SIPHRE Shock Index in Pregnancy: Haemorrhage Risk Evaluation. Shock index as a predictor of adverse outcome in maternal haemorrhage

800 women die in pregnancy and childbirth each day, 99% of these occur in low-middle income countries (LMICs). Haemorrhage can cause death within a few hours if not recognised and managed quickly and is the leading cause of maternal death worldwide. Early detection and effective management rely on monitoring of vital signs , including heart rate and blood pressure. We have effective management to prevent death from bleeding, but most deaths or severe morbidities occur because of delayed or substandard care in the diagnosis and management; in many low- middle income countries (LMIC) vital signs are not measured due to lack of staff/ equipment or poorly functioning devices, leaving the recognition of compromised patients to crude visual estimation of blood loss alone or late signs such as collapse.

Our project is a validation study of a vital signs ratio called shock index (SI)- a simple calculation using heart rate divided by blood pressure. Our previous work on SI in pregnancy and haemorrhage suggests it is a superior way of recognising compromised patients. Since SI is simple to measure and easy to apply to daily practice, we hope to demonstrate it is an effective tool for early recognition of bleeding patients at risk of poor outcome (major surgery, major blood loss or death) with the aim to reduce maternal mortality and morbidity in LMIC.

The CRADLE VSA (CVSA) device accurately measures blood pressure and heart rate, displaying the results on a screen. It shows a red/amber/green traffic light warning system consistent with shock index thresholds to alert users to the patient's risk compromise, thus allowing prompt referral or escalation of management by even untrained health workers. The CVSA device has been designed and extensively validated by our team over 15 years to become the cheapest and most accurate device to measure BP at the high and low range we see in pregnancy. It is recognised as one of the top thirty high impact innovations in global health. It has been developed for use in LMIC being light, inexpensive, robust and easy to use. The use of the CVSA was extensively researched in the CRADLE 3 trial. In country data from Sierra Leone (SL) showed its implementation reduced maternal death rate by 60% (RR 0.37 CI 0.25 to 0.550). Following its success, the Department of Health in Sierra Leone are now rolling out the CRADLE VSA device nationwide, supported by DIFD, WHO, UNICEF and JICA.

We propose a prospective validation study of shock index in Sierra Leone, where adverse outcomes are so prevalent, in order to obtain large data to support SI as a tool for earlier recognition of compromise in bleeding mothers. SL has the world's highest mortality rate (1,165 deaths per 100,000 live births. In the UK this is 9 per 100,000). Data including first and worst vital signs and clinical outcomes will be collected on all pregnant/postpartum women who suffer with haemorrhage (blood loss over 500mls). The performance of shock index to predict adverse clinical outcomes will be evaluated and compared to existing vital signs scores.

A standardised early warning system for recognition of compromised patients does not exist for use in LMIC. Following validation on SI we propose a trial to evaluate the effect of SI in triage and management of bleeding patients in LIC. It is anticipated that a care bundle that incorporates SI as a triage/referral criteria will improve care. Since avoidable maternal deaths and near-misses are often related to delays in recognition of hemodynamic compromise this simple automated early warning system could be used in LIC worldwide.

We propose a validation of shock index (SI) as an earlier predictor of adverse outcome from obstetric haemorrhage in LIC. This data could support a future prospective trial to evaluate the use of SI in triage and management of bleeding patients. Since avoidable maternal deaths and near-misses are often related to delays in recognition of hemodynamic compromise it is anticipated that a care bundle that incorporates SI as a triage/referral criteria will improve care.

Our prior research has evaluated SI (ratio of HR/systolic BP) in obstetric haemorrhage and sepsis. A prospective study in South Africa validated previously determined thresholds: < 0.9 worked well as a rule out test, >0.9 indicated need for referral, and >1.7 indicated high chance of adverse outcome. Use of the CVSA (a vital signs measurement device which incorporates shock index thresholds) in management was associated with an 80% reduction in hysterectomy (Or 0.21, 95% CI 0.07- 0.66), and in- country analysis in Sierra Leone (SL) showed a reduction in maternal mortality by 60%. Use of SI appears promising- we seek to validate SI on a larger scale in a LIC, where adverse outcomes are more prevalent, to confirm mechanism of benefit. This allows us to trial its prospective use in management with the capacity to impact national and international guidance on the management of obstetric haemorrhage.

This is a prospective cohort study, supported by the Ministry of Health and Sanitation in SL, in the main maternity hospital (PCMH) and surrounding units. PCMH has around 7500 deliveries per year, with 95 maternal deaths in 2018. Local researchers will recruit patients diagnosed with obstetric haemorrhage, comparing vital signs at diagnosis and during their admission to outcomes. Area under the receiver-operator-curve predicts whether SI is a superior predictor of adverse outcome (hysterectomy, transfusion>4 units, critical care admission, Hb<70) compared to conventional vital signs.

The CVSA device is uniquely validated as a BP monitor for that is accurate in both hypertension and shock in pregnancy and in the non-pregnant population. Our data suggests SI can reduce maternal mortality and morbidity. It also has an role in diagnosing haemorrhagic shock in all specialties. The ambulance service in SL have expressed interest in the CVSA as its robust and portable design makes it suited for emergency situations.

SIPHRE will work alongside our existing project (SCALE CRADLE) that provides all government maternity facilities in SL with BP devices and adequate training. We propose that a validated vital signs device and accurate scoring system will reduce maternal complications from bleeding. A preliminary study demonstrated that SL has enough medical supplies to treat bleeding patients but suffers with insufficiently trained healthcare staff and inadequate monitoring. This project aims to bridge that gap through quick and effective triage of bleeding patients. Following this validation we aim to trial shock index pathways of management and eventually scale up in other LMIC including South Africa and Uganda.

Morbidity
Haemorrhage prolongs recovery time and hospital admission and increases complications impacting physical and mental health of patients and those around them. It is well recognized that neonatal and infant morbidity and mortality are inextricably linked with maternal health; infants of mothers who die are 3 to 10 times more likely to die within two years (3). Our objective is to support long-term sustainable change that will improve the lives of all women and children, including the most vulnerable and marginalized.
Reducing blood loss will address post partum anaemia and the health inequalities related to its treatment. In many LMIC where blood products are scarce, relatives must donate or it is purchased at cost to the patient, making it out of reach for the poor. Those with poor diet or antenatal care often have a lower baseline haemoglobin and therefore suffer more at similar blood loss thresholds. Use of SI is more useful in these anaemic patients than current practice which focuses on visual estimation of blood loss.

Health workers and policy makers
We will work to address the challenge of how to embed patient monitoring into the daily activities of healthcare workers in a method reproducible for wider scale up. We propose SI can make recognition of the unwell patient faster and easier- a triage tool that focuses care onto the patients that need it most. We have seen cases of inappropriate referrals using up scarce resources eg. ambulances, leading to complications including death. Reliable methods for recognition of compromise will allow for optimal resource allocation, improved local, regional and global decision making, and better planning of interventions.

Public Health
We have extensive longstanding networks throughout LMIC from our global health research which facilitates spread of new findings. Through these, use of the CVSA device has been recommended in the hypertension guidelines in South Africa. Through existing communication networks and data publication our findings we aim ti inform guidelines on management of bleeding based on shock indices to reduce maternal mortality.

Researchers
Lessons learnt from this trial will cement relationships built through longstanding support between KCL, the Ministry of Health, Welbodi Partnership, COMAHS and PCMH. It will provide methodological experience that can be emulated by others working on interventions in a global health context. We encourage multi-disciplinary working and employ junior researches who will benefit of the support of the in-country support in SL as well as the global academic experience of the PI. Our local researcher and UK co-ordinator will synchronise interests of policy makers, government bodies, health professionals, managers, patients and their families to work to achieve combined goals.