Annals of Oncology 2: 9-12, 1991. O 1991 Kluwer Academic Publishers. Primed in the Netherlands.
Special article Results of the first ESMO examination in medical oncology, London 1989 P. Alberto,1 B. Mermillod,2 P. G. Cocconi,3 M. Schneider,4 S. Seeber,5 D. J. T. Wagener,6 J. M. A. Whitehouse7 & H. Cortes-Funes8 ESMO Educational Committee (P.A. chairman); 'Division of Onco-Hematology, University Hospital, Geneva, Switzerland; 2Unit of Biostatistics, University Hospital, Geneva, Switzerland; }Sen>. Oncologia Medica, Ospedale Regionale, Parma, Italy; ^Centre Antoine Lacassagne, Nice, France; 5 Oncology-Hematology, Stadt Klinik Leverkuscn, Germany; ''Division of Medical Oncology, University Hospital, Nijmegen, The Netherlands; 'Medical Oncology Unit, Southampton General Hospital, Southampton, U.K.; *Serv. de Oncologia Medica, Hosp. '12 de Octubre] Madrid, Spain
Summary. Since 1989, a Certification in Medical Oncology is offered to ESMO members on the basis of their professional curriculum vitae and of their scores in a multiple choice examination. The first session took place in London, U.K., in September 1989, during ECCO 5. One hundred and twenty-five ESMO members were evaluated by means of 60 multiple choice questions covering various aspects of medical oncology, such as tumor diagnosis, prognosis and treatment, drug pharmacology and toxicity, histology and cytology, epidemiology, carcinogenesis and tumor biology. The mean percent of correct answers was 77.4. The best results were obtained with questions dealing with chemotherapy. Scores of 40% or less were obtained in 8 questions, including 2 questions on AIDS related tumors. A similar examination is available in 1990 and will be prepared yearly in the future. It is hoped that this European ESMO Certification will contribute to lessen the professional discrepancies between oncologists of European countries and improve the level of oncological training in Europe. Key words: ESMO certification, educational program, multiple choice examination Introduction
Material and methods
In 1987, the ESMO Educational Committee was appointed by the ESMO Board to evaluate the feasibility of a European Certification in Medical Oncology. For this purpose, the Educational Committee was also due to prepare a general evaluation of the training facilities, level of medical education in oncology, and characteristics of professional activities of medical oncologists in various European countries. The Educational Committee presented a report at the ESMO General Assembly in November 1987. The conclusions of this report were 1. that important differences exist among medical oncologists in Europe, either concerning pre- and post-graduate training, or regarding the recognition of medical oncology as a speciality of internal medicine, either in hospitals or institutes, or in private medical practices, and 2. that a Certification in Medical Oncology organized by ESMO and offered to its members in the European countries, could contribute to equalize the level of knowledge among oncologists in Europe. A project was presented for a European Examination, using the technique of multiplechoice questions. This project was accepted by the general assembly, and the first session of the ESMO Examination in Medical Oncology was scheduled for September 1989 during ECCO 5 in London. The results of the first examination are presented in this paper.
Only ESMO members were eligible for the examination. In addition, it was stipulated that all applicants should work on a full time basis as medical oncologist, either in a hospital or institute, or on a private basis, and that they should have had a previous training in internal medicine and medical oncology for at least six years. Before being registered by the Educational Committee, applications had to be reviewed by the National ESMO Representatives. Hundred eighty applicants were registered. Hundred twenty-five of them attended the examination. 60 multiple-choice questions were prepared, including 30 questions type A, 8 questions type B, 10 questions type E, and 12 questions type K'. It was assumed that many senior oncologists would like to take the examination and that they might be less familiar with the technique of multiple-choice questions than their younger colleagues. For this reason, a detailed explanation concerning the examination was published in the European Cancer News well before the examination, with examples of questions types A, B, E and K' (see addendum 1). The 60 questions were generated in order to cover as homogeneously as possible the various aspects of medical oncology. There were 8 questions on tumor biology and laboratory tests, 4 questions on histology and cytology, 8 questions on pharmacology, 19 questions on clinical therapeutics and toxicities, 14 questions on clinical diagnosis and
10 Table 3. General results and results to questions A, B, E and K'.
Table 1. Number of questions dealing with various tumor types. Lymphoma, leukemia, myeloma Urology and testicular tumors Breast cancer Skin, soft tissues, bone Digestive tract Lung, pleura, head and neck Gynecology Children tumors AIDS related tumors Miscellaneous
14 12 10 10 9 7 5 4 2 7
Number of questions Mean score (percent) Mean score (no. of points) (standard deviation) Worst score (no. of points) Best score (no. of points)
Total
A
B
E
59 77.4
29 81.9
8 87.3
10
12
67.0
68.8
45.7
23.7
5.0 32
2.4 18
6.7 1.6 3
8.3 1.3 5.5
53.5
29
7.0 1.0 2.5 8.0
10.0
11.0
K'
and the results to questions A, B, E and K' are shown in Table 3. The overall mean score of correct answer is 77.4%. The best mean result was observed for questions B, and the worse mean percent for question E. For all types of questions, the score was superior to 66%. The performance of candidates in the various fields of medical oncology are indicated in Table 4 and Fig. 1. The best scores were obtained with questions dealing with cancer treatment, in particular chemotherapy and drug pharmacology, with response rates of 88 and 87%. Lower rates, about 70%, were recorded for clinical diagnosis and prognosis, laboratory data interpretation, and basic questions on tumor biology, epidemiology and carcinogenesis. The poorest responses were given to questions dealing with histology and cytology. The analysis of the results for each individual question showed that the percentage of correct answers ranged from 100% (in one question only), and 10.4% (in one question). 14 questions (11 type A, 3 type B) had a percentage of correct answers between 95 and 100%. 3 questions Results were correctly answered by less than 25% of the parThe distribution of the 125 participants by age and ticipants, all belonging to type K'. Overall, the percencountry is shown in Table 2. There were 18 women and tage of correct answers equal or superior to 50% was 107 men coming from fourteen countries. The mean obtained in 48 out of 59 questions, or 81%. A short age of participants was 40.8 years, slightly younger for description of the 8 questions with response rate of women (38.0) than for men (41.3). The overall results 40% and lower is given in Table 5. Two of these questions concern AIDS related tumors and 3 focused on hematological malignancies. prognosis and 7 questions on epidemiology and carcinogenesis. As shown in Table 1, many different tumor types were used in the questions, including hematological malignancies, pediatric tumors, various solid tumors, and AIDS related tumors. The mean level of difficulty was median, in order that the examination would be within the reach of all well trained medical oncologists without special preparation. The allotted time was 90 minutes, and the questions were distributed to participants at the beginning of the examination session. All data were introduced in a computer program for complete analysis. Each correct answer counted for one point. In addition, a half point was given to questions B and K' when one error only was made in the four or five possible answers. One question (on epidemiology) seemed inadequately formulated and was withdrawn, leaving a maximum of 59 points.
Table 2. Distribution of participants by age, sex and country. Country Austria Belgium Denmark Germany France Ireland Israel Italy Netherlands Portugal Spain Switzerland Tunisia United Kingdom Total
Participants 2 9 5 23 9 1 1 15 5
8 33 9 1 4 125
Male
Female
Mean age (years)
5 30 7
3 3 2
1 4
0
0
36.5 40.3 43.2 43.7 41.6 (42) (47) 38.1 47.2 38.9 38.4 41.0 (34) 43.0
107
18
40.8
2
8 5 20 8 1 1 10 5
0 1
0 3 1
0 0 5 0
Discussion As of December 1989, 973 oncologists coming from 17 countries were ESMO members. This first Examination in Medical Oncology attracted 125 ESMO members, corresponding to 13%, coming from 14 nations. This clearly demonstrates a very large interest in a professional qualification in Medical Oncology at the European level. An interesting observation is that an approximately equal number of oncologists older or younger than 40 years took the examination. This shows that senior oncologists, despite having already an established professional status, have also been attracted by the possibility of a European Certification in Medical Oncology. Also interesting is the fact that the mean scores of junior and senior oncologists were
11 Table 4. Results in subgroups.
Number of questions Mean score (percent) Mean score (no. of points) (standard deviation) Worst score (no. of points) Best score (no. of points)
Total
a1
b2
c3
d4
e3
f6
59 77.4
19 88.1
14 68.5
8 86.8
8 74.0
6 70.9
4 56.2
45.7 5.0 32 53.5
16.7 1.7 11.5 19
9.6 1.8 4.5 14
6.9 1.1 4 8
5.9 1.2 2.5 8
4.3 0.9 2 6
2.2 0.9 1 4
'a: Clinical therapeutics and toxicity. b: Clinical diagnosis and prognosis. 3 c: Drug pharmacology. 4 d: Tumor biology and laboratory tests. 5 e: Tumor epidemiology and carcinogenesis. 6 f: Tumor histology and cytology. 2
nearly identical. Although the number of multiplechoice questions was relatively small, it was sufficient for a practical assessment of the results. The overall results of the examination are satisfactory with 77.4% of correct answers. The best individual answers were registered for questions dealing with cancer treatment, particularly chemotherapy. This reflects the predominance of cancer treatment, especially with antitumor agents, in the daily activity of oncologists. The negative counterpart of this observation is that many medical oncologists do not keep fully abreast and fully trained in the wider aspects of medical oncology. In particular, poor performances were registered for what concerns hematological oncology, AIDS related tumors and tumor histology. Another failure reason could have been the lack of a broad training programme for medical oncologists, or the hyperspecialized activity of some candidates in limited aspects of oncology, such as hematological malignancies or solid tumors. However, all candidates obtained results superior to 50%, so that they all were eligible for the 1989 European qualification. In 1990, the ESMO Examination in Medical Oncology has been prepared on the same basis as for 1989. Although questions were changed, the general shape of the examination has been kept as similar as possible to the previous one, to enable an eventual evaluation of Table 5. Questions with correct answers < 40%. Type
% Correct answers
Description
K' K' K' E
10.4 11.2 16.8 28.8
E
35.2
A K' K'
39.2 39.2 40.0
Prognostic classification of lymphomas Risk factors for cervix cancer AIDS related lymphomas Medullary breast cancer, histology and prognosis Androgen production in postmenopausal women Toxicity of doxorubicin Diagnosis of myelofibrosis versus CML AIDS related Kaposi sarcoma
results over a period of several years, with a larger number of observations. A second session took place during the Lugano Lymphoma Conference in May 1990, and the third session will take place in December 1990, during the Annual ESMO Meeting in Copenhagen. Addendum I
General informations concerning the examination for the European Certification in Medical Oncology 1. The first examination will take place during ECCO 5, on September 1989, in London. 2. The written examination will consist of 60 multiple choice questions, to be answered in 90 minutes. 3. The examination will include questions of type A, B, E and K'. 4. The evaluation of the results will be made by the Educational Committee of the European Society for Medical Oncology. Examples of questions Type A Select the answer or completion which is best among 5 lettered answers. An increased incidence of breast cancer is most commonly associated with: A) early castration B) early age of menarche C) a history of taking estrogen-containing medication D) multiple pregnancies E) a history of endometrial cancer The correct answer is B.
12
Percentage of Correct Response. ESMO 1989
TypeB Put together the headings, words or statements which are most closely associated. A) pulmonary fibrosis B) peripheral neuropathy C) ototoxicity D) hyperglycaemia
60 -,
1. cisplatin 2. bleomycin 3. vincristin 4. prednimustin 5. L-asparagjnase The correct answer is: A-2, B-3, C-l, D-5.
60 -i
Type E Select whether both statements are right, and in case they are right, whether they are causatively related. Elevated serum concentration of neuronspecific enolase is frequent in advanced stages of small cell lung cancer /because/ brain metastases are frequent in this tumor type. 1. + because + 2. + / + 3. + / 4. / + 5. / The correct answer is 2: both statements are correct but not causatively related.
Total 30-
30-
60-i
30-
60 -i
Tumor Biology and Laboratory Tests
Histopathology and Cytology
Drug Pharmacology
30-
TypeK' Answer yes or no to each of the given alternatives. All, some or none may be correct. Chemotherapeutic agents which might cause lung Clin Therapeutics toxicity include: and Toxicity A) methotrexate 30 B) mitomycin C C) busulfan D) 5-fluorouracil E) cyclophosphamide The correct answer is: A-yes, B-yes, C-yes, D-no, 60 Clin Diagnostic E-yes. Received 20 September 1990; accepted 20 September 1990. Correspondence to: Dr. Pierre Alberto Division of Onco-Hematology University Hospital Rue Micheli-du-Crest 24 1211 Geneva 4, Switzerland
30-
60
n
30-
Epidemiology and Carcinogenesis
0%
20%
40%
60%
80%
100%
Percentage Fig. 1. Distribution of percentage of correct answers in subgroups. Stars indicate the mean value, circles indicate the median value.