Abstract
Background:
Emergency contraception (EC) has the potential to play a vital role in preventing unintended pregnancies after unprotected sexual intercourse or contraceptive failure. Residency training can influence practice behaviors, however, the extent to which EC-related information is taught in training programs remains unknown. This study examined where residents obtain information about EC and whether knowledge differs by resident program characteristics.
Materials and Methods:
Program coordinators of Obstetrics and Gynecology (OB/GYN) and Family Medicine residency programs (n = 689) were emailed and requested to forward the survey link to residents. The survey included measures of EC education (hours, sources, including lectures, grand rounds), and EC-related knowledge. EC knowledge items assessed the three methods of EC (copper intrauterine device, ulipristal acetate, and oral levonorgestrel), effectiveness, mechanism of action, contraindications, and side effects. t-Tests and analysis of variances were used to compare mean knowledge scores (maximum = 20; higher scores indicating higher knowledge).
Results:
Among participants (n = 676), 61% were Family Medicine residents, 66% were white, and 72% were female. Overall, 34% received <1 hour of EC education, with OB/GYN residents receiving significantly more hours than Family Medicine residents. OB/GYN residents (mean = 14.40, standard deviation [SD] = 2.69) had a significantly higher mean knowledge score than Family Medicine residents (12.12, SD = 2.63; p < 0.000). Mean knowledge score differences were found by region of residency program, with residents in the Northeast reporting higher knowledge.
Conclusions:
Overall, residents received very little EC education, with OB/GYN residents receiving more training and having higher knowledge than their Family Medicine counterparts. Additional training is needed to ensure that residents are knowledgeable about this effective method to decrease unintended pregnancies.
Introduction
Almost half of all pregnancies in the United States are unintended. 1 Of these, 95% occur in women who use contraception inconsistently or not at all. 2 Emergency contraception (EC) is the only way of preventing unintended pregnancy after unprotected sexual intercourse or in the event of failure of planned contraception methods. 3 However, only 11% of sexually active women report ever using EC. The three most effective types of EC, in order of effectiveness, include the copper intrauterine device (IUD), ulipristal acetate, and oral levonorgestrel. 4 Despite being the least effective method, levonorgestrel pills remain the most commonly used method of EC. 4
As health care providers play an important role in EC access, provider-level barriers can influence the use of the most effective methods of EC, and many of these come from a lack of education on available EC methods. While studies have reported positive attitudes toward EC among providers, potential barriers still exist in terms of limited knowledge regarding EC availability, timing of administration, and efficacy of EC in preventing unintended pregnancies. 5 These barriers result in a low prevalence of EC counseling and prescription by providers. 6,7
Residency training can influence counseling and practice behaviors, however, the extent to which EC-related information is taught in training programs remains unknown. In addition, there is little research regarding residents' perceptions about EC. One study of emergency medicine residents noted that those who were first year residents or reported a lack of knowledge about EC were significantly less likely to prescribe EC to their patients. 8 Thus, gaps remain to explore the EC-related knowledge among residents who commonly see reproductive-age women: Family Medicine residents and Obstetrics and Gynecology (OB/GYN) residents.
The most effective EC methods are available only through providers; the copper IUD requires a clinician visit for insertion, and ulipristal acetate requires a prescription. Because of this, understanding and improving EC knowledge among providers are essential to increasing access and use of the most effective EC methods. Residency training can influence practice behaviors, thus exploring the EC-related information taught in training programs is important. This study aims to examine where OB/GYN and Family Medicine residents obtain information about EC and whether knowledge differs by resident program characteristics.
Materials and Methods
The data presented in this article are derived from a larger study aimed at exploring OB/GYN and Family Medicine resident's knowledge, attitudes, and practices regarding EC. The survey instrument, including knowledge items, was developed based on previous literature focused on EC beliefs, knowledge, and counseling and prescribing practices among physicians and residents. 7,9 –11 The instrument was evaluated for content validity by experts in the field piloted with three health care providers, and modified based on their suggestions. The study was considered exempt from the University of South Florida's Institutional Review Board as no identifying information was linked to survey responses.
Sample
For recruitment, a list of 689 OB/GYN and Family Medicine residency programs across the United States was compiled and included 237 OB/GYN programs and 452 Family Medicine programs. The list included contact information for the residency program coordinator of each program. Recruitment for survey participation was guided by the Dillman Method as follows 12 ; a prerecruitment email was sent to the program coordinators, introducing the study and stating that they would soon receive an email with the link to the survey and information for participants, and we asked that they distribute that email to the residents in their program. Two days later, the survey link was sent to the coordinators and included information regarding the study to be sent to residents. Follow-up emails were sent to the program coordinators 2 weeks later and a final email was sent 3 weeks after the initial email. We were unable to calculate the response rate for this study, as it is unclear how many program coordinators received our emails and forwarded our request to their residents. Residents participating in the study were given the option to leave their contact information in a separate survey link to be entered into a raffle for one of four iPads.
Measures
Demographic variables were measured, including resident specialty (OB/GYN or Family Medicine), race, sex, year of residency program, and religious affiliation of residency program (yes, no, unsure; if yes, Catholicism, Judaism, or other). Participants were also asked to select the region of the United States where their residency is located, which included six responses: Northeast, Southeast, Midwest, West, Southwest, and outside of the United States.
To describe resident training regarding EC, participants were asked how many hours of EC education they had received (less than one, one, two, or more than two, which would roughly correspond to the number of lectures received). Participants were also asked to indicate the sources of EC information in their residency program (“Where in your residency have you learned about EC?”), including lectures, grand rounds, required readings, or other sources. Twenty knowledge items were developed regarding three methods of EC (the copper IUD, ulipristal acetate, and oral levonorgestrel), and included effectiveness, mechanism of action, contraindications, and side effects. Eight of the twenty items were multiple choice with five response options, eight items were true/false, and four items were yes/no responses.
Analysis
Descriptive statistics were calculated for demographic variables, hours of EC education, sources of EC information, and each individual knowledge item. Bivariate testing, including chi-square tests and z-tests to compare column proportions, were conducted to analyze resident specialty, religious-affiliation of residency program, and region of residency and their relationship between hours of education, sources of information, and knowledge (correct or incorrect response). A p-value of 0.05 was used to indicate significance, and Bonferroni adjustments were applied to p-values when multiple tests were conducted (residency region). Some items had missing responses and therefore the analytic sample size varied based on the item. A knowledge score was calculated with one point for each correct answer and no points for an incorrect answer (minimum = 0, maximum = 20) and t-test and analysis of variance tests were used to compare mean knowledge score by specialty, religious-affiliation, and residency region. To compare mean knowledge scores by region, post hoc Tukey Honest Significant Difference (HSD) tests were calculated for significant pairwise comparisons. All analyses were conducted in SPSS.
Results
Of the 676 participants, 61% were Family Medicine residents, 66% were white, 72% were female, most were in their first, second, or third year of residency, and 83% were in residency programs that were not religiously affiliated (Table 1).
Demographic, Resident, and Residency Program Characteristics of Sample
Missing three responses for this variable.
Of those who reported yes to religious program.
PGY, postgraduate year; OB/GYN, Obstetrics and Gynecology.
Overall, hours of EC education received in residency programs varied, with 34% of the sample receiving <1 hour of education (Table 2). There were significant differences in hours of education received by specialty, with 39% of OB/GYN residents reporting more than 2 hours of EC education compared to 18% of Family Medicine residents. Significant differences in hours of education received were found by religious affiliation of program, but no significant differences were found by residency region.
Hours of Emergency Contraception Education and Sources of Emergency Contraception Education (Yes to Receiving Information from this Source) by Resident Specialty, Religious Affiliation of Residency Program, and Residency Region, n (%)
Subset to those who responded yes or no to religious program.
Indicates significant difference between OB/GYN and Family Medicine residents at p < 0.05.
Indicates significant difference between religious and non-religious programs at p < 0.05.
Indicates significant difference by residency region at p < 0.05.
FM, family medicine; NE, northeast; SE, southeast; MW, midwest; SW, southwest.
Almost 74% of participants in the sample reported that they received EC information from lectures and 71% received information on EC from other sources. OB/GYN residents were significantly more likely than Family Medicine to report receiving information from all four sources (lectures, grand rounds, required reading, and other sources). Residents in nonreligiously affiliated programs received significantly more education from lectures, grand rounds, and required readings compared with residents in religiously affiliated programs, but there were no differences in receipt of information from other sources. Residents who received EC information from required readings and other sources differed by region.
The overall mean knowledge score for the sample was 13.01 (standard deviation [SD] = 2.87). Knowledge items, response options, and percentages for each response option overall and by specialty are presented in Table 3. The items most participants answered incorrectly include how long after intercourse levonorgestrel is considered effective (18% correct), ranking the methods in order of effectiveness (23% correct), mechanism of action of ulipristal acetate (29% correct), length of time after intercourse ulipristal acetate is considered effective (30% correct), and mechanism of action for levonorgestrel (32% correct).
Mean Knowledge Score, Knowledge Scale Items, and Response Options for Each Item Among the Overall Sample and by Specialty, n (%)
Indicates correct answer.
Indicates significant difference in correct response between OB/GYN and Family Medicine residents at p < 0.05.
SD, standard deviation; EC, emergency contraception; IUD, intrauterine device; FDA, Food and Drug Administration; DVT, deep vein thrombosis.
When comparing resident knowledge by specialty, OB/GYN residents (mean = 14.40, SD = 2.69) had a significantly higher mean knowledge score than Family Medicine residents (12.12, SD = 2.63; p < 0.000). For 13 of the 20 items, OB/GYN residents were significantly more likely to respond correctly than Family Medicine residents (p < 0.05), including items regarding mechanism of action, time after intercourse methods are effective, contraindications, and side effects. However, Family Medicine residents were significantly more likely to correctly indicate that a common side effect of EC is a headache than OB/GYN residents (90% and 80%, p = 0.000). There were no significant differences in correct responses to the knowledge items based on religious affiliation of residency program (data not shown).
In addition to differences in knowledge by specialty, mean knowledge score differences existed by region of residency program. Mean knowledge scores were significantly different (p < 0.000), and post hoc testing with Tukey's HSD tests indicated significant differences between knowledge scores of residents in the Northeast (13.55, SD = 2.90) and the Southeast (12.22, SD = 2.67; p < 0.000); residents in the West (13.77, SD = 2.75) compared with residents in the Southeast (p < 0.000); and residents in the West compared with residents in the Midwest (12.82, SD = 2.90; p = 0.04).
In addition, there were differences in correct responses for 5 of the 20 knowledge items, however, only three remained significant after Bonferroni adjustments (p < 0.01) (Table 4). Responses to the item regarding the primary mechanism of action of levonorgestrel indicated residents in the Northeast (37% correct), Midwest (34% correct), and West (36% correct) were significantly more likely to respond correctly than residents in the Southeast (20% correct). There were no significant differences between these regions and the Southwest on this item. Residents in the Northeast were significantly more likely to correctly respond to the length of time after intercourse ulipristal acetate is considered effective than residents in the Southeast (38% compared to 22%). Similarly, residents in the Northeast (55%) were significantly more likely to correctly identify the length of time after intercourse that the copper IUD can be inserted compared to residents in the Southeast (36%) and residents in the Midwest (40%). There were significant differences in correct responses to items on the primary mechanism of action of the copper IUD and that a common side effect of EC is infertility, however, findings were no longer significant after Bonferroni adjustments in post hoc testing.
Knowledge Items, Percent Correct by Residency Region
Subset analysis–removed two missing observations and two observations who were in residency outside of the United States, n = 672.
Tukey HSD indicates significant difference (p < 0.05) between NE and SE.
Tukey HSD indicates significant difference (p < 0.05) between SE and West.
Tukey HSD indicates significant difference (p < 0.05) between MW and West.
Indicates significant difference by residency region at p < 0.05.
Indicates significant difference by residency region after Bonferroni Adjustment (p < 0.01).
HSD, Honest Significant Difference.
Discussion
OB/GYN and Family Medicine physicians are the primary providers of contraceptive services and therefore need to be informed about the most effective contraceptive methods. EC is an important contraceptive option in the case of unprotected sex or contraceptive failure, and provider knowledge is necessary for appropriate patient contraceptive counseling. Our study showed that there are significant gaps in knowledge in residency training with respect to EC.
The finding that Family Medicine residents received significantly less education in EC than OB/GYN residents is reflected in their knowledge scores, which were significantly lower. In addition to Family Medicine residents receiving less education, those residents attending religiously affiliated programs also received less training, with almost half of them receiving <1 hour of training on EC. This result is similar to those in other studies of family practice providers and their hospital affiliation, which show that those in religiously affiliated programs prescribed less EC and reported receiving just 1 hour of education on contraception. 13 In this study, residents in religiously affiliated programs were less likely to learn about EC during a grand rounds lecture. This is also similar to a study of residents in religiously affiliated hospitals, which found that no teaching afternoons, grand rounds, or formal education on EC was provided. 13 These findings suggest a need to work within religiously affiliated programs to ensure that residents receive appropriate education on the most effective methods of contraception, including EC. While knowledge alone may not directly influence practice behaviors, it is a necessary component. Educating residents on EC is needed so that they are able to provide accurate information and contraceptive services in their future professions.
Residents who receive less training and have lower knowledge scores may be less likely to provide EC once becoming practicing physicians. 3 Even if they discuss and provide EC, there may be factual inaccuracies in their patient education, negating the goal of preventing unintended pregnancies. 9 Specifically, over half of residents in this study believe that copper IUDs are the least effective emergency contraceptive, while in fact they are the most effective. 4 Knowledge gaps such as this may prevent physicians from discussing and offering the copper IUD as a form of EC. 9
Education in EC is an Accreditation Council for Graduate Medical Education (ACGME) requirement for both OB/GYN and Family Medicine residency programs. In 1999, the Kenneth J. Ryan Residency Training Program in Abortion and Family Planning (Ryan Program) was created as a privately funded national initiative to help meet the ACGME mandate for comprehensive training in abortion and contraception in OB/GYN residencies. 14 In addition, the Reproductive Health Education in Family Medicine (RHEDI) Program was developed in 2004 to provide similar training in Family Medicine residency programs. 15 Continued support and expansion of these programs are needed to increase education on contraception, including EC.
For residencies without a Ryan or RHEDI program, ongoing review of the curriculum is needed to guarantee that there is appropriate time within the didactic curriculum dedicated to EC. Residencies may consider bringing in experts in the field to give a grand rounds lecture on EC, particularly those programs providing <1 hour of education or those without a grand rounds addressing EC. In addition, future research should identify the various barriers as to why EC training remains a gap in residency programs. Results from this study can inform the development and testing of innovative methods to deliver EC training to reach all residents. Such methods could include technology-assisted training (i.e., virtual training or other applications) to ensure that the EC education requirement is met. 16
Findings of this study must be considered in light of limitations. Due to convenience sampling, a response rate was not calculated for this survey. Response bias is an additional limitation; those with strong positive or negative feelings toward EC may have been more likely to answer the survey questions potentially confounding the results. In addition, we did not link identifying characteristics of participants with their survey responses, including their specific residency program, and therefore cannot stratify respondents from these programs for comparison. Future studies should explore these residency programs as influential factors. However, the strengths of this study are that we were able to survey a large sample of residents by contacting the residency coordinators of all accredited OB/GYN and Family Medicine residencies in the United States. Subsequently, participants represent a diverse population, including all regions of the country and year of residency.
Conclusions
In summary, training gaps in EC remain for OB/GYN and Family Medicine residents. This is reflected in their EC-specific knowledge, which can contribute to inefficient or incorrect patient counseling. Additional training is needed so that residents are sufficiently knowledgeable about EC as an effective contraceptive method under circumstances where women are at high risk for unintended pregnancy.
Footnotes
Acknowledgments
This work was supported by the University of South Florida Women's Health Collaborative Grant. The funders did not play a role in data collection, analysis, interpretation of data, the writing of the report, or the decision to submit the article for publication.
Author Disclosure Statement
No competing financial interests exist.