Oncology education for Canadian internal medicine residents : the value of participating in a medical oncology elective rotation

Background Despite the high incidence and burden of cancer in Canadians, medical oncology (mo) rotations are not mandatory in most Canadian internal medicine (im) residency training programs. Methods All im residents scheduled for a mo rotation at 4 Canadian teaching cancer centres between 1 January 2013 and 31 December 2015 were invited to complete an online survey before and after their rotation. The survey was designed to evaluate perceptions of oncology, comfort in managing cancer patients, and basic oncology knowledge. Results The survey was completed by 68 im residents pre-rotation and by 48 (71%) post-rotation. Cancer-related learning was acquired mostly from mo physicians in clinic (35%). Self-directed learning, didactic teaching, and resident or fellow teaching accounted for 31%, 26%, and 10% respectively of learning acquisition. Comfort level in dealing with cancer patients and patients at end of life improved to 4.0/5 from 3.2/5 (p < 0.001) and to 4.0/5 from 3.6/5 (p = 0.003) respectively. Mean knowledge assessment score improved to 83% post-rotation from 76% pre-rotation (p = 0.003), with the greatest increase observed in general knowledge of common malignancies. The 3 topics ranked as most important to learn during a mo rotation were oncologic emergencies, common complications of treatment, and approach to diagnosis of cancer. Conclusions A rotation in mo improves the perceptions of im residents about oncology and their comfort level in dealing with cancer patients and patients at end of life. Overall cancer knowledge is also improved. Given those benefits, im residency programs should encourage most of their residents to complete a mo rotation.


BACKGROUND
Cancer is a leading cause of morbidity and mortality in the Canadian population, and the number of new cases continues to rise as the population grows and ages.Almost half of Canadians will develop cancer in their lifetime, and 1 in 4 will die of the disease 1 .Nevertheless, only 38% of internal medicine (im) training programs in Canada have a mandatory medical oncology (mo) rotation 2 during their 3 years of formal training.In most training programs, im residents have the option to pursue selectives or electives in mo.
Previous studies have consistently shown a deficiency in mo teaching in the undergraduate setting [2][3][4][5][6] .Given that problem at the undergraduate level, comprehensive cancer education at the postgraduate level is important because graduating residents will be moving on to independent practice.Studies examining oncology teaching in postgraduate training programs are limited, but demonstrate that oncology is underrepresented compared with other subspecialty topics 2,5 .In a recent study by Tam et al. 2 , medical students, im residents, family medicine residents, and program directors were asked to rank the adequacy of teaching of 10 medical subspecialty topics.Oncology was identified by both im residents and program directors as the least adequately taught compared with the 9 other topics.The study also found that 50% of im program directors and Current Oncology, Vol. 25, No. 3, June 2018 © 2018 Multimed Inc.
63% of the residents perceived oncology teaching in their respective training programs as inadequate.
Because education in im training programs is structured into clinical rotations, we designed the present study to clarify the value of a mo rotation for Canadian im residents.We aimed to compare perceptions of oncology as a specialty, comfort in the management of cancer patients, and level of basic oncology knowledge before and after the mo rotation.As a secondary objective, we compared the characteristics of oncology rotations at 4 participating teaching cancer centres.

Participant Recruitment
Between 1 January 2013 and 31 December 2015, all im residents scheduled to complete a mo rotation at 4 Canadian cancer centres-BC Cancer, Vancouver; Tom Baker Cancer Centre, Calgary; and Princess Margaret Cancer Centre and Sunnybrook Odette Cancer Centre, Toronto-were invited by e-mail to participate in this survey study.A resident researcher recruited the im residents such that the oncologists responsible for the resident evaluations were not aware of who participated.

Survey Distribution
Residents were sent a recruitment e-mail message 2 weeks before the start of their rotation.Each participant provided informed consent by responding to the initial e-mail message.Upon receipt of the consent, a link to the pre-rotation survey was sent, together with a unique identification number.Identification numbers were assigned to each participant to ensure anonymity and to allow for the pre-and post-rotation surveys to be compared.Within a week of completing their mo rotation, im residents were again sent an e-mail message asking that they complete the post-rotation survey.The Web site used to conduct this survey was http://www.surveymonkey.com.To enhance response rates, each participant was mailed a coffee card upon completion of the post-rotation survey.

Survey Content
The pre-rotation and post-rotation survey instruments both contained questions about demographics, the resident's perceptions of oncology as a subspecialty, and the resident's comfort level with managing cancer patients.In the pre-rotation survey, a comprehensive list of oncology topics taken from the Objectives of Training in the Subspecialty of Internal Medicine set out by the Royal College of Physicians and Surgeons of Canada (rcpsc) 7 and the Canadian Oncology Goals and Objectives for Medical Students defined by the Canadian Oncology Education Group 8 were provided, and residents were asked to rank the importance of each topic.The survey also contained 20 basic cancer knowledge-assessment questions related to cancer epidemiology, screening, diagnosis, complications of cancer and treatment, prognosis, and follow-up care.The questions were written by a Fellow of the rcpsc certified as an internal medicine educator and were approved by attending mo physicians at participating centres to have good construct validity for assessing cancer knowledge.Questions were designed to address cancer-related objectives of training for im according to the rcpsc 7 .The pre-and post-rotation knowledge assessment questions were identical in content, but the order of the questions was changed to avoid recall bias.The post-rotation survey also included questions about specific details of the recently completed mo rotation and sought opinions about the quality, duration, and other characteristics of the rotation.The comprehensive list of oncology topics ranked in the pre-rotation survey were again provided, and residents were asked to rate how well each topic was taught during their oncology rotation.

Statistical Analysis
Descriptive statistics were calculated for survey responses.Paired t-tests were used to compare the mean scores for questions about perceptions of, and comfort with, mo and for knowledge assessment questions pre-and post-rotation.Additional exploratory analyses compared the mean scores between various groups.A 2-sided p value less than 0.05 was considered statistically significant for all outcomes.All analyses were performed using the IBM SPSS Statistics software application (version 19.0: IBM, Armonk, NY, U.S.A.).

RESULTS
Of 192 im residents, 66 (34%) agreed to complete the pre-rotation survey.Subsequently, 48 of those participants (73%) completed the post-rotation survey.Table i shows the characteristics of the 48 residents who completed both surveys.Most were 25-29 years of age, male, and at the PGY-1 training level.Almost 45% had completed a previous rotation in mo.

Characteristics and Perceptions of the MO Rotation
Table ii summarizes and compares the characteristics of the mo rotations at the participating cancer centres that were evaluated post-rotation.Most rotations (75%) were 4 weeks in duration.Overall, the residents felt that the duration of their mo rotation was adequate.Residents spent approximately 80% of their rotation in outpatient clinics.Residents participating in mo rotations at the Tom Baker Cancer Centre, compared with those participating at the other sites, spent a greater proportion of their time performing emergency room assessments.The average number of overnight call shifts was 4-5 per rotation, a number that was similar at the Tom Baker Cancer Centre, BC Cancer, and the Princess Margaret Cancer Centre.Only 1 overnight call shift was completed by im residents during their mo rotation at Sunnybrook Odette Cancer Centre.Overall, the quality of teaching both in didactic sessions and in clinics was rated 4.0 on a 5-point Likert scale.
The most frequently used resource from the list provided was, at all 4 cancer centres, http://www.UpToDate.com.
Other resources that were used by the residents and that had not been listed included the BC Cancer Web site, Alberta Health Services cancer treatment guidelines, Cancer Care Ontario guidelines, and the American Society of Clinical Oncology SEP book.
Table iii summarizes the proportion of learning delivered using various approaches during mo rotations.Overall, the im residents obtained most of their learning from staff teaching in clinic and self-directed learning or reading.The exception was at Princess Margaret Cancer Centre, where 51% of learning came from didactic teaching sessions.
The perceived importance of various cancer topics and the quality of teaching of each of those topics was evaluated at the end of the rotation (Table iv).Oncologic emergencies, complications of cancer treatment, and approach to diagnosis of cancer were rated as the most important topics.However, in terms of quality of teaching, only the complications of treatment topic was rated well (4.0/5).General principles of treating cancer and principles of systemic therapy were also taught well during the mo rotations.
In terms of the mo rotation providing cancer knowledge in preparation for the rcpsc im examination, the im residents gave a mean rating of 3.9/5, which corresponds to moderate-to-great benefit.
Overall, 54% of residents felt that mo should be a mandatory rotation during im training, 42% believed that it should be a selective rotation, and only 2% believed that mo should be an elective only.

Perceptions of MO and Cancer Patients
Table v summarizes the perceptions of the residents about mo as a specialty and their comfort in dealing with mo-related patients.Comparing the pre-and post-rotation scores, im residents were more likely to consider mo as a career after the rotation (p = 0.002).Level of comfort in dealing with cancer patients and patients at end of life was also significantly improved.After a mo rotation, im residents were also less likely to be concerned that a career in mo would be overly emotionally draining (p = 0.032).

Cancer Knowledge Assessment
The mean test score pre-rotation was 15.2/20 (76%).Post-rotation, the score increased to 16.7/20 (84%), which represented a statistically significant improvement (p < 0.001).The greatest improvements were observed in questions related to cancer epidemiology, screening, prognosis, and follow-up care for common cancers (surveillance).No improvements were observed in the questions dealing with the approach to diagnosis, complications of treatment, and oncologic emergencies.An improvement in mean test score was observed only for the 27 students who had not completed a previous rotation, with the mean score improving to 17.5/20 from 15.3/20.In contrast, the mean score was 15.2/20 both pre-and post-rotation for the 21 students who had completed a prior rotation.

DISCUSSION
Our study is the first to demonstrate the educational value of a mo rotation for im residents.The im residents showed an improvement in comfort level when dealing with cancer patients and patients at end of life.Experiencing the mo rotation also improved their perceptions of mo as a specialty and their general knowledge of cancer and its treatment.Few qualitative differences were found between the mo rotations at the 4 participating Canadian cancer centres.This study also served as an example that multicentric education research is possible in Canada.
Overall, a statistically significant improvement in the mean knowledge assessment score was observed, but that improvement was limited to residents for whom the mo rotation was their first in that specialty.The scores of the residents who had previously completed mo rotations as a medical student or earlier in residency were the same pre-and post-rotation.That finding suggests that most of the knowledge benefit might be attained after one rotation, and that afterward, little change in basic oncology knowledge occurs.The greatest improvements in knowledge were related to cancer epidemiology, screening, prognosis, and follow-up care for common cancers (surveillance).
Although some of those improvements can be attributed to the quality of the teaching during the oncology rotation, it is interesting to note that the least improvement on the knowledge questionnaire was seen for complications of treatment, and yet the quality of the teaching for that topic was still subjectively rated 4.0/5.That observation suggests that perceptions of quality of teaching might not always correlate with knowledge retained.Other factors such as the knowledge attained from self-directed learning during the rotation might also play a role.
A number of studies have examined oncology education in undergraduate medical education 2,9 , but we are aware of only two other studies dealing with oncology education for im residents 9,10 .The study by Chen et al. 9 used a short survey instrument to determine the opinions of im residents about a hematology/oncology rotation at the State University of New York at Buffalo.Some of their findings were similar, in that they found that approximately one third of the residents reported improvement in knowledge about pain management, hospice, palliative care, and end-of-life care.However, unlike our survey, their survey included no objective measurement of improvement.Of their residents, 64% reported an improved general opinion of hematology/oncology, and 40% were more motivated to pursue that specialty.McFarland et al. 10 conducted a survey study of interns and residents at a single institution rotating through an inpatient hematology/oncology ward in New York.They found that their trainees showed decreased interest in the field of hematology/oncology after the inpatient rotation.They suggested that a similar study in outpatient hematology/oncology should be conducted, given that oncology is increasingly an outpatient practice and that residents would thereby obtain a more accurate perspective of oncology practice.To our knowledge, ours is the only Canadian study to examine the value of an oncology rotation at the postgraduate im training level.Additionally, our study included multiple centres across the country, minimizing institutional bias.
Observed improvements in the perception of oncology and in comfort in dealing with oncology patients are likely to be explained by the exposure to cancer patients that a mo rotation provides.Misconceptions about mo and  cancer patients are likely, because im residents are typically exposed to cancer inpatients who are unwell and admitted to medical teaching units for diagnosis or for complications of treatment.A mo rotation provides im residents with the opportunity to experience cancer care in the outpatient clinic setting where most cancer patients are managed.It also allows them to see that, in addition to palliative treatments, some patients are treated with adjuvant therapy with curative intent.The im residents are also able to see the full spectrum of cancer patients, most of whom tolerate their treatment well and do not have to be hospitalized.
Although im residents have likely seen cancer diagnosis and end-of-life care in hospitals on other rotations, the mo rotation gives them perspective about a significant part of a patient's cancer journey during which they are being treated and building relationships with the cancer centre team.That perspective might lead to the observed decrease in the perception that a career in oncology is emotionally draining and undesirable as a subspecialty career choice.However, despite the improved level of comfort in dealing with cancer patients and patients at end of life, relatively poor ratings were given to the quality of teaching for palliative care and breaking bad news (3.0 and 2.7 respectively on the 5-point Likert scale).That observation supports the hypothesis that comfort in dealing with cancer patients is more likely derived from exposure to patients than from didactic lectures or clinical teaching by staff oncologists.Regardless of whether mo is pursued as a career choice, nearly all physicians will deal with cancer patients in their practice and will encounter questions about cancer on their im examination.All specialties within medicine will have to consider the effects of cancer, because both the disease and its treatments have the potential to affect almost every organ system.It is therefore imperative that im residency training prepare physicians for encountering cancer patients in their practice.Our knowledge assessment questions were able to show that overall cancer knowledge improved after a mo rotation-an improvement that was likely gained from direct patient encounters, clinical teaching, and self-study during mo rotations.Only modest improvements were observed in the category of oncologic emergencies, a topic that was ranked most important by the im residents.The quality of teaching of oncologic emergencies was rated only 3.4/5 and could therefore be improved.It is unlikely that, during a 4-week mo rotation, im residents would be exposed to a wide variety of oncologic emergencies in clinics (where they spend little time on emergency assessments).That situation emphasizes the importance of including uncommon, but fundamental, topics as part of didactic teaching sessions and increasing the proportion of learning that occurs using that approach.

Limitations
Despite the strengths of our multicentre survey approach, our study has several limitations.The sites included in the survey were mainly large academic cancer centres.The results might be less applicable to mo rotations for im residents at smaller cancer centres.Sampling bias might also be an issue, given that the response rate for the pre-rotation survey was only 34%.It is likely that im residents who were more interested in oncology were more motivated to complete the survey.Strategies to enhance the response rate were already in place and included sending subsequent reminder messages and offering a coffee card as an incentive to those who completed the surveys.Another limitation of our study was the use of a survey instrument that has not been validated in other studies.To help improve validity, an identical survey was distributed at all centres.Although we attempted to capture important characteristics of rotations at various sites that could affect learning, certainly some were not captured.For example, we did not collect information about the details of the overnight call shifts (for example, in-house vs. home) and whether additional services (hematology or bone marrow transplantation, for instance) were covered as well.In the future, we plan to use the same survey to expand the study to more centres, including smaller centres within Canada.

CONCLUSIONS
Our survey study provides evidence that a predominantly outpatient rotation in mo can improve a resident's perception of oncology as a specialty also their comfort in dealing with cancer patients and patients at end of life.General oncology knowledge was also improved.Based on those benefits, im residency programs should consider promoting mo rotations for their residents.
Ethics approval was obtained from the Research Ethics Board at the University of British Columbia, the Conjoint Health Research Ethics Board at the University of Calgary, and the University Health Network Research Ethics Board and Sunnybrook Research Institute Research Ethics Board in Toronto.

TABLE I
Characteristics of the internal medicine residents who participated in the survey

TABLE II
Characteristics of the rotation a Likert scale: 1 = not enough to 5 = too much.b Likert scale: 1 = poor to 5 = excellent.

TABLE III
Proportion of oncology learning during the rotation a Because of rounding, column totals might not exactly equal 100%.Current Oncology, Vol. 25, No. 3, June 2018 © 2018 Multimed Inc.

TABLE V
Perceptions of medical oncology (MO) and cancer patients