Decision-case mix model for analyzing variation in cesarean rates

Regional variations in childbirth interventions in the Netherlands: a nationwide explorative study

BACKGROUND

Although interventions in childbirth are important in order to prevent neonatal and maternal morbidity and mortality, non-indicated use may cause avoidable harm. Regional variations in intervention rates, which cannot be explained by maternal characteristics, may indicate over- and underuse. The aim of this study is to explore regional variations in childbirth interventions in the Netherlands and their associations with interventions and adverse outcomes, controlled for maternal characteristics.

METHODS

Childbirth intervention rates were compared between twelve Dutch regions, using data from the national perinatal birth register for 2010-2013. All single childbirths from 37 weeks' gestation onwards were included. Primary outcomes were induction and augmentation of labour, pain medication, instrumental birth, caesarean section (prelabour, intrapartum) and paediatric involvement. Secondary outcomes were adverse neonatal and maternal outcomes. Multivariable logistic regression analyses were used to adjust for maternal characteristics. Associations were expressed in Spearman's rank correlation coefficients.

RESULTS

Most variation was found for type of pain medication and paediatric involvement. Epidural analgesia rates varied from between 12 and 38% (nulliparous) and from between 5 and 14% (multiparous women). These rates were negatively correlated with rates of other pharmacological pain relief, which varied from between 15 and 43% (nulliparous) and from between 10 and 27% (multiparous). Rates of paediatric involvement varied from between 37 and 60% (nulliparous) and from between 26 and 43% (multiparous). For instrumental vaginal births, rates varied from between 16 and 19% (nulliparous) and from between 3 and 4% (multiparous). For intrapartum caesarean section, the variation was 13-15% and 5-6%, respectively. A positive correlation was found between intervention rates in midwife-led and obstetrician-led care at the onset of labour within the same region. Adverse neonatal and maternal outcomes were not lower in regions with higher intervention rates. Higher augmentation of labour rates correlated with higher rates of severe postpartum haemorrhage.

CONCLUSIONS

Most variation was found for type of pain medication and paediatric involvement, and least for instrumental vaginal births and intrapartum caesarean sections. Care providers and policy makers should critically audit remarkable variations, since these may be unwarranted. Limited variation for some interventions may indicate consensus for their use. Further research should focus on variations in evidence-based interventions and indications for the use of interventions in childbirth.

Variation in referrals to secondary obstetrician-led care among primary midwifery care practices in the Netherlands: a nationwide cohort study

BACKGROUND

The primary aim of this study was to describe the variation in intrapartum referral rates in midwifery practices in the Netherlands. Secondly, we wanted to explore the association between the practice referral rate and a woman's chance of an instrumental birth (caesarean section or vaginal instrumental birth).

METHODS

We performed an observational study, using the Dutch national perinatal database. Low risk births in all primary care midwifery practices over the period 2008-2010 were selected. Intrapartum referral rates were calculated. The referral rate among nulliparous women was used to divide the practices in three tertile groups. In a multilevel logistic regression analysis the association between the referral rate and the chance of an instrumental birth was examined.

RESULTS

The intrapartum referral rate varied from 9.7 to 63.7 percent (mean 37.8; SD 7.0), and for nulliparous women from 13.8 to 78.1 percent (mean 56.8; SD 8.4). The variation occurred predominantly in non-urgent referrals in the first stage of labour. In the practices in the lowest tertile group more nulliparous women had a spontaneous vaginal birth compared to the middle and highest tertile group (T1: 77.3%, T2:73.5%, T3: 72.0%). For multiparous women the spontaneous vaginal birth rate was 97%. Compared to the lowest tertile group the odds ratios for nulliparous women for an instrumental birth were 1.22 (CI 1.16-1.31) and 1.33 (CI 1.25-1.41) in the middle and high tertile groups. This association was no longer significant after controlling for obstetric interventions (pain relief or augmentation).

CONCLUSIONS

The wide variation between referral rates may not be explained by medical factors or client characteristics alone. A high intrapartum referral rate in a midwifery practice is associated with an increased chance of an instrumental birth for nulliparous women, which is mediated by the increased use of obstetric interventions. Midwives should critically evaluate their referral behaviour. A high referral rate may indicate that more interventions are applied than necessary. This may lead to a lower chance of a spontaneous vaginal birth and a higher risk on a PPH. However, a low referral rate should not be achieved at the cost of perinatal safety.

Complications of labor induction among multiparous women in a community-based hospital system

OBJECTIVE

The purpose of this study was to examine complications of labor induction compared to spontaneous labor in multiparas.

STUDY DESIGN

This was a retrospective cohort study of multiparous women with live, singleton pregnancies at term, who had no contraindications to labor or labor induction. Cesarean delivery was the primary outcome.

RESULTS

Of the study subjects, 7208 experienced spontaneous labor, 2190 underwent labor induction with oxytocin, and 239 underwent labor induction requiring cervical ripening agents. Oxytocin-induced multiparas were 37% more likely to require cesarean compared to those with spontaneous labor (OR, 1.37; 95% CI, 1.10-1.71) and nearly 3 times more likely to undergo cesarean when cervical ripening agents were used (OR, 2.82; 95% CI, 1.84-4.53). Women requiring cervical ripening were also 10 times more likely to spend more than 12 hours in labor than those with spontaneous labor.

CONCLUSION

Multiparas undergoing labor induction are at increased risk for obstetric complications compared to spontaneous labor.

Clinical case mix adjustment of cesarean delivery rates in U.S. military hospitals, 2002

OBJECTIVE

To assess whether significant variations in observed cesarean rates in U.S. military hospitals may be attributed to differences in clinical case mix.

METHODS

Hospital discharge records for births in U.S. military hospitals in 2002 were grouped into mutually exclusive clinical strata to calculate predicted cesarean rates for subgroups defined by maternal race, health plan, hospital location, delivery volume, teaching status, and neonatal intensive care unit (NICU) status. The 95% confidence interval (CI) around each standardized ratio (SR) of the observed-to-predicted cesarean rate was used to assess statistical significance.

RESULTS

Observed cesarean rates were significantly higher than predicted rates for small hospitals (23.1% and 20.4%, respectively, SR 1.13, 95% CI 1.08-1.19), teaching hospitals (23.7% and 22.5%, respectively, SR 1.05, 95% CI 1.02-1.08), black women (25.1% and 22.8%, respectively, SR 1.10, 95% CI 1.05-1.14), and other minorities (22.7%, and 21.6%, respectively, SR 1.05, 95% CI 1.01-1.09). No significant differences between observed and predicted cesarean rates were found across hospital locations or NICU status. Significant differences found for non-managed care beneficiaries were attributed to teaching status of the hospitals in which they delivered.

CONCLUSION

Clinical case mix does not adequately account for the relatively high rates of cesarean delivery observed for small hospitals and teaching hospitals and among black women in the study population. Further study is recommended to identify additional clinical and nonclinical factors that should be considered when comparing performance across institutions, health plans, or individual providers.

Variability in interventions with pulmonary artery catheter data

OBJECTIVES

To determine if intensivists given PAC data from critically ill patients make uniform management choices.

DESIGN

Cross-sectional survey of board-certified intensivists.

SETTING

Medical intensive care unit.

PARTICIPANTS

Board-certified intensivists who are members of the American College of Chest Physicians and Society of Critical Care Medicine.

INTERVENTIONS

A survey questionnaire containing three medical intensive care clinical vignettes was mailed to critical care physicians. Each vignette contained PAC data and one-half of the surveys contained echocardiographic (Echo) information. Every respondent was asked to select one of six interventions for each vignette.

MEASUREMENTS

There were 126 evaluable surveys returned. In vignette 1 an intervention (none of the above) was selected by more than 50% of respondents. In vignettes 2 and 3, the most frequent selection was chosen only 44 and 37% of the times, respectively. There was a significant difference in the distribution of management choices between te Echo and the non-Echo subgroups.

CONCLUSIONS

There is a significant heterogeneity in selecting an intervention based on PAC data among intensivists. The presence of Echo information may change the intervention selected but does not reduce heterogeneity. Any randomized trial evaluating efficacy of PAC's will have to have strict treatment protocols.