Objective
We sought to reduce scheduled births between 36 0/7 -38 6/7 weeks that lack appropriate medical indication.
Study Design
Twenty Ohio maternity hospitals collected baseline data for 60 days and then selected locally appropriate Institute for Healthcare Improvement Breakthrough Series interventions to reduce the incidence of scheduled births. Deidentified birth data were analyzed centrally. Rates of scheduled births without a documented indication, birth certificate data, and implementation issues were shared regularly among sites.
Results
The rate of scheduled births between 36 0/7 -38 6/7 weeks without a documented medical indication declined from 25% to <5% ( P < .05) in participating hospitals. Birth certificate data showed inductions without an indication declined from a mean of 13% to 8% ( P < .0027). Dating criteria were documented in 99% of charts.
Conclusion
A statewide quality collaborative was associated with fewer scheduled births lacking a documented medical indication.
Scheduled induction of labor or cesarean birth is recognized by the American College of Obstetricians and Gynecologists (ACOG) as appropriate only when justified by medical or obstetric complications. Some indications for scheduled birth require prompt delivery regardless of gestational age for maternal and/or fetal safety (eg, severe preeclampsia), while others are scheduled to reduce the likelihood of complications (eg, maternal anticoagulant thromboprophylaxis) or because of a fetal anomaly expected to need immediate intervention after delivery. Because infants born <39 weeks’ gestation more often require neonatal intensive care unit (NICU) admission and have increased rates of neonatal and infant morbidity and mortality, ACOG recognizes social or soft indications (eg, a history of fast labor, long distance to the hospital, maternal psychosocial discomfort) as appropriate only when the gestational age has been firmly established as ≥39 weeks, and the mother has been thoroughly informed about the risks and alternatives to scheduled birth. Births for nonmedical reasons such as the convenience of the patient, her family, and/or her caregivers should be scheduled only ≥39 0/7th weeks because maternal and infant morbidity and mortality are significantly lower than at 36-38 weeks of pregnancy.
For Editors’ Commentary, see Table of Contents
Birth certificate data provide uncertain estimates of the frequency and indications for scheduled births <39 weeks, but the frequency of scheduled births between 36 0/7th -38 6/7th weeks increased substantially between 1990-2006. The Ohio Perinatal Quality Collaborative (OPQC) is a consortium of Ohio perinatal clinicians, hospitals, and policy makers founded in 2007 to pursue a mission of using collaborative improvement science methods to reduce preterm births and improve outcomes of preterm newborns in Ohio as rapidly as possible. OPQC was funded in part by the Ohio Department of Jobs and Family Services to develop a statewide collaborative network of perinatal care sites aided by a central staff with expertise in quality improvement, data management, perinatal vital statistics, neonatology, and maternal fetal medicine. OPQC was charged to establish an ongoing statewide quality collaborative that would promote rapid adoption of care strategies known to reduce perinatal and infant morbidity and mortality in Ohio, and be reflected in state vital statistics. The growing number of NICU admissions of infants born at 36-38 weeks prompted selection of this topic by OPQC obstetric participants.
Materials and Methods
The OPQC
Twenty maternity and neonatal care hospitals in the 6 major metropolitan areas of Ohio accounting for 47% of all Ohio births agreed to share patient-level data from limited datasets that contain no identifiers other than month and hospital of birth. Participating sites signed data-sharing agreements with OPQC that specified common confidentiality and privacy principles. OPQC and member sites all obtained institutional review board approval to share the results of deidentified data housed in a central perinatal quality improvement database at Cincinnati Children’s Hospital Medical Center, Cincinnati, OH. Encrypted data are transferred electronically to OPQC via a secure, password-protected, World Wide Web-based extranet. The Health Insurance Portability and Accountability Act – specified, limited data set procedures are used for all OPQC improvement projects.
The initial project was chosen based on the following weighted criteria: documented geographic variation in population-based Ohio outcomes, available high-quality evidence, feasible interventions, demonstrated change by others using improvement methods, population impact (defined as the proportion of births affected and significant associated morbidity), and enthusiasm from clinicians. Topics considered included optimal use of antenatal corticosteroids, appropriate hospital for delivery of very low birthweight infants, and reduction of scheduled deliveries without apparent medical or obstetric indication at 36-38 weeks gestational age. Reduction of inappropriate scheduled births at 36-38 weeks was chosen because of the large number of affected pregnancies, the existence of a clear practice benchmark (ACOG practice bulletin no. 10 ), and strong enthusiasm from clinicians. The project goal was a 60% reduction in the rate of scheduled births that lacked documentation of an appropriate indication for the aggregate of all sites within 1 year.
Sites were asked to create an improvement team that included at least 1 nurse, data manager, and physician. Teams agreed to participate in monthly telephone calls and 3 face-to-face learning sessions. The OPQC Scheduled Birth Initiative was formally introduced at each site in September 2008 after the first learning session, where obstetric teams met for 2 days to learn about the rationale for the project and the Institute for Healthcare Improvement Breakthrough Series techniques of introducing and sustaining health improvement projects ( http://www.ihi.org/ihi ).
The key drivers, including a list of interventions that might enable achieving the goal, were developed by OPQC Project Team faculty. These included promotion of optimal determination of gestational age with ultrasound; use of ACOG criteria for the indication and timing of scheduled births; increased awareness among pregnant women, nurses, and physicians of the risks and benefits of births between 36-38 weeks; improved communication between obstetricians and pediatricians; and inclusion of scheduled births as part of an overall culture of safety. Practices recommended to facilitate these steps are shown in Table 1 .
|
Each site selected interventions based on the key drivers and modified them as appropriate for local use. Sites were encouraged to adopt as many of these interventions in whatever order they deemed appropriate to their site, to accomplish the goal of a 60% reduction in scheduled births that lacked documentation of a medical or obstetric indication.
A scheduled birth was defined as one in which induction of labor or cesarean birth was scheduled in advance. This definition thus included women admitted to the hospital specifically for delivery, and women already hospitalized whose birth was scheduled prior to the day of delivery.
Data were collected on a standard OPQC Scheduled Birth Data Form ( Appendix 1 ), deidentified, and then reported electronically to OPQC by a data abstractor/coder at each site who reviewed inpatient and outpatient medical records and hospital data to determine the number of scheduled births each month between 36 0/7 weeks and 38 6/7 weeks. The OPQC Scheduled Birth Data Form listed most standard indications for scheduled birth (eg, maternal diabetes, hypertension, fetal growth restriction or anomalies). The initial form lacked specificity and proved incomplete, so that coders too often entered free text instead of checking a listed indication (eg, maternal prophylactic anticoagulation was not listed as a reason for scheduled birth). A modified version of the form that included a more comprehensive list of potential indications for scheduled birth was introduced in July 2009. The revised form generates far fewer “other” or free text entries, and includes medical conditions such as seizure disorder, substance abuse, and advanced maternal age, that were entered in free text on the initial version of the form. We wanted to monitor the actual reasons listed by physicians as the rationale for scheduling a birth. Notably, ACOG does not offer a proscriptive list of appropriate vs inappropriate reasons. The responsibility to assess the appropriateness of the listed indications remained with the local quality assurance committees. For example, maternal seizure disorder is a condition that does not mandate a scheduled birth, yet some women might benefit from it. Epilepsy is a disorder that may be exacerbated by disturbed sleep or increased stress, and anticonvulsant medications may complicate administration of labor analgesia and anesthesia. Charts that listed the indication as elective , or that lacked any documentation of a reason for scheduled birth were coded as lacking a medical or obstetric indication for scheduled birth.
OPQC staff generated monthly aggregate and site-specific reports in which outcomes and process measures ( Table 2 ) were tracked for each site and the aggregate of all sites. Each site received its own data and the aggregate report each month, and could share some or all of their data with other sites if they wished during monthly conference calls, periodic webinars, and the second and third learning sessions.
