Materials and Methods
We recruited MFM fellows by contacting all associate members of the Society for Maternal Fetal Medicine (SMFM) and inviting fellows who were enrolled in 1 of the ABOG-approved MFM fellowship sites in 2010. The names were obtained from a purchased list that is available through the SMFM, the ABOG subspecialty handbook, and a search of the institutional websites of the fellowship sites. When we were able to obtain email addresses from the institutional websites or from PubMed, we sent an email invitation with a link to the online survey. Fellows for whom an email address was not available received a mailed invitation that contained the survey and a link to the online version. If names of the fellows were not available publicly from institutional websites, we sent paper surveys addressed to “MFM Fellow” at their respective institutions. Those recruited by email received 3 separate email invitations; those who received the paper survey received a postcard reminder once after the original mailing. As an incentive, respondents were offered a $5 gift card that was not contingent on completion of the survey. To preserve anonymity, the respondent’s name and address were entered separately from the survey responses. All responses were entered into KeySurvey, an online software program.
The survey included 75 questions on topics that included demographics and training opportunities that were available at their fellowship. Training opportunities were classified as “routine” (a required fellowship rotation), “opt-in” (available for interested fellows to arrange), or “not available.” We also asked questions about previous experience with D&E and about abortion attitudes with 5 questions with Likert scale responses. Abortion attitude scores ranged from 5–25; higher scores reflected attitudes more supportive of abortion. This survey was completely anonymous and asked only about geographic location but not about fellowship institution.
To complement the individual fellow’s perspective, we then conducted a follow-up study that was directed at MFM fellowship program directors to confirm the availability of training opportunities. We sent a link to a 10-question online survey to the 79 ABOG-approved MFM fellowship program directors whose names and email addresses are publicly available on the SMFM website. Two reminder invitations were sent, and a $5 gift card was offered as an incentive. These questions focused directly on the D&E training opportunities that were available to MFM fellows and the estimated proportion of fellows who participated. Fellowship directors were asked about the availability of formal routine or opt-in rotations (identical to the options given to the fellows themselves); a third category that addressed informal training opportunities (“no formal training, but fellows can participate in D&E when they occur”) was added based on feedback from the fellows’ survey. Both surveys were approved by the Committee on Human Research at the University of California, San Francisco.
We performed descriptive statistics with χ 2 tests and multivariable logistic regression to identify correlates of D&E provision and training. We compared geographic characteristics with publicly available information about MFM and family planning fellowships. All analyses were performed with Stata software (version 12; Stata Corporation, College Station, TX). Results with a P value of < .1 were examined for possible statistical significance because of our small sample size and our desire to avoid a type II error of concluding no difference where one does truly exist.
Results
Characteristics of respondents
Of the 270 MFM fellows in 2010, we obtained the names of 190 and the email addresses of 156. A total of 126 paper surveys were mailed, 80 of which were addressed to “MFM Fellow” because names were not available. Ninety-two fellows responded to our survey for an overall response of 34%, with those invited by email more likely to respond (48%; 75/156). Most survey respondents were female (78%), and most lived either in the Northeast or the West ( Table 1 ). Forty-four of the 79 MFM fellowship program directors completed our survey, for a response rate of 56%. These fellowship directors are at programs that train 168 fellows, which is 57% of all current MFM fellows. Most fellowship directors who responded to our survey were from the Northeast or South/Southwest ( Table 1 ). Nationally, most fellows and fellowship programs are located in the Northeast (33%), followed by the South and the Midwest (23% each) and the West (18%). However, there were no statistically significant differences between the geographic distribution of the fellows or the fellowship directors when compared with the national distribution of MFM fellowships (χ 2 test: P = .358 and .801, respectively).
| Characteristic | Maternal-fetal medicine, n (%) | ||
|---|---|---|---|
| Fellows (n = 92) | Fellowship directors (n = 44) | All fellowships a (n = 79) | |
| Region | |||
| Northeast | 37 (40) | 16 (37) | 26 (33) |
| West | 22 (24) | 6 (14) | 14 (18) |
| Midwest | 17 (18) | 9 (21) | 20 (25) |
| South | 16 (17) | 13 (30) | 19 (24) |
| Family planning fellowship at institution | 59 (64) b | 15 (34) | 24 (30) |
a Data from the Society for Maternal-Fetal Medicine and the Fellowship in Family Planning
b P < .001 for χ 2 test of difference that compared the prevalence of family planning fellowship among maternal-fetal medicine fellow respondents and the national prevalence; no other comparisons were statistically significant.
Results from fellows’ survey
Two-thirds of MFM fellow respondents are in fellowships that offer D&E training (n = 60; 65%), and three-quarters of those fellows have training that is opt-in rather than routine (n = 45, 75%). Training is equally likely to be supervised by an MFM or a family planning subspecialist (45% vs 42%). D&E training is most likely to occur in the hospital operating room (60%) compared with labor and delivery (15%) or in an outpatient setting (22%).
When asked about their personal provision of second-trimester abortions, 87% of fellows provide either induction termination or D&E; 37% of respondents perform D&E. Most fellows who provide D&E were trained in residency; however, 22% of them report that they received training only during their fellowships. Just over one-third of fellows who responded to the survey (37%) intend to provide D&E services for their patients after fellowship. After adjustment for abortion attitude and having been trained in D&E, being at a fellowship that offers D&E training (either opt-in or routine) is associated with 7.5 times higher odds of the intention to provide D&E later in their career ( P = .005; 95% confidence interval, 1.8–30; Table 2 ).
| Correlate | Fellows who intend to provide dilation and evacuation (n = 36) | Fellows who do not intend to provide dilation and evacuation (n = 56) | Odds ratio | |
|---|---|---|---|---|
| Unadjusted (95% CI) | Adjusted (95% CI) | |||
| Age, y a | 34 ± 3.6 | 34 ± 3.5 | 0.99 (0.9–1.1) b | — |
| Location: South vs other, n (%) | 5 (14) | 11 (20) | 0.65 (0.2–2.1) | — |
| Male, n (%) | 14 (39) | 6 (11) | 0.2 (0.06–0.6) | — |
| Family planning fellowship at current institution, n (%) | 26 (72) | 33 (59) | 1.8 (0.7–4.5) | — |
| Family planning fellowship at past institution, n (%) | 13 (36) | 22 (41) | 0.8 (0.3–1.9) | — |
| Dilation and evacuation training offered in fellowship, n (%) | 32 (89) | 28 (50) | 8 (2.5–25) | 7.5 (1.8–30) |
| Abortion attitude score a,c | 19 ± 2.9 | 16 ± 4.6 | 1.2 (1.1–1.4) d | 1.2 (1.0–1.4) |
| Trained in dilation and evacuation at any time, n (%) | 32 (89) | 22 (39) | 12.4 (3.9–39) | 7.6 (2.1–26) |
| Trained in residency, n (%) | 20 (56) | 20 (35) | 2.3 (0.9–5.3) | — |
b Odds ratio for each increase in year
c Abortion attitude scores can range from 5–25; higher scores reflect attitudes more supportive of abortion
d Odds ratio for 1 point increase in abortion attitude score.
Of the 58 fellows who do not currently provide D&E, 31% reported lack of training as the major barrier to provision, and 34% would like to provide D&E. Of the 38 fellows who have not been trained in D&E, one-third (n = 13, 35%) of them desire training. Almost one-third of fellows currently at fellowships that do not offer D&E training would like to be trained (29%). The overwhelming majority of respondents think that D&E training should be offered during MFM fellowship (86%), with routine training identified as the preferred strategy of fellows at institutions both with and without D&E training ( Figure ).
Results from fellowship directors’ survey
According to the fellowship directors who responded to our survey, 46% of programs offer some type of organized D&E training; routine training, which is the least common training strategy, is seen in only 18% of all programs (n = 8). The most common training environment for MFM fellows is an informal one, where, although there is no formalized training elective, fellows can participate in D&Es that are done on labor and delivery or in the operating room rather than in a dedicated family planning clinical location (n = 18; 41%). Eleven percent (n = 5) of MFM fellowships do not offer any type of D&E training; the remainder of programs have optional training (n = 13; 30%). Consistent with what was reported by the fellows, training most often occurs in the hospital operating room (52%) and is staffed by an MFM attending (51%).
Compared with programs that do not offer organized D&E training, programs that do offer training are more likely to be located in the West (24% vs 4%; P = .06) and to collaborate with family planning subspecialists to provide training (43% vs 17%; P = .06). Our data suggested that programs with informal training opportunities may be more likely to have most fellows participate in D&E training (50% vs 23%; P = .09) compared with programs with more formalized opportunities ( Table 3 ). However, none of these results had a P value of < .05, so that may be due to chance alone.
