Teaching Women's Health Skills

Abstract

OBJECTIVE

This study assesses the readiness of academic general internists to perform and precept a commonly utilized women's health examination, and procedural and management skills.

DESIGN

Full-time faculty from divisions of general internal medicine and departments of family practice in 9 states reported their encounter frequency with, comfort precepting, and the importance they ascribe to several examination, procedural, and management skills relevant to women's health care; and their attitudes toward performing the pelvic exam and obtaining a Pap smear.

MEASUREMENTS AND MAIN RESULTS

A total of 331 general internal medicine physicians (GIMs) and 271 family medicine physicians (FPs) completed questionnaires, with response rates of 57% and 64%, respectively. More than 90% of GIMs and FPs indicated they were confident precepting the breast and Pap/pelvic examinations. A relatively small percentage of GIMs expressed confidence precepting the management of dysfunctional uterine bleeding (22%), initiating Depo-Provera (21%), and initiating oral contraceptives (45%), while a substantially larger percentage indicated that these skills were important to primary care practice (43%, 44%, and 85%, respectively). Although GIMs indicated they were confident precepting the Pap/pelvic exam, they were less likely than FPs to agree with the following statements: “Performing routine Pap smears is a good use of my time” (GIMs 65%, FPs 84%); “It is a waste of health care dollars for primary care physicians to refer patients to gynecologists for routine Pap/pelvic exams” (GIMs 69%, FPs 90%); “I feel very well trained to do a routine bimanual exam” (GIMs 71%, FPs 98%), and “The clinic where I practice is well equipped to do a Pap smear” (GIMs 78%, FPs 94%).

CONCLUSIONS

Although most academic GIMs are confident precepting the breast and pelvic examination, only a minority are confident precepting the management of dysfunctional uterine bleeding, initiating Depo-Provera, and initiating oral contraceptives. These findings suggest that a number of academic GIMs may not be prepared or willing to perform or precept important women's health skills.

The American Board of Internal Medicine (ABIM) acknowledges that the residency training environment does not always facilitate optimal care of women,¹ and that it has not insisted that all trainees be competent in examining the female urogenital tract,² and capable of diagnosing and treating common gynecologic problems.^3–5 Recognizing that training in women's health must improve in order to correct deficiencies in the medical care provided to women, the ABIM's commentary, “Training Internists in Women's Health,”² recommends assessing trainee competence in women's health, including women's health topics in didactic presentations to residents, and exploring the benefits of cross training between obstetrics/gynecology and family medicine. Furthermore, the Board's subcommittee on Clinical Competence in Women's Health has recently identified a set of core competencies that all internists seeking board certification should demonstrate.⁶ These competencies include management of common gynecologic disorders and family planning, as well as the following essential skills: breast exam, pelvic exam/rectal exam, and obtaining a Pap smear. New Internal Medicine Residency Review Committee gender-specific special requirements⁷ mandate that those who commence training subsequent to July 1997 demonstrate competence with the breast, pelvic, and rectal exams. These requirements also suggest that residents, depending on their practice expectations, should be trained in several additional women's health management and procedural skills.

Expanding residency training in women's health likely will require faculty development⁸ for those who received little exposure to these skills during their own training or for faculty who do not routinely utilize these skills.⁹ This study assessed the readiness of academic general internal medicine physicians (GIMs) and family medicine physicians (FPs) to perform and precept a commonly utilized women's health care examination and procedural and management skills, and assessed their attitudes toward performing the pelvic exam and obtaining a Pap smear. Contrasting GIM and FP faculty responses allowed us to compare faculty from a discipline that has recently emphasized women's health residency education, internal medicine, with a discipline that has emphasized women's health education for many years, family practice. The American Academy of Family Practice has had Women's Health Core Educational Guidelines since 1994,¹⁰ has required obstetrics/gynecology training since its inception, and has always required a substantial amount of ambulatory training experience in which women's health issues are most often encountered.

METHODS

Study Participants

Full-time faculty from the divisions of general internal medicine and the departments of family practice at all 37 medical schools in the 9 states (Fla, Ga, Ky, NC, Ohio, Pa, Ind, Va, and WVa) represented by the investigators were invited to participate in this survey. Two medical schools declined to participate. A questionnaire was mailed to GIMs and FPs at the 35 remaining schools. Nonrespondents received a reminder postcard 2 weeks after the initial mailing and a second letter and questionnaire at 6 weeks.

Questionnaire Design

Items for the questionnaire were selected to represent a sample of women's health skills considered important, or expected to grow in importance for the practice of general internal medicine. The skills were chosen based on a review of the literature and author consensus. Most of these skills have since been included as part of the ABIM's Core Competencies in Women's Health⁶ and the Federated Council of Internal Medicine Task Force recommendations for in-depth knowledge of conditions in women's health.¹¹

The questionnaire addressed the examination and procedural and management skills (see Table 2). For each skill, respondents estimated the number they performed or precepted each month or year, their confidence precepting these skills, and the importance of these skills to primary care practice. Confidence and importance were measured using Likert scales where 1 = not confident at all or not important at all, and 6 = very confident or very important. In addition, using Likert scales where 1 = strongly disagree and 6 = strongly agree, participants were asked to indicate their level of agreement with 19 statements relevant to performing the pelvic examination and obtaining a Pap smear. Each statement regarding the Pap/pelvic exam is designed to assess either the importance respondents place on performing the exam, the quality of their Pap/pelvic training, their confidence performing the exam, the extent to which their practice setting facilitates performing the exam, or their opinion on patient preferences for who performs this exam.

Table 2.

Examinations, Procedures, and Management Encounters per Month or Year: Median Numbers Performed and Percentages of Physicians Performing Any and 10 or More^*

Examination, Procedure, or Management Encounter	Physician Specialty		Physicians Performing Any, %	Physicians Performing ≥10, %
Median number/mon
Routine PAP/pelvic	GIM	10†	92	53
	FP	20	98	83
Routine breast exam	GIM	20	96	73
	FP	20	99	86
Median number/y
Endometrial biopsy	GIM	0†	3	1
	FP	5	71	25
Initiating oral contraceptive therapy	GIM	5†	83	43
	FP	20	98	85
Initiating Depo-Provera	GIM	0†	42	16
	FP	6	93	47
Initiating hormone replacement therapy	GIM	24	98	84
	FP	24	100	85
Managing dysfunctional uterine bleeding	GIM	5†	88	30
	FP	6	99	46
Managing major depression in a middle-aged woman without suicidal ideations	GIM	15	98	65
	FP	12	99	65

^*Median numbers reported include all respondents; statistical comparisons between the 2 groups were performed by the Wilcoxon-Mann-Whitney rank sum test for median numbers only; differences are not significant unless otherwise noted.

^†P < .005.

GIM, general internal medicine; FP, family practice.

Statistical Analysis

The Wilcoxon-Mann-Whitney rank sum test was used to compare the median number of procedural and clinical encounters between GIMs and FPs. The Fisher's exact test was used to compare by discipline importance of performing/managing, confidence precepting, early referral, and the percentage of physicians who strongly agreed with the 19 Pap/pelvic statements. Binary logistic regression was used to model the probability of having a high level of confidence precepting women's health management skills. High confidence was defined as a Likert score of 5 or 6, while low/medium confidence was defined as a Likert score of 1 to 4. The main independent variables (all categorical) included: number of procedures/conditions the physician performed/encountered (grouped into 0, 1–9 and ≥10 procedures per month or year); self-reported importance of performing/managing conditions that were classified as having a high level of importance (Likert score 5 or 6) versus a low level of importance (Likert score 1–4); physician type (GIM vs FP); frequency of early referral to another health care professional; and gender. Additional independent variables regarding physician attitudes concerning the Pap/pelvic exam were included when the binary logistic regression model was built for that examination. A forward selection procedure was used, with a P value of ≤.05 as the criterion for entry into the model. Goodness-of-fit tests were performed, and the results supported the adequacy of the final models.¹² All analyses were conducted using SAS software, Version 6.12 (SAS Institute, Cary, NC). Two-sided P values and 95% confidence intervals (95% CIs) are reported.

RESULTS

Demographics

Surveys were mailed to 571 GIMs and 434 FPs. A total of 331 GIMs and 271 FPs returned completed confidential questionnaires for response rates of 57% and 65%, respectively. Forty-five GIMs who identified a Veterans Affairs hospital clinic as their primary ambulatory practice site, where the majority of patient encounters were assumed to be with men, were excluded from the analyses. As seen in Table 1, 40% and 30% of the GIM and FP respondents, respectively, were female. The majority of physicians from both groups described their current practice setting as university based, and 87% of GIMs and 54% of FPs reported training in a university program. Both groups indicated that just over 50% of their time was spent in ambulatory practice, and that one half of this time was spent practicing with residents.

Table 1.

Respondent Characteristics by Specialty

Characteristic	Internal Medicine (N = 286)	Family Practice (N = 271)
Mean age, y (SD)	41 (8.3)	44 (9.2)
Women, %	40	30
Average year residency completed (SD)	1985 (8.1)	1983 (9.0)
Post-graduate program location, %*
University	87	54
Community	13	45
Current practice type, %†
University-based practice	80	69
Private practice	16	18
Community health center	3	6
Public clinic	6	2
HMO	2	2
Other	6	13
Current practice location (population), %
Large city/metropolitan area (>200,000)	64	46
Large town/small city (50,000–200,000)	24	33
Rural/small town (<50,000)	12	21
Time spent practicing, %
Outpatient
With residents	24	26
Without residents	27	27
Inpatient
With residents	15	8
Without residents	4	3
Payor mix, %
Medicare	33	24
Indemnity insurance HMO	35	46
Indigent	16	10
Medicaid	18	19

^*Because of rounding, percentages may not all total 100.

^†Respondents could select all that apply.

Skills

The median number of women's health examination, procedures, management encounters by physician specialty, and the percentage of physicians reporting these encounters at all or ≥10 times per month or year is shown in Table 2. The percentage of physicians by specialty who attribute importance to, report confidence precepting, or refer early for these same women's health skills is shown in Figure 1.

FIGURE 1.

Percentage of respondents choosing 5 or 6 (1 = strongly disagree, 6 = strongly agree) for importance of performing, confidence precepting, and almost always refer early for specific women's health care skills. ^†P, not significant; ^‡P < .05; ^§P < .01; P < .001 unless otherwise noted.

Examination Skills

Both groups reported similar frequencies for performing routine breast exams (median 20/mo), while FPs performed routine Pap/pelvic exams more frequently (FP, 20/mo; GIM, 10/mo). Over 95% of both groups (Fig. 1) indicated that performing these exams was an important primary care skill, and over 90% of GIMs and FPs reported they were confident precepting the breast and Pap/pelvic exams.

Procedural Skill

Only 1% of GIMs reported performing endometrial sampling 10 or more times per year, and only 13% indicated that this was an important primary care skill. In contrast, 25% of FPs reported performing endometrial sampling 10 or more times per year, and the majority of FPs indicated that this was an important primary care skill that they were confident precepting.

Management Skills

Less than half (43%) of GIMs reported initiating oral contraceptives 10 or more times per year and even fewer (16%) reported initiating Depo-Provera 10 or more times per year. In contrast, 85% of FPs reported initiating oral contraceptives and 47% reported initiating Depo-Provera 10 or more times per year. The percentage of GIMs indicating they were confident with these skills (oral contraceptives, 45%; Depo-Provera, 21%) was substantially less than the percentage indicating that the skill was important (oral contraceptives, 84%; Depo-Provera, 44%). The median number of hormone replacement therapy encounters was 24 per year for both FPs and GIMs, and both physician groups believed this was an important primary care skill that they were confident precepting.

Dysfunctional uterine bleeding was encountered more commonly by FPs. Eighty-nine percent of FPs indicated that evaluating dysfunctional uterine bleeding was an important skill, and 78% of FPs reported confidence precepting the skill. In contrast, 43% of GIMs believed this skill was important, and only 22% reported confidence precepting this presentation. Sixty-one percent of GIMs compared to 5% of FPs would seek early referral for a patient with this condition.

Physician Attitudes Toward Pap/pelvic Exam

The percentage of physicians by discipline who strongly agreed (5 or 6 on a scale of 1 to 6) with a series of statements pertaining to the Pap/pelvic exam is shown in Table 3.

Table 3.

Attitudes Regarding Pap Smear/Pelvic Exam: Percentage of Physicians Ranking Highest Scores (5 or 6) for Each Item (1 = Strongly Disagree; 6 = Strongly Agree)^*

	GIM	FP
Priority and importance
It is the primary care provider's responsibility to perform routine Pap smears.	90	99
My patients tend to have so many other medical problems that a routine pelvic exam is low on the list of priorities.	13	4
I wish I could skip the pelvic exam.	17	5
Performing routine Pap smears is a good use of my time.	65	84
Routine Pap smears are an important part of female preventive health services.	98	99†
It is a waste of health care dollars for primary care physicians to refer patients to gynecologists for routine Pap smears.	69	90
Training
I feel very well trained to do a routine Pap smear.	86	99
I feel very well trained to do a routine bimanual exam.	71	98
By the completion of my residency, I had performed enough pelvic exams under supervision to feel confident doing them.	49	90
Confidence
I am confident I can obtain a sufficient endocervical sample most of the time when I do a Pap smear.	88	97
I worry that I may be missing important findings when I do a pelvic exam.	15	9‡
I know female anatomy well enough to detect common abnormalities on a pelvic exam.	66	96
I can reliably determine which abnormal Pap smear results need to be referred to a gynecologist.	73	96
Gender and discipline
Women prefer female physicians to perform their pelvic exam.	45	33
Female physicians are better at doing pelvic exams because they are more aware of patient concerns.	20	9
Most of my patients prefer to get their routine Pap smears done by a gynecologist.	13	4
Facilities
The clinic where I practice is well equipped to do a Pap smear.	78	94
My clinic has ample nursing staff available to assist me in obtaining a Pap smear.	64	75
I usually schedule a separate office visit to obtain a routine Pap smear.	18	36

^*P < .001 unless noted otherwise; P values were calculated using Fisher's Exact test.

^†P, not significant.

^‡P = .025.

GIM, general internal medicine; FP, family practice.

Priority/importance

The vast majority (≥90%) of respondents from both groups strongly agreed that the Pap smear was an important component of female preventive health services, and that it was the primary care physician's responsibility to perform routine Pap smears. Substantially fewer internists (GIMs 65%, FPs 84%) agreed that performing the routine Pap smear was a good use of their time. Furthermore, 90% of FPs, but only 69% of GIMs, agreed that “it is a waste of health care dollars for primary care physicians to refer patients to a gynecologist for routine Pap smears.”

Training

General internists expressed substantially less agreement than FPs for each of the statements concerning their training to perform the pelvic exam and obtain a Pap smear. Only half of GIMs agreed with the statement that “by the end of residency, I had performed enough pelvic exams under supervision to feel confident performing this exam.”

Confidence

Although only 66% of the GIMs strongly agreed they know female anatomy well enough to detect common abnormalities on pelvic exam, and 73% that they could reliably determine which abnormal Pap smear results need to be referred to a gynecologist, 96% of FPs expressed strong agreement with both statements.

Gender

General internists were more likely to agree that “women prefer female physicians to perform their pelvic exam” (GIMs, 45%; FPs, 33%) and that “female physicians are better at pelvic exams because they are more aware of patient concerns” (GIMs, 20%; FPs, 9%). When the respondents were compared by gender and not physician type (data not shown), 69% of female physicians compared with 23% of male physicians agreed that female patients prefer female physicians to perform their pelvic exams. Thirty percent of female respondents and only 9% of males agreed that female physicians were better at pelvic exams.

Facilities

General internists were less likely to agree that their practice facility was well equipped to perform a Pap smear (GIMs, 78%; FPs, 94%) and that their practice facility had ample nursing staff to assist with the pelvic exam (GIMs, 64%; FPs, 75%). FPs (36%) were twice as likely to agree with the statement, “I usually schedule a separate office visit to obtain a routine Pap smear.”

Binary Logistic Regression Model

Binary logistic regression was used to model the probability of having a high level of confidence precepting 6 women's health care skills (Table 4). Attributing importance to a skill was predictive of confidence precepting the skill, except for initiating oral contraception. FPs were more likely to have a high level of confidence precepting each skill, except endometrial sampling. The nonsignificance of physician type for endometrial sampling was due to its high correlations with 1 or more other variables in the model, such as importance and number of procedures performed. In general, odds ratios (ORs) were highest for performing 10 or more procedures per year, emphasizing the contribution of performing procedures frequently to a high level of confidence teaching the procedures. Female gender was predictive of having a high level of confidence precepting the skills of endometrial sampling, evaluation of dysfunctional uterine bleeding, initiating Depo-Provera, and initiating oral contraceptives. There was an inverse relationship between early referral and a high level of confidence precepting initiating oral contraceptives, initiating Depo-Provera, initiating hormone replacement therapy, and management of depression.

Table 4.

Predictors of High Confidence in Precepting (Likert Score 5 or 6) Specific Women's Health Care Skills: Final Binary Logistic Regression Models

		Adjusted* Odds Ratios (95% CI)
Predictor Variables	Reference Group	Endometrial Sampling	Initiation of OCT	Initiation of Depo-Provera	Managing DUB	Manage Major Depression	Initiation of HRT
Importance, Likert 5 or 6	Likert 1–4	7.9 (2.7 to 23.3)	†	6.8 (3.0 to 15.2)	11.6 (6.0 to 22.4)	6.8 (2.7 to 17.1)	6.1 (1.5 to 25.6)
Physician type, FP	GIM	1.3 (0.3 to 6.6)	20.2 (9.3 to 44.0)	7.7 (3.8 to 15.8)	7.4 (4.3 to 12.7)	2.8 (1.4 to 5.5)	2.4 (1.1 to 5.2)
Number performed, 1–9/y	0	22.0 (4.3 to 111.6)	5.6 (1.4 to 21.6)	4.7 (1.7 to 12.7)	11.3 (1.2 to 106.2)	1.9 (0.1 to 26.9)	10.8 (0.8 to 149.1)
Number performed, ≥10/y	0	98.4 (10.6 to 909.7)	18.6 (4.9 to 70.8)	9.1 (3.1 to 26.5)	18.1 (1.9 to 172.0)	4.3 (0.3 to 61.2)	41.4 (3.1 to 545.7)
Gender, female	Male	5.4 (1.5 to 18.8)	3.4 (1.8 to 6.4)	3.8 (1.7 to 8.1)	2.4 (1.4 to 4.3)	†	2.4 (1.0 to 5.6)
Early referral, Likert 5 or 6	Likert 1–4	‡	0.08 (0.01 to 0.47)	0.05 (0.01 to 0.21)	†	0.37 (0.17 to 0.80)	0.018 (0.001 to 0.223)

^*Adjusted for the other variables in the final model; only primary independent variables are reported.

^†Indicates predictors that are not included in the final model.

^‡Early referral not assessed for this procedure.

CI, confidence interval; GIM, general internal medicine; FP, family practice; OCT, oral contraceptive therapy; IM, intramuscular; DUB, dysfunctional uterine bleeding; HRT, hormone replacement therapy.

Similar binary logistic regression modeling was employed to explore predictors of a high level of confidence precepting PAP/pelvic exams (data not shown). Significant predictors of a high level of confidence precepting the Pap/pelvic exam included feeling well prepared to perform a Pap smear (OR, 5.6; 95% CI, 1.6 to 19.7), feeling well trained to do a bimanual exam (OR, 12.4; 95% CI, 2.8 to 54.4), and indicating that performing routine Pap smears is a good use of time (OR, 7.9; 95% CI, 2.4 to 25.5).

DISCUSSION

Recently, the ABIM endorsed a set of women's health core competencies that should be demonstrated by trainees who seek board certification.⁶ Studies showing that GIMs often lack the skills or knowledge to detect pelvic masses, switch oral contraceptives,⁵ or manage aspects of the menopause¹³ serve to amplify concerns that many internal medicine residents may receive inadequate training in women's health. In order to strengthen women's health training, it is important to understand GIM faculty comfort with, attitudes toward, and ascribed importance to a variety of women's health care skills. Our study begins to address these issues by surveying a sample of GIM and FP faculty who spend a significant amount of time precepting residents in ambulatory practice.

Our results indicate that a large percentage of GIMs (≥90%) expressed confidence precepting the breast, pelvic, and Pap smear exams. In contrast, although the vast majority of GIMs indicated that initiating oral contraceptive therapy is an important primary care skill, and a substantial number of GIMs acknowledged that the evaluation of dysfunctional uterine bleeding and initiating Depo-Provera are important primary care skills, a much smaller percentage expressed confidence precepting each of these skills. We refer to this disparity in ascribed importance and reported confidence precepting as an importance/confidence mismatch. While this lack of confidence might reflect inadequate training, it also may reflect too few clinical opportunities, available or taken advantage of, to maintain a sense of proficiency. It should be noted that in our binary logistic model, the frequency with which a provider performs the skill is a strong predictor of confidence precepting the skill. One might speculate that this importance/confidence mismatch also reflects a growing perception within the academic community that educational and marketplace demands necessitate that GIMs manage aspects of women's health care that previously were referred to other providers.

A substantial difference exists between FPs and GIMs in the importance they ascribe to managing dysfunctional uterine bleeding that cannot be attributed to the small, albeit significant, difference in encounter frequency between disciplines. Moreover, only 13% of GIMs considered the procedural skill, endometrial sampling, to be an important primary care skill. This difference between disciplines in the importance placed on the management of dysfunctional uterine bleeding, which in some cases requires endometrial sampling, may reflect the medical orientation of internists as opposed to the procedural orientation of family physicians and gynecologists.

The responses by GIMs to the statements “performing the Pap smear is a good use of my time” and “it is a waste of health care dollars to refer a routine Pap smear to a gynecologist” suggest that GIMs place less of a priority, relative to FPs, on being the provider who actually performs this exam. GIMs responses to several statements regarding their training to perform the Pap/pelvic exam and their confidence performing this exam may explain this finding. The perception GIMs have of competing demands from medically complex patients and less than optimal facility support also might diminish their enthusiasm for performing the Pap/pelvic exam.

Others have reported that patients often express a preference for a physician of the same gender particularly when a rectal or genital exam is required,^14,15 and in our survey, 45% of the GIM respondents in contrast to 33% of FM respondents agreed with the statement “women prefer female physicians to perform their pelvic exams.” Interestingly, 69% of female respondents and 23% of male respondents agreed with this statement. The impact of these faculty beliefs, as a function of discipline and gender, on the opportunity for male trainees to obtain clinical experience performing pelvic exams is not known.

There are several limitations to this study. The sample is limited to full-time FP and GIM medicine faculty at academic health centers in 9 Eastern states, and no attempt has been made to corroborate the self-reported frequencies of women's health skill encounters with data from the medical record. In addition, demographic data were not available for nonresponders, so comparisons between responders and nonresponders could not be performed. Nonresponders may be different with regard to their practices and attitudes concerning women's health skills. Although the survey respondents assessed their own confidence precepting the various skills, we did not measure their competence. Obtaining additional patient demographic information might have permitted us to better understand some of the differences in encounter frequencies noted by discipline.

The challenge that the discipline of internal medicine faces in its efforts to enhance training in women's health is exemplified by noting that as recently as 1994, less than 60% of primary care internal medicine program directors believed that the majority of their own residents had mastered the Pap smear.¹⁶ Indeed, our own results suggest that barriers exist that interfere with GIM faculty performing the Pap smear and pelvic exam. We anticipate that the opportunity for GIM faculty to cultivate expertise in providing and teaching a broad range of women's health skills may be hindered if these barriers are not addressed. Although we found that the vast majority of academic GIMs are confident precepting the breast exam, pelvic exam, initiating hormone replacement therapy, and managing major depression, an importance/confidence mismatch was identified among GIM but not FP respondents for 3 of the 8 women's health skills we evaluated. This finding suggests that a number of academic GIMs may not be prepared to perform or precept certain women's health skills. It will be important for residency programs to consider whether similar mismatches exist among faculty for the larger set of ABIM women's health core competencies.⁶

As the discipline of internal medicine seeks to improve training in women's health, residency programs will need to avoid relying primarily on training experiences provided by other disciplines. We believe it is important that residents' continuity clinics, under the direction of GIMs, reinforce the women's health skills learned by residents during other rotations and allow GIMs to role-model the essential procedures (breast exam, pelvic/rectal exam, and obtaining a Pap smear) along with many of the other ABIM-recommended core competencies in women's health.⁶ The extent of the importance/confidence mismatch for the ABIMs more extensive list of women's health skills and the ability of academic GIMs to address this mismatch through additional focused training will be important factors determining which skills can be taught by GIMs in resident clinics. It will be important to carefully evaluate the impact of the various approaches taken to enhance the ability of GIM faculty to teach women's health skills.