Perioperative Serum Procalcitonin and High-sensitivity C-reactive Protein as Prognostic Markers of Postoperative Outcomes in Surgical Neonates: A Prospective Observational Study

ABSTRACT

Introduction:

Neonatal sepsis is a major cause of morbidity and mortality. Procalcitonin (PCT) and high-sensitivity C-reactive protein (hs-CRP) are increasingly evaluated as early markers of infection. This study assessed the predictive value of PCT and hs-CRP at admission and 24 h postsurgery for adverse outcomes.

Materials and Methods:

A total of 120 operated neonates admitted to our surgical neonatal intensive care unit were prospectively enrolled. Demographic data, clinical parameters, blood investigations, blood culture results, and biomarker levels were recorded at admission and 24 h postoperatively. Note was also made of any surgical or anesthetic events. Outcomes were assessed at discharge or death and classified as favorable (discharge) or adverse (Clavien–Dindo Grade 4 or death). Receiver operating characteristic curves determined sensitivity, specificity, and optimal cutoffs; Spearman’s correlation assessed associations of outcome with biomarkers, total leukocyte count (TLC), and blood culture. Multivariable logistic regression (backward stepwise) was performed to identify independent predictors of adverse outcome, with model fit assessed using the Hosmer–Lemeshow test and Nagelkerke R².

Results:

Of 120 neonates (mean age: 3.5 ± 4.5 days; male = 69, female = 51), 97% had congenital conditions pertaining to the gastrointestinal tract and required either bowel anastomosis or stoma. Culture-proven sepsis was present in 23% of neonates at admission and increased to 39% at 24 h postsurgery (almost 60% rise). Clinically suspected sepsis was observed in 41% of neonates at admission and 57% at 24 h after surgery (about 40% increase). Abnormal TLC was noted in 22.5% at admission, which increased marginally after surgery, while absolute neutrophil count remained comparable across time points. At admission, elevated hs-CRP and PCT were detected in 58% and 54% of neonates, respectively; these incidences increased to 82% and 61.6% at 24 h postsurgery, indicating a 20% postoperative rise in biomarker levels. Adverse outcomes occurred in 59% of neonates, and most of these patients had elevated PCT and hs-CRP (70% and 65%, respectively), which had a correlating trend at admission and 24 h (P < 0.001). Using standard laboratory thresholds, hs-CRP demonstrated a marked increase in sensitivity from 57.7% at admission to 91.5% at 24 h postsurgery (50% increase), while specificity declined marginally from 40.8% to 30.6% for predicting adverse outcomes. In contrast, PCT showed modest improvement in postoperative performance, with sensitivity increasing from 57.7% to 73.2% (35% increase), and specificity remained roughly the same (51.0% to 55.1%). All neonates with intraoperative events suffered poor outcomes, and all of them had high biomarker levels at admission. On multivariable analysis, lower birth weight (odds ratio [OR]: 0.29, 95% confidence interval [CI]: 0.12–0.68, P = 0.004) and blood culture positivity at 24 h (OR: 6.41, 95% CI: 2.18–18.85, P = 0.001) were independent predictors of adverse outcome, while biomarker levels were not independently associated.

Conclusions:

More than half of surgical neonates had elevated biomarkers at admission, indicating a high burden of sepsis at admission. Higher values of biomarkers increasingly suggest the likelihood of intraoperative complications and adverse overall outcomes. Hs-CRP emerged as the more sensitive and cost-effective screening and risk-stratification tool as an adjunct to clinical and microbiological data for the prediction of postoperative sepsis or adverse outcomes.

INTRODUCTION

Sepsis continues to be the leading cause of adverse postoperative outcomes, accounting for nearly 25%–65% of neonatal deaths worldwide.[1] While neonatal sepsis has been extensively studied in medical settings, evidence focusing specifically on surgical neonates remains scarce, particularly in resource-limited regions.[1] Early identification of neonates at risk of sepsis is essential for initiating prompt action and curtailing the neonatal mortality rate, which is an important national health indicator. Delayed diagnosis and transport may lead to neonates harboring sepsis at presentation itself, while exposure to anesthesia and surgical procedures may superadd to it by initiating inflammation. This makes surgical neonates a special subset of neonates altogether, marking a need to study sepsis in them to cover up for major time points of concern. To add to this, surgical neonates with gastrointestinal (GI) are likely to be exposed to gut environments which may be directly breached, leading to contamination, infection, and sepsis.

The American Association of Pediatrics (AAP) defines culture-negative sepsis as those with signs and symptoms of systemic inflammatory response syndrome (SIRS) attributed to a bacterial etiology, in whom no organism is isolated from properly collected blood, cerebrospinal fluid (CSF), or urine cultures.[2] Culture-positive sepsis is easy to diagnose and monitor using total leukocyte count (TLC), absolute neutrophil count (ANC), and blood cultures, and such cases receive appropriate and timely treatment.[3] These hematological markers prove unreliable for the detection of culture-negative sepsis, and hence, there is a role for novel biomarkers in such cases. Both procalcitonin (PCT) and high-sensitivity C-reactive protein (hs-CRP) have shown promising results in medical conditions, with PCT rising early after bacterial insult and hs-CRP peaking later. However, even sterile inflammation may also lead to elevated levels of these biomarkers, and distinguishing noninfective inflammation from sepsis is often challenging due to significant physiological overlap.

The predictive role of biomarkers as indicators of sepsis in surgical neonates remains only partially explored. The operating room environment can precipitate both inflammatory and infective responses as a result of surgical and anesthetic stress, as well as potential environmental contamination.[4] A recent Indian study reported that elevated PCT following esophageal atresia repair correlated with postoperative complications, highlighting its prognostic potential for inflammation and sepsis.[5]

Maintaining a high index of suspicion in surgical neonates based on clinical signs supported by biomarkers is essential for early identification of at-risk patients and timely intervention to improve outcomes. This study was therefore undertaken to assess the predictive role of serum PCT and hs-CRP, at admission and 24 h after surgery, in determining adverse postoperative outcomes among surgical neonates. In addition, the study evaluated associations of these biomarkers with TLC and blood culture results. Further, the aim was also to explore whether combined biomarker panels could improve diagnostic sensitivity and specificity for predicting sepsis and postoperative complications.

MATERIALS AND METHODS

Study design and setting

This prospective observational study was conducted over 36 months in the Department of Pediatric Surgery at our tertiary care center, after approval by the Institutional Ethics Committee (No. F-32, 2539/2024-25/Research Project/MC/Pur/30145-48) and informed parental consent. The study aimed to assess perioperative changes in serum PCT and hs-CRP and their predictive value for postoperative outcomes in surgical neonates.

Sample size

Sample size was calculated using the formula N = Z²pq/d², based on a reported postoperative sepsis incidence of 73.75% and an assumed complication rate of 50%.[6] Using a 95% confidence level and 10% precision, the required sample size was 100; with 10% attrition, the final sample size was 120 neonates.

Study population

A total of 120 neonates admitted to the surgical neonatal intensive care unit (NICU) and undergoing operative intervention were prospectively enrolled. Of these, 97% of neonates had congenital conditions pertaining to the GI tract and required either bowel anastomosis or stoma [Annexure 1]. Unless dictated by clinical condition, most neonates were operated upon within 24–48 h of admission. Neonates with major syndromic anomalies, or those managed conservatively, were excluded. For the purpose of this study, sepsis was studied as per AAP definitions as either blood culture-proven sepsis or clinically suspected sepsis (based on laboratory values and clinical parameters), which may be culture negative.[2]

Annexure 1.

Diagnosis and treatment summary (n=120)

Diagnosis	Treatment procedure	Number of cases, n (%)
GI ANOMALIES
Tracheoesophageal fistula (TEF) and esophageal atresia (EA)/pure EA	TEF ligation with primary esophageal repair	43 (36)
	Esophageal diversion (cervical esophagostomy + feeding gastrostomy)
Anorectal malformations (ARM)/pouch colon	Diversion colostomy	30 (25)
	Anoplasty
Intestinal atresia (jejunal/ileal/jejunoileal/colonic/duodenal)	Ileostomy	17 (14)
	Colostomy
	Resection and anastomosis
	Kimura’s duodenoduodenostomy
Gastroschisis	Silo creation ventral hernia creation	8 (6.67)
Hirschsprung disease	Mapping biopsy with colostomy/ileostomy	4 (3.5)
Intestinal malrotation	Ladd’s procedure	5 (4)
Necrotizing enterocolitis (NEC)	Exploratory laparotomy with stoma creation or repair of perforation	7 (6)
Meconium ileus	Ileostomy (divided/Bishop–Koop)	3 (2.5)
Total		117 (97.5)
OTHERS
Ruptured meningomyelocele (MMC)	Excision and primary repair	1 (1)
Congenital diaphragmatic hernia (CDH)	Primary repair	2 (1.5)
Total		120 (100)

Clinical evaluation and data collection

Each neonate underwent a detailed admission evaluation, including perinatal history, vital parameters, and clinical signs of sepsis such as respiratory distress, poor activity, delayed capillary refill, or hypotension. Demographic, surgical, and intraoperative details were recorded in a structured pro forma. Blood samples were collected at admission and 24 h postsurgery for complete blood count, ANC, serum biomarkers (PCT and hs-CRP), and blood culture. Any major associated anomalies or major surgical or anesthetic events were also noted.

Laboratory investigations

Serum PCT was measured by chemiluminescence immunoassay (>2 ng/mL = elevated) and hs-CRP by immunoturbidimetric assay (>5 mg/L = abnormal). Blood cultures were processed under standard microbiological protocols. Reference ranges were TLC: 5000–20,000 cells/mm³, ANC: 1000–5400 cells/mm³, and capillary refill time <3 s.

Outcome measures

Postoperative outcomes were classified as favorable or adverse. Adverse outcome was defined as the occurrence of postoperative complications and severe morbidity (Clavien–Dindo Grade 4) or death (Clavien–Dindo grade 5).[7] These outcomes were assessed at discharge or death and analyzed collectively to reflect clinically relevant postoperative morbidity, acknowledging heterogeneity in severity and to present the results in the simplest manner.

Statistical analysis

Data were analyzed using SPSS v31 (IBM Corp., Armonk, NY, USA). Continuous variables were expressed as mean ± standard deviation or median (interquartile range [IQR]) and categorical variables as frequencies and percentages. Receiver operating characteristic (ROC) analysis (Youden’s index) determined biomarker cutoffs, sensitivity, and specificity; while Spearman’s correlation assessed associations among biomarkers, TLC, and culture results. Combined biomarker accuracy was evaluated by cross-tabulation, with P < 0.05 considered significant. Multivariable logistic regression analysis was performed to identify independent predictors of adverse outcome (discharge vs. death). A stepwise backward conditional method was applied, including gestational age, birth weight, blood culture status (at admission and 24 h postsurgery), and biomarker levels (PCT and hs-CRP at admission and 24 h postsurgery) as independent variables. The probability for variable removal was set at 0.20. Model fit was assessed using the Hosmer–Lemeshow goodness-of-fit test, and explanatory power was evaluated using Nagelkerke R².

RESULTS

A total of 120 surgical neonates were included in the study, with a mean age of 3.7 ± 4.5 days. Most were term neonates, with a mean birth weight of 2.41 kg and a mean gestational age of 37.7 weeks. The baseline demographic, clinical, and laboratory characteristics of the cohort are summarized in Table 1.

Table 1.

Clinical, demographic, and laboratory profile of surgical neonates (n=120)*

Parameters	Findings (n=120)
Demographic details
Mean age (days)	3.7±4.5
Gender distribution (male: female)	71:49
Mean birth weight (kg)	2.48±0.55
Mean gestational age (weeks)	37.6±2.2
Mean time to surgery after admission (hours)	46.2±20.8
Clinical parameters at admission
Mean heart rate (beats/min)	135.5±14.8
Mean respiratory rate (breaths/min)	40.7±12.5
Mean SpO2 at admission (%)	96.4±4.8
Abnormal anterior fontanelle (depressed/raised) (%)	34.2
Abnormal color (dull or pale) (%)	4.2
Abnormal cry or activity (%)	27.5
Prolonged capillary refill (>3 s) (%)	17.5
Shock at admission (%)	14.2
Biomarker Profile
Median hs-CRP at admission, mg/L (IQR)	8.05 (1.75–35.40)
Median hs-CRP at 24 h postsurgery, mg/L (IQR)	19.5 (7.07–88.47)
Median procalcitonin (PCT) at admission, ng/mL (IQR)	2.25 (0.72–5.75)
Median procalcitonin (PCT) at 24 h postsurgery, ng/mL (IQR)	3.56 (0.94–7.92)
Hematological parameters
Median TLC at admission,/ mm3 (IQR)	12,415 (8500–17,730)
Median TLC at 24 h postsurgery,/mm3 (IQR)	11,250 (7175–15,175)
Abnormal TLC (<5000 or>20,000) (admission/24 h postoperative)	27 (22.5)/34 (28)
Median ANC at admission,/mm3 (IQR)	6710 (3475–9350)
Median ANC at 24 h postsurgery,/mm3 (IQR)	6500 (3000–8500)
Abnormal ANC (<1000 or>5400)	72 (60)/73 (61)
Microbiological findings
Blood culture positive (admission/24 h postoperative)	28 (23)/47 (39)
Outcomes, n (%)
Discharge	49 (40.8)
Severe morbidity/Clavien–Dindo Grade 4	29 (24)
Mortality/Clavien–Dindo Grade 5	42 (35)

*Data expressed as median (IQR) or mean±SD as appropriate.

Both the clinical parameters suggestive of sepsis (temperature, heart rate, respiratory rate, color, cry or activity, and capillary refill time), as well as blood culture positivity, were found about 1.5-folds higher 24 h postoperatively. Gram-negative organisms predominated, although four cases of methicillin-resistant Staphylococcus aureus were identified. Adverse postoperative outcomes occurred in 59% of neonates (29 had severe morbidity/Clavien–Dindo Grade 4 and 42 suffered mortality), among whom 70% had elevated PCT and 65% had elevated hs-CRP levels. Hematological markers mirrored this trend. The median hs-CRP level at admission was 8.05 mg/L (IQR: 1.75–35.40), which increased approximately 2.5-fold to 19.5 mg/L (IQR: 7.07–88.47) at 24 h postsurgery. The median PCT levels also demonstrated a postoperative rise, increasing 1.5-fold from 2.25 ng/mL (IQR: 0.72–5.75) at admission to 3.56 ng/mL (IQR: 0.94–7.92) at 24 h [Table 2].

Table 2.

Variations in parameters at admission and at 24 h after surgery (n=120)

Time point of elevated biomarker levels	Abnormal clinical signs, n (%)	Positive blood cultures, n (%)	Abnormal TLC values (<5000 or>20,000/mm3), n (%)	Elevated hs-CRP (>5 mg/L), n (%)	Elevated PCT (>2 ng/mL), n (%)
At admission	49 (41)	28 (23)	27 (22.5)	70 (58)	65 (54)
24 h after surgery	68 (56.6)	47 (39)	34 (28)	98 (82)	74 (61.6)

Among neonates with clinical evidence of a systemic inflammatory state, elevation of biomarkers was more frequently noted (hs-CRP and PCT elevated in 58% and 54% of cases, respectively), compared with abnormal TLC in 28% and blood culture positivity in 23%.

Intraoperative complications were encountered in a subset of patients, including two cases of intraoperative desaturation, eight episodes of significant hemorrhage (despite normal coagulation profiles), and two iatrogenic bowel injuries. All of these subjects had elevated hs-CRP and PCT levels both at admission as well as 24 h after surgery, while laboratory parameters did not show these changes [Table 3]. None of the neonates in our study cohort died within the first 24 h of surgery.

Table 3.

Parameters of subset of neonates with intraoperative complications (n=12)

Parameters	Findings (n=12)
Abnormal clinical signs (n,%)	11 (92)
Median TLC at admission,/mm≥ (IQR)	16,600 (10,225–18,230)
Median TLC at 24 h postsurgery,/mm3 (IQR)	8200 (5300–10,275)
Abnormal TLC (<5,000 or>20,000) (admission/24 h post-op)	4 (33.3)/5 (41.7)
Blood culture positive (admission/24 h postoperative)	8 (66.7)/9 (75)
Median hs-CRP at admission, mg/L (IQR)	12.0 (2.2–16.4)
Median hs-CRP at 24 h postsurgery, mg/L (IQR)	13.5 (7.07–88.47)
Median procalcitonin (PCT) at admission, ng/mL (IQR)	5.26 (3.02–8.25)
Median procalcitonin (PCT) at 24 h postsurgery, ng/mL (IQR)	6.59 (9.84–38.8)
Elevated PCT (admission/24 h postoperative)	11 (92), 10 (83)
Elevated hs-CRP (admission/24 h postoperative)	7 (58), 11 (92)

Using standard laboratory thresholds, hs-CRP demonstrated increasing sensitivity for predicting poor outcome from 57.7% at admission to 91.5% at 24 h, though specificity declined from 40.8% to 30.6%. PCT showed improved predictive accuracy postoperatively, with sensitivity rising from 57.7% to 73.2% and specificity from 51.0% to 55.1% [Table 4].

Table 4.

Sensitivity and specificity of hs-CRP and procalcitonin (using standard laboratory cutoffs) at admission and 24 hours postsurgery for predicting postoperative complications

Biomarkers, timepoint and laboratory threshold	Favorable outcome (n=49)		Unfavorable outcome (n=71)		Total	Sensitivity (%)	Specificity (%)
	Less than threshold	More than threshold	Less than threshold	More than threshold
hs-CRP at admission (>5mg/L)	20	29	30	41	120	57.7	40.8
hs-CRP at 24 hours postsurgery (>5mg/L)	15	34	6	65	120	91.5	30.6
PCT at admission (>2ng/mL)	25	24	30	41	120	57.7	51.0
PCT at 24 hours postsurgery (>2ng/mL)	27	22	19	52	120	73.2	55.1

For further enhancing performance and to determine optimal biomarker thresholds for predicting adverse outcomes, ROC curve analysis was performed, which indicated that hs-CRP has higher overall sensitivity (85.9% at 24 h after surgery), whereas PCT offers better postoperative specificity (56.3%) [Figure 1].

Figure 1.

Receiver operating characteristic curves and area under the curve comparing diagnostic accuracy of high-sensitivity C-reactive protein and procalcitonin at admission and 24 h postsurgery for predicting adverse outcomes in surgical neonates, with cutoffs derived using Youden’s J statistic. ROC: Receiver operating characteristic, AUC: Area under the curve, CI: Confidence interval, CRP: C-reactive protein, PCT: Procalcitonin

It is worthwhile to mention that ROC analysis for culture positivity demonstrated that hs-CRP demonstrated better discriminatory ability for heightened systemic response, which may be attributed to bloodstream infection, particularly in the preoperative phase [Figure 2]. Any value of hs-CRP above 5.5 mg/L was a threshold to alert to the likelihood of an infective organism to be detected on culture.

Figure 2.

Receiver operating characteristic curves comparing diagnostic performance of high-sensitivity C-reactive protein and procalcitonin at admission and 24 h postsurgery for predicting blood culture positivity in surgical neonates. ROC: Receiver operating characteristic, AUC: Area under the curve, CI: Confidence interval, CRP: C-reactive protein, PCT: Procalcitonin

Spearman’s correlation analysis revealed a strong positive correlation between biomarkers across time points, demonstrating parallel perioperative behavior [Figure 3]. Admission and postoperative PCT levels were significantly correlated (ρ = 0.398, P < 0.001), as were hs-CRP levels (ρ = 0.389, P < 0.001). Admission PCT correlated positively with admission hs-CRP (ρ = 0.481, P < 0.001) and postoperative hs-CRP (ρ = 0.250, P = 0.006). Similarly, postoperative PCT correlated with both admission (ρ = 0.259, P = 0.004) and postoperative hs-CRP (ρ = 0.415, P < 0.001).

Figure 3.

Scatter plots depicting strong positive correlations between high-sensitivity C-reactive protein and procalcitonin levels at (a) admission and (b) 24 h postsurgery, showing synchronous perioperative trends. Hs-CRP: High-sensitivity C-reactive protein

TLC showed a weak but significant association with postoperative PCT (ρ =0.195, P = 0.033), though no meaningful correlation was found with hs-CRP. Blood culture positivity did not correlate significantly with either biomarker. The diagnostic yield of combined biomarker panels was evaluated by integrating hs-CRP, PCT, and TLC at admission and 24 h postsurgery.

At admission, combinations of hs-CRP or PCT with TLC achieved specificities above 93%, while the triple combination (hs-CRP + PCT + TLC) reached 98% specificity, albeit with low sensitivity (<20%), making it a strong confirmatory but poor screening model.

At 24 h postsurgery, the hs-CRP + PCT panel achieved 67.6% sensitivity and 65.3% specificity, which was lower than hs-CRP alone (85.9% sensitivity, similar specificity). This reaffirmed hs-CRP as a more sensitive screening biomarker for predicting early postoperative systemic inflammatory response and adverse outcomes in surgical neonates.

On multivariable logistic regression analysis, lower birth weight was independently associated with higher mortality (odds ratio [OR]: 0.287, 95% confidence interval [CI]: 0.121–0.678, P = 0.004) [Table 5]. Blood culture positivity at 24 h postsurgery was also a strong independent predictor of adverse outcome (OR: 6.41, 95% CI: 2.18–18.85, P = 0.001). Blood culture positivity at admission showed a trend toward association but did not reach statistical significance (P = 0.077). Biomarker levels, including PCT at admission and hs-CRP at 24 h, were not independently associated with outcome after adjustment. The model demonstrated good fit (Hosmer–Lemeshow P = 0.505) and explained 28.2% of the variability in outcome (Nagelkerke R² = 0.282).

Table 5.

Multivariable logistic regression analysis for predictors of adverse outcome in surgical neonates

Variables	Beta	SE	Wald Chi-square	P	OR	95% CI
Birth weight (kg)	−1.250	0.440	8.080	0.004	0.287	0.121–0.678
Blood culture at admission	−1.054	0.597	3.121	0.077	0.348	0.108–1.122
Blood culture 24 hours postsurgery	1.858	0.550	11.398	0.001	6.410	2.180–18.847
PCT at admission	−0.021	0.017	1.677	0.195	0.979	0.948–1.011
hs-CRP 24 hours postsurgery	0.003	0.003	1.335	0.248	1.003	0.998–1.009
Constant	2.954	1.144	6.674	0.010	19.190	-

DISCUSSION

Neonatal sepsis reflects an overlap between infection and systemic inflammation. As per AAP guidelines, it is defined as SIRS with suspected or proven infection, while culture-negative sepsis denotes SIRS without microbiological confirmation. Culture-negative sepsis refers to infants with clinical features of SIRS attributed to a presumed bacterial etiology in whom no organism is isolated from appropriately collected cultures.[2] This suspicion of infection, apart from clinical parameters, is well established.

By noting the hematological parameters such as immature leukocytes and recently developed biomarkers, there is an improved diagnostic approach. The biomarkers commonly used in the past were erythrocyte sedimentation rate and CRP (of late, hs-CRP is considered a better marker). More recently, PCT has joined this league of diagnostic markers. Since sepsis is defined by AAP as including SIRS with proven cultures or culture-negative sepsis; hence, the role of these markers has been perceived to be even more important. The immature immune system, limited metabolic reserve, and invasive procedures make surgical neonates highly susceptible to inflammation and sepsis. Despite advances in perioperative care, postoperative morbidity and mortality remain high.

Data on biomarker performance in surgical neonates remain limited. Our study adds to existing evidence by specifically evaluating perioperative biomarker dynamics using dual timepoint assessment and combined biomarker analysis, thereby providing clinically relevant insights into risk stratification in this unique and understudied population.

Of 120 surgical neonates, 40% had clinical features of SIRS. Culture-proven sepsis was present in about one-fourth, and biomarkers were elevated in half of these. This shows that biomarkers have greater sensitivity in detecting the inflammatory processes, even while infectious agents may not be detectable. As expected, the incidence of SIRS increased 24 hours postoperatively (~60%), accompanied by a rise in culture positivity to 40%. This increase can be attributed to both delayed detection of pre-existent bacterial sepsis and also to the operating room environment which can vehemently add to infection and inflammation. Correspondingly, both these factors may have led to a greater incidence of elevated biomarkers (more so for hs-CRP (82% for hs-CRP vs 60% for PCT).

Furthermore, the incidence of elevated biomarkers at admission was comparable to adverse outcomes postoperatively, and reciprocally, all the neonates with adverse events periperatively had elevated biomarkers to begin with. Thus, elevated biomarkers may point to underlying infection/inflammation that increases the risk of perioperative adverse events and also poor outcome. Our study objectively supports the above two rationales, as elevated biomarker levels showed a higher incidence in patients with poor outcomes. However, a significant incidence of at admission elevated levels may reflect greater underlying illness severity or preexisting systemic inflammatory response, and hence, there is a need for exertion of more care in handling and transport of such neonates and also imparting necessary education to the public health workers for this purpose.

More than half of the surgical neonates in our cohort had elevated biomarker levels at admission, and 40% demonstrated at least one abnormal clinical parameter, strongly suggesting that a substantial proportion were already harboring sepsis at the time of admission. Interestingly, only one-fourth of these pathogens could be detected in blood. Similar trends have been documented in other low- and middle-income settings. Similar trends have been reported in low- and middle-income settings, with sepsis prevalence ranging from 33.8% to 39.5% in Ethiopia, Nepal, and Nigeria.[8,9,10] These findings highlight the consistently high burden of neonatal sepsis and are comparable to our cohort. Notably, biomarker elevation in more than half of our patients suggests greater sensitivity for detecting early systemic inflammatory states compared to conventional parameters.

In our cohort, both biomarkers rose significantly 24 h postsurgery, but hs-CRP exhibited higher sensitivity (91.5%) for predicting postoperative complications, compared with PCT (73.2%). Similar findings were documented by Altunhan et al., who found PCT to be more specific but less sensitive than CRP for early-onset sepsis, reflecting the slower kinetics and longer half-life of CRP.[11] In contrast, Shravya et al. reported that CRP demonstrated the highest combined sensitivity and specificity (65.3% and 85.4%) for predicting neonatal sepsis, a trend echoed in our results, confirming its robustness as a screening biomarker even in surgical neonates.[12] Several large meta-analyses have also supported the diagnostic utility of CRP and PCT, albeit with moderate accuracy. The Cochrane review concluded that while both markers improve diagnostic confidence, neither can singularly confirm sepsis without supportive clinical or microbiological evidence.[13] Our data align with these findings, demonstrating that although PCT and hs-CRP associate well with adverse outcomes, their elevation is an interplay between postoperative inflammation and infection.

Higher biomarker levels were more frequently observed in neonates with positive blood cultures, especially at 24 h after surgery; however, this did not reach statistical significance. Hence, not only care during transportation, but also stringent operating room practices can help curtail adverse outcomes in surgical neonates.

Possibly testing of other body fluids for bacterial agents would have improved the detection rate.[14] Undoubtedly, the practice of taking multiple site swabs upon admission of neonates has to be upheld for early detection of sepsis. Urine culture evaluation can be a routine practice, but there are no specific guidelines to perform urine and CSF routinely in all neonates, especially since lumbar puncture is an invasive procedure.[14,15,16]

A strong positive association was observed between PCT and hs-CRP both at admission and 24 h postsurgery (ρ = 0.481 and ρ = 0.415; P < 0.001), suggesting synchronous activation of the systemic inflammatory cascade. This phenomenon parallels the observations of Eichberger et al., who emphasized that CRP and PCT often exhibit parallel kinetics reflecting generalized inflammation rather than infection alone.[17] Nedeva et al. further explained this overlap through mechanistic insights: PCT expression is driven by bacterial endotoxins and cytokines (interleukin-6 [IL-6] and tumor necrosis factor-alpha), while CRP synthesis is IL-6 dependent, causing near-simultaneous surges during systemic inflammation.[18]

PCT and CRP are inflammatory markers with distinct kinetics. PCT rises rapidly within 2–4 h of a bacterial trigger, peaking early at 12–24 h, and has a ~24-h half-life. Conversely, CRP rises within 4–6 h, but peaks later, usually at 48–72 h.[19,20] Biomarkers are usually elevated at presentation and decline over 3–5 days of treatment.[21]

Our findings also demonstrated that TLC correlated weakly but significantly with postoperative PCT (ρ = 0.195, P = 0.033), consistent with Abdelhamid, who reported that PCT provided higher predictive value over TLC and ANC in culture-proven sepsis.[22] We did not study the immature to mature ratio of leukocytes, which is taken as an important parameter and can be counted as a limitation of our study.[23]

In our cohort, neonates with abnormal bedside clinical parameters at admission and hence early physiological instability were identified to have elevated biomarker values more frequently than traditional parameters, such as abnormal TLC and culture-proven sepsis. This alignment suggests that hs-CRP and PCT may function as more sensitive early indicators of physiological compromise in surgical neonates and may outperform TLC and blood cultures in early postoperative risk stratification for adverse outcomes.[24] Furthermore, elevated clinical markers marginally improved the sepsis detection over biomarkers, which underscores the relevance of clinical parameters, which is still very important, but detection of the pathogen alone may lead to underdiagnosis of sepsis.

Blood cultures are a widely employed tool for the assessment of infection. Blood culture positivity was identified in one-fourth at admission and about 40% at 24 h postsurgery, a range comparable to previous NICU series.[25,26] More than half of the surgical neonates already had elevated biomarker levels at admission, and this frequency rose nearly 2.5-fold after surgery. Pathogen detection rate increased by 1.5-fold in these cases; however, no significant correlation was established between the two. However, this might be because only blood cultures were relied upon. Since reliance on blood culture alone would have missed nearly half of these cases which emphasizes the need to include additional culture sources, particularly CSF and urine, in sepsis evaluation, with CSF culture serving as the gold standard for confirming neonatal sepsis.[27]

ROC analysis demonstrated that hs-CRP had better sensitivity for blood culture positivity at admission (68%) compared with PCT (67.9%. These results mirror the multicenter data from Cantey, which questioned the reliability of CRP in isolation but acknowledged its utility as a trend marker when serially measured.[28]

Interestingly, our study also highlighted that culture-negative sepsis comprised nearly two-thirds of adverse outcome cases, a finding supported by Kurma et al. and Kayange et al., where inflammatory biomarkers remained elevated even in the absence of culture-proven infection.[29,30] This reinforces the concept that biomarker-guided clinical judgment may prove more beneficial than reliance on culture results alone, particularly given the 24–48 h delay and variable sensitivity of blood cultures in neonates.[13] Reciprocally, it may suggest that bacterial nonisolation does not rule out that the sepsis cascade would not have been initiated by bacterial products like endotoxins.

To enhance predictive precision, we analyzed combined biomarker panels. The combination of hs-CRP and PCT at 24 h yielded 67.6% sensitivity and 65.3% specificity, while the addition of TLC increased specificity to 94% but markedly reduced sensitivity. Similar synergistic effects have been reported in multiple studies: Herzlich et al. found that combining PCT with CRP and IL-6 significantly improved specificity for neonatal sepsis, whereas Kumar et al. emphasized that CRP-based combinations are cost-effective for low-resource settings.[31,32] Our findings thus support a stepwise diagnostic algorithm: initial screening with hs-CRP (for its high sensitivity), followed by combined assessment with PCT or TLC for specificity enhancement, which can guide us to appropriate therapy.

Our median PCT levels (admission 2.25 ng/mL; 24 h 3.56 ng/mL) were comparable to those in Altunhan et al. and Jatsho et al., both reporting similar peri-inflammatory increases in PCT among septic neonates.[11,33] Similarly, postoperative hs-CRP elevation observed in our study aligns with Wuni et al., where CRP > 10 mg/L was associated with significantly higher risk of complications.[34] Beaumont et al. underscored the role of early serial CRP/PCT monitoring in optimizing antibiotic stewardship – minimizing overtreatment without compromising safety.[35]

In a broader context, He et al. and Nedeva et al. proposed that the exaggerated inflammatory response, not bacterial burden alone, drives morbidity in neonatal sepsis.[18,36] Our observation of biomarker elevation in neonates experiencing intraoperative hemorrhage or bowel injury supports this theory; noninfective inflammatory triggers can mimic or amplify biomarker responses, complicating postoperative interpretation. In addition, the elevation of biomarker levels suggests that factors other than infection, such as surgical stress or tissue injury, may also contribute to their rise.

Our study highlights that while PCT is a more specific indicator of a heightened systemic inflammatory response, which may be due to infection, overall, hs-CRP remains a more sensitive screening biomarker owing to its high sensitivity, affordability, and availability and can guide early suspicion of sepsis or postoperative deterioration.[37]

Moreover, the low culture positivity rate underscores the importance of integrated biomarker-based sepsis assessment models. Predictive models incorporating PCT, CRP, and clinical parameters, such as those proposed by Shravya et al. and Ali et al., could improve diagnostic accuracy and reduce empiric antibiotic overuse, a critical step toward combating antimicrobial resistance in NICUs.[12,38]

Multivariable analysis demonstrated that biomarker levels were not independent predictors of outcome after adjustment. However, their consistent association with adverse outcomes on univariate analysis suggests that they remain valuable adjuncts for early risk stratification when interpreted alongside clinical and microbiological parameters.

Limitations

This study is limited by its single-center design and inclusion of heterogeneous surgical pathologies, which may have influenced biomarker kinetics and limited generalizability. Evaluation of culture-proven sepsis was done on the basis of blood culture alone instead of a complete panel of urine, CSF, and other swabs. Other inflammatory biomarkers, such as ILs (e.g., IL-6 and IL-8), which have shown promise in early sepsis detection,[39] and detailed hematologic indices like immature to total neutrophil ratio were not assessed due to logistical constraints.[23] Biomarkers were evaluated at only two perioperative time points, restricting assessment of dynamic trends. Finally, adjustment for potential confounders, including gestational age, birth weight, urgency, and type of surgery, was not performed due to sample size limitations.

CONCLUSIONS

Our study underscores the substantial burden of sepsis present at admission in surgical neonates, highlighting the urgent need for strengthened neonatal transport systems and aseptic referral practices in resource-constrained settings. Higher values of biomarkers (especially levels above laboratory thresholds of PCT >2 ng/mL and hs-CRP >5 mg/L) increase the likelihood of intraoperative complications and adverse overall outcomes. hs-CRP, owing to its high sensitivity, affordability, and accessibility, may be useful as a sensitive adjunctive screening tool for risk stratification alongside clinical assessment and microbiological data, though limited specificity mandates caution. Serial trends appear more informative than isolated values. Future studies incorporating standardized sepsis definitions and multivariable models are needed to validate these findings.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

A total grant of Rs. 258,085 was sanctioned for the project by the competent authority, i.e., Secretary, Health and Family Welfare, Government of NCT of Delhi, vide order no. F.1 (360)/MC/Plg./2023-24/CD no. 112738015 dated September 20, 2024.