SKILLS IN CLINICAL COMMUNICATION: ARE WE CORRECTLY ASSESSING THEM AT UNDERGRADUATE LEVEL?

Traditonal learning and assessment systems are overwhelmed when it comes to addressing the complex and mult-dimensional problems of clinical communicaton and professional practce. This paper shows results of a training program in clinical communicaton under Problem Based Learning (PBL) methodology and correlaton between student self-assessment and teachers assessment. This involves a qualitatve-quanttatve cross-sectonal study in usual practce in the 2nd year of the degree in Medicine. Teaching methodology is PBL, including 15 associate professors and 90 students. Educatonal tools for learning: PBL cases and seminars (video recorded, theoretcal-practcal lectures). Assessment tools: Tutorials on those cases worked on PBL (40%), knowledge test (30%), assessment of a case with PBL methodology (20%) and video recording report (10%). Communicaton skills are evidenced by CICCA-D scale (Connect-Understand-IdentfyAgree-Help-Decision). Variables: academic performance, score on CICCA-D and academic methodological assessment. The analysis is carried out using descriptve statstcs, calculatng the intra-class correlaton coefcients and weighted Kappa index with quadratc weights. 92.2% of students passed the course on the frst round. In a range between 0 and 34 points students' self-assessment scored 13 (SD ± 5) points and teachers' 16 (SD ±7). A weak (21% 41%) or poor (< 20%) correlaton was obtained between teachers and students for all questons on CICCA-D. The authors suggest a summatve assessment using diferent instruments and techniques to assess clinical communicaton skills from the frst year onwards, and highlight the key role of self-assessment, peer assessment and the use of video recording techniques along with feedback in formatve assessment.


INTRODUCTION
Communicaton is an essental component of the skill required from medical professionals. Communicatng with the patent in clinical practce refers to the way in which the doctor and the patent interact both verbally and nonverbally in order to achieve a shared understanding of problems and solutons. Basic communicatve tasks in a clinical setng could be summarised as follows: empathising with the patent and family, defning health problems, agreeing on the decisions to be made and the actons to be taken in order to address their health problems, helping the patent and their families how to understand, make choices and act at all tmes. Good communicaton in the doctor-patent relatonship is associated with beter clinical outcomes, increased patent and professional satsfacton and, ultmately, good professional practce (Dwamena et al., 2012;Cannarella Lorenzet, Jacques, Donovan, Cotrell and Buck, 2013;Fawole et al., 2013;Schofeld, Green & Creed, 2008;Street, Makoul, Arora, & Epstein, 2009;General Medical Council Tomorrow's Doctors, 2001;Prat et al., 2004). Clinical communicaton skills as such are likely to be taught, learned and assessed (Cleries, 2010).
Clinical communicaton has been considered as one of the essental skills to be developed by doctors for the last quarter of a century (Brown, 2008) and this has been introduced sporadically in a number of university educaton programs over the last 20 years. However, the development and implementaton of the European Higher Educaton Area (EHEA) has presented an opportunity in relaton to the need to include communicaton aspects in medical degree training programs (Cleries, 2010;Michaud, 2012;Kiessling & Langewitz, 2013). An ofcial recommendaton has been in place in Spain since 2008, highlightng the importance of incorporatng clinical communicaton content into the development of medicine degree curricula (Order-Ministry of Educaton and Science/332/2008). Only 15 out of the 32 Facultes of Medicine in Spain include training in clinical communicaton on the curriculum, while there is also no objectve set for standardised assessment teaching methodology. A European consensus for teaching clinical communicaton to health professionals was recently published in an atempt to highlight its importance as a clinical skill and to avoid variability in its teaching (Bachmann et al., 2013).
The best strategies for the learning of clinical communicaton seem to be those that include role playing (with and without simulated patents), teacher feedback with videotaping of consultatons (with and without simulated patents) and discussion in small groups (Bachmann, et al., 2011;Ruiz -Moral, 2003;Moore, Gómez & Kurtz, 2012;Deveugele, Derese, De Maesschalck, Willems, Van Driel & De Maeseneer, 2005). The teaching community has the commitment and the challenge of obtaining evidence of how students develop clinical skills that are not measurable as a simple sum of knowledge, skills and attudes. The student must show "what he/she knows" (basic knowledge of clinical communicaton theory), "that he/she knows how to" (applied knowledge), "that he/she shows how" ("in vitro" with simulated patents or the Clinical Skills Laboratory) and fnally "what he/she does" (clinical skills "in vivo" with patents and real situatons). Another very relevant aspect is the inclusion of formatve assessment actvites ("feedback") as a means of guiding and enhancing learning. The characteristcs of the clinical skills assessments and the "medical professionalism" should be those that are required for any assessment: validity, reliability, transparency, acceptability, feasibility and having educatonal impact. In this regard, we have designed, validated and implemented various instruments that reveal, among other things, the skills acquired in clinical communicaton: portolios (Figueras & Martnez Carretero, 2006), objectve and structured clinical evaluaton (OSCE) (Toledo García, Fernández Ortega, Trejo Mejía, Grijalva & Gómez Clavelina, 2002;Kronfy, Ricarte, Juncosa & Martnez-Carretero, 2007), direct observaton of practce (with real or simulated patents), analysis of video recordings (Baribeau, Mukovozov, Sabljic, Eva & Delotnville, 2012), evaluatve scales and checklist (Cleries 2010;Gavilán, Ruiz-Moral, Pérula de Torres & Parras Rejano, 2010;Ruiz-Moral, Prados Castllejo, Alba Jurado, Bellón Saameño & Pérula de Torres, 2001). One of the challenges faced by the teaching community is determining which or what combinaton of these instruments allows us to efectvely assess the degree of communicaton skills acquired in future medical professionals at each stage of learning.
The aim of this paper is to share the innovatve teaching experience in teaching and assessment of communicaton skills and clinical interviewing in medical degree courses at the University of Girona (UdG) among the teaching community, demonstratng the learning system learning-assessment design and results of the same.

METHOD DESCRIPTION
The learning methodology used and the results of student assessment of Clinical Communicaton Module of 2nd year Medical Degree at the UDG, Catalonia, Spain (year 2011 are presented. This involves a qualitatve and quanttatve transversal descriptve study under normal practce conditons.

Subjects
The communicaton skills and clinical interview module at the University of Girona is taught in the 2nd year of the degree in Medicine with a study load of 6 ECTS (European Credit Transfer and Accumulaton System credits) Vol. 4(2), 2014, pp 90 per student and 24 credits for teaching and research staf (PDI in Spanish), which are shared among 15 associate professors who perform the work of facilitator tutor. There are between 90 and 130 students in each year. The study load is spread out over four weeks. The methodology used is Problem Based Learning (PBL) (Branda, 2009). The study load is taught over four weeks. Analysis of 90 students were included (N: 90). During the study there were no losses to follow up.

Learning and Assessment System 2.2.1 The learning educatonal instruments used are:
• PBL Cases: Cases will be worked on using the PBL methodology in groups of 10 students during three 2-hour sessions. A total of 4 cases with diferent communicaton scenarios will be worked on during the year. Each PBL case has defned learning objectves in relaton to the skills students should develop during the module.
• Video recorded lectures: Each student is flmed in a clinical setng recreated in the Clinical Skills Centre, where they are presented with a clinical interview with simulated patents. Each student makes a critcal analysis of both the positve points and those parts of his/her interventon that could be improved. A later session is carried out with the tutor and the PBL group in which each student can voluntarily analyse his/her interview and carry out a feedback session. Subsequently, students are ofered the opton of personalised feedback for those who did not partcipate in this in the group.
• Theory-practcal lectures: Viewing video recordings, role-playing and feedback group sessions.

The assessment educatonal instruments used are:
The assessment of acquired skills forms part of the learning process itself and consists of the following: • Tutorials on those cases worked on using PBL methodology: The PBL group, the students themselves and the tutor evaluate the learning skills, communicaton skills, responsibility of teamwork and interprofessional relatonships. The assessment is consists of a series of 20 items evaluated in a Likert scale from 0 to 5 points. As a result, each student manages to gain perspectve from the self-assessment, peer assessment and assessment received from the tutor.
• Skills exam: Afer viewing a video recording, a skills development test is carried out using short questons.
• Assessment of a case with PBL methodology: A case is presented via a video recording. The student must choose two topics, justfying the study relatng it to the objectves content and the case. The following day, the student is asked questons about those areas of interest selected.
• Video recording report: Students draf a self-evaluatve report of the communicatve aspects of the clinical interview performed in a clinical simulaton setng during which, they had to conduct an interview with a simulated patent.
The fnal assessment was obtained from the sum of (I+II+III+IV): Contnuous evaluaton of PBL Cases (40%), exam test afer viewing a video recording (30%), assessment of a case with PBL methodology (20%) and a selfevaluatve report of a video recording report (10%). In order to pass, the student must obtain a pass mark in each of the 4 assessment tests performed, with the opton of retaking each of the four tasks proposed for the assessment if the minimum grade required is not frst achieved.

Test used for research purposes
In order to assess the use of a communicaton skills evaluatve questonnaire in our context, students and tutors were voluntarily invited to use the CICCA-D scale (Connect-Understand-Identfy-Agree-Help-decision) when assessing the video recording. The CICCA-D scale comprises 17 items and consists of a tool focused on the assessment of the patent's partcipaton in the decision-making process (Gavilán et al., 2010;Moral & Pérula, 2006). The CICCA-D is based on the patent-centred interview model. The 17 items of the scale are grouped into three components: • Component 1-IDENTIFYING AND UNDERSTANDING THE PROBLEMS • Component 2-AGREEING AND HELPING TO ACT Vol. 4(2), 2014, pp 91

• Component 3-DECISIONS WITH OPTIONS
Each item is assigned a value of between 0 (no presence of the item in the video recording) and 2 (intense or consolidated presence).
Students and tutors are voluntarily invited to use the CICCA-D scale in the feedback training session of the video recording that they will later use for this work. This test was for research purposes only rather than being considered for the summatve assessment of the students and this was explicitly explained to students and teachers alike.
Scores and reports contributed by the teachers and students during the PBL tutorials were used to assess the level of satsfacton with the teaching methodology throughout the 4-week module.

Variables and measurements:
• Academic performance: Percentage of passes and scores obtained • Academic methodology assessment: the teachers and coordinators of the module analysed the conclusions exposed at the end of module assessment meetngs.
• Communicatonal and clinical relatonship aspects: Score obtained in the CICCA-D scale

Statstcal analysis
A database was built in ACCESS-Microsof for the processing of data from the CICCA-D survey and questonnaires were recorded by a research assistant.
The analysis is carried out using descriptve statstcs, calculatng the intra-class correlaton coefcients and weighted kappa index with quadratc weights. The Stata / SE Version 12.1 I.T. program was used (StataCorp, Collage Staton, TX, USA).

Ethical aspects
The confdentality of personal data was respected during the handling of all the material and verbal consent was sought from students, teachers and simulated patents to be used for research purposes. The analytcal processing of the results of the CICCA-D questonnaires was carried out on an anonymised basis, making it impossible to relate the answers with students who provided them. The video recordings used were destroyed once the study was completed.

EXPERIMENTAL DATES AND RESULTS
90 students enrolled and completed the full module. The percentage of passes obtained in the frst round (2011-2012 academic year) was 92.2% (83 students) (Table 1). A systematc and literal transcripton of the scores and comments from the tutoring records was carried out to evaluate the academic methodology (Table 2).

Student opinions Tutor opinions
Uncertainty in terms of self-learning management and the development of minimum skills.
The difculty in objectve assessment with the Likert scales from the PBL tutorials stands out.
Difculty in addressing the PBL exam, especially in the justfcaton secton of topics to be developed.
Positve assessment of the formatve nature of self-assessment.
The CICCA-D questonnaire is perceived to be of litle use in the frst few academic years.
The evaluator model used requires a signifcant amount of organizatonal structure.
The assessment model used implies a greater amount of tme used.
The assessment model used requires prior training of teachers.
Difculty performing summatve assessment of intangible skills.
Ra ise s t he ne e d for a sp e c ifc c lin ica l communicaton scale for undergraduate level.

Table 2. Qualitatve assessment of students and tutors and teaching and evaluatve methodology employed
We conducted a narratve analysis of the informaton and the results were discussed with the entre research team. 49 student self-assessment questonnaires (54.4%) and 57 teachers assessment (63.3%) were recovered in terms of the CICCA-D questonnaire. In a range between 0 and 34 points, the student self-assessments registered mean of 13 (SD ± 5) points, while the assessments carried out by the tutors showed a mean of 16 (SD ± 7) points. The agreement between students and teachers could only be measured in the 47 evaluatons that were available from both evaluatons. A weak (21% -41%) or poor (<20%) correlaton was obtained for all the questons. No correlaton between teachers and students was found in 4 items (Table 3). Table 4 shows the best and worst aspects assessed by students and teachers.  In this study, it is suggested a dynamic assessment deals with summatve and formatve aspects to assess competence in clinical communicaton skills in the early years of the medical degree. This study shows how the combined applicaton of diferent assessment instruments (PBL tutorials assessment, PBL exam, short answers test and video recording assessment with simulated patents) could be a feasible combinaton for the assessment of skills in clinical communicaton in students enrolled in the 2nd year of medicine. The limitatons detected are the requirement for tutors to have prior training, the need for a signifcant organizatonal structure, the requirement for major involvement of teachers and the difculty in carrying out a summatve assessment of intangible skills. It highlights the role of self-assessment, peer assessment and feedback from the tutor during formatve assessment. Despite the formatve value of self-assessment using specifc assessment scales, its inclusion in summatve assessment was ruled out.

STUDENT-TEACHER CORRELATION (Score= 47)
Authors of this work agree with previous studies that noted the need for the simultaneous use of diferent instruments to assess "the knowledge", "the know-how," the demonstraton "how to" "and the do" in clinica communicaton skills (Borrell-Carrió, Clèries, Paredes-Zapata, Borrás-Andrés, Sans-Corrales & Mascort-Roca, 2012; Kiessling & Langewitz, 2013, Street & Hanneke, 2013, Berkhof, van Rijssen, Schellart, Anema & van der Beek, 2011. The challenge involves defning the combinaton of more efectve assessment instruments for each stage of learning. In the European Consensus on learning of Clinical Communicaton recently published there are a list of embraces more of an individual perspectve focusing on what skills the individual student should learn (Bachmann et al., 2013). The authors believe that the combinaton of assessment tools suggested to allow evaluate individual clinical communicaton skills in the early years of the medical degree. In the review made, authors have not found any work with this combinaton of assessment elements.
The PBL learning method is efectve in that it is student-centred, has a constructvist approach, it allows the development of generic skills and facilitates the development of an integrated curriculum. The PBL method has the added advantage to allow students became aware of their mistakes especially in areas of communicaton and knowledge sharing. The PBL method requires a certain level of prior training by teachers and, can generate some uncertainty among students above all in the early stages, in relaton to the learning objectves and how they will be assessed. Another difculty with this teaching method for tutors is the development of a valid, reliable and objectve summatve assessment of the work developed and knowledge acquired (Branda, 2009;Gavilán et al., 2010, Schmidt, Rotgans & Yew, 2011. In this work, a Likert scale is used in the PBL tutorial assessment by the students themselves, classmates and the tutor to assess aspects such as learning skills, group communicaton, responsibility and interpersonal relatonships (ability to make constructve critcism, cooperatve behaviour and collaboratve work). The positve aspects of this assessment system is how it's equally fosters self-assessment and peer-assessment and facilitates contnuous assessment by the tutor. Contnuous assessment is associated with a learning efort distributed in tme and more in-depth learning and greater motvaton (Delgado & Oliver, 2006). The authors of this study emphasize the crucial role of these techniques in formatve assessment. The assessment should go beyond the mere reproducton of knowledge and focus on the student's ability to meet new challenges and learning tasks: problem solving, constructon of meaning and the development of self-learning strategies. This approach touches on the PBL exam but the problem is found in the development and validaton of systems that allow the objectve scoring and integraton aspects that are ofen intangible and difcult to evaluate in summatve assessment, such as cooperatve work, for example. In this work, both students and tutors indicate the difculty of carrying out a single summatve assessment from PBL group work. In this work, the value assigned to the contnuous assessment of PBL is 40% and this value is the same for all modules and all medical degree. Another issue to be resolved is the value that the assessment of the PBL tutorials should have in the overall assessment. Another difculty raised by tutors in this work is the enormous consumpton of tme needed for this method of assessment. This percepton has already been noted by other authors (Tai & Yuen, 2007).
In order to assess theoretcal skills ("knowledge") in communicaton skills, the authors of this work propose the use of short questons once the video recording has been viewed, together with a PBL exam. This system is fexible and open but the problem lies in the fact that these types of questons are difcult to develop and score in a reliable manner. The authors propose peer correcton and the use of correcton series to increase the level of reliability, unfortunately at the cost of increased tme spent working by teachers (Carreras-Barnés, 2009 used by the authors can assess aspects such as creatvity, capacity to search for and analyse informaton and a capacity for synthesis, skills hat have had to be worked during module development. However, their development also requires a high level of involvement by teachers while peer correcton and correcton on consensual templates between all tutors is equally important to increase its reliability (Norman & Schmidt, 2000). The use of the video recording of the student in a simulated situaton can help assess a higher level of skills, "applicaton knowledge" and demonstrates in vitro ("demonstratng how") their knowledge of clinical communicaton. This technique plays a major role in formatve assessment partcularly if complemented by selfassessment, peer assessment and feedback from the tutor (Jamtvedt, Young, Kritofersen, O´Brien & Oxman, 2006). Its impact on summatve assessment should be increasing as the student advances in the study of medicine and, above all, when he/she frst makes contact with real clinical situatons (Orientale et al., 2008).
The use of validated scales for the assessment of skills in clinical communicaton is a feld of great interest because they allow the student and the tutor to individually detect learning aspects for improvement and plan a personalized learning. The fundamental problem involved in applying it to undergraduate students is that most of these scales are only validated at a post-graduate level. Another signifcant difculty in the use of assessment scales is that they require prior training of both teachers and the students, which in turn requires signifcant tme consumpton for its correct applicaton. CICCA-D is a validated test in the feld at graduate level and is focused on promotng patent partcipaton in decision-making (Gavilán et al., 2010). In our study, CICCA-D has been applied to students who have made frst clinical contact. Our results indicate that by using selfefcacy scores, students studying the second year of Medicine underestmate their communicatons skills with simulated patents. These results are consistent with other published studies using self-assessment instruments in comparison with external scores (Lipset, Harris & Downing, 2011;Lundquist, Shogbon, Momary & Rogers 2013, Ammentorp, Thomsen, Jarb, Holst, Holm Øvrehus & Kofoed, 2013. We believe that it cannot be used as a tool for summatve assessments for testng the individual students as self-efcacy assessment. One issue to be resolved is how to involve students in the design and validaton of instruments to assess clinical skills.
All these arguments suggest that the assessment combinaton proposed by the authors in this work may have important educatonal efects. The academic data and the results of opinion surveys to students and teachers are consistent with this statement.
From a functonal perspectve, efectve communicaton is not just what an individual does, but what interactons achieve. Successful communicaton may difer from one person to another, depending on one´s perspectve and situaton (Street, Makoul, Arora & Epstein, 2009). This could be problematc if the evaluaton of the clinical communicaton skill depends primarily on a checklist of demonstrated behaviours (Mazor et al., 2005). The present work was carried out in student of second course of medical degree without contact with patents therefore it was not be able to assess the real impact of clinical communicaton.
Thus, in response to the ttle of the study, the authors propose that the combinaton of assessments tools is very useful in assessing clinical skills in students of the frst years of de degree of Medicine but insufcient if we want to assess the efects of clinical communicaton. In later years of medical studies involving real contact with clinical practce, it is already possible to assess whether the student communicates efectvely in a real environment (the "doing") and in more complex situatons, such as the delivery of bad news (Schildmann, Kupfer, Burchardi & Volmann, 2012).
The most important limitaton of our study design was the non-consttuton of comparison groups. For practcal reasons, we were unable to randomly assign students from the same year to an interventon and a control group.

CONCLUSIONS
To conclude this study, the authors have suggested a dynamic assessment deals with summatve and formatve aspects to assess competence in clinical communicaton skills in the early years of the medical degree. The authors have suggested an efectve summatve assessment using diferent instruments (Contnuous evaluaton of PBL Cases, test exam afer viewing a video recording, assessment of a case with PBL methodology and a selfevaluatve report of a video recording report with simulated patents).. They authors believe that CICCA-D test cannot be used as a tool for summatve assessments for testng the individual students as self-efcacy assessment. The authors highlight the key role of self-assessment, peer assessment and the use of video recording techniques along with feedback in formatve assessment. With this methodology we can measure the degree of competton in clinical communicaton skills but not its efects. All in all, implementaton of authentc Vol. 4(2), 2014, pp 96 assessment strategies is seen as a tedious process to evaluate students' learning, so a more efcient assessment strategy is needed. To evaluate the efects of efectve clinical communicaton, especially in senior students, is necessary to design and validate assessments system that involved real patents and clinical situatons.

TEACHING INVOLVEMENT
Based on this work, the authors present some recommendatons for the assessment of clinical communicaton skills during the frst years of degree of Medicine. Assessment should be "fanned out" with the use of diferent instruments in a "spiral" efect, where there is increasing difculty with increasing contact with the overall and clinical practce, with the involvement of all stakeholders involved. Self-assessment, peer assessment and assessment by the tutor along with feedback techniques are essental in the formatve evaluaton. It is necessary to train teachers in learning and assessment of clinical communicaton skills.
In the clinical setng, assessment by colleagues and other professionals who share care work with students such as colleagues, nurses, doctors or medical assistants (360 º assessment) can be of enormous educatonal value (Quest, Ander & Ratclif, 2006). Its inclusion in the summatve assessment requires the design and validaton of assessment scales to decrease variability and to increase valuaton objectvity (Norcini & Burch 2007). A point of partcular interest is the inclusion of the assessment of communicaton in the doctor-patent relatonship by patents using validated scales (Ruiz- Moral, Perual de Torres & Ramillo Martn, 2007). Another aspect to consider in the assessment of clinical communicaton and other generic skills is the role to be played Medical Educaton Units (Rugiero et al., 2010).
Research lines proposed are the determinaton of which combinaton of assessment instruments is right and what value each of the instruments should have in the summatve assessment as a whole through randomised studies to evaluate clinical communicaton skills in undergraduate students. Another proposal would be to enhance the development of validaton studies of assessment scales in clinical communicaton skills at undergraduate level that would be able to measure the impact of these educatonal interventons.