OBJECTIVE: To evaluate the impact of diabetes on provision of contraceptive counseling.
METHODS: We compared counseling provided to diabetic and nondiabetic women on 40,304 visits made to U.S. ambulatory practices by nonpregnant women, 14–44 years of age, included in the National Ambulatory Medical Care Survey and National Hospital Ambulatory Medical Care Survey, 1997–2000. Logistic regression, adjusting for age, race, ethnicity, insurance status, and provider characteristics was used to estimate the relationship between provision of contraceptive counseling and diabetes.
RESULTS: Visits made by diabetic women of reproductive age were significantly less likely to include contraceptive counseling than visits made by nondiabetic women of reproductive age (odds ratio [OR] 0.42, 95% confidence interval [CI] 0.21–0.81). Visits made by diabetic women under 25 years of age were less likely to include contraceptive counseling than visits made by older diabetic women (OR 0.17, 95% CI 0.06–0.54). Overall, only 4% of visits made by diabetic women documented contraceptive counseling. When family planning was the primary reason for a visit (OR 34.4, 95% CI 20.8–56.9) or women visited a gynecologist (OR 24.3, 95% CI 16.7–35.2), women were significantly more likely to receive contraceptive counseling. However, diabetic women made only 0.3% of all visits to gynecologists.
CONCLUSION: Ambulatory physicians in the United States rarely provide contraceptive counseling to diabetic women. This may contribute to adverse birth outcomes for some diabetic mothers who conceive before optimal glucose control is obtained.
LEVEL OF EVIDENCE: II-3
Diabetic women in the United States receive less contraceptive counseling than nondiabetic women.
From the 1Department of Medicine, Division of General Internal Medicine, Center for Research on Health Care, University of Pittsburgh, Pittsburgh, Pennsylvania; and 2Department of Medicine, Division of General Internal Medicine, University of California, San Francisco, California.
The National Center for Health Statistics designed and conducted the National Ambulatory Medical Care Survey (NAMCS) and National Hospital Ambulatory Medical Care Survey (NHAMCS). All data used in this analysis were collected by the National Center for Health Statistics.
Corresponding author: Eleanor Bimla Schwarz, MD, MS, Center for Research on Health Care, 230 McKee Place, Suite 600, Pittsburgh, PA 15213; e-mail: Schwarzeb@upmc.edu.
Individuals with diabetes require counseling regarding the impact of diet, exercise, and medications on disease progression and clinical outcomes. Diabetic women of reproductive age require additional counseling because birth defects occur in about 5–8% of their offspring,1 approximately twice the rate in the general population.2–5 With tight glycemic control before and during pregnancy, rates of congenital malformations and birth trauma can be minimized.6–8 Although about half of all pregnancies in the United States are unintended,9 nearly two thirds of pregnancies in diabetic women are unplanned.10,11 The goal of our study was to estimate whether contraceptive counseling provided to diabetic women of reproductive age differs from that provided to nondiabetic women of reproductive age.
MATERIALS AND METHODS
The National Ambulatory Medical Care Survey is an annual survey of approximately 3,000 office-based U.S. physicians (including obstetrician-gynecologists) who are not employed by the federal government.12 The National Ambulatory Medical Care Survey uses a 3-stage probability sampling procedure to allow extrapolation of survey results to the U.S. population. The first stage contains 112 geographic primary sampling units. The second stage consists of a probability sample of practicing physicians selected from each geographic sampling unit. The third stage involves the random selection of 25 visits to the sample physician during a randomly assigned 1-week reporting period.13 Participants are instructed to record data by using a standardized form shortly after each patient visit to minimize the potential for recall bias. Data collected includes demographic information, the reason for patient’s visit (up to 3 entries), whether the provider or anyone in the practice has seen this patient before, physician’s diagnoses for this visit (up to 3 entries), and medications ordered, supplied, administered, or continued during this visit (up to 6 entries). In addition check boxes are used to indicate whether any of 12 counseling or education services were provided. The time frame of 1997–2000 was selected because it was the most recent period during which all participants were asked if the counseling/education provided included “family planning/contraception.” Information about the patients’ race and ethnicity is based on physician perception. By combining data from the 1997–2000 surveys, we obtained a sample of 17,322 visits made to 2,146 physicians by nonpregnant women aged 14–44 years.
The National Hospital Ambulatory Medical Care Survey (NHAMCS) is an annual survey of approximately 500 U.S. outpatient departments operated by noninstitutional, nonfederal hospitals.14,15 The National Hospital Ambulatory Medical Care Survey uses a 4-stage probability sampling procedure. The first-stage sample contains the same 112 geographic primary sampling units as the National Ambulatory Medical Care Survey. The second stage consists of a probability sample of nonfederal, short-stay or general hospitals with emergency departments, outpatient departments, or both, within the sampled primary sampling units. The third stage involves selecting clinics within outpatient departments. The fourth stage consists of sampling patient visits within clinics during a randomly assigned 4-week reporting period. Data are collected by using an instrument that is very similar to that used for the National Ambulatory Medical Care Survey. However, although the National Ambulatory Medical Care Survey identifies the specialty of the provider seen at a given visit, the National Hospital Ambulatory Medical Care Survey identifies only the specialty of the clinic visited. By combining data from the 1997–2000 surveys, we identified 22,982 visits made by nonpregnant women aged 14–44 years.
Of the 40,304 visits described in the surveys, 787 listed a diagnosis of diabetes, and an additional 131 listed the prescription of oral hypoglycemic medications. We considered these 918 visits to be made by diabetic women. We repeated all analyses without the 29 visits that listed metformin but not a diagnosis of diabetes.
The National Ambulatory Medical Care Survey and National Hospital Ambulatory Medical Care Survey use checklists to collect data about preventive counseling services. We considered contraceptive counseling to have occurred if it was noted on a checklist or if a family planning diagnosis code was listed in any one of the 3 fields available for physician’s diagnoses for this visit. In addition, we considered visits that included oral or injectable hormonal contraception on a woman’s medication list to have included contraceptive counseling. Although it is not possible to distinguish whether women received a contraceptive prescription at the study visit or a prior visit, women known to be using a hormonal contraceptive who did not receive contraceptive counseling at the study visit likely received contraceptive counseling in the prior year.
To estimate the number of visits made by diabetic women for each visit made by a nondiabetic woman of reproductive age, we divided the proportion of visits made by diabetic women of reproductive age by 1.1%, the estimated prevalence of diagnosed diabetes among women of reproductive age.16 To account for the fact that diabetic women made more visits each year, we multiplied the proportion of visits made by diabetic women that included counseling by the number of visits made by diabetic women per visit made by a nondiabetic woman. To examine patient, provider, and visit characteristics predictive of providing contraceptive counseling to all women of reproductive age, we used univariable and multivariable models that used weights provided by the National Center for Health Statistics to account for the multistage probability sampling used to collect National Ambulatory Medical Care Survey and National Hospital Ambulatory Medical Care Survey data.
For all analyses, we used SAS 9.1 (SAS Institute, Cary, NC) with SUDAAN 9.0.1 (RTI International, Research Triangle Park, NC) and weights generated by the National Ambulatory Medical Care and National Hospital Ambulatory Medical Care surveys to produce estimates with 95% confidence intervals that reflect population-based values.15 This work was exempt from review by the University of California, San Francisco Committee on Human Research because data contained no individually identifiable protected health information.
Between 1997 and 2000, women of childbearing age made an estimated 164 million (95% confidence interval [CI] 162–166 million) office visits annually. Of these, 1.8% or about 2.9 million (95% CI 2.8–2.9 million) visits were made by diabetic women. Because it is estimated that 1.1% of women of reproductive age have been diagnosed with diabetes,16 diabetic women made an estimated 1.61 ambulatory care visits for every visit made by a nondiabetic woman. Diabetic women were more likely to visit a generalist and less likely to visit a gynecologist than nondiabetic women (Table 1).
A comparison of visits of diabetic and nondiabetic women showed that a larger proportion of visits made by diabetic women included counseling about diet/nutrition (43.2% versus 10.9% of visits; P < .001) and exercise (18.3% versus 9.4%; P = .002), but that a smaller proportion of visits made by diabetic women included counseling about contraception (4.0% versus 11.5%; P < .001). After adjusting for the fact that diabetic women made 1.61 visits to ambulatory clinicians for every visit made by nondiabetic women, we still found that a smaller proportion of diabetic women (6.4% versus 11.5%) received counseling about contraception per fixed unit of time.
In multivariable analyses adjusting for age, race, ethnicity, insurance, physician or clinic specialty, seeing a primary care provider, having a follow-up visit, and listing family planning as the primary reason for the visit, we found that diabetic women were still less likely than nondiabetic women to receive contraceptive counseling (odds ratio [OR] 0.42, 95% CI 0.21–0.81). In addition, we found that diabetic women under 25 years of age were less likely to receive contraceptive counseling than were older diabetic women (OR 0.17, 95% CI 0.06–0.54). Women were more likely to receive contraceptive counseling if their primary reason for visiting was family planning (OR 34.4, 95% CI 20.8–56.9), but only one visit by a diabetic woman listed this as the primary reason (Table 2). Visits to gynecologists were most likely to list family planning as the primary reason for visit, but only 0.3% of visits to gynecologists were made by women with diabetes. We found no significant change in these results in analyses that excluded the 29 visits that listed a metformin prescription but not a diagnosis of diabetes.
Although diabetes is affecting a growing number of women17 and poses an increased risk of adverse birth outcomes if glucose levels are not tightly controlled, outpatient physicians are providing contraceptive counseling less frequently to diabetic women than to nondiabetic women. Although counseling services have been shown to be underreported by the National Ambulatory Medical Care Survey and National Hospital Ambulatory Medical Care Survey,18 this underreporting should not differ by whether or not a woman has diabetes. For women who have had diabetes less than 20 years and do not have end-organ disease, estrogen-containing contraceptives are safe.19 For diabetic women with vascular complications, contraceptive options are more limited. Nevertheless, it is particularly important for women with medical complications to avoid unintended pregnancies. Our results suggest that diabetic women require more contraceptive counseling and perhaps need to have visits scheduled primarily to address family planning.
Several factors limit the generalizability of our results. First, the data sources do not include information about women’s desire to conceive, sexual orientation, sterilization history, and use of barrier contraceptive methods or intrauterine devices, nor did the sources include information about whether previous health care visits included contraceptive counseling. However, other work has shown that diabetic women use contraception less than nondiabetic women.20 In addition, more recent data were unavailable because contraceptive counseling was removed from the surveys’ preventive services checklists after 2000.
For diabetic women of reproductive age, postponing pregnancy until glucose control is optimized can minimize the risk of adverse birth outcomes.21 To maximize the health of diabetic women and their offspring, physicians must provide counseling about the importance of prepregnancy glucose control and effective contraceptive options. Increasing involvement of gynecologists in the care of diabetic women may have significant public health benefit.
1. Weintrob N, Karp M, Hod M. Short- and long-range complications in offspring of diabetic mothers. J Diabetes Complications 1996;10:294–301.
3. Clausen TD, Mathiesen E, Ekbom P, Hellmuth E, Mandrup-Poulsen T, Damm P. Poor pregnancy outcome in women with type 2 diabetes. Diabetes Care 2005;28:323–8.
4. Verheijen EC, Critchley JA, Whitelaw DC, Tuffnell DJ. Outcomes of pregnancies in women with pre-existing type 1 or type 2 diabetes, in an ethnically mixed population. BJOG 2005;112:1500–3.
5. Boulot P, Chabbert-Buffet N, d’Ercole C, Floriot M, Fontaine P, Fournier A, et al. French multicentric survey of outcome of pregnancy in women with pregestational diabetes. Diabetes Care 2003;26:2990–3.
6. Rosenn B, Miodovnik M, Combs CA, Khoury J, Siddiqi TA. Pre-conception management of insulin-dependent diabetes. improvement of pregnancy outcome. Obstet Gynecol 1991;77:846–9.
7. Willhoite MB, Bennert Jr, HW Palomaki GE, Zaremba MM, Herman WH, Williams JR, et al. The impact of preconception counseling on pregnancy outcomes: the experience of the Maine Diabetes in Pregnancy Program. Diabetes Care 1993;16:450–5.
8. Kitzmiller JL, Buchanan TA, Kjos S, Combs CA, Ratner RE. Preconception care of diabetes, congenital malformations, and spontaneous abortions. Diabetes Care 1996;19:514–41.
10. American Diabetes Association. Standards of medical care in diabetes. Diabetes Care 2005;28 suppl:S4–36.
11. American Diabetes Association. Medical management of pregnancy complicated by diabetes. 3rd ed. Alexandria (VA): American Diabetes Association; 2000.
13. Bryant E, Shimizu I. Sample design, sampling variance, and estimation procedures for the National Ambulatory Medical Care Survey. Vital Health Stat 2 1988;108:1–39.
14. McCaig LF, McLemore T. Plan and operation of the National Hospital Ambulatory Medical Care Survey: programs and collection procedures. Vital Health Stat 1 1994;34:1–78.
16. Harris MI, Flegal KM, Cowie CC, Eberhardt MS, Goldstein DE, Little RR, et al. Prevalence of diabetes, impaired fasting glucose, and impaired glucose tolerance in U.S. adults. The Third National Health and Nutrition Examination Survey, 1988–1994 Diabetes Care 1998;21: 518–24.
17. King H, Aubert RE, Herman WH. Global burden of diabetes, 1995-2025: prevalence, numerical estimates, and projections. Diabetes Care 1998;21:1414–31.
18. Gilchrist VJ, Stange KC, Flocke SA, McCord G, Bourguet CC. A comparison of the National Ambulatory Medical Care Survey (NAMCS) measurement approach with direct observation of outpatient visits. Med Care 2004;42:276–80.
20. Chuang CH, Chase GA, Bensyl DM, Weisman CS. Contraceptive use by diabetic and obese women. Womens Health Issues 2005;15:167–73.
21. Goldman JA, Dicker D, Feldberg D, Yeshaya A, Samuel N, Karp M. Pregnancy outcome in patients with insulin-dependent diabetes mellitus with preconceptional diabetic control: a comparative study. Am J Obstet Gynecol 1986;155:293–7.
Figure. No caption available.