- Top of page
- Materials and methods
The effect of a gynecological oncology fellow on obstetrics and gynecology resident education and training is uncertain. The objective is to assess the effect of gynecological oncology fellows on the surgical training of residents in obstetrics and gynecology. Fourth year residents in obstetrics and gynecology in the United States were identified and stratified as to the presence or absence of an oncology fellowship program. Demographics, surgical volume, procedures performed, and self-assessment of surgical proficiency were collected. Responses were compared between residency programs with and without fellowships. Responses were received from 40% of programs. Residents at programs without a fellowship more frequently operated with attendings than did residents at programs with fellows, 91% vs 77%, P= 0.016, and more frequently were responsible for complicated cases, 39% vs 22%, P < 0.0001. Over 90% of residents in both groups reported surgical training as positive and valuable; both groups reported a similar perceived lack of proficiency in radical hysterectomy and lymphadenectomy. Attitudes toward the fellows were generally positive; however, competition for cancer cases was reported by over 66% of residents from programs with fellows. While fellows are often thought of as a detracting factor to residency training, they do not appear to affect the perception of the quality of resident surgical training.
Residency in obstetrics and gynecology continues a physician’s medical education by direct involvement in the primary, obstetric, and gynecological care of patients under the supervision of faculty members or mentors. When his or her training is completed, the physician should have developed the knowledge and skills to practice independently. As a surgical subspecialty, operative training in obstetrics and gynecology is a very important part of residency training as a whole. The specific surgical training that occurs during residency involves both didactic sessions and one-on-one technical instruction at the operating table from a variety of mentors. Lipscomb et al.(1) surveyed residents in an attempt to assess the adequacy of this training and to identify what aspects in particular are most beneficial. These authors reported a statistically significant correlation between the residents’ perception of surgical skill and the number of uncommon or specialty-related procedures they performed. Of the particular procedures that reached statistical significance, more than half were procedures that residents would most likely be exposed to during their rotations in gynecological oncology (GO).
As a subspecialty of obstetrics and gynecology, GO requires fellowship training after residency. These fellowships are generally in institutions that have obstetrics and gynecology residency programs; thus, residents and fellows work together to provide patient care. The impact of this resident–fellow relationship on training has been previously examined. Metheny and Sherline surveyed residents and program directors and found that, in general, fellows are felt to improve the quality of resident training(2,3). However, in regard to GO fellows, residents specifically felt that didactic teaching and research were improved, but that surgical experience was unaffected or even worsened through the competition for procedures. Thus, we set out to further examine the role of the GO fellow in the surgical training of obstetrics and gynecology residents.
Materials and methods
- Top of page
- Materials and methods
Following approval from the behavioral institutional review board, all of the 35 American Board of Obstetrics and Gynecology–approved GO fellowship programs were identified, as were their affiliated obstetrics and gynecology residency programs, if applicable. Geographically matched (according to major population centers within the northeast, south, midwest, southwest, and west regions) university-based obstetrics and gynecology residency programs were then matched in a 2:1 ratio to each residency affiliated with a GO fellowship program. Within these regions, residency programs without GO fellowships were randomly selected (without knowledge of the demographic factors of the residents) and enrolled in this study, based on their willingness to participate in this study and similarities in demographics of the city, and number of residents when matched with the residency with an associated GO fellowship. When more than two academic residency programs without a GO fellowship were available in the geographic region, matching was done in a 3:1 ratio to maximize the number of potential residents in this study. The total number of programs in each group was 72 and 35, respectively. This sample size was chosen to reflect all residencies with an associated GO program and at least one program without a GO fellowship in the same geographic region. Following approval from the institutional review board, programs were contacted via email to describe our study and ask the program director or coordinator to answer four questions that confirmed the presence of any affiliated fellowships. Programs who did not respond after two email attempts were then contacted by telephone. The web address of the survey was forwarded by email to all 4th year residents at programs that agreed to participate. Repeat mailings to nonrespondents were done at 2 and 4 weeks after the first.
The survey collected demographic data regarding age, race, gender, and practice plans after completion of residency. Residents who did not complete their entire residency at their current program were excluded. General questions regarding residency structure, and specifically regarding the usual level of surgeon with whom the resident worked, were used to determine the perceived level of supervision. Three separate general questions regarding the quality of surgical training were asked, with answers being provided on a 5-scale continuum from favorable to unfavorable, positive to negative, and valuable to worthless. Surgical proficiency in general gynecological techniques as well as those encountered on a GO service were asked using a categorical variable scale ranging from 1 (extremely proficient) to 5 (not at all proficient). The frequency of performing such procedures was reported using a categorical variable scale ranging from 1 (<1 case per year) to 5 (>25 cases per year). For residents in programs with a GO fellowship, the opinions regarding their interactions with the fellows were queried. Specifically, questions such as “how well do you work with the GO fellows,”“how important is the GO fellow to your (research or surgical) training,” and “to what extent do GO fellows and residents compete for surgical cases” were reported using a categorical variable scale ranging from 1 to 5. Narrative responses regarding the role of GO fellows in their programs were recorded verbatim. The survey was designed specifically for this study and has not been validated previously.
Responses were converted to a comma separated value file, and the data was then analyzed. Two-sample t-tests or nonparametric Wilcoxon rank sum tests were used for statistical comparisons when appropriate. Categorical covariates were compared using Chi-square tests of association. The sample size reported was limited by the number of American Board of Obstetrics and Gynecology fellowship programs, and not specifically powered to answer specific hypothetical questions.
- Top of page
- Materials and methods
Agreement to participate in the study was received from all 107 of the residency programs contacted. This included 35 programs with a GO fellowship and 72 without a GO fellowship. Of these programs, at least one response was received from 13/35 (37%) with a GO fellowship and 30/72 (42%) without a GO fellowship. Of the programs without a GO fellowship who had at least one resident respond, 12 had no fellowships, while the remaining 18 had one or more fellowships that included maternal fetal medicine, reproductive endocrinology and infertility, female pelvic medicine and reconstructive surgery, or family planning. Responses were received from 22/232 (10%) of residents from programs with affiliated GO fellowships and 83/404 (21%) of residents from programs without an affiliated GO fellowship. Sixteen responses were excluded due to previous training received at a residency program other than the resident’s current program or minimal completion of the survey (Table 1). Demographics were similar between respondents in residencies with and without a GO fellowship, including a similar percentage (approximately 30%) from each group who planned to enter a fellowship upon completion of residency training. However, the subspecialty breakdown was quite different among them; 50% (3/6) of residents planning a career in a subspecialty coming from programs with a GO fellowship chose to pursue a fellowship in GO, compared with only 15% (3/20) of residents planning to pursue fellowship training from programs without a GO fellowship program, P= 0.1 (Table 2).
Table 1. Response rates in programs with and without a gynecological oncology fellowship programs
| ||No GO fellowship||GO fellowship|
|Number of programs||72||35|
|Number of residents||404||232|
|Total responses||83 (21%)||22 (10%)|
Table 2. Demographic information from respondents from programs with and without gynecological oncology fellowship programs
| ||No GO fellowship (%)||GO fellowship (%)|
|Mean age (years)||31||33|
| Private practice||56||56|
| Academic practice||13||11|
When asked general questions about their surgical training, both groups of residents rated both the quality and volume of surgical training as satisfactory, and all agreed that they worked well with the faculty and other residents at their program (Table 3). Over 90% of residents from both groups reported their surgical training as positive and valuable. An attending surgeon (and not another resident or fellow) was identified as the person with whom a resident most frequently operated in 83% of each group. However, residents at programs without a GO fellowship were more likely to identify attendings as the physician from whom they received the majority of their surgical training (91% vs 77%, P= 0.016, Table 3). Residents at programs without a GO fellowship were more likely to feel that they were responsible for the majority of complicated cases compared with the residents from programs associated with a GO fellowship (39% vs 22%, P < 0.0001, Table 3).
Table 3. Ratings of surgical experiences and training reported by residents from programs with and without gynecological oncology fellowship programs
| ||No GO fellowship (%)||GO fellowship (%)||P value|
|Quality of surgical training|
|Volume of surgical training|
|Majority of surgical training from attending||91||77||0.016|
|Primary responsibility for complicated surgical cases|
Residents at a program with a GO fellowship program more commonly reported that they assisted in greater than or equal to six radical hysterectomies (56% vs 28%, P= 0.047) and greater than or equal to six laparoscopic oophorectomies (100% vs 79%, P= 0.035) than residents training at a program without a GO fellowship program. No significant differences were noted between the rates of assisting in or being the primary surgeon during the remainder of the procedures compared, including retroperitoneal lymphadenectomy, intestinal and urinary tract surgeries, and laser surgery (data not shown).
Residents from both types of program reported similar perceived proficiency in performing the breadth of gynecological procedures performed during a residency. As seen in Table 4, the mean proficiency scores and standard deviations are not significantly different between the groups, and in general reflect proficiency with standard procedures such as abdominal and vaginal hysterectomy, and lack of self-reported proficiency with the radical surgical procedures characteristic of gynecological cancer surgery such as radical hysterectomy and retroperitoneal lymphadenectomy. Despite the lack of statistical significance, there did appear to be a trend toward slightly higher assessments of proficiency for the subspecialty procedures reported by those residents at programs without a fellowship. A greater than 10% difference in the mean self-reported proficiency scores was noted for pelvic lymphadenectomy (26%), para-aortic lymphadenectomy (10%), radical hysterectomy (16%), and LASER surgery (13%), with residents in programs with a GO fellowship reporting increased proficiency compared with residents from programs without a GO fellowship (Table 4).
Table 4. Mean self-reported proficiency scores (+/− standard deviation) from residents in programs with and without gynecological oncology fellowship programs. Procedures are generally organized in order of decreasing proficiency. Responses reflect a numerical score from 5 (extremely proficient) to 1 (not at all proficient)
| ||No GO fellowship||GO fellowship|
|Total abdominal hysterectomy||4.53 +/− 0.77||4.67 +/− 0.49|
|Bilateral salpingo-oophorectomy||4.54 +/− 0.77||4.78 +/− 0.43|
|Laparoscopic oophorectomy +/− salpingectomy||4.27 +/− 0.86||4.17 +/− 0.92|
|Retroperitoneal ureteral identification||4.11 +/− 0.90||4.17 +/− 0.79|
|Total vaginal hysterectomy||3.75 +/− 1.12||3.83 +/− 0.86|
|Cystoscopy||3.66 +/− 1.13||4.00 +/− 0.91|
|Enterolysis||3.18 +/− 1.06||3.44 +/− 1.15|
|Cystotomy repair||3.13 +/− 1.03||3.11 +/− 1.18|
|Wide local excision of the vulva||2.99 +/− 1.18||3.28 +/− 1.18|
|Appendectomy||2.86 +/− 1.15||2.83 +/− 1.20|
|Laser surgery||2.82 +/− 1.13||2.44 +/− 1.15|
|Simple enterotomy repair||2.62 +/− 1.02||2.78 +/− 1.17|
|Pelvic lymphadenectomy||2.16 +/− 0.96||1.61 +/− 0.92|
|Radical hysterectomy||1.92 +/− 0.79||1.61 +/− 0.92|
|Para-aortic lymphadenectomy||1.79 +/− 0.83||1.61 +/− 0.92|
When residents from programs associated with GO fellowships were asked about their relationship with GO fellows (Table 5), 83% of residents viewed the presence of a GO fellow as positive, and 68% rated the fellows’ influence on their surgical training as valuable. GO fellows were frequently reported to be at least somewhat important to a resident’s research (48%), surgical (84%), and clinical training (89%). Comments such as “I owe most of my surgical training to the GO fellows” were common from residents in programs with a GO fellowship. Despite these favorable ratings, the majority of residents (72%) felt that there was at least a moderate amount of competition for surgical cases between residents and fellows. In support of these results, unstructured comments from residents in programs associated with a GO fellowship included statements such as “residents would have greater surgical experience in the absence of fellows.” However, we have no data on the competition felt between residents in programs with and without a GO fellowship, so the overall competition for cancer cases cannot be assessed.
Table 5. Responses from residents at programs with a gynecological oncology fellowship program regarding the resident–fellow relationship
| ||Responses (%)|
|What is your impression of the influence of the GO fellow?|
| Positive or very good||83|
|How do you interact (work) with the GO fellow?|
| Extremely or very well||78|
| Moderately well||22|
|What is the importance of the GO fellow to surgical training?|
| Extremely or very important||45|
| Somewhat important||39|
| Slightly or not at all important||16|
|What is your impression of the distinction of clinical responsibilities between you and a GO fellow?|
| Extremely or very distinct||67|
| Somewhat or slightly distinct||23|
|What is the level of competition for surgical cases with the GO fellow?|
| A great deal or quite a bit||33|
| A moderate amount||39|
| A little or not at all||28|
- Top of page
- Materials and methods
Overall, the collected data shows that residents at programs both with and without GO fellowships are generally satisfied with the surgical training they receive. Each group rated the quality and volume of their surgical training positively. Residents at programs without an associated GO fellowship did report an increase in the amount of surgical training received from attending physicians, as well as an increase in primary responsibility for complicated cases when compared to residents at programs with a fellowship. Despite this distinction, both groups of residents reported statistically similar proficiency ratings and frequencies of procedures. However, the small differences between the two groups’ mean scores of proficiency ratings for the subspecialty procedures could potentially be attributed to the increased amount of instruction received directly from an attending physician.
For the two procedures that did show a significant difference in frequency between the groups, it is not unexpected for those residents at programs with an affiliated GO fellowship to be exposed to a greater number of radical type procedures as an assistant. As the fellows take on the primary surgeon or first assistant role, residents are often left to first or second assist during the procedure. While teaching does occur in this setting, the confidence of performing the procedure will not improve without some primary responsibility being taken by the resident. It is important to note, however, that despite the more frequent role as assistant, residents in programs with associated GO programs still reported similar frequencies in the role of primary surgeon and proficiency with all of the procedures. They also reported overall satisfaction with their surgical training, despite the prevailing feeling of competition for procedures.
These same seemingly discrepant findings have been reported previously. Cundiff et al.(4) showed that when a new fellowship in Female Pelvic Medicine and Reconstructive Surgery was approved, residents initially anticipated that the fellowship would detract from their educational experience. However, over a 3-year period, these residents persistently increased their positive rating of the impact of the fellowship, as well as increasing their self-assessments of quality of education. In another study of fellowship and residency directors, Perler et al.(5) found that, despite a decrease in resident case load, there was an increase in the quality of vascular surgical training. Therefore, this study concluded that vascular surgery fellowships had not adversely affected general surgery training, and actually often enhanced it.
It was somewhat surprising to find a statistical difference in the number of laparoscopic oophorectomies each group had performed as an assistant, as this procedure is typically performed for benign indications and would not be expected to increase or decrease in frequency based on the presence of a GO fellowship. It is possible that this discrepancy is explained by the referral patterns of patients with an increased risk for hereditary ovarian cancer going to programs with a GO fellowship program rather than a program without such a fellowship program. Another explanation is that the GO surgeons are more willing to perform advanced laparoscopic procedures, for example, the resection of a large complex ovarian mass, which gynecologists might otherwise refer to a subspecialist or perform via laparotomy.
Though not statistically significant, it was interesting to note the considerably higher percentage of residents from programs with an affiliated GO fellowship who also chose to pursue a GO fellowship themselves. This could be explained in two ways: medical students who know they have an interest in this subspecialty choose a residency program because of the presence of the affiliated fellowship, or the exposure to the field and interaction with the fellows during residency fosters a greater interest in the subspecialty.
The main limitation of this study was the poor response rate. This did limit the ability to find statistical significance between much of the data collected from the two groups, including the residents’ ratings of proficiency and the number of procedures performed. Future studies could further address these questions, perhaps using a database such as the Accreditation Council for Graduate Medical Education’s Resident Case Log System. Obviously, this remains limited by its reliance on resident participation, which is understandably lacking due to the many demands of a resident’s schedule independent of the presence or absence of a GO fellowship. The influence of this low response rate on the results of this study is unclear. It is possible that less busy residents were more likely to respond to the survey; however, there is no reason to assume that residents in programs with a GO fellowship are any busier (or less busy) than those in programs without a GO fellowship. As such, any bias resulting from the low response rate would likely affect the results in an unbiased fashion. Another source of bias in this study is the nonrandom selection of residency programs. Although unintentional, a selection bias that enriched the population for gender, race, or other demographic features may have occurred; again, the effect of such bias is uncertain.
In summary, while the perception that a GO fellowship is a detracting factor from residency training is still pervasive, these data suggest that a GO fellow does not detract from a resident’s surgical training and patient management.