Psychometric Properties of Calgary Cambridge Guides PDF

International Journal of Medical Education. 2014;5:212-218 ISSN: 2042-6372 DOI: 10.5116/ijme.5454.c665 Psychometric properties of the Calgary Cambridge guides to assess communication skills of undergraduate medical students Anne Simmenroth-Nayda, Stephanie Heinemann, Catharina Nolte, Thomas Fischer...

International Journal of Medical Education. 2014;5:212-218 ISSN: 2042-6372 DOI: 10.5116/ijme.5454.c665 Psychometric properties of the Calgary Cambridge guides to assess communication skills of undergraduate medical students Anne Simmenroth-Nayda, Stephanie Heinemann, Catharina Nolte, Thomas Fischer, Wolfgang Himmel Department of General Practice, Family Medicine, University of Göttingen, Germany Correspondence: Anne Simmenroth-Nayda, Department of General Practice, Family Medicine, University of Göttingen, Humboldtallee 38, 37073 Göttingen, Germany. Email: [email protected] Accepted: November 01, 2014 Abstract Objectives: The aim of this study was to analyse the psy- from the factor analysis represented important constructs of chometric properties of the short version of the Calgary doctor-patient communication The ratings for the first and Cambridge Guides and to decide whether it can be recom- second round of assessing the videos correlated at 0.75 (p < mended for use in the assessment of communications skills 0.0001). Intraclass correlation coefficients for each item in young undergraduate medical students. ranged were moderate and ranged from 0.05 to 0.57. Methods: Using a translated version of the Guide, 30 Conclusions: Reasonable score distributions of most items members from the Department of General Practice rated 5 without ceiling or floor effects as well as a good test-retest videotaped encounters between students and simulated reliability and construct validity recommend the C-CG as patients twice. Item analysis should detect possible floor an instrument for assessing communication skills in under- and/or ceiling effects. The construct validity was investigat- graduate medical students. Some deficiencies in inter-rater ed using exploratory factor analysis. Intra-rater reliability reliability are a clear indication that raters need a thorough was measured in an interval of 3 months, inter-rater relia- instruction before using the C-CG. bility was assessed by the intraclass correlation coefficient. Keywords: Undergraduate medical education, question- Results: The score distribution of the items showed no naires, physician-patient relations, teaching, observer ceiling or floor effects. Four of the five factors extracted variation Introduction Acquiring communicative competence is an important goal nication Skills Assessment Scale (LIV-MAAS), the Liver- of medical education. Especially history-taking, developing pool Communication Skills Assessment Scale (LCAS) and the doctor-patient-relationship, sensitive counselling, the Calgary-Cambridge Guide (C-CG), have become well- shared decision-making and breaking bad news are consid- established in many countries.7-10 ered to be essential skills. Many medical faculties worldwide These instruments were often developed as observation have integrated communication topics in a longitudinal guides for the purposes of delineating evidence-based skills curriculum.1-5 Similar to initiatives in many other countries, and enhancing detailed, descriptive, verbal feedback during the revision of the German Medical Licensure Act in 2004 the teaching and learning process. In addition, they have emphasised the importance of teaching communicative and frequently been adapted to measure performance on social skills in the medical curricula. Such skills should summative exams such as OSCEs and are used to compare already be learned by younger students6 when they begin learner performance before and after a defined teaching their clinical education. term. To measure whether communication skills are success- The instruments differ in form, scope and objectives. fully taught, reliable instruments are needed. Several as- The MAAS-Global Rating List,7 a comprehensive scale, sessment instruments for communicative skills, such as the includes 47 items with a 7-point-scale, divided into 3 Maastricht History-taking and Advice Scoring list consist- sections consisting of items for assessing both communica- ing of global items (MAAS-Global), the Liverpool Commu- tion and clinical examination skills. It was developed and 212 © 2014 Anne Simmenroth-Nayda et al. This is an Open Access article distributed under the terms of the Creative Commons Attribution License which permits unrestrict- ed use of work provided the original work is properly cited. http://creativecommons.org/licenses/by/3.0 validated in Dutch and in English and―after adding 27 dents attend this course in the beginning of their 3rd year. items―converted into the LIV-MAAS Scale8, especially for The instrument British purposes. The Liverpool Communication Skills Assessment Scale (LCSAS)9 is a rather short instrument We chose the C-CG version with 28 items, designed for with 12-items and a 4-point-scale, mainly designed for assessing the history-taking interview.13 This version is has a assessing OSCEs and giving student feedback during 3-point scale (“no”, “yes, but”, “yes”) and is sub-divided into teaching. Other instruments focus on specific patient 6 parts: ‘initiating the session’, ‘gathering information’, groups, such as the Structured Communication Adolescent ‘understanding the patient perspective’, ‘providing a struc- Guide (SCAG)14 for training communication with adoles- ture for the consultation’, ‘building a relationship’, and cents and their parents. These instruments did not meet our ‘closing the session’. needs for assessing a younger student’s communication After consulting Suzanne Kurtz, author of the C-CG, 3 skill, due to their size and scope. In contrast, the Calgary researchers with a good command of English independently Cambridge Guide (C-CG)13 first published in 1996 in translated this version into German (“forward” translation). Canada, seemed to fit for our purposes. Then, a native speaker (SH) translated this preliminary The C-CG was developed for several reasons: first of all, instrument “backward” into English. Two senior lecturers it was the basis for curricular planning and defining teach- (AS, TF), reviewed all translations and developed the pre- ing goals in communication skills. The C-CG covers the final version. If the versions disagreed, they consulted WH whole medical interview and was used as an observation and CN. guide during teaching. It is also used as an assessment tool, The final version was pre-tested with a group of student typically in short versions of the original instrument. In tutors in our department. The raters reported major diffi- 2001, the C-CG became part of the Kalamazoo Consensus culties with the 3-point scale in the original version. They Statement.15,16 This underlines the acceptance of the instru- had the feeling a larger selection of ratings would make ment within an international leading declaration for teach- assessment easier. As a consequence, a 5-point scale (based ing communication skills.17 Especially for a basic skill course upon the typical German grading structure with 1 = excel- which does not include physical examination, the 28-item lent and 5 = deficient) was implemented. version of the C-CG seems appropriate. Although the Preparation of the material original C-CG has already been introduced and validated in From a pool of 117 SP consultation videos that are routinely several translated versions,18 its psychometric properties generated by our “basic medical skills” course, a sample of 5 have not been analysed when used in educational contexts videos was selected to represent the range of the quality of with younger medical students. student performance between “excellent” and “deficient”. The aim of this study was to analyse the psychometric Two authors (AS and TF) screened the video material and properties of the short version of C-CG and to decide selected 5 video consultations which showed a stepwise whether it can be recommended for use in the assessment of grading from excellent to deficient performances. The communications skills in young undergraduate medical videotapes were converted to digitised files on DVD. students. Especially four aspects should be studied in detail: 1. Item distribution, i. e. does the C-CG provide a dif- Participants and training ferentiated assessment? Members from the Institute of General Practice (medical 2. Construct validity, i. e. does the C-CG represent doctors, sociologists, psychologists, and student tutors) meaningful aspects of communication? were asked to take part in the study as raters. The group was 3. Test-retest reliability, i. e. can the C-CG be used reli- trained in a 90-minute session, including a short presenta- ably from semester to semester? tion of the experiment and the C-CG. Afterwards, an 8- 4. Inter-rater reliability, i. e. can the C-CG be used in- minute-video, presenting a consultation between an SP and tuitively by raters? a student of the current course was shown and the group- members carried out an individual rating with the C-CG. Methods These individual ratings were then discussed item per item with the whole group; the aim was a best possible consensus Context about scoring. The training was conducted by AS and TF. At Göttingen University Medical School the “basic clinical After this instruction, all raters received a DVD with the 5 skills course” includes manual skills (e.g. injections, EKG, selected SP-consultations and the C-CG in printed form. wound-suturing) and communication skills (such as They were instructed to score the videos within the follow- history-taking and basic communication techniques). We ing 4 weeks. We reminded them by e-mail and telephone use, among others, role plays and consultations with simu- call. After 3 months, the rating procedure was repeated. lated patients (SP) in small-group learning sessions. The The ethical review board of the University of Göttingen course extends over 12 weeks with 3-hour modules. Stu reviewed and approved the study protocol (No. 27714An). Int J Med Educ. 2014;5:212-218 213 Simmenroth-Nayda et al. Psychometric properties of the Calgary Cambridge guides Statistical analysis rotation. One factor comprised only 1 item (‘negotiates All analyses were performed using SAS 9.3. Several methods agenda’). The four other factors seem to represent im- were applied to assess the psychometric qualities of the C- portant constructs of doctor-patient communication: CG: technicalities of opening and closing a session with a patient, structuring the consultation, formal aspects of Item analysis communication and patient orientation. However, the Mean scores, standard deviations (SD), ranges, and per- number of items of each factor is far from being optimal. centages of the scores given by the raters were calculated to While there were many items loading on factor 1, especially evaluate score distributions, especially to detect possible most or all items of the scale ‘gathering information’ and floor and/or ceiling effects. the scale ‘understanding patient’s perspective’, only three or Construct validity fewer items loaded on factor 3 and 4. The 5-factor solution explained 74.1% of the whole variance (factor 1: 30.9%; The validity of the C-CG construct was investigated by an factor 2: 15.75%; factor 3: 14.6%; factor 4: 7.4%; factor 5: exploratory factor analysis.19 The underlying factors were 5.5%). identified by means of varimax rotation. Test-retest reliability Test-retest reliability Intra-rater reliability was measured within an interval of 3 The raters’ mean total score at the first assessment was 2.37 months. The correlation between the two rating rounds was (SD 0.7, median 2.3, range 1.2 to 4.2). The raters gave assessed with 3 different statistical measures: (1) Pearson’s r, somewhat lower, i. e. better, scores at the second assessment (2) a t-test for dependent samples to analyse whether the (mean 2.26, SD 0.7, median 2.2, range 1.1 to 4.1). The t-test difference between the two assessments was significantly for the difference between the 2 assessments, although different from zero, and (3) a descriptive analysis of how marginal (0.11, 95%CI [0.01, 0.17]), was statistically signifi- often a rater gave the same score at the 2 assessments, how cant (p = 0.023). often the assessments differed by 1 point and how often by 2 The ratings at the first and second rating round corre- points or more. lated at 0.75 (Pearson’s r, p1.0), we were or the second or at both rating rounds: ‘negotiates agenda’, able to extract 5 factors. This solution is shown in Table 2 ‘clarifies patient’s statements’, ‘determines and acknowledg- with the corresponding factor scorings after varimax es patient’s ideas’. 214 Table 1. Item characteristics Item Mean SD Min Max IQR* 10% - 90%** Greets patient 1.7 0.9 1 5 1 1-3 Introduces self and role 2.4 1.3 1 5 2 1-5 Demonstrates respect 1.8 1.0 1 5 1 1-3 Identifies and confirms problems list 2.4 1.2 1 5 2 1-4 Negotiates agenda 3.9 1.2 1 5 2 2-5 Scale “beginning the session” 2.4 0.7 1.0 4.2 1.0 1.5 - 2.5 Encourages patient to tell story 2.0 1.2 1 5 2 1-4 Appropriately moves from open to closed questions 2.4 1.2 1 5 2 1-4 Listens attentively 1.8 1.0 1 5 1 1-3 Facilitates patient’s responses verbally and non-verbally 2.1 1.0 1 5 2 1-4 Uses easily understood questions and comments 1.7 0.9 1 5 1 1-3 Clarifies patient’s statements 2.4 1.1 1 5 2 1-4 Establishes dates 2.0 1.0 1 5 2 1-3 Scale “gathering information” 2.1 0.9 1.0 4.5 1.1 1.0 - 3.6 Determines and acknowledges patient’s ideas re cause 2.7 1.4 1 5 3 1-5 Explores patient’s concerns re problem 2.4 1.2 1 5 2 1-4 Encourages expression of emotions 2.5 1.2 1 5 1 1-4 Picks up/responds to verbal and non-verbal clues 2.6 1.0 1 5 1 1-4 Scale “understanding patient’s perspective” 2.6 2.6 1.0 5.0 1.5 1.3 - 4.3 Summarises at end of a specific line of inquiry 3.1 1.2 1 5 2 1-5 Progresses using transitional statements 2.6 1.1 1 5 1 1-4 Structures logical sequence 2.4 1.0 1 5 1 1-4 Attends to timing 2.1 1.0 1 5 2 1-4 Scale “providing structure to consultation” 2.6 0.9 1.0 5.0 1.2 1.5 - 3.8 Demonstrates appropriate non-verbal behaviour 2.0 1.0 1 5 2 1-3 If reads or writes, doesn’t interfere with dialogue/rapport 2.1 1.1 1 5 2 1-4 Is not judgemental 2.0 1.1 1 5 2 1-4 Empathises with and supports patient 2.3 1.2 1 5 2 1-4 Appears confident 2.1 1.0 1 5 2 1-3 Scale “building relationship” 2.1 0.9 1.0 4.2 1.4 1.0 - 3.4 Encourages patient to discuss any additional points 2.5 1.4 1 5 3 1-5 Closes interview by summarising briefly 3.0 1.3 1 5 2 1-5 Contracts with patient re next steps 1.9 1.0 1 5 1 1-4 Scale “closing the session” 2.5 1.0 1.0 5.0 1.7 1.4 - 3.7 Overall score 2.4 0.7 - - 1.1 1.5 - 3.5 * ** Interquartile range (difference between upper and lower quartile), 10% to 90% interval poor performers. Several items were assessed with high Discussion reliability. For example, the item ‘demonstrates respect’ Reasonable score distributions of most items without ceiling showed only a 10% disagreement by more than 1 point or floor effects as well as a good test-retest reliability and between first and second assessment. The ICCs for this item construct validity recommend the C-CG as an instrument were higher than 0.40 and it clearly differentiated between for assessing communication skills in undergraduate the 5 videos and had a high scoring on the first factor. medical students and for regularly monitoring the success Similar effects were also observed for the items such as of the communication skills curriculum. Some deficiencies ‘greets patient’, ‘introduces self and role’, ‘empathises with in inter-rater reliability are a clear indication that raters and supports patient’ and ‘closes interview by summarising need a thorough instruction before using the C-CG. briefly’. On the basis of this psychometric analysis, we found that the C-CG is able to assess and reproduce the Comparison with literature and meaning of the results main learning goals in this early stage of medical education: Using the C-CG, the raters exploited the range of scores for to build a relationship and to keep in touch with a new nearly all items (Table 1). The wide range between 1 and 5 patient by means of empathic listening and sensitive ques- shows the ability of the C-CG to detect differences and tioning. graduations within single communicative skills. We con- Other items, however, reduced the psychometric quali- clude the C-CG is well-suited to identify good compared to ties of inter-rater and intra-rater reliability as well as con- Int J Med Educ. 2014;5:212-218 215 Simmenroth-Nayda et al. Psychometric properties of the Calgary Cambridge guides struct validity. A quite exceptional item is ‘negotiates to assess this item in particular. Interestingly, neither the agenda’. In 27% of instances, the raters differed from the MAAS-Global, LIV-Maas, LCSAS nor the SEGUE- first to the second rating round by more than 1 point. This Framework contain a comparable item.7-9,21 item also has a fairly low ICC value (Table 2), which means Similar problems, although less distinct, were observed that raters scored the same performance quite differently. In with the item ‘identifies and confirms problem list’ that addition, this item was the only one that scored high on the showed a poor intra-rater reliability, a poor ICC and a fifth factor (Table 3). Scheffer et al. described the same rather low scoring on the first factor. Obviously, the raters problem with this item in their validation of a global rating had also difficulties with the item ‘If reads or writes, doesn’t instrument that they compared with the C-CG. When we interfere with dialogue/rapport’. In more than 30% of simulate a first consultation, we assume that this is a new instances, they differed by more than 1 point between first patient whose reason for coming to see the ‘GP’ is not yet and second assessment. These items measure skills that not known. Therefore, it must have been difficult for the raters so important for younger students who are just beginning Table 2. Factor loadings for the final 5-factor solution Component Items 1 2 3 4 5 Greets patient 0.11 0.25 0.08 0.83 -0.11 Introduces self and role 0.06 -0.08 0.03 0.86 0.25 Demonstrates respect 0.72 0.18 0.43 0.25 -0.15 Identifies and confirms problems list 0.47 0.35 0.37 0.19 0.07 Negotiates agenda -0.10 0.23 0.05 0.09 0.81 Encourages patient to tell story 0.64 0.34 0.29 0.30 -0.15 Appropriately moves from open to closed questions 0.64 0.36 0.43 0.19 -0.07 Listens attentively 0.66 0.24 0.44 0.12 -0.26 Facilitates patient’s responses verbally and non-verbally 0.64 0.27 0.45 0.20 -0.15 Uses easily understood questions and comments 0.65 0.04 0.50 0.10 -0.11 Clarifies patient’s statements 0.74 0.21 0.27 -0.01 0.27 Establishes dates 0.35 0.03 0.74 0.14 0.15 Determines and acknowledges patient’s ideas re cause 0.79 -0.12 0.07 -0.17 0.35 Explores patient’s concerns re problem 0.83 0.37 0.03 0.09 0.00 Encourages expression of emotions 0.86 0.21 0.13 0.08 -0.03 Picks up/responds to verbal and non-verbal clues 0.78 0.30 0.25 0.05 0.11 Summarises at end of a specific line of inquiry 0.31 0.65 0.24 0.07 0.41 Progresses using transitional statements 0.43 0.61 0.46 0.17 0.06 Structures logical sequence 0.29 0.38 0.75 -0.02 0.14 Attends to timing 0.14 0.30 0.71 -0.04 0.01 Demonstrates appropriate non-verbal behaviour 0.63 0.41 0.44 0.11 -0.13 If reads or writes, doesn’t interfere with dialogue/rapport 0.42 0.58 0.38 -0.16 -0.21 Is not judgemental 0.67 0.28 0.42 -0.03 -0.17 Empathises with and supports patient 0.78 0.41 0.19 0.11 -0.07 Appears confident 0.15 0.59 0.42 0.07 0.07 Encourages patient to discuss any additional points 0.45 0.56 0.01 0.40 -0.10 Closes interview by summarising briefly 0.10 0.72 0.16 0.15 0.35 Contracts with patient re next steps 0.46 0.68 0.18 0.00 0.04 Bold values indicates variables with significant scorings of at least 0.47. to learn to build relations with new patients. History-taking helpful while observing the different stages of the encoun- is a first step in this learning process. We often encourage ter. In contrast or additionally, our factor analysis accentu- undergraduate students to avoid writing and reading in ates the different aspects of communication which re-occur order to fully concentrate on the patient’s verbal and throughout the different chronological stages of the en- nonverbal signs. These items, therefore, do not play a major counter. Our first factor comprises those aspects that role in undergraduate medical education, at least in Germa- characterise the typical aspects of patient-oriented behav- ny, and seemed difficult for raters to assess. iour such as patient encouragement, exploring patient’s Kurtz and Silverman13 suggested dividing the C-CG into concerns or demonstrating non-verbal behaviour. The different sections such as ‘beginning’, ‘gathering infor- second factor reflects the ability to structure the communi- mation’, or ‘closing the session’. This structure follows the cation. The third factor focuses on formal aspects of the typical course of a doctor-patient encounter. This is very communication, including dates and timing. The fourth 216 factor considers the technicalities of beginning a session the sample size of five videos was small. For a valid inter- with a patient. In our opinion, this structure is truly valua- rater assessment, it would have been better to ask raters to ble because it reflects the different aspects of communica- evaluate a large number of student consultations which tion behaviour. We assume that the factors identified in our differed only marginally. However, such a procedure would factor analysis may also be valid for other language versions have exceeded our resources. Inter-rater reliability, though of the C-CG. With regards to the one-item factor ‘negotiates not optimal according to our results, may have even been agenda’, there may be cultural differences between commu- overestimated due to the small number of different consul- nication expectations or the way history-taking is integrated tation videos. into the healthcare system which may alter the validity of Implications for practice this item in other cultural settings. The C-CG seems to be an adequate instrument to assess Strengths and limitations skills and abilities that medical younger students should The raters came from a wide variety of backgrounds (stu- learn in communication courses and to assess whether dent tutors, medical doctors, sociologists, psychologists), teachers have successfully taught these skills. But three which reflects the interdisciplinary teaching staff in our caveats are required: medical school. The group or raters was balanced in terms 1. Some items may limit the validity of the instrument if is of gender. This mix of raters helped to assess how the C-CG intended to assess younger students or the quality of performs in real life. Although we analysed the C-CG using communication courses for these students. The item 300 rating assessments (5 videos x 30 raters x 2 time points), ‘negotiating agenda’ proved to be such a case. Table 3. Intraclass Correlation (ICC) First assessment Second assessment Items ICC 95%CI ICC 95%CI Greets patient 0.21 (0.07; 0.71) 0.31 (0.12; 0.80) Introduces self and role 0.25 (0.09; 0.75) 0.28 (0.11; 0.77) Demonstrates respect 0.43 (0.21; 0.87) 0.44 (0.21; 0.87) Identifies and confirms problems list 0.27 (0.10; 0.76) 0.19 (0.06; 0.68) Negotiates agenda 0.10 (0.03; 0.51) 0.15 (0.05; 0.61) Encourages patient to tell story 0.57 (0.31; 0.92) 0.37 (0.17; 0.84) Appropriately moves from open to closed questions 0.42 (0.20; 0.86) 0.30 (0.12; 0.79) Listens attentively 0.52 (0.27; 0.90) 0.38 (0.17; 0.84) Facilitates patient’s responses verbally and non-verbally 0.38 (0.16; 0.84) 0.27 (0.11; 0.77) Uses easily understood questions and comments 0.25 (0.09; 0.74) 0.17 (0.05; 0.65) Clarifies patient’s statements 0.16 (0.05; 0.63) 0.05 (0.00; 0.36) Establishes dates 0.17 (0.05; 0.65) 0.08 (0.02; 0.48) Determines and acknowledges patient’s ideas re cause 0.18 (0.06; 0.67) 0.15 (0.04; 0.62) Explores patient’s concerns re problem 0.27 (0.10; 0.76) 0.19 (0.07; 0.68) Encourages expression of emotions 0.26 (0.10; 0.75) 0.20 (0.07; 0.68) Picks up/responds to verbal and non-verbal clues 0.26 (0.10; 0.75) 0.25 (0.09; 0.75) Summarises at end of a specific line of inquiry 0.25 (0.09; 0.75) 0.19 (0.07; 0.68) Progresses using transitional statements 0.37 (0.16; 0.83) 0.28 (0.11; 0.78) Structures logical sequence 0.34 (0.14; 0.82) 0.24 (0.09; 0.74) Attends to timing 0.40 (0.18; 0.85) 0.16 (0.05; 0.63) Demonstrates appropriate non-verbal behaviour 0.49 (0.25; 0.89) 0.33 (0.14; 0.81) If reads or writes, doesn’t interfere with dialogue/rapport 0.44 (0.21; 0.87) 0.30 (0.12; 0.79) Is not judgemental 0.45 (0.22; 0.88) 0.27 (0.10; 0.77) Empathises with and supports patient 0.41 (0.19; 0.86) 0.37 (0.16; 0.83) Appears confident 0.28 (0.11; 0.78) 0.17 (0.05; 0.63) Encourages patient to discuss any additional points 0.44 (0.21; 0.87) 0.28 (0.11; 0.77) Closes interview by summarising briefly 0.17 (0.05; 0.65) 0.28 (0.11; 0.77) Contracts with patient re next steps 0.30 (0.12; 0.79) 0.35 (0.15; 0.82) We suggest deleting this item when the C-CG is used in it measures typical tasks in doctor-patient encounters, early stages of the medical curriculum where younger which include not only talking with the patient, but also students are beginning to learn their professional role in a structured procedure in diagnostics, treatment, refer- communicating with patients. In later phases of the cur- ral and other clinical activities. riculum, it may be important to include this item since 2. Although teachers usually prefer to sum up similar Int J Med Educ. 2014;5:212-218 217 Simmenroth-Nayda et al. Psychometric properties of the Calgary Cambridge guides items and, thus, to calculate a sum score, it could be patient consultations. misleading to sum up the items of the six scales of the Acknowledgments original version because the items in each scale com- prise different skills and abilities and do not represent We thank all members of the Department of General consistent and coherent concepts, as could be shown in Practice and our GP-teachers for participating in our study. the factor analysis. If teachers and raters are interested Conflict of Interest to learn whether students have a good command of cer- The authors declare that they have no conflict of interest. tain communication skills, e.g. patient-orientation, and to find out a student’s strengths and weaknesses for later References interventions, they should rather look at the items of the 1. Van Dalen J, Bartholomeus P, Kerkhofs E, Lulofs R, Van Thiel J, Rethans factors that we extracted. JJ, et al. Teaching and assessing communication skills in Maastricht: the first 3. Although we trained raters to use the C-CG adequately twenty years. Med Teach. 2001;23(3):245-251. 2. Maguire P, Pitceathly C. Key communication skills and how to acquire and although most of the items of the C-CG seemed to them. BMJ. 2002;325(7366):697-700. be self-explanatory, raters had problems with several 3. Hook KM, Pfeiffer CA. Impact of a new curriculum on medical students’ items. A more thorough training may be appropriate, interpersonal and interviewing skills. Med Educ. 2007;4(2):154-59. especially for those items that are more difficult to as- 4. Schirmer JM, Mauksch L, Lang F, Marvel MK, Zoppi K, Epstein RM, et al. Assessing communication competence: a review of current tools. Fam Med. sess. Since the ICCs for the second assessment were al- 2005;37(3):184-92. most constantly poorer than for the first assessment, it 5. Kurtz SM, Silverman JD. The Calgary-Cambridge referenced observation may also be necessary the repeat the training, or at least guides: an aid to the defining the curriculum and organizing the teaching in communication training programmes. Med Educ. 1996;30(2):83-89. to provide a refreshment. If the C-CG will be later used 6. Simmenroth-Nayda A, Chenot JF, Fischer T, Scherer M, Stanske B, as basis for official grading, a better inter-rater reliability Kochen MM. Mit Laienschauspielern das ärztliche Gespräch trainieren. is important, not least to ensure fairness towards the Dtsch Ärztebl. 2007;104(13):A847-52. students of an entire semester and to avoid that an indi- 7. Van Dalen J, Prince CJ, Scherpbier AJ, Van Der Vleuten CP. Evaluating communication skills. Adv Health Sci Educ Theory Pract. 1998;3(3): vidual’s assessment is dependent on a rater’s personal 187-95. interpretation of the C-CG items. 8. Robinson J, Walley T, Pearson M, Taylor D, Barton S. Measuring consultation skills in primary care in England: evaluation and development Future research of content of the MAAS scale. Br J Gen Pract. 2002;52(484):889-93. Future research should focus on the construction of the C- 9. Humphris GM, Kaney S. Assessing the development of communication skills in undergraduate medical students. Med Educ. 2001;35(3):225-31. CG and try to re-assess and refine the underlying factor 10. Kurtz SM, Silverman J, Benson J, Draper J. Marrying content and structure of this instrument. It may be necessary to create a progress in clinical method teaching: enhancing the Calgary-Cambridge version of the C-CG which is focussed on a subset of items guides. Acad Med. 2003;78(8):802-09. especially relevant for this earlier study phase. Although it 11. Froehmel A, Burger W, Ortwein H. Integration of simulated patients into the study of human medicine in Germany. [in German] Dtsch Med is difficult to find or establish a sort of gold standard, the C- Wochenschr. 2007;132(11):549-54. CG should be validated against such a standard in the future 12. Jünger J, Köllner V. Integration eines Kommunikationstrainings in die to determine construct validity, especially convergent medizinische Lehre. Psychother Psych Med. 2002;53(2):56-64. validity. 13. Kurtz S, Silverman J, Draper J. Teaching and learning communication skills in medicine. 2nd ed. Oxford: Radcliffe, 2005. Conclusion 14. Blake K, Vincent N, Wakefield S, Murphy J, Mann K, Kutcher M. A structured communication adolescent guide (SCAG): assessment of Originally created for use in curricular planning and to reliability and validity. Med Educ. 2005;39(5):482-491. define teaching goals for communication skills, the C-CG 15. Papageorgiou A, Miles S, Fromage M, Kemmy J, Leinster SJ. Cross sectional evaluation of a longitudinal consultation skills course at the new short version can also be recommended for evaluation UK medical school. BMC Med Educ. 2011;11:55. purposes. A student’s communications skills can be reliably 16. Makoul G. Essential elements of communication in medical encounters: assessed with the C-CG. In addition, teachers can be regu- the Kalamazoo consensus statement. Acad Medi. 2001;76(4):390-393. 17. Lang F, McCord R, Harvill L, Anderson DS. Communication assessment larly informed whether they have reached their training using the common ground instrument: psychometric properties. Fam Med. goals and whether they have become better or worse, 2004;36(3):189-98. compared to the previous semester. However, it is of 18. Scheffer S, Muehlinghaus I, Froehmel A, Ortwein H. Assessing stu- upmost importance that raters be well-trained in the use of dents´communication skills: validation of a global rating. Adv Health Sci Educ Theory Pract. 2008;13(5):583-92. the instrument for results to be reliable. Our factor analysis 19. Consulting group of the division of statistics and scientific computing at indicated four separate latent concepts: patient-orientation, the University of Texas/Austin. SAS library: factor analysis using SAS PROC communication structure, formal aspects and technicalities FACTOR. [cited 12 September 2014]; available from: http://www.ats.ucla. of beginning a session with a patient. These concepts edu/stat/sas/library/factor_ut.htm. 20. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater represent important features of the medical encounter and reliability. Psychol Bull. 1979; 86(2): 420-428. are relevant even for undergraduate students, just beginning 21. Makoul G. The SEGUE framework for teaching and assessing commu- to learn the basic communication skills involved in doctor- nication skills. Patient Educ Couns. 2001;45(1):23-34. 218