Examining the effect of hospital-level factors on mortality of very low birth weight infants using multilevel modeling

OBJECTIVE

The objective of this study was to examine the effect of hospital-level factors on mortality of very low birth weight infants using multilevel modeling.

STUDY DESIGN

This is a secondary data analysis of California maternal-infant hospital discharge data from 1997 to 2002. The study population was limited to singleton, non-anomalous, very low birth weight infants, who delivered in hospitals providing neonatal intensive care services (level-2 and higher). Hierarchical generalized linear modeling, also known as multilevel modeling, was used to adjust for individual-level confounders.

RESULT

In a multilevel model, increasing hospital volume of very low birth weight deliveries was associated with lower odds of very low birth weight mortality. Characteristics of a particular hospital's obstetrical and neonatal services (the presence of residency and fellowship training programs and the availability of perinatal and neonatal services) had no independent effect.

CONCLUSION

Using multilevel modeling, hospital volume of very low birth weight deliveries appears to be the primary driver of reduced mortality among very low birth weight infants.

Variation in elective primary cesarean delivery by patient and hospital factors

OBJECTIVE

Our purpose was to describe variation in elective primary cesarean rates by nonclinical factors.

STUDY DESIGN

With use of California discharge data and American Hospital Association data for 1995, patients were classified into 13 mutually exclusive categories for elective primary cesarean delivery. With use of recursive partitioning algorithms, women in each category were then studied to determine whether nonclinical factors were associated with elective primary cesarean delivery.

RESULTS

A total of 463,196 women were delivered at 288 hospitals, and the elective primary cesarean delivery rate was 4.25% (19,664/463,196). Risk for elective primary cesarean delivery varied by clinical condition. The most discriminant risk factors were hospital type (malpresentation, multiple gestation, macrosomia, other hypertension), maternal age (antepartum bleeding, uterine scar, soft tissue disorder, preterm, unspecified), and teaching status (herpes, severe hypertension, unengaged head).

CONCLUSION

This article presents methods that use administrative data to isolate and monitor the impact of nonclinical factors on the use of elective primary cesarean.

Variation in vaginal breech delivery rates by hospital type

OBJECTIVE

To relate vaginal breech delivery rates to the following hospital types: public, health maintenance organization, private teaching, or private nonteaching.

METHODS

In a retrospective study using administrative discharge data from Los Angeles County, California, we calculated the vaginal breech delivery rates of singleton breech deliveries during calendar years 1988 and 1991.

RESULTS

Ten thousand four hundred breech deliveries were identified, 8988 (86.4%) term and 1412 (13.6%) preterm. Twelve percent (1252 of 10,400) were vaginal deliveries (10.1% term and 24.5% preterm). Term vaginal breech deliveries varied by hospital type and were more frequent in public hospitals (28.4%, 95% confidence interval [CI] 26.1%, 30.7%) and less frequent in private nonteaching hospitals (5.4%, 95% CI 4.8%, 5.9%). Term vaginal deliveries were 2.4 to 11.3 times more likely among black women and 1.3 to 6.3 times more likely for Hispanic women across all hospital types, compared with white women in private nonteaching hospitals. There was no difference in the proportion of preterm vaginal breech deliveries by hospital type (mean 24.5%). However, with the exception of public hospitals, the proportion of vaginal breech deliveries for both term and preterm deliveries varied significantly by ethnicity.

CONCLUSION

The use of vaginal breech delivery varied by hospital type and patient ethnicity. Within private teaching and nonteaching hospitals, vaginal breech delivery was more likely for black women than for women of other ethnic groups. Further study is needed to understand the hospital policies or organizational factors, as well as the patient-related sociocultural and clinical factors, that contribute to those differences.

Monitoring, risk adjustment and strategies to decrease cesarean rates

Cesarean rate as a clinical indicator for health care quality continues to be a focus of discussion and research among clinicians and health policy advocates. Over the review period, there were several studies regarding statistical strategies for monitoring and reporting cesarean rates, clinical and nonclinical risk factors for cesarean, and clinical interventions related to the management of labor that may help to decrease the likelihood of cesarean delivery. Future research should focus on developing and refining the statistical strategies for monitoring and adjusting cesarean rates to allow for meaningful comparisons.

Fetal echogenic bowel: parameters to be considered in differential diagnosis

OBJECTIVES

To evaluate the extent that associated findings aid in the differential diagnosis and/or prognosis of fetal echogenic bowel.

METHODS

Medical history, obstetric records and outcome details were examined for 131 consecutive pregnancies with fetal hyperechogenic bowel.

RESULTS

In 62 (47%) cases, there were no visible anomalies other than hyperechogenic bowel and no evidence of growth restriction. This group included four (7%) pregnancies with Down syndrome, 15 (24%) with infection or a recent episode of influenza and eight (13%) with blood staining of amniotic fluid. In the remaining 69 (53%) cases, hyperechogenic bowel was accompanied by hydrops or nuchal edema (n = 16, 12.2%), growth restriction (n = 9, 6.9%), other markers for chromosome anomalies (n = 33, 25.2%) or multiple structural anomalies (n = 11, 8.4%). In this group, the prevalence of Down syndrome was 12%, infection or influenza was reported in 14 (20%) cases and there was blood staining of amniotic fluid in seven (10%). Cystic fibrosis screening was performed in 65 (50%) pregnancies; the results were negative in all cases and clinical assessment did not indicate cystic fibrosis in any of the 91 infants who were born alive. Maternal serum screening was performed in 41 (31%) pregnancies. High alpha-fetoprotein levels were associated with multiple abnormalities or severe growth restriction.

CONCLUSIONS

In many pregnancies with fetal hyperechogenic bowel, there are multiple factors that may explain these findings. Thus identification of one potential underlying cause should not preclude further testing. Once chromosome defects, cystic fibrosis, structural abnormalities, infection and growth restriction have been excluded, parents can be counseled that the prognosis is good, irrespective of the presence or absence of blood stained amniotic fluid.

Vaginal birth after cesarean and uterine rupture rates in California

OBJECTIVE

To describe attempted and successful vaginal birth after cesarean (VBAC) rates and uterine rupture rates for women with and without prior cesareans, and compare delivery outcomes in hospitals with different attempted VBAC rates.

METHODS

We used California hospital discharge summary data for 1995 to calculate attempted and successful VBAC rates and uterine rupture rates. We used multivariate logistic regression models to evaluate and adjust for age, ethnicity, and payment source. We report the relative risk (RR), attributable fraction, and 95% confidence intervals (CIs) for uterine rupture.

RESULTS

There were 536,785 delivery discharges during 1995. The cesarean rate was 20.8%, and 12.5% of women had histories of cesareans. Of women with histories of cesareans, 61.4% attempted VBAC and 34.8% were successful. There were 392 uterine ruptures (0.07%). Women with prior cesareans were 16.98 (95% CI 13.51, 21.43) times more likely to experience uterine rupture, attributable fraction 66% (95% CI 60%, 73%). Among women with prior cesareans, those who attempted VBAC were 1.88 (95% CI 1.45, 2.44) times as likely to have uterine rupture, attributable fraction 34% (95% CI 21%, 46%). Women who delivered in hospitals with high attempted VBAC rates were less likely to have cesarean deliveries, more likely to have successful VBACs, and more likely to experience uterine ruptures.

CONCLUSION

Uterine rupture occurs at a low rate in women with and without prior cesarean delivery. Risk of rupture is increased among women with prior cesarean delivery and among those who attempt VBAC.

Using the continuous quality improvement process to safely lower the cesarean section rate

BACKGROUND

In 1994 a five-year prospective observational study (including 38,541 singleton live-born deliveries) based on maternal and neonatal hospital administrative discharge data for DRGs 370-375 was launched at Cedars Sinai Medical Center (CSMC) in Los Angeles. In 1993 a cesarean section (C-section) reduction task force was first convened and several interventions were conducted and monitored during a two-year period. In 1995 CSMC joined the Institute for Healthcare Improvement's (IHI's) national collaborative on lowering C-section rates.

RESULTS

The first intervention involved physician education (grand rounds) and occurred during the preintervention baseline period. Providing physician-specific data had been implemented before participation in the IHI collaborative. Two other interventions were implemented before the collaborative versus 13 interventions after. The C-section rate decreased from 26.0% in the baseline period in 1993 to 20.5% in 1997, a 21.2% reduction. During the postintervention period, the C-section rate increased to 23.5%. There was no statistically or clinically significant increase in clavicular fractures, brachial plexus injuries, or cerebral hemorrhage in the four study years, compared to the baseline period.

DISCUSSION

It is possible to safely reduce C-section delivery rates. Activities are now under way to involve additional private physician leaders in the continuous quality improvement effort. Although the small increase in the C-section rate during the postintervention period may represent statistical variation, and in itself may not be clinically significant, it supports the thesis that ongoing, continuous organizational support is required to achieve and maintain gains.

Cesarean deliveries for medicaid patients: a comparison in public and private hospitals in Los Angeles county

OBJECTIVE

The aims of the study were to describe the difference in cesarean delivery rates for Medicaid patients according to hospital type and adjusted for case mix and to determine cost implications for additional cesarean deliveries.

STUDY DESIGN

This retrospective study used California discharge data for 92,800 patients delivered in 78 hospitals in Los Angeles County during 1991. Multivariable logistic regression was used to adjust for case mix and to calculate adjusted cesarean delivery rates according to hospital type. Cost estimates assumed $821 per day hospital reimbursement.

RESULTS

The unadjusted cesarean delivery rate in private nonteaching hospitals (reference group) was 24.5%, compared with 13.2%, 17.4%, and 16.5% in public, health maintenance organization, and private teaching hospitals, respectively. Adjustment for case mix decreased the cesarean delivery rate in public (9.0%), health maintenance organization (12.0%), and private teaching hospitals (8.0%). Cesarean deliveries performed on patients in private nonteaching hospitals result in an additional $13.6 million in Medicaid health care expenses.

CONCLUSIONS

There are increased health care costs related to increased cesarean deliveries performed on Medicaid patients in private nonteaching hospitals.

Maternal and infant complications in high and normal weight infants by method of delivery

OBJECTIVE

To estimate the population risks of maternal and infant complications with the birth of macrosomic (at least 4000 g) compared with normal weight infants.

METHODS

Term, singleton infants were identified from the state of Washington's birth event records database for 1990. Diagnosis codes from the Internal Classification of Diseases (9th revision) were used to identify delivery method and previously defined complications. We adjusted for maternal demographic and clinical factors using multivariable logistic regression to derive the risk of each maternal and infant complication.

RESULTS

The incidence of macrosomia was 13% (8815 of 66,086). Vaginal birth of macrosomic infants was associated with low incidence of complications except for shoulder dystocia (11%) and postpartum hemorrhage (5%). Postpartum infection was the most common complication for women who had cesarean delivery after labor (5%), and complications for women who had cesarean without labor were rare (less than 3%). Neonatal complications were rare. Among infants with shoulder dystocia, the risks of asphyxia (adjusted relative risk [RR] 1.2, 95% confidence interval [CI] 0.6, 2.3), birth trauma (RR 0.6, 95% CI 0.2, 1.6), long-bone injury (RR 1.2, 95% CI 0.6, 2.4), seizures (RR 1.0, 95% CI 0.0, 25.0), and facial palsy (RR 2.2, 95% CI 0.2, 44.4) were not significantly different for macrosomic and normal weight infants; however, macrosomic infants had a significantly increased risk of Erb palsy (RR 3.5, 95% CI 1.8, 7.5).

CONCLUSION

This population-based study showed that most macrosomic infants are delivered vaginally with low rates of maternal and neonatal complications. Macrosomic infants have higher rates of Erb palsy, but similar rates of other serious complications of shoulder dystocia when compared with normal weight infants.

Changes in indications for cesarean delivery: United States, 1985 and 1994

OBJECTIVES

The percentages of cesarean deliveries attributable to specific indications (breech, dystocia, fetal distress, and elective repeat cesarean) were computed for 1985 and 1994.

METHODS

Data were derived from the 1985 and 1994 National Hospital Discharge Surveys.

RESULTS

Dystocia was the leading indication for cesarean delivery in both years. In comparison with 1985, cesareans performed in 1994 that were attributable to dystocia and breech presentation increased, those attributable to fetal distress did not change significantly, and elective repeat cesareans declined.

CONCLUSIONS

Studying indications for cesareans can be useful for hospitals, clinicians, and researchers in determining strategies to lower primary and repeat cesarean rates.

The isolated choroid plexus cyst

OBJECTIVE

To describe the karyotypes of a population of fetuses with choroid plexus cysts and compare affected fetuses with and without additional ultrasonographic findings.

METHODS

The study population included all patients undergoing second-trimester ultrasound examination in a prenatal diagnostic program between January 1993 and October 1995. The records of all cases in which a choroid plexus cyst was found were reviewed, and information was abstracted regarding the fetal karyotype and the presence of other sonographic abnormalities.

RESULTS

Two hundred ten cases of choroid plexus cysts were identified among 7617 patients (2.8%) who underwent second-trimester ultrasound examination. The majority of the cases (181, or 86%) involved isolated choroid plexus cysts and the remaining 29 (14%) were associated with additional ultrasonographic findings. Autosomal aneuploidy was found in one patient with an isolated choroid plexus cyst (trisomy 21) and in another with additional findings (trisomy 18); the mothers of both of these patients were at least 35 years old. For those fetuses with known outcome, the risk of aneuploidy with isolated choroid plexus cyst (one in 180) was not statistically significantly different from that associated with choroid plexus cyst accompanied by other sonographic findings (one in 26). More than 1000 fetuses with choroid plexus cysts would have to be studied to determine whether such a difference was real.

CONCLUSION

Because of the rarity of aneuploidy, the reported risk for a fetus with an isolated choroid plexus cyst must be interpreted cautiously and should include the baseline risk.

Using ICD-9 codes to identify indications for primary and repeat cesarean sections: agreement with clinical records

Aggregate databases are increasingly being used to evaluate appropriateness of care, and, for cesarean sections, Anderson and Lomas' International Classification of Diseases, 9th Revision (ICD-9), coding hierarchy is a widely used tool. The aim of this study was to assess the validity of the hierarchy and expand its applicability to repeat cesareans. Hospital records of 1885 singleton cesareans were reviewed. Clinical indications and ICD-9 hierarchical codes were concordant for 83% of primary and 86% of repeat cesareans; modification allowed elective repeat cesareans to be distinguished from indicated procedures. The Anderson and Lomas ICD-9 hierarchy is a valid tool for assessing indications for cesarean. The current modification improves its clinical utility and expands its application to repeat procedures.

Repeat cesareans: how many are elective?

OBJECTIVE

To describe the clinical indications for repeat cesarean delivery and to compare these with indications for primary procedures.

METHODS

We reviewed cesarean deliveries at our academic nonprofit hospital during 1992 (n = 1885). The indication for the procedure was abstracted based on surgeon operative reports and discharge ICD-9 codes (International Classification of Diseases, Clinical Modification, 9th Edition).

RESULTS

The hospital cesarean rate was 28.7%; 34% of these were repeat procedures (n = 643). Elective cesarean delivery was the leading cause of repeat cesareans, followed by "other" indications, dystocia, breech, and fetal distress. In contrast, dystocia was the leading cause for primary cesarean, followed by fetal distress, "other," and breech presentation. One hundred women (15.6%) undergoing repeat cesarean had absolute or relative contraindications to a trial of labor.

CONCLUSIONS

Indications for cesareans using hierarchies based on ICD-9 codes do not attempt to differentiate categories of indications for repeat cesarean. Current recommendations for lowering cesarean rates by increasing vaginal birth after previous cesarean are based on aggregate data and do not recognize that some repeat cesareans are clinically indicated. A coding system designed to distinguish elective from indicated repeat cesareans would be useful for future prospective studies.

Cost of non-insulin-dependent diabetes in women with a history of gestational diabetes: implications for prevention

Women with gestational diabetes have a 50% risk of developing non-insulin-dependent diabetes mellitus within 10 years of delivery and thus constitute a well-defined target population for primary prevention. Current obstetric standards advocate universal screening of all pregnant women for gestational diabetes. Therefore, approximately half the reproductive-age United States population is screened for carbohydrate intolerance before the onset of overt disease. Continuation of dietary and behavioral changes initiated during pregnancy theoretically could delay or prevent progression to overt diabetes. We present an economic model of the health care dollars that could be saved by promoting postpartum life-style changes in women diagnosed with gestational diabetes. Assuming the incidence of diabetes could be reduced by 10, 25, or 50% in a national cohort of women with gestational diabetes, then 32, 140, or 331 million health care dollars could be saved over 10 years.

Risk factors for Erb-Duchenne palsy

The risk factors associated with the occurrence of Erb-Duchenne palsy were examined. Of 22 palsies, 18 were noted among 32,088 nondiabetic gravidas (0.56 per 1000) compared with four among 380 diabetic gravidas (10.5 per 1000), a statistically significant difference. One in six infants of diabetic gravidas who sustained shoulder dystocia experienced an Erb-Duchenne palsy. The incidence of precipitate second-stage labors was high (31.8%) among those infants who experienced the neurologic complication. This labor abnormality is not preventable and may contribute, in many ways, to the neurologic complication. Although recently graduated (less than four years' postresidency training) obstetricians, especially if placed in a high-volume practice, were more likely to experience this adverse outcome than more experienced physicians, even the most senior clinicians delivered infants who were affected.