If you’ve ever wondered how much could be accomplished if a hospital made a concerted effort to reduce cesareans—and not because an insurance company held a metaphorical gun to its head—this study is for you. EurekAlert reports that Boston’s Beth Deaconess Hospital reduced their cesarean rate in low-risk 1st-time mothers from 35% to 21% and the overall cesarean rate from 40% to 29%. Let’s look at the study to see how they did it as well as get an eye-opening look on how non-medical factors drive cesarean rates.
Background & Strategies
Beth Deaconess is a large, Level 3, meaning it is equipped and staffed to serve high-risk women and babies, teaching hospital. From 2008 through 2015, a quality improvement team phased in a series of measures designed to reduce the cesarean rate in low-risk (full-term, one head-down baby), 1st-time mothers, A.K.A. nulliparous-term-singleton-vertex or NTSV. While the effort was led by a physician champion, it involved other leaders, planners, and stakeholders in the process, which the study authors believe was essential to the project’s success and sustainability.
Of note, it inaugurated a midwifery service in the spring of 2014, which means that cesarean rate decreases prior to that date are entirely attributable to modifications in obstetric management and that decreases in the final year, in which midwives were available to about one-third of the women, might be at least partially attributable to their different model of care.
The team identified and devised strategies for five factors that impacted cesarean rates in their hospital:
- Interpretation and management of fetal heart rate tracing: In light of the knowledge that tracings that are neither clearly normal or grossly abnormal have little predictive value for newborn outcomes, the team sought to reduce cesareans for non-reassuring tracing by removing “failure to perform a cesarean for non-reassuring tracing” from their peer review committee’s quality measures and requiring that more detailed, standardized language be used to document tracings of concern beyond merely “non-reassuring.”
- Tolerance for labor: The team addressed several aspects that they felt affected willingness to allow labor to continue: They educated staff on the modern understanding of normal labor progress, redefined “arrest of labor” to align with that understanding, and provided leaders on the labor & delivery unit to support doctors in having patience. They mandated attendance at simulation training in managing shoulder dystocia (the head is born, but the shoulders hang up behind the pubic bone) and conducting instrumental deliveries. In addition, they encouraged vaginal birth after cesarean (VBAC), which doesn’t apply to their target population, but it was felt that a culture preferencing vaginal birth whenever safely possible could affect attitudes towards first births as well.
- Labor induction: The team began by putting a hard stop on inductions prior to 39 wks gestation and progressed from there to the current guidelines, which consider gestational age, whether this is a first baby, and cervical readiness in determining timing of induction and which limit most inductions prior to 41 wks. Inductions that don’t meet guideline criteria must be approved by the Director of Labor & Delivery. Induction guidelines also cover appropriate use of cervical ripening techniques. In addition, the team developed a guideline to address too frequent contractions (tachysystole), a complication that can lead to avoidable cesareans when labor is being induced or augmented.
- Clinical practice environment: This, too, encompassed several elements. With respect to cesarean management, the team changed the system for scheduling cesareans because:
We theorized that an inability to reliably predict when scheduled cesarean deliveries would occur and how long they would take resulted in pressure on the physicians to schedule planned cesarean deliveries for days when they were on call so that they would not interfere with clinical responsibilities off of the Labor and Delivery Unit. As a result, an individual provider would be responsible for more deliveries in a given shift. We hypothesized that the increase in workload, need to multitask, and the need to coordinate multiple surgical cases may drive providers to perform intrapartum [during labor], unplanned cesarean deliveries earlier than they would otherwise.
Accordingly, they changed the scheduling system so that scheduled cesareans would start on time consistently and standardized the steps for carrying them out to increase efficiency. They also instituted emergency cesarean drills with the intent that doctors could feel confident that urgent cesareans could be performed expeditiously and therefore feel less pressure to proceed to cesarean pre-emptively. In addition, they adopted “more liberal guidelines” (undefined) for labor companions to increase the likelihood of women having continuous support.
- Awareness of the NTSV cesarean rate: To draw attention to the effects of individual practice style, the team implemented an audit and feedback process, providing all physicians with their personal and the departmental rates twice a year.
Results
Beth Deaconess had 15,144 NTSV deliveries from 2008 to 2015. As noted above, the cesarean rate in this population fell from 35% to 21% as the strategies came online, and the hospital’s overall cesarean rate fell from 40% to 29%. (No further data are provided on overall rates since that was not the study’s intent.) As also noted, the inauguration of a midwifery service in 2014 may have played a role in further reducing cesarean rates in the final years. The instrumental vaginal delivery rate remained the same over the time period.
Looking at outcomes, the mean gestational age at delivery increased slightly (39.3 to 39.6 wk), and there were no differences in the incidence of 5-min Apgar < 5 (a score indicating the baby’s condition at birth), admission to neonatal intensive care for more than 24 hr, or the incidence of shoulder dystocia. The incidence of meconium aspiration syndrome (breathing difficulties arising from inhaling particles of stool) rose from 1 per 1000 to 9 per 1000, but study authors note that this is still half the rate reported in deliveries at term. Despite the rise in the percentage of vaginal births, the incidence of episiotomy (cutting the vaginal opening to enlarge it for birth) declined (16% to 3%) as did the incidence of tears into, but not through, the anal sphincter (3rd-degree tears) (4% to 2%). The incidence of tears through the anal sphincter (4th-degree tears) was unchanged (range: 2-4 per 1000). Blood transfusions increased compared with the baseline years before the study began, but held steady (range: 11-16 per 1000) during the study years, and as the study’s authors point out, the incidence with 1st cesarean is 20-40 per 1000. In short, the fall in cesarean rate did no clinically significant harm and had some benefits, not the least of which was avoiding the consequences for future pregnancies of having a prior cesarean.
A “significant unintended consequence,” as the authors termed it, was an increase in occupied beds in the labor unit because more women were spending more time in labor. The authors don’t say so, but slower bed turnover poses a potent disincentive for decreasing cesareans.
The Take-Away
Beth Deaconess is to be congratulated for beating the Healthy People 2020 24% NTSV goal by three percentage points and dropping their overall rate from 40% to 29%. It bears reminding, however, that these are by no means optimal rates.
A little history is in order: Healthy People 2000, the first set of U.S. government benchmarks, set a cesarean rate goal of 15%. This was, mind you, an overall cesarean rate, not just a rate in low-risk 1st-time mothers. Healthy People 2010 amended this goal to 15% in NTSV women, probably in light of the fact that VBAC had nearly disappeared, and Healthy People 2020 weakened it to a 10% decrease from the 2007 rate of 27% in NTSV women, doubtless because a 15% rate in NTSV women was so far in the rear-view mirror as to be unachievable. However, 15% overall is still what the research supports for countries or regions because it is the rate at which no further improvements are seen in newborn or maternal mortality.
While Beth Deaconess should be cut some slack because they are a high-risk institution, an overall rate twice 15% is still much higher than it should be. As for NTSV women, we know, too, from out-of-hospital studies that the cesarean rate in healthy, low-risk 1st-time mothers ranges from 7% to 13% (Birthplace in England 2011; Hutton 2015; Janssen 2009; Johnson 2005; Rooks 1989; Van der Hulst 2004), and while some NTSV women at Beth Deaconess had health complications (6-9% high blood pressure; 3-4% pre-eclampsia/eclampsia; 3-6% gestational diabetes), that isn’t enough to explain the large difference between Beth Deaconess and even the highest rate in women planning out-of-hospital birth.
So, what’s our take-away? On the one hand, this study illustrates both the degree to which non-medical factors embedded in the system and medical-management thinking drive the cesarean rate in the wrong direction while at the same time it shows the degree to which a committed team can walk that cesarean rate back if they set their minds to it. On the other hand, their effort, admirable as it is, doesn’t go nearly far enough. Sometimes, the reward for a job well done is a harder job. Let us hope that the staff at Beth Deaconess see it that way too.
###
