BACKGROUND AND OBJECTIVES:

To explore the microbiologic etiology and trends in incidence and survival of nonneonatal pediatric sepsis in the United States by using the 2006, 2009, and 2012 Kids’ Inpatient Database.

METHODS:

Children with sepsis were identified by using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes for severe sepsis and septic shock (ICD-9-CM cohort) and by the modified Angus method, which incorporates ICD-9-CM codes for infection and organ dysfunction (Angus cohort). Temporal trends in incidence and microbiologic etiology were evaluated.

RESULTS:

Among 8 830 057 discharges, 26 470 patients in the ICD-9-CM cohort were diagnosed with severe sepsis and septic shock (29.97 per 10 000 discharges) and 89 505 patients in the Angus cohort (101.34 per 10 000 discharges). The incidence of sepsis increased in both cohorts from 2006 to 2012 (P < .01). In the Angus cohort, the case-fatality rate was the highest for methicillin-resistant Staphylococcus aureus (14.42%, P < .01) among Gram-positive organisms and for Pseudomonas (21.49%; odds ratio: 2.58 [95% confidence interval: 1.88–3.54]; P < .01) among Gram-negative organisms.

CONCLUSIONS:

The incidence of sepsis has increased, and the sepsis case-fatality rate has decreased, without a decrease in the overall sepsis-associated mortality rate among hospitalized children. Also, bacterial and fungal organisms associated with the pediatric sepsis have changed over these years. These findings are important for focusing the allocation of health care resources and guiding the direction of future studies.

Sepsis is a leading cause of morbidity and mortality in hospitalized children in the United States.1,2  A global prospective study (the Sepsis Prevalence, Outcomes, and Therapy [SPROUT] study) revealed that hospital mortality rate was 25% in pediatric sepsis, which was higher than what was previously estimated in retrospective studies that used administrative databases.3  In the recent years, several studies have revealed an increase in the incidence rate but decrease in the mortality rate related to severe sepsis and septic shock.47 

In the past, the definition of sepsis was vague without precise guidelines for determining the diagnoses of sepsis and septic shock. In 1992, a consensus conference sponsored by the American College of Chest Physicians and the Society of Critical Care Medicine introduced definitions for the terms “systemic inflammatory response syndrome” and “multiple organ dysfunction syndrome” as they relate to infection.8  In 2005, the International Pediatric Sepsis Consensus Conference published a modified systemic inflammatory response syndrome criteria and definitions for severe sepsis and septic shock for the pediatric population.9  In 2016, the Third International Consensus Definitions for Sepsis and Septic Shock redefined sepsis definitions for adults.10  Sepsis is, now, defined as life-threatening organ dysfunction caused by a dysregulated host response to infection.10  No such definition of dysregulated host response has been adopted for pediatric patients, although a consensus of doing that has been growing.11  Organ dysfunction is identified as an acute change in the Sequential Organ Failure Assessment (SOFA) score consequent to an infection.12 

In 2001, Angus et al13  developed criteria to identify patients with severe sepsis using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes, which defined severe sepsis as presence of bacterial or fungal infection and at least 1 organ dysfunction, whereas the ICD-9-CM cohort was defined as a set of explicit ICD-9-CM codes for severe sepsis (995.92) and septic shock (785.52).14  Although billing codes for the sepsis definition provided moderate reliability (consistency of a measure) in comparison to SOFA scores with high reliability, the billing codes did have higher content validity (reflects clinician judgement) when compared to SOFA scores.15  There are few studies in which researchers have investigated the microbiologic etiologies responsible for severe sepsis in children.4,16,17  In the current study, the incidence of sepsis was compared between the 2 cohorts in the hospitalized pediatric patients by using the 2006, 2009, and 2012 Kids’ Inpatient Databases (KIDs). Patients were identified by using explicit ICD-9-CM codes for severe sepsis and septic shock (ICD-9-CM cohort) and by the modified Angus criteria (Angus cohort). We, then, compared the trends in incidence of sepsis from 2006 to 2012 using these 2 criteria. Finally, we described the underlying microbiologic etiology and associated sepsis case-fatality rate in these children.

A retrospective cross-sectional database study was conducted by using the 2006, 2009, and 2012 KID from Healthcare Cost and Utilization Project of the Agency for Healthcare Research and Quality. The KID is the largest publicly available all-payer pediatric inpatient care database in the United States, containing data from 2 to 3 million pediatric hospital discharges, excluding neonates, each year.18  We used 2 methods to identify pediatric patients with sepsis; this was based on methodology used by Balamuth et al,19  which formed the foundation of this study’s analysis. In ICD-9-CM cohort, all patients from 1 month old to 20 years old with diagnoses of severe sepsis and septic shock (ICD-9-CM codes 995.92 and 785.52, respectively) were included in the analysis. The incidence, sepsis case-fatality rate, length of stay (LOS), and associated comorbidities were determined in this cohort. In the Angus cohort, the method described by Angus et al13  was used to identify patients with sepsis in the 2006, 2009, and 2012 KID with diagnoses of bacterial and fungal infections, along with the diagnoses indicating acute organ dysfunction. Clinical Classifications Software (CCS) 2015 version was used to identify pediatric sepsis by using modified Angus criteria (Supplemental Tables 57).20  Total hospital charges were adjusted for inflation to the year 2012, by using the Consumer Price Index inflation calculator from the US Department of Labor Bureau of Labor Statistics.21  The Institutional Review Board at Nicklaus Children’s Hospital determined this study to be exempt.

The overall incidence of sepsis was presented as cases per 10 000 discharges, whereas the incidence of sepsis caused by specific microorganisms was calculated in the Angus cohort and presented as percentages. χ2 for trend analysis was used to analyze trends in incidence of sepsis in both cohorts and for each individual microbiologic etiology. Epi Info StatCalc (Centers for Disease Control and Prevention, Atlanta, GA) was used for trend analysis to compare the changes in incidence and sepsis case-fatality rate from 2006 to 2012. Children were grouped by age into 4 groups: infant (1 month to 1 year), toddler (2–5 years), school-aged (6–12 years), and adolescent (13–20 years). χ2 was used to compare categorical variables, and the Kruskal-Wallis test was used for comparing continuous variables. The odds ratio (OR) of the sepsis case-fatality rate (Table 1) was calculated by using Pearson’s χ2, which used the reference value as odds of sepsis case-fatality rate from all causes of sepsis in comparison to that of a particular microorganism. In the ICD-9-CM cohort, trend analysis was done for the proportion of patients coded for septic shock, All Patients Refined Diagnosis Related Group (APR DRG) severity of illness (SOI) category 4 (extreme category), use of mechanical ventilation (invasive or noninvasive) or vasopressors, and discharge from children’s hospitals. Binary regression analysis was performed in the ICD-9-CM cohort to determine the effect of calendar year of admission on the case-fatality rate adjusting for other variables, which revealed a significant linear trend. For regression analysis, we have recoded APR DRG SOI into 3 categories, category 4 (extreme), category 3 (major), and others (minor and moderate). Data were analyzed by using SPSS version 17 (SPSS Inc, Chicago, IL). Sample weighting was used to present national estimates.

TABLE 1

The Trend of SOI Indicators and Discharges From Children’s Hospitals in the ICD-9-CM Cohort From 2006 to 2012

Variable200620092012TotalSignificance for Trend
APR DRG SOI 4, % 87.6 86.5 76.0 82.8 <.001 
Septic shock ICD-9-CM code, % 67.9 72.8 72.6 71.4 <.001 
Mechanical ventilation or vasopressor use, % 56.5 55.5 49.2 53.4 <.001 
Children’s hospital discharges, % 27.4 29.9 31.9 30.0 <.001 
Variable200620092012TotalSignificance for Trend
APR DRG SOI 4, % 87.6 86.5 76.0 82.8 <.001 
Septic shock ICD-9-CM code, % 67.9 72.8 72.6 71.4 <.001 
Mechanical ventilation or vasopressor use, % 56.5 55.5 49.2 53.4 <.001 
Children’s hospital discharges, % 27.4 29.9 31.9 30.0 <.001 

There were 8 830 057 nonneonatal pediatric patient discharges from the US hospitals during the years 2006, 2009, and 2012. In total, 26 470 patients were diagnosed with severe sepsis or septic shock (29.97 per 10 000 discharges) according to ICD-9-CM criteria and 89 505 patients (101.34 per 10 000 discharges) according to the modified Angus criteria. There were 18 258 patients who qualified in both the ICD-9-CM and Angus cohorts. The age distribution was significantly different in the ICD-9-CM cohort compared with the Angus cohort (Table 2). The relative proportion of infants and toddlers was greater in the Angus cohort, whereas there was a greater proportion of adolescents in the ICD-9-CM cohort. The sex distribution was similar in both cohorts. Racial distribution was similar in both cohorts, when compared to other hospital discharges.

TABLE 2

Demographic Characteristic of Pediatric Sepsis

ICD-9-CM Code CohortAngus Cohort
2006 (n = 7781)2009 (n = 9178)2012 (n = 9511)Total (26, 470)2006 (n = 25 418)2009 (n = 32 170)2012 (n = 31 917)Total (n = 89 505)
Prevalence per 10 000 discharges 25.47a 30.15a 34.81a 29.97a 83.22a 105.68a 116.82a 101.34a 
Sepsis case-fatality rate, % 19.8a 16.9a 12.9a 16.52a 8.9a 7.8a 6.9a 7.87a 
Sepsis-associated mortality per 10 000 discharges 5.04 5.10 4.49 4.95 7.41a 8.24a 8.06a 7.98a 
Age, median (IQR) 11 (2–17) 12 (2–18) 13 (3–18) 12 (2–18) 9 (1–17) 10 (2–18) 10 (2–18) 10 (2–17) 
Age groups, n (%)         
 Infants (1 mo–1 y) 1010 (13) 1210 (13.2) 1336 (14) 3556 (13.4)a 4762 (18.7) 5108 (15.9) 5734 (18) 15 603 (17.3)a 
 Toddlers and preschoolers (2–5 y) 1289 (16.6) 1828 (19.9) 1746 (18.4) 4864 (18.3)a 5821 (22.9) 7534 (23.4) 6943 (21.8) 20 298 (22.5)a 
 School-aged children (6–12 y) 1137 (14.6) 1663 (18.1) 1620 (17) 4420 (16.7)a 4168 (16.4) 5315 (16.5) 5145 (16.1) 14 628 (16.3)a 
 Adolescents and young adults (13–20 y) 4345 (55.8) 4477 (48.8) 4809 (50.6) 13 630 (51.5)a 10 667 (42) 14 213 (44.2) 14 095 (44.2) 38 974 (43.3)a 
Female sex, % 46.5 49.2 50.1 48.6 48.7 50.1 48.8 49.2 
LOS, d, median (IQR) 13 (6–27) 12 (5–26) 10 (5–22) 11 (5–25) 12 (5–26) 11 (5–25) 11 (5–24) 11 (5–25) 
Total chargesa, $, median (IQR) 93 824 (39 675–2,15 841) 110 929 (45 622–274 962) 111 653 (46 608–292 745) 112 397 (46 976–277 537) 77 627 (29 032–190 337) 89 183 (32 452–237 707) 106 777 (37 571–287 297) 97 178 (35 239–253 721) 
Race and/or ethnicity, %         
 White 49.3 49.0 48.0 48.9 49.3 48.7 48.7 48.9 
 Black 16.4 16.4 17.9 16.9 18.6 18.6 19.0 18.6 
 Hispanic 22.9 23.6 22.4 22.8 21.3 22.2 22.0 22 
 Others 11.4 11.1 11.8 11.4 10.8 10.4 10.2 10.5 
ICD-9-CM Code CohortAngus Cohort
2006 (n = 7781)2009 (n = 9178)2012 (n = 9511)Total (26, 470)2006 (n = 25 418)2009 (n = 32 170)2012 (n = 31 917)Total (n = 89 505)
Prevalence per 10 000 discharges 25.47a 30.15a 34.81a 29.97a 83.22a 105.68a 116.82a 101.34a 
Sepsis case-fatality rate, % 19.8a 16.9a 12.9a 16.52a 8.9a 7.8a 6.9a 7.87a 
Sepsis-associated mortality per 10 000 discharges 5.04 5.10 4.49 4.95 7.41a 8.24a 8.06a 7.98a 
Age, median (IQR) 11 (2–17) 12 (2–18) 13 (3–18) 12 (2–18) 9 (1–17) 10 (2–18) 10 (2–18) 10 (2–17) 
Age groups, n (%)         
 Infants (1 mo–1 y) 1010 (13) 1210 (13.2) 1336 (14) 3556 (13.4)a 4762 (18.7) 5108 (15.9) 5734 (18) 15 603 (17.3)a 
 Toddlers and preschoolers (2–5 y) 1289 (16.6) 1828 (19.9) 1746 (18.4) 4864 (18.3)a 5821 (22.9) 7534 (23.4) 6943 (21.8) 20 298 (22.5)a 
 School-aged children (6–12 y) 1137 (14.6) 1663 (18.1) 1620 (17) 4420 (16.7)a 4168 (16.4) 5315 (16.5) 5145 (16.1) 14 628 (16.3)a 
 Adolescents and young adults (13–20 y) 4345 (55.8) 4477 (48.8) 4809 (50.6) 13 630 (51.5)a 10 667 (42) 14 213 (44.2) 14 095 (44.2) 38 974 (43.3)a 
Female sex, % 46.5 49.2 50.1 48.6 48.7 50.1 48.8 49.2 
LOS, d, median (IQR) 13 (6–27) 12 (5–26) 10 (5–22) 11 (5–25) 12 (5–26) 11 (5–25) 11 (5–24) 11 (5–25) 
Total chargesa, $, median (IQR) 93 824 (39 675–2,15 841) 110 929 (45 622–274 962) 111 653 (46 608–292 745) 112 397 (46 976–277 537) 77 627 (29 032–190 337) 89 183 (32 452–237 707) 106 777 (37 571–287 297) 97 178 (35 239–253 721) 
Race and/or ethnicity, %         
 White 49.3 49.0 48.0 48.9 49.3 48.7 48.7 48.9 
 Black 16.4 16.4 17.9 16.9 18.6 18.6 19.0 18.6 
 Hispanic 22.9 23.6 22.4 22.8 21.3 22.2 22.0 22 
 Others 11.4 11.1 11.8 11.4 10.8 10.4 10.2 10.5 
a

Prevalence values and mortality rates, when analyzed by using Epi Info trend analysis, revealed that the trends over 3 y had a statistically significant difference; the P value was <.05.

b

Charges are expressed as 2012 dollars.

Although the incidence of sepsis by using either definition increased during the study period (P < .01), the sepsis case-fatality rate decreased during the same period (P < .01; Fig 1). The overall case-fatality rate during the 3 years was 16.5% in the ICD-9-CM cohort and 7.8% in the Angus cohort. However, the sepsis-associated mortality rate (per 10 000 discharges) remained unchanged in the ICD-9-CM cohort (P = .55 for trend) and increased in the Angus cohort (P < .001 for trend; Table 2). In the Angus cohort, the age-specific sepsis case-fatality rate was highest among infants, compared with the other age groups (10.2% in infants, 7.6% in toddlers, 8.0% in school-aged children, and 7.0% in adolescents; P < .01; Fig 2) When the SOI was compared by using the APR DRG, we found that ICD-9-CM cohort was more likely to have extreme loss of function compared with the Angus cohort (OR: 4.28 [95% confidence interval (CI): 4.10–4.47]; P < .001). In the ICD-9-CM cohort, trend analyses revealed that there was an increase in the coding of septic shock but a decrease in the proportion of APR DRG SOI 4 and use of mechanical ventilation or vasopressors from 2006 to 2012 (Table 1). There was an increase in total charges (adjusted for inflation to 2012) from 2006 to 2012 in both cohorts (P < .01; Kruskal-Wallis test; Table 2). The LOS for nonsurvivors was higher in comparison with those who survived (12 days [interquartile range (IQR) 3–36] vs 11 days [IQR 5–25]; P < .01) and nonsurvivors had higher total charges compared with survivors ($186 790 IQR [65 679–496 026] vs $97 178 [IQR 35 239–253 721]; P < .01). Among nonsurvivors, 7.9% of patients died on the day of admission, 12.2% on day 1, and 53.3% within 10 days of admission (Fig 3).

FIGURE 1

Sepsis case-fatality rate and incidence trends in ICD-9-CM cohort and Angus cohort. Incidence increased and mortality decreased from 2006 to 2012 (P < .01; Epi Info).

FIGURE 1

Sepsis case-fatality rate and incidence trends in ICD-9-CM cohort and Angus cohort. Incidence increased and mortality decreased from 2006 to 2012 (P < .01; Epi Info).

Close modal
FIGURE 2

Age-specific sepsis case-fatality rates in Angus cohort. The sepsis case-fatality rate in infants (10.2%) was significantly higher than other groups. P < .01 compared with other groups.

FIGURE 2

Age-specific sepsis case-fatality rates in Angus cohort. The sepsis case-fatality rate in infants (10.2%) was significantly higher than other groups. P < .01 compared with other groups.

Close modal
FIGURE 3

LOS of nonsurvivors. Within 10 days of admission, 53.3% of the nonsurvivors died. Most patients died on the first day (12.2%) of admission rather than the day of admission (7.9%). For purpose of demonstration, only the first 20 days of the LOS are depicted.

FIGURE 3

LOS of nonsurvivors. Within 10 days of admission, 53.3% of the nonsurvivors died. Most patients died on the first day (12.2%) of admission rather than the day of admission (7.9%). For purpose of demonstration, only the first 20 days of the LOS are depicted.

Close modal

Among the bacterial etiologies we investigated, the prevalence decreased for all except infections with Escherichia coli (6.4%–7.3%, P = .02). The ICD-CM code for methicillin-resistant Staphylococcus aureus (MRSA) (ICD-9-CM code 038.12) was not available until 2008, and, therefore, its prevalence could not be determined for 2006.22  In the Angus cohort, MRSA prevalence decreased from 2009 to 2012 (1.3%–1.1%; P = .02), additionally the “other” Gram-negative sepsis (ICD-9-CM code 38.49) had the highest prevalence (7.55%), followed by Streptococcal sepsis (7.11%), and E coli sepsis (5.7%; Table 3).

TABLE 3

The Incidence and Sepsis Case-Fatality Rate due to Various Bacterial Causes in the Angus Cohort

2006, OR (95% CI)2009, OR (95% CI)2012, OR (95% CI)Prevalence of Bacterial Isolationa, %Case-Fatality Ratea, %
Streptococcal sepsis 1.63 (1.33–1.99) 2.15 (1.78–2.60) 1.90 (1.53–2.36) 7.11 13.63 
MSSA sepsis 1.54 (1.25–1.90) 1.53 (1.15–2.04) 1.69 (1.29–2.20) 5.12 11.98 
MRSA sepsisb — 2.46 (1.91–3.17) 1.75 (1.27–2.43) 1.24 14.42 
Other Staphylococcus sepsis 2.00 (1.57–2.53) 1.26 (0.90–1.76) 2.20 (1.57–3.08) 2.47 13.46 
Pneumococcal sepsis 1.23 (0.86–1.77) 1.00 (1.00–1.01) 1.88 (1.22–2.88) 2.72 10.64 
Bacteroides fragilis sepsis 1.63 (0.81–3.30) 2.11 (1.20–3.74) 1.51 (0.73–3.14) 0.2 13.08 
H influenzae sepsis 1.16 (0.41–3.26) 1.08 (0.39–3.01) 0.85 (0.31–2.34) 0.39 3.48 
E coli sepsis 1.88 (1.46–2.42) 1.55 (1.21–1.99) 1.05 (0.79–1.41) 5.7 10.93 
Pseudomonas sepsis 3.79 (2.91–4.94) 3.39 (2.64–4.36) 2.58 (1.88–3.54) 2.64 21.49 
Meningococcal sepsis 1.87 (1.20–2.93) 1.56 (0.88–2.79) 2.59 (1.36–4.95) 1.04 14.32 
Serratia sepsis 1.71 (1.02–2.86) 1.49 (0.82–2.72) 0.94 (0.34–2.59) 0.57 10 
Other Gram-negative sepsis 2.11 (1.75–2.54) 1.61 (1.32–1.96) 1.79 (1.44–2.21) 7.55 13.22 
NOS Gram-negative sepsis 2.37 (1.69–3.33) 3.70 (2.68–5.10) 2.90 (2.02–4.16) 1.88 19.96 
Bacteremia NOS 0.70 (0.60–0.82) 0.70 (0.60–0.82) 0.80 (0.68–0.96) 1.26 6.04 
2006, OR (95% CI)2009, OR (95% CI)2012, OR (95% CI)Prevalence of Bacterial Isolationa, %Case-Fatality Ratea, %
Streptococcal sepsis 1.63 (1.33–1.99) 2.15 (1.78–2.60) 1.90 (1.53–2.36) 7.11 13.63 
MSSA sepsis 1.54 (1.25–1.90) 1.53 (1.15–2.04) 1.69 (1.29–2.20) 5.12 11.98 
MRSA sepsisb — 2.46 (1.91–3.17) 1.75 (1.27–2.43) 1.24 14.42 
Other Staphylococcus sepsis 2.00 (1.57–2.53) 1.26 (0.90–1.76) 2.20 (1.57–3.08) 2.47 13.46 
Pneumococcal sepsis 1.23 (0.86–1.77) 1.00 (1.00–1.01) 1.88 (1.22–2.88) 2.72 10.64 
Bacteroides fragilis sepsis 1.63 (0.81–3.30) 2.11 (1.20–3.74) 1.51 (0.73–3.14) 0.2 13.08 
H influenzae sepsis 1.16 (0.41–3.26) 1.08 (0.39–3.01) 0.85 (0.31–2.34) 0.39 3.48 
E coli sepsis 1.88 (1.46–2.42) 1.55 (1.21–1.99) 1.05 (0.79–1.41) 5.7 10.93 
Pseudomonas sepsis 3.79 (2.91–4.94) 3.39 (2.64–4.36) 2.58 (1.88–3.54) 2.64 21.49 
Meningococcal sepsis 1.87 (1.20–2.93) 1.56 (0.88–2.79) 2.59 (1.36–4.95) 1.04 14.32 
Serratia sepsis 1.71 (1.02–2.86) 1.49 (0.82–2.72) 0.94 (0.34–2.59) 0.57 10 
Other Gram-negative sepsis 2.11 (1.75–2.54) 1.61 (1.32–1.96) 1.79 (1.44–2.21) 7.55 13.22 
NOS Gram-negative sepsis 2.37 (1.69–3.33) 3.70 (2.68–5.10) 2.90 (2.02–4.16) 1.88 19.96 
Bacteremia NOS 0.70 (0.60–0.82) 0.70 (0.60–0.82) 0.80 (0.68–0.96) 1.26 6.04 

Data are presented as odds of sepsis case-fatality rate due to the listed causes of infection as compared with other causes of infection in Angus Cohort. Prevalence is percentage of patients due to given cause of infection as compared all other causes of sepsis in Angus cohort. MSSA, methicillin-sensitive S aureus; NOS, not otherwise specified; —, not applicable.

a

Prevalence and case-fatality rate represents all 3 y (2006, 2009, and 2012).

b

The MRSA sepsis ICD-9-CM code was introduced after 2008 and, therefore, was not calculated for the 2006 database.

Overall, there was a decrease in sepsis case-fatality rate due to most bacterial organisms during the study period; the only exception was sepsis caused by meningococcus, in which the case-fatality rate increased from 15.54% to 16.17% (P = .01). The median LOS for patients with meningococcal sepsis who did not survive was much lower compared with the patients who died of other causes of sepsis (1 day [IQR 0–3 days] vs 12 days [IQR 3–35 days]; P < .01). The case-fatality rate of sepsis due to Pseudomonas (21.49%) was the highest when compared to the other bacterial pathogens among the patients in the Angus cohort for all 3 years (Table 3). The case-fatality rate of all Staphylococcal infections (methicillin-sensitive S aureus, MRSA, and other staph species) was 12.8% (OR: 1.78 [95% CI: 1.61–1.96]).

The overall prevalence of fungal infection was 5.04% during the study period. The prevalence did not change significantly from 2006 to 2012, but the case-fatality rate decreased significantly (19.6%–11.6%; P < .01; Fig 4). Among fungal infections, candida was the most commonly associated organism with severe sepsis (3.77%), followed by Aspergillus (0.27%; Table 4). Invasive aspergillosis had the highest case-fatality rate among all microbiologic organism found in the study (28.2%; P < .01).

FIGURE 4

Incidence and sepsis case-fatality rate of fungal infections among patients with severe sepsis.

FIGURE 4

Incidence and sepsis case-fatality rate of fungal infections among patients with severe sepsis.

Close modal
TABLE 4

Fungal Infections Among Patients With Severe Sepsis in the Angus Cohort

Fungal InfectionNo. PatientsDeceasedCase-Fatality Rate, %
Candidiasis 3455 403 11.7a 
Aspergillosis 887 250 28.2a 
Coccidioidomycosis 82 11.0 
Dermatophytosis 38 5.3 
Histoplasmosis 22 9.1 
Blastomycosis 25 12.0 
Fungal InfectionNo. PatientsDeceasedCase-Fatality Rate, %
Candidiasis 3455 403 11.7a 
Aspergillosis 887 250 28.2a 
Coccidioidomycosis 82 11.0 
Dermatophytosis 38 5.3 
Histoplasmosis 22 9.1 
Blastomycosis 25 12.0 
a

Sepsis case-fatality rate dues to these fungal infections were found to be statistically significant when compared to other causes of infection among patients with severe sepsis.

Overall, 30% of children in ICD-9-CM cohort and 31.8% in Angus cohort were discharged from children’s hospitals. The overall incidence of sepsis with ICD-9-CM criteria in children’s hospitals was 50.8 (40.4 in 2006 to 59.5 in 2012) per 10 000 discharges and with Angus criteria was 183.9 (157.3 in 2006 to 205.8 in 2012) per 10 000 discharges. There was an increase in the trend of discharge from children’s hospitals in the ICD-9-CM cohort during the study period (Table 1). The case-fatality rate was not significantly different in those discharged from children’s hospital in comparison to nonchildren’s hospitals in the ICD-9-CM cohort (16.4% vs 15.9%; OR: 1.04 [95% CI: 0.97–1.12]) but was higher in the Angus cohort (8.4% vs 7.6%; OR: 1.12 [95% CI: 1.06–1.18]).

In the ICD-9-CM cohort, a binary regression analysis was performed, with mortality as an independent variable and calendar year of admission, septic shock, APR DRG SOI, use of mechanical ventilation or vasopressors, and discharge from children’s hospitals as dependent factors. Compared to 2006, the case-fatality rate from sepsis was lower in 2009 (OR: 0.860 [95% CI 0.792–0.935]) and 2012 (OR: 0.716 [95% CI: 0.657–0.781]).

This study is focused on sepsis in the nonneonatal pediatric population. The disease process, etiology, pathogenesis, and outcomes of sepsis in these patients differ substantially from neonates.23  The KID provides a large sample size and includes data from 4200 US community hospitals, and national estimates gave us an opportunity to evaluate >8 million nonneonatal hospital discharges.18  This is a larger sample size compared to the Public Health Information System (PHIS), which includes data from 45 children’s hospitals.24  Studies have revealed that mortality rates and other parameters, like medical complexity and LOS, differ between large teaching hospitals and smaller hospitals.25 

There was an increase in the incidence and decrease in the case-fatality rate from severe sepsis and septic shock during the study period by using both the ICD-9-CM and modified Angus criteria. As expected, the incidence of sepsis was higher by using the modified Angus criteria compared to the ICD-9-CM criteria because the modified Angus method is a more sensitive method of identifying the patients with sepsis.26  However, the sepsis case-fatality rate was higher when ICD-9-CM specific codes for severe sepsis and septic shock were used to identify sepsis. This suggests higher specificity but lower sensitivity in identifying patients with severe illness. The increase in the incidence and decrease in the sepsis case-fatality rate is similar to the other published pediatric sepsis studies.6,19  The increased incidence of sepsis may be due to a true increase in sepsis incidence or due to changes in coding and/or documentation practices or a combination of both. The increased incidence may be attributed to sicker patients being hospitalized, increased survival of patients with complex conditions, and improved sepsis detection. A recent study revealed increased prevalence of in-hospital cardiac arrests, suggesting an increased proportion of higher acuity patients being admitted to the US hospitals.27  However, the unchanged sepsis-associated mortality rate (per 10 000 total discharges) in the ICD-9-CM cohort and increased sepsis-associated mortality rate in the Angus cohort in the face of the decreasing sepsis case-fatality rate may suggest overdiagnosis. Similarly, patients in the ICD-9-CM cohort were much sicker in comparison with the Angus cohort, on the basis of APR DRG SOI, which could again point toward overdiagnosis of pediatric sepsis by using the Angus criteria.28,29  The increase in coding of septic shock but decrease in the proportion of APR DRG SOI 4 and use of mechanical ventilation or vasopressors from 2006 to 2012 (Table 1) can be due to a change in coding and/or documentation practice or due to early recognition and rescue of sepsis episodes, thereby decreasing SOI and critical interventions, or, most likely, due to a combination of both.

Over the study period, the gap in incidence of severe sepsis between the 2 cohorts seemed to decrease, which may indicate better identification (or documentation) over the time. We identified sepsis 3.4 times more often using the modified Angus criteria than we did using only the ICD-9-CM codes for severe sepsis and septic shock. In a recent study, researchers suggested a need for improvement in the way we code pediatric organ dysfunction by incorporating age-specific organ dysfunction thresholds, like the pediatric SOFA score, Pediatric Logistic Organ Dysfunction score, or Pediatric Multiple Organ Dysfunction Score.30 

When this study was compared to a similar study by Hartman et al,4  both studies revealed an increase in the incidence of sepsis, whereas culture-positive sepsis organisms continue to decrease in incidence. These trends could be explained by the increase in the coding of sepsis in the setting of organ dysfunction, for which infection was suspected but not confirmed by culture data (so called “culture-negative sepsis”). Similarly, in other studies in adults, researchers have found that coding for sepsis has become more inclusive, with increased application of these codes to those without positive blood culture results.31,32  The only exception to this trend that was found in this study was the sepsis due to E coli, in which incidence increased from 5.7% to 6.2% (P = .02). In this study, we also found that Streptococcal sepsis had the highest incidence among the patients who had a positive culture result recorded (7.11%; Table 1); similarly, other studies had found that in nonneonatal pediatric patients with bacteremia, Streptococcal pneumoniae was the most commonly identified organism, followed by Neisseria meningitidis.33,34 

The sepsis case-fatality rate was almost double in the ICD-9-CM cohort compared with Angus cohort, which could be due to more specific codes for severe sepsis and septic shock used for this cohort, thus representing a much sicker population in the former. Although there was a decreasing trend in the case-fatality rate from severe sepsis during the study period, the sepsis-associated mortality rate per all discharges either remained flat or increased (Table 2). Therefore, the increase in sepsis incidence can point toward misattribution or overdiagnosis of sepsis due to an emphasis on early recognition of sepsis or a real increase in sepsis due to an increase in complex medical conditions and SOI among hospitalized patients. As previously mentioned, the Angus cohort identified 3.4 times as many patients with sepsis than the ICD-9-CM cohort. Therefore, we used the Angus cohort to calculate the case-fatality rate because we felt that it was a complete representation of pediatric sepsis. In fact, a 2014 study by Iwashyna et al26  on adult patients, revealed that the Angus cohort had higher sensitivity (50.3% vs 9.2%) in identification of sepsis, although they had a lower positive predictive value (70.7% vs 100%) in comparison to ICD-9-CM diagnosis codes. This could again be due to overestimation of sepsis by using administrative data. Similarly, identifying microbiologic etiology by using ICD-9-CM codes had its own inherent limitations and sensitivities, ranging from 23% to 78%, according to various studies.22,35,36  Sepsis from Gram-negative organisms had a higher case-fatality rate, compared to sepsis from Gram-positive organisms. The case-fatality rate declined for all bacterial etiologies of sepsis that were identified in the study except for meningococcemia. The incidence of meningococcal sepsis decreased significantly because of the introduction of the meningococcal vaccine for adolescents in 2005.37,38  However, the case-fatality rate for meningococcal sepsis increased from 2006 to 2012. Interestingly, those patients who died of meningococcemia had a short LOS, indicating that they died relatively quickly after admission. It appears that the current strategies for early recognition and management of sepsis have not made an impact on the early case-fatality rate from meningococcemia. This may indicate that these strategies are inadequate in patients presenting with meningococcemia or that the serotype of N meningitidis leading to an increased case-fatality rate is not covered by current vaccination.

Regression analysis revealed that after adjusting for SOI, use of critical interventions, and discharge from children’s hospitals, the case-fatality rate was lower in 2009 and 2012 compared to that in 2006. Although there may be a component of increased coding as a reason for an increased incidence of sepsis, the finding of reduction in case-fatality rate still persisted after adjusting for some of the variables. Various factors have played a role in reducing the sepsis case-fatality rate. The Surviving Sepsis Campaign began in 2001, with a goal of reducing mortality from sepsis in adults by 25%. Guidelines specific to pediatric sepsis were introduced in 2008, and, later, a pediatric specific algorithm and recommendations were developed in 2020, as a part of the Surviving Sepsis Campaign.3941  Current guidelines are focused on early recognition of sepsis and the use of age-specific therapies to attain time-sensitive goals.

Less than one-third of all children in both cohorts in our study were discharged from children’s hospitals. There was an increase in the trend of discharges from children’s hospitals during the study period. The case-fatality rate was similar in children’s hospital compared to nonchildren’s hospitals in the ICD-9-CM cohort but was significantly higher in the children’s hospital (the absolute difference in case-fatality rate was 0.8%) among those in the Angus cohort. Balamuth et al19  have included data from children’s hospitals during a 9-year period from 2004 to 2012, whereas our study included national data of children discharged during 3 years in similar time frame. The other key demographic difference between the 2 studies is that we have included children <21 years of age but excluded neonates, whereas the PHIS study included all patients <18 years of age, including neonates. The case-fatality rate was similar in the Angus cohort in our study (8.4% vs 8.1%) compared to that in the study by Balamuth et al,19  which used the PHIS database. However, the overall case-fatality rate is 16.4%, compared to 21.2% in the PHIS database study in the ICD-9-CM cohort. The PHIS database includes data from a limited number of children’s hospitals, whereas the KID data set includes data from all children’s hospitals. The ratio of the Angus to ICD-9-CM cohort in the Balmuth et al19  study is 7.1, whereas the ratio is 3.6 in our study among discharges from children’s hospitals. The age distribution and characteristics of children’s hospitals may have contributed to the differences in case-fatality rates. The findings of increased incidence and decreased case-fatality rate of sepsis in hospitalized children is similar in both studies. Similar trends are noted in children’s hospitals, as reported in Balamuth et al’s19  study, as well as in all hospitals as in our study.

Sepsis continues to be a leading cause of morbidity and mortality among pediatric patients. Although the exact cost in the pediatric population is unclear, sepsis in adults has been found to account for 13% of the total US hospital costs ($24 billion in 2013), making it the highest among the admissions for all disease states.42  As of 2003, the estimated cost of sepsis in children in the US was $1.97 billion.16  In this study, it was determined that the total hospital charges for nonneonatal sepsis in children were $8.38 billion in 2012. Although the sepsis case-fatality rate had decreased, it was found that the LOS in nonsurvivors was 38% higher, with almost double the charges, than those who survived, almost similar to the study by Odetola et al.43  Similar correlations between mortality and increased charges have been found in other studies as well.6 

Fungal infections continue to be an important player in the pediatric sepsis epidemiology.44  Candida sepsis is the leading cause of fungal sepsis in our study and other similar studies.45  Ever since the introduction of amphotericin B and its use as an antifungal agent since 1958, there has been a decline in mortality due to invasive fungemia.46,47  Although the case-fatality rate of the fungal sepsis decreased in our study, the overall prevalence of fungal infections remained stable (∼5%) from 2006 to 2012. Invasive aspergillosis had the highest case-fatality rate (28.2%) among any other cause of sepsis in the study. This may be due to underlying serious illnesses, such as immunosuppression and transplant. This study revealed the case-fatality rate of invasive candidiasis was 11.7%, compared to the other studies, which had case-fatality rates ranging from 10% to 28%.4850  The higher case-fatality rates in other studies were probably due to a larger proportion of neonatal population. The decreased case-fatality rate in fungal sepsis can be explained by the advent of newer antifungal medications and early recognition of sepsis.

There are many limitations to the study. First, there is a lack of standardization of sepsis codes because they vary from 1 institution to another. There are chances that there may be errors while coding and misinterpretation of the clinical status of the patient. This may lead to over- or underestimation of sepsis.

Secondly, sensitivity of only a few codes has been validated and has variable results in different studies. Therefore, chances are that, in some cases, sepsis may not have been identified when it was present. Similarly, overestimation of sepsis, on the basis of the recent studies revealing an increase in sepsis coding, is also a possibility.32 

Lastly, as mentioned earlier, bacterial and fungal infection ICD-9-CM code sensitivity varies a lot, depending on the organism. This may skew the results for certain organisms.

In this study, we show that the incidence of sepsis has increased, and the case-fatality rate has decreased among children hospitalized with sepsis in the United States. Current strategies for early recognition and management of sepsis have not made an impact on the case-fatality rate of meningococcemia. These findings are important for focusing the allocation of health care resources and guiding the direction of future studies.

We thank Ricardo Meneses and Natasha Strump for their help with data collection. We also thank Taruna Sehgal, PhD, for helping us with editing the article.

Drs Sehgal and Ladd conceptualized and designed the study, drafted the initial manuscript, and reviewed and revised the manuscript; Dr Totapally conceptualized and designed the study, coordinated and supervised data collection, and critically reviewed the manuscript for important intellectual content; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

FUNDING: No external funding.

1
Centers for Disease Control and Prevention
. National vital statistics system, mortality. Available at: https://www.cdc.gov/nchs/data/dvs/LCWK1_2015.pdf. Accessed October 14, 2020
2
Fleischmann
C
,
Scherag
A
,
Adhikari
NKJ
, et al
.
Assessment of global incidence and mortality of hospital-treated sepsis. Current estimates and limitations
.
Am J Respir Crit Care Med
.
2016
;
193
(
3
):
259
272
3
Weiss
SL
,
Fitzgerald
JC
,
Pappachan
J
, et al
;
Sepsis Prevalence, Outcomes, and Therapies Study Investigators and Pediatric Acute Lung Injury and Sepsis Investigators Network
.
Global epidemiology of pediatric severe sepsis: the sepsis prevalence, outcomes, and therapies study
.
Am J Respir Crit Care Med
.
2015
;
191
(
10
):
1147
1157
4
Hartman
ME
,
Linde-Zwirble
WT
,
Angus
DC
,
Watson
RS
.
Trends in the epidemiology of pediatric severe sepsis*
.
Pediatr Crit Care Med
.
2013
;
14
(
7
):
686
693
5
Angus
DC
,
van der Poll
T
.
Severe sepsis and septic shock
. [
published correction appears in N Engl J Med. 2013;369(21):2069
].
N Engl J Med
.
2013
;
369
(
9
):
840
851
6
Ruth
A
,
McCracken
CE
,
Fortenberry
JD
,
Hall
M
,
Simon
HK
,
Hebbar
KB
.
Pediatric severe sepsis: current trends and outcomes from the Pediatric Health Information Systems database
.
Pediatr Crit Care Med
.
2014
;
15
(
9
):
828
838
7
Samransamruajkit
R
,
Limprayoon
K
,
Lertbunrian
R
, et al
.
The utilization of the surviving sepsis campaign care bundles in the treatment of pediatric patients with severe sepsis or septic shock in a resource-limited environment: a prospective multicenter trial
.
Indian J Crit Care Med
.
2018
;
22
(
12
):
846
851
8
Bone
RC
,
Balk
RA
,
Cerra
FB
, et al
.
Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis
.
Chest
.
1992
;
101
(
6
):
1644
1655
9
Goldstein
B
,
Giroir
B
,
Randolph
A
;
International Consensus Conference on Pediatric Sepsis
.
International pediatric sepsis consensus conference: definitions for sepsis and organ dysfunction in pediatrics
.
Pediatr Crit Care Med
.
2005
;
6
(
1
):
2
8
10
Singer
M
,
Deutschman
CS
,
Seymour
CW
, et al
.
The Third International Consensus Definitions for Sepsis and Septic Shock (sepsis-3)
.
JAMA
.
2016
;
315
(
8
):
801
810
11
Schlapbach
LJ
,
Kissoon
N
.
Defining pediatric sepsis
.
JAMA Pediatr
.
2018
;
172
(
4
):
312
314
12
Jones
AE
,
Trzeciak
S
,
Kline
JA
.
The Sequential Organ Failure Assessment score for predicting outcome in patients with severe sepsis and evidence of hypoperfusion at the time of emergency department presentation
.
Crit Care Med
.
2009
;
37
(
5
):
1649
1654
13
Angus
DC
,
Linde-Zwirble
WT
,
Lidicker
J
,
Clermont
G
,
Carcillo
J
,
Pinsky
MR
.
Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care
.
Crit Care Med
.
2001
;
29
(
7
):
1303
1310
14
ICD-9-CM coordination and maintenance committee meeting
.
Available at: https://www.cdc.gov/nchs/data/icd/agendadec02.pdf. Accessed February 28, 2020
15
Saria
S
,
Henry
KE
.
Too many definitions of sepsis: can machine learning leverage the electronic health record to increase accuracy and bring consensus?
Crit Care Med
.
2020
;
48
(
2
):
137
141
16
Watson
RS
,
Carcillo
JA
,
Linde-Zwirble
WT
,
Clermont
G
,
Lidicker
J
,
Angus
DC
.
The epidemiology of severe sepsis in children in the United States
.
Am J Respir Crit Care Med
.
2003
;
167
(
5
):
695
701
17
Prout
AJ
,
Talisa
VB
,
Carcillo
JA
,
Decker
BK
,
Yende
S
.
Bacterial and fungal etiology of sepsis in children in the United States: reconsidering empiric therapy
.
Crit Care Med
.
2020
;
48
(
3
):
e192
e199
18
Agency for Healthcare Research and Quality
.
Overview of the Kids’ Inpatient Database (KID)
. Available at: www.hcup-us.ahrq.gov/kidoverview.jsp. Accessed February 28, 2020
19
Balamuth
F
,
Weiss
SL
,
Neuman
MI
, et al
.
Pediatric severe sepsis in U.S. children’s hospitals
.
Pediatr Crit Care Med
.
2014
;
15
(
9
):
798
805
20
Agency for Healthcare Research and Quality
.
Clinical Classifications Software (CCS) for ICD-9-CM
. Available at: www.hcup-us.ahrq.gov/toolssoftware/ccs/ccs.jsp. Accessed February 28, 2020
21
US Bureau of Labor Statistics
.
CPI inflation calculator
. Available at: https://data.bls.gov/cgi-bin/cpicalc.pl. Accessed February 28, 2020
22
Schweizer
ML
,
Eber
MR
,
Laxminarayan
R
, et al
.
Validity of ICD-9-CM coding for identifying incident methicillin-resistant Staphylococcus aureus (MRSA) infections: is MRSA infection coded as a chronic disease?
Infect Control Hosp Epidemiol
.
2011
;
32
(
2
):
148
154
23
Wynn
JL
,
Wong
HR
.
Pathophysiology and treatment of septic shock in neonates
.
Clin Perinatol
.
2010
;
37
(
2
):
439
479
24
Rice
HE
,
Englum
BR
,
Gulack
BC
, et al
.
Use of patient registries and administrative datasets for the study of pediatric cancer
.
Pediatr Blood Cancer
.
2015
;
62
(
9
):
1495
1500
25
Burke
LG
,
Frakt
AB
,
Khullar
D
,
Orav
EJ
,
Jha
AK
.
Association between teaching status and mortality in US hospitals
.
JAMA
.
2017
;
317
(
20
):
2105
2113
26
Iwashyna
TJ
,
Odden
A
,
Rohde
J
, et al
.
Identifying patients with severe sepsis using administrative claims: patient-level validation of the Angus implementation of the international consensus conference definition of severe sepsis
.
Med Care
.
2014
;
52
(
6
):
e39
e43
27
Martinez
PA
,
Totapally
BR
.
The epidemiology and outcomes of pediatric in-hospital cardiopulmonary arrest in the United States during 1997 to 2012
.
Resuscitation
.
2016
;
105
:
177
181
28
Baram
D
,
Daroowalla
F
,
Garcia
R
, et al
.
Use of the All Patient Refined-Diagnosis Related Group (APR-DRG) risk of mortality score as a severity adjustor in the medical ICU
.
Clin Med Circ Respirat Pulm Med
.
2008
;
2
:
19
25
29
Fontaine
P
.
Using Severity Adjustment Classification for Hospital Internal and External Benchmarking
.
Chicago, IL
:
American Health Information Management Association
;
2004
30
Hsu
HE
,
Abanyie
F
,
Agus
MSD
, et al
.
A national approach to pediatric sepsis surveillance
.
Pediatrics
.
2019
;
144
(
6
):
e20191790
31
Rhee
C
,
Murphy
MV
,
Li
L
,
Platt
R
,
Klompas
M
;
Centers for Disease Control and Prevention Epicenters Program
.
Comparison of trends in sepsis incidence and coding using administrative claims versus objective clinical data
.
Clin Infect Dis
.
2015
;
60
(
1
):
88
95
32
Fleischmann-Struzek
C
,
Thomas-Rüddel
DO
,
Schettler
A
, et al
.
Comparing the validity of different ICD coding abstraction strategies for sepsis case identification in German claims data
.
PLoS One
.
2018
;
13
(
7
):
e0198847
33
Jafri
RZ
,
Ali
A
,
Messonnier
NE
, et al
.
Global epidemiology of invasive meningococcal disease
.
Popul Health Metr
.
2013
;
11
(
1
):
17
34
Gomez
B
,
Hernandez-Bou
S
,
Garcia-Garcia
JJ
,
Mintegi
S
;
Bacteraemia Study Working Group from the Infectious Diseases Working Group, Spanish Society of Pediatric Emergencies (SEUP)
.
Bacteremia in previously healthy children in emergency departments: clinical and microbiological characteristics and outcome
.
Eur J Clin Microbiol Infect Dis
.
2015
;
34
(
3
):
453
460
35
Olsen
MA
,
Fraser
VJ
.
Use of diagnosis codes and/or wound culture results for surveillance of surgical site infection after mastectomy and breast reconstruction
.
Infect Control Hosp Epidemiol
.
2010
;
31
(
5
):
544
547
36
Erik
RD
,
Kimberly
R
,
McDonald
LC
,
Victoria
F
.
ICD-9 codes and surveillance for Clostridium difficile-associated disease
.
Emerg Infect Dis J
.
2006
;
12
(
10
):
1576
–1576
37
Harrison
LH
.
Epidemiological profile of meningococcal disease in the United States
.
Clin Infect Dis
.
2010
;
50
(
suppl 2
):
S37
S44
38
Hamborsky
J
,
Kroger
A
,
Wolfe
CS
, eds.
Epidemiology and Prevention of Vaccine-Preventable Diseases.
13th ed.
Washington, DC
:
Public Health Foundation
;
2015
39
Levy
MM
,
Fink
MP
,
Marshall
JC
, et al
;
Society of Critical Care Medicine
European Society of Intensive Care Medicine
American College of Chest Physicians
American Thoracic Society
Surgical Infection Society
.
2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference
.
Crit Care Med
.
2003
;
31
(
4
):
1250
1256
40
Dellinger
RP
,
Levy
MM
,
Rhodes
A
, et al
;
Surviving Sepsis Campaign Guidelines Committee including The Pediatric Subgroup
.
Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock, 2012
.
Intensive Care Med
.
2013
;
39
(
2
):
165
228
41
Weiss
SL
,
Peters
MJ
,
Alhazzani
W
, et al
.
Surviving Sepsis Campaign international guidelines for the management of septic shock and sepsis-associated organ dysfunction in children
.
Intensive Care Med
.
2020
;
46
(
suppl 1
):
10
67
42
Paoli
CJ
,
Reynolds
MA
,
Sinha
M
,
Gitlin
M
,
Crouser
E
.
Epidemiology and costs of sepsis in the United States-an analysis based on timing of diagnosis and severity level
.
Crit Care Med
.
2018
;
46
(
12
):
1889
1897
43
Odetola
FO
,
Gebremariam
A
,
Freed
GL
.
Patient and hospital correlates of clinical outcomes and resource utilization in severe pediatric sepsis
.
Pediatrics
.
2007
;
119
(
3
):
487
494
44
Wisplinghoff
H
,
Seifert
H
,
Tallent
SM
,
Bischoff
T
,
Wenzel
RP
,
Edmond
MB
.
Nosocomial bloodstream infections in pediatric patients in United States hospitals: epidemiology, clinical features and susceptibilities
.
Pediatr Infect Dis J
.
2003
;
22
(
8
):
686
691
45
Zaoutis
TE
,
Coffin
SE
,
Chu
JH
, et al
.
Risk factors for mortality in children with candidemia
.
Pediatr Infect Dis J
.
2005
;
24
(
8
):
736
739
46
Gallis
HA
,
Drew
RH
,
Pickard
WW
.
Amphotericin B: 30 years of clinical experience
.
Rev Infect Dis
.
1990
;
12
(
2
):
308
329
47
Kaushik
A
,
Kest
H
.
The role of antifungals in pediatric critical care invasive fungal infections
.
Crit Care Res Pract
.
2018
;
2018
:
8469585
48
Cleveland
AA
,
Farley
MM
,
Harrison
LH
, et al
.
Changes in incidence and antifungal drug resistance in candidemia: results from population-based laboratory surveillance in Atlanta and Baltimore, 2008-2011
.
Clin Infect Dis
.
2012
;
55
(
10
):
1352
1361
49
Zaoutis
TE
,
Argon
J
,
Chu
J
,
Berlin
JA
,
Walsh
TJ
,
Feudtner
C
.
The epidemiology and attributable outcomes of candidemia in adults and children hospitalized in the United States: a propensity analysis
.
Clin Infect Dis
.
2005
;
41
(
9
):
1232
1239
50
Blyth
CC
,
Chen
SC
,
Slavin
MA
, et al
;
Australian Candidemia Study
.
Not just little adults: candidemia epidemiology, molecular characterization, and antifungal susceptibility in neonatal and pediatric patients
.
Pediatrics
.
2009
;
123
(
5
):
1360
1368

Competing Interests

POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.

Supplementary data