Racial Disparities in Perioperative Outcomes in Children: Where Do We Go From Here?

Several previous studies have shown a higher risk of adverse postoperative events and mortality for African American compared to White children.1–3 In this issue of Anesthesia & Analgesia, Willer et al4 adds to the current body of literature examining disparity in postsurgical outcomes between African American and White children undergoing noncardiac surgical procedures. The postsurgical outcome of interest in this study is failure to rescue (FTR) defined as mortality among patients who required a reoperation following an index inpatient surgical procedure.4 Racial disparity in perioperative outcomes in children is real and should be of concern to the anesthesia community. While a substantial amount of research has been focused on disparities between African American and White children, it must be recognized that 30% of children in the United States are neither African American nor White and that substantial inequities exist for other racial and ethnic groups. To fully understand the depth and breadth of this issue, the reader is referred to data from the Health Opportunity and Equity (HOPE) Initiative.5 This interactive interface graphically and vividly illustrates disparities in 10 different categories of health and opportunity across 6 racial and ethnic groups for each of the 50 states. African American children are more likely to die after congenital heart surgery or neurologic procedures and are more likely to have complications related to appendicitis.6–8 Analysis of the Kids’ Inpatient Database (KID) revealed that African American children undergoing cerebrospinal fluid shunting had a 56% increased likelihood of inpatient death (odds ratio [OR] = 1.56, 95% confidence interval [CI], 1.02-2.40, P < .05) and 30% higher likelihood of nonroutine hospital discharge compared with White children (OR = 1.30, 95% CI, 1.14-1.48; P < .05).7 Some, but not all studies, suggest that these disparities may be related to a higher incidence in African American children of well-recognized risk factors associated with adverse outcomes. An analysis of the American College of Surgeons National Surgical Quality Improvement (ACS NSQIP) database using network analyses showed that the prevalence of preoperative risk factors associated with postoperative mortality was significantly higher for African American than for White children. The OR for death after surgery among African American compared with White children was 2.22 (95% CI, 1.83-2.70). In addition, many of the risk factors associated with death after surgery (such as mechanical ventilation, oxygen support, wound infection, and neonatal status) carried a higher risk when they occurred in African American children.1 Nonetheless, it has been shown that even in healthy pediatric patients without major preoperative risk factors, postoperative mortality was higher for African American compared to White children. (crude OR = 3.48; 95% CI, 1.76-6.87).9 Other possible explanations for these disparate outcomes include differences in the type of surgical procedures performed, socioeconomic factors, cultural and family factors, differential access to prehospital and hospital care, and the quality and capabilities of prehospital and hospital care.10–14 Administrative databases, with large patient cohorts, have been used to analyze these factors and to develop and validate risk models in patients undergoing cardiac and noncardiac procedures.15,16 Comparison of studies using these databases is complicated by the reality that the same data fields are not collected in all databases and the demographics of the cohorts are not identical. The authors have completed an important first step in the analysis of the difference in the probability of FTR between African American and White children. The ACS NSQIP-Pediatric database is a multi-institutional database of academic and community hospitals. It collects more than 129 variables including demographics, preoperative status, surgical procedures, and postoperative outcomes. This database has a reliable record of data accuracy secondary to regular training and audits of the database. It has been used to report the incidence of postoperative outcomes in both adult and pediatric population, as well as identifying preoperative and surgical risk factors. An advantage of the NSQIP database is the ability to use current procedural terminology (CPT) codes, as well as the reoperation CPT codes, to identify the surgical procedures that a patient has undergone. There is also a flag called “related unplanned reoperation” that allows identification of an unplanned reoperation. Analysis of this data would permit evaluation of potential differences in distribution of surgical cases among the different race groups. Differences in distribution of surgical procedures between African American and White children may very well play a role in the higher rate of reoperation and subsequent FTR. The next logical step, using the NSQIP database, will be assessment of differences in the distribution of surgical cases between the 2 groups. The ACS NSQIP-Pediatric database includes more than 800 unique CPT codes. The risk of mortality for each procedure each with a unique CPT code has been studied and 4 different risk quartiles determined.15 Unfortunately, a limitation of the ACS NSQIP-Pediatric database is the lack of information (type, location, and size) about individual hospitals. This prevents the comparison of institutional outcomes and evaluation of potential clustering during analysis. In addition, the NSQIP-Pediatric database does not include information on social risk factors and socioeconomic status which may differ among racial groups. There are other large administrative databases available for analysis, but each has inherent limitations. The Pediatric Health Information System (PHIS) database is a large all-payer hospital administrative dataset, capturing billing information on all patients hospitalized at 52 participating institutions. The 52 freestanding children’s hospitals are part of the Children’s Hospital Association. The PHIS database includes granular data on the geographic location of the hospital, income linked to zip code, and payor mix. Analysis of this data could be used to assess the association between economic status and postoperative outcomes. A major disadvantage of the PHIS database is that it is a billing database, which does not include granular details of patients’ underlying disease. In addition, it is not possible to associate a surgical procedure with reoperation in this database. The Agency for Healthcare Research and Quality (AHRQ) Healthcare Utilization Project KID database includes income-based quartile data as well as the hospital type, region, and volume. It also includes the zip code income quartile data. These data can be used to control for clustering based on hospital. However, this database is limited to inpatients and does not include ambulatory procedures. The authors have made a valuable contribution to the growing body of knowledge regarding health care outcome disparities in African American as compared to White children. At this point, however, we need a more nuanced understanding of the causes of these disparities and we need to begin addressing those that are modifiable. In addition, we need to extend this analysis to other groups of non-White children. Without the ability to link databases, this is going to be a difficult task given the inherent limitations of each large administrative database. Smaller, institution-based databases with granular hospital and patient demographic, socioeconomic, and disease processes data exist but meaningful analysis is hampered by small cohort sizes. The best solution is likely systematic collection of granular data from multiple institutions to strike a balance between case numbers and data quality. Perhaps most importantly, it is essential that we as individual anesthesiologists work to recognize our own implicit biases and perceptions of racial groups. Ideally there should be no inequities in the delivery of the components of health care over which we have direct control such as anxiolysis and pain management.17 We are innovators and leaders in medical education, health care delivery, and research; our recognition that racial and ethnic disparities exist in all these areas and our daily engagement in addressing them will make a world of difference. DISCLOSURES Name: Viviane G. Nasr, MD. Contribution: This author wrote the manuscript and approved the final version. Name: James A. DiNardo, MD, FAAP. Contribution: This author helped edit the manuscript and approved the final version. This manuscript was handled by: Edward C. Nemergut, MD.