Study design
We conducted a prospective, case-control study with patients younger than 18 years with SCD from June 2015 to June 2018. Children with SCD and fever were eligible to participate in the study as cases and asymptomatic steady-state SCD children were included as controls. Exclusion criteria of cases and controls included age older than 18 years, hematopoietic stem cell transplantation, incomplete diagnostic tests and patients whose parents or legal guardians did not sign the informed consent form for the study. Controls were also excluded if they had had a febrile episode in the previous month. The study received approval from the Institutional Review Board.
Study setting
This study was conducted at the Hospital General Universitario Gregorio Marañón (HGUGM) in Madrid, Spain, a tertiary hospital with 120 pediatric beds and around 7000 pediatric inpatient admissions per year. HGUGM is a national reference center for children with SCD in Spain, with about 300 SCD pediatric patients upon follow-up.
Study protocol and definitions
Patients included in the study as cases were those with an axillary temperature of 38ºC or more, documented at home or within the first 24 hours of hospitalization. Controls were asymptomatic steady-state SCD patients recruited during their follow-up attendance in the Pediatric Hematology clinic. Patients recruited as cases could also be included in the study as controls if at least one month had passed since their last fever episode. The following samples were collected in all cases: blood tests (including complete blood count, biochemistry, CRP and procalcitonin), blood cultures and nasopharyngeal samples for viral detection by a multiplex polymerase chain reaction (multiplex-PCR) assay [including influenza A, B and C, respiratory syncytial virus (RSV), human metapneumovirus, adenovirus, human bocavirus, rhinovirus, human parainfluenza virus and coronavirus]. Other diagnostic tests and the patients’ management were performed under the criteria of the treating physicians. In all cases and controls a blood sample was collected and stored at the HGUGM Biobank for subsequent cytokine analysis. This analysis was performed by the DIAplex Human Th1 / Th2 / Inflammation kit (bioNova científica S. L., Spain), a multiplexed fluorescent bead-based immunoassay for the quantification of multiple human cytokines in serum and culture supernatants by flow cytometry, including the following: interleukin (IL)-1β, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12p70, IL-17a, interferon-γ (IFN-γ) and tumor necrosis factor- α (TNF-α). The limit of detection for the different cytokines was: IL-1β 3.5 pg/mL, IL-2 12.4 pg/mL, IL-4 4.3 pg/mL, IL-6 1.4 pg/mL, IL-8 1.3 pg/mL, IL-10 1.7 pg/mL, IL-12p70 3.4 pg/mL, IL-17a 8.7 pg/mL, TNFα 9.8 pg/mL and IFN-γ 0.8 pg/mL. For statistical analysis, when a single value was non-detectable, it was considered to be half the limit of detection of that cytokine.
SBI was defined as bacteremia, meningitis, pneumonia, osteomyelitis, urinary tract infection (UTI) or any other severe infection with the identification of a plausible microorganism in a normally sterile site (only confirmed SBI were included in this group). For the diagnosis of UTI, pyuria in urine analysis was required in addition to the positive urine culture. Viral infection (VI) was defined as a positive result in the multiplex-PCR from respiratory samples. Patients with a bacterial infection and a viral detection in the respiratory sample at the same time were classified in the group of bacterial-viral coinfections and excluded from the subanalyses. Patients with no proven infection (NPI) were those not included in the previous groups.
Variables evaluated
Data were collected in a database specifically designed for the study. Data related to cases and controls included baseline characteristics such as age, gender, SCD type, diagnostic method, parents’ continent of origin, immunization status (including vaccination against Haemophilus influenzae type b, Streptococcus pneumoniae and Neisseria meningitidis serogroups A, B, C, W, Y), penicillin prophylaxis, treatment with hydroxyurea, other treatments, long-term central venous catheter (CVC), splenectomy and history of previous hospital admissions. Clinical data during the episode (only in cases) included duration of fever before admission, maximum axillary temperature during the episode, respiratory symptoms, hemodynamic instability (defined as hypotension or persistent tachycardia requiring fluid resuscitation or inotropic drugs) and hypoxemia (< 92%). Laboratory parameters included hemoglobin, WBC, neutrophils and platelets on admission, initial and maximum CRP and procalcitonin during the episode, blood cultures, as well as other cultures and viral tests. Other recorded data were inpatient/outpatient management, antimicrobial treatment, need for antibiotic change because of unfavorable clinical course, diagnosis of VOC or acute chest syndrome (ACS), duration of fever and of hospitalization, pediatric intensive care unit (PICU) admission and final outcome (hospital discharge or exitus). Cytokine levels in all cases and controls were also evaluated.
Statistical analysis
Qualitative variables were expressed as frequencies or rates. Quantitative variables were reported as means (standard deviation) or medians [interquartile range (IQR)], according to their distribution. The X2 test was used to compare categorical variables; Yates correction was applied to any contingency table and Fisher’s exact test was used when at least one cell had a value <5. T-test and ANOVA were used to compare the means of normally distributed variables whereas Mann-Whitney U test and Kruskal-Wallis test were performed to compare the medians of variables not normally distributed. Receiver operating characteristics (ROC) curves and area under the ROC curve (AUC) were calculated for significant parameters. The optimal cut-off point of each biomarker was selected using the Youden Index, based on the maximum sum of sensitivity and specificity. Sensitivity, specificity, positive predictive values (VPP) and negative predictive values (NPV) were calculated according to various possible prevalence rates (PR) of SBI. A p value <0.05 was considered significant. Statistical analysis was performed using IBM SPSS Statistics software version 22.0 (IBM Corp., Armonk, N.Y., USA).