Hispanic Race Is Associated with Increased Risk of Severe Sepsis in 1913 Hospitalizations of Children with AML and Infection

Background

The development of severe sepsis is a life-threatening risk facing children undergoing the intensive therapy necessary to treat Acute Myeloid Leukemia (AML). Although all children receiving such therapies are vulnerable to treatment-related neutropenia, non-White children experience invasive infections at higher rates than White children. However, racial differences in the development of severe sepsis in children with AML has not been examined. This study, a secondary analysis of publicly available data, tested the hypothesis that, once children with AML develop an infection, non-White race is associated with an increased risk of severe sepsis.

Method

The 2016 Healthcare Cost and Utilization Project Kids' Inpatient Database (HCUP KID) contains data representing over three million individual hospital stays collected in 4200 U.S. hospitals including 80% of the hospitalizations of children. Each case file contains up to 30 discharge diagnosis codes in the International Classification of Diseases, 10^thRevision, Clinical Modification format (ICD-10-CM) and 15 Procedure Coding System codes (ICD-10-PCS). When race and Hispanic ethnicity were both reported by a hospital, these cases were recorded in the HCUP KID file as Hispanic race. For this study, cases were included if any of the 30 discharge diagnoses contained both an ICD-10-CM codes for AML (C9200, C9201, or C9102) and a code indicating infection, with fever being most commonly documented in over 60% of the cases. Uncommon forms of AML, cases with ICD-10-CM codes related to pregnancy or perinatal care, or any case in which the ICD-10-CM or ICD-10-PCS codes indicated bone marrow or stem cell transplant took place were removed. In order to avoid duplicate counting, cases with evidence that the child was transferred to another facility were also removed. Severe sepsis was defined as the presence of an infection in addition to the ICD-10-CM diagnosis codes for systemic inflammatory response syndrome (SIRS), severe sepsis, septic shock, or any case with documented evidence in the ICD-10-CM or ICD-10-PCS codes that mechanical ventilatory support or vasoactive medications were required. Logistic regression models were used to examine the association between race and sepsis. The final adjusted multivariate model included relevant covariates and incorporated bootstrapping procedures to estimate effects.

Results

The sample consisted of 1913 cases with documentation of both AML and at least one infectious complication, representing 66% of the total number of cases of AML. Of those cases with infection 49% were White, 27% Hispanic, 12% Black, 6% Asian, and 7% other race. Fourteen percent had documentation indicating severe sepsis occurred. The average age was 9.7 years (S.D. 6.5 years). Covariates meeting the threshold for inclusion in the final model were age, presence of neutropenia, obesity, and care provided in a designated children's hospital.

In the unadjusted model Hispanic children were more likely to have documented sepsis than White children (OR 1.87, p<.001). After potential confounding variables were controlled, the Hispanic racial group was associated with a 61% increased risk of developing sepsis (p=.003). There were no significant differences in the likelihood of the development of sepsis in Black (AOR = 1.27, p=.276), Asian (AOR=0.63, p=.193), or other race (AOR=1.48, p=.145) children when compared to White children in the adjusted model.

Conclusion

The risk of sepsis among Hispanic children hospitalized with AML and infection in the U.S. during 2016 was 61% greater than the risk for White children. Further analyses of available data and prospective studies are needed to better understand the unmodifiable (e.g. genetic) or modifiable (e.g. environmental, cultural, dietary, economic, etc.) factors that lead to the increased risk of severe sepsis in Hispanic children. Such research can be used to inform the monitoring and delivering of supportive care for children with AML to reduce treatment-related morbidity and mortality.

Background

The development of severe sepsis is a life-threatening risk facing children undergoing the intensive therapy necessary to treat Acute Myeloid Leukemia (AML). Although all children receiving such therapies are vulnerable to treatment-related neutropenia, non-White children experience invasive infections at higher rates than White children. However, racial differences in the development of severe sepsis in children with AML has not been examined. This study, a secondary analysis of publicly available data, tested the hypothesis that, once children with AML develop an infection, non-White race is associated with an increased risk of severe sepsis.

Method

The 2016 Healthcare Cost and Utilization Project Kids' Inpatient Database (HCUP KID) contains data representing over three million individual hospital stays collected in 4200 U.S. hospitals including 80% of the hospitalizations of children. Each case file contains up to 30 discharge diagnosis codes in the International Classification of Diseases, 10^thRevision, Clinical Modification format (ICD-10-CM) and 15 Procedure Coding System codes (ICD-10-PCS). When race and Hispanic ethnicity were both reported by a hospital, these cases were recorded in the HCUP KID file as Hispanic race. For this study, cases were included if any of the 30 discharge diagnoses contained both an ICD-10-CM codes for AML (C9200, C9201, or C9102) and a code indicating infection, with fever being most commonly documented in over 60% of the cases. Uncommon forms of AML, cases with ICD-10-CM codes related to pregnancy or perinatal care, or any case in which the ICD-10-CM or ICD-10-PCS codes indicated bone marrow or stem cell transplant took place were removed. In order to avoid duplicate counting, cases with evidence that the child was transferred to another facility were also removed. Severe sepsis was defined as the presence of an infection in addition to the ICD-10-CM diagnosis codes for systemic inflammatory response syndrome (SIRS), severe sepsis, septic shock, or any case with documented evidence in the ICD-10-CM or ICD-10-PCS codes that mechanical ventilatory support or vasoactive medications were required. Logistic regression models were used to examine the association between race and sepsis. The final adjusted multivariate model included relevant covariates and incorporated bootstrapping procedures to estimate effects.

Results

The sample consisted of 1913 cases with documentation of both AML and at least one infectious complication, representing 66% of the total number of cases of AML. Of those cases with infection 49% were White, 27% Hispanic, 12% Black, 6% Asian, and 7% other race. Fourteen percent had documentation indicating severe sepsis occurred. The average age was 9.7 years (S.D. 6.5 years). Covariates meeting the threshold for inclusion in the final model were age, presence of neutropenia, obesity, and care provided in a designated children's hospital.

In the unadjusted model Hispanic children were more likely to have documented sepsis than White children (OR 1.87, p<.001). After potential confounding variables were controlled, the Hispanic racial group was associated with a 61% increased risk of developing sepsis (p=.003). There were no significant differences in the likelihood of the development of sepsis in Black (AOR = 1.27, p=.276), Asian (AOR=0.63, p=.193), or other race (AOR=1.48, p=.145) children when compared to White children in the adjusted model.

Conclusion

The risk of sepsis among Hispanic children hospitalized with AML and infection in the U.S. during 2016 was 61% greater than the risk for White children. Further analyses of available data and prospective studies are needed to better understand the unmodifiable (e.g. genetic) or modifiable (e.g. environmental, cultural, dietary, economic, etc.) factors that lead to the increased risk of severe sepsis in Hispanic children. Such research can be used to inform the monitoring and delivering of supportive care for children with AML to reduce treatment-related morbidity and mortality.