Shock remains a major cause of morbidity and mortality in critically ill children worldwide, accounting for up to 20–30% of pediatric intensive care unit (PICU) admissions and contributing significantly to adverse outcomes [1,2]. Early recognition and prompt management are crucial to improve survival. However, conventional clinical and laboratory parameters used to detect shock are often unreliable, subjective, or become abnormal only at later stages of circulatory compromise. Capillary refill time (CRT), one of the most widely used bedside assessments, is subjective and influenced by environmental temperature, lighting, and observer variability [3] . Blood pressure, another key parameter, is a late marker in pediatric shock because children are able to maintain normotension until advanced stages through compensatory mechanisms [4]. Serum lactate, although a useful biochemical marker, requires invasive sampling and laboratory facilities, which may not always be feasible in resource-limited settings[5]. Hence, there is a need for an objective, continuous, and non-invasive tool that can detect early perfusion abnormalities before overt clinical deterioration occurs. The perfusion index (PI), derived from pulse oximetry, is an emerging parameter that reflects the ratio of pulsatile to non-pulsatile blood flow and thus provides a quantitative assessment of peripheral perfusion [6]. PI can be continuously monitored using the same pulse oximetry probe routinely applied in the PICU. It has the advantages of being simple, non-invasive, real-time, and cost-effective. The physiological rationale is that during circulatory compromise, such as in shock, peripheral vasoconstriction occurs as part of compensatory redistribution of blood flow, leading to reduced pulsatile signal and consequently a lower PI value [7]. Studies in adults and neonates have shown that PI correlates with tissue perfusion, hemodynamic instability, and clinical outcomes. Lima and Bakker demonstrated that low PI values were strongly associated with poor perfusion and adverse outcomes in adult intensive care patients [6]. De Felice et al. reported that PI values less than 0.7 predicted mortality in neonates with high sensitivity [8]. In pediatric populations, Bazaraa et al. found that PI correlated with lactate levels and could detect impaired perfusion in critically ill children [9] . Similarly, Ullah et al. validated age-specific cut-offs for PI in neonates and children, highlighting its clinical applicability [10] Despite these promising findings, there is limited data on the use of PI as a predictor of shock in critically ill children in the Indian context. Pediatric shock presentations vary with infectious burden, nutritional status, and healthcare access, and hence validation in this setting is essential. Moreover, while PI has been studied in relation to mortality and general severity of illness, focused evaluation of its utility in predicting shock specifically is relatively scarce [11]. Rationale: Given the high burden of pediatric shock, the limitations of conventional clinical markers, and the need for a reliable, continuous, and non-invasive monitoring tool, this study was designed to evaluate PI as a predictor of shock in critically ill children. Aims and Objective: To assess the predictive value of perfusion index in identifying shock among critically ill children admitted to a tertiary care PICU.

Study Design and Setting This was a prospective observational study conducted in the Pediatric Intensive Care Unit (PICU) of Private Medical College,Bangalore,India, over a two-year period. The PICU is a tertiary care referral unit catering to critically ill children from urban and rural areas of Eastern Bangalore. Participants A total of 120 critically ill children aged 1 month to 18 years who were admitted to the PICU during the study period were included. • Inclusion criteria: All children admitted to the PICU who required continuous monitoring were eligible. • Exclusion criteria: Children with congenital heart disease, peripheral vascular disorders, or those who left against medical advice (DAMA) were excluded, as these conditions could confound perfusion index (PI) values. Data Collection At admission (0 hr), demographic and clinical details such as age, sex, primary diagnosis, comorbidities, and presence of shock were documented. Perfusion index (PI) was measured using a standardized Masimo™ pulse oximeter probe applied to the index finger or toe, ensuring proper positioning and minimal motion artifacts. PI was automatically displayed by the device as the ratio of pulsatile to non-pulsatile blood flow. PI measurements were repeated at 6 hr after admission. The average of three stable readings taken over one minute was recorded to minimize variability. Definition of shock: Shock was defined clinically as per PICU protocol—presence of tachycardia, cold peripheries, prolonged capillary refill time, poor pulse volume, hypotension for age, or need for fluid resuscitation/inotropes (Dellinger et al., 2013; Carcillo & Davis, 2017). Statistical Analysis Data were analyzed using SPSS version 22. Continuous variables were expressed as mean ± standard deviation (SD). Categorical variables were expressed as frequencies and percentages. Independent samples t-test was used to compare mean PI between shock and non-shock groups. Receiver Operating Characteristic (ROC) curve analysis was performed to evaluate the diagnostic accuracy of PI for predicting shock at 0 hr and 6 hr, and to determine optimal cut-off values. A p-value <0.05 was considered statistically significant. Ethical clearance was obtained from the Institutional Ethics Committee prior to commencement of the study. Written informed consent was obtained from parents/guardians.

A total of 120 critically ill children were included. Of these, 70 (58.3%) were male and 50 (41.7%) female. The mean age was 6.4 ± 4.1 years, with the largest proportion (53%) belonging to the 1–12 years group. The most common admitting diagnoses were lower respiratory tract infections (37.5%), sepsis (15%), dengue fever (12.5%), and central nervous system (CNS) infections (10%). At admission, 37 children (30.8%) presented with clinical features of shock, while 83 (69.2%) did not. Table 1. Demographic and Clinical Characteristics of Study Population Variable Total (N=120) Shock (n=37) No Shock (n=83) Sex Male 70 (58.3%) 21 (56.8%) 49 (59.0%) Female 50 (41.7%) 16 (43.2%) 34 (41.0%) Age Group <1 year 18 (15%) 6 (16.2%) 12 (14.5%) 1–12 years 64 (53%) 21 (56.8%) 43 (51.8%) >12 years 38 (31.7%) 10 (27.0%) 28 (33.7%) Primary Diagnosis LRTI/Pneumonia 45 (37.5%) 12 (32.4%) 33 (39.8%) Sepsis 18 (15%) 8 (21.6%) 10 (12.0%) Dengue Fever 15 (12.5%) 7 (18.9%) 8 (9.6%) CNS infections 12 (10%) 4 (10.8%) 8 (9.6%) Others 30 (25%) 6 (16.2%) 24 (28.9%) Perfusion Index and Shock Status at Admission Mean PI at admission was 0.87 ± 0.32 in the shock group and 1.52 ± 0.47 in the non-shock group, showing a highly significant difference (p<0.001). Table 2. Association of PI with Shock at Admission (0 hr) Shock status Mean PI ± SD p-value Shock (n=37) 0.87 ± 0.32 <0.001 No shock (n=83) 1.52 ± 0.47 Perfusion Index and Shock Status at 6 Hours At 6 hr, 29 children remained in shock, while 91 were without shock. Mean PI remained significantly lower in those with shock (0.92 ± 0.38) compared to non-shock patients (1.66 ± 0.55, p<0.001). Diagnostic Accuracy of PI for Predicting Shock ROC analysis demonstrated excellent predictive performance of PI at both admission and 6 hr. • At 0 hr, the Area Under Curve (AUC) was 0.876 (95% CI: 0.789–0.962). An optimal cut-off of PI ≤1.1 gave sensitivity of 84% and specificity of 81%. • At 6 hr, AUC was 0.892 (95% CI: 0.801–0.967). The optimal cut-off of PI ≤1.2 yielded sensitivity of 87% and specificity of 83%. Table 3. Association of PI with Shock at 6 hr Shock status Mean PI ± SD p-value Shock (n=29) 0.92 ± 0.38 <0.001 No shock (n=91) 1.66 ± 0.55 Table 4. Diagnostic Accuracy of PI for Shock (ROC Analysis) Time point AUC (95% CI) Optimal cut-off Sensitivity (%) Specificity (%) 0 hr 0.876 (0.789–0.962) ≤1.1 84 81 6 hr 0.892 (0.801–0.967) ≤1.2 87 83 Figure 1. ROC Curves of PI at Admission and 6 hr

In this prospective observational study of 120 critically ill children, we found that the perfusion index (PI) was significantly lower in children with shock compared to those without shock, both at admission and at 6 hours. Moreover, PI demonstrated excellent predictive accuracy, with an AUC of 0.876 at admission and 0.892 at 6 hours, reinforcing its role as a reliable non-invasive tool for early identification and monitoring of shock. Our findings are consistent with prior studies that have highlighted PI as a surrogate marker of peripheral perfusion and circulatory status. Lima et al. first described the correlation between peripheral perfusion and outcomes in critically ill patients, noting that poor peripheral perfusion was associated with adverse outcomes [6] . Later, De Felice et al. showed that PI measured from pulse oximetry could reliably assess perfusion in neonates, especially during hemodynamic compromise [8]. In pediatric populations, PI has been studied in sepsis, dehydration, and perioperative settings. Ullah et al. reported that PI was significantly reduced in neonates with septic shock, and cut-off values ≤1.1 yielded good sensitivity and specificity, which aligns closely with our cut-off thresholds of 1.1 (admission) and 1.2 (6 hr) [10]. Similarly, Dumas et al. highlighted that PI trends correlated with fluid responsiveness and could serve as an early marker of hemodynamic instability [12]. Our results also resonate with studies in emergency and critical care settings, where PI has been proposed as an adjunct to capillary refill time and lactate measurement. Studies by Darnall and Hakanen et al. emphasized that PI is not only rapid and non-invasive but also provides continuous feedback, which is particularly advantageous in resource-constrained environments such as Indian PICUs [13,7] The high sensitivity and specificity demonstrated in our study underscore the clinical utility of PI as a screening and monitoring tool for shock in critically ill children. In busy pediatric intensive care settings, PI can complement traditional clinical assessment, helping clinicians identify shock earlier than with conventional signs alone. Unlike blood pressure, which may remain preserved in compensated shock, PI detects subtle reductions in peripheral perfusion, making it especially valuable in the “golden hour” of resuscitation [4] Furthermore, PI values tracked over time provided additional insights into treatment response. Children whose PI improved after fluid resuscitation and inotropic support showed parallel clinical recovery, suggesting that PI can be used for dynamic monitoring of resuscitation effectiveness. This reinforces the role of PI not just as a diagnostic marker, but also as a prognostic tool.

Perfusion index (PI), a simple and non-invasive parameter derived from pulse oximetry, demonstrated excellent accuracy in identifying and predicting shock in critically ill children. PI values ≤1.1 at admission and ≤1.2 at 6 hours were strongly associated with shock, with both high sensitivity and specificity. Importantly, serial PI measurements reflected clinical improvement following resuscitation, highlighting its role in dynamic monitoring. Incorporating PI into pediatric intensive care assessment may enhance early detection of shock, guide timely interventions, and improve outcomes, particularly in resource-constrained settings where rapid, reliable, and low-cost tools are urgently needed.

1. Carcillo JA, Davis AL. Pediatric shock: Definitions and recognition. Pediatr Crit Care Med. 2017;18(2S Suppl 1):S17–23. https://doi.org/10.1097/PCC.0000000000001056 2. Ranjit S, Kissoon N. Prompt recognition and aggressive early management of septic shock in resource-limited settings: What is new? Indian J Pediatr. 2013;80(3):222–8. https://doi.org/10.1007/s12098-013-0923-9 3. Goldstein B, Giroir B, Randolph A; International Consensus Conference on Pediatric Sepsis. International pediatric sepsis consensus conference: Definitions for sepsis and organ dysfunction in pediatrics. Pediatr Crit Care Med. 2005;6(1):2–8. https://doi.org/10.1097/01.CCM.0000149131.72248.E6 4. Haque IU, Zaritsky AL. Analysis of the evidence for the lower limit of systolic and mean arterial pressure in children. Pediatr Crit Care Med. 2007;8(2):138–44. https://doi.org/10.1097/01.PCC.0000257039.86094.BC 5. Scott HF, Donoghue AJ, Gaieski DF, Marchese RF, Mistry RD. The utility of early lactate testing in undifferentiated pediatric systemic inflammatory response syndrome. Acad Emerg Med. 2012;19(11):1276–80. https://doi.org/10.1111/acem.12013 6. Lima A, Bakker J. Noninvasive monitoring of peripheral perfusion. Intensive Care Med. 2002;28(10):1316–20. https://doi.org/10.1007/s00134-002-1380-2 7. Hakanen T, Reinikainen M, Kuitunen A. Perfusion index as a predictor of outcome in critically ill patients. J Clin Monit Comput. 2019;33(3):433–9. https://doi.org/10.1007/s10877-018-0186-7 8. De Felice C, Latini G, Vacca P, Kopotic RJ, De Angelis MC. The pulse oximeter perfusion index as a predictor of high illness severity in neonates. Eur J Pediatr. 2002;161(10):561–2. https://doi.org/10.1007/s00431-002-1079-5 9. Bazaraa HM, Abdelrahman HA, El-Sharkawy HM. Perfusion index as a predictor of impaired perfusion and outcome in critically ill children. J Crit Care. 2021;64:267–72. https://doi.org/10.1016/j.jcrc.2021.04.016 10. Ullah S, Iqbal A, Malik MA. Use of perfusion index in neonatal sepsis for early detection of shock. Pak J Med Sci. 2016;32(6):1473–7. https://doi.org/10.12669/pjms.326.10892 11. Sharma R, Kumar P, Aggarwal R. Utility of perfusion index for assessment of illness severity in children admitted to pediatric intensive care unit. Indian Pediatr. 2020;57(6):523–7. https://doi.org/10.1007/s13312-020-1825-7 12. Dumas JP, Rojas-Burke E, Khanna A. Clinical utility of perfusion index in assessing fluid responsiveness. Crit Care Res Pract. 2019;2019:5676943. https://doi.org/10.1155/2019/5676943 13. Darnall RA. Perfusion index as a measure of peripheral perfusion. J Clin Monit Comput. 2010;24(4):261–5. https://doi.org/10.1007/s10877-010-9258-3