Objective
We sought to explore how racial/ethnic minority-serving hospitals perform on 15 delivery-related indicators, and examine whether indicators vary by race/ethnicity within the same type of hospitals.
Study Design
We used 2008 through 2011 linked State Inpatient Database and American Hospital Association data from 7 states, and designated hospitals with >50% of deliveries to non-Hispanic white, non-Hispanic black, and Hispanic women as white-, black-, and Hispanic-serving, respectively. We calculated indicator rates per 1000 deliveries by hospital type and, separately, for non-Hispanic white, non-Hispanic black, and Hispanic women within each hospital type. We fitted multivariate Poisson regression models to examine associations between delivery-related indicators and patient and hospital characteristics by hospital type.
Results
White-serving hospitals offer obstetric care to an older and wealthier population than black- or Hispanic-serving hospitals. Rates of the most prevalent indicators examined (complicated vaginal delivery, complicated cesarean delivery, obstetric trauma) were lowest in Hispanic-serving hospitals. Generally, indicator rates were similar in Hispanic- and white-serving hospitals. Black-serving hospitals performed worse than other hospitals on 12 of 15 indicators. Indicator rates varied greatly by race/ethnicity in white- and Hispanic-serving hospitals, with non-Hispanic blacks having 1.19-3.27 and 1.15-2.68 times higher rates than non-Hispanic whites, respectively, for 11 of 15 indicators. Conversely, there were few indicator rate differences by race/ethnicity in black-serving hospitals, suggesting an overall lower performance of these hospitals compared to white- and Hispanic-serving hospitals.
Conclusion
We found considerable differences in delivery-related indicators by hospital type and patients’ race/ethnicity. Obstetric care quality measures are needed to track racial/ethnic disparities at the facility and population levels.
Significant racial and ethnic disparities exist in maternal health in the United States. Compared with non-Hispanic white women, higher proportions of racial/ethnic minority women experience severe pregnancy and delivery complications and death during pregnancy or in the postpartum period. Also, health insurance coverage and utilization of health care are significantly different for racial/ethnic minority women. For example, compared with non-Hispanic white women, non-Hispanic black and Hispanic women are more likely to be uninsured, begin prenatal care in the third trimester of pregnancy, or obtain no prenatal care. The association between being a racial/ethnic minority woman and being uninsured or having publicly funded health insurance may affect how she is treated in medical settings. For instance, evidence suggests that non-Hispanic black women receive substandard prenatal care compared with non-Hispanic white women.
Recent research has focused on the site of care as a potential explanation for racial/ethnic disparities in health. Ly et al used national Medicare data to evaluate the performance of hospitals on 11 medical and surgical patient safety indicators. They found that primarily black-serving hospitals performed worse than other hospitals on 6 of the 11 indicators, and that, by and large, both white and black patients had higher rates of potential safety events in primarily black-serving hospitals than in non-black-serving hospitals. Little is known about potential disparities in the quality of obstetric care offered in US hospitals with different patient population distributions by race/ethnicity. However, with the country becoming more racially and ethnically diverse, examining and understanding these disparities has become a high priority for stakeholders in obstetric care. This analysis aims to explore how racial/ethnic minority-serving hospitals perform on delivery-related indicators with a quality-of-care component, and to examine whether these indicators vary by race/ethnicity within the same type of hospital.
Materials and Methods
We used pooled 2008 through 2011 data from Healthcare Cost and Utilization Project’s State Inpatient Database (SID) in the 7 US states with >80,000 annual live births (ie, the mean number of 2008 through 2011 births in the 50 US states and District of Columbia) that report race and ethnicity data: Arizona, California, Florida, Michigan, New Jersey, New York, and North Carolina. Of note, 2008 and 2009 data from North Carolina did not include race/ethnicity information, and 2011 data from Arizona, Michigan, New York, and North Carolina were not available at the time of analysis; thus, these state- and year-specific data were not included in the analysis. The SID contains all inpatient discharges in participating states, translated into a uniform set of variables to facilitate multistate comparisons. To obtain data on hospital characteristics, we linked SID and American Hospital Association (AHA) data from 6 of the 7 states (the AHA hospital identifier was missing in the Michigan SID data); the 6-state linkage rate was 99.3%. AHA is a national source of proprietary hospital and health system data collected annually by the AHA. Because this analysis involved the use of publicly available data lacking direct personal identifiers, the Centers for Disease Control and Prevention Institutional Review Board determined it to be research not involving human subjects.
We identified all delivery hospitalization records using the algorithm developed by Kuklina et al. To identify delivery hospitalization records, this algorithm uses International Classification of Diseases, Ninth Revision, Clinical Modification ( ICD-9-CM ) diagnosis codes for delivery outcomes, procedure codes for selected delivery-related procedures, and diagnosis-related group delivery codes. To ensure consistency among states, records were restricted to the first 15 diagnoses and 15 procedures listed. We excluded data from hospitals with <30 deliveries in a given year and where >50% of delivery records lacked data on race/ethnicity, and all records with a missing or invalid value for race/ethnicity. After these exclusions, the sample consisted of 4,456,426 delivery hospitalizations representing 88.9-95.3% of all state- and year-specific births in Arizona, California, Florida, New Jersey, New York, and North Carolina, and 72.3-72.9% of year-specific births in Michigan (Appendix; Supplementary Table 1 ). We ranked all hospitals by their proportion of deliveries to non-Hispanic white, non-Hispanic black, and Hispanic women. Hospitals with a majority (>50%) of deliveries to non-Hispanic white women were designated as white-serving hospitals and, similarly, those with a majority of deliveries to non-Hispanic black or Hispanic women were designated as black- or Hispanic-serving, respectively. We identified 1021 white-serving, 56 black-serving, and 530 Hispanic-serving hospitals in the 7 states with a median annual number of 2291, 2922, and 2749 deliveries, respectively ( Table 1 ). Deliveries in other hospitals (ie, not primarily serving any single racial/ethnic group) represented 32.9% of all deliveries in the sample and were excluded from the analysis.
| Characteristic | White-serving hospitals, n = 1021 | Black-serving hospitals, n = 56 | Hispanic-serving hospitals, n = 530 |
|---|---|---|---|
| Total no. of deliveries | 1,714,765 | 119,146 | 1,154,629 |
| Mean (SD) annual deliveries | 2540.25 (1652.38) | 2834.99 (1350.83) | 3062.47 (1585.25) |
| Median (range) annual deliveries | 2291 (32–7548) | 2922 (381–5489) | 2749 (57–7431) |
| Patient characteristics | |||
| Race, % | |||
| White | 69.18 | 12.81 | 16.63 |
| Black | 9.87 | 66.41 | 6.21 |
| Hispanic | 12.15 | 12.48 | 70.39 |
| Other | 8.80 | 8.30 | 6.77 |
| Age, % | |||
| <20 | 6.97 | 13.78 | 11.45 |
| 20-34 | 74.65 | 72.88 | 74.62 |
| ≥35 y | 18.38 | 13.34 | 13.92 |
| State-level household income quartile for patient ZIP code, a % | |||
| First (poorest) | 21.87 | 53.38 | 40.19 |
| Second | 24.51 | 17.39 | 28.58 |
| Third | 26.07 | 13.42 | 19.62 |
| Fourth (richest) | 25.91 | 6.61 | 9.02 |
| Insurance coverage, % | |||
| Medicaid | 35.05 | 64.06 | 65.68 |
| Private insurance | 59.92 | 23.97 | 28.51 |
| Self-pay | 1.90 | 8.60 | 3.66 |
| Other | 3.13 | 3.37 | 2.16 |
| Medical history, % | |||
| Preexisting diabetes | 0.75 | 1.26 | 0.99 |
| Chronic hypertensive disease | 1.62 | 4.13 | 1.39 |
| Chronic heart disease | 0.49 | 0.37 | 0.18 |
| Chronic respiratory disease | 3.40 | 5.88 | 2.16 |
| Chronic renal disease | 0.23 | 0.26 | 0.20 |
| Chronic liver disease | 0.02 | 0.03 | 0.03 |
| HIV/AIDS | 0.07 | 0.80 | 0.05 |
| Any chronic condition b | 6.10 | 11.50 | 4.62 |
| Mode of delivery, % | |||
| Vaginal | 65.99 | 65.38 | 65.45 |
| Primary cesarean | 19.16 | 20.14 | 17.62 |
| Repeat cesarean | 14.85 | 14.48 | 16.93 |
| Hospitalization characteristics | |||
| Mean (SD) hospital length of stay (days) | 2.65 (2.15) | 2.95 (2.71) | 2.57 (1.97) |
| Median (range) hospital length of stay (days) | 2 (0–310) | 3 (0–159) | 2 (0–217) |
| Mean (SD) time to first procedure (days) | 0.41 (1.82) | 0.56 (2.25) | 0.45 (1.68) |
| Mean (SD) no. of diagnoses | 4.66 (2.46) | 5.19 (2.90) | 4.28 (2.38) |
| Mean (SD) no. of procedures | 2.26 (1.24) | 2.32 (1.35) | 2.21 (1.16) |
| Routine discharge, % | 97.75 | 98.83 | 99.23 |
| Hospital characteristics c | |||
| Hospital location | |||
| Urban | 96.63 | 100.0 | 97.52 |
| Rural | 3.37 | 0.00 | 2.48 |
| Hospital teaching status, % | |||
| Teaching | 48.52 | 78.80 | 44.34 |
| Nonteaching | 51.48 | 21.20 | 55.66 |
| Multihospital system membership, % | |||
| Yes | 33.56 | 22.87 | 36.36 |
| No | 66.44 | 77.13 | 63.64 |
| Hospital ownership, % | |||
| Public | 8.2 | 44.3 | 22.8 |
| Private not-for-profit | 79.8 | 43.4 | 59.5 |
| Private for profit | 12.0 | 12.3 | 17.6 |
a Information missing for 1.64%, 9.21%, and 2.58% of deliveries in white-, black-, and Hispanic-serving hospitals, respectively
b At least 1 of the conditions considered: preexisting diabetes, chronic hypertensive disease, chronic heart disease, chronic respiratory disease, chronic renal disease, chronic liver disease, HIV/AIDS
c Information not available for hospitals in Michigan; P values for all hospital comparisons shown are statistically significant at P < .005.
We examined patient, hospitalization, and hospital characteristics by type of hospital (white-, black-, and Hispanic-serving). The patient characteristics of interest were race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, and other); age (<20, 20-34, and ≥35 years); state-level household income quartile for patient ZIP code; insurance coverage for delivery (Medicaid, private, self-pay, or other); and presence of chronic medical conditions including preexisting diabetes, chronic hypertensive disease, chronic heart disease, chronic respiratory disease, chronic renal disease, chronic liver disease, and human immunodeficiency virus/acquired immune deficiency syndrome (corresponding ICD-9-CM codes in Supplementary Table 2 ). The hospitalization-related characteristics examined were length of stay (days), time to the first hospital procedure (days), number of diagnoses and procedures during the delivery hospitalization, and proportion of routine hospital discharges. Hospital characteristics of interest were location (urban/rural), teaching status (yes/no), multiple-hospital system membership (yes/no), and ownership (public, private for-profit, private not-for-profit).
Few indicators of obstetric care quality and patient safety exist in the published literature. Following review of the available published and unpublished literature on proposed obstetric care indicators amenable to assessment using administrative data, we arrived at 15 delivery-related indicators with a quality-of-care component: complicated vaginal delivery, complicated cesarean delivery, obstetric trauma, obstetric wound complications following vaginal delivery, obstetric wound complications following cesarean delivery, uterine rupture among patients with a trial of labor, obstetric thrombosis or embolism, a larger group of postpartum vascular complications including obstetric thrombosis or embolism, peripheral vascular complications and nonthrombotic pulmonary emboli, puerperal infection excluding antepartum infection of amniotic cavity, postpartum urinary tract infection, other obstetric complications (ie, pulmonary, cardiac, central nervous system, anesthesia, shock, major complications of obstetric procedures, acute postpartum renal failure), peripartum hysterectomy among patients with postpartum hemorrhage, puerperal cerebrovascular disorders among patients with hypertensive disorders of pregnancy and without coagulopathy, blood transfusion, and in-hospital mortality (corresponding ICD-9-CM /diagnosis-related group codes in Supplementary Table 2 ). We calculated delivery-related indicator rates per 1000 deliveries for the 15 indicators by type of hospital and, separately, for non-Hispanic white, non-Hispanic black, and Hispanic women within each hospital type. To estimate relative risks for non-Hispanic black and Hispanic women compared with non-Hispanic white women in each type of hospital, we fitted separate Poisson regression models for the 15 indicators. Next, we selected the 4 most relevant indicators at the population level (ie, the 2 most prevalent indicators [complicated vaginal delivery and complicated cesarean delivery], the most common indicator of severe maternal morbidity for all racial and ethnic groups in the United States [blood transfusion ], and in-hospital mortality), and further assessed their associations with patient and hospital characteristics by fitting 3 additional sets of Poisson regression models stratified by type of hospital: model I was adjusted for race/ethnicity, annual hospital delivery volume, state, and year; model II was further adjusted for age, insurance coverage for delivery, presence of a chronic condition, and state-level household income quartile for patient ZIP code; and model III was even further adjusted for hospital location, teaching status, multiple-hospital system membership, and ownership. All regression models in this analysis were adjusted for clustering at the hospital level with generalized estimating equations, and analyses were performed using software (Stata, version 12; StataCorp, College Station, TX).
Given the lower percentage of state births with race/ethnicity data captured in our analytic data set for Michigan relative to the other 6 states and because AHA data were not available for Michigan and, by implication, model III Poisson regressions could not be fitted on the Michigan data, we conducted sensitivity analyses to assess the potential influence of including the Michigan data in our analyses. Results from Poisson regression models I and II fitted before and after exclusion of Michigan data did not differ considerably in magnitude and statistical significance. Therefore, we present delivery-related indicator rates based on data from all 7 states, and models I-III regression results after excluding data from Michigan.
Results
More than two thirds of deliveries in all 3 types of hospitals were to the predominantly served racial/ethnic group ( Table 1 ). Similar proportions of women aged 20-34 years delivered in all types of hospitals, but fewer women aged <20 years did so in white-serving (6.97%) than in black-serving (13.78%) or Hispanic-serving (11.45%) hospitals. Slightly more than a third (35.05%) of women delivering in white-serving hospitals received Medicaid funding for delivery compared to 64.01% and 65.68% of those delivering in black- and Hispanic-serving hospitals, respectively. Overall, 6.10%, 11.50%, and 4.62% of women giving birth in white-, black-, and Hispanic-serving hospitals, respectively, had at least 1 of the chronic conditions considered. About two-thirds of deliveries in all 3 types of hospitals were vaginal; more deliveries to women in Hispanic-serving than in either white- or black-serving hospitals were repeat cesareans. The median length of delivery hospitalization was 3 days for women in black-serving hospitals and 2 days for those in white- and Hispanic-serving hospitals. The numbers of diagnoses and procedures during delivery hospitalizations were slightly higher for women in black-serving than in either white- or Hispanic-serving hospitals.
In white-serving hospitals ( Table 2 ), delivery-related indicator rates varied greatly, and 3 indicators had rates >10 per 1000 deliveries: 37.64 women with obstetric trauma per 1000 deliveries, 74.19 complicated vaginal deliveries per 1000 vaginal deliveries, and 110.73 complicated cesarean deliveries per 1000 cesarean deliveries. Compared to corresponding rates among non-Hispanic white women, rates for 11 of 15 indicators were significantly higher among non-Hispanic black women, and rates for 6 of 15 indicators were significantly higher among Hispanic women. Yet, rates of obstetric trauma were significantly lower in both black and Hispanic women than in white women delivering in white-serving hospitals, as were rates of obstetric wound complications among Hispanic women.
| Indicator | Hospital rate (95% CI) | Whites | Blacks | Hispanics | |
|---|---|---|---|---|---|
| Rate (95% CI) | RR | RR (95% CI) | RR (95% CI) | ||
| White-serving hospitals | |||||
| Complicated vaginal delivery | 74.19 (74.11–74.27) | 71.01 (70.96–71.06) | 1.00 | 1.36 (1.34–1.38) e | 1.05 (1.03–1.07) e |
| Complicated cesarean delivery | 110.73 (108.96–111.14) | 107.34 (106.75–107.95) | 1.00 | 1.37 (1.35–1.39) e | 0.94 (0.92–0.95) e |
| Obstetric trauma | 37.64 (37.34–37.94) | 38.23 (37.85–38.57) | 1.00 | 0.68 (0.66–0.70) e | 0.87 (0.85–0.89) e |
| Blood transfusion | 8.80 (8.64–8.96) | 7.44 (7.26–7.62) | 1.00 | 2.14 (2.04–2.25) e | 1.36 (1.29–1.43) e |
| Other obstetric complications a | 5.81 (5.69–5.93) | 5.68 (5.53–5.83) | 1.00 | 1.19 (1.12–1.26) e | 0.98 (0.92–1.04) |
| Obstetric wound complications postvaginal delivery | 4.51 (4.41–4.62) | 4.48 (4.36–4.60) | 1.00 | 0.95 (0.88–1.03) | 0.92 (0.86–0.99) e |
| Obstetric wound complications postcesarean delivery | 3.40 (3.32–3.48) | 3.40 (3.30–3.51) | 1.00 | 1.04 (0.95–1.13) | 0.84 (0.77–0.92) e |
| Puerperal infection b | 1.17 (1.12–1.22) | 0.83 (0.78–0.88) | 1.00 | 2.85 (2.53–3.20) e | 1.94 (1.72–2.20) e |
| Postpartum urinary tract infection | 1.15 (1.10–1.25) | 0.91 (0.86–0.96) | 1.00 | 2.62 (2.34–2.94) e | 1.57 (1.38–1.79) e |
| Peripartum hysterectomy if patients had PPH | 0.65 (0.61–0.69) | 0.55 (0.50–0.59) | 1.00 | 1.70 (1.43–2.02) e | 1.33 (1.11–1.58) e |
| Postpartum vascular complications c | 0.29 (0.27–0.31) | 0.26 (0.23–0.30) | 1.00 | 2.21 (1.76–2.79) e | 1.14 (0.89–1.50) |
| Obstetric thrombosis or embolism | 0.23 (0.20–0.25) | 0.22 (0.20–0.24) | 1.00 | 2.08 (1.61–2.68) e | 1.11 (0.82–1.50) |
| Uterine rupture if patients had trial of labor | 0.16 (0.12–0.20) | 0.15 (0.13–0.17) | 1.00 | 1.22 (0.83–1.79) | 0.96 (0.65–1.41) |
| Puerperal cerebrovascular disorders if patients had HDP d | 0.10 (0.08–0.12) | 0.08 (0.06–0.10) | 1.00 | 3.27 (2.27–4.72) e | 1.96 (1.30–2.95) e |
| In-hospital mortality | 0.05 (0.03–0.07) | 0.04 (0.02–0.06) | 1.00 | 3.05 (1.80–5.15) e | 0.99 (0.47–2.10) |
| Black-serving hospitals | |||||
| Complicated vaginal delivery | 106.89 (105.29–108.49) | 98.66 (93.89–103.49) | 1.00 | 1.19 (1.13–1.26) e | 0.75 (0.69–0.81) e |
| Complicated cesarean delivery | 153.69 (151.09–155.50) | 144.13 (138.80–149.73) | 1.00 | 1.15 (1.10–1.21) e | 0.76 (0.72–0.82) e |
| Obstetric trauma | 32.43 (31.41–33.65) | 34.92 (31.90–37.96) | 1.00 | 0.88 (0.80–0.97) e | 0.87 (0.76–0.98) e |
| Blood transfusion | 24.09 (23.11–25.07) | 15.36 (13.12–17.52) | 1.00 | 1.80 (1.56–2.09) e | 1.16 (0.95–1.41) |
| Other obstetric complications a | 8.54 (8.02–9.06) | 8.12 (6.84–9.52) | 1.00 | 1.12 (0.93–1.36) | 0.70 (0.53–0.92) e |
| Puerperal infection b | 5.58 (5.16–6.00) | 2.10 (1.62–2.60) | 1.00 | 3.17 (2.22–4.53) e | 0.89 (0.54–1.49) |
| Obstetric wound complications postvaginal delivery | 3.63 (3.29–3.97) | 3.67 (2.68–4.58) | 1.00 | 1.11 (0.75–1.33) | 0.72 (0.48–1.08) |
| Obstetric wound complications postcesarean delivery | 3.00 (2.70–3.30) | 3.21 (2.32–4.11) | 1.00 | 0.91 (0.67–1.24) | 0.73 (0.48–1.13) |
| Postpartum urinary tract infection | 2.35 (2.02–2.64) | 1.11 (0.62–1.60) | 1.00 | 2.52 (1.54–4.13) e | 1.33 (0.71–2.50) |
| Peripartum hysterectomy if patients had PPH | 1.15 (0.96–1.26) | 1.18 (0.66–1.72) | 1.00 | 0.99 (0.59–1.63) | 0.91 (0.47–1.79) |
| Postpartum vascular complications c | 0.69 (0.55–0.83) | 0.47 (0.15–0.79) | 1.00 | 2.09 (0.91–4.82) | 0.86 (0.26–2.81) |
| Obstetric thrombosis or embolism | 0.64 (0.50–0.78) | 0.39 (0.04–0.71) | 1.00 | 1.90 (0.82–4.39) | 0.86 (0.26–2.80) |
| Uterine rupture if patients had trial of labor | 0.22 (0.21–0.24) | 0.26 (0.01–0.51) | 1.00 | 0.96 (0.33–2.82) | N/A f |
| Puerperal cerebrovascular disorders if patients had HDP d | 0.22 (0.14–0.30) | 3.28 (2.86–3.68) | 1.00 | 0.73 (0.27–1.96) | 0.21 (0.02–1.76) |
| In-hospital mortality | 0.17 (0.10–0.24) | 0.33 (0.04–0.63) | 1.00 | 0.54 (0.19–1.50) | N/A f |
| Hispanic-serving hospitals | |||||
| Complicated vaginal delivery | 65.76 (65.30–66.12) | 69.49 (68.28–70.71) | 1.00 | 1.15 (1.12–1.19) e | 0.91 (0.89–0.93) e |
| Complicated cesarean delivery | 102.36 (101.80–102.96) | 113.26 (111.89–114.63) | 1.00 | 1.33 (1.30–1.36) e | 0.84 (0.82–0.85) e |
| Obstetric trauma | 31.21 (30.88–31.53) | 36.52 (35.72–37.30) | 1.00 | 0.62 (0.59–0.65) e | 0.81 (0.79–0.83) e |
| Blood transfusion | 10.27 (10.05–10.49) | 8.77 (8.39–9.15) | 1.00 | 1.99 (1.83–2.16) e | 1.14 (1.07–1.20) e |
| Other obstetric complications a | 5.11 (4.99–5.24) | 5.53 (5.23–5.83) | 1.00 | 1.19 (1.07–1.33) e | 0.87 (0.82–0.94) e |
| Obstetric wound complications postvaginal delivery | 3.93 (3.82–4.04) | 3.88 (3.58–4.18) | 1.00 | 1.05 (0.91–1.20) | 1.00 (0.93–1.09) |
| Obstetric wound complications postcesarean delivery | 2.82 (2.71–2.92) | 2.89 (2.68–3.08) | 1.00 | 1.20 (1.03–1.39) e | 0.93 (0.85–1.03) |
| Puerperal infection b | 1.39 (1.32–1.46) | 0.93 (0.80–1.06) | 1.00 | 2.66 (2.16–3.28) e | 1.50 (1.28–1.75) e |
| Postpartum urinary tract infection | 1.24 (1.18–1.30) | 0.94 (0.82–1.10) | 1.00 | 1.95 (1.56–2.44) e | 1.35 (1.15–1.58) e |
| Peripartum hysterectomy if patients had PPH | 0.76 (0.71–0.81) | 0.67 (0.55–0.78) | 1.00 | 1.51 (1.13–2.01) e | 1.09 (0.90–1.32) |
| Postpartum vascular complications c | 0.23 (0.20–0.26) | 0.24 (0.17–0.31) | 1.00 | 1.69 (1.07–2.69) e | 0.93 (0.68–1.29) |
| Obstetric thrombosis or embolism | 0.19 (0.17–0.22) | 0.21 (0.15–0.27) | 1.00 | 1.50 (0.90–2.50) | 0.85 (0.60–1.21) |
| Uterine rupture if patients had trial of labor | 0.13 (0.11–0.15) | 0.15 (0.09–0.20) | 1.00 | 0.96 (0.46–1.97) | 0.88 (0.58–1.33) |
| Puerperal cerebrovascular disorders if patients had HDP d | 0.11 (0.09–0.13) | 0.11 (0.05–0.18) | 1.00 | 2.68 (1.46–4.90) e | 0.93 (0.58–1.51) |
| In-hospital mortality | 0.09 (0.07–0.10) | 0.07 (0.03–0.11) | 1.00 | 2.49 (1.17–5.29) e | 1.16 (0.66–2.07) |
a Pulmonary, cardiac, central nervous system, anesthesia, shock, major complications of obstetric procedures, acute postpartum renal failure
b Excludes patients with diagnosis of antepartum infection of amniotic cavity
c Includes obstetric thrombosis or embolism, peripheral vascular complications, and nonthrombotic pulmonary emboli
d Excludes patients with coagulopathy
e Statistically significant at P < .05
Twelve of 15 indicator rates were highest in black-serving hospitals, with rates of puerperal infection, urinary tract infection, obstetric embolism, puerperal cerebrovascular disorders, blood transfusion, and in-hospital mortality being considerably higher than in either white- or Hispanic-serving hospitals. There were fewer significant rate differences by race/ethnicity in black-serving than in white- or Hispanic-serving hospitals; compared to non-Hispanic white women, non-Hispanic black women had significantly higher rates of puerperal infection, urinary tract infection, blood transfusion, and complicated vaginal and complicated cesarean delivery, and a significantly lower rate of obstetric trauma, while Hispanic women had significantly lower rates of complicated vaginal and complicated cesarean delivery, obstetric trauma, and other obstetric complications.
Only 3 indicator rates were significantly higher in Hispanic- than in white-serving hospitals: puerperal infection, peripartum hysterectomy, and blood transfusion. Rates of 11 indicators were significantly higher for non-Hispanic black than non-Hispanic white women delivering in Hispanic-serving hospitals. Hispanic women delivering in Hispanic-serving hospitals had significantly higher rates of puerperal infection, urinary tract infection, and blood transfusion, but lower rates of complicated vaginal and cesarean delivery, obstetric trauma, and other obstetric complications than non-Hispanic white women giving birth in the same hospitals.
After adjustment for patient and hospital characteristics and when compared with non-Hispanic white women, non-Hispanic black women had a higher risk of a complicated vaginal or cesarean delivery, while Hispanic women had a lower risk of a complicated cesarean delivery in all types of hospitals ( Table 3 ). Age <20 or ≥35 years was associated with only slightly higher risks of complicated vaginal and complicated cesarean delivery. Yet, presence of a chronic condition increased the risk of a complicated vaginal delivery by a factor of 3.78 in black-serving hospitals, 4.35 in white-serving hospitals, and 4.86 in Hispanic-serving hospitals, and the risk of a complicated cesarean delivery by a factor of 1.92 in black-serving hospitals, 2.14 in white-serving hospitals, and 2.43 in Hispanic-serving hospitals. We also found higher risks of complicated vaginal and complicated cesarean delivery in teaching than nonteaching white- and Hispanic-serving hospitals, and in white-serving hospitals that are members of multihospital systems; conversely, there appears to be a 33% lower risk of a complicated cesarean in black-serving hospitals that are part of a multihospital system. When compared with deliveries in private not-for-profit hospitals, deliveries in public Hispanic-serving hospitals were associated with higher risks of complicated vaginal and complicated cesarean delivery, while deliveries in both white- and Hispanic-serving private for-profit hospitals were associated with a lower risk of complications following both vaginal and cesarean deliveries.
| Covariates | Complicated vaginal delivery | Complicated cesarean delivery | ||||
|---|---|---|---|---|---|---|
| Model I a RR (95% CI) | Model II b RR (95% CI) | Model III c RR (95% CI) | Model I a IRR (95% CI) | Model II b RR (95% CI) | Model III c RR (95% CI) | |
| White-serving hospitals | ||||||
| Race (white = ref) | ||||||
| Black | 1.37 (1.31–1.44) d | 1.25 (1.20–1.31) d | 1.24 (1.19–1.29) d | 1.25 (1.22–1.28) d | 1.15 (1.13–1.18) d | 1.15 (1.12–1.17) d |
| Hispanic | 1.02 (0.96–1.07) | 1.03 (0.99–1.08) | 1.04 (0.99–1.08) | 0.97 (0.94–0.99) d | 0.96 (0.93–0.98) d | 0.96 (0.93–0.99) d |
| Age (20-34 y = ref) | ||||||
| <20 | 1.09 (1.05–1.12) d | 1.09 (1.05–1.12) d | 1.12 (1.09–1.15) d | 1.12 (1.09–1.15) d | ||
| ≥35 | 1.12 (1.10–1.15) d | 1.12 (1.10–1.14) d | 1.08 (1.06–1.10) d | 1.07 (1.05–1.09) d | ||
| Chronic condition (no = ref) | ||||||
| Yes | 4.39 (4.13–4.66) d | 4.35 (4.09–4.62) d | 2.13 (2.07–2.23) d | 2.14 (2.06–2.22) d | ||
| Teaching hospital (no = ref) | ||||||
| Yes | 1.15 (1.05–1.26) d | 1.11 (1.03–1.20) d | ||||
| Multihospital system membership (no = ref) | ||||||
| Yes | 1.09 (1.01–1.180) d | 1.06 (1.01–1.13) d | ||||
| Hospital ownership (private not-for-profit = ref) | ||||||
| Public | 0.93 (0.80–1.07) d | 1.1 (0.91–1.12) | ||||
| Private for-profit | 0.09 (0.07–0.10) d | 0.89 (0.81–0.98) d | ||||
| Black-serving hospitals | ||||||
| Race (white = ref) | ||||||
| Black | 1.15 (0.95–1.39) d | 1.12 (1.01–1.27) d | 1.16 (1.04–1.28) d | 1.16 (1.03–1.30) d | 1.11 (1.00–1.23) d | 1.09 (1.02–1.17) d |
| Hispanic | 0.73 (0.54–0.98) | 0.80 (0.65–1.00) | 0.82 (0.68–0.99) d | 0.77 (0.65–0.93) d | 0.79 (0.68–0.92) d | 0.78 (0.68–0.90) d |
| Age (20-34 y = ref) | ||||||
| <20 | 1.14 (1.05–1.23) d | 1.13 (1.05–1.22) d | 1.08 (1.03–1.13) d | 1.09 (1.04–1.12) d | ||
| ≥35 | 1.14 (1.01–1.29) d | 1.16 (1.04–1.29) d | 1.07 (1.01–1.13) d | 1.07 (1.02–1.12) d | ||
| Chronic condition (no = ref) | ||||||
| Yes | 3.94 (3.12–4.98) d | 3.78 (3.02–4.73) d | 1.99 (1.70–2.34) d | 1.92 (1.62–2.28) d | ||
| Teaching hospital (no = ref) | ||||||
| Yes | 1.18 (0.40–3.51) | 1.31 (0.56–3.11) | ||||
| Multihospital system membership (no = ref) | ||||||
| Yes | 0.87 (0.54–1.40) | 0.67 (0.46–0.97) d | ||||
| Hospital ownership (private not-for-profit = ref) | ||||||
| Public | 0.81 (0.52–1.25) | 0.93 (0.39–2.78) | ||||
| Private for-profit | 0.60 (0.20–1.79) | 1.04 (0.39–2.78) | ||||
| Hispanic-serving hospitals | ||||||
| Race (white = ref) | ||||||
| Black | 1.25 (1.15–1.37) d | 1.15 (1.07–1.23) d | 1.13 (1.08–1.19) d | 1.21 (1.15–1.27) d | 1.15 (1.11–1.21) d | 1.15 (1.10–1.20) d |
| Hispanic | 0.91 (0.85–0.97) d | 0.98 (0.93–1.04) | 0.99 (0.96–1.03) | 0.88 (0.85–0.92) d | 0.93 (0.89–0.96) d | 0.94 (0.91–0.97) d |
| Age (20-34 y = ref) | ||||||
| <20 | 1.19 (1.15–1.23) d | 1.19 (1.15–1.23) d | 1.20 (1.17–1.24) d | 1.21 (1.17–1.24) d | ||
| ≥35 | 1.12 (1.08–1.15) d | 1.10 (1.07–1.14) d | 1.11 (1.09–1.13) d | 1.10 (1.09–1.12) d | ||
| Chronic condition (no = ref) | ||||||
| Yes | 5.15 (4.57–5.81) d | 4.86 (4.33–5.46) d | 2.51 (2.34–2.69) d | 2.43 (2.26–2.62) d | ||
| Teaching hospital (no = ref) | ||||||
| Yes | 1.27 (1.14–1.41) d | 1.17 (1.08–1.26) d | ||||
| Multihospital system membership (no = ref) | ||||||
| Yes | 1.09 (0.96–1.24) | 1.05 (0.96–1.15) | ||||
| Hospital ownership (private not-for-profit = ref) | ||||||
| Public | 1.35 (1.20–1.54) d | 1.16 (1.04–1.30) d | ||||
| Private for-profit | 0.80 (0.66–0.96) d | 0.85 (0.76–0.96) d | ||||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
