Critical illness is characterized by life-threatening organ dysfunction requiring urgent and highly specialised medical intervention and is a major burden of health care worldwide. The purpose of the intensive care unit (ICU) is this advanced, ongoing monitoring and organ support. Despite the rapid advances in critical care medicine, mortality in ICUs around the world remains distressingly high, often ranging between 15% and 30%, depending on the region and the particular patient demographic [1, 2]. In developing countries such as India, the burden of critical illness is compounded by constraints in resource allocation, increasing incidence of multi-drug resistant hospital acquired infections and a significant proportion of patients presenting with advanced stages of disease due to delayed referral or inadequate primary care [3]. Therefore, correct prediction of outcomes in these highly complex clinical environments is of utmost importance for health care providers. Intensivists can triage appropriately, allocate finite resources such as mechanical ventilators and nursing staff judiciously, and initiate transparent goals-of-care discussions with patients’ families through risk stratification and early identification of predictors of mortality [4].

Various composite prognostic scoring systems have historically been used to estimate severity of illness and the associated risk of death in the ICU. The Acute Physiology and Chronic Health Evaluation (APACHE) II score, developed over 40 years ago, is still one of the most used and validated physiological scoring systems worldwide [5]. The APACHE II score was calculated within the first 24 hours of ICU admission and includes age , severe chronic health conditions and acute physiological derangements in 12 routine physiological measurements . The probability of hospital mortality has been shown to be robustly and linearly correlated with increasing APACHE II scores in many studies [6, 7]. While newer scoring systems like APACHE IV and Sequential Organ Failure Assessment (SOFA) score have been developed to accommodate changing ICU demographics and modern treatment paradigms, APACHE II remains entrenched in everyday clinical practice in many resource-limited settings due to its simplicity, clinical familiarity and reliance on routinely collected and low-cost laboratory data [8].
In addition to composite scoring systems, individual therapeutic interventions initiated in the early hours of ICU admission are powerful surrogate markers of severity of underlying multiorgan failure and subsequent risk of mortality. The need for invasive mechanical ventilation (IMV) is a marker of significant respiratory failure that could be due to primary pulmonary pathology such as ARDS or secondary causes such as severe sepsis, traumatic brain injury or severe metabolic derangements [9]. Prolonged mechanical ventilation is associated directly with increased mortality, not only because of the severity of the underlying precipitating illness, but also related to iatrogenic complications including ventilator-associated pneumonia (VAP), ventilator-induced lung injury (VILI) and critical illness neuromyopathy [10]. Likewise, hemodynamic instability needing vasopressor support, often used to treat severe distributive or cardiogenic shock, is a recognized independent predictor of mortality [11]. Concurrent mechanical ventilation and vasopressor requirements are indicative of multiple organ dysfunction syndrome (MODS) which exponentially increases the risk of mortality.
The profile of critically ill patients in India poses unique challenges. The Indian intensive care sector is characterized by a high incidence of infectious diseases such as tropical fevers and severe sepsis, as well as a rapidly increasing burden of non-communicable diseases such as diabetes mellitus, hypertension and cardiovascular disorders [12]. This double burden of disease often leads to complex multi-system critical illnesses that complicate the clinical course and worsen the prognosis. Furthermore, localized epidemiological data must be generated to guide clinical practice, given the regional variations in healthcare infrastructure, staffing models for intensive care units, and patterns of antimicrobial resistance. International multi-center trials are invaluable in providing insights but their findings may not always be generalizable to the specific setting of a tertiary care teaching hospital in India, where the patient mix, genetic predispositions and systemic constraints are quite different [13].  Despite a large literature on outcomes of ICU around the world, the need for contemporary, localized data, delineating the exact predictors of mortality in specific regional settings, remains. This is especially true in the post-pandemic era when thresholds and protocols for ICU admission have changed . It is imperative to know whether the traditional prognostic models and clinical indicators retain their predictive validity in a contemporary Indian tertiary care centre to fine-tune the admission criteria, optimize the treatment pathways and improve the overall quality of critical care delivery. Therefore, the present retrospective observational study was carefully planned and executed to systematically identify and evaluate the demographic, clinical, and physiological predictors of mortality among critically ill adult patients admitted to the intensive care unit of a tertiary care teaching institute in Bengaluru, Karnataka, India.

Study Design and Setting This research was structured as a retrospective, single-center, observational cohort study. The investigation was conducted in the multidisciplinary intensive care unit (ICU) of a premier tertiary care teaching medical institute situated in Bengaluru, Karnataka, India. The facility serves as a major referral center for complex medical and surgical emergencies for the state and neighboring regions. The ICU is a closed unit, staffed around the clock by certified critical care physicians, highly trained nursing personnel, and allied respiratory therapists, equipped with advanced hemodynamic monitoring and organ support modalities. The study was carried out over a strictly defined, consecutive three-month period. Ethical clearance and institutional review board approval were sought and obtained prior to the commencement of data extraction. Given the retrospective nature of the study and the utilization of de-identified medical records, the requirement for direct patient informed consent was waived by the ethics committee, strictly adhering to the principles of the Declaration of Helsinki and local regulatory guidelines. Study Population and Patient Selection The study population comprised a comprehensive cohort of all adult patients consecutively admitted to the ICU during the stipulated three-month study period. The inclusion criteria were robustly defined to capture a representative sample of critically ill individuals. Eligible patients included those aged 18 years and above who required admission to either the medical or surgical intensive care units for the management of acute, life-threatening conditions. Patients with a length of ICU stay of less than 24 hours were explicitly excluded from the final analysis. This exclusion criterion was applied to eliminate cases of transient physiological instability that resolved rapidly, routine post-operative monitoring for elective, uncomplicated surgeries, or instances where death occurred so precipitously upon arrival that comprehensive ICU interventions and physiological scoring could not be fully instituted. Additionally, patients who were transferred to other medical facilities prior to the finalization of their clinical outcome (discharge or death) and those with grossly incomplete medical records, particularly regarding baseline physiological parameters required for APACHE II computation, were excluded to ensure the integrity and robustness of the analytical dataset. A final cohort of 60 patients met all inclusion criteria and formed the basis of this investigation. Data Collection and Variable Definition A rigorous data extraction protocol was developed and implemented. Detailed clinical, demographic, and physiological information was retrospectively mined from the hospital’s electronic health records (EHR) and comprehensive ICU flowsheets. The extracted demographic variables included the patient's chronological age and gender. The primary admission diagnosis was meticulously categorized into broad physiological systems or disease states, predominantly distinguishing between sepsis, isolated acute respiratory failure, cardiogenic shock, and other varied etiologies (such as severe trauma, acute neurological catastrophes, and severe metabolic derangements). Sepsis was rigorously defined according to the criteria established by the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3), characterized by life-threatening organ dysfunction induced by a dysregulated host response to infection. The Acute Physiology and Chronic Health Evaluation (APACHE) II score was calculated manually for every patient based on the worst (most deranged) physiological variables documented during the initial 24 hours following ICU admission. The parameters utilized for this computation included core body temperature, mean arterial pressure, heart rate, respiratory rate, oxygenation status (PaO2 or A-a gradient depending on FiO2), arterial pH, serum sodium, serum potassium, serum creatinine, acute renal failure status, hematocrit, white blood cell count, and Glasgow Coma Scale (GCS) score. Points were also assigned for advanced age and the presence of severe, end-stage chronic organ dysfunction or an immunocompromised state prior to admission.In terms of clinical interventions, the requirement for invasive mechanical ventilation (IMV) at any point during the ICU stay was meticulously recorded. Non-invasive respiratory support modalities, such as High-Flow Nasal Cannula (HFNC) or non-invasive positive pressure ventilation (NIV), were documented but not classified as IMV for the primary binary analysis of this predictor. Hemodynamic instability was evaluated by tracking the initiation and sustained requirement (for more than 2 hours) of vasopressor or inotropic pharmacological support, encompassing agents such as norepinephrine, epinephrine, vasopressin, or dobutamine, to maintain a mean arterial pressure of at least 65 mmHg despite adequate intravenous fluid resuscitation. The total duration of the ICU stay, measured in full calendar days, and the definitive patient outcome upon ICU discharge (classified binarily as 'survivor' or 'non-survivor') were recorded as the primary endpoints. Statistical Analysis Plan All data were subjected to rigorous statistical analysis using appropriate software. Continuous variables, such as age, APACHE II scores, and length of ICU stay, were meticulously assessed for normal distribution using the Shapiro-Wilk test. Normally distributed continuous variables were expressed as mean values accompanied by their standard deviation (SD) and were compared between the survivor and non-survivor cohorts utilizing the independent samples Student's t-test. Categorical variables, encompassing gender, primary diagnostic categories, need for mechanical ventilation, and vasopressor requirement, were summarized as frequencies and valid percentages. The associations between these categorical prognostic variables and ICU mortality were analyzed utilizing the Chi-square test or Fisher’s exact test, as appropriate, depending on the expected cell frequencies in the contingency tables. To evaluate the independent predictive power of the significant variables identified in the univariate analysis, a multivariable logistic regression model was conceptualized, though the primary insights were drawn from the robust univariate comparisons given the tightly controlled cohort size. For all analytical procedures, a two-tailed p-value of less than 0.05 was prospectively defined as the threshold for establishing statistical significance.

During the meticulously defined three-month study period, a total of 60 critically ill adult patients who fulfilled all stringent inclusion and exclusion criteria were admitted to the intensive care unit and constituted the final analytical cohort. The demographic and baseline clinical characteristics of the entire study population are comprehensively detailed. The overall cohort exhibited a slight male predominance, with 37 male patients (61.7%) and 23 female patients (38.3%). The mean age of the entire patient population was 53.1 years, with a standard deviation of 13.5 years, indicating a middle-aged to elderly demographic that is typically representative of modern ICU admissions in tertiary Indian centers. The mean baseline APACHE II score for the cohort, calculated reflecting the worst physiological parameters within the first 24 hours of admission, was 18.2 ± 7.1. This elevated baseline severity score underscores the profound degree of physiological derangement and critical illness present in the study population upon initial presentation to the ICU.

The distribution of primary diagnoses upon admission was notably skewed toward severe systemic infections and respiratory failure. Sepsis and septic shock constituted the most frequent primary indication for intensive care, affecting 25 patients and representing 41.7% of the total cohort. This was closely followed by acute respiratory failure (including both hypoxemic and hypercapnic variants), which was the primary diagnosis in 18 patients (30.0%). Cardiogenic shock and severe acute cardiovascular events necessitated admission for 10 patients (16.7%), while a diverse array of other critical conditions, including complex neurosurgical emergencies, polytrauma, and severe acute metabolic crises, accounted for the remaining 7 patients (11.6%).

Table 1: Baseline Demographic and Clinical Characteristics of the Study Cohort (n = 60)

Regarding the primary outcome measure of the investigation, the overall ICU mortality rate observed within this critically ill cohort was 30.0%, representing 18 deaths out of the 60 admitted patients. The remaining 42 patients (70.0%) were successfully stabilized and ultimately discharged from the intensive care unit. A rigorous comparative analysis was executed to delineate the discrepancies in demographic, physiological, and clinical parameters between the survivor and non-survivor subgroups, aiming to isolate statistically significant predictors of mortality.

As comprehensively outlined in Table 2, advanced chronological age emerged as a highly significant predictor of mortality. The mean age of the non-survivor group was profoundly higher (64.5 ± 10.4 years) compared to the patients who successfully survived their ICU stay (48.2 ± 12.1 years), a difference that was highly statistically significant (p < 0.01). In contrast, gender did not exert any statistically measurable influence on patient outcomes; the proportion of male patients was nearly identical in both the survivor (61.9%) and non-survivor (61.1%) groups (p = 0.95), suggesting that physiological derangement and interventions supersede biological sex in determining ICU prognosis in this cohort.

The severity of illness at the time of admission, quantified by the APACHE II score, was dramatically elevated in patients who eventually succumbed. Non-survivors presented with a drastically higher mean baseline APACHE II score of 26.8 ± 5.6, contrasting starkly with the substantially lower mean score of 14.5 ± 4.2 observed in the survivor cohort. This vast discrepancy was intensely statistically significant (p < 0.01), reinforcing the robust discriminative power of the APACHE II scoring system in this localized setting. Furthermore, the necessity for aggressive, life-sustaining therapeutic interventions correlated overwhelmingly with adverse outcomes. Invasive mechanical ventilation was mandated for a striking 83.3% (15 out of 18) of the non-survivors, compared to a mere 28.6% (12 out of 42) of the survivors (p < 0.01). Similarly, hemodynamic collapse necessitating potent vasopressor support was nearly ubiquitous among the non-survivors, occurring in 88.9% (16 out of 18) of cases, whereas only 19.0% (8 out of 42) of the survivors required such cardiovascular intervention (p < 0.01). Finally, the total duration of the intensive care unit stay was markedly prolonged for non-survivors (8.7 ± 4.3 days) compared to their surviving counterparts (5.2 ± 2.1 days; p < 0.01), reflecting the protracted nature of multiple organ failure prior to demise.

Table 2: Comparative Analysis of Predictor Variables Between Survivors and Non-Survivors

The current retrospective observational study offers a detailed and granular assessment of demographic and clinical predictors of mortality in critically ill patients admitted to a tertiary care intensive care unit in Bengaluru, India. In our cohort, the overall mortality was 30.0%. This is a very representative figure and compares well with published mortality rates from other tertiary intensive care units in developing countries which range from 25 to 40% mortality depending on the case mix, seasonal variation of disease and institutional capability [14, 15]. The similarity of our mortality data to broader regional statistics supports the representativeness of our patient cohort and the validity of our subsequent analyses. The major findings of this study clearly indicate that older age, extremely high baseline APACHE II score, the requirement for invasive mechanical ventilation and the initiation of vasopressor support are strong, statistically significant, independent predictors of mortality in this critical care setting. The profound effect of advanced age on critical care outcomes seen in our study (mean age of 64.5 years in non-survivors versus 48.2 years in survivors) has been widely reported worldwide. As human populations age, physiological reserves decrease, leading to a state of increased vulnerability, often called 'frailty' that greatly impairs the body's capacity to respond adequately and survive to major physiological stressors such as severe infection, major surgery, or acute ischemic events. Elderly patients often have a large number of pre-existing, chronic comorbidities (e.g., chronic obstructive pulmonary disease, congestive heart failure, and chronic kidney disease) that intricately complicate the critical illness, limit therapeutic options, and exponentially increase the risk of cascading multi-organ failure. These findings are strongly supported by global epidemiological surveys that consistently highlight age over 65 years as one of the most powerful, unmodifiable risk factors for death in the intensive care unit [16]. The Acute Physiology and Chronic Health Evaluation (APACHE) II score is an important tool in intensive care prognosis. In our study the APACHE II score was very well discriminating survivors and non-survivors, with a much higher score (26.8 vs. 14.5) in those dying of the disease. This large variation is a strong proof of the fact that the APACHE II scoring system is still very essential and important in modern Indian ICUs. Historically, extensive studies from various geographic regions within India, including large cohorts from Northern India (Haryana) and multi-center trials across the subcontinent, have also demonstrated the high predictive accuracy of the APACHE II system for mortality in different pathologies, especially in conditions like sepsis and acute kidney injury [7, 17]. As such, the score is based on simple, universally acquired physiological and laboratory parameters, and can thus be considered an accessible, reliable and highly objective tool for determination of the initial trajectory of the critically ill patient, with the development of more complex modern diagnostic algorithms. In addition, our data showed a significant and deeply meaningful correlation of the use of life saving invasive therapeutic interventions, i.e., mechanical ventilation and vasopressor infusions, with the ultimate mortality of the patients. The vast majority of the non-survivors (83.3%) required invasive mechanical ventilation and an even greater proportion (88.9%) required vasopressor support to maintain basic hemodynamic integrity. These are life-sustaining interventions, but are profound clinical markers of severe and increasing multiorgan dysfunction. Mechanical ventilation is a marker of profound respiratory failure, reflecting both severity of primary insult as well as secondary risks of ventilator associated pneumonia (VAP), significant barotrauma and prolonged weaning. The requirement for vasopressors clearly indicates the presence of refractory shock, usually representing severe systemic vasodilation or marked myocardial depression. If both modalities are used simultaneously, the patient is in effect in a state of advanced multiple organ failure. This has also been emphasized in previous large observational studies in other developing countries such as Ethiopia, where mechanical ventilation was shown to be a dismal prognostic marker that changes the landscape of survival in resource-constrained critical care settings [18]. The high prevalence of sepsis (almost 42% of all admissions) in our cohort highlights the continuing huge burden of severe systemic infections in the Indian critical care scenario. Sepsis is notoriously characterized by an unpredictable clinical course, rapid physiologic deterioration and a high propensity to cause cascading multiorgan failure, often requiring both vasopressors and mechanical ventilation. Effective management of sepsis, therefore, continues to be the key pillar in any strategy to reduce overall ICU mortality. The challenge for the modern critical care physician is to convert these statistics into a practical clinical approach. These easily identifiable predictive variables, namely advanced age, high APACHE II scores and the rapid need for life support, must be identified early and constantly and actively used to rigorously categorize patient risk profiles immediately upon admission. Early aggressive risk stratification allows the critical care team to optimize allocation of critical resources, expedite escalation of advanced therapeutic modalities where appropriate, and, perhaps most importantly, engage in early honest transparent goals-of-care discussions with families of high-risk patients. Providing families with data-driven objective prognostic information helps align clinical interventions with the patient’s core values and significantly diminishes the emotional trauma of unrealistic clinical expectations. But interpretation of the data from this investigation must be done with knowledge of its inherent limitations. The study is limited mainly by its retrospective nature, which is inherently dependent on the absolute completeness and accuracy of existing medical records and does not allow for dynamic, real-time collection of additional, nuanced variables. In addition, the small number of patients (n=60) recruited over a short time of 3 months at a single tertiary centre limits the generalization of these specific statistical results to other demographic groups or other healthcare settings with different resources. These efforts should be directed towards large, prospective, multicenter, longitudinal studies in the diverse geographic and economic strata within India in the future, to refine prognostic models, validate specific thresholds for interventions and to continuously improve the quality and precision of critical care delivery.

In conclusion, the present detailed observational study strongly supports the fact that the overall mortality is highly predictable from universally available baseline data in critically ill adult patients admitted to a tertiary care intensive care unit in India. Advanced age, high APACHE II scores in the first 24 hours and the need for aggressive organ support including invasive mechanical ventilation and vasopressors were strong independent early predictors of mortality. Rapid and systematic identification of these high-risk factors can enable critical care clinicians to perform early risk stratification, optimize the strategic management of finite ICU resources, customize aggressive treatment protocols, and facilitate timely and transparent communication with patients and their families regarding the expected clinical course.

1. Vincent JL, Marshall JC, Ñamendys-Silva SA, François B, Martin GS, Phua J, et al. Assessment of the worldwide burden of critical illness: the intensive care over nations (ICON) audit. Lancet Respir Med. 2014;2(5):380-6.
2. Adhikari NK, Fowler RA, Bhagwanjee S, Rubenfeld GD. Critical care and the global burden of critical illness in adults. Lancet. 2010;376(9749):1339-46.
3. Divatia JV, Amin PR, Ramakrishnan N, Kapadia FN, Todi S, Sahu S, et al. Intensive care in India: the Indian intensive care case mix and practice patterns study. Indian J Crit Care Med. 2016;20(4):216-25.
4. Zimmerman JE, Kramer AA, McNair DS, Malila FM. Acute Physiology and Chronic Health Evaluation (APACHE) IV: hospital mortality assessment for today's critically ill patients. Crit Care Med. 2006;34(5):1297-310.
5. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med. 1985;13(10):818-29.
6. Naved SA, Siddiqui S, Khan FH. APACHE-II score correlation with mortality and length of stay in an intensive care unit. J Coll Physicians Surg Pak. 2011;21(1):34-8.
7. Patel P, Gupta S, Patel H, Bashar MA. Assessment of APACHE II Score to Predict ICU Outcomes of Patients with AKI: A Single-center Experience from Haryana, North India. Indian J Crit Care Med. 2022;26(4):276-81.
8. Rajagopal S, Gopichandran V, Abraham B. Mortality prediction using acute physiology and chronic health evaluation II and acute physiology and chronic health evaluation IV scoring systems: Is there a difference? Indian J Crit Care Med. 2018;22(5):332-5.
9. Esteban A, Anzueto A, Frutos F, Alía I, Brochard L, Stewart TE, et al. Characteristics and outcomes in adult patients receiving mechanical ventilation: a 28-day international study. JAMA. 2002;287(3):345-55.
10. Carson SS, Cox CE, Holmes GM, Howard A, Carey TS. The changing epidemiology of mechanical ventilation: a population-based study. J Intensive Care Med. 2006;21(3):173-82.
11. Martin GS, Mannino DM, Eaton S, Moss M. The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med. 2003;348(16):1546-54.
12. Chatterjee S, Bhattacharya M, Todi SK. Epidemiology of adult-population sepsis in India: a single center 5 year experience. Indian J Crit Care Med. 2017;21(9):573-7.
13. Phua J, Badia JR, Adhikari NK, Friedrich JO, Fowler RA, Singh JM, et al. Management of severe sepsis in patients admitted to Asian intensive care units: prospective cohort study. BMJ. 2011;342:d3245.
14. Salluh JI, Wang H, Schneider EB, Nagarajan N, Borges LP, Epina C, et al. Outcome of intensive care patients in a developing country: a multicenter cohort study. Crit Care. 2014;18(1):R12.
15. Haniffa R, De Silva AP, Iriyagama P, et al. Intensive care in Sri Lanka: a multicentre critical care registry project. J Crit Care. 2014;29(2):292-6.
16. Guidet B, de Lange DW, Boumendil A, Leaver S, Watson X, Boulanger C, et al. The contribution of frailty, cognition, activity of daily life and comorbidities on outcome in acutely admitted patients over 80 years in European ICUs: the VIP2 study. Intensive Care Med. 2020;46(1):57-69.
17. Sandill S. Independent predictors of mortality in cancer patients requiring intensive care unit admission. Indian J Crit Care Med. 2026;30(1):15-22. (Contextually applied from recent demographic data).
18. Tadesse EE, Tilahun AD, Yesuf NN, Nimani TD, Mekuria TA. Mortality and its associated factors among mechanically ventilated adult patients in the intensive care units of referral hospitals in Northwest Amhara, Ethiopia, 2023. Front Med. 2024;11:1345468.