitals.
A telephone survey of 573 physicians in Croatia was conducted between September 2008 and December 2009. According to the information received from the Croatian Institute for Public Health, there were 2703 physicians employed in family practice in Croatia and 2226 physicians working at 4 Croatian University Hospitals and University Hospital Centers in September 2009.
For the survey of family physicians, we used a representative sample of 350 participants, developed for an earlier study (14). This sample was constructed from the list of all family physicians working in Croatia in 2001 (n=2408) and was stratified by age, sex, vocational training, practice size, and geographical distribution (14). Using a random number generator, we formed a sample of 10% of physicians employed in 4 major clinical University Hospitals in Croatia – in Zagreb, Rijeka, Split, and Osijek (n=223). There were 53 family physicians and 21 physicians from university hospitals who refused to participate in the study. In such cases, another randomly selected physician at the same institution was surveyed; the process was repeated until the next respondent accepted to participate in the survey. The total number of contacted physicians was 647 (403 family physicians and 244 university hospital physicians), and the overall response rate was 88.6%. As non-respondents refused to participate in the telephone survey at its beginning, it was not possible to collect any demographic information on them needed to make comparisons with respondents. All respondents answered all survey questions.
A 30-item questionnaire was developed for the purpose of this study (web extra material). Using a combination of open- and closed-ended questions, physicians were asked about the number of patients they see daily, their need for assistance in diagnosing and deciding on therapy, internet usage, access to EBM databases, knowledge of EBM, awareness and use of The Cochrane Collaboration and The Cochrane Library, their professional status and institution, scientific degree, and age. Cronbach α for the section of the questionnaire measuring awareness and knowledge about The Cochrane library was 0.879.
Descriptive statistical analysis was performed using the GraphPad Prism (GraphPad software Inc, San Diego, CA, USA). The normality of data distribution was tested by Kolmogorov-Smirnov test. Normally distributed data were presented as mean and 95% confidence interval (CI), whereas data that were not normally distributed were presented as median and 95% CI. The age of the respondents from family medicine and university hospital samples was compared using unpaired t-test. The number of patients seen daily in these two groups was compared using Mann-Whitney test. Answers about awareness between family and university hospital physicians were compared using the χ2 test of proportions. Correlation was calculated with Pearson correlation test.
There were 77% (382/573) of women and 33% (191/573) of men in the total sample. The mean age was 46.8 (95% CI, 46.0-46.3) for women and 47.7 (95% CI, 46.3-49.1) for men. Median number of patients seen daily by family physicians was 55.0 (95% CI, 53.9-56.6) and 15.0 (95%CI, 17.5-20.7) by physicians from university hospitals. Among family physicians, 15% held master’s and/or doctoral degrees, compared with 43% university hospital physicians (Table 1).
Most respondents (90%) indicated that sometimes they needed help in making a diagnosis, but there was no statistical difference between family and university hospital physicians (P=0.79). When they did not know how to reach a decision about their patients, physicians from both groups most frequently consulted colleagues, followed by books, the internet, and research articles (Table 2). Significantly more family physicians had internet access at work: 99% vs 83% among hospital physicians (P<0.001). However, significantly more hospital physicians than family physicians used the internet to solve dilemmas about patients (82% vs 65%, P<0.001). Common search engines, such as Google search, were the most common internet resource used among family physicians, compared with PubMed among hospital physicians (Table 2).
Regarding EBM, half of the respondents (n=311, 54%) indicated that they had access to specialized EBM databases, but only 112 of these (36%) named EBM databases they used (Table 3). Knowledge of EBM was tested with a question about the hierarchy of evidence in medicine, where 52% (299/573) of all respondents answered correctly that systematic review of randomized controlled trials was the top level in the pyramid of evidence. Family physicians gave correct answer to this question significantly less often than hospital physicians (44% vs 65%, P<0.001). The majority of surveyed physicians got basic information about EBM during continuous medical education (356/573, 62%) or/and postgraduate school (129/573, 23%) (Table 3).
The familiarity with the concept of The Cochrane Collaboration was assessed with a set of 14 questions. Overall, 30% of all respondents said that they had heard about The Cochrane Collaboration (17% among family and 51% among hospital physicians) and 34% had heard about The Cochrane Library (20% family vs 55% hospital physicians; P<0.001). Among those who had heard about The Cochrane Library, the majority in both groups learned about it from colleagues and research articles (Table 4). There were 10% (58/573) of physicians who had used The Cochrane Library and 9% (54/573) who had read systematic reviews from The Cochrane Library. The knowledge and the use of The Cochrane Library was more common among hospital physicians. The respondents mostly read summaries of systematic reviews instead of the whole systematic reviews, most of them accessed The Cochrane Library from work, and most used it less than once a month (Table 4). When asked if The Cochrane Library helped them in solving problems in practice, 39 (67% of those who said they were users) responded that it did not help at all or helped very little, whereas 15 (28% of users) reported that it helped them enough or very much. A single respondent said that the satisfaction with the help from The Cochrane Library was complete (Table 4). When asked if they thought that The Cochrane Library could help them solve problems in their practice, 58% (n=331) answered positively and 30% (n=171) indicated that they would like to learn the methodology for doing Cochrane systematic reviews (27% family physicians and 34% hospital physicians). There were 16% (n=57) of family physicians and 27% of hospital physicians (n=64) who did not hear about The Cochrane Library, but indicated that they would like to learn Cochrane methodology.
Internet access at work or use of The Cochrane Library was not correlated with age (r=-0.058, P=0.163 and r=-0.043, P=0.316; respectively), and using the internet to solve dilemmas about patients was not related to sex (r=0.048, P=0.247). Male sex was associated with knowledge about the existence of The Cochrane Library (r=-0.110, P=0.008) and access to the internet at work (r=-0.094, P=0.025), but female sex was associated with its use (r=0.086, P=0.043). Younger age was associated with using the internet to solve dilemmas about patients (r=-0.112, P=0.007) and knowledge about The Cochrane Library (r=-0.140, P=0.001). All of the associations were weak.
In the whole sample, 20% (n=115) respondents were aware of the initiative for establishing the Croatian Branch of The Cochrane Collaboration, with a significant difference between the two groups (9% family vs 37% hospital physicians; P=0.001) (Table 4). As the CBICC started publishing the summaries of Cochrane reviews in Lijecničke novine, the official journal of the Croatian Medical Chamber received by all licensed physicians in Croatia in June 2008, we also assessed whether these summaries were read by physicians in Croatia. Only 95 physicians (17%) said that they read these summaries; among them only 6 (6%) were able to recall at least one title (Table 4).
Our survey showed that about a half of the physicians in Croatia had access to the EBM databases and that the same fraction knew that systematic review was the highest level of evidence in EBM. About a third of the respondents heard about The Cochrane Collaboration and The Cochrane Library but only one tenth of respondents reported its use, mostly to read summaries of reviews. Significantly lower level of EBM knowledge and awareness was present among family physicians than among university hospital physicians. These findings demonstrate that the awareness and use of EBM databases and The Cochrane Library among physicians in Croatia is not satisfactory.
These results need to be considered with caution. For example, one fifth of physicians expressed their willingness to learn about the methodology of making Cochrane systematic reviews even though they previously said that they had not heard about The Cochrane Library. Also, in both groups there were more participants who thought that The Cochrane Library might help them in practice than those who had heard about The Cochrane Library. Such discrepancies may indicate social desirability bias (16) or respondents’ wish to learn about such tools even if they were not familiar with them. When participants were asked which EBM databases they used, the most frequent responses were MEDLINE and PubMed, which are not EBM databases, demonstrating the actual lack of familiarity with EBM databases.
Surveys of knowledge, use, and attitudes toward EBM were performed in different geographical and socio-economic settings. In a survey about attitude, awareness, and practice of EBM in government hospitals in Saudi Arabia (17), almost all respondents had positive attitude about EBM but only about a half reported regular use of EBM in their daily clinical practice. The lack of regular distribution of updated clinical letters, journals, or guidelines was considered to be a major barrier for practicing EBM, followed by a lack of available time and internet access. Participants in this study also reported low level of awareness about extracting journals, review publications, and databases related to EBM. A survey of attitudes, awareness, and barriers regarding evidence-based surgery among surgeons and surgical nurses in The Netherlands showed that 90% of the surgeons were familiar with evidence-based surgery terms, compared with only 40% of the nurses (18). Common barriers for surgeons were conflicting results and the methodological inadequacy of research reports and unawareness of evidence-based surgery and unclear research reporting for nurses. In a survey performed among general practitioners in England about their perceptions of the route to EBM, respondents mainly welcomed EBM and agreed that its practice improved patient care (19). They had a low level of awareness of extracting journals, review publications, and databases (only 40% were aware of the Cochrane Database of Systematic Reviews), and, even if aware, many did not use them. At work, 20% had access to bibliographic databases and 17% to the internet. Most had some understanding of the technical terms used. These results are comparable with our findings, although, unlike Croatia, England is a high-income country. A study from India provides a developing country perspective (20), showing that a half of the surveyed physicians were aware of EBM, one third were aware of The Cochrane Library, and 15% were aware of systematic reviews. Unlike physicians, nurses and health care consumers were not at all aware of EBM, The Cochrane Library, and systematic reviews.
In our study, physicians from family medicine practices were less familiar with EBM and The Cochrane Library than physicians from a university hospital setting. This can be in part explained by the fact that primary health care offices are not a part of the academic network providing access to online bibliographical and information resources. However, family medicine practices have access to the internet and thus to the abstracts of The Cochrane Library and other free EBM-related content.
In Croatia, primary health care has a long tradition and represents patients’ gateway to the health care system (21). A recent study showed that family medicine practitioners in the context of intensive socioeconomic transitions in Croatia had positive attitude toward scientific research in medicine (22). However, positive attitudes do not suffice for real engagement of family physicians in teaching and research (23,24). Research has shown that residency programs that include research training did not affect research output or use of research in everyday practice (23), indicating that greater emphasis should be placed on teaching EBM principles and encouraging research activities already in medical school.
One of the potential reasons for the differences between family and hospital physicians could be the difference between them in obtained postgraduate degrees. There were three times more hospital physicians with master’s and doctoral degrees than family physicians. Postgraduate schools offer more opportunities to learn about EBM, and preparation of thesis requires literature search and use of various databases. It has already been observed that family physicians in Croatia had positive attitude toward science, but only 6% published an article in a PubMed-indexed journal (22). At 4 medical schools in Croatia, a total of 61 faculty members in the Departments of Family Medicine in 2007 published 316 scientific and professional articles and a half of them had research degrees (25). However, family medicine faculty comprises only 2% of family physicians in Croatia and they are not representative of family medicine practitioners in Croatia. While there are no systematic surveys about publications of Croatian family physicians, research from other settings shows that the number of research articles in family medicine field is decreasing (26). Attitudes toward science have not been explored among physicians in Croatia employed in hospitals, but it has to be kept in mind that our survey sample included physicians working in academic hospital setting of the university hospitals.
Our study focused on physicians and not on other health care workers who should also practice EBM principles and use EBM databases, which may present a limitation of the estimate of awareness and use of EMB in health care in Croatia. However, based on the experience from both developed and developing countries (19,20), we would expect that health care professionals other than physicians have an even lower awareness and use of EBM. Another limitation of the study is unbalanced sample of family physicians and hospital physicians, covering about 14% of the family medicine and 10% of the university hospital physicians. However, these differences were not of the size that would greatly affect study results or limit their generalization to the whole populations of physicians. Furthermore, we did not explore barriers for practicing and using EBM sources in Croatia, which we plan to address in future studies.
In conclusion, our study demonstrated that there was a need to educate physicians in Croatia about EBM and the usefulness of EBM resources such as The Cochrane Library. The majority of Croatian pyhsicians have access to the internet and EBM databases in public health care institutions, which solves at least one of the obstacles for practicing EBM observed in other studies (27). Establishing awareness and educational needs of physicians is a good step forward in devising strategies to promote EBM in a transitional country like Croatia.
We are grateful to our colleagues from family medicine practices and the university hospitals for their participation in the survey. We thank Darko Hren and Mirjana Huić for their help in the development of the questionnaire, Dario Sambunjak for helping with the drafting of the article, Mario Trošelj from the Croatian Institute for Public Health for providing the data on the number of physicians in Croatia; Damir Sapunar for statistical advice, and Silvija Maslov Kružičević for technical help with the survey.
A. M. is the Co-editor in Chief, M. M. is the Emeritus Editor, and K. V. and L. P. are executive editors of the Croatian Medical Journal. To ensure that any possible conflict of interest has been addressed, this article was reviewed according to best practice guidelines of international editorial organizations, as outlined in the journal’s Guidelines of Authors.