The California Hospital Outcomes Project is an initiative mandated by the State of California and conducted by the Office of Statewide Health Planning and Development (OSHPD), to develop public reports comparing hospital outcomes for selected medical conditions and surgical procedures for patients treated in hospitals throughout California. Over the last decade, CHOP has reported risk-adjusted hospital mortality rates for heart attack and community-acquired pneumonia. In 2005, OSHPD is releasing its first report on obstetric care.
Delivery was selected as an important topic for public reporting because it is the most frequent single reason for hospitalization in California, and because complications of delivery are associated with substantial health care costs and impairment of function. In 2003, for example, there were 170,465 repairs of obstetric lacerations and 147,084 cesarean deliveries performed in California hospitals – more than any other surgical procedure. Although most women who require these procedures have excellent outcomes, a small minority experience complications that cause pain, weakness, impaired bonding with their new child, bowel or bladder problems, sexual dysfunction, rehospitalization, and even death.
This technical report, prepared for OSHPD, summarizes the validation of multiple potential measures of inpatient obstetric quality of care. These measures include two that have been endorsed by OSHPD for public reporting: risk-adjusted postpartum maternal readmission rates and risk-adjusted perineal laceration rates. Other measures were also evaluated in this validation study, but are not recommended for public reporting, including risk-adjusted rates of endometritis, wound infection, hemorrhage, and urinary tract infection. This validation study was designed by the UC Davis research team in collaboration with OSHPD staff and the AB 524 Technical Advisory Committee. It was designed to address a variety of concerns, specified in detail later in this report, about the validity of using hospital-reported ICD-9-CM codes in the California Patient Discharge Data Set to report publicly on hospital performance.
The original methodology for estimating and analyzing risk-adjusted postpartum maternal readmission rates was developed in 1996, using data on deliveries performed in 1992-1993. This developmental work is fully described in a report that was published by the OSHPD in December 1996 (Section 2, citation 19). We were subsequently asked by the OSHPD to validate the data and methodology used in this 1996 report. To simplify the task, we selected a subsample of the same records for this validation study. As described in detail in later sections of this report, we collected records from hospitals in 1998, recoded and abstracted them in 1999, and performed analyses in 2000-2001. Although some results from these analyses have already appeared in print elsewhere, this report compiles all relevant findings in a single document. We believe that the findings are still informative, despite their age, because there is no evidence of statewide improvement in the coding of obstetric records over the past decade. In addition, the mean postpartum length of stay and readmission rate have remained relatively stable over time (after some decrease in length of stay during the 1990s), suggesting that the clinical factors driving readmissions have also been relatively stable. This study remains the most comprehensive published analysis of the accuracy of ICD-9-CM coded inpatient obstetric data. However, if the OSHPD continues to use the same datasets in the same manner, it would be prudent to repeat this validation study in the future.