Abstract
Severe sepsis (SS) in pediatric oncology patients is a leading cause of morbidity and mortality. We investigated the incidence of and risk factors for morbidity and mortality among children diagnosed with cancer from 2008 to 2012, and admitted with SS during the 3 years following cancer diagnosis. A total of 1,002 children with cancer were included, 8% of whom required pediatric intensive care unit (PICU) admission with SS. Death and/or multiple organ dysfunction syndrome occurred in 34 out of 99 PICU encounters (34%). Lactate level and history of stem-cell transplantation were significantly associated with the development of death and/or organ dysfunction ( p < 0.05).
Keywords: sepsis, septic shock, oncology, morbidity and mortality
Introduction
The significant improvement in childhood cancer outcomes over the past several decades is due in part to advanced diagnostic modalities, aggressive treatment protocols, and vigilant supportive care. 1 2 3 However, increasingly intensive cancer therapies are associated with greater risk of life-threatening complications. Complications attributed to cancer treatment contribute to 16% of the 5-year mortality in children, of which infections continue to be the most common. 3 4 5 6 7 8 9 10
Severe sepsis (SS) is a leading cause for hospitalization, pediatric intensive care unit (PICU) admission, and morbidity and mortality among children with cancer. 6 7 9 In recent years, an increasing number of pediatric cancer patients required admission to the PICU for SS. 4 5 11 The available data regarding mortality rates from SS among children with cancer is highly variable, with mortality rates ranging from 8% to as high as 41% in a recent study. 5 12 13 14 15 16 In addition to mortality, multiorgan dysfunction syndrome (MODS) is a common complication of sepsis, reported in up to 60% of children with SS. 16 The development of MODS correlates with mortality 17 and poor functional outcome, 18 and has been suggested as another measure of outcome. However, the majority of patients with sepsis have MODS on admission to the PICU, which makes it a poor measure of PICU management. 19 Therefore, the development of new or progressive MODS (NPMODS) after the first day of PICU admission has been suggested as a more accurate measure of PICU outcome. 19
Because of the variability in reported mortality from SS in pediatric oncology patients, and the limited data on NPMODS in this population, we undertook the present study with the following objectives 1 : to better understand the baseline incidence and outcomes of SS among pediatric oncology patients, and 2 to identify risk factors for PICU-associated morbidity and mortality in this population. By identifying high-risk patients, this study could help inform opportunities to improve sepsis screening, surveillance and recognition.
Methods
Study Design and Setting
To determine the rates of NPMODS and mortality among children with cancer admitted to the PICU with SS, and to identify risk factors associated with this poor outcome, we conducted a retrospective cohort study of pediatric cancer patients admitted to the PICU. The study was conducted at Intermountain Primary Children's Hospital (IPCH), a tertiary care children's hospital with 289 beds, and over 2,000 PICU admissions per year. The IPCH oncology team diagnoses and treats ∼200 new cases of pediatric cancer per year and has ∼700 admissions per year to the oncology unit. The University of Utah Institutional Review Board approved the study and waived the requirement for consent because of the retrospective nature of the study and its determination as minimal risk.
Participant Selection
The IPCH oncology team maintains a registry of all patients diagnosed with cancer and cared for at IPCH. From this registry, we included all children less than 18 years of age who had a new cancer diagnosis between January 2008 and December 2012, and who subsequently required hospitalization for suspected or proven sepsis within 3 years of their initial cancer diagnosis (January 2008–December 2015). Patients with suspected or proven sepsis were identified by one of the following two criteria 1 : having a discharge International Classification of Diseases , 9th Revision (ICD-9) diagnosis code for sepsis, SS or septic shock (ICD-9: 995.91, 995.92, and 785.82, respectively), or 2 flagged for suspected sepsis with fever and neutropenia, based on concurrently having an absolute neutrophil count (ANC) < 1,000, a blood culture obtained and/or an antipseudomonal antibiotic started. The ANC cut point of 1,000 was based on our institution's febrile neutropenia protocol. Fever was defined as temperature > 38.5°C. The following agents were considered antipseudomonal antibiotics: cefepime, ceftazidime, meropenem, imipenem, piperacillin/tazobactam, ciprofloxacin, and levofloxacin.
From the list of children with suspected or proven sepsis, we further identified children who required PICU admission and manually reviewed the patient charts to determine if sepsis was the primary indication for the PICU admission using the 2005 Consensus Definitions. 20 All cases that met the 2005 Consensus criteria for SS or septic shock upon PICU admission were defined as cases with SS. For patients with repeated admissions, each encounter was analyzed as a separate case because cancer patients can develop SS at different stages of their disease/treatment and thus the risk of morbidity/mortality may differ.
Data Collection
Demographic and patient characteristic data were extracted from the IPCH oncology registry and the Intermountain Healthcare enterprise data warehouse. The primary investigator conducted a manual chart review of all patients requiring PICU admission. Data elements collected included age on admission, time since diagnosis with cancer, sepsis category upon PICU admit, transferring unit, disease status at the time of admission (i.e., active disease vs. remission), type of central line, history of hematopoietic stem cell transplantation (HSCT), pediatric index of mortality II score upon the PICU admission, PICU resource utilization, source of infection, causative organism, survival to hospital discharge, and daily organ dysfunction measures. Patients were considered in remission if the primary oncologist stated this in the medical charts.
Outcomes
The primary outcome measure was the composite of the development of NPMODS and/or in-hospital mortality. NPMODS was defined as the development of two or more organ dysfunctions after the first day of the PICU admission. Organ dysfunction was defined using the criteria suggested by Proulx et al. 21 Secondary outcomes included PICU resource utilization (requirement for positive pressure ventilation, vasoactive agents, dialysis, and insulin infusion) and PICU length of stay.
Statistical Analysis
Standard descriptive statistics were used to describe the patients' characteristics. We used χ 2 (Fisher's exact test when appropriate) to compare categorical variables, and student t -test (Wilcoxon's rank-sign test when appropriate) to compare continuous variables. Odds ratios (OR) and 95% confidence intervals (CI) were calculated. A p -value <0.05 was considered significant. A logistic regression multivariable model for the development of NPMODS and/or death was constructed. All variables with p < 0.25 in the univariate analysis were included in the model, and least significant variables were eliminated. At each step, variables with marginal p < 0.05 after stepwise logistic regression analysis were kept in the model. The process terminated when no more variables could be eliminated. The statistical software SAS (V9.4) was used for this analysis.
Results
There were 1,002 children diagnosed with cancer at IPCH between January 2008 and December 2012. In the 3 years following their diagnosis, 453 patients (45%) were hospitalized at least once with suspected or proven sepsis, and 77 patients (8%) were admitted to the PICU during 99 separate encounters with SS ( Fig. 1 ). The 1,002 patients had a total of 5,156 hospital encounters, of whom 1,510 were for suspected or proven sepsis.
Fig. 1.
Number of patients hospitalized with proven or at risk of sepsis and admitted to the PICU with severe sepsis. Abbreviations: PICU, pediatric intensive care unit; SS, severe sepsis.
Patient demographics and clinical characteristics of the study patients are outlined in Table 1 . The most common underlying malignancies were leukemia (27%), central nervous system (CNS) tumors (26%), and lymphoma (10%). Demographic features associated with hospital and PICU admission are displayed in Table 2 . Children hospitalized for suspected or proven sepsis were younger at cancer diagnosis compared with those who were not (6.7 years vs. 8.1 years, p < 0.01), but age at cancer diagnosis was not associated with PICU admission for SS (8.4 vs. 7.5 years, p = 0.95). Children with hematopoietic malignancies were more likely to be hospitalized for suspected or proven sepsis (OR: 8.7, 95% CI: 6.4–11.6, p < 0.01), and to be admitted to the PICU for SS (OR: 1.57, 95% CI: 1.02–2.4) compared with children with other malignancies.
Table 1. Demographic and clinical characteristics of the patients.
| Characteristics | n = 1,002 |
|---|---|
| Age at diagnosis (years) | 6 (2–13) |
| Gender | |
| Male | 514 (51) |
| Female | 488 (49) |
| Race | |
| African American | 11 (1) |
| Asian | 10 (1) |
| Native Hawaiian or Pacific Island | 11 (1) |
| White | 890(89) |
| Unknown | 80 (8) |
| Malignancy | |
| Leukemia | 275 (27) |
| CNS | 262 (26) |
| Lymphoma | 103 (10) |
| Soft tissue sarcoma | 75 (7) |
| Bone tumors | 56 (6) |
| Neuroblastoma | 54 (5) |
| Germ cell | 54 (5) |
| Renal tumors | 27 (3) |
| Hepatic tumors | 16 (2) |
| Retinoblastoma | 14 (1) |
| Other | 66 (7) |
| Hospitalized with suspected or proven sepsis | 453 (45) |
| Admitted to the PICU with SS | 77 (8) |
Abbreviations: CNS, central nervous system; PICU, pediatric intensive care unit; SS, severe sepsis.
Note: Data presented as n (%) or median (interquartile range) for each group. Fourteen patients had repeated PICU admissions with SS. Ten patients were admitted twice and four patients were admitted three to five times.
Table 2. Risk factors for hospitalization and PICU admission with SS.
| Risk Factors | Hospitalized with suspected or proven sepsis | Admitted to PICU with SS | ||||
|---|---|---|---|---|---|---|
| Yes n = 453 |
No n = 549 |
p -Value | Yes n = 77 |
No n = 925 |
p -Value | |
| Age at diagnosis | 6.7 (5.3) | 8.1 (6.0) | < 0.001 | 8.4 (6.0) | 7.5 (5.7) | 0.18 |
| Gender | 0.33 | 0.90 | ||||
| Female | 213 (47) | 275 (50) | 38 (49) | 450 (49) | ||
| Male | 240 (53) | 274 (50) | 39 (51) | 475 (51) | ||
| Type of malignancy | < 0.01 | < 0.01 | ||||
| Hematopoietic malignancies | 287 (63) | 91 (17) | 51 (66) | 327 (35) | ||
| Solid tumors | 166 (37) | 458 (83) | 26 (34) | 598 (65) | ||
Abbreviations: PICU, pediatric intensive care unit; SD, standard deviation; SS, severe sepsis.
Note: Data listed as n (%) or mean (SD).
Demographics and clinical characteristics of the 99 SS PICU admissions are highlighted in Table 3 . The 99 PICU admissions (cases) for SS included 77 individual patients ( Fig. 1 ), of which 14 had repeated PICU admission for SS, and 63 had a single SS PICU admission. All 14 patients with repeated admissions had leukemia (9 with acute lymphoblastic leukemia [ALL] and 5 with acute myeloid leukemia [AML]). Median age at PICU admission was 8 years, interquartile range (IQR) 2 to 14 years, and 58% were male. The majority of cases were admitted either from the oncology unit (80%) or the emergency department at IPCH (12%).
Table 3. Clinical characteristics of patients admitted to the PICU.
| Clinical Characteristics | Encounters n = 99 |
|---|---|
| Age at PICU admission (years) | 8 (2–14) |
| Gender | |
| Male | 57 (58) |
| Female | 42 (42) |
| Race | |
| African American | 3 (3) |
| Native Hawaiian or Pacific Island | 1 (1) |
| White | 81 (82) |
| Unknown | 14 (14) |
| Malignancy | |
| Hematopoietic malignancies | 73 (74) |
| Solid tumors | 26 (26) |
| HSCT recipient | |
| Yes | 14 (14) |
| No | 85 (8) |
| Type of central line | |
| Broviac | 51 (52) |
| Port | 31 (31) |
| PICC | 14 (14) |
| None | 3 (3) |
| Source of infection | |
| CLABSI | 40 (40) |
| Pneumonia | 18 (18) |
| Not identified | 17 (17) |
| Skin | 8 (8) |
| Abdominal infection | 7 (7) |
| Others | 9 (9) |
| Causative organism | |
| Bacterial | 50 (50) |
| Gram positive | 22 (22) |
| Gram negative | 28 (28) |
| Fungal | 5 (5) |
| Viral | 4 (4) |
| Not isolated | 40 (40) |
| Transferring unit | |
| Oncology unit | 79 (80) |
| ER | 12 (12) |
| OSH | 6 (6) |
| Clinic | 2 (2) |
Abbreviations: CLABSI, central line associated bloodstream infection; ER, emergency room; HSCT, hematopoietic stem cell transplantation; OSH, outside hospital; PICC, peripherally inserted central catheter; PICU, pediatric intensive care unit.
Note: Data presented as n (%) or median (interquartile range) for each group.
Fourteen of the 99 cases (14%) had a history of HSCT, 11 with allogeneic HSCT, and 3 with autologous HSCT. The underlying malignancy among the HSCT patients was leukemia in nine cases; CNS tumor in two cases; and neuroblastoma, myelodysplastic syndrome, and Langerhans cell histiocytosis in one case each. The median (IQR) time from transplantation to admission with SS was 44 days (25–110). Due to the small number of HSCT cases and the heterogeneity in this subgroup, no analysis was done to compare the outcome between different HSCT subclasses.
Central line-associated bloodstream infection (CLABSI) was the most common identified infection (40%), followed by pneumonia (18%) ( Table 3 ). Patients with an implantable venous access device (e.g., Portacath) for their central venous access were less likely to have CLABSI, compared with those with a tunneled catheter (e.g., Broviac) or a peripherally inserted central catheter line (OR 0.22, 95% CI, 0.01–0.5). The source of infection was not identified in 39 cases (39%) ( Table 3 ). The causative organism was isolated in 60% of the cases, the majority of which were bacteria ( Table 3 ). A list of organisms isolated from patients in the cohort is provided in Supplementary Table S1 (online only).
With respect to resource utilization, 66 (67%) required at least one vasoactive agent, and 37 (37%) required positive pressure ventilation. Norepinephrine was the most commonly used vasoactive agent, followed by dopamine (57 and 23%, respectively). Median PICU length of stay was 2.2 days (IQR: 1–6.5) ( Table 4 ).
Table 4. Resource utilization of patients admitted to the PICU.
| Resources |
n
(99) |
|---|---|
| Vasoactive agents | |
| Yes | 66 (67) |
| o 1 agent | 45 (45 |
| o 2 agents | 12 (12) |
| o > 2 | 11 (11) |
| No | 33 (33) |
| Norepinephrine | 56 (57) |
| Dopamine | 23 (23) |
| Epinephrine | 12 (12) |
| Vasopressin | 6 (6) |
| Mechanical ventilation | |
| IPPV | 30 (30) |
| NIPPV | 26 (26) |
| Either | 37 (37) |
| Insulin infusion | 14 (14) |
| Surgery: | |
| Source control | 7 (7) |
| Line removal | 11 (11) |
| Diagnostic | 3 (3) |
| Dialysis | 9 (9) |
| PICU LOS (days) | 2.2 (1–6.5) |
Abbreviations: IPPV, invasive positive pressure ventilation; LOS, length of stay; NIPPV, noninvasive positive pressure ventilation; PICU, pediatric intensive care unit.
Note: Data presented as n (%) or median (interquartile range) for each group.
The overall in-hospital mortality was 23% ( n = 18). The composite unfavorable outcome of NPMODS and/or death before discharge occurred during 34 encounters (34%). Of these, both NPMODS and subsequent death occurred in 13 encounters, NPMODS with survival to hospital discharge occurred in 16, and death without prior NPMODS occurred in 5. The average maximum number of organ failures was 1.63 ± 1.0. The maximum number of organ failures was significantly higher among patients that died (2.9 ± 1.1 vs. 1.4 ± 0.8, p < 0.01). The frequency of specific organ system dysfunctions and maximum number of organ dysfunction and the associated mortality are shown in Supplementary Tables S2 and S3 (online only), respectively. Nine cases required renal replacement therapy. Of these, all developed NPMODS and five died. The following factors correlated with having the composite unfavorable outcome by univariate analysis ( p < 0.05): History of HSCT, source of infection, and initial lactate level. In multivariate analysis, only history of HSCT and initial lactate level remained significantly associated with unfavorable outcome (OR: 7.7, 95% CI: 1.2–52.2, and OR: 1.6, 95% CI: 1.2–2.2, respectively) ( Table 5 ).
Table 5. Risk factors for the development of NPMODS and/or death.
| Risk Factors | NPMODS and/or death | Univariate analysis | Multivariate analysis | |||
|---|---|---|---|---|---|---|
| Yes n = 34 |
No n = 65 |
p- Value | OR | 95% CI | p- Value | |
| Age at diagnosis | 8.3 (6.6) | 7.8 (5.8) | 0.70 | |||
| Gender (female) | 16 (47) | 26 (40) | 0.49 | – | – | – |
| Race | 0.58 | – | – | |||
| African American | 1 (3) | 2 (3) | ||||
| Native Hawaiian or Pacific Island | 1 (3) | 0 (0) | ||||
| White | 27 (79) | 54 (83) | ||||
| Unknown | 5 (15) | 9 (14) | ||||
| Hematopoietic malignancies | 27 (80) | 46 (70) | 0.35 | |||
| In remission (yes) | 13 (38) | 31 (48) | 0.36 | |||
| HSCT recipient (yes) | 9 (26) | 5 (8) | 0.01 | 7.7 | 1.2–52.2 | 0.036 |
| Central line type | 0.84 | – | – | – | ||
| Broviac | 18 (51) | 33 (51) | ||||
| Port | 9 (26) | 22 (34) | ||||
| PICC | 6 (18) | 8 (12) | ||||
| None | 1 (3) | 2 (3) | ||||
| Source of infection | 0.02 | |||||
| CLABSI | 10 (29) | 30 (46) | ||||
| Pneumonia | 12 (35) | 6 (9) | ||||
| Abdominal infection | 2 (6) | 5 (8) | ||||
| Skin | 3 (9) | 5 (8) | ||||
| Unknown | 3 (9) | 14 (22) | ||||
| Others | 4 (12) | 5 (8) | ||||
| Organism | 0.07 | |||||
| Bacterial | 13 (38) | 38 (58) | ||||
| Negative cx | 15 (44) | 24 (37) | ||||
| Fungal | 3 (9) | 2 (3) | ||||
| Viral | 3 (9) | 1 (2) | ||||
| Transferring unit (OU) | 28 (82) | 51 (78) | 0.64 | |||
| Required surgical intervention (yes) | 8 (24) | 12 (18) | 0.52 | |||
| Initial lactate | 4.18 (2.8) | 2.25 (1.6) | 0.007 | 1.6 | 1.2–2.2 | 0.002 |
| ANC | 1.9 (3.6) | 1.3 (2.5) | 0.43 | |||
| PIM II | 39 (37) | 24 (34) | 0.05 | |||
Abbreviations: ANC, absolute neutrophil count; CLABSI, central line associated bloodstream infection; CI, confidence interval; HSCT, hematopoietic stem cell transplantation; NPMODS, new or progressive multiorgan dysfunction syndrome; OR, odds ratio; OU, oncology unit; PICC, peripherally inserted central catheter; PIM II, pediatric index of mortality, second revision.
Note: Data presented as n (%) or mean ± standard deviation for each group.
Of note, ANC did not differ significantly between patients with unfavorable outcome compared with those with favorable outcome (median [IQR] 100 (0–2,400) vs. 200 [0–1,350], p 0.43). Similarly, there was no correlation between the time from HSCT to admission with SS and having the unfavorable outcome. Additionally, history of previous PICU admission for SS was not associated with increased rates of NPMODS and/or death (36 vs. 39%, p = 0.78). Finally, in the group of patients with identified bacterial infections, gram stain class of the causative bacteria was not associated with increased risk of NPMODS and/or death (gram positive: 32%, gram negative: 25%, p = 0.57).
Discussion
In this study, we assessed the incidence of and risk factors for morbidity and mortality in children with cancer admitted to the PICU with SS. Our results indicate high rates of hospitalization for suspected or proven sepsis among children with cancer. Of all oncology patients, 45% were admitted at least once for concerns of sepsis and 8% required PICU admission for SS. Among those admitted to the PICU for SS, 34% developed NPMODS and/or died. Lactate level on admission to PICU and history of HSCT were associated with increased risk of death and/or NPMODS.
Children with cancer are at significantly higher risk of infection and sepsis due to multiple factors: prolonged bone marrow aplasia and disruption of the mucosal barrier secondary to intensive chemotherapy and radiotherapy, prolonged courses of high-dose steroids for leukemia, and the long-term need for central venous access. More aggressive primary cancer treatment has improved the overall survival of pediatric cancer, but has also increased the number of patients at risk of developing SS and septic shock. 5 22 23 24 We report a high incidence rate of admission for suspected or proven sepsis among this group of patients; suspected or proven sepsis was the reason for admission in 1 out of every 3.4 admissions.
Not surprisingly, the type of malignancy was significantly associated with hospitalization and PICU admission for SS, with hematopoietic cancers carrying higher risk. The inherent immunosuppression from bone marrow infiltration and more intensive chemotherapy regimens have been postulated as explanations for this high risk. 8 22 25 26 27 However, despite increased risk of SS, type of malignancy did not affect PICU outcome. Over the last two decades, survival among children with cancer admitted to the PICU with SS has significantly increased. This has been attributed to the improvement in recognition and treatment of sepsis rather than changes in the immunological status of the oncology patients. 12 14 Similarly, ANC as a marker of immunosuppression did not affect PICU outcome in our cohort. This is consistent with previous studies that found no association between ANC and PICU outcome among children with cancer admitted for SS. 9 14 15 Thus, overall improvement in care of patients with SS might explain why the type of malignancy and the ANC did not affect morbidity and mortality in our study. 6
Mortality in the present study (23%) is slightly higher than what has been reported in the majority of previous investigations of SS in pediatric oncology patients, with rates ranging 12 to 17%. 12 14 15 However, there are important differences between our investigation and these earlier studies. The current patient population is likely different than that of previous studies given the advancement in cancer treatment, improvement in overall pediatric cancer survival rates, and increased population of cancer survivors at risk of treatment complication. 5 15 Additionally, we used diagnostic criteria based on the 2005 International Consensus Conference Definition 20 that may have differed from the sepsis diagnostic criteria used in previous studies.
The Sepsis Prevalence, Outcomes, and Therapies (SPROUT) study, a recent international point prevalence study of SS in PICU patients, included 80 children with malignancy, who experienced a 41% mortality, compared with 25% for the entire SPROUT cohort. 16 The SPROUT study also used the 2005 Consensus Definition for SS. The reported mortality rates from the SPROUT study are much higher than what has been reported from our institution, both from the present study of oncology patients (23%) and from a previous study of all PICU patients (5%). 28 However, the SPROUT study included 128 PICUs from 26 countries. Health care practice and available resources are highly variable among those PICUs and may explain the higher mortality compared with our institution. Additionally, our patient population is less diverse than the SPROUT study ( Table 3 ). Despite these differences, the mortality reported in both the SPROUT study and the present study are higher than what has previously been reported for children with cancer who develop SS. 12 13 14 15 16 It is possible that this is related to the use of the 2005 Consensus Definition for cohort identification, which may be identifying a sicker cohort of children. While it seems that there is still marked variability in the reported mortality rates, these studies demonstrate higher mortality rates from SS among children hospitalized with cancer compared with the general pediatric population.
HSCT recipients have been historically recognized to carry a higher risk of mortality from SS, due to their prolonged period of immunosuppression, risk of graft-versus-host disease, and increased numbers of comorbidities. 26 29 We report higher rates of death and/or NPMODS among the subgroup of patients with a history of HSCT. Our results are consistent with previous reports. 5 6 14 15 A larger study with more HSCT recipients is needed to accurately investigate the rates of death and NPMODS among this subgroup of patients.
We found serum lactate level upon PICU admission to be significantly associated with development of NPMODS and/or death. Serum lactate as a marker of poor tissue perfusion and its utility in recognizing patients at higher risk of progression to septic shock, organ failure, and mortality has been extensively studied. 30 31 The results from the present study support these findings.
Recognizing patients at increased risk of septic shock and worse outcome could prompt an early and more vigilant approach, potentially improving outcome. Similarly, recognizing patients at low risk may limit the frequency and duration of hospitalizations that impact the quality of life and exhaust hospital resources. 32 33 34 Risk factors associated with worse outcome, such as lactate level and history of HSCT, could be utilized to create a predictive model that identifies patients at higher risk at early stages of their illness. A large prospective study is needed to better characterize the predictors of morbidity and mortality among this group of patients.
Our study has several limitations. First, while using the clinical diagnosis of febrile neutropenia provided greater capture of patients at risk of SS who might have otherwise been missed by administrative codes, there are likely some cases that were missed by our identification strategy. We also did not screen the patients who did not require PICU transfer for SS, although it is unlikely for patients with SS to be managed out of the ICU at IPCH. Second, data regarding resuscitation prior to PICU admission was not available and may have influenced the patient's outcome. Third, lactate level was not obtained on all patients. Patients with lactate level obtained on admission might represent a sicker subgroup of patients. Thus, including lactate in our analysis may have introduced selection bias. Nevertheless, in the group of patients for whom the lactate levels are known, the positive association between lactate and poor outcome supports its prognostic power and utility in identifying patients at risk of worse outcome. Finally, although our center follows standard protocols for cancer therapy and provides standard management for suspected infections, results from this single-center cohort may not be generalizable to other pediatric centers.
Conclusion
The development of SS and septic shock is common among children with cancer. Consistent with another recent paper by Weiss et al, we report higher rates of morbidity and mortality among this group of patients than has previously been reported. These higher rates may represent a more accurate understanding of risk for pediatric oncology patients, despite recent improvements in treatment and supportive care. In our cohort of oncology patients admitted to the PICU with SS, 34% developed NPMODS and/or died. Hematopoietic cancer and low ANC did not increase the risk of NPMODS and/or death. Not surprisingly, lactate level upon PICU admission and history of HSCT were associated with poor outcome and could be utilized clinically for identifying patients at higher risk. However, a more effective approach may be the implementation of an early septic shock recognition and treatment program tailored for the inpatient pediatric oncology patient.
Footnotes
Conflict of Interest None declared.
Supplementary Material
References
- 1.National Cancer Institute.Surveillance, Epidemiology, and End Results (SEER) Program Research Data (1973–2014)DCCPS, Surveillance Research Program, Surveillance Systems Branch. 2014. doi:http://www.seer.cancer.gov
- 2.Pritchard-Jones K, Kaatsch P, Steliarova-Foucher E, Stiller C A, Coebergh J W. Cancer in children and adolescents in Europe: developments over 20 years and future challenges. Eur J Cancer. 2006;42(13):2183–2190. doi: 10.1016/j.ejca.2006.06.006. [DOI] [PubMed] [Google Scholar]
- 3.Mertens A C, Yong J, Dietz A C et al. Conditional survival in pediatric malignancies: analysis of data from the Childhood Cancer Survivor Study and the Surveillance, Epidemiology, and End Results Program. Cancer. 2015;121(07):1108–1117. doi: 10.1002/cncr.29170. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Phillips R S, Sutton A J, Riley R D, Chisholm J C, Picton S V, Stewart L A; PICNICC Collaboration.Predicting infectious complications in neutropenic children and young people with cancer (IPD protocol) Syst Rev 201218. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Demaret P, Pettersen G, Hubert P, Teira P, Emeriaud G. The critically-ill pediatric hemato-oncology patient: epidemiology, management, and strategy of transfer to the pediatric intensive care unit. Ann Intensive Care. 2012;2(01):14. doi: 10.1186/2110-5820-2-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Haase R, Lieser U, Kramm C et al. Management of oncology patients admitted to the paediatric intensive care unit of a general children's hospital - a single center analysis. Klin Padiatr. 2011;223(03):142–146. doi: 10.1055/s-0031-1275291. [DOI] [PubMed] [Google Scholar]
- 7.Meyer S, Gottschling S, Biran T et al. Assessing the risk of mortality in paediatric cancer patients admitted to the paediatric intensive care unit: a novel risk score? Eur J Pediatr. 2005;164(09):563–567. doi: 10.1007/s00431-005-1695-y. [DOI] [PubMed] [Google Scholar]
- 8.Zinter M S, DuBois S G, Spicer A, Matthay K, Sapru A. Pediatric cancer type predicts infection rate, need for critical care intervention, and mortality in the pediatric intensive care unit. Intensive Care Med. 2014;40(10):1536–1544. doi: 10.1007/s00134-014-3389-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Singer K, Subbaiah P, Hutchinson R, Odetola F, Shanley T P. Clinical course of sepsis in children with acute leukemia admitted to the pediatric intensive care unit. Pediatr Crit Care Med. 2011;12(06):649–654. doi: 10.1097/PCC.0b013e31821927f1. [DOI] [PubMed] [Google Scholar]
- 10.Russell H V, Okcu M F, Kamdar K et al. Algorithm for analysis of administrative pediatric cancer hospitalization data according to indication for admission. BMC Med Inform Decis Mak. 2014;14:88. doi: 10.1186/1472-6947-14-88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Turkoglu M, Mirza E, Tunçcan O G et al. Acinetobacter baumannii infection in patients with hematologic malignancies in intensive care unit: risk factors and impact on mortality. J Crit Care. 2011;26(05):460–467. doi: 10.1016/j.jcrc.2011.04.007. [DOI] [PubMed] [Google Scholar]
- 12.Pound C M, Johnston D L, Armstrong R, Gaboury I, Menon K. The morbidity and mortality of pediatric oncology patients presenting to the intensive care unit with septic shock. Pediatr Blood Cancer. 2008;51(05):584–588. doi: 10.1002/pbc.21670. [DOI] [PubMed] [Google Scholar]
- 13.Hallahan A R, Shaw P J, Rowell G, O'Connell A, Schell D, Gillis J. Improved outcomes of children with malignancy admitted to a pediatric intensive care unit. Crit Care Med. 2000;28(11):3718–3721. doi: 10.1097/00003246-200011000-00030. [DOI] [PubMed] [Google Scholar]
- 14.Fiser R T, West N K, Bush A J, Sillos E M, Schmidt J E, Tamburro R F. Outcome of severe sepsis in pediatric oncology patients. Pediatr Crit Care Med. 2005;6(05):531–536. doi: 10.1097/01.pcc.0000165560.90814.59. [DOI] [PubMed] [Google Scholar]
- 15.Kutko M C, Calarco M P, Flaherty M B et al. Mortality rates in pediatric septic shock with and without multiple organ system failure. Pediatr Crit Care Med. 2003;4(03):333–337. doi: 10.1097/01.PCC.0000074266.10576.9B. [DOI] [PubMed] [Google Scholar]
- 16.Weiss S L, Fitzgerald J C, Pappachan J et al. Global epidemiology of pediatric severe sepsis: the sepsis prevalence, outcomes, and therapies study. Am J Respir Crit Care Med. 2015;191(10):1147–1157. doi: 10.1164/rccm.201412-2323OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Dervan L A, Watson R S. Incorporating morbidity into PICU quality measures: a “TOPICC” of critical importance. Crit Care Med. 2015;43(08):1781–1782. doi: 10.1097/CCM.0000000000001103. [DOI] [PubMed] [Google Scholar]
- 18.Farris R W, Weiss N S, Zimmerman J J. Functional outcomes in pediatric severe sepsis: further analysis of the researching severe sepsis and organ dysfunction in children: a global perspective trial. Pediatr Crit Care Med. 2013;14(09):835–842. doi: 10.1097/PCC.0b013e3182a551c8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Lin J C, Spinella P C, Fitzgerald J C et al. New or progressive multiple organ dysfunction syndrome in pediatric severe sepsis: a sepsis phenotype with higher morbidity and mortality. Pediatr Crit Care Med. 2017;18(01):8–16. doi: 10.1097/PCC.0000000000000978. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.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 20056012–8. [DOI] [PubMed] [Google Scholar]
- 21.Proulx F, Fayon M, Farrell C A, Lacroix J, Gauthier M. Epidemiology of sepsis and multiple organ dysfunction syndrome in children. Chest. 1996;109(04):1033–1037. doi: 10.1378/chest.109.4.1033. [DOI] [PubMed] [Google Scholar]
- 22.Ammann R A, Laws H J, Schrey D et al. Bloodstream infection in paediatric cancer centres--leukaemia and relapsed malignancies are independent risk factors. Eur J Pediatr. 2015;174(05):675–686. doi: 10.1007/s00431-015-2525-5. [DOI] [PubMed] [Google Scholar]
- 23.Ammann R A, Bodmer N, Hirt A et al. Predicting adverse events in children with fever and chemotherapy-induced neutropenia: the prospective multicenter SPOG 2003 FN study. J Clin Oncol. 2010;28(12):2008–2014. doi: 10.1200/JCO.2009.25.8988. [DOI] [PubMed] [Google Scholar]
- 24.Rosolem M M, Rabello L S, Lisboa T et al. Critically ill patients with cancer and sepsis: clinical course and prognostic factors. J Crit Care. 2012;27(03):301–307. doi: 10.1016/j.jcrc.2011.06.014. [DOI] [PubMed] [Google Scholar]
- 25.Viscoli C, Castagnola E, Giacchino M et al. Bloodstream infections in children with cancer: a multicentre surveillance study of the Italian Association of Paediatric Haematology and Oncology. Supportive Therapy Group-Infectious Diseases Section. Eur J Cancer. 1999;35(05):770–774. doi: 10.1016/s0959-8049(99)00052-0. [DOI] [PubMed] [Google Scholar]
- 26.Tamburro R.Pediatric cancer patients in clinical trials of sepsis: factors that predispose to sepsis and stratify outcome Pediatr Crit Care Med 20056(3, Suppl):S87–S91. [DOI] [PubMed] [Google Scholar]
- 27.Costa PdeO, Atta E H, Silva A R. Predictors of 7- and 30-day mortality in pediatric intensive care unit patients with cancer and hematologic malignancy infected with Gram-negative bacteria. Braz J Infect Dis. 2014;18(06):591–599. doi: 10.1016/j.bjid.2014.05.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Workman J K, Ames S G, Reeder R W et al. Treatment of pediatric septic shock with the surviving sepsis campaign guidelines and PICU patient outcomes. Pediatr Crit Care Med. 2016;17(10):e451–e458. doi: 10.1097/PCC.0000000000000906. [DOI] [PubMed] [Google Scholar]
- 29.Kapoor N. Graft-versus-host disease and immunosuppression. Immunol Res. 2008;41(01):34–44. doi: 10.1007/s12026-007-0026-y. [DOI] [PubMed] [Google Scholar]
- 30.Jat K R, Jhamb U, Gupta V K. Serum lactate levels as the predictor of outcome in pediatric septic shock. Indian J Crit Care Med. 2011;15(02):102–107. doi: 10.4103/0972-5229.83017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Bai Z, Zhu X, Li M et al. Effectiveness of predicting in-hospital mortality in critically ill children by assessing blood lactate levels at admission. BMC Pediatr. 2014;14:83. doi: 10.1186/1471-2431-14-83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Miedema K G, Tissing W J, Abbink F C et al. Risk-adapted approach for fever and neutropenia in paediatric cancer patients--a national multicentre study. Eur J Cancer. 2016;53:16–24. doi: 10.1016/j.ejca.2015.10.065. [DOI] [PubMed] [Google Scholar]
- 33.Hakim H, Flynn P M, Srivastava D K et al. Risk prediction in pediatric cancer patients with fever and neutropenia. Pediatr Infect Dis J. 2010;29(01):53–59. doi: 10.1097/INF.0b013e3181c3f6f0. [DOI] [PubMed] [Google Scholar]
- 34.Baorto E P, Aquino V M, Mullen C A, Buchanan G R, DeBaun M R. Clinical parameters associated with low bacteremia risk in 1100 pediatric oncology patients with fever and neutropenia. Cancer. 2001;92(04):909–913. doi: 10.1002/1097-0142(20010815)92:4<909::aid-cncr1400>3.0.co;2-h. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.