Validity of Procalcitonin as a Single Biomarker for Predicting Mortality in Pediatric Patients after Acute Traumatic or Post-Surgical Infections: A Retrospective Study


 Background Procalcitonin (PCT) levels are elevated in pediatric patients with systemic inflammation. This study was aimed to evaluate the use of Procalcitonin as a single biomarker in the prediction of mortality in pediatric patients after acute traumatic events or post-surgical infection.MethodsThis retrospective longitudinal study involved pediatric patients aged more than two weeks who were admitted to Pediatric Intensive Care Unit (PICU) of King Fahd Hospital of the University from January 2017 to December 2019 due to post-surgical infection. The data includes age, gender, comorbidity, severity of trauma, Pediatric Risk of Mortality Score (PRISM III) score in 24h of admission, PCT level at 24h and 48-72h, day-5 admission, blood culture, PICU stay, multi-organ failure, number of inotropes and intubation requirement were retrieved.ResultsThe mortalities after surgical infection was of 15.9%. Significantly, the elevated mean levels including PRISM III score, PCT level at 24h, 48-72h and day-5 admission were found in patients who were expired during their hospital stay. The PCT level at 48-72h was 0.89 (best cutoff point >1.35) revealed as highly predictive validity of PCT compared to PCT level at 24h which was 0.76 (best cutoff point >0.15) respectively.ConclusionThe serum PCT level as a single laboratory biomarker is remarkably useful in the prediction of mortality after surgical infection for commencing of an earliest treatment plan as a lifesaving attempt. A blended approach based on the PCT level with other clinical biomarkers can be developed for monitoring a treatment given.


Background
Incidence of post-surgical systematic in ammatory response syndrome (SIRS) in pediatric age group is persistently increase in spite of a continuous advancement in the sterilization of surgical equipment and environmental hygiene to perform invasive diagnostic and therapeutic procedures in high-resource countries like Saudi Arabia [1,2]. In Intensive Care Unit (ICU), availability of an instant, accurate, and speci c laboratory tests regardless of cost or invasiveness are extremely desired by clinicians in identifying infected patients when some speci c gold standard is controversial, also deprives bene t from prompt empirical antibiotic therapy [3][4][5].
Use of biomarkers and predictive scoring is a novel approach to assess infection, make diagnosis, estimate disease severity and forecast of patient's outcome. At present, C-reactive protein (CRP) and procalcitonin (PCT) levels are widely used biomarkers in the routine clinical practice. Predictive mortality using PCT level at a cut-off value of 6.38 ng/dl has a sensitivity and speci city 81.8% and 80.8% respectively, ROC area = 0.838, was the highest as compared to CRP and leucocyte [6], but predictive e ciency of PCT in detecting acute bacterial infection and severity of outcome in terms of sensitivity and speci city on speci c cutoff values are still to be standardized. Another study consisting of 105 neurotrauma patients [7] reported prediction of mortality based on procalcitonin using area under ROC on admission was 0.76 (p < 0.05) and found it comparatively better than CRP area of ROC = 0.73 (p < 0.05).
Overall rate of sepsis in ICU admitted patients in hospital-based study in Saudi Arabia has described 16% and of SIRS 8.1%, whereas 40.3% mortality rate associated with sepsis was reported [8]. However, there is no single study reported about the predictive e cacy of PCT level in post-surgical sepsis and associated mortality in pediatric patients, except the recent study in the same set up in Saudi Arabia relevant to some extent where ve years' experience of predictive mortality using procalcitonin level biomarker based on 400 patients admitted in PICU on mechanical ventilator was reported [9].
The present study was a step ahead to explore the use of procalcitonin as a biomarker in the prediction of mortality in pediatric patients after acute traumatic or post-surgical infection. An increased predictive value of PCT in the estimation of in-hospital outcome based on post-surgical events, data may provide a su cient evidence to adopt procalcitonin biomarker as baseline standard procedure.

Methods
This was a retrospective longitudinal study based on a cohort of critically ill pediatric patients from our previous study participants [9] admitted in PICU of King Fahd Hospital of the University (KFHU) from January 2017 to December 2019 to evaluate predictive validity of biomarkers in prediction of in-hospital outcome in terms of survival or mortality. It includes pediatric patients older than two weeks to 14 years of age admitted in PICU due to signi cant trauma or post-acute surgical events having at least two of the following four criteria, one of which must be either abnormal temperature or leukocytes count: 1) Temperature of > 38.5°C or < 36°C, 2) Tachycardia, de ned as a mean heart rate > 2 Standard Deviation (SD) above the normal for age, or bradycardia, de ned as a mean heart rate < 10th percentile for age, 3) Mean respiratory rate > 2 SD above the normal for age, or mechanical ventilation for an acute insult that is not related to an underlying neuromuscular disease or the administration of general anesthesia, 4) Elevated or depressed Leukocytes count for age (excluding chemotherapy-induced leukopenia) or > 10% immature neutrophils.
Patients admitted for < 24 h, who died within the rst 24 h, those who experienced cardiac arrest before PICU admission, those admitted to the PICU due to burn or chemical trauma, and those with incomplete or missing data were excluded from this study. All clinical and laboratory data were registered. The type and severity of trauma was classi ed according to the Injury Severity Score (ISS). Predictive validity of biomarkers in the estimation of in-hospital outcome were compared. These data including age, gender, comorbidity, severity of trauma based on the ISS and Pediatric Risk of Mortality Score (PRISM) III score at requirement were retrieved as per required outcomes.
Data analysis was performed through SPSS-20.0 (IBM product, Chicago-USA). All categorical data were presented as frequencies and percentages. Chi-square test or Fisher's exact test was used for comparison of proportions in relation to expired versus survived outcomes. Numeric data based on age, weight, PICU stay, mechanical ventilation stay and levels of biomarkers were presented as Mean ± SD. Unpaired t-test in case of data distributed normally or non-parametric Wilcoxon Mann-Whitney U-test for non-Gaussian distribution were applied for comparison of numeric variables between expired versus survived outcomes.
ROC analysis was performed to estimate predictive validity of PRISM-III and PCT level at 24hr 48/72 hour and 5 days of admission. Kaplan Meier's survival analysis was performed to show probability of survival in relation to length of PICU stay and mechanical ventilator. P-value of ≤ 0.05 was considered statistically signi cant.

Results
Total of 245 pediatric patients ful lling inclusion criteria were reviewed, having the mean age of 5.27 ± 4.20 years. Male preponderance in the study (58.8%).
Sepsis was seen in 76 (31.0%) patients. Total 39 (15.9%) mortalities after surgical infection took place, out of them 22/39 (56.4%) were diagnosed with sepsis. In-hospital outcome of the pediatric patient was non-signi cant with regards to gender (p = 0.461), need of mechanical ventilator (p = 0.371), but inhospital outcome was highly signi cant associated with type of etiology, blood culture positivity and comorbidity (p < 0.001) as presented in Table 1.
Signi cantly elevated mean levels of biomarkers including PRISM III score, PCT at 24 h, PCT 48-72 h and PCT at day-5 were found in patients who were expired during the hospital stay (p < 0.001). There were also signi cantly increased mean length of PICU stay, ventilation stay and number of inotropic drugs in expired than survived group (p < 0.001), except mean age and weight were signi cantly lesser in expired group (p < 0.05) as detailed in Table 2.

Discussion
Total 39 (15.9%) mortalities after surgical infection took place, out of them 22/39 (56.4%) were diagnosed with sepsis, which is comparatively higher than that reported in local study [8], also higher than mortality rates reported in a systematic review [10], 44% mortality rate was the highest [11]. Turkish study reported highest predictive validity of procalcitonin (PCT) than C-reactive protein (CRP) and leukocyte in prediction of mortality in critically ill patients in PICU, reported area under ROC was 0.838 (0.711-0.966), sensitivity and speci city at best cutoff point 6.38 ng/ml were 81.8% and 80.8% respectively [6], which was consistent with our ndings of PCT at 48/72 h the sensitivity and speci city were respectively 82.1% and 82.0%, even AUC = 0.89 (95% CI: 0.85-0.93) in our study was higher.
However, sensitivity and speci city of PCT level at 24 h were 66.7%, 74.3% in this study were comparatively low as well as the area under ROC = 0.76 (95% CI: 0.66-0.85), the best cutoff point > 0.15. Likewise results were reported in a study based on neuro-traumatic adult patients done by Deng et al [7], areas under the ROC at admission for the prediction of mortality was 0.76. The ndings were also remarkably higher than that reported by Masson et al [12], who compared PCT with Presepsin for mortality prediction in sepsis and exhibited presepsin as better prognostic biomarker than PCT in the range of Sequential Organ Failure Assessment score (area under the curve (AUC) from 0.64 to 0.75 vs. AUC 0.53 to 0.65). Another study [13] on adult patients in ICU following multiple trauma, initial PCT yielded AUC = 0.805.
This study exhibited signi cantly elevated mean levels of biomarkers including PRISM III level, PCT at 24 h, PCT 48-72 h and PCT at day-5 of those patients with poor outcome (p < 0.001), which is consistent with the ndings of all studies included in the review article [14][15][16]. After pediatric liver transplant (PLT) observed serial PCT measurements during the rst week was not useful to identify patients with bacterial infections, but they proposed that serum PCT may be useful after the rst week post PLT [17]. The evidence support that procalcitonin can provide invaluable information when viewed as one piece of a clinical puzzle, and is most powerful when the interpreting physician is aware of how values are in uenced by the different clinical scenarios presented in this article [5].
This study is a novel study that demonstrated predictive validity of PCT level as a single and instant approach to estimate in-hospital outcome of postsurgical trauma patients who likely to suffer sepsis and other bacterial infections. However, there is no single likewise study in literature to compare in the use of PCT level in PICU patients and best cutoff point to project in-hospital outcome. Our study also emerged an additional feature related to average survival time of patients as 25 days in ICU with a 90% probability of mortality within rst 15 days, whereas average survival time was 17 days on mechanical ventilator and 95% probability of mortality in rst 7-8 day. Sole-Rebalta et al described their limitation as these were only three deaths and could not be strati ed before multiple organ failure to compare predictive validity related to outcome, however they reported an average stay in the PICU about 5 days and for the total length of hospitalization was 14.5 days [18].
Despite having various cutoff points of PCT level on various interval, there is no even a single study to compare and x it as typical cutoff to be used as goldstandard biomarker for estimation of severity of in-hospital outcome. Another limitation of the study is retrospective design of study which does not con rm PCT level on a certain point was used to label severity of in-hospital outcome, rather a prospective cross-sectional study can con rm if based on elevated PCT level to some extent really indicated poor outcome and provided effect of an instant management plan based on single non-invasive serum biomarker.

Conclusions
Serum procalcitonin level as a single laboratory biomarker is remarkably useful in prediction of in-hospital outcome of pediatric patients after surgical infection for commencement of earliest treatment plan as lifesaving attempt. In a longer duration of hospital stay, a blended approach based on PCT level with other clinical biomarkers can be developed for monitoring of given treatment as well as switching to more effective remedies. The role of PCT level can be further explored in prospective cross sectional studies restricted to homogeneous post-surgical trauma category.

Declarations
Ethics approval and consent to participate: The author con rms this study was in accordance to institutional guidelines and regulation and con rms that informed consent was obtained from all participants and/or their legal guardians. The ethical approval for this study was obtained from the Institutional review board (IRB) of Imam Abdulrahman Bin Faisal University with an IRB number 2021-01-078 Consent for publication: Not applicable Availability of data and materials' statement: The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Competing Interests: The author declares no competing interests.
Funding: This research did not receive any speci c grant from funding agencies in the public, commercial, or not-for-pro t sectors.
Author's contributions: The author solely made the study conception and design, data acquisition, analysis and data interpretation, drafting of the manuscript and critical revision. The mean survival time in PICU was 24.6 (95% CI: 22.8-26.3) days, highest probability of mortalities (90% or above) within rst 15 days The mean survival time on mechanical ventilator was 16.8 (95% CI: 15.4-18.2) days, highest probability of mortalities (95% or above) within rst 7-8 days