Impact of Expansion of Specialty Consultation Practice on Use of Specialty Care for Medicaid High Risk Pregnancies Janet M. Bronstein University of Alabama at Birmingham jbronstein@uab.edu A N G E L S Antenatal & Neonatal Guidelines, Education & Learning System A N G E L S Antenatal & Neonatal Guidelines, Education & Learning System • Co-authors at the University of Arkansas for Medical Sciences – – – – Songthip T. Ounpraseuth, Ph.D. Jeffrey Jonkman, Ph.D. David Fletcher, M.B.A Curtis Lowery, M.D. • Co-authors at the State of Arkansas – Judith McGhee, M.D. (Medicaid Agency) – Richard Nugent, M.D., M.P.H. (Department of Public Health) Medicaid covers a large portion of U.S. pregnancies – many are high risk ANGELS – Arkansas Medicaid supported expansion of UAMS MFM consulting service Case Management Education & Support for OB Providers Evidence Based Guidelines Clinical Research 24/7 Consultation Call Center Expansion of Telemedicine Network & Clinics Evaluation of ANGELS (Phase one, 20012005, intervention started April 2003) • Did ANGELS increase the proportion of – Maternal Fetal Medicine (MFM) consults for targeted diagnoses? – MFM consults for other reasons? – Deliveries of infants younger than 35 weeks at neonatology-staffed NICU? Consults and NICU deliveries are special cases of physician referral behavior • Making the Referral • Norms for referral (boundaries of expertise), professional networks, pressures to retain patients – ANGELS and health care market factors • Completing the Referral • Enabling characteristics, awareness of pregnancy risk, relationship with primary provider – Demographics and care factors Consults and NICU deliveries are special cases of physician referral behavior • ANGELS and health care market factors – Guidelines, calls to call center, site for MD tele-rounds, remote patient consult site – Presence of OB or GP or no maternity, beds and MDs per capita, sites claiming to do high risk OB and have NICUs • Demographics and care factors – Education, age, race, parity, ethnicity, partner status, county poverty level, % Black, rural, distance – Adequacy of prenatal care, initial care in health dept, gestation at delivery, any MFM consult (for delivery at NICU) Methodology Data • Medicaid claims for pregnancy 2001-2005, clustered into episodes, linked to birth certificates. Total 91,902 live births. • County ARF and AHA data for residence and contiguous counties linked to the episodes. • MFM analysis excludes Little Rock. NICU analysis includes only births < 35 weeks gestation. Statistical Approach • Time series logistic regression with standard errors corrected for county clustering. • Time trend, distance to specialist and distance to closest NICU hospital controlled Trends in Outcome Measures 2001 2002 2003 Angels began 2004 Angels 2005 Angels MFM with 28.7% targeted DX (n=30,364) 30.4% 27.5% 25.6% 24.3% 17.8% MFM no targeted DX (n=47,410) 18.4% 16.0% 12.2% 12.3% 40.3% 40.5% 45.3% 45.0% 43.6% NICU delivery (n=4,531) Impact of ANGELS measures on care delivery (O.R., 95% C.I., p value) MFM for targeted DX MFM for other DX Delivery at NICU Guidelines in active draft 0.99 (0.99,1.00) 1.00 (0.99,1.01) 1.02 (1.01,1.04)** Cumulative guidelines completed 0.98 (0.98,0.99)*** 0.98 (0.98,0.99)*** 1.00 (0.99, 1.01) MD calls from county 0.91 (0.74,1.13) 0.71 (0.54,0.93)* 1.22 (0.94,1.57) PT calls from county 1.68 (1.33,2.11)*** 2.59 1.60 (1.82, 3.69)*** (1.13,2.27)** Teleconference site 1.03 (0.71,1.48) 1.16 (0.80, 0.70) 0.29 (0.18,0.45)*** Clinical remote site 1.26 (0.90,1.77) 1.36 (0.80,2.32) 1.88 (1.22,2.89)** * P < .05, ** p<.01, *** p< .001 Impact of Health Care Market measures on care delivery (sign, p value) MFM for targeted DX MFM for other DX Delivery at NICU Other MD vs OB - (p=.06) -*** +*** No maternity vs OB +* +* +*** Reports NICU (no (ns) (ns) -*** in cty +*** contig cty Reports high risk OB (no MFM) (ns) - (p=.05) -** MDs per capita -* - (p=.07) +** in cty, -*** contig cty Beds per capita +*** +*** +*** neonatologist) * P < .05, ** p < .01, *** p < .001 Impact of individual demographic measures on care delivery (sign, p value) MFM for targeted DX MFM for other DX Delivery at NICU Black/other vs white (ns) (ns) (ns) Hispanic vs not -*** -* +* Education (ns) + low (p=.07) - High ** -low* + high*** Age (ns) - >34 *** -Teen* Partner status vs married + without* + without*** (ns) % cnty Black +* +* +* % cnty < FPL -*** -** -*** (ns) -** +*** Rural county * P < .05, ** p < .01, *** p < .001 Impact of individual care measures on care delivery (sign, p value) MFM for targeted DX Multiple birth MFM for other DX Delivery at NICU +** + p=.05 (ns) + 25 -31wks** + 32-33 wks p=.06 (ns) +*** (most common for 25-27 wks) (ns) + high parity* + first birth*** Adequacy of Prenatal care + intensive*** + intensive*** -Intermed.*** + inadeq.*** Initial care in Health Dept +*** +*** -*** Health Dept X time -*** -*** +*** MFM visit ------ ----- +*** Gestational age at delivery vs oldest Parity * P < .05, ** p < .01, *** p <.001 Conclusions • Medicaid-supported expansion of academic MFM consulting service is altering delivery of maternity care in Arkansas – Increasing local expertise, decreasing specialty visits – Increasing use of neonatology-staffed NICU for preterm deliveries • Local expertise competes with more distant specialists, may lead to less appropriate care. Thank You A N G E L S Antenatal & Neonatal Guidelines, Education & Learning System