Professional Documents
Culture Documents
OBJECTIVE: To evaluate whether there was an associ- RESULTS: Of 1,574 patients delivering twins, 897 (57%)
ation between the systematic promotion of twin vaginal were included, with 479 in the 3 years before and 418 in the
delivery and an increase in the rates of twin vaginal 3 years after the intervention. Adjusted rates of vaginal
birth. delivery increased from 32.1% (n5153) to 44.2% (n5185)
METHODS: We conducted a retrospective cohort study. before and after the intervention (P,.01), with a decrease in
We implemented a quality improvement initiative pro- elective cesarean delivery from 54.6% (n5479) to 44.3%
moting twin vaginal delivery at an academic tertiary care (n5185) (P,.01). Rates of breech extraction increased after
center in 2013. The program included a needs assessment, the intervention (5.7% vs 9.3%, P5.04). However, there was
simulation of vaginal twin delivery, online educational no difference in the rate of change in twin vaginal birth in
material, and the expansion of a dedicated twin clinic. We the time period before (1.35% annual increase, P5.76) or
analyzed rates of twin vaginal birth in pregnancies at or after (5.8% annual increase, P5.40) the intervention.
beyond 24 weeks of gestation without a contraindication CONCLUSIONS: Although we observed an increased
to labor, prior uterine surgery, or a demise or lethal rate of twin vaginal birth in the time period after our
anomaly of either twin. Using linear regression, we intervention, because the rates of increase before and
calculated annual adjusted rates of twin vaginal birth from after the intervention were not statistically different, the
2010 to 2015 and in the 3 years before and after our increase is not attributable to our intervention and is
intervention. We performed an interrupted time-series more properly attributed to secular trend.
analysis estimating rates of change before and after the (Obstet Gynecol 2018;132:85–93)
intervention to account for the influence of secular trend. DOI: 10.1097/AOG.0000000000002680
VOL. 132, NO. 1, JULY 2018 Easter et al Intervention to Decrease Twin Cesarean Deliveries 87
exposure to the patient- and health care provider- increase in rates of prenatal care by maternal–fetal
focused interventions. They also cared for patients medicine subspecialists in general or in the special-
with higher rates of maternal or fetal comorbidities ized twin clinic specifically across the study period.
that may differentially affect their probability of a suc- When considering patient characteristics stratified by
cessful vaginal birth. As opposed to adjusting for pre- practice type, patients in the faculty practice tended to
natal care provider, we chose to perform a stratified have higher rates of monochorionicity (27.4% vs 7.8%,
analysis according to whether a patient received care P,.01), spontaneous conception (46.9% vs 31.6%,
in the faculty practice compared with a private prac- P,.01), and fetal anomalies (8.7% vs 1.5%, P,.01)
tice. This study was approved by the Partners Human compared with patients seen by private practitioners
Research Committee (Protocol #2012P001737, (Appendix 1, available online at http://links.lww.com/
approved September 12, 2012). All analyses were per- AOG/B101). Faculty patients delivered at a lower
formed with SAS 9.4. median gestational age (36 weeks of gestation, inter-
quartile range 33–37 weeks of gestation) compared
RESULTS with patients of private health care providers (median
Of 1,574 patients delivering twins at 24 weeks of 37 weeks of gestation, interquartile range 35–38 weeks
gestation or greater, 897 patients (57%) were eligible of gestation, P,.01).
for inclusion in the present study. Table 1 compares the The frequency and adjusted rates of the primary
maternal, fetal, and health care provider characteristics and secondary outcomes of interest according to year
according to year of delivery. There were no significant of delivery are presented in Table 2. In an analysis
changes in rates of nulliparity, fetal presentation, cho- adjusted for maternal age and parity, adjusted rates of
rionicity, or other maternal or fetal characteristics vaginal delivery increased over the study period, with
across the study period. There was an expected a nadir in 2010 (28.2%) to a peak in 2015 (55.7%), with
a statistically significant difference in rates of vaginal and breech extraction increased among faculty patients
birth across years (P,.01). Conversely, rates of elective but not private patients. Similarly, rates of elective cesar-
cesarean delivery showed a stable decrease over the ean delivery decreased for faculty patients while remain-
study period from a peak 59.5% in 2010 to a nadir of ing stable for patients of private health care providers.
27.6% in 2015 (P,.01 across years). Delivery by breech Even after controlling for gestational age, adjusted rates
extraction increased over the time period from a nadir of neonatal morbidity decreased among patients of pri-
of 4.3% in 2010 to a peak of 13.3% in 2014 (P5.03 vate practitioners after the 2013 intervention. This may
across years). Rates of combined delivery, operative reflect a decrease in the number of high-risk patients
vaginal delivery, and composite maternal morbidity such as monochorionic twins seen by private health care
were similar over time. Rates of composite neonatal providers after promotion of the specialized twin clinic.
morbidity for the nonpresenting twin specifically Figure 1 displays the annual adjusted rate of vagi-
decreased over time even after controlling for gesta- nal birth for the cohort overall and stratified by health
tional age. Table 3 presents these outcomes of interest care provider type before and after the 2013 interven-
stratified by health care provider type. The adjusted rate tion (unadjusted rates demonstrated in Appendix 2,
of vaginal delivery increased substantially for patients available online at http://links.lww.com/AOG/B101).
seen in the faculty practice (24.8% in 2010 vs 60.9% in In the interrupted time-series analysis controlling for
2016, P,.01 across years) with no statistically significant secular trend, there was no significant effect of calendar
difference in the adjusted rate of vaginal delivery year on the rate of twin vaginal birth in the time period
(P5.21) or elective cesarean delivery (P5.39) for lower before (1.35% annual increase, P5.76) or after (5.8%
risk patients cared for in the private practices. annual increase, P5.40) the intervention. Similarly,
We next analyzed outcomes in the 3 years before the rate of change of twin vaginal birth was similar
and after our 2013 intervention. The adjusted rates of before and after the intervention for faculty patients
primary and secondary outcomes of interest for the (4.7%, P5.31; 3.2%, P5.58) or private patients
group overall and stratified by health care provider (3.6%, P5.28; 1.6%, P5.70). There was no statisti-
type are presented in Table 4. Rates of vaginal birth cally significant difference in the slope of the
VOL. 132, NO. 1, JULY 2018 Easter et al Intervention to Decrease Twin Cesarean Deliveries 89
Table 3. Adjusted* Rates of Primary and Select Secondary Outcomes According to Year of Delivery and
Affiliation of Delivering Provider
Faculty Provider
n 82 74 63 75 75 66
Vaginal delivery 20 (24.8) 28 (36.9) 21 (34.1) 32 (45.3) 37 (49.2) 37 (60.9) ,.01
Operative vaginal delivery 3 (3.5) 6 (8.2) 3 (4.5) 6 (8.3) 11 (14.4) 5 (7.8) .17
Breech extraction 4 (5.1) 7 (9.1) 2 (3.6) 5 (7.3) 14 (18.8) 8 (12.8) .01
Any cesarean delivery 62 (75.3) 46 (63.3) 42 (65.9) 43 (54.9) 38 (50.9) 29 (46.4) ,.01
Elective cesarean delivery 52 (63.0) 35 (48.1) 36 (56.5) 31 (39.2) 30 (40.1) 18 (19.0) ,.01
Combined delivery 2 (2.5) 5 (6.7) 1 (1.6) 4 (5.4) 3 (4.0) 2 (2.9) .62
Composite maternal morbidity 11 (13.5) 11 (15.0) 7 (11.2) 9 (11.2) 14 (19.0) 7 (12.7) .76
Any composite neonatal morbidity‡ 18 (15.5) 9 (10.3) 7 (10.4) 6 (7.0) 8 (11.6) 7 (13.6) .54
Private provider
n 87 72 101 83 74 45
Vaginal delivery 27 (31.4) 29 (39.5) 28 (28.6) 33 (38.0) 27 (39.1) 19 (48.3) 0.21
Operative vaginal delivery 4 (4.6) 6 (8.4) 5 (4.9) 3 (3.7) 7 (9.3) 4 (8.9) 0.56
Breech extraction 3 (3.6) 6 (8.2) 5 (5.2) 5 (5.5) 5 (7.5) 2 (4.5) 0.82
Any cesarean delivery 60 (68.8) 43 (60.7) 73 (71.5) 50 (62.2) 47 (61.0) 26 (68.9) 0.41
Elective cesarean delivery 49 (56.1) 32 (45.2) 58 (56.6) 43 (53.2) 41 (53.2) 22 (40.3) 0.39
Combined delivery 3 (3.5) 3 (4.0) 6 (6.1) 2 (2.1) 2 (3.2) 0 (0) 0.47
Composite maternal morbidity 8 (9.1) 10 (13.9) 5 (4.8) 5 (6.3) 9 (11.8) 3 (6.0) 0.27
Any composite neonatal morbidity‡ 7 (5.3) 7 (5.8) 6 (2.5) 3 (0.14) 3 (0.35) 0 (0.83) 0.36
* Adjusted for maternal age and parity.
†
Omnibus P value across years.
‡
Adjusted for gestational age.
Vaginal delivery 153 (32.1) 185 (44.2) ,.01 69 (31.5) 106 (49.1) ,.01 84 (32.4) 79 (39.0) .12
Operative 27 (5.6) 36 (8.7) .07 12 (5.4) 22 (10.3) .06 15 (5.8) 14 (6.9) .61
vaginal
delivery
Breech 27 (5.7) 39 (9.3) .04 13 (6.0) 27 (12.5) .02 14 (5.4) 12 (5.9) .83
extraction
Any cesarean 326 (68.0) 233 (55.8) ,.01 150 (68.5) 110 (50.9) ,.01 176 (67.6) 123 (61.0) .12
delivery
Elective 262 (54.6) 185 (44.3) ,.01 123 (56.2) 79 (36.6) ,.01 139 (53.4) 106 (52.5) .84
cesarean
delivery
Combined 20 (4.2) 13 (3.1) .38 8 (3.7) 9 (4.2) .79 12 (4.6) 4 (1.9) .11
delivery
Maternal morbidity 52 (10.9) 47 (11.3) .85 29 (13.3) 30 (13.8) .89 23 (8.8) 17 (8.5) .91
Neonatal 54 (10.5) 27 (7.3) .06 34 (13.7) 21 (11.5) .42 20 (7.4) 6 (3.4) .04
morbidity†
Morbidity of twin 40 (7.7) 23 (6.3) .35 23 (8.9) 18 (9.9) .67 17 (6.3) 5 (2.8) .06
A†
Morbidity of twin 37 (7.1) 14 (4.0) .02 26 (10.6) 11 (6.4) .08 11 (4.0) 14 (1.8) .13
B†
* Rates adjusted for maternal age and parity unless otherwise indicated.
†
Rates of neonatal outcomes of interest adjusted for gestational age.
elective cesarean delivery rather than a greater our targeted intervention had no effect on the
likelihood of success among women laboring with naturally rising institutional rate of twin vaginal
twins. Notably, this increase in attempted vaginal birth, these observed secular changes do have
birth was accompanied by higher rates of breech important implications for clinicians and policy-
extraction and operative vaginal delivery without makers interested in decreasing rates of primary
a rise in morbidity or combined delivery. Although cesarean delivery.
VOL. 132, NO. 1, JULY 2018 Easter et al Intervention to Decrease Twin Cesarean Deliveries 91
VOL. 132, NO. 1, JULY 2018 Easter et al Intervention to Decrease Twin Cesarean Deliveries 93