To reduce misclassification of births, discharge records with ungroupable DRGs, defined as a DRG=999 or missing (n=1,463 in 2009), were excluded; and to reduce double-counting of hospital stays (n=4,512 in 2009), discharge records with an indication of transfers to or from another hospital were also excluded. Less than 0.2 percent of encounters were excluded from each year. Analysis National estimates selleck chemicals llc were calculated using HCUP-supplied weights, based on the NIS sampling frame. The unit of analysis is the hospital discharge (i.e., the hospital stay), not an infant. Trends from 2002–2009 were displayed graphically and significance of trend was ascertained through a chi-square test for

trend (Snedecor & Cochran, 1989). To determine likely payer source in 2009, three logistic regression models were conducted for each of the three most prevalent diagnoses (preterm birth/low birth weight, respiratory distress, and jaundice). Presence of the diagnosis was the outcome variable, while the main predictor variable in each model was expected payer source (Medicaid, Private, or Uninsured).

Odds ratios were adjusted for infant’s gender, race/ethnicity, community-level median household income, location of residence, and hospital characteristics. Results Overall Trends for Complicated Births Exhibit 1 shows the number of complicated newborn stays split out by the total number of complicated births and the total number of neonate admissions within 30 days after the birth, and the percent of complicated

births across all expected payers from 2002 to 2009, (P = .08 for trend). By 2009, there were 4,154,637 total births, and complicated newborn stays reached 859,853 (21% of all births). Of the complicated newborn stays, 143,975 (17%) were for neonate admissions within 30 days of birth. Exhibit 1. Total Hospital Discharges for Complicated Newborn Stays from 2002–2009 Exhibit 2 displays trends by expected payer, showing that the proportion of complicated newborn stays billed to Medicaid increased between 2006 and 2009 (P<.001 for trend), while the proportion of these stays Drug_discovery billed to private insurance decreased (P<.001 for trend). By 2009, the trend lines crossed and Medicaid was billed for a higher proportion of complicated newborn stays than private payers. There were far fewer uninsured complicated newborn stays than those billed to Medicaid or private payers and the proportion remained unchanged between 2002 and 2009. Exhibit 2. Percent of Complicated Newborn Stays by Expected Payer Source from 2002–2009 Exhibit 3 shows the proportion of normal (uncomplicated) Medicaid births, the proportion of complicated Medicaid births, and the proportion of women in the U.S. 15–44 years old who are covered by Medicaid. The proportion of normal and complicated births followed a similar projection over time, and there was an increase over time in the proportion of women 15–44 years old who were covered by Medicaid.