Racial Disparity in Severe Maternal Morbidity Associated with Hypertensive Disorders in Washington State: A Retrospective Cohort Study.

OBJECTIVES: To evaluate the relationship between hypertensive (HTN) disorders and severe maternal morbidity (SMM). To understand whether there is differential prevalence of HTN disorders by race and whether the relationship between HTN disorders and SMM is modified by race and ethnicity. METHODS: We performed a retrospective cohort study using patient-level rates of SMM for pregnancies at all 61 non-military hospitals in Washington State from 10/2015 to 9/2016. Data were obtained from the Washington State Comprehensive Hospital Abstract Reporting System. Adjusted odds ratios (OR) and 95% confidence intervals (CI) were calculated to evaluate the association of HTN disorders and SMM (with and without transfusion) overall and by race. The population-attributable fraction of HTN disorders on SMM within each racial/ethnic group was calculated. RESULTS: Of 76,965 deliveries, 864 (1.1%) had any SMM diagnosis or procedure. All racial and ethnic minorities, except white and Asian, were disproportionally affected by preeclampsia with severe features (SF) and SMM. Overall, and within each racial/ethnic group, the SMM rate was higher among pregnancies with any HTN disorder compared to no HTN disorder (2.8 vs. 0.9%, OR 3.1, 95% CI 2.7-3.6). Race and ethnicity significantly modified the association. Overall and within each racial/ethnic group, there was a dose-response relationship between the type of HTN disorder and SMM, with more severe HTN disorders leading to a greater risk of SMM. The population-attributable fraction of HTN disorders on SMM was 20.6% for Black individuals versus 17.5% overall. The findings were similar when reclassifying transfusion-only SMM as no SMM. CONCLUSIONS: In Washington, HTN disorders are associated with SMM in a dose-dependent fashion with the greatest impact among Black individuals.

Severe Maternal Morbidity Associated with Hospital NICU Level in Washington State.

OBJECTIVE: Rising maternal mortality and severe maternal morbidity (SMM) rates have drawn increasing public health attention. We evaluated patterns of SMM across the Washington State Perinatal Regional Network, in which neonatal intensive care unit (NICU) levels correlate with maternal level of care. STUDY DESIGN: Retrospective cohort study using de-identified patient and hospital-level rates of SMM diagnoses and procedures for all women who delivered at 58 hospitals from October 2015 to September 2016. Data were obtained from the Washington State Comprehensive Hospital Abstract Reporting System, which includes inpatient diagnosis with associated Present on Admission flags, procedure, and discharge information derived from hospital billing systems. Deliveries were stratified by having or not having SMM. For each SMM diagnosis, POA rates were tabulated. Hospital SMM rates (all SMM, transfusion only, and SMM excluding transfusion) were grouped according to their NICU level of care (critical access [CA] and 1-4). Odds ratios and 95% confidence intervals (CI) were calculated. RESULTS: Of 76,961 deliveries, 908 women (1.2%) had any SMM including 533 with transfusion only and 375 with all other SMM diagnoses/procedures. Rates of SMM were highest at level 1 and level 4 hospitals at 1.3 and 1.5%, respectively. Level 1 and CA hospitals had the highest transfusion rate (1.0%), while level 2, 3, and 4 hospitals had progressively lower rates (0.8, 0.7, and 0.5%, respectively; p < 0.01). Level 4 hospitals had the highest rate of SMM diagnoses/procedures (1.0%). Among SMM diagnoses, the percentage with POA was lowest in level 1/CA hospitals (23%) and similar across level 2, 3, and 4 hospitals (39%). CONCLUSION: SMM diagnoses occur most frequently at the centers providing the highest level of care, likely attributable to the regional referral system. However, transfusion rates are increased in level 1/CA hospitals. Efforts to decrease SMM should focus on equipping level 1/CA hospitals with tools to decrease maternal morbidity and improve referral systems. KEY POINTS: · SMM occurs most frequently at highest level of care.. · Higher transfusion rates occur at lower care level hospitals.. · Most SMM POA occurs at higher level of care..

Contribution of hypertension to severe maternal morbidity.

BACKGROUND: Maternal mortality and severe maternal morbidity are growing public health concerns in the United States. The Centers for Disease Control and Prevention Severe Maternal Morbidity measure provides insight into processes underlying maternal mortality and may highlight modifiable risk factors for adverse maternal health outcomes. OBJECTIVE: The primary objective of this study was to evaluate the association between hypertensive disorders and severe maternal morbidity at a regional perinatal referral center. We hypothesized that women with preeclampsia with severe features would have a higher rate of severe maternal morbidity compared to normotensive women. We also assessed the proportion of severe maternal morbidity diagnoses that were present on admission, in contrast to those arising during the delivery hospitalization. STUDY DESIGN: In this retrospective cross-sectional analysis, we assessed rates of severe maternal morbidity diagnoses (eg, renal insufficiency, shock, and sepsis) and procedures (eg, transfusion and hysterectomy) for all 7025 women who delivered at the University of Washington Medical Center from Oct. 1, 2013, through May 31, 2017. Severe maternal morbidity was determined from prespecified International Classification of Diseases diagnosis and procedure codes; all diagnoses were confirmed by chart review. Present-on-admission rates were calculated for each diagnosis through hospital administrative data provided by the Vizient University Health System Consortium. Maternal demographic and clinical characteristics were compared for women with and without severe maternal morbidity. The χ2 and Fisher exact tests were used to determine statistical significance. Odds ratios and 95% confidence intervals were calculated for the associations between maternal demographic and clinical characteristics and severe maternal morbidity. RESULTS: Of 7025 deliveries, 284 (4%) had severe maternal morbidity; 154 had transfusion only, 27 had other procedures, and 103 women had 149 severe maternal morbidity diagnoses (26 women had multiple diagnoses). Severe preeclampsia occurred in 438 deliveries (6.2%). Notably, hypertension was associated with severe maternal morbidity in a dose-dependent fashion, with the strongest association observed for preeclampsia with severe features (odds ratio, 5.4; 95% confidence interval, 3.9-7.3). Severe maternal morbidity was also significantly associated with preeclampsia without severe features, chronic hypertension, preterm delivery, pregestational diabetes, and multiple gestation. Among women with severe maternal morbidity, over one third of preterm births were associated with maternal hypertension. American Indian/Alaskan Native women had significantly higher severe maternal morbidity rates compared to other racial/ethnic groups (11.7% vs 3.9% for Whites, P < .01). Overall, 39.6% of severe maternal morbidity diagnoses were present on admission. CONCLUSION: Hypertensive disorders in pregnancy are strongly associated with severe maternal morbidity in a dose-dependent relationship, suggesting that strategies to address rising maternal morbidity rates should include early recognition and management of hypertension. Prevention strategies focused on hypertension might also impact medically indicated preterm deliveries. The finding of increased severe maternal morbidity among American Indian/Alaskan Native women, a disadvantaged population in Washington State, underscores the role that socioeconomic factors may play in adverse maternal health outcomes. As 39% of severe maternal morbidity diagnoses were present on admission, this measure should be risk-adjusted if used as a quality metric for comparison between hospitals.

A Comparison of the Nulliparous-Term-Singleton-Vertex and Society of Maternal-Fetal Medicine Cesarean Birth Metrics Based on Hospital Size.

OBJECTIVE: The purpose of this study was to compare the nulliparous-term-singleton-vertex (NTSV) and the Society of Maternal-Fetal Medicine (SMFM) cesarean birth metrics as tools for quality improvement efforts based on hospital size. MATERIALS AND METHODS: Cesarean birth rates from 275 hospitals from six states were used to evaluate the NTSV metric and 81 hospitals from four states for the SMFM metric. Data were assessed based on delivery volume, their use as an effective tool for ongoing quality improvement programs, and their ability to serve as performance-based payline indicators. RESULTS: The average NTSV and SMFM cesarean birth rates were 25.6 and 13.0%, respectively. The number of deliveries included in the NTSV metric was stable across all hospital sizes (33.1-36.2%). With the SMFM metric, there was a progressive decline in the number of deliveries included, 90.0 versus 69.6%, in relatively small to large facilities. Variability was less and precision increased with the SMFM metric, which reduced the number of hospitals that could be incorrectly categorized when using performance-based predefined cesarean birth rate paylines. CONCLUSION: The SMFM metric appears to be better suited as a tool for rapid process improvement programs aimed at reducing cesarean birth rates in low-risk patients.

Effect of severity of illness on cesarean delivery rates in Washington State.

BACKGROUND: Hospitals and providers are increasingly held accountable for their cesarean delivery rates. In the perinatal quality improvement arena, there is vigorous debate about whether all hospitals can be held to the same benchmark for an acceptable cesarean rate regardless of patient acuity. However, the causes of variation in hospital cesarean delivery rates are not well understood. OBJECTIVE: We sought to evaluate the association and temporal trends between severity of illness at admission and the primary term singleton vertex cesarean delivery rate among hospitals in Washington State. We hypothesized that hospitals with higher patient acuity would have higher cesarean delivery rates and that this pattern would persist over time. STUDY DESIGN: In this cross-sectional analysis, we analyzed aggregate hospital-level data for all nonmilitary hospitals in Washington State with ≥100 deliveries/y during federal fiscal years 2010 through 2014 (287,031 deliveries). Data were obtained from the Washington State Comprehensive Hospital Abstract Reporting System, which includes inpatient demographic, diagnosis, procedure, and discharge information derived from hospital billing systems. Age, admission diagnoses and procedure codes were converted to patient-level admission severity-of-illness scores using the All Patient Refined Diagnosis Related Groups classification system. This system is widely used throughout the United States to adjust hospital data for severity of illness. Mean admission hospital-level severity-of-illness scores were calculated for each fiscal year among the term singleton vertex population with no history of cesarean delivery. We used linear regression to evaluate the association between hospital admission severity of illness and the primary term singleton vertex cesarean delivery rate, calculated Pearson correlation coefficients, and compared regression line slopes and 95% confidence intervals for each fiscal year. RESULTS: Hospitals were diverse with respect to delivery volume, level of care, and geographic location within Washington. Hospital aggregate admission severity-of-illness score correlated with primary term singleton vertex cesarean delivery rate in all fiscal years (R2 0.38-0.58, P < .001). For every year in the study interval, as admission severity of illness increased so did the primary term singleton vertex cesarean rate. The slope of the regression line decreased during the study interval, suggesting that statewide decrease in primary term singleton vertex cesarean rate occurred across the range of severity of illness. CONCLUSION: Admission severity-of-illness score is strongly associated with the primary term singleton vertex cesarean delivery rate among hospitals in Washington State. Approximately 50% of variation in hospital primary term singleton vertex cesarean delivery rates appeared to be related to admission severity of illness. This relationship persisted over time despite a statewide decrease in cesarean delivery, suggesting that patient acuity will likely continue to contribute to hospital variation in cesarean delivery rates despite perinatal quality improvement efforts. The major implication of this study is that patient acuity should be considered when determining optimal cesarean delivery rates. High-acuity hospitals are likely to have high cesarean rates because they provide a specific role in serving regional needs. To hold these centers to an arbitrary benchmark may jeopardize the funding necessary to support regional safety net institutions.

Accuracy of the Adverse Outcome Index: An Obstetrical Quality Measure.

BACKGROUND: In obstetrics, a nationally accepted set of quality indicators for patient safety was not available in the United States until the development of a set of 10 adverse outcome measures-the Adverse Outcome Index (AOI). The National Perinatal Information Center (NPIC) developed hospital discharge data-based algorithms combined with a small set of supplemental patient data for calculation of the AOI. A study was conducted to determine the specificity, sensitivity, positive predictive value (PPV), and negative predictive value (NPV) of the AOI by using the National Perinatal Information Center (NPIC) algorithm. METHODS: A retrospective chart review of 4,252 obstetrical and neonatal charts from 2003 through 2007 was performed. NPIC definitions were compared with the "gold standard"-chart review. RESULTS: A total of 229 deliveries among the 4,000 randomly selected charts had at least one adverse outcome, reflecting an AOI of 5.7%. For detection of the 10 adverse outcomes within the AOI, the overall sensitivity of the AOI was 81.7%, specificity was 98.2%, PPV was 86.3%, and NPV was 97.4%. The Kappa value for agreement between the coded charts and the chart review was 0.82 (standard deviation=0.01, 95% confidence interval [CI]=0.80-0.85), which is considered very good. DISCUSSION: The AOI is highly reliant on accurate coding and provider documentation and requires validation with manual chart review. Concurrent chart review improves the accuracy of the AOI. Caution is advised when using the AOI as an exclusive measure of assessing obstetric quality because it may be heavily influenced by a single outcome measure; perineal laceration rates represented twice the frequency of all other outcomes combined. The AOI should be modified to better measure preventable adverse events and include a means of accounting for preexisting conditions.

In search of perinatal quality outcome measures: 1 hospital's in-depth analysis of the Adverse Outcomes Index.

OBJECTIVE: The purpose of this study was to assess the Adverse Outcome Index perinatal quality indicator system that was derived from administrative data. STUDY DESIGN: Adverse events were identified for 10 component measures; the Adverse Outcome Index was calculated by the National Perinatal Information Center from 42 months of administrative data. After retrospective chart review, we estimated positive predictive value for 10 measures that were obtained by corrected calculations of Adverse Outcome Index. RESULTS: Positive predictive values were 86-100% in 7 indicators, with lower values in 3 indicators: neonatal death, 0/2 fetuses; inborn birth trauma, 22/33 infants (67%); and maternal return to the operating room, 16/33 women (48.5%). In term admission to the neonatal intensive care unit, 107 false negatives were identified, with a negative predictive value of 45%. CONCLUSION: Indicator positive predictive value was variable. Performance can be strengthened by methods to identify both false-positive and false-negative adverse events that would include chart review and some measure specification revisions to improve alignment with original indicator intent. Interhospital comparison application requires further study.