Clinical Profile and Outcomes of Obstetrics ICU Admission in a North India Tertiary Care Centre a Retrospective Study

Introduction

Pregnancy, though a physiological process, carries inherent risks that can escalate to life-threatening emergencies requiring intensive care. The intensive care unit (ICU) serves as the last line of defence for critically ill patients experiencing circulatory failure, respiratory compromise, renal dysfunction, disseminated intravascular coagulation, or septicaemia. In obstetric practice, ICU admission has emerged as an objective marker of severe maternal morbidity, reflecting the gravity of complications arising during the peripartum period [1].

Globally, obstetric patients account for a small but significant proportion of ICU admissions, with reported rates ranging between 0.7% and 13.5%, though this figure varies considerably across countries and institutions [2]. In developing nations, pregnant women may constitute up to 10% of all ICU admissions, largely because of late referrals, inadequate antenatal care, low socioeconomic status, and poor health infrastructure at the primary level [3]. In India specifically, obstetric ICU admissions contribute approximately 0.4-4% of total ICU admissions in public hospitals, with maternal mortality rates reported between 12% and 33% in various Indian studies [4].

The reasons for ICU care in critically ill obstetric patients fall broadly into three categories. The first encompasses conditions unique to pregnancy such as eclampsia, preeclampsia, obstetric haemorrhage, and puerperal sepsis. The second includes pre-existing medical illnesses aggravated by pregnancy, such as rheumatic heart disease, hypertension, and diabetes. The third involves conditions that, while manageable in non-pregnant individuals, carry disproportionately high mortality in pregnant women, such as hepatitis E [5]. Across most Indian studies, obstetric haemorrhage and hypertensive disorders of pregnancy consistently emerge as the two leading indications for ICU admission [4,6].

Managing critically ill obstetric patients presents unique challenges to intensivists and obstetricians alike. The altered maternal physiology of pregnancy affects every organ system, complicating both diagnosis and therapeutic decision-making. Additionally, the simultaneous consideration of fetal wellbeing adds a layer of complexity that is absent in standard critical care [5]. Standard severity scoring systems such as APACHE II and SAPS II may overestimate disease severity in obstetric patients due to pregnancy-related physiological changes, limiting their applicability in this population [7].

Despite growing recognition of obstetric critical care as a subspecialty, data from tertiary care centres in north India remain limited. The present retrospective study was therefore undertaken to analyse the incidence, clinical characteristics, interventions performed, and outcomes of obstetric patients admitted to the ICU of a tertiary care centre in north India, with the aim of identifying preventable causes and informing better clinical practice.

Materials and Methods

This retrospective observational study was conducted in the Department of Obstetrics and Gynaecology along with the Obstetric Intensive Care Unit of a tertiary care centre in north India over a period of 18 months. All pregnant women and postpartum women up to six weeks following delivery who required ICU admission during the study period were included. Patients admitted for gynaecological conditions unrelated to pregnancy and non-critical obstetric patients were excluded. Readmissions within 30 days were counted only once to avoid duplication of data. Ethical approval was obtained from the institutional ethics committee prior to commencement of data collection, and since this was a record-based retrospective analysis, the requirement for individual patient consent was waived.

Data were retrieved from the medical records section of the hospital. A predesigned proforma was used to systematically collect relevant information for each patient. Variables recorded included patient demographics such as age, parity, residential background, socioeconomic status, obstetric details including gestational age at admission, booking status, mode of delivery, adequacy of antenatal care received. Indications for ICU admission were divided into obstetric causes (pregnancy-related causes that occurred during pregnancy or within 42 days of termination of pregnancy) and non-obstetric causes (medical or surgical causes not directly related to pregnancy). Standard severity scoring systems such as APACHE II and SAPS were not used, as it is known that physiological changes associated with pregnancy lead to spuriously elevated scores, thus limiting their validity in this population.

All interventions performed during the ICU stay were meticulously recorded. These were the need of mechanical ventilation and its duration, inotropic and vasopressor support, blood and blood product transfusions, central venous line placement, renal replacement therapy, administration of anticonvulsants and antihypertensives and use of broad spectrum antibiotics. We also noted surgical procedures including cesarean section, cesarean hysterectomy, uterine artery ligation and balloon tamponade. Length of stay in ICU was calculated in days from date of admission to date of discharge/death.

Maternal outcome was recorded as survival or death and causes of mortality were recorded on the basis of clinical findings and investigation reports. Perinatal outcome in terms of intrauterine death, still birth and early neonatal death was also recorded. Data were entered into Microsoft Excel and analyzed using standard statistical software. Continuous variables were reported as mean with standard deviation and categorical variables as frequency and percentage. Continuous variables were compared between survivors and non-survivors using independent samples t-test and categorical variables using Fisher’s Exact test. Statistical significance was considered at p < 0.05.

Results

The mean age of obstetric ICU admissions was 28.9 ± 6.7 years, with the majority (70%) belonging to the 21–35 years age group. Most patients were from rural areas (68%) and were referred from other healthcare facilities (65%). A large proportion had inadequate antenatal care (61%). The overwhelming majority of admissions occurred during the postpartum period (92%). Caesarean section (63%) was the most common mode of delivery among ICU admissions. The mean gravida was 2.7 ± 1.5, mean parity was 1.1 ± 1.0, and mean gestational age at admission was 35.6 ± 2.3 weeks.

Obstetric hemorrhage (32%) was the leading indication for ICU admission, closely followed by hypertensive disorders of pregnancy (30%). Sepsis, heart disease, and acute kidney injury accounted for 10%, 10%, and 8% of admissions, respectively. Among associated comorbid conditions, acute kidney injury (46%) and hypertensive disorders (42%) were the most frequent. At admission, patients exhibited marked physiological derangement, with a mean pulse rate of 117.2 ± 18.7 beats/min and mean oxygen saturation of 93.2 ± 5.2%. Laboratory evaluation revealed significant anemia (Hb 8.2 ± 1.8 g/dL), thrombocytopenia (platelet count 125.6 ± 58.7 ×10³/µL), and impaired renal function (serum creatinine 2.00 ± 1.16 mg/dL), reflecting the severity of illness among obstetric ICU admissions.

Blood transfusion was the most frequently utilized ICU intervention, administered to 69% of patients, highlighting the significant burden of obstetric hemorrhage and anemia. More than half of the patients (52%) required vasopressor support, while 34% underwent mechanical ventilation. Renal replacement therapy was needed in 21% of cases, reflecting the high prevalence of acute kidney injury and multi-organ dysfunction.

The mean ICU stay was 4.0 ± 1.7 days (median 4 days, IQR 3–5), whereas the mean total hospital stay was 10.5 ± 3.0 days (median 10 days, IQR 8–13). Patients who expired had a slightly longer ICU stay (4.2 ± 2.1 days) compared with survivors (4.0 ± 1.7 days). Additionally, 22% of neonates required NICU admission, indicating substantial neonatal morbidity associated with severe maternal illness requiring ICU care.

Out of 100 obstetric admissions to ICU 92 women survived and discharged and 8 patients died with maternal mortality rate of 8%. Haemorrhagic shock (50% of all deaths) was the most common cause of maternal death followed by cardiorespiratory failure (25%). Septicemia with MODS and acute renal failure with MODS each accounted for 12.5% of maternal deaths.

In terms of neonatal outcomes, 80% resulted in live birth and 20% resulted in stillbirth or intra-uterine deaths (IUDs). Of the total neonates 74% were discharged alive and healthy while 6% died early neonatally. Twenty-two percent of the newborns required NICU admission again emphasizing the significant neonatal morbidity associated with severe maternal illness requiring intensive care. The findings suggest that obstetric emergencies were a significant contributor to maternal and neonatal adverse outcomes, despite high maternal survival.

* Statistically significant (p < 0.05)

Comparison of survivors and non-survivors showed that none of the demographic characteristics, admission laboratory parameters, or vital signs were significantly associated with maternal mortality (all p > 0.05). Women who expired were slightly younger and had marginally higher serum creatinine levels; however, these differences did not reach statistical significance.

Among the categorical variables, requirement of dialysis/renal replacement therapy (RRT) was the only significant predictor of maternal mortality. Dialysis was required in 62.5% of women who died compared with 17.4% of survivors (Fisher's Exact test, p = 0.01), indicating a strong association between severe renal dysfunction and adverse maternal outcome.

Although mechanical ventilation was more common among women who expired (62.5% vs. 31.5%), the association was not statistically significant (p = 0.117). Similarly, rural residence, referral status, inadequate antenatal care, vasopressor support, and blood transfusion were not significantly associated with mortality.

Discussion

The present retrospective study analysed the clinical profile, interventions, and outcomes of 100 obstetric patients admitted to the ICU of a tertiary care centre in north India. Our findings are broadly consistent with existing literature from India and other developing countries, while also revealing certain centre-specific patterns that warrant consideration.

The ICU admission rate in our study was comparable to that reported in other Indian tertiary care institutions. Nigeen et al. from Srinagar reported an admission rate of 1.02% [4], while Panda et al. from Varanasi found a rate of 1.84% [7], and Thakyal et al. from Jammu reported 1.28% [8]. Higher rates have been documented in studies from Bihar and Bhopal, where Anamika et al. and Parihar et al reported rates of 12.47% and 3.48% respectively, which probably reflect the increased burden of late referrals, limited access to peripheral healthcare and a larger catchment of patients [9,10]. Differences in admission rates point to the influence of institutional referral patterns, bed availability and the presence or absence of a high dependency unit (HDU) as an intermediate step prior to escalation to the ICU.

The mean age of patients in our study was 28.9 ± 6.7 years, which is consistent with other Indian studies. Miglani et al. in their north Indian cohort found the mean age to be 26.03 years6 while Gupta et al. found the mean age to be 26 ± 2.31 years [11]. This is a pointer towards the relatively young age at which women in India, particularly those from rural areas, have pregnancies and complications related to them. The predominance of rural patients (68%) and referred cases (65%) in our study mirrors national trends, where geographical inaccessibility, poor transport infrastructure, and delayed decision-making at peripheral centres contribute to late presentation [11] Parihar et al. similarly found that 83.9% of their ICU admissions were referred patients, suggesting that tertiary care obstetric ICUs in India predominantly serve as referral endpoints rather than primary admission destinations [10].

A significant finding was that 61% of our patients had inadequate antenatal care, which is a well-recognised risk factor for obstetric morbidity. Panda et al. reported similar findings, with 60.86% of their patients admitted to the ICU having inadequate ANC [7].

In our study 92% admissions were in postpartum period which is in concordance with the observations of Parihar et al. (79.31%) and Miglani et al. (88.7%). This could be due to haemodynamic changes after delivery, acute blood loss and unmasking of pre-existing conditions after childbirth [10,11]. The most common diagnosis for ICU admission was obstetric haemorrhage (32%), closely followed by hypertensive disorders of pregnancy (30%). Gupta et al. found that haemorrhage constituted 37.79% and hypertensive disorders 28.35% of all ICU admissions [6], while Thakyal et al. similarly reported haemorrhage in 38.08% and hypertensive disorders in 29.77% [8]. Nigeen et al. and Miglani et al. also confirmed obstetric haemorrhage and eclampsia/preeclampsia as the two dominant indications [4,11]. In contrast, some studies from Bhopal and Bihar identified hypertensive disorders as the primary indication, reflecting regional differences in disease burden and referral bias [9,10]. Sepsis accounted for 10% of our admissions, lower than the 9.91% to 17.39% [7,8], possibly reflecting differences in prevalence of septic abortions and intrauterine deaths with infection.

Among comorbid conditions, acute kidney injury (AKI) was present in 46% of our patients, representing one of the most striking findings. This markedly high prevalence likely reflects the downstream consequences of obstetric haemorrhage, sepsis, and severe preeclampsia, all of which compromise renal perfusion. A previous study found AKI in 33.1% of their critically ill obstetric patients, reinforcing its significance as a marker of disease severity [12]. Critically, AKI requiring renal replacement therapy was the only statistically significant predictor of maternal mortality in our study (p = 0.01), with 62.5% of women who died requiring dialysis compared to only 17.4% of survivors. This finding suggests that early identification and aggressive management of renal dysfunction may be a key determinant of maternal survival.

With respect to ICU interventions, blood transfusion was required in 69% of patients, reflecting the heavy burden of obstetric haemorrhage and coexisting anaemia. Vasopressor support was needed in 52% of cases, while mechanical ventilation was required in 34%. These figures are broadly comparable with other Indian studies; Gupta et al. reported mechanical ventilation in 38.58% and inotropic support in 50.39% [6], while Thakyal et al. documented similar rates [8]. Renal replacement therapy was administered to 21% of our patients, higher than in most comparable studies, consistent with the high burden of AKI in our cohort.

The maternal mortality rate in our study was 8%, with haemorrhagic shock as the leading cause of death (50%), followed by cardiorespiratory failure (25%). This compares favourably with rates reported from Bihar (11.12%) and Bhopal (11.95%), as well as the significantly higher rate of 34.78% from Varanasi [7,9,10]. In contrast, Nigeen et al. reported a similar mortality of 8% [4] and Gupta et al. reported 7.87%,[6] suggesting that dedicated obstetric ICU infrastructure at tertiary centres in north India has improved maternal survival compared to earlier data. Although mechanical ventilation, vasopressor use, and blood transfusion were more common among women who died, these associations were not statistically significant, possibly due to the small number of deaths (n=8) limiting statistical power.

Neonatal outcomes were also a matter of concern, with 20% stillbirths or intrauterine deaths and 22% of neonates requiring NICU admission. Nigeen et al. reported a fetal and neonatal mortality rate of 22%,[4] while Miglani et al. found perinatal mortality of 31.06% [11]. These figures collectively highlight that obstetric emergencies requiring ICU care carry significant risks not only for the mother but also for the fetus, emphasising the need for coordinated obstetric and neonatal critical care.

The findings of this study have important implications for healthcare planning. The high proportion of referred patients with inadequate ANC and rural residence points toward systemic failures at the primary and secondary care levels. Strengthening peripheral health infrastructure, improving antenatal care outreach, training frontline healthcare workers to identify danger signs, and establishing clear referral pathways are essential steps toward reducing the burden on tertiary ICUs and improving both maternal and neonatal outcomes.

Conclusion

Obstetric haemorrhage and hypertensive disorders of pregnancy remain the dominant indications for ICU admission at tertiary care centres in north India. Acute kidney injury is the most common comorbidity, and the requirement for renal replacement therapy is the strongest independent predictor of maternal mortality. The majority of patients admitted are rural, referred, and have had inadequate antenatal care, pointing to addressable systemic gaps. Strengthening primary antenatal care, improving timely referral pathways, and establishing dedicated obstetric ICUs with multidisciplinary teams are essential to reduce preventable maternal and neonatal morbidity and mortality.

Declarations

Ethics approval and consent to participate

Approved by the Institutional Ethics Committee.

Authors' Contributions

All author contributor equally.

Funding Statement

No funding and not a part of the employment of the authors.

Conflict of Interest

There is no conflict of interest among authors.

Acknowledgements

None