Abstract
Objectives. We sought to determine rates and factors associated with screening for type 2 diabetes mellitus (DM) in women with a history of gestational diabetes mellitus.
Methods. We retrospectively studied women with diagnosed gestational diabetes mellitus who delivered at a university-affiliated hospital (n=570). Data sources included medical and administrative record review. Main outcome measures were the frequency of any type of glucose testing at least 6 weeks after delivery and the frequency of recommended glucose testing. We assessed demographic data, past medical history, and prenatal and postpartum care characteristics.
Results. Rates of glucose testing after delivery were low. Any type of glucose testing was performed at least once after 38% of deliveries, and recommended glucose testing was performed at least once after 23% of deliveries. Among women with at least 1 visit to the health care system after delivery (n=447), 42% received any type of glucose test at least once, and 35% received a recommended glucose test at least once. Factors associated with testing were being married, having a visit with an endocrinologist after delivery, and having more visits after delivery.
Conclusions. These findings suggest that most women with gestational diabetes mellitus are not screened for type 2 DM after delivery. Opportunities for DM prevention and early treatment are being missed.
Gestational diabetes mellitus (GDM), or glucose intolerance first diagnosed during pregnancy, affects more than 135 000 women each year—between 1% and 14% of all pregnancies.1 GDM is a major risk factor for the development of maternal type 2 diabetes mellitus (DM) and impaired glucose tolerance (IGT).2 Therefore, both the American College of Obstetricians and Gynecologists and the American Diabetes Association recommend screening women with a history of GDM for type 2 DM approximately 6 weeks after delivery and periodically thereafter.3,4
Unfortunately, such screening may not occur. One survey of 66 rural patients with a history of GDM showed that 40% of women who delivered in the early 1990s were not tested at all in the 5 years after delivery.5 For women who were tested, providers initiated testing only about 60% of the time, with the women themselves initiating testing in the remaining instances. No significant difference in DM testing was found between medical specialties.
To our knowledge, no other studies in the United States have examined the frequency and predictors of type 2 DM screening in women with a history of GDM. Currently, no national or population-based registries collect information on women with GDM and their follow-up, in part because pregnancy is not a permanent state and because of variable diagnostic criteria for GDM.3,4 Therefore, it is unclear why rates of screening were so low and whether screening rates have improved since the early 1990s.6
The timely identification of persons at high risk for type 2 DM has taken on new urgency with the publication of the results of the Diabetes Prevention Program. This major clinical trial demonstrated that lifestyle interventions and administration of metformin could delay or prevent the development of type 2 DM in people with prediabetes, or IGT.7 Currently, persons with IGT are not easily identified, as universal screening for IGT is not recommended. Because screening for GDM is performed almost universally and identifies women at high risk for IGT,3,4 the population with GDM is also an ideal one to screen for IGT. We therefore sought to determine screening rates for type 2 DM and the predictors of screening in a cohort of women with GDM cared for in the University of Michigan Healthcare System, which was previously documented to have excellent quality of care for persons with DM.8
METHODS
Study Setting and Population
Using the hospital tracking system and codes from the International Classification of Diseases, Ninth Revision,9 we identified 639 women with a history of GDM who delivered at the 872-bed University of Michigan Hospital between 1997 and 2002. The hospital, which employs electronic medical records, had more than 35 000 inpatient admissions, 9500 obstetrician/gynecologist outpatient visits, and 3000 deliveries in 2003. Women were included if their medical records documented any pregnancy with a first prenatal visit between October 17, 1997, and February 18, 2002, including at least 1 pregnancy with a documented GDM diagnosis. We reasoned that women for whom the diagnosis of GDM had been missed would still be at risk for future type 2 DM, but that these women would not necessarily have received recommendations for further testing. We excluded 86 women (13%) after medical record review because they had a history of pregestational DM, either type 1 or type 2 (n = 17); did not have a documented history of GDM (n = 49); or did not make a prenatal visit (n = 20).
Data Collection
Three reviewers reviewed the subjects’ medical records to collect information on the women’s demographic data (age, race, marital status), medical history (gravidity; parity; previous diagnosis of GDM, preeclampsia, or eclampsia; family history of DM), physiological data (height and weight over the course of the pregnancy), prenatal care (date of prenatal visit, provider type at first prenatal visit, week of pregnancy at first presentation, insulin use during the affected pregnancy, number and type of contacts with the health care system, complications of pregnancy), information on delivery (birthweight, type of health care provider, gestational age at delivery), and postpartum care (type of provider, extent of contact with the health care system). We also obtained administrative data including subjects’ laboratory, pharmacy, and health care utilization records from 1997 to 2002; glucose test results; and extent and type of contacts with the health care system.
We abstracted the values of glucose tolerance tests performed during pregnancy to validate the diagnosis of GDM and to confirm whether a formal diagnosis of type 2 DM was made after pregnancy. We randomly selected 5% of the records and double-abstracted them in a blind fashion; that is, the initial reviewers were not aware of which subjects were selected for double abstraction. Agreement was between 61% and 100% for the independent variables listed in Table 1 ▶. For the subset of double-abstracted records, the study population women had correctly received the diagnosis of GDM. That is, they had at least 2 elevated values on the 3-hour 100-g glucose tolerance test recommended during pregnancy: a fasting glucose ≥ 95 mg/dL, a 1-hour value ≥180 mg/dL, a 2-hour value ≥155 mg/dL, or a 3-hour value ≥ 140 mg/dL.3,4
TABLE 1—
Characteristics of Women with GDM, by Screening Status
| Patient Characteristics | Screeneda (n = 204) | Not Screeneda (n = 329) | P |
| Age at delivery, y, mean (SE) | 33 (0.4) | 32 (0.3) | .13 |
| Race | .19 | ||
| Non-Hispanic White, % | 78% | 85% | . . . |
| Non-Hispanic Black, % | 15% | 9% | . . . |
| Other, % | 7% | 6% | . . . |
| Married, % | 87% | 78% | .005 |
| Prior history of GDM, % | 20% | 19% | .89 |
| Gravidity, mean (SE) | 3 (0.1) | 3 (0.1) | .17 |
| Prior history of preeclampsia or eclampsia, % | 6% | 7% | .68 |
| Family history of DM, % | 31% | 29% | .67 |
| Body mass index at first prenatal visit, kg/m2, mean (SE) | 30 (0.6) | 31 (0.5) | .73 |
| Weight gain over pregnancy, lb, mean (SE) | 30 (1.4) | 29 (1.2) | .62 |
| Prenatal insulin use, % | 52% | 43% | .04 |
| Gestational age at delivery, wk, mean (SE) | 38 (0.2) | 38 (0.1) | .92 |
| Most frequent type of prenatal care provider | .33 | ||
| Obstetrician, % | 85% | 80% | . . . |
| Family/general physician, % | 9% | 13% | . . . |
| Nurse-midwife/nurse practitioner, % | 6% | 7% | . . . |
| Any visit with an endocrinologist during pregnancy, % | 88% | 76% | <.001 |
| Total number of prenatal visits, mean (SE) | 17 (0.4) | 15 (0.3) | .004 |
| Live birth, % | 98% | 98% | .91 |
| Multiple births with affected pregnancy, % | 6% | 6% | .93 |
| Birthweight > 9 lb, % | 2% | 4% | .27 |
| Obstetrician present at delivery, % | 94% | 92% | .60 |
| Any visit with an endocrinologist after delivery, % | 21% | 11% | <.001 |
| Any visit with an obstetrician-gynecologist after delivery, % | 72% | 57% | <.001 |
| Total number of visits after delivery, mean (SE) | 4 (0.3) | 3 (0.2) | <.001 |
Note. GDM = gestational diabetes mellitus; DM = diabetes mellitus.
aScreening is defined as any glucose test at least 6 weeks after delivery.
Main Outcome Measures
The primary outcome measure was whether any glucose testing occurred 6 or more weeks after delivery. We used the cutoff of 6 weeks that is currently recommended by the American Diabetes Association and the American College of Obstetricians and Gynecologists based on the observation that maternal glucose may require a period of several weeks to normalize after pregnancy ends.3,4 Of the women who had any glucose testing performed, the majority (79%) had their testing performed 6 or more weeks after the termination of pregnancy, with the remaining glucose tests performed primarily in the days immediately postpartum. Because of the previously reported low rate of testing,6 we considered all glucose testing including whole blood capillary; random venous glucose; glucoses obtained as part of panels of laboratory tests; and hemoglobin A1c, fasting venous glucose, and oral glucose tolerance tests.
We also wanted to determine how many women underwent appropriate postdelivery testing.3,4 Therefore, we also examined how many women underwent the currently recommended fasting venous glucose (either alone or as part of a more comprehensive panel) or an oral 2-hour glucose tolerance test, which is typically performed in nonpregnant women, as opposed to the 3-hour test, which is performed in pregnant women.3,4
Statistical Analyses
The primary comparison was between women who underwent any screening 6 or more weeks after delivery (yes/no) and women who did not. We first conducted bivariate comparisons using χ2 tests for categorical variables and log-rank or t tests for continuous variables. Independent variables statistically associated with screening in bivariate analysis at P< .05 were then forced in multivariable logistic regression models with screening (yes/no) as the dependent variable. Independent variables were evaluated for collinearity before inclusion in the models.
As expected, use of insulin was associated with elevated values on the diagnostic 3-hour test. Use of insulin was particularly associated with a fasting glucose ≥ 95 mg/dL (P< .001), 1-hour value ≥ 180 mg/dL (P< .0001), and 3-hour value ≥ 140 mg/dL (P= .04), and to a lesser degree with a 2-hour value ≥ 155 mg/dL (P= .39), so the variable of insulin use rather than the glucose values was included in the models. We also constructed the models using stepwise logistic regression for comparison; because we found similar results, we did not include those data here. Candidate variables for the multivariate models included the factors listed in Table 1 ▶.
We conducted several additional analyses. We hypothesized that fragmentation of care between providers from before to after delivery could contribute to low rates of testing, but that seeing an obstetrician after delivery could compensate for this; we expected obstetricians would be more aware of recommendations for glucose testing. Therefore, we also evaluated an interaction term between screening, primary health care provider type before delivery, and obstetrician visit after delivery. This interaction term was not significant in any of the models and was not included in the final models. We also determined the proportion of women who had the same physician provide both prenatal and postpartum care. In the instances where patients had several prenatal providers, “contact with the same physician in the prenatal and postpartum periods” was defined as any contact with a specific prenatal provider in the postpartum period. On occasion, the medical record noted postpartum care, but the specific provider was not clearly documented because of the bundling of care. These were noted as “unknown if the same physician provided prenatal and postpartum care.” There were 202 women (38%) who had postpartum contact with 1 of their prenatal providers, 222 (42%) who did not have postpartum contact, and 110 (20%) for whom contact was unknown. This variable was included in the multivariate models. We also constructed regression models in which contact “unknown” was coded as “contact occurred with the same physician in the prenatal and postpartum periods” and regression models in which contact “unknown” was coded as “contact did not occur with the same physician in the prenatal and postpartum periods.”
We also wanted to determine if the factors associated with screening differed between women who had no contact with the health care system after delivery versus women who had at least 1 contact. Therefore, we excluded women who had no contact with the health care system and reconstructed the multivariate regression models as described above. Finally, we reconstructed the multivariate regression models for only the population of women whose postpartum contacts occurred at least 6 weeks after delivery. All analyses were performed using SAS version 8.0 (SAS Institute Inc, Cary, NC).
RESULTS
The rate of screening after delivery using any marker of glucose was 38%. Women who were screened were more often married, were more likely to have used insulin during the pregnancy, were more likely to have had any visit with an endocrinologist during the pregnancy, had had a greater number of prenatal visits, and were more likely to have seen an endocrinologist or obstetrician after delivery (Table 1 ▶). Screened women had a median of 39 days (interquartile range [IQR]=12–45 days) to their first postpartum contact whereas unscreened women had a median of 37 days (IQR=12–44 days) to their first postpartum contact (P=.32). There were no differences in mean 3-hour glucose tolerance test values for women who were screened compared with those who were not screened, for the fasting value (95 mg/dL vs 97 mg/dL; P=.13), 1-hour value (201 mg/dL vs 199 mg/dL; P= .61), 2-hour value (181 mg/dL vs 180 mg/dL; P=.68), or 3-hour value (134 mg/dL vs 131 mg/dL; P=.48). When we compared women who were screened with the appropriate test (i.e., a fasting glucose or oral glucose tolerance test) to women who did not receive the appropriate test, factors predicting screening were similar (results not shown), but the rate of screening dropped to 23%.
In a multivariate model including the variables significant in bivariate analysis, we found that only marital status (odds ratio [OR] = 1.73; 95% confidence interval [CI] = 1.01, 2.98), visit with an endocrinologist after delivery (OR = 1.85; 95% CI = 1.04, 3.28), and total number of visits after delivery (OR = 1.07; 95% CI = 1.01, 1.14) were associated with any glucose screening. Prenatal insulin use, total number of prenatal visits, and visit with an obstetrician after delivery were not associated with glucose screening. When we defined screening as a fasting blood glucose or oral glucose tolerance test, only marital status (OR = 3.79; 95% CI = 1.62, 8.85), visit with an endocrinologist after delivery (OR = 2.10; 95% CI = 1.12, 3.92), and number of prenatal visits (OR = 1.05; 95% CI = 1.01, 1.10) were associated with glucose screening. Any visit with an obstetrician after delivery was not associated with appropriate glucose screening.
We next examined only the women who had at least 1 contact with the health care system after delivery (n = 447). Among those women, only 42% were screened with any glucose test (Table 2 ▶). Women who were screened did not have different mean 3-hour glucose tolerance test values than women who were not screened, for the fasting value (94 mg/dL vs 95 mg/dL; P = .56), 1-hour value (200 mg/dL vs 200 mg/dL; P = .81), 2-hour value (181 mg/dL vs 181 mg/dL; P = .97), or 3-hour value (134 mg/dL vs 131 mg/dL; P = .40). Marital status, visiting an endocrinologist after delivery, and having a greater number of visits after delivery were associated with greater rates of screening. Screened women had a median of 39 days (IQR = 12–45 days) to their first postpartum contact whereas unscreened women had a median of 36 days (IQR = 11–43 days) to their first postpartum contact (P = .30). When we compared women who were screened with the appropriate test (i.e., a fasting glucose or oral glucose tolerance test) to women who did not receive the appropriate test, factors predicting screening were similar (results not shown), but the rate of appropriate screening dropped to 35%.
TABLE 2—
Characteristics of Women with GDM Who Had at Least 1 Contact with the Health Care System After Delivery, by Screening Status
| Patient Characteristics | Screeneda (n = 192) | Not Screeneda (n = 255) | P |
| Age at delivery, y, mean (SE) | 33 (0.4) | 32 (0.3) | .16 |
| Race | .45 | ||
| Non-Hispanic White, % | 79% | 84% | . . . |
| Non-Hispanic Black, % | 14% | 10% | . . . |
| Other, % | 7% | 6% | . . . |
| Married, % | 88% | 80% | .02 |
| Prior history of GDM, % | 18% | 18% | .93 |
| Gravidity, mean (SE) | 3 (0.2) | 3 (0.1) | .3 |
| Prior history of preeclampsia or eclampsia, % | 6% | 8% | .38 |
| Family history of DM, % | 31% | 27% | .44 |
| Body mass index and first prenatal visit, kg/m2, mean (SE) | 30 (0.6) | 30 (0.5) | .92 |
| Weight gain over pregnancy, lbs, mean (SE) | 30 (1.5) | 29 (1.3) | .68 |
| Prenatal insulin use, % | 53% | 46% | .12 |
| Gestational age at delivery, wk, mean (SE) | 38 (0.2) | 38 (0.2) | .69 |
| Most frequent type of prenatal care provider | .28 | ||
| Obstetrician, % | 85% | 80% | . . . |
| Family/general physician, % | 9% | 13% | . . . |
| Nurse-midwife/nurse practitioner, % | 6% | 7% | . . . |
| Any visit with an endocrinologist during pregnancy, % | 89% | 86% | .31 |
| Total number of prenatal visits, mean (SE) | 17 (0.4) | 16 (0.4) | .06 |
| Live birth | 98% | 98% | .23 |
| Multiple births with affected pregnancy, % | 6% | 5% | .60 |
| Birthweight > 9 lb, % | 3% | 4% | .34 |
| Obstetrician present at delivery, % | 94% | 92% | .45 |
| Any visit with an endocrinologist after delivery, % | 22% | 14% | .02 |
| Any visit with an obstetrician-gynecologist after delivery, % | 77% | 73% | .43 |
| Total number of visits after delivery, mean (SE) | 4 (0.3) | 3 (0.2) | .01 |
Note. GDM = gestational diabetes mellitus; DM = diabetes mellitus.
a Screening is defined as any glucose test at least 6 weeks after delivery.
In multivariate models, factors predicting any screening among women with at least 1 postpartum contact with the health care system included only marital status (OR = 1.91; 95% CI = 1.12, 3.29), visit with an endocrinologist after delivery (OR = 1.75; 95% CI = 1.06, 2.88), and total number of visits after delivery (OR = 1.07; 95% CI = 1.01, 1.14). When we defined screening as women receiving the appropriate glucose test, factors associated with screening included marital status (OR = 5.92; 95% CI = 2.04, 10.15), visit with an endocrinologist after delivery (OR = 1.97; 95% CI = 1.07, 3.61), and total number of prenatal visits (OR = 1.06; 95% CI = 1.01, 1.11).
When we further restricted the study population to women who had a postpartum visit at least 6 weeks after delivery (n = 395), we found that the factors predicting screening in multivariate models were similar. Namely, marital status (OR = 1.85; 95% CI = 1.01, 3.35), visit with an endocrinologist after delivery (OR = 1.81; 95% CI = 1.04, 3.15), and total number of visits after delivery (OR = 1.05; 95% CI = 1.01, 1.11) were associated with any glucose screening.
Postpartum contact with a specific prenatal provider was not associated with screening in multivariate models (OR = 1.49; 95% CI = 0.91, 2.42). However, when the women for whom contact was “unknown” were reclassified as “contact occurred with the same physician in the prenatal and postpartum periods,” contact was significantly associated with screening in multivariate models (OR = 1.68; 95% CI = 1.09, 2.59).
Only 2% of women were identified in the medical records as having a new diagnosis of type 2 DM within 1 year after delivery. Of the women who were not diagnosed with type 2 DM, only 3% were diagnosed with impaired fasting glucose (defined as ≥ 110 mg/dL) or IGT (defined as a fasting value ≥ 110 mg/dL or a 2-hour value > 140 and < 199 mg/dL) after delivery.3,4
DISCUSSION
In a university-affiliated health care system previously demonstrated to have excellent quality of DM care, we found that quality of care did not extend to women at high risk for type 2 DM. Screening rates for type 2 DM after delivery were low, whether screening was defined as any type of glucose test or the recommended fasting glucose or glucose tolerance test.
There are several possible reasons for low testing rates. Women with GDM may have limited knowledge of the significance of their diagnosis, in that they do not realize they are at increased risk of type 2 DM after delivery.10 Another reason may be lack of health care provider knowledge of the association between diseases of pregnancy and maternal health after delivery. A 1998 survey of American College of Obstetricians and Gynecologists fellows showed that only 62% believed that women with GDM were at risk for developing type 2 DM and only 71% would recommend testing for this disease.11 By 2004, the latter percentage had improved only to 75%.12
Fragmentation of medical care from the prenatal to postpartum periods may contribute to the low testing rates, in that women with GDM are typically diagnosed by an obstetrician, sometimes referred to an endocrinologist for care during the pregnancy, and then referred back to the primary care provider after delivery. However, we did not find that continuity between specialties from before to after delivery seemed to impact screening rates. Finally, testing may be low because of limited interaction between these women and the health care system. Many women with a history of GDM are asymptomatic and otherwise healthy, decreasing the opportunities to screen at other visits. These women may not present to the health care system for routine checkups. We found that even among women who did follow up with the health care system for other reasons, glucose screening by any definition remained low.
The only factors consistently associated with better screening rates were marital status and any visit with an endocrinologist after delivery. It may be that a visit with an endocrinologist is a proxy for greater disease severity, and testing may occur primarily for women who are perceived to be at higher risk, although the current recommendations are to screen all women with the diagnosis of GDM. The fact that prenatal insulin treatment dropped out of the model when endocrinologist visit was entered supports this hypothesis. Alternatively, endocrinologists may be more focused than non–diabetes specialists on the ramifications of GDM for maternal health after delivery.
The survey of obstetricians mentioned above found a lack of universal support for glucose testing after delivery.12 This may reflect ambivalence regarding screening for GDM itself. More specifically, diagnosis and treatment of GDM may lead to increased cesarean delivery rates without other benefits on pregnancy outcomes.13 In addition, the diagnosis of GDM may lead to negative self-perceptions of health status.14 Such issues may blur providers’ perceptions about the value of testing for type 2 DM after delivery in women with GDM. Primary care physicians may be even less aware of the implications of GDM for future type 2 DM in the mother, although to our knowledge no studies have examined this issue.
There may have been several reasons why marital status was associated with improved screening. Married women may, through their spouses, have had better insurance coverage for postpartum care—including health care visits and glucose screening tests—than was available to unmarried women. It is also possible that marital status was a proxy for other factors, such as instrumental social support, that were associated with preventive behaviors.
We also noted that postpartum contact with a specific prenatal provider was associated with screening, although the precision of the definition was limited. This suggests that lack of continuity, and therefore postpartum provider lack of awareness of the GDM diagnosis, may have contributed to the lack of screening. Although the university hospital employs electronic medical records, and our study population was identified by a diagnosis of GDM upon hospital discharge, fragmentation of care could still have occurred for several reasons. Medical record documentation of a GDM diagnosis may not be as salient to a physician as actually providing care for GDM in the prenatal period. In addition, resident clinics may contribute to greater discontinuity because frequent service changes lead to lower availability of a specific resident. Although attending physicians are required to monitor residents, prenatal care given by both an attending physician and a resident could lead to a less thorough knowledge of a woman’s history by an attending physician.
Our study has several limitations. We examined screening rates in a single university health care system. However, we believe screening rates are most likely to be even lower in other settings, particularly those with lower quality of care for patients with DM. Obstetricians provided almost all of the prenatal care for women with GDM, and our power to detect differences between obstetricians and other types of providers was limited. It is possible that women received glucose testing from physicians outside the health care system, thus artificially reducing the screening rates, although we found that among women who continued to seek care within the university system the screening rates were still not optimal. Our assessment of continuity of care from the prenatal to the postpartum periods was limited in precision, as noted above. In addition, we assessed only any contact, rather than extent of contact. This relationship would have been difficult to quantify further, given the nature of prenatal care delivery in the university hospital setting, which requires that both residents and attending physicians provide care. As our study was confined to medical chart review, we could not determine whether defects in testing were attributable to patient, provider, or health care system barriers, or to a combination of the 3.
We conclude that the follow-up of women with GDM, and screening for type 2 DM in particular, could be vastly improved. Failure to screen for type 2 DM may result in missed opportunities for disease prevention. Further research should examine the patient, provider, and health care system barriers to screening, including the impact of patient and provider perception of risk and cost. Future study of the strength of the relationship between continuity of care and adequate screening could determine ways to improve communication between the prenatal and postpartum periods.
Acknowledgments
C. Kim was supported by an American Diabetes Association Junior Faculty Award and a Collaborative Health Services Research for Practice Improvements grant from the Executive Vice President for Medical Affairs, University of Michigan. B. P. Tabaei, L. N. McEwen, R. Burke, and W.H. Herman were supported by the Centers for Disease Control and Prevention (grant U58/CCU523525-03).
Human Participant Protection This protocol was reviewed and approved by the institutional review board of the University of Michigan.
Peer Reviewed
Contributors C. Kim conceptualized and supervised the study, and led the writing. B. P. Tabaei and L. N. McEwen completed the analysis. R. Burke, R. W. Lash, S. L. Johnson, K. L. Schwartz, S. J. Bernstein, and W.H. Herman assisted with the study.
References
- 1.Ferrara A, Kahn HS, Quesenberry CP, Riley C, Hedderson MM. An increase in the incidence of gestational diabetes mellitus: Northern California, 1991–2000. Obstet Gynecol. 2004;103:526–533. [DOI] [PubMed] [Google Scholar]
- 2.Kim C, Newton KM, Knopp RH. Gestational diabetes and incidence of type 2 diabetes mellitus: a systematic review. Diabetes Care. 2002;25:1862–1868. [DOI] [PubMed] [Google Scholar]
- 3.American College of Obstetricians and Gynecologists Committee on Practice Bulletins—Obstetrics. ACOG Practice Bulletin. Clinical management guidelines for obstetrician-gynecologists. Number 30, September 2001 (replaces Technical Bulletin Number 200, December 1994). Gestational diabetes. Obstet Gynecol. 2001;98:525–538. [PubMed] [Google Scholar]
- 4.American Diabetes Association. Gestational diabetes mellitus. Diabetes Care. 2000;23(suppl 1): S77–S79. [PubMed] [Google Scholar]
- 5.Kaufmann RC, Smith T, Bochantin T, Khardori R, Evans MS, Steahly L. Failure to obtain follow-up testing for gestational diabetic patients in a rural population. Obstet Gynecol. 1999;93:734–737. [DOI] [PubMed] [Google Scholar]
- 6.Kaufmann RC, Schleyhahn FT, Huffman DG, Amankwah KS. Gestational diabetes diagnostic criteria: long-term maternal follow-up. Am J Obstet Gynecol. 1995;172:621–625. [DOI] [PubMed] [Google Scholar]
- 7.Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002;346: 393–403. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kim C, Williamson DF, Mangione CM, Safford MM, Selby JV, Marrero DG, et al. Managed care organization and the quality of diabetes care: the Translating Research Into Action for Diabetes (TRIAD) study. Diabetes Care. 2004;27:1529–1534. [DOI] [PubMed] [Google Scholar]
- 9.International Classification of Diseases, Ninth Revision. Geneva, Switzerland: World Health Organization; 1980.
- 10.Kieffer EC, Willis SK, Arellano N, Guzman R. Perspectives of pregnant and postpartum Latino women on diabetes, physical activity, and health. Health Educ Behav. 2001;29:542–556. [DOI] [PubMed] [Google Scholar]
- 11.Gabbe S, Hill L, Schmidt L, Schulkin J. Management of diabetes by obstetrician-gynecologists. Obstet Gynecol. 1998;91:643–647. [DOI] [PubMed] [Google Scholar]
- 12.Gabbe SG, Gregory RP, Power ML, Williams SB, Schulkin J. Management of diabetes mellitus by obstetrician-gynecologists. Obstet Gynecol. 2004;103: 1229–1234. [DOI] [PubMed] [Google Scholar]
- 13.Naylor CD, Sermer M, Chen E, Sykora K. Cesarean delivery in relation to birthweight and gestational glucose tolerance: pathophysiology or practice style? Toronto Trihospital Gestational Diabetes Investigators. JAMA. 1996;275:1165–1170. [PubMed] [Google Scholar]
- 14.Feig DS, Chen E, Naylor CD. Self-perceived health status of women three to five years after the diagnosis of gestational diabetes: a survey of cases and matched controls. Am J Obstet Gynecol. 1998;178:386–393. [DOI] [PubMed] [Google Scholar]