Objective
Maternal morbidity is increasing in the United States. Our objectives were to examine whether a labor and delivery (L&D) provider model with regular maternal-fetal medicine (MFM) coverage decreases the rates of maternal morbidity during delivery hospitalizations and has an impact on obstetrician-gynecologist residents’ perceptions of safety and education.
Study Design
We performed a retrospective cohort study to compare the rates of maternal morbidity before and after the implementation of an MFM-centered coverage model on L&D. Outcomes were identified using International Classification of Diseases , ninth revision, codes. The primary outcome was a composite of severe maternal morbidity. Additionally, obstetrician-gynecologist residents completed an anonymous survey asking them to compare coverage models, and their Council on Resident Education in Obstetrics and Gynecology examination scores were compared.
Results
Data from 4715 deliveries were included. There were no differences in composite morbidity or individual adverse outcomes. Most residents (81.3%) preferred the new provider model, with median 5-point Likert scores indicating perceived increases in safety and education. Mean Council on Resident Education in Obstetrics and Gynecology scores improved in the 18 residents exposed to both models.
Conclusion
Although the MFM-centered provider model appears to have had a positive impact on residents’ perceptions of safety and education, it was not associated with significant changes in severe maternal morbidity.
In the United States, epidemiologic trends suggest that maternal mortality has not decreased for nearly 30 years and may actually be increasing. Based on the monitoring of the Centers for Disease Control and Prevention’s Pregnancy Mortality Surveillance System, the pregnancy-related mortality ratio from 1998 to 2005 was 14.5 per 100,000 live births. This represents the highest reported mortality ratio since the Surveillance System’s inception in 1986. Among African Americans, the maternal mortality ratio reached a peak of 37.7 per 100,000 in 2005, which is more than 3-fold higher rate than among white women.
The mortality ratio has been compared with the tip of the iceberg. For every maternal death in the United States, approximately 50 women experience severe morbidity. There are approximately 52,000 cases of severe morbidity among pregnant American women each year.
Given the magnitude of this problem, perinatologists were charged to renew their attention to issues related to maternal mortality and morbidity by the 2010 article, “Where is the ‘M’ in maternal-fetal medicine?” As a result, a consensus panel of national experts convened in 2012 to suggest strategies for the improvement of maternal outcomes in the United States. One suggestion was to increase the maternal-fetal medicine (MFM) provider’s role in inpatient care by working alongside laborists and generalists as key team members in the delivery of high-risk maternal and fetal care. To date, however, there remains a lack of evidence regarding the efficacy of such an intervention.
To that end, the primary objective of this study was to examine whether an MFM-centered labor and delivery (L&D) coverage model improves maternal outcomes during delivery hospitalizations compared with a generalist-based model. A second objective was to examine whether obstetrics-gynecology residents’ perceptions about safety and their educational experiences in L&D had a positive impact by an MFM-centered coverage model. We hypothesized that there would be improved maternal outcomes with an MFM-centered L&D coverage model because of the greater role of MFM providers in inpatient care and that residents would report improvement in the safety culture as well as improved educational experiences. With a greater didactic role on L&D, we hypothesized that MFM’s impact would translate to improved outcomes, even when MFM providers were not directly involved in inpatient care, such as nights or weekends staffed by generalists.
Materials and Methods
We performed a retrospective cohort study of all women who delivered at the Hospital of the University of Pennsylvania (HUP) during an 18-month period. HUP is an urban tertiary care hospital located in Philadelphia, PA, that has approximately 4200 deliveries annually. The hospital has an obstetrics-gynecology residency program as well as subspecialty fellowships in MFM, reproductive endocrinology and infertility, gynecological oncology, family planning, and urogynecology. The study was approved by the HUP Institutional Review Board.
Prior to July 1, 2012, all laboring patients at our institution were primarily managed by generalists or hospitalists who staffed L&D. An MFM provider team, which was comprised of an attending physician and fellow, staffed the antepartum service and covered L&D 1 day and 1 night per week. MFM providers were also available for L&D consultations throughout the week, but their presence on the labor floor was limited.
On July 1, 2012, the L&D provider model was modified such that MFM providers began to staff L&D 7 days per week from 7:00 am to 6:00 pm . In addition to directing care of the antepartum service, an MFM attending physician and fellow worked in conjunction with a generalist attending physician or hospitalist to provide intrapartum care during the postexposure period. During the evenings, 2 generalist attending physicians and/or hospitalists staffed L&D, and the MFM fellow and attending physician were available for phone consultation via pager call.
We identified all women who delivered at our institution during the 18-month period from July 1, 2011, to Feb. 1, 2013. Women were divided into 2 groups based on the timing of their delivery relative to the change in the L&D provider coverage model. Specifically, we defined the period prior to MFM coverage (PRE) to be July 1, 2011, to Feb. 1, 2012, and the period with increased MFM coverage (POST) to be July 1, 2012, to Feb. 1, 2013. These months were chosen to ensure that relative differences in resident educational experience associated with a later time period in the academic year would not influence outcomes.
Because of our belief that practice patterns would change rapidly as a result of the new coverage model, no washout period was included. Patients who delivered from Feb. 1, 2012, to June 30, 2012 were excluded from analyses. No other significant systems changes were implemented on L&D during the study period.
Outcome data were abstracted from prepopulated International Classification of Diseases , ninth revision (ICD-9) codes and from the electronic medical record. The primary outcome was derived exclusively from ICD-9 codes. Secondary outcomes were derived from both ICD-9 codes and our institution’s electronic medical record on L&D. The electronic medical record, as completed by physician and nursing staff in clinical settings, is a continuous database that may be queried for research purposes. Charts from patients admitted to the intensive care unit (ICU) were verified for accuracy by the primary author (J.S.B.).
The primary outcome was defined as a composite of maternal mortality and severe maternal morbidity ( Table 1 ). The composite outcome, which reflected ICD-9 codes of severe maternal morbidity, was based on previously published work and was defined as 1 or more of the following: maternal death, acute renal failure, liver failure, respiratory failure, obstetric shock, cerebrovascular accident, pulmonary embolism, amniotic fluid embolism, eclampsia, septicemia, complications of anesthesia, cardiac events/procedures, mechanical ventilation, blood transfusion, and invasive hemodynamic monitoring.
| Morbidity | ICD-9 code(s) |
|---|---|
| Diagnosis-based codes | |
| Renal failure | 584, 586, 669.30,2,4 |
| Liver failure | 570 |
| Respiratory failure a | 518.4, 518.5, 518.81,2,4, 799.1 |
| Obstetric shock | 669.10,1,2,3,4 |
| Cerebrovascular accident | 430, 431, 432, 433, 434, 436, 671.50,1,2,3,4, 674.00,1,2,3,4 |
| Embolism | 673.00,1,2,3,4, 673.20,1,2,3,4, 673.30,1,2,3,4, 673.80,1,2,3,4, 415.11, 415.19, 673.10,1,2,3,4 |
| Eclampsia | 642.60,1,2,3,4 |
| Septicemia | 038 |
| Complications of anesthesia | 668.00,1,2,3,4 |
| Procedure-based codes | |
| Cardiac events/procedures | 428, 425, 427.5, 410, 99.60, 99.62, 99.62, 99.63, 99.64, 99.69, 89.60-64 |
| Mechanical ventilation | 96.70, 96.71, 96.72 |
| Transfusion | 99.03, 99.04 |
a Respiratory failure includes mechanical ventilation codes.
Secondary maternal outcomes included intraoperative bowel and bladder injury, complicated perineal laceration (eg, third/fourth-degree lacerations), ICU admission, and hospital length of stay as well as infectious complications such as wound infection, wound dehiscence, chorioamnionitis, and endometritis. The terms wound infection, wound dehiscence, chorioamnionitis, and endometritis, were defined clinically by the providers managing each patient. If the provider believed that the patient met criteria for one of these diagnoses, it was noted in the medical record.
Generally, providers at our institution define wound infection as a postoperative fever of 100.4°F or higher in the setting of incisional skin changes such as erythema and exudate. Wound dehiscence is defined as a wound infection plus skin separation requiring debridement and/or wound packing. Chorioamnionitis is defined as a maternal fever of 100.4°F or higher in the setting of maternal or fetal tachycardia, uterine tenderness, and/or foul-smelling lochia. Endometritis is defined as a maternal fever of 100.4°F or higher in the setting of postpartum fundal tenderness.
Additionally, we hypothesized that the increased MFM presence on the labor floor could have an impact on practice patterns. We therefore also included rates of induction of labor, trials of labor after cesarean deliveries (TOLAC), vaginal birth after cesarean deliveries (VBAC), operative vaginal deliveries, and the administration of medications (eg, magnesium sulfate and betamethasone) as secondary outcomes.
We developed an online survey to assess residents’ opinions about educational experiences on L&D. The survey was anonymously disseminated via surveymonkey.com (SurveyMonkey Inc, Palo Alto, CA). The survey included questions about maternal safety, communication, and educational opportunities on L&D. We also queried perceptions about rates of various obstetrical procedures including vaginal breech extractions, forceps assisted vaginal deliveries, and obstetrical ultrasounds as well as residents’ comfort performing these procedures. Finally, residents were asked whether they preferred the new L&D coverage model. Only residents who had been exposed to both provider models (eg, postgraduate years 2, 3, and 4) were invited to participate in the survey. Mean overall and obstetric Council on Resident Education in Obstetrics and Gynecology (CREOG) scores from PRE to POST were also compared.
Statistical analyses were performed using Stata version 10.1 (StataCorp LP, College Station, TX). Categorical data were compared across PRE and POST using χ 2 tests. Means and medians of continuous data were compared using Student t tests and Wilcoxon rank-sum tests as indicated. An a priori power calculation was not performed because this cohort represents a sample size of convenience.
Results
There were 4715 deliveries during the study period. A total of 2286 deliveries were performed in PRE, and 2429 deliveries were performed in POST. MFM providers performed significantly more deliveries in POST after the implementation of the MFM-centered provider model (41% vs 24%, P < .001).
Demographic information is presented in Table 2 . Although patients in POST were slightly older, there were no other significant differences in the medical comorbidities between the groups.
| Variable | PRE (n = 2286) | POST (n = 2429) | P value a |
|---|---|---|---|
| Age, y | 26 (22–31) | 27 (22–32) | .03 |
| Gravida | 2 (1–4) | 2 (1–4) | .44 |
| Parity | 1 (0–2) | 1 (0–2) | .61 |
| Race, African American | 1632 (72.4) | 1674 (70.9) | .39 |
| Maternal medical problems | |||
| Diabetes | 69 (3) | 53 (2.2) | .07 |
| Asthma | 400 (17.6) | 418 (17.3) | .76 |
| Hypertension | 94 (4.2) | 1014 (4.3) | .78 |
| Heart disease | 35 (1.6) | 28 (1.2) | .26 |
| Prior cesarean delivery | 322 (14.1) | 374 (15.4) | .2 |
| Prior blood transfusion | 52 (2.3) | 55 (2.3) | .98 |
| Obesity (BMI >30 kg/m 2 ) | 53 (2.3) | 74 (3.1) | .12 |
| Tobacco smoker | 143 (6.3) | 138 (5.7) | .41 |
a P values were determined by a χ 2 test (categorical data) and Wilcoxon rank sum (nonnormally distributed continuous data).
There were no maternal mortalities in the cohort, and the overall risk of maternal morbidity was 27.1 per 1000 deliveries. Despite the increased presence of MFM subspecialists on L&D in POST, there was not a significant difference in the primary outcome ( Table 3 ). The maternal morbidity rate was 30 per 1000 deliveries in PRE compared with 24 per 1000 deliveries in POST ( P = .21), which corresponds to a relative risk of 0.8 (95% confidence interval, 0.57–1.13).
| Morbidity | PRE (n = 2286) | POST (n = 2429) | P value a |
|---|---|---|---|
| Overall composite | 69 (3) | 59 (2.4) | .21 |
| Diagnosis-based codes | |||
| Renal failure | 5 (0.2) | 2 (0.1) | .22 |
| Liver failure | 0 | 0 | NA |
| Respiratory failure (includes mechanical ventilation codes) | 1 (0.04) | 3 (0.1) | .35 |
| Obstetric shock | 1 (0.04) | 0 | .3 |
| Cerebrovascular accident | 0 | 1 (0.04) | .33 |
| Embolism | 1 (0.04) | 1 (0.04) | .97 |
| Eclampsia | 3 (0.13) | 2 (0.1) | .61 |
| Septicemia | 0 | 0 | NA |
| Complications of anesthesia | 2 (0.1) | 0 | .15 |
| Procedure-based codes | |||
| Cardiac events/procedures | 6 (0.3) | 5 (0.2) | .69 |
| Mechanical ventilation | 1 (0.04) | 1 (0.04) | .97 |
| Transfusion | 56 (2.5) | 52 (2.1) | .48 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
