Development and psychometric testing of an instrument to evaluate cognitive skills of evidence based practice in student health professionals
© Lewis et al; licensee BioMed Central Ltd. 2011
Received: 25 March 2011
Accepted: 3 October 2011
Published: 3 October 2011
Health educators need rigorously developed instruments to evaluate cognitive skills relating to evidence based practice (EBP). Previous EBP evaluation instruments have focused on the acquisition and appraisal of the evidence and are largely based in the medical profession. The aim of this study was to develop and validate an EBP evaluation instrument to assess EBP cognitive skills for entry-level health professional disciplines.
The Fresno test of competence in evidence based medicine was considered in the development of the 'Knowledge of Research Evidence Competencies' instrument (K-REC). The K-REC was reviewed for content validity. Two cohorts of entry-level students were recruited for the pilot study, those who had been exposed to EBP training (physiotherapy students, n = 24), and who had not been exposed to EBP training (human movement students, n = 76). The K-REC was administered to one cohort of students (n = 24) on two testing occasions to evaluate test-retest reliability. Two raters independently scored the first test occasion (n = 24) to evaluate the inter-rater reliability of the marking guidelines. Construct validity was assessed by comparison of the two groups, 'exposed' and 'non-exposed', and the percentage of students achieving a 'pass' score in each of these groups. Item difficulty was established.
Among the 100 participants (24 EBP 'exposed', and 76 EBP 'non-exposed' students), there was a statistically significant (p < 0.0001) difference in the total K-REC scores. The test-retest and inter-rater reliability of the individual items and total scores ranged from moderate to excellent (measured by Cohen's Kappa and ICC, range: 0.62 to perfect agreement).
The K-REC instrument is a valid and reliable evaluation instrument of cognitive skills of EBP in entry-level student health professionals. The instrument is quick to disseminate and easy to score, making it a suitable instrument for health educators to employ to evaluate students' knowledge of EBP or in the evaluation of entry-level EBP training.
The most widely accepted definition of evidence based practice (EBP) involves three components: the integration of the best research evidence with patient values and clinical expertise . Evidence based practice is now well established in the health and social care professions . To date, the main focus of EBP has been on the behaviours (application and practice) of health professionals in the clinical and research environments . More recently, the EBP movement has expanded to include the educational processes (content, delivery and assessment) involved in entry-level training of health professionals .
It has been proposed that EBP consists of a five step process [1, 5, 6]. Previous authors have suggested that this five step model should form the basis not only for health professionals' clinical decision making and practice but for curricular content underpinning EBP training of health professionals . The first four steps of this model involve asking a clinical question, acquiring and appraising the evidence, and applying the evidence into clinical practice. The fifth step encourages individuals to reflect upon the process undertaken in the first four steps. This final step may also provide an opportunity for EBP training providers to undertake formal evaluation procedures to assess the effectiveness of EBP training. The recommendations provided in the Sicily Statement highlight a need for future educational research evaluating EBP training, and the development and application of evaluation instruments to assess EBP training for each of the five steps .
A previous systematic review  identified 104 instruments for evaluating the effectiveness of EBP training, the majority of which were developed or tested with medical students or practitioners and the minority (n = 13) developed or tested on other health professions. Seven of the 104 instruments identified in this review were recognised as 'Level 1 instruments' (supported by established inter-rater reliability, objective outcome measures, and three or more types of established validity). The 'Fresno test of competence in evidence based medicine'  and the Berlin Questionnaire  were the only 'Level 1' instruments identified as evaluating all aspects of EBP competence . The 'Fresno test' consists of two clinical scenarios and open ended questions to measure medical professionals' skills and knowledge across the four main steps of the EBP process. The Berlin Questionnaire consists of 15 multiple choice questions which mainly focus on participants' epidemiological skills and knowledge. While the Berlin questionnaire is described as an instrument that assesses EBP competence , it only comprehensively evaluates the third step (appraisal) of the EBP process .
EBP evaluation instruments identified in an updated systematic search (post 2006) and comparison to the Fresno test
No. of knowledge items
EBP content assessed
Levels of evidence
Ramos et al 2003
12 (short answer/essay)
Inter-rater reliability, IC, content, discriminative and responsive validity
Caspi et al 2006 
IC, content validity
KA, KSR, A
Krueger 2006 
not reported (m/c)
Content and discriminative validity
Novack et al 2006 
not reported (m/c)
Content and discriminative validity
Meyer et al 2007 
13 (not reported)
Inter-rater reliability, content and responsive validity
Shuval et al 2007 
written assignment + online exam
not reported (short answer/essay)
Inter-rater reliability, content validity
KA, A, U, B
Siriwardena et al 2007 
Manchester short EBM survey
IC, content and discriminative validity
Davis et al 2008 
5 (m/c, short answer)
IC, content, discriminative and responsive validity
Ahmadi-Abhari et al 2008 
IC, content and discriminative validity
McCluskey & Bishop 2009 
Adapted Fresno test of competence in EBP
7 (short answer/essay)
Inter-rater reliability, IC and responsive validity
Tilson 2010 
Modified Fresno Test
13 (short answer/essay)
Inter- and intra-rater reliability, IC, content and discriminative validity
At the time of the current study (2008), the instrument developed in this study, the 'Knowledge of Research Evidence Competencies' instrument (K-REC) was the first known tool prospectively designed to collect data on cognitive EBP skills of entry-level allied health students. Since the initial literature search and development of the K-REC, two further instruments have been published that evaluated cognitive skills of EBP in the allied health disciplines [20, 21]. The 'Adapted Fresno test of competence in evidence-based practice' (AFT) was developed for occupational therapists, based on the first seven items of the Fresno test. Inter-rater reliability (ICC 0.91-0.96) and internal consistency (Cronbach's α 0.74) was confirmed with a sample of 10 occupational therapists representative of the wider sample studied (75% graduated > 5 years) and the AFT demonstrated responsiveness before and after a two day EBP workshop . The 'Modified Fresno test' (MFT) for physical therapists considered all 12 items of the Fresno test, with item stems almost identical to the original test . Two further items were added to the modified test regarding the other pillars of EBP; patient preferences and clinical expertise, with the clinical expertise item removed following poor psychometric performance. Three cohorts of respondents were recruited for this study, first year physical therapy students (novices n = 31), third year Doctor of Physical Therapy students (EBP trained students n = 50) and physical therapy faculty (EBP experts n = 27). The 'Modified Fresno test' for physical therapists demonstrated inter- and intra-rater reliability (inter-rater ICC 0.91, intra-rater ICC 0.95, 0.96), internal consistency (Cronbach's α 0.78) content and discriminative validity (significant difference in total score corresponding to training level p < 0.0001) .
While the Fresno test has been recognised as the current best available instrument to comprehensively assess cognitive skills of EBP , the instrument is not without limitations as noted by developers of two modified versions [20, 21]. The Fresno test includes open ended questions requiring short essay style answers. The answers are then assessed using a standard scoring template with points awarded for key components of each answer. The authors of the Fresno test estimate that respondents require up to 60 minutes for completion of the test . The test is therefore lengthy to complete and difficult to score, with training necessary to score the test. The original test was not relevant to health professions outside of medicine, with medical based clinical scenarios, and reference to issues of diagnosis which would be unlikely to be raised in other health professions . The original Fresno test has been modified for single profession use in occupational therapy  and physical therapy . Both of the modified Fresno tests have retained the open ended style of questions and complex scoring template of the original test. Given the absence of an instrument specifically developed for students across entry-level health professional programs, there was a need for the development of an instrument based on the Fresno test as a means to assess cognitive EBP skills across a range of entry-level health professional disciplines.
Due to the focus of this study on the assessment of entry-level education (novice rather than postgraduate learners or clinicians), the instrument was intended to evaluate the first three fundamental steps of the EBP process model (ASK, ACQUIRE and APPRAISE), rather than assessment of the fourth step involving the application of knowledge and skills in clinical practice. The aim of this study was to develop an instrument to evaluate entry-level respondents' cognitive skills regarding the research evidence component of EBP that:
could be used across the health professions;
was quick to complete and easy to score; and
was underpinned by a transparent and defensible psychometric testing process.
The development of the instrument was completed in two stages. The first stage involved the development of the instrument and the second stage comprised the processes used to psychometrically evaluate the instrument.
Ethical approval for the study was obtained from the University of South Australia Human Research Ethics Committee (protocol number P067/08).
Stage 1: Development of the instrument
Firstly, the research team considered the content of the Fresno test in the drafting of the instrument. The first seven items of the Fresno test were relevant for application to health professions external to medicine and were considered in the development of the K-REC instrument. The remaining Fresno items (8-12) involved advanced statistical calculations (eg predictive values, number needed to treat, risk reduction), and questions about diagnosis and prognosis. While these skills are pertinent to all health professions, the instrument was intended for use with entry-level health professional students with varying degrees of exposure to EBP education (novice learners rather than graduates of professions). Therefore, a choice to focus on what might reasonably be expected to be taught generically across a range of degrees/stages of programs and across disciplines was required. In both forms of the modified Fresno test [20, 21], the specific statistical items of the original Fresno test have been altered or removed. The choice was made to limit statistical knowledge items in the instrument to interpretation of metrics of significance and focus upon questions of intervention. Knowledge of levels of evidence was not assessed in the original Fresno test but was considered an essential component and was therefore included.
One clinical scenario template was designed for the K-REC, with specific inclusion of each of the PICO components. The clinical scenario needed to be relevant to a variety of health professional disciplines and was therefore deliberately designed around the topic of a chronic condition (cystic fibrosis) and two possible types of intervention for this condition. The interventions included in the scenario (exercise and breathing exercises) were intentionally chosen to be understood by a number of health professional disciplines (e.g. physiotherapy, occupational therapy, podiatry, human movement).
Given the time and scoring related limitations of the Fresno test, the K-REC was designed to be relatively quick for respondents to complete and easy to score. The K-REC therefore consisted of a combination of short answer, multiple choice and true/false type questions rather than conforming to the open ended short essay answer style of the Fresno test. The final draft of the K-REC instrument was distributed to three senior academics with expertise in research methods and EBP and two entry-level physiotherapy students at the University of South Australia to review the clarity and formatting of the items and instructions. The instrument underwent three rounds of review and minor revision to the wording, instructions and layout of the instrument.
The final K-REC instrument
The K-REC instrument was designed to evaluate cognitive skills of EBP, with a combination of items evaluating either knowledge, or skills in the application of knowledge (for example, the application of knowledge to write a relevant research question based on the clinical scenario). The K-REC instrument was intended to assess entry-level respondents' cognitive skills regarding the research evidence component of EBP. The instrument therefore covered the first three steps of the five step EBP process model relevant to cognitive skill (ASK, ACQUIRE and APPRAISE), but did not consider the last two steps involving the application of evidence into practice (APPLY and ASSESS) . The final draft of the K-REC consisted of a clinical scenario template and nine items relating to that scenario (Additional file 1).
The 'correct' answers for each item of the K-REC were determined through review of the Fresno test, and consultation and consensus of the research team and senior academics during the development procedure. The two open-ended items (1 and 7) were scored according to the presence or absence of set criteria. For item one, respondents were allocated half a mark for each of the PICO (Population, Intervention, Control or Comparator, Outcome) components provided in their question. In item seven, respondents were required to list four characteristics of randomised controlled trials that would increase their confidence that the research was methodologically sound. Half a mark was awarded for any characteristic that was consistent with the Physiotherapy Evidence Database (PEDro) scale  which is an appraisal of methodological bias tool specifically for randomised controlled trials. Two other characteristics of randomised controlled trials were added as correct answers for this item as a result of answers provided by students during the survey development procedure (refer to the K-REC marking guidelines in Additional file 2).
Stage 2: Pilot study: psychometric evaluation of the K-REC instrument
The aims of Stage 2 of the instrument development process were to explore the practicalities of the survey process (recruitment, dissemination, workload required of participants, and response rate), establish test-retest and inter-rater reliability, investigate the validity of the K-REC, and seek informal feedback from the participants about the appearance, layout and general user friendliness of the instrument.
The recruitment strategy aimed to enlist two main types of participants:
'Exposed': participants who were representative of the target population for the K-REC and who could be reasonably expected to have a good understanding of EBP through exposure to formal EBP training, and,
'Non-exposed': participants who could be reasonably expected to have minimal understanding of EBP through minimal or no exposure to formal EBP training.
Third year physiotherapy students (n = 24) were invited to participate as the 'exposed' group. At the time of the study (2008), these students had completed either a mandatory stand-alone EBP course (13 weeks) or an Honours research preparation course (2 weeks and ongoing supervision within a research team) where they were taught and assessed on the different components of EBP. Therefore, these students had prior exposure to formal training and were expected to have a foundation knowledge and understanding of EBP.
Human movement students (n = 89) were invited to participate as the 'non-exposed group'. Rather than compare first and third year students from the same professional discipline (physiotherapy), human movement students were selected as they would have been exposed to similar information technology literacy (databases, searching) and were at the same time point in their entry-level programs as the physiotherapy students (third year). The human movement students were expected to have minimal knowledge of EBP, as at the time of the study these students were not required to enroll or be assessed in any EBP or research training course and were therefore not exposed to any formal standardised EBP or research instruction.
Pilot study procedure
The entire group of 24 physiotherapy students was invited to complete the K-REC instrument at a face to face class. Students were requested to write their student identification number on the survey to be used to match the second round of surveys (test-retest). A second administration to the participating physiotherapy students (n = 24) was completed three days after the first testing occasion. The completed surveys from the first testing occasion (n = 24) were independently scored by two raters to establish the inter-rater reliability of the instrument. In order to establish the discriminative validity of the K-REC, the 'non-exposed' cohort of human movement (n = 89) students were invited to participate at a face to face class. The mean item and total K-REC scores of each of the student groups ('exposed' and 'non-exposed') were compared to establish discriminant validity.
Test-retest and inter-rater reliability was assessed by using the intraclass correlation coefficient (ICC) for items with interval data or those with ordinal data where the intervals between measurements were assumed as equivalent (item 7 and total scores). The reliability of the remaining items was assessed using Cohen's Kappa coefficient (un-weighted) for categorical data (items 2, 4, 5, 6, 8 and 9) and percentage agreement (items 1 and 3). Agreement of less than 0.50 was classified as poor, between 0.50 and 0.75 as moderate, and greater than 0.75 was classified as a good level of reliability .
Discriminative validity was assessed by comparison of the individual item and total scores from the two student groups (human movement and physiotherapy) with calculation of unpaired t tests, and the z-test for comparing proportions (two-tailed) for the differences between the percentage of students achieving a pass (≥ 50% score) mark for each of the survey items in both student groups . Probability values of less than 0.05 were deemed statistically significant.
The K-REC achieved an overall response rate of 88 per cent (physiotherapy students: 100%, human movement students: 85%). The average completion time for the K-REC was approximately 10 minutes.
Test-retest and inter-rater reliability of the K-REC instrument
Item no. and content assessed
1 Research question (PICO)
2 Sources of information
K 0.91 (0.85 - 0.97)
3 Study design knowledge
4 Search strategy (MeSH)
K 0.93 (0.79 - 1.00)
5 Search strategy (Boolean)
K 0.77 (0.54 - 1.00)
K 0.83 (0.44 - 1.22)
6 Critical appraisal
K 0.71 (0.32 - 1.00)
7 Critical appraisal
8a Research evidence statistics
K 0.86 (0.68 - 1.00)
8b Research evidence statistics
K 0.94 (0.82 - 1.00)
9 Levels of evidence
K 0.62 (0.32 - 0.92)
K-REC total scores
Comparison of scores obtained from the 'exposed' and 'non-exposed' groups and the percentage of students who achieved a ≥50% pass mark for each K-REC item
Item no. and content
Max. possible score
pvalue (unpaired t test)
Percentage of students passing (≥50%)
(z score, 2 tailed)
n = 24
n = 76
n = 24
'Non-exposed' n = 76
1 Research question (PICO)
2 Sources of information
3 Study design
4 Search strategy (MeSH)
5 Search strategy (Boolean)
6 Critical appraisal
7 Critical appraisal
8a Research evidence statistics
8b Research evidence statistics
9 Levels of evidence
Mean total score
(range 6.5 - 11.0)
(range 1.0 - 11.5)
Considering the mean scores of each group of students for each K-REC item, for nine out of 10 (item eight has two sections which were considered separately) instrument items a higher proportion of 'exposed' than 'non-exposed' students passed. A z-test for comparing proportions (two-tailed) was performed to investigate the differences between the groups resulting in a statistically significant difference between groups for seven of the 10 instrument items (Table 3).
The values presented in Table 3 of the percentage of pilot study participants who achieved a pass mark (≥50% score) for each K-REC item and the total scores were a representation of the relative difficulty of each of the items. The relative difficulty of each item ranged from 'difficult' with only 26 per cent of the total pilot study participants passing the first section of the research evidence statistics question (8a) and 15 per cent passing the second section (question 8b), to 'moderate' with 84 per cent of participants achieving a pass mark for question two on the identification of sources of information.
The K-REC met the basic psychometric requirements for reliability, validity and usability of a standard instrument for collecting data on respondents' cognitive skills of EBP. The information gained through testing the instrument on two health professional discipline areas at one institution suggests that if the instrument is disseminated in a similar fashion, a reasonable response rate should be achieved.
There are a variety of instruments available to assess EBP cognitive skills, some of which have been demonstrated to discriminate between novice and expert learners. The issue with pre-existing tools is that for novice learners, the relevance of the scenario questions and a floor effect is likely to be the issue especially with entry-level students in the early years of their programs. The K-REC instrument was designed to assess the cognitive EBP skills of novice learners (entry-level students) irrespective of their professional discipline. That is, the anticipated target audience of this instrument was predominantly entry-level health professional students, who could range from pre-novice (no exposure to EBP principles at all), novice and intermediate EBP learners. While the K-REC could be used for expert learners, we suspect that a ceiling effect is likely to occur which may make it a useful tool for comparison between participants of varying EBP exposure but it is likely to be insensitive to change in EBP experts (which is likely to be the case for most, if not all other EBP knowledge instruments).
The K-REC differed from existing instruments in terms of design, length, content and psychometric properties. The K-REC contains a clinical scenario template that may be relevant to a variety of health professional disciplines and individual items relating to the scenario. The K-REC therefore has the potential to be applicable to a variety of health disciplines, rather than being specific to one discipline. In contrast to the original Fresno test , and two modified versions [20, 21], the K-REC was designed to be short, easy to complete and disseminate. The short answer and multiple choice design ensured the usability of the instrument as reflected in the short completion times.
Summary of the psychometric properties of the K-REC*
(comparison of scores on an initial test to scores by the same participant on a retest)
(the degree of agreement between two independent raters)
Cohen's Kappa, ICC and percentage agreement
At least a moderate level of agreement (0.50) between testing occasions (test-retest) or raters (inter-rater)
Each of the items and total scores achieved at least a moderate level of agreement (range: 0.62 to perfect agreement)
Each of the items and total scores achieved a very good level of agreement between raters (range: 0.83 to perfect agreement)
Content validity (instrument covers entire topic of interest)
Test covers all of the main aspects of EBP
Content and revisions based on experts' suggestions
Item difficulty (relative difficulty of each item)
The % of candidates who answer achieve a passing score
A wide range of difficulties allows a test to be used with both 'exposed' and 'not exposed' groups
Ranged from moderate (84% question 2) to difficult (15% question 8b)
Construct validity (evidence that the instrument measures the construct that it intends to)
Mean scores of 'exposed' and 'not-exposed' compared by t test
Significant difference, higher 'exposed' student scores
'Not exposed' mean (human movement) was 4.2 and 'exposed' mean (physiotherapy) was 8.4 (p < 0.0001)
% passing for 'exposed' and 'not-exposed' groups compared by the z-test for comparing proportions
Higher % of 'exposed' students passing
For 9 out of 10 instrument items a higher proportion of 'exposed' (physiotherapy students) than 'not exposed' (human movement students) passed. There was a significant difference between groups for 7 of the 10 instrument items.
Developers of instruments assessing EBP knowledge face a number of philosophical and practical issues. While the Sicily Statement clearly conveys the five fundamental steps of EBP, no similar consensus statements exist to guide which specific knowledge and skills might be considered essential to all learners regardless of stage of training (novice or expert) or professional discipline versus optional, advanced or profession specific knowledge. This might explain why there is such a profusion of derivative instruments from the original Fresno test which modify, exclude or add items concerning statistical concepts, scenarios relevant to various professional groups, or focus upon questions of intervention rather than giving equal weight to questions of prognosis or diagnosis. Where instruments have been developed to assess the learning outcomes of specific EBP training courses, it is likely that the key items included within the assessment are reflected within the specific EBP education program. For example, the content of the Fresno test  was designed "to assess the effectiveness of a comprehensive evidence based medicine curriculum in the University of California, San Francisco's Fresno family practice residency programme" (pg 319) and as such, the content of the instrument reflected the learning objectives of a specific evidence based medicine course reviewed and revised by family practice residents, faculty and self identified evidence based medicine experts. In contrast, the Adapted Fresno Test  was developed to assess change in skills and knowledge after completion of a two day EBP workshop for occupational therapists. The curriculum of the workshop did not include number needed to treat, risk reduction or how to interpret studies of diagnostic accuracy, which were considered to be more suitable for advanced workshops.
While it is clear that entry-level students should be able to critically analyse and evaluate the usefulness of research in disciplines such as physiotherapy and occupational therapy [26, 27], there is currently no consensus about what statistical content should be included in entry-level health professional curricula. In medicine, it has been clearly documented that the processes used to teach, facilitate learning and assessment of medical statistics should be student centred and clinically integrated [28, 29]. However, there are currently no clear guidelines about exactly what content should be taught in undergraduate medical programs, with studies on statistics training showing differing content between institutions [30, 31]. Our intent in developing the K-REC was to include items which could be reasonably expected to be included within entry-level curricula relating to EBP education regardless of year level across a range of health professional programs, the majority of which are targeted at providing health care interventions. Consequently rather than include more advanced statistics (risk ratio, number needed to treat) we opted to include metrics of significance (p value, effect size and confidence intervals) and questions relating to intervention (rather than prognosis/diagnosis).
An additional issue concerning development of EBP evaluation instruments is whether the instrument is intended for repeated use with the same group of participants (eg pre and post EBP training). Where the identical instrument is used on repeated occasions with the same group, concerns arise over respondent familiarity with the items and expected answers. In a number of instruments, provision has been made for alternate items to be used on repeated occasions (Berlin: Set A or B, APT: two clinical scenarios, MFT: three clinical scenarios). The clinical scenario for a question of intervention developed for the K-REC was intended to comprise a template where the specific condition, intervention, comparator and outcome could be modified without altering the intent of the questions relating to the scenario (Q1: write a PICO format search question, Q2: identify sources of information likely to provide reliable/valid information Q3: most appropriate design to answer search question etc). For example, the PICO components relevant in the scenario included within this tested version of the K-REC were:
"Jane is a 16 year old girl who has cystic fibrosis (POPULATION) and she has recently been admitted to hospital with a chest infection. Jane normally self treats at home with breathing exercises (INTERVENTION) taught to her by a physiotherapist. One of her friends also has cystic fibrosis (POPULATION) but she treats herself with exercise (COMPARATOR), not breathing exercises. Jane wants to know whether her lung condition (OUTCOME) would be more effectively managed with an exercise program. You have no experience of either breathing exercises or exercise programs for cystic fibrosis and are not sure what to recommend."
This scenario could be modified to be relevant to any health professional discipline by altering the PICO components, for example, based on one of the clinical scenarios within the AFT , the scenario could be modified to:
"Jane is a 16 year old girl who has a traumatic brain injury (POPULATION), currently completing her final year exams and referred to you for memory rehabilitation. Jane has a number of usual strategies (INTERVENTION) but one of her friends also has had a traumatic brain injury (POPULATION) but she manages her memory issues with a diary (COMPARATOR). Jane wants to know whether she would function better at school (OUTCOME) with a diary. You have no experience of the usual strategies or diaries for memory problems in traumatic brain injury and are not sure what to recommend."
When the K-REC instrument with the above alternate clinical scenario was completed by a sample of entry-level physiotherapy students (n = 15), the inter-rater reliability of the marking guidelines remained high (2 raters, item 1: 93% agreement, items 2-6 and 8-9: 100%, item 7: ICC 0.97, and total scores: ICC 0.97). The K-REC currently consists of a clinical scenario template, with two 'tested' scenarios. If the instrument were to be applied longitudinally, it would be necessary to develop and test a 'bank' of alternate clinical scenarios. While the option is there for users of the K-REC instrument to modify the clinical scenario template to suit various health professional disciplines, it is important to note that any modified scenario would need to be tested appropriately. It is also worth considering that the clinical scenarios contained in the original Fresno test, two modified versions and the instrument developed in the current study were all based on intervention style questions. The Centre of Evidence Based Medicine Levels of Evidence  currently recognises six different types of EBP questions (pre-test probabilities, diagnostic accuracy, prognosis, treatment benefits, treatment harms, and screening). While the original and the modified Fresno test for physical therapy [9, 21] included two final open ended items unrelated to the scenarios relating to the identification of the best study design for questions of diagnosis and prognosis, it is interesting to note that none of the instruments contained clinical scenarios or the majority of items relating to different types of questions.
It is possible that the three days between test occasions resulted in a degree of respondent familiarity with the items in the K-REC instrument for the test-retest reliability testing. The choice of how long to wait before retesting knowledge items requires a trade-off between the likelihood of participant recall and learning. In test-retest situations where the intent is to determine stability of respondent's answers, a longer duration is likely to reduce the potential for participants to simply recall the answers they provided during the first test occasion but increases the possibility of ongoing learning of the specific knowledge concepts included in the original test. In the current study, participants were university students currently engaged in courses where it was likely that there would be ongoing exposure to EBP principles and knowledge. The choice of a shorter period between test occasions was selected as a compromise between recall and learning.
The predominant multiple choice question design of the K-REC has both advantages and disadvantages. By providing a list of possible answers, it may encourage respondents to complete the question (rather than omit the question) due to the fact that the 'correct' answer is somewhere in the list of options. The answers in the list may also act to prompt respondents if they are unsure or having difficulty remembering the answer to a particular question. It is possible that if the same items were designed as 'fill in the blank' type questions, respondents may be less likely to guess the answer. It is not possible to know from the data collected in this study whether respondents were influenced by the provision of answers in the multiple choice items in the K-REC.
The items in the K-REC that related to the second step of the EBP process (ACQUIRE) did not discriminate between the groups with differing exposures to formal EBP training (items 2, 4 and 5). The ability to effectively search the literature is an essential step in the EBP process . Identifying sources of information such as databases and journals, and completing an efficient and effective search strategy requires skills and knowledge that are applicable to most areas of health professional training, rather than being specific to EBP. It is likely that the students involved in this study (third year students), had previously obtained skills and knowledge in searching the literature effectively in earlier years of their entry-level programs (eg for assignments), regardless of the presence or absence of formal EBP training. It is also possible that the true/false design of items 4 and 5 may have resulted in an inability of these items to discriminate between leaners who 'know about the concept' of search strategy, as opposed to learners who 'know how to apply' the concept . As the K-REC instrument was designed to evaluate skills and knowledge of EBP, rather than behaviours (actual performance of EBP in practice), the underlying reason for the lack of discrimination found in these items is unclear.
The K-REC instrument was designed to evaluate the EBP cognitive skills of novice learners (with a focus on entry-level health professionals). It was therefore appropriate that the testing of the instrument compared the two extremes that could be present in entry-level training of health professionals, that is, those with no exposure to formal EBP training versus those with exposure to formal EBP training. While comparisons between novice, intermediate and expert learners would be useful to undertake with the K-REC instrument, it is unknown whether a ceiling effect might exist for the expert group, rendering this instrument inappropriate for this group.
The Sicily Consensus Statement on EBP  highlighted a need for effective training in each of the five steps of EBP, and future research into valid and reliable instruments to evaluate this training. While it seems reasonable in theory to advocate the need for EBP evaluation instruments that cover all of the five steps of EBP, this may be a difficult undertaking within any single instrument. For example, instruments which evaluate cognitive EBP skills such as the Fresno test, and the instrument developed in the current study are not placed to evaluate the application of the evidence into practice (step 4: APPLY), or to promote reflection of individual performance of the entire EBP process (step 5: ASSESS). Rather than aiming to develop instruments which assess all of the five steps of EBP, it may be that instruments are designed for specific purposes, such as the evaluation of EBP theoretical courses (instrument to assess cognitive skills), or the evaluation of EBP in clinical practice (instruments to assess performance based skills and application).
Previous EBP instruments have most commonly contained items evaluating the 'ACQUIRE' and 'APPRAISE' steps of EBP . The ten instruments found in the updated systematic search for the current study (Table 1) were also consistent with this finding. The two instruments that were developed for occupational and physical therapy [20, 21] have closely followed the original Fresno test and thus evaluated cognitive skills pertaining to the first three steps of EBP. While the Sicily Statement  has recommended that instruments cover the five step process model, it is currently not established which core competencies should be included in each of the EBP steps, and whether these competencies are different between professions. For example, while it may be appropriate for a medical doctor involved in clinical research to understand how to calculate number needed to treat, is this a skill that an allied health professional student needs or should be taught during their entry-level training? The question of which competencies should be included for each of the steps in EBP evaluation instruments and whether this differs between professions or levels of training (eg undergraduate or postgraduate) warrants further exploration.
The K-REC instrument captures cognitive EBP skills relating to the first three steps of the five step EBP process model (ASK, ACQUIRE, APPRAISE). The instrument has demonstrated very good reliability, and the validity findings show promise in the application of the instrument for evaluating change at an undergraduate health professional level.
LKL is a Lecturer in the School of Health Sciences at the University of South Australia. She has recently completed a PhD in the area of EBP and entry-level physiotherapy education. Her research interests are in health professional education, and the teaching of EBP process and best research evidence to undergraduate students.
MTW is an Associate Professor in the School of Health Sciences at the University of South Australia. Her research interests included EBP education, with particular reference to educational processes for best research evidence, and the sensation of breathlessness in people with chronic pulmonary conditions.
TSO is a Professor in the School of Health Sciences at the University of South Australia. His research interests include the links between time use and health outcomes, mathematical modelling of sports performance, anthropometry and historical trends in fitness, fatness, food intake and sleep of children.
The authors would like to acknowledge Ms Anna Phillips from the School of Health Sciences, University of South Australia, for her valuable contribution in the comparison of the instrument developed in the current study to the previous instruments developed for the allied health disciplines.
- Sackett D, Straus S, Richardson W, Rosenberg W, Haynes R: Evidence-based Medicine: How to Practice and Teach EBM. 2000, Edinburgh: Churchill Livingstone, secondGoogle Scholar
- Herbert RD, Sherrington C, Maher C, Moseley AM: Evidence-based practice - imperfect but necessary. Physiother Theory Pract. 2001, 17: 201-211. 10.1080/095939801317077650.View ArticleGoogle Scholar
- Chipchase L, Dalton M, Williams M, Scutter S: Is education immune from evidence-based scrutiny?. Aust J Physiother. 2004, 50: 133-135.View ArticleGoogle Scholar
- Chipchase L, Williams M, Robertson V: Factors affecting curriculum content and the integration of evidence-based practice in entry-level physiotherapy programs. J Allied Health. 2007, 36: 17-23.Google Scholar
- Cook DJ, Jaeschke R, Guyatt GH: Critical appraisal of therapeutic interventions in the intensive care unit: human monoclonal antibody treatment in sepsis: Journal Club of the Hamilton Regional Critical Care Group. J Intensive Care Med. 1992, 7: 275-282.View ArticleGoogle Scholar
- Rosenberg W, Donald A: Evidence based medicine: an approach to clinical problem-solving. Br Med J. 1995, 310: 1122-1126.View ArticleGoogle Scholar
- Dawes M, Summerskill W, Glasziou P, Cartabellotta A, Martin J, Hopayian K, Porzsolt F, Burls A, Osborne J: Sicily Statement on Evidence-Based Practice. BMC Med Educ. 2005, 5: 1-10.1186/1472-6920-5-1. [http://www.biomedcentral.com/1472-6920//5/1]View ArticleGoogle Scholar
- Shaneyfelt T, Baum KD, Bell D, Feldstein D, Houston TK, Kaatz S, Whelan C, Green M: Instruments for evaluating education in evidence-based practice: a systematic review. JAMA. 2006, 296: 1116-1127. 10.1001/jama.296.9.1116.View ArticleGoogle Scholar
- Ramos KD, Schafer S, Tracz SM: Validation of the Fresno test of competence in evidence based medicine. Br Med J. 2003, 326: 319-321. 10.1136/bmj.326.7384.319.View ArticleGoogle Scholar
- Fritsche L, Greenhalgh T, Falck-Ytter Y, Neumayer HH, Kunz R: Do short courses in evidence based medicine improve knowledge and skills? Validation of the Berlin Questionnaire and before and after study of courses in evidence based medicine. Br Med J. 2002, 325: 1338-1341. 10.1136/bmj.325.7376.1338.View ArticleGoogle Scholar
- Ilic D: Assessing competency in evidence based practice: strengths and limitations of current tools in practice. BMC Med Educ. 2009, 9: 53-10.1186/1472-6920-9-53. [http://www.biomedcentral.com/1472-6920/9/53]View ArticleGoogle Scholar
- Caspi O, Mcknight P, Kruse L, Cunningham V, Figueredo AJ, Sechrest L: Evidence-based medicine: discrepancy between perceived competence and actual performance among graduating medical students. Med Teach. 2006, 28: 318-325. 10.1080/01421590600624422.View ArticleGoogle Scholar
- Krueger PM: Teaching critical appraisal: a pilot randomized controlled outcomes trial in undergraduate osteopathic medical education. J Am Osteopath Assoc. 2006, 106: 658-662.Google Scholar
- Novack L, Jotkowiiz A, Knyazer B, Novack V: Evidence-based medicine: assessment of knowledge of basic epidemiological and research methods among medical doctors. Postgrad Med J. 2006, 82: 817-822. 10.1136/pgmj.2006.049262.View ArticleGoogle Scholar
- Meyer G, Köpke S, Lenz M, Kasper J, Mühlhauser I: Evidence-based medicine for diabetes educators: a pilot study. Diabet Med. 2007, 24: 901-905. 10.1111/j.1464-5491.2007.02185.x.View ArticleGoogle Scholar
- Shuval K, Shachak A, Linn S, Brezis M, Reis S: Evaluating primary care doctors' evidence-based medicine skills in a busy clinical setting. J Eval Clin Pract. 2007, 13: 576-580. 10.1111/j.1365-2753.2007.00805.x.View ArticleGoogle Scholar
- Siriwardena AN, Sandars J, Gemmell I, Rashid A: Teaching and learning evidence-based medicine: cross-sectional survey investigating knowledge and attitudes of teachers and learners in primary and secondary care. Educ Prim Care. 2007, 18: 45-57.View ArticleGoogle Scholar
- Davis J, Crabb S, Rogers E, Zamora J, Khan K: Computer-based teaching is as good as face to face lecture-based teaching of evidence based medicine: a randomized controlled trial. Med Teach. 2008, 30: 302-307. 10.1080/01421590701784349.View ArticleGoogle Scholar
- Ahmadi-Abhari S, Soltani A, Hosseinpanah F: Knowledge and attitudes of trainee physicians regarding evidence-based medicine: a questionnaire survey in Tehran, Iran. J Eval Clin Pract. 2008, 14: 775-779. 10.1111/j.1365-2753.2008.01073.x.View ArticleGoogle Scholar
- McCluskey A, Bishop B: The adapted Fresno test of competence in evidence-based practice. Journal of Continuing Education in the Health Professions. 2009, 29: 119-126. 10.1002/chp.20021.View ArticleGoogle Scholar
- Tilson JK: Validation of the modified Fresno Test: assessing physical therapists' evidence based practice knowledge and skills. BMC Med Educ. 2010, 10: 38-10.1186/1472-6920-10-38. [http://www.biomedcentral.com/1472-6920/10/38.]View ArticleGoogle Scholar
- Physiotherapy Evidence Database (PEDro). PEDro scale, [http://www.pedro.org.au/english/downloads/pedro-scale/]
- Portney L, Watkins M: Foundations of Clinical Research Applications to Practice. 2000, New Jersey: Prentice-Hall, secondGoogle Scholar
- Glantz SA: Primer of Biostatistics. 1992, New York: McGraw HillGoogle Scholar
- Long K, McEvoy M, Lewis LK, Wiles L, Williams M, Olds T: Entry level training in evidence-based practice - Does it change knowledge and attitudes in physiotherapy students? A longitudinal study. The Internet Journal of Allied Health Sciences and Practice. 2011, 9: 3.Google Scholar
- Australian Physiotherapy Council (APC): Australian Physiotherapy Council Accreditation Requirements 2009. [http://www.physiocouncil.com.au/files/accreditation-manual]
- Occupational Therapy Australia: Australian minimum competency standards for new graduate occupational therapists 2010. [http://www.ausot.com.au/images/Australian%20Competency%20Standards%20Entry%20Level%20Elec%20Copy.pdf]
- Bland JM: Teaching statistics to medical students using problem-based learning: the Australian experience. BMC Med Educ. 2004, 4: 31-10.1186/1472-6920-4-31. [http://www.biomedcentral.com/1472-6920/4/31]View ArticleGoogle Scholar
- General Medical Council: Tomorrow's Doctors: Outcomes and standards for undergraduate medical education 2009. [http://www.gmc-uk.org/education/undergraduate/tomorrows_doctors_2009_outcomes1.asp]
- Dommisse J, Joubert G: Profile of research methodology and statistics training of undergraduate medical students at South African universities. SA Fam Pract. 2009, 51: 158-161.Google Scholar
- Sami W: Biostatistics education for undergraduate medical students. Biomedica. 2010, 26: 80-84.Google Scholar
- OCEBM Levels of Evidence Working Group: The Oxford 2011 Levels of Evidence. Oxford Centre for Evidence-Based Medicine, [http://www.cebm.net/index.aspx?o%20=%205653]
- Gruppen L, Rana G, Arndt T: A controlled comparison study of the efficacy of training medical students in evidence-based medicine literature searching skills. Acad Med. 2005, 80: 940-944. 10.1097/00001888-200510000-00014.View ArticleGoogle Scholar
- Miller G: The assessment of clinical skills/competence/performance. Acad Med. 1990, 65: S63-67. 10.1097/00001888-199009000-00045.View ArticleGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1472-6920/11/77/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.