Teaching Public and Population Health in Medical Education: An Evaluation Framework.

Curriculum models and training activities in medical education have been markedly enhanced to prepare physicians to address the health needs of diverse populations and to advance health equity. While different teaching and experiential learning activities in the public health and population health sciences have been implemented, there is no existing framework to measure the effectiveness of public and population health (PPH) education in medical education programs. In 2015, the Association of American Medical Colleges established the Expert Panel on Public and Population Health in Medical Education, which convened 20 U.S. medical faculty members whose goal was to develop an evaluation framework adapted from the New World Kirkpatrick Model. Institutional leaders can use this framework to assess the effectiveness of PPH curricula for learners, faculty, and community partners. It may also assist institutions with identifying opportunities to improve the integration of PPH content into medical education programs. In this article, the authors present outcomes metrics and practical curricular or institutional illustrations at each Kirkpatrick training evaluation level to assist institutions with the measurement of (1) reaction to the PPH education content, (2) learning accomplished, (3) application of knowledge and skills to practice, and (4) outcomes achieved as a result of PPH education and practice. A fifth level was added to measure the benefit of PPH curricula on the health system and population health. The framework may assist with developing a locally relevant evaluation to further integrate and support PPH education at U.S. medical schools and teaching hospitals.

Peripartum Maternal Hepatitis B Care in a US Nationwide Data Set.

BACKGROUND: Hepatitis B virus (HBV) screening during pregnancy is standard of care to prevent vertical transmission to infants, yet the mothers themselves may not receive appropriate follow-up. GOALS: Using a national database, we sought to determine rates of maternal peripartum follow-up with a HBV specialist and identify factors associated with a lack of follow-up. MATERIALS AND METHODS: We identified women who delivered in 2000 to 2012 and were diagnosed with HBV according to International Classification of Diseases-9 codes using a national database (Optum) derived from commercial insurance claims with ∼46 million members ages 0 to 64 in all 50 states. Our primary outcome was follow-up during or after pregnancy with a HBV specialist (gastroenterology/infectious diseases). RESULTS: The prevalence of HBV was 0.27% (2558/959,747 pregnancies), and median follow-up was 45 months. Only 21% of women had peripartum HBV specialist follow-up. On multivariable regression, predictors of peripartum follow-up at 1-year included younger age [odds ratio (OR), 0.97/y; 95% confidence interval (CI), 0.94, 0.99], Asian race/ethnicity (OR, 1.56 vs. white; 95% CI, 1.13, 2.17), and residing in the Northeast (OR, 1.70; 95% CI, 1.09, 2.66) and Midwest (OR, 1.73; 95% CI, 1.07, 2.81) versus West. Predictors of testing for HBV DNA and alanine aminotransferase at 1 year included Asian race (OR, 1.72; 95% CI, 1.23, 2.41), a primary care physician visit within 2 years of delivery (OR, 1.63; 95% CI, 1.19, 2.22), and peripartum HBV specialist follow-up within 1 year (OR, 15.68; 95% CI, 11.38, 21.60). CONCLUSIONS: Maternal HBV specialist follow-up rates were extremely low in this large, diverse cohort representing all United States regions. Referral to a HBV specialist was the strongest predictor of appropriate postpartum HBV laboratory testing. Follow-up rates may be even lower in uninsured populations.

Disparities in Care for Publicly Insured Women With Pregestational Diabetes.

OBJECTIVE: To investigate the association among public health insurance, preconception care, and pregnancy outcomes in pregnant women with pregestational diabetes. METHODS: This is a retrospective cohort of pregnant women with pregestational type 1 or type 2 diabetes from 2006 to 2011 in Massachusetts-a state with universal insurance coverage since 2006. Women delivering after 24 weeks of gestation and receiving endocrinology and obstetric care in a multidisciplinary clinic were included. Rates of preconception consultation, our primary outcome of interest, were then compared between publicly and privately insured women. We used univariate analysis followed by logistic regression to compare receipt of preconception consultation and other secondary diabetes care measures and pregnancy outcomes according to insurance status. RESULTS: Fifty-four percent (n=106) of 197 women had public insurance. Publicly insured women were younger (median age 30.4 compared with 35.3 years, P<.01) with lower rates of college education (12.3% compared with 45.1%, P<.01). Women with public insurance were less likely to receive a preconception consult (5.7% compared with 31.9%, P<.01), had lower rates of hemoglobin A1C less than 6% at the onset of pregnancy (37.2% compared with 58.4%, P=.01), and experienced higher rates of pregnancies affected by congenital anomalies (10.4% compared with 2.2%, P=.02) compared with those with private insurance. In adjusted analyses controlling for educational attainment, maternal age, and body mass index, women with public insurance were less likely to receive a preconception consult (adjusted odds ratio [OR] 0.21, 95% CI 0.08-0.58), although the odds of achieving the target hemoglobin A1C (adjusted OR 0.45, 95% CI 0.20-1.02) and congenital anomaly (adjusted OR 2.23, 95% CI 0.37-13.41) were similar after adjustment. CONCLUSION: Despite continuous access to health insurance, publicly insured women were less likely than privately insured women to receive a preconception consult-an evidence-based intervention known to improve pregnancy outcomes. Improving use of preconception care among publicly insured women with diabetes is critical to reducing disparities in outcomes.

Contraception and postpartum follow-up in patients with gestational diabetes.

OBJECTIVES: We sought to review the effect of immediate provision of postpartum contraception on postpartum follow-up and screening for type 2 diabetes in women with gestational diabetes mellitus (GDM). STUDY DESIGN: We performed multivariate logistic regression to identify the association between immediate provision of postpartum contraception and attendance at the 6-week postpartum visit and performance of a 6- to 12-week oral glucose tolerance test, controlling for age, type of insurance, parity and race. RESULTS: Women who received contraception prior to hospital discharge were less likely to attend their postpartum visit [adjusted odds ratio (aOR) 0.59, 95% confidence interval (CI) 0.39-0.88], but just as likely to participate in postpartum diabetes screening (adjusted odds ratio (aOR) 1.27, 95% confidence interval (CI) 0.78-2.06). Attendance at the postpartum visit was associated with private insurance (OR 1.93, 95% CI 1.31-2.99). CONCLUSION: Receiving contraception while admitted postpartum did not affect follow-up for diabetes screening for women with GDM.

Sex-Specific Associations of Maternal Gestational Glycemia with Hormones in Umbilical Cord Blood at Delivery.

Background Few studies have examined sex-specific associations of maternal gestational glycemia with cord blood hormones, which might predict later health. Methods In 976 women without pre-existing diabetes in the Project Viva cohort, we used linear regression to examine associations of maternal gestational glycemia with cord hormone concentrations, adjusted for maternal characteristics and stratified by infant sex. Results A total of 6.1% of women had gestational diabetes mellitus (GDM), 8.8% isolated hyperglycemia, 3.2% gestational impaired glucose tolerance, and 81.9% were normoglycemic. In boys, compared with infants of normoglycemic mothers, infants of GDM mothers had higher cord levels of IGF-2 (ß 35.55 ng/mL; 95% CI: 2.60, 68.50), IGFBP-3 (111.2 ng/mL; 5.53, 216.8), insulin (4.66 uU/mL; 2.38, 6.95), C-peptide (0.46 ng/mL; 0.25, 0.67), and leptin (3.51 ng/mL; 1.37, 5.64), but lower IGF-1 (-6.71 ng/mL; -12.7, - 0.76, adjusted for IGFBP-3). In girls, GDM offspring had higher cord blood levels of IGF-1 adjusted for IGFBP-3 (12.45 ng/mL; 4.85, 20.04). Boys, but not girls, of mothers with abnormal glucose tolerance but not GDM also had higher levels of some hormones. Conclusion GDM was associated with growth factors and adipokines in cord blood from boys, but only IGF-1 in girls. These findings suggest sex differences in responses to fetal overnutrition.