Vasoactive inotropic score to predict mortality in neonatal septic shock

INTRODUCTION

Neonatal septic shock is a critical condition marked by systemic infection and cardiovascular collapse in newborns, often with high rates of mortality and morbidity. It presents with variable and non-specific clinical signs, making early diagnosis challenging. Common symptoms include tachycardia, respiratory distress, poor perfusion and hypotension. The pathophysiology involves reduced cardiac output with increased systemic vascular resistance. In preterm neonates, the immature immune system and adrenal insufficiency contribute to a heightened risk of infection and haemodynamic instability.^[1]

Septic shock is classified into cold and warm types based on peripheral perfusion. Cold shock presents with vasoconstriction and poor perfusion, while warm shock involves vasodilation and increased cardiac output. These are further classified into compensated, uncompensated and irreversible shock, depending on the progression of clinical deterioration.^[2] Pre-term neonates are particularly susceptible due to low IgG levels, immature barriers and frequent use of invasive medical devices.^[3] Sepsis is categorised into early-onset sepsis (EOS-within 72 h of birth) and late-onset sepsis (LOS −>72 h after birth). EOS typically results from maternal infections and vertical transmission, while LOS stems from environmental sources post-delivery, including healthcare settings. The most common LOS pathogens include coagulase-negative staphylococci, Gram-negative bacilli and Candida species. Multidrug-resistant strains contribute to increased neonatal mortality and present a major concern in neonatal intensive care units (NICUs).

Clinical features of neonatal septic shock include abnormalities in heart rate, temperature, perfusion, neurologic status, respiratory function and gastrointestinal symptoms. Laboratory indicators such as metabolic acidosis, elevated lactate, leucocyte count abnormalities and raised inflammatory markers such as C-reactive protein, procalcitonin and interleukin-6 are essential for diagnosis.^[4] Functional echocardiography often shows preserved ejection fraction but elevated cardiac output in septic shock, suggesting a vasodilatory cause.^[5]

The management of neonatal septic shock involves prompt fluid resuscitation and use of vasoactive and inotropic agents. Crystalloids are preferred for initial resuscitation.^[6] Dopamine is commonly used due to its safety, while dobutamine, epinephrine, norepinephrine and vasopressin are selected based on the patient’s specific haemodynamic profile.^[7] Milrinone and newer agents such as levosimendan are being explored for advanced management.

The vasoactive inotropic score (VIS) is a tool that quantifies the total vasoactive support a patient is receiving. It incorporates dosages of various agents such as dopamine, epinephrine and norepinephrine. VIS has proven to be a reliable predictor of outcomes in neonates and children with septic shock or post-cardiac surgery.^[8] Higher VIS and VIS_max values are associated with increased risk of mortality, prolonged intensive care unit stay and longer ventilation requirements.

Multiple studies support VIS’s prognostic role across different neonatal conditions, including congenital diaphragmatic hernia,^[9] congenital heart disease (CHD) surgery,^[10] and pulmonary hypertension. VIS has also shown correlation with biomarkers like troponin and has the potential to serve as an early indicator of cardiac dysfunction when lab access is limited.^[11]

Despite variability in VIS calculation methods and cutoff thresholds across studies, VIS remains a practical and objective measure for risk stratification in neonatal intensive care. It allows clinicians to assess the severity of illness, monitor therapy effectiveness and guide escalation of care. Moreover, integrating VIS with biomarkers such as Mid regional pro-adrenomedullin (MR-proADM) and predictive models such as the EuroSCORE has shown enhanced mortality prediction in surgical settings.^[12]

Emerging evidence suggests that VIS can be integrated into electronic health systems for real-time monitoring and alerts. Research also indicates that combining VIS with other indices, such as lactate levels and lactate clearance, can provide a more comprehensive assessment of patient status. High VIS values correlate with lower lactate clearance and higher mortality, making it a useful early warning tool.^[13]

The future of VIS may involve expansion into modified scores such as the levosimendan VIS, which accounts for newer vasoactive agents.^[14] The continued refinement and validation of VIS and its extensions will enhance their utility in both clinical practice and research. Neonatal septic shock is a complex and rapidly progressing condition. Early identification and management are crucial to improving outcomes. VIS has emerged as a valuable tool in this regard, offering a standardised, quantifiable approach to evaluate cardiovascular support and predict outcomes. Its incorporation into neonatal sepsis management protocols holds promise for advancing neonatal critical care and improving survival rates among the most vulnerable patients. This study was conducted with the objective of studying the association of VIS with mortality of late-onset neonatal sepsis and septic shock and to classify neonates into high-risk and low-risk based on VIS score.

MATERIAL AND METHODS

This observational study was conducted in the NICUs of Sangameshwar Teaching and General Hospital and Basaveshwar Teaching and General Hospital, both affiliated with Mahadevappa Rampure Medical College in Kalaburagi, Karnataka. A total of 60 neonates meeting the specified inclusion and exclusion criteria were enrolled using a simple purposive sampling method. The study participants were selected based on clinical relevance rather than randomisation, ensuring the inclusion of appropriate cases for the research objectives. Data collection was performed using a semi-structured pro forma designed to capture relevant clinical and demographic information from each neonate admitted to the NICU.

The study included all neonates – pre-term, term and post-term – with culture-positive LOS and septic shock unresponsive to fluid resuscitation. Exclusion criteria comprised neonates with culture-positive EOS, those who had received inotropes before the onset of sepsis and infants with major congenital malformations or CHD.

Newborns with EOS were excluded from this study due to significant differences in pathophysiology compared to LOS and paediatric sepsis. In the early neonatal period, haemodynamic physiology is uniquely influenced by factors such as a physiologically open ductus arteriosus, immature myocardium and underdeveloped vasomotor tone. In addition, respiratory distress syndrome is more prevalent, particularly in pre-term infants. These conditions are major contributors to neonatal hypotension, making it difficult to identify its underlying causes during this early stage accurately.

Late-onset sepsis is diagnosed with clinical and laboratory parameters and septic shock is managed. The standard institutional protocol is used to treat hypotension and shock. Initially, dopamine and dobutamine were started, and if no response adrenaline and noradrenaline were added sequentially.^[2] Four VIS values are calculated for each sepsis episode in the first 48 h of septic shock. VIS0 is the value when the first inotrope was started.VIS24 and VIS48 at 24 and 48 h after ionotropic treatment commenced. Necessary blood investigations were carried out; however, echocardiography was not performed before initiating treatment.

Formula: VIS = epinephrine (mcg/kg/min) × 100 + norepinephrine(mcg/kg/min) × 100 + milrinone (mcg/kg/min) × 10 + vasopressin (U/kg/min) × 10000 + Dopamine (mcg/kg/min) + dobutamine (mcg/kg/min)

Informed, written consent was taken from the patient attenders in their own vernacular language. The identity of the patient was kept confidential. Statistical data were analysed using IBM Statistical Package for the Social Sciences version 27.0 software. For qualitative data analysis, the Chi-square test was applied. For quantitative data analysis, the Z-test, t-test and analysis of variance test were applied for statistical significance. A P < 0.05 was considered statistically significant.

RESULTS

Gestational age, birth weight, pathogen type and serum lactate levels were all found to have statistically significant associations with mortality in neonatal septic shock. Lower gestational age (P = 0.004) and lower birth weight (P = 0.003) were more common among non-survivors, indicating their role as major risk factors [Table 1].

Infections with multidrug-resistant Gram-negative organisms such as Acinetobacter baumannii and Klebsiella pneumoniae were significantly associated with higher mortality (P = 0.032), while less virulent organisms were more frequent in survivors. In, non-survivors had markedly elevated serum lactate levels (P = 0.004), suggesting that high lactate may serve as a valuable biomarker for poor outcomes in these patients. Invasive ventilation was strongly associated with non-survival, suggesting that patients requiring it were likely more critically ill [Table 2].

Among the non-survivors, multiple inotrope combinations – particularly those including adrenaline and noradrenaline – were more frequently used, whereas survivors predominantly received single-agent therapy such as dobutamine or adrenaline alone. Overall, 50 survivors and 10 non-survivors required inotropic support, with dobutamine being the most commonly used agent [Table 3].

The mean inotropic score was found to be significantly higher in non-survivors compared to survivors. Specifically, non-survivors had a mean score of 51.00 ± 12.13, while survivors had a mean of 21.98 ± 10.49. This difference was statistically significant, with a Mann-Whitney U value of 20.000 and a P = 0.001. This suggests that a higher inotropic requirement is strongly associated with increased mortality risk. The inotropic score reflects the intensity of pharmacological support needed to maintain cardiovascular function, and its elevation in non-survivors indicates more severe haemodynamic compromise [Table 4].

Similarly, the duration of vasoactive support, measured as VIS hours, was also significantly associated with mortality. Non-survivors had a mean VIS of 33.00 ± 11.05 h, while survivors had a shorter duration at 22.12 ± 10.02 h. This difference was statistically significant (Mann–Whitney U = 101.000, P = 0.001), suggesting that prolonged use of vasoactive agents correlates with poorer outcomes. This could be due to prolonged exposure to severe disease or complications requiring extended pharmacological intervention [Table 5].

The predictive value of these parameters was further examined through receiver operating characteristic (ROC) analysis. The inotropic score demonstrated excellent discriminatory power with an area under the curve (AUC) of 0.96, a sensitivity of 100% and a specificity of 84% at a cutoff value of ≥32. This indicates that the inotropic score is both a highly sensitive and specific marker for identifying neonates at an increased risk of mortality. Group analysis reinforced this, with patients having inotropic scores ≥32 showing a significantly higher mean score of 46.39 ± 10.64 compared to those with scores <32 [Figure 1].

Close

Figure 1:

Ionotropic score for mortality.

Export to PPT

The ROC analysis for VIS hours also showed good predictive utility, with an AUC of 0.80. The analysis identified a cutoff value of ≥30 h, yielding a sensitivity of 80% and specificity of 88%. Although not as precise as the inotropic score, VIS hours still provide valuable insight into patient prognosis. In the subset analysis, neonates with VIS ≥30 h had a mean duration of 39.43 ± 5.63 h, further highlighting the association between prolonged vasoactive support and increased mortality risk [Figure 2].

Close

Figure 2:

Receiver operating characteristic (ROC) of vasoactive inotropic score and its maximum value hours, for mortality.

Export to PPT

DISCUSSION

The findings from our study, which identified a VIS >32 as a significant predictor of mortality in neonatal septic shock, aligning closely with a growing body of research that highlights the prognostic value of VIS in critically ill paediatric patients. Similar studies, including Amer et al. and Gaies et al., have demonstrated that elevated VIS is strongly associated with adverse outcomes in neonatal and paediatric septic shock populations.^[15,16] While Amer et al. identified a VIS cutoff of ≥25 in pre-term neonates and Gaies et al. reported a VIS >20 as predictive of 100% mortality in fluid-refractory shock.^[15,16] Our study’s threshold of 32.5 sits comfortably between these values, suggesting that neonates with severe septic shock require more aggressive inotropic support, which is associated with a higher risk of mortality. Interestingly, studies focusing on cardiac surgery populations, such as Gaies et al. (VIS >15.5), show lower VIS cutoffs, but the context of post-operative recovery further underscores the variability of VIS thresholds across different patient groups.^[16] The Neonatal Septic Shock Study (2022) found a VIS_max cutoff >20, supporting the robustness of VIS as a predictor, even at lower thresholds [Table 6].

One of the key limitations of this study is its single-centre design. Data were collected from a specific hospital or NICU, which may limit the generalisability of the findings to other institutions, particularly those with different patient demographics, care practices or resource availability. The VIS cutoff of 32.5 identified in your study may require further external validation in independent cohorts to confirm its predictive ability. There is inherent variability in the calculation of VIS based on clinical practice, as different hospitals or clinicians may use slightly different protocols for dosing inotropes and vasopressors. This could lead to variability in VIS scores, affecting the consistency and comparability of the results across institutions.

CONCLUSION

This study concludes that the VIS and its maximum value (VIS_max) are strong and statistically significant predictors of mortality in neonates with septic shock. Specifically, an Inotropic Score ≥32 and VIS_max ≥30 are associated with significantly increased mortality risk, demonstrating excellent (AUC 96%) and strong (AUC 80%) predictive accuracy, respectively. These findings support the use of VIS as a practical and effective tool for early risk stratification, allowing for timely clinical interventions in critically ill neonates. However, due to limitations such as the single-centre design and small sample size, further multicentre studies are warranted to validate and generalise these results.