DISCUSSION
In the present study, which was done over a period of 18 months, incidence of ICU admission was 0.77% which is comparable to the results of Begum and Padmavat8. However, Ozumba et al9 in 2018 in their study found a higher incidence of 1.7% probably due to different admission criteria or due to a large catchment area.
It is obvious that booked status of the patients is associated with better outcome as is also reflected in the present study. In contrast to the present study, Joseph et al10 found increased ICU admission among booked cases. Probably this was due to the referral of these booked cases in complicated stages.10
In the present study, referred cases comprised the major part (55.6%) of the ICU admissions and their outcomes were significantly worse than the direct admissions.
This may be attributed to the fact that the patients who are referred are generally high risk or in critical condition as they could not be managed at the periphery hospital. If this factor is also associated with delay in the transport then the outcome of the referred patients further deteriorates.
Maternal health behavior varies with socioeconomic status and it is also affected by education level. Low socioeconomic status is usually associated with low education status, poor health seeking behavior, unintended and unplanned pregnancy, inadequate antenatal visits, lack of knowledge regarding available health facility benefits and their cultural stigma. These issues are clubbed with logistic problems like inaccessibility to health care and form a vicious circle in this strata. The same is clearly reflected in the present study where patients of lower socioeconomic strata had worse outcomes. Concordant results were found by Panda et al, 64.13% of the patients admitted to ICU in their study belonged to lower socioeconomic status.11
To improve the medical care in obstetric emergency, time is a crucial factor in life threatening conditions. Delay at any level worsens the prognosis of patients because life-threatening conditions may develop without any warning and require prompt treatment.
In the present study, first level delay was present in 65.4% cases admitted in ICU.
Similar to present study, Ghumare et al found that 27% delays were at the first level. In 19% cases mixed delay was present.12 Kumari K et al also observed that first delay was present in 81.8% of cases and level 1 delay was the most common delay found.13
In the present study most, common delay found was the delay at level 1, which is the delay in deciding to seek care. It was mainly due to socioeconomic and cultural characteristics. They were also not educated about the warning sign of any complication by the front-line provider (ASHA) in some cases. An accredited social health activist (ASHA ) is a community health worker instituted by the government of India’s Ministry of Health and Family Welfare (MOHFW) as a part of the National Rural Health Mission (NRHM). Their antenatal visits were limited to nutritional supplements and a general check up at the ground level, which might have missed preexisting medical and also obstetrical complications like malpresentation and cephalopelvic disproportion.
The hesitation to seek health care was compounded by cultural taboos and gender bias which further increased the duration of delay. The outcomes worsened significantly with increasing duration of level 1 delay. This again emphasizes the importance of intervention in the first few golden hours.
Second level delay is the delay in reaching the appropriate health care facility and in the present study, it was present in 50% cases.
Similar to our study Kumari K, et al also found that contribution of second level delay was present in 54.5% cases.13
Second level delay of <4hr is present in 33.1% which was mainly due to geographic distribution of referral center, cost of transportation and unavailability of transport. Second level delay of >4hr was found in 16.9% and was associated with delayed decision of referral by peripheral hospital. Main reasons of referral were non availability of NICU, blood bank, ICU facility and facility of cesarean section (lack of functional operation theatre or trained personnel). This results in lack of active intervention in the first few golden hours thus worsening the condition of the patients. Increased duration of level 2 delay was significantly associated with worse outcome.
Third level delay was present in 9.7% cases in the present study. Presence of third level delay in our tertiary care hospital could probably be explained by disproportionate infrastructural facility in comparison to the patient load. (only single functional emergency operation theatre, no availability of dedicated obstetric ICU facility, a smaller number of beds available in general ICU).
In contrast to present study, Ghumare et al found that third level delay was present in 21% cases12 and Kumari K et al also found third level delay in 45.5% cases.13 It was higher than the present study which may have been contributed by superadded effect of inadequate specialist services and inadequate blood component transfusion facility.12,13
Presence of any delay was significantly associated with worse outcomes. In a multicentric cross sectional study done by Pacagnella et al, any type of delay was observed in 53.8% of subjects and there was positive association between the presence of any delay and severity of maternal outcome.14Kumari K et al observed that most of the deaths were associated with multiple levels of delay.13
It was observed that the most common indication for ICU admission was obstetrical hemorrhage (37.1%), followed by hypertensive disorder of pregnancy (25.8%). In concordance with the present study, Sodhi et al15 and Joseph et al10 had similar results but Ozumba et al 9 found that rupture uterus was the most common indication of ICU admission. Uterine rupture has been remarkably eliminated in most parts of the world but probably low socioeconomic status and poor health-seeking behavior of the subjects in the study (South Ease Nigeria) contributed to this finding.9
Mean APACHE II score was 14.77±6.85, According to this predicted mortality was 25% but observed mortality was found to be higher i.e. 30.6%. %. This was probably due to the infrastructure and logistic constraints of our study area, which is a government organization. There was significant (p<0.01) difference in APACHE II score in the presence of 1st and 2nd level delay. APACHE II score was significantly (p<0.05) high in patients having delay of ≥24 hrs. Also, APACHE II score was significantly higher in the presence of 2nd level delay and it was significantly increased as the duration of delay increased (p=0.0001).
In contrast to the present study, Sodhi et al found observed mortality rate (OMR) to be too low as compared to the predicted mortality of 24%.15 This variability can probably be explained as this study was conducted in a private hospital having most of the modern equipment. Our study was conducted in a government setup having limited and conventional resources.
Strengths: It is one of the few studies conducted in a tertiary center of northern India correlating all three levels of delay to fetomaternal outcome thus allowing a genuine root cause analysis at the ground level.