Maternal morbidity and mortality remain major public health challenges globally, particularly in low- and middle-income countries. Despite advancements in obstetric care, a subset of women still experience severe complications during pregnancy, childbirth, or the postpartum period, necessitating admission to the Intensive Care Unit (ICU). These ICU admissions provide a critical insight into the “near miss” events that could potentially result in maternal death if not managed promptly and effectively.

Obstetric admissions to the ICU are relatively infrequent but are often associated with severe life-threatening conditions such as hypertensive disorders, sepsis, haemorrhage, and cardiac complications. In a landmark study by Mahutte et al., obstetric ICU admissions accounted for a small proportion of total admissions but involved a high degree of morbidity and clinical complexity [1]. Similarly, a South African study underscored the burden of critical illness among obstetric patients, revealing key patterns in patient characteristics and outcomes [2].

Severity and outcomes of these ICU admissions are influenced by multiple factors including the timeliness of referral, availability of multidisciplinary care, and the presence of pre-existing medical conditions. Gilbert et al. found that most ICU admissions in obstetrics involve significant interventions and are linked with measurable outcomes, highlighting the need for robust ICU support in obstetric units [3].

The World Health Organization emphasizes that for every maternal death, 20–30 women suffer acute or chronic morbidity, with critical care support being a pivotal component in averting mortality in many such cases [4]. Studies like Waterstone et al. and Heinonen et al. further demonstrate the predictive factors and population-level need for maternal critical care, advocating for structured risk identification and intensive monitoring protocols [5][6].

In the South Asian context, where healthcare infrastructure is often constrained, ICU utilization for obstetric care is particularly telling. A study from Pakistan indicated high maternal morbidity in obstetric ICUs, with hypertensive disorders and sepsis among the most frequent causes of admission [7].

This study, conducted in the Government General Hospital, Siddipet, aims to assess the causes, characteristics, and outcomes of obstetric ICU admissions over a one-year period. The findings will contribute to a better understanding of critical care needs in obstetrics and guide clinical and administrative strategies to improve maternal health outcomes.

Aims and Objectives

The study aimed to evaluate the causes and clinical outcomes of obstetric ICU admissions at Government General Hospital, Siddipet. Specifically, it assessed the pregnancy status at the time of ICU admission (antepartum or postpartum), identified the comorbid conditions and complications that necessitated critical care, and examined clinical outcomes such as survival, mortality, and referral patterns. The relationship between MEOWS scores and patient outcomes was also analyzed.

Study Design and Setting

We conducted a retrospective observational study at the Department of Obstetrics and Gynaecology, Government Medical College and General Hospital, Siddipet. The study period extended from January 1, 2023, to December 31, 2023.

Inclusion and Exclusion Criteria

We included all pregnant women or women within six weeks postpartum who were admitted to the Intensive Care Unit (ICU) during the study period. Women admitted for non-obstetric reasons and cases with incomplete medical records were excluded from the analysis.

Data Collection and Variables

Clinical records were reviewed to extract relevant information. The data collected included:

• Demographic variables: Age, parity, gestational age at admission, booking status.
• Clinical variables: Antepartum or postpartum status, primary diagnosis, comorbidities, type of interventions (e.g., mechanical ventilation, surgery), MEOWS score, and duration of ICU stay.
• Outcome variables: ICU survival or mortality, referral to higher centres, discharge status.

Data were recorded at the time of ICU admission and during the course of ICU stay, including final outcome status.

Statistical Analysis

Data analysis was conducted using IBM SPSS Statistics version 29.0 and R version 4.3.3. Descriptive statistics were used to summarize the data. Continuous variables were expressed as means (±SD) or medians (IQR), and categorical variables as frequencies and percentages. The Chi-square or Fisher’s exact test was applied to analyze categorical variables. Independent t-test or Mann-Whitney U test was used for comparison of continuous variables. A p-value of <0.05 was considered statistically significant. Where applicable, logistic regression and ROC analysis were conducted to evaluate predictive associations.

Ethical Considerations

Institutional Scientific Review Committee and Ethics Committee approvals were obtained prior to data retrieval. Patient confidentiality and data integrity were maintained in accordance with institutional guidelines.

1. Baseline Characteristics of ICU-Admitted Obstetric Patients

A total of 38 obstetric patients required ICU admission during the study period. The mean age of patients was 30.05 ± 6.95 years, with an age range of 18 to 40 years. Regarding parity, 44.7% (n=17) were primigravida, 36.8% (n=14) had one prior delivery, and 18.4% (n=7) were multiparous (parity ≥2).

Of the ICU-admitted patients, 57.9% (n=22) were in the antepartum period at the time of admission, while 42.1% (n=16) were in the postpartum period (within six weeks after delivery). The mean gestational age at admission among antepartum patients was 32.6 ± 3.81 weeks. Among postpartum cases, the mean duration since delivery was 18.1 ± 11.19 days.

With respect to antenatal care, 73.7% (n=28) of the patients were booked, having received at least one documented antenatal visit, while 26.3% (n=10) were unbooked. A Chi-square test revealed no statistically significant difference in mortality between booked and unbooked patients (χ²(1) = 0.00, p = 1.000).

Table 1: Baseline Characteristics of ICU-Admitted Obstetric Patients

2. Distribution of ICU Admissions by Obstetric Status

Among the 38 obstetric ICU admissions, 22 patients (57.9%) were in the antepartum period, while 16 patients (42.1%) were admitted during the postpartum period. The distribution was statistically analyzed using a Chi-square goodness-of-fit test, assuming an equal expected distribution between the two groups.

Chi-square statistic: χ²(1) = 0.21
p-value: 0.645

This difference was not statistically significant, indicating that the observed variation in ICU admissions between antepartum and postpartum status may be due to chance.

3. Clinical Diagnoses and ICU Interventions

The most common diagnosis among ICU-admitted obstetric patients was preeclampsia/eclampsia, accounting for 31.6% (n=12) of cases. Other major indications included obstetric hemorrhage (26.3%, n=10), sepsis (18.4%, n=7), severe anemia (13.2%, n=5), and cardiac disease (10.5%, n=4).

A Fisher’s exact test was used to compare diagnosis categories between survivors and non-survivors. None of the individual diagnosis categories showed a statistically significant association with mortality.

Table 2: Primary Diagnoses Leading to ICU Admission

Table 3: Association between ICU Interventions and Mortality

4. MEOWS Scores and Clinical Outcome Correlation

The Modified Early Obstetric Warning Score (MEOWS) at ICU admission ranged from 3 to 13, with a mean of 6.66 ± 2.17. Based on commonly used thresholds, patients were categorized as follows:
- Low (≤4): 5 patients (13.2%)
- Moderate (5–7): 21 patients (55.3%)
- High (≥8): 12 patients (31.6%)

The distribution of MEOWS scores was compared between survivors and non-survivors using the Mann–Whitney U test, which yielded a U statistic of 120.0 and a p-value of 0.674. No statistically significant difference was observed.

To assess the predictive ability of MEOWS scores for mortality, a Receiver Operating Characteristic (ROC) curve analysis was conducted. The Area Under the Curve (AUC) was 0.45, indicating poor discriminatory ability.

Table 4: MEOWS Score Categories at ICU Admission

5. ICU Interventions and Duration of Stay

Among the 38 obstetric patients admitted to the ICU, mechanical ventilation was required in 44.7% (n=17) of cases, while surgical intervention was performed in 31.6% (n=12). Additionally, inotropic support was administered to 26.3% (n=10) of patients, reflecting the severity of maternal conditions requiring multi-modal critical care. The overall median duration of ICU stay was 3 days, with a range spanning from 1 to 9 days.

Table 5: Median ICU Stay by Intervention Type

A Kruskal–Wallis test was used to evaluate differences in ICU stay across intervention groups:
- Mechanical ventilation: H = 0.10, p = 0.754
- Surgery: H = 0.05, p = 0.823
- Inotropes: H = 0.21, p = 0.649
None of the differences were statistically significant.

Chi-square tests were conducted to assess associations between each intervention and patient outcomes (survival vs. mortality):
- Mechanical ventilation: χ²(1) = 0.28, p = 0.595
- Surgery: χ²(1) = 0.00, p = 1.000
- Inotropes: χ²(1) = 1.11, p = 0.292

6. Outcomes: Survival, Mortality, and Referral

Of the 38 obstetric patients admitted to the ICU, 31 (81.6%) survived, while 7 patients (18.4%) died during the ICU stay. Additionally, 2 patients (5.3%) required referral to a higher-level center for advanced care.
To explore potential associations between clinical outcomes and various risk indicators, patient outcomes were stratified by primary diagnosis, ICU interventions, and MEOWS score categories. Appropriate statistical tests, including Chi-square, Fisher’s Exact Test, and logistic regression, were applied to assess relationships between these factors and mortality.

6.1 Stratification by Primary Diagnosis

6.2 Stratification by Interventions

6.3 Stratification by MEOWS Score Categories

Chi-square test showed no significant association between MEOWS score category and mortality (χ²(2) = 1.81, p = 0.404).

6.4 Multivariate Logistic Regression

A multivariate logistic regression model was constructed to evaluate whether MEOWS score and interventions jointly predicted mortality.

ICU Admission Rate and Mortality

In this retrospective study conducted at a government teaching hospital, we observed an ICU admission rate of 7.6 per 1000 obstetric admissions (38/5009), which is consistent with other reports from low- and middle-income countries (LMICs). For example, Veerabhadrappa et al. [8] documented an ICU admission rate of 9.4 per 1000 obstetric admissions in a South Indian tertiary centre, while Anane-Fenin et al. [10] reported 8.1 per 1000 over a ten-year period in Ghana. Our observed maternal ICU mortality rate of 18.4% (7/38) is within the range reported in similar resource-limited settings: 16% in Sierra Leone (Marotta et al. [12]) and 20% in Malawi (Prin et al. [14]). However, this contrasts with lower mortality figures (<10%) in resource-rich environments [15, 20], underlining the persistent disparities in maternal critical care outcomes across settings.

Patient Demographics and Booking Status

The mean age of ICU-admitted patients in our study was 30.05 ± 6.95 years, aligning with the profile reported by Pattanaik et al. [9] (mean age: 28.9 years) and Abrar et al. [18], who also found high ICU risk among women aged 25–35. A majority (73.7%) were booked cases, suggesting that while antenatal contact is crucial, it does not always prevent complications requiring critical care. This proportion differs from studies in ultra-low-resource zones like rural Malawi, where >60% of ICU cases are unbooked [14].

Primary Diagnoses Leading to ICU Admission

Hypertensive disorders of pregnancy (31.6%) and obstetric haemorrhage (26.3%) were the most common reasons for ICU admission in our cohort, comparable to findings from Anane-Fenin et al. [10] (preeclampsia/eclampsia: 29.2%, haemorrhage: 25.4%) and Hofmeyr et al. [13], who cited hypertensive and haemorrhagic complications as global leading causes of maternal critical illness. Sepsis was the third most frequent diagnosis (18.4%), similar to rates reported by Marotta et al. [12] and Putoto et al. [17], highlighting ongoing infection control challenges in obstetric practice.

Fisher's exact test showed no statistically significant association between diagnosis and mortality in our cohort (p > 0.05 for all diagnoses). However, the odds ratio for mortality in patients with preeclampsia/eclampsia was 3.30, suggesting a trend toward increased risk, though this did not reach statistical significance (p = 0.395).

ICU Interventions and Their Outcomes

A substantial proportion of patients required mechanical ventilation (44.7%) or surgical intervention (31.6%), consistent with findings from Veerabhadrappa et al. [8], where 42% required ventilation. Use of inotropes in 26.3% of patients indicates high acuity, and is notably higher than the 15% reported by Prin et al. [14].

None of the interventions (ventilation, surgery, or inotropes) were statistically associated with mortality in bivariate analysis (p > 0.05 for all), although inotrope use showed a higher odd of mortality (χ²(1) = 1.11, p = 0.292). Kruskal-Wallis tests indicated no significant difference in ICU length of stay based on any single intervention (e.g., mechanical ventilation: H = 0.10, p = 0.754), suggesting that duration may be influenced by multifactorial complexity rather than a single modality.

Predictive Utility of MEOWS Score

The MEOWS score at admission had a mean of 6.66 ± 2.17, with 31.6% of patients classified as high-risk (score ≥8). However, the MEOWS did not significantly correlate with mortality in either univariate (Mann–Whitney U = 120.0, p = 0.674) or multivariate logistic regression (p = 0.848 for MEOWS). The ROC AUC = 0.45, confirming poor discriminatory ability in this cohort.

This contrasts with validation studies such as Kheswa et al. [20], who found an AUC of 0.72 for MEOWS in South African ICUs, and suggests that either the MEOWS threshold needs local calibration, or additional parameters (e.g., lactate, organ dysfunction) may improve prediction. The lack of predictive power in our study is also echoed by findings from Marotta et al. [17], who noted similar limitations of early warning scores in high-dependency obstetric units.

Referral and Resource Limitations

Only 2 patients (5.3%) were referred to higher centres, likely reflecting the infrastructural limitations of tiered referral in many government systems. As discussed by Vasco et al. [11] and Rojas-Suarez et al. [15], ICU triage and escalation remain system-level challenges in LMICs. Hofmeyr et al. [13] emphasize that improving resource deployment, task shifting, and guideline adherence can meaningfully reduce maternal morbidity and mortality even in low-resource contexts.

Limitations

This study was conducted at a single-centre government hospital, which may limit the generalizability of findings to other settings with differing levels of infrastructure or referral capacity. The retrospective design depended on the completeness and accuracy of medical records, introducing potential data omissions or misclassification. Additionally, the small sample size (n = 38) limited the power to detect statistically significant associations, particularly in subgroup analyses and multivariate models. Finally, some variables such as lactate levels, APACHE II scores, and long-term maternal outcomes were not available, which may have further enriched the analysis.

This study provides a focused snapshot of the causes, interventions, and outcomes among obstetric patients admitted to a government hospital ICU in a low-resource setting. Hypertensive disorders, haemorrhage, and sepsis emerged as the predominant indications for critical care, reflecting patterns observed across other LMIC-based studies. Despite the relatively high utilization of mechanical ventilation and surgical interventions, overall survival was 81.6%, and the maternal ICU mortality rate was 18.4%.

Importantly, commonly used clinical tools such as the MEOWS score demonstrated limited predictive value in this cohort, with no significant correlation to mortality. This underscores the need for locally validated, context-sensitive triage and monitoring systems.

Strengthening early recognition, timely referral, and critical care infrastructure—combined with regular outcome audits—remains essential to improving maternal survival in resource-constrained public health systems.

1. Mahutte NG,Murphy-Kaulbeck L,Le Q,Solomon J,Benjamin A,Boyd ME.obstetric admissions to the intensive care unit.Obstet Gynecol. 1999,94:263-6
2. Platteau P,Engelhardt T,Moodley J,Muckart Obstetric and gynaecologocal patients in an intensive care unit:A 1 year review.Trop Doct.1997:27:202-6
3. Gilbert TT,Smulian JC,Martin AA,etal.Obstetric admissions to the intensive care unit:Outcomes and severity of illness.Obstet Gynecol.2003;102:897_903
4. WHO Statistical information system(WHOSIS). WHO website: htpp://www.who.int/whisis/indicators/maternalmortratio/en/.Accessed Sep 2008
5. WaterstoneM,Bewley S,Wolfe C.Incidence and predictors of severe obstetric morbidity: case control study.BMJ2001; 1089-1094
6. Heinonen S,Tyrvainen E,Saarikoski S,Ruokonen E,Need for maternal critical care in obstetrics:A population based analysis.Int J Obstet Nesth 2002:1:260-4
7. Qureshi R, Ahmed SI,Raza A,Khurshid A,Chishti U.Obstetric patients in intensive care unit: Perspective from a teaching hospital in Pakistan.JRSM Open 2016:7:2054270416663569.
8. Veerabhadrappa VK,Shivanagappa M Mahadevaiah M,Srikanth SM et.al. Maternal outcome in obstetric ICU and HDU: a study from a teaching hospital in south india.Int. J Reprod Contacept Obstet Gynaecol 2019;8:862-8.
9. Pattanaik T, Samal S,Behuria S.Obstetric admissions to the intensive care units:A five year review,International Journal of reproduction,Contraception,Obstetric and Gynaecology.2015 Dec;6:1914_1917
10. Anane-Fenin, B., Agbeno, E. K., Osarfo, J., Opoku Anning, D. A., Boateng, A. S., Ken-Amoah, S., ... & Jeffery, E. (2021). A ten-year review of indications and outcomes of obstetric admissions to an intensive care unit in a low-resource country. Plos one, 16(12), e0261974.
11. Vasco, M., Pandya, S., Van Dyk, D., Bishop, D. G., Wise, R., & Dyer, R. A. (2019). Maternal critical care in resource-limited settings. Narrative review. International journal of obstetric anesthesia, 37, 86-95.
12. Marotta, C., Pisani, L., Di Gennaro, F., Cavallin, F., Bah, S., Pisani, V., ... & Putoto, G. (2020). Epidemiology, outcomes, and risk factors for mortality in critically ill women admitted to an obstetric high-dependency unit in Sierra Leone. The American journal of tropical medicine and hygiene, 103(5), 2142.
13. Hofmeyr, G. J., Haws, R. A., Bergström, S., Lee, A. C., Okong, P., Darmstadt, G. L., ... & Lawn, J. E. (2009). Obstetric care in low-resource settings: what, who, and how to overcome challenges to scale up?. International Journal of Gynecology & Obstetrics, 107, S21-S45.
14. Prin, M., Kadyaudzu, C., Aagaard, K., & Charles, A. (2019). Obstetric admissions and outcomes in an intensive care unit in Malawi. International journal of obstetric anesthesia, 39, 99-104.
15. Rojas-Suarez, J., & Paruk, F. (2024). Maternal high-care and intensive care units in low-and middle-income countries. Best Practice & Research Clinical Obstetrics & Gynaecology, 93, 102474.
16. Abie, A., Getie Mehari, M., Eseyneh Dagnew, T., Mebrat Delie, A., Melese, M., Workie Limenh, L., ... & Hailu, M. (2025). Obstetric admission and maternal mortality in the intensive care unit in Africa: A systematic review and meta-analysis. Plos one, 20(4), e0320254.
17. Marotta, C., Pisani, L., Di Gennaro, F., Cavallin, F., Bah, S., Pisani, V., ... & Putoto, G. (2020). Epidemiology, Outcomes and Risk Factors for Mortality in Critically–Ill Women Admitted to an Obstetric High Dependency Unit in a Resource–Limited Setting.
18. Abrar, S., Abrar, T., Sayyed, E., & Naqvi, S. A. (2022). Ruptured uterus: Frequency, risk factors and feto-maternal outcome: Current scenario in a low-resource setup. Plos one, 17(4), e0266062.
19. Poli, P. A., Orang’o, E. O., Mwangi, A., & Barasa, F. A. (2020). Factors related to maternal adverse outcomes in pregnant women with cardiac disease in low-resource settings. European Cardiology Review, 15, e68.
20. Kheswa, N., Rodseth, R., Wise, R., & Bishop, D. G. (2024). Predictive scores for critically ill obstetric patients in a resource-limited setting: A retrospective validation of the Obstetric Early Warning Score. South African Journal of Obstetrics and Gynaecology, 30(1), 16-20.