Obstetric emergency pathway to Indian referral units

It does not establish that an obstetric emergency pathway functions reliably at district level.

A woman with eclampsia, haemorrhage, obstructed labour, or a previous caesarean section does not benefit from an average national indicator. Her outcome depends on a sequence of operational events: recognition of risk, clinical stabilisation, transport, referral acceptance, communication with the receiving facility, access to surgery, blood, anaesthesia, and newborn support. Failure at any point converts an institutional delivery into a transfer chain.

The emergency obstetric referral pathway in India is therefore not a single transport problem. It is a capacity-matching problem across facilities operating at unequal levels of readiness. Referral is appropriate when a lower-level facility identifies a complication beyond its mandate. It becomes hazardous when the next facility cannot deliver Comprehensive Emergency Obstetric Care, or when the transfer begins without a confirmed receiving team.

The anatomy of delay: referral is not one event

The Three Delays model remains useful because it separates maternal risk into three phases:

1. Delay in deciding to seek care. This includes delayed recognition of danger signs, household decision-making constraints, and financial uncertainty.

2. Delay in reaching an appropriate facility. This includes ambulance availability, road travel, transfer arrangements, and distance between facilities.

3. Delay in receiving adequate treatment after arrival. This includes triage, specialist availability, operating theatre access, blood storage, and neonatal support.

The model is often presented as a linear sequence. District referral data suggest a more complex pattern. A woman may reach a public facility promptly, be assessed, and still enter Phase II and Phase III delays repeatedly through onward referral. The first contact has occurred. Definitive treatment has not.

Studies of maternal deaths have found that 25% of deceased women were referred to three different facilities before death. Mean travel time between facilities ranged from one hour to 1.6 hours. These averages conceal the clinical significance of each transfer. In postpartum haemorrhage, an additional hour is not an administrative inconvenience. It is an interval in which blood loss, shock, and coagulopathy progress while the patient remains outside definitive care.

The practical pathway is better understood as five linked questions:

• Was the obstetric complication detected early enough?
• Could the initial facility stabilise the woman before transfer?
• Was a receiving facility identified and informed?
• Did transport depart without delay and with suitable clinical support?
• Could the receiving unit provide surgery, blood, anaesthesia, and newborn care immediately?

A “yes” to the first question does not compensate for a “no” to the last four. Referral documentation may record that a woman was transferred. It rarely captures whether the transfer produced timely definitive care.

A referral is not a clinical outcome. It is a handover between two capacity levels, and the second level must be real rather than nominal.

First Referral Units: designation is not functionality

A First Referral Unit, or FRU, is expected to provide round-the-clock Emergency Obstetric and Newborn Care. In operational terms, that includes the capacity for caesarean delivery, blood storage, newborn care, and the clinical workforce required to use those resources at any hour.

This standard places the FRU at the centre of rural pregnancy complications referral. Sub-centres and Primary Health Centres are designed to identify risk, provide basic care, and transfer cases that exceed their clinical scope. An FRU is intended to be the first facility able to convert referral into treatment.

The distinction between a formally designated unit and a functioning unit is decisive. A facility may have a building, labour room, beds, and a referral designation while lacking an anaesthetist, operative team, working theatre, blood storage, or neonatal capacity. In that setting, the patient is not arriving at the endpoint of the pathway. She is arriving at an intermediate stop.

Available facility evidence identifies several non-medical drivers of referral:

These figures require careful interpretation. They do not mean that every referral has one isolated cause. Obstetric readiness is composite. An operating theatre without anaesthesia is not a functioning surgical service. A caesarean capability without blood storage is incomplete in a haemorrhage-prone setting. A mother may survive surgery while her newborn requires care that the facility cannot provide.

The workforce gap reinforces the structural problem. Community Health Centres have recorded a 65.1% shortfall of obstetricians and gynaecologists. This does not imply that every CHC is unable to deliver emergency care. It does indicate that the system cannot assume uniform FRU functionality from facility type alone.

The operational implication is direct: maternal health referral guidelines must map actual service availability, not only facility classification. District control rooms and referring clinicians need current information on which unit can perform emergency caesarean section, has blood storage available, can manage severe pre-eclampsia, and can receive a compromised newborn. Static referral maps become unsafe when staffing and service availability vary by shift.

Referral rates reveal where clinical capacity is concentrated

Referral frequency is not automatically evidence of poor care. A facility with limited staff should refer cases it cannot safely manage. Low referral rates can also be misleading if complications are not detected or recorded. The more useful question is whether the pattern of referral matches the clinical capacity of the originating facility.

Evidence from nurse-run sub-centres and Primary Health Centres shows obstetric referral rates ranging from 14.3% to 36.3%. Doctor-run centres report substantially lower rates, from 2% to 7.5%. The difference is expected. It reflects variation in diagnostic authority, confidence in managing complications, staffing, and availability of immediate intervention.

A public-facility study in urban Maharashtra, covering 2016 to 2019, found that 14% of deliveries in Level I and II facilities resulted in referral to a higher facility. Pregnancy-induced hypertension and eclampsia accounted for 17% of referrals. Previous caesarean section accounted for 12%.

These are not marginal categories. Hypertensive disorders can deteriorate rapidly and may require magnesium sulphate, monitoring, urgent delivery, anaesthesia, and neonatal care. A previous caesarean section is not always an emergency, but it changes the risk profile of labour and requires access to surgical intervention if uterine rupture, fetal distress, or labour obstruction emerges.

The referral burden should therefore be stratified rather than viewed as a single percentage. A district dashboard needs, at minimum, the following categories:

• Clinical indication: haemorrhage, hypertensive disorder, sepsis, obstructed labour, fetal distress, previous caesarean section, or other indication.
• Originating facility type: sub-centre, PHC, CHC, urban facility, or private facility.
• Reason for referral: clinical escalation, unavailable theatre, absent specialist, unavailable blood, unavailable neonatal care, or patient preference.
• Transfer interval: time from referral decision to departure, travel duration, and time to treatment at the receiving unit.
• Final destination: whether the first receiving facility delivered definitive care or referred the patient onward.
• Maternal and newborn outcome: discharge status, surgery performed, intensive care requirement, stillbirth, neonatal admission, and mortality.

Without these variables, referral counts remain administrative totals. They cannot distinguish appropriate risk escalation from avoidable system failure.

High referral volume is not the primary signal of weakness. Repeated referral, uncommunicated transfer, and referral for absent basic capacity are the stronger signals.

The communication void between facilities

Transport is visible. Communication failure is less visible and often more consequential.

In fewer than half of obstetric emergency cases—47%—the referring facility initiated phone-based communication with the receiving facility. That leaves a large proportion of transfers in which the destination may receive a woman without prior clinical briefing, estimated arrival time, or confirmation that the required service is available.

A referral slip does not solve this problem. Paper documentation travels with the patient. It does not alert an operating theatre, mobilise blood, confirm bed capacity, or establish whether an anaesthetist is on site. In unstable obstetric cases, the receiving team needs clinical information before arrival.

A functional pre-referral communication sequence is short:

1. The originating clinician identifies the emergency and begins stabilisation within available capacity.

2. The clinician contacts the receiving FRU or higher facility directly.

3. The receiving facility confirms service readiness and identifies the point of entry.

4. The referring facility records the accepted destination, clinical indication, treatment already given, and departure time.

5. Transport departs with a defined escalation plan if the woman deteriorates.

6. The receiving facility logs arrival, treatment initiation, and whether onward referral occurred.

This is not a digital-health aspiration. It is a minimum coordination protocol. Phone-based communication has limitations: networks fail, staff change shifts, and receiving units may be overloaded. But no communication creates a known information deficit before a high-risk patient begins travel.

The consequences are most acute when referral is driven by facility absence rather than patient complexity. If the first hospital lacks an emergency theatre, the referral destination must be selected for operative readiness. If the first hospital lacks neonatal intensive care, the pathway must consider both maternal and newborn needs. These are different destination decisions. A generic instruction to “refer to higher centre” is operationally weak.

District systems also need feedback loops. Referring facilities should know whether the patient arrived, what intervention was performed, and whether the referral diagnosis was confirmed. Without feedback, clinical learning remains fragmented. Repeated referrals for hypertension, for example, may indicate late screening in antenatal care, inadequate availability of antihypertensive treatment, lack of magnesium sulphate, or absence of a facility able to terminate pregnancy safely. The referral register alone cannot identify which explanation applies.

Obstetric emergency transport: entitlement versus operational reach

India’s maternal health policy has progressively reduced direct financial barriers to facility-based care. Janani Suraksha Yojana began in 2005. Janani Shishu Suraksha Karyakram followed in 2011. The Surakshit Matritva Aashwasan initiative, known as SUMAN, was launched on 10 October 2019.

SUMAN guarantees zero-expense, dignified, and quality maternal and newborn care. Its stated package includes free transport from home to facility, inter-facility transfer, and drop-back services. The policy logic is sound. Emergency transport costs should not determine whether a woman follows referral advice.

Implementation is not equivalent to entitlement. Available evidence indicates that only 70% of women complied with referral advice, with transport and cost barriers contributing to non-compliance. Private transport and out-of-pocket spending remain part of the real pathway in some settings. The relevant measure is not whether free transport exists in policy documents. It is whether a vehicle arrives when a referral is issued, whether it can carry a woman safely, and whether it delivers her to a facility ready to treat her.

For district managers, transport performance should be measured against referral time intervals rather than fleet counts. A large ambulance inventory is not a maternal health indicator. The useful metrics are more specific:

• median time from referral decision to vehicle dispatch;
• median time from dispatch to facility departure;
• travel time by referral route and season;
• share of cases transported directly to a definitive care facility;
• proportion of inter-facility transfers accompanied by documented clinical handover;
• proportion of women who reach the first receiving facility but require a second referral.

No nationwide estimate is currently available for the share of obstetric emergencies that reach a Comprehensive Emergency Obstetric Care facility within a defined “golden hour.” That absence is itself material. Systems cannot optimise a time threshold they do not routinely measure.

From financial access to pathway accountability

The policy sequence from JSY to JSSK and SUMAN shows a change in emphasis. Earlier programmes concentrated heavily on increasing institutional delivery and reducing household expenditure. This was necessary. Maternal survival cannot improve where women deliver without access to skilled care.

The next constraint is quality at the point of referral. Institutional delivery is a process indicator. Emergency surgical readiness, blood access, clinical communication, and transfer completion are outcome-relevant conditions.

In June 2026, the Union Health Minister launched the SUMAN Roadmap 2030 alongside new National Ambulance Guidelines. The policy direction is aligned with the observed gaps: transport must be integrated with maternal and newborn service readiness rather than managed as an isolated logistical service.

Its effect cannot yet be assumed. The specific impact of the Roadmap 2030 on Phase II and Phase III delays has not been established in real-time data. The appropriate evaluation framework will need district-level baseline and follow-up measures, stratified by rurality, facility type, and referral indication.

Three indicators would provide a more credible test than headline transport coverage:

A fourth measure should track mortality and severe maternal morbidity among referred women separately from women admitted directly. This cohort is clinically different. Aggregating the two can obscure risk concentration within the referral pathway.

What a district-level pathway should do

The first question in emergency obstetric care is not whether a woman has been referred. It is whether the system has moved her to the right capability level without creating new delays.

A district pathway should begin with a live inventory of functional emergency services. It should then link each lower-level facility to a verified receiving unit by clinical indication and time of day. It should require pre-arrival contact, record transfer intervals, and audit every second referral. These are not additional reporting burdens if they replace incomplete referral counts with decision-relevant data.

India has demonstrated that maternal mortality can decline at national scale. The remaining challenge is more granular. The emergency pathway must be analysed as a chain of facilities, staff, transport decisions, and time intervals. A referral unit that cannot operate, a vehicle that departs without destination confirmation, or a patient transferred twice is not a minor service gap. It is a measurable exposure to maternal and newborn risk.

The policy implication is therefore narrow and testable: by 2030, maternal referral systems should be assessed not by the number of ambulances deployed or facilities designated, but by the proportion of obstetric emergencies that reach verified definitive care on the first transfer.