Each NIS contains approximately 8 million discharges, weighted to approximate 38 to 39 million discharges. Each discharge contains patient demographics, up to 15 International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) diagnoses and procedures, Diagnosis-Related Groups (DRG) codes, total charges, length of stay, and expected payer source from billing Bak apoptosis or hospital discharge abstract data. Variables and Data Definitions The primary variables of interest for monitoring trends were complicated newborn stays, defined as hospital discharge records with an ICD-9-CM diagnosis indicating a complicated birth at delivery or a neonate

admission to the hospital within 30 days after the birth (this does not include the birth event). Other variables of interest were expected primary payer, classified as private insurance, Medicaid, other types of insurance (including Medicare) or uninsured (self-pay, no charge); costs of complicated newborn stays; length of stay in the hospital; and most prevalent

principal diagnosis. Total hospital charges were converted to costs using a year-specific HCUP Cost-to-Charge Ratio (HCUP, 2006–2009). Costs were adjusted to 2009 dollars using the overall consumer price index (CPI Inflation Calculator, 2013). Hospital length of stay is calculated by subtracting the hospital discharge date from the hospital admission date. Additional covariates of interest in assessing expected payer for complicated

newborn stays were infant’s gender race/ethnicity (Black, Hispanic, White, Other), community-level median household income based on the ZIP Code of the patient, location of residence, and hospital characteristics (ownership, region, teaching status, bed size, urban or rural location). To further understand the types of complications that newborns experienced at birth or within 30 days of birth in 2009, all-listed ICD-9-CM diagnoses were classified into clinically meaningful categories using a modified version of the Clinical Classification Software (CCS) (HCUP CCS, 2009, December). The principal diagnosis was used, unless this diagnosis was not plausible as a reason for a complicated hospital stay, Batimastat in which case the first clinically relevant diagnosis code was used. For example, if the principal diagnosis was “other perinatal conditions” the next diagnosis indicating a specific condition was used. See Appendix A for list of clinically meaningful categories through a modified CCS. The majority of stays were classifiable within the first four diagnoses. Study Sample The study sample for this analysis was created from HCUP NIS and comprised all complicated newborn stays drawing from approximately 4.2 million newborn discharge records per year, in years 2002–2009.