Background
Approximately 800 women die in pregnancy or childbirth every day [
1]. Obstetric haemorrhage (antepartum and postpartum), sepsis and hypertension in pregnancy account for more than 50% of maternal deaths worldwide, 99% of which occur in low- and middle-income countries (LMICs) [
1]. Early detection and effective management of these conditions relies on vital sign monitoring, including pulse and blood pressure [
2]. Strategies aiming to improve the detection of haemodynamic instability are vital in recognising women in need of urgent medical care.
There are simple, cost-effective and established interventions to save lives for each of these pregnancy-related conditions, but the barrier to initiating treatment lies in recognition of maternal haemodynamic compromise and access to interventions [
3,
4]. Accurate blood pressure measurement is essential for detecting and monitoring pre-eclampsia, enabling antihypertensive and prophylactic anticonvulsant therapy, appropriate transfer to higher-care facilities for timed delivery and, thereby, prevention of maternal and perinatal mortality and morbidity; however, the majority of blood pressure devices that are commercially available either fail validation in pregnancy or their accuracy is unknown [
5]. In LMICs, hypertension is frequently under-detected, not only because of poor availability of working and accurate blood pressure devices, but also because of inadequate training [
6,
7]. Vital sign measurement is equally critical in the management of obstetric haemorrhage and sepsis [
2]. Postpartum haemorrhage, the leading cause of maternal mortality in LMICs [
1], can cause death within a few hours, but effective interventions are available and early recognition facilitating immediate intervention is often lifesaving [
8].
The CRADLE Research Group proposes that the introduction of the CRADLE Vital Sign Alert (VSA) device, a vital-sign-measuring device and alert tool, integrated with the CRADLE training package, will aid the early recognition of haemodynamic instability secondary to haemorrhage, sepsis and pre-eclampsia in LMICs. The CRADLE VSA device has been validated as accurate for use in pregnancy, including in women with pre-eclampsia [
9] and is the first device to be validated as accurate in pregnant women with low blood pressure; from example, from haemorrhage or sepsis [
10]. It has been designed to be robust and longer lasting for use in community and health facility environments in LMIC settings [
9,
10]. The device is affordable, easy-to-use (including for community healthcare providers), reliable and portable. The CRADLE Group has incorporated a traffic-light early warning system into the device alerting users to hypertension and shock.
This protocol describes the stepped-wedge cluster-randomised controlled trial to assess the implementation and clinical usefulness of the CRADLE intervention in LMICs.
Discussion
We have described the protocol for a mixed-methods, multicentre, stepped-wedge cluster-randomised controlled trial. The main objective of this trial is to determine whether the introduction of the CRADLE intervention (CRADLE VSA device and CRADLE training package) to maternity care settings in LMIC communities and facilities reduces maternal mortality and morbidity. A stepped-wedge cluster-randomised controlled trial design has been chosen to evaluate the intervention in a pragmatic fashion. Individual randomisation would be logistically difficult and would not measure impact and transferability at a population level. The stepped-wedge design is useful where phased implementation is preferable because of logistical and practical constraints.
The CRADLE intervention is intended to be beneficial to all types of practice in a wide variety of settings. In keeping with the pragmatic trial design, and to ensure that results are generalisable, efforts were made to ensure that participating clusters represented diverse settings, at both a country level with different healthcare systems and at a facility level. Our study clusters, therefore, included academic/specialist facilities, private facilities and primary and secondary-level government facilities. We hypothesise that the intervention will work by facilitating earlier detection, referral and treatment of pregnancy complications.
A composite of maternal mortality or morbidity (one of maternal death, eclampsia or emergency hysterectomy with no double counting) was chosen as the primary outcome as powering a trial for maternal death alone would require a prohibitively large sample size. These are robust and meaningful clinical endpoints that are unambiguous and, therefore, feasible to collect in these environments. Maternal complications were selected, each of which is associated with significant acute and chronic morbidity. Therefore, a reduction in this composite would be highly desirable. Each component of the primary outcome will be individually interrogated to ensure no paradoxical effects within the composite, i.e. an increase in morbidity may be associated with a reduction in mortality, as a consequence of intervention. Therefore, should the CRADLE intervention result in appropriate referral and intervention (e.g. hysterectomy) this will be recognised as beneficial even though the overall effect on the composite outcome may be neutral. We have ensured that all primary outcome measures are feasible to collect at primary to tertiary facilities, as determined through the feasibility study. It is, therefore, not necessary to conduct costly household surveys. In the unlikely event that a primary outcome occurs without HCP contact, this is likely to be unchanged pre and post intervention and should not affect our ability to assess the impact of intervention implementation.
It is recognised that the CRADLE intervention may reduce perinatal mortality and morbidity. Perinatal outcomes have not been included as primary outcomes as the intervention is designed specifically to identify maternal health complications. Many of these occur postpartum and will not directly influence perinatal outcomes. Acquisition of detailed perinatal data within LMIC settings would be a substantial additional cost. However, the secondary perinatal outcomes will be collected in women who experience a primary outcome.
The potential for this trial to demonstrate reductions in maternal mortality and morbidity and a positive impact on the working lives of HCPs will be of interest to local, national and international health-policy makers. The CRADLE VSA could become a key tool in achieving the post-2015 global maternal health goals, as well as facilitating a basic recommended standard of care, i.e. accurate blood pressure determination in all pregnant women. Given the low cost, reliability and accuracy of the device, if the trial is successful, a low regulatory hurdle and rapid progression to adoption is anticipated. The trial results will be generalisable beyond the immediate research setting, as the trial will be carried out in a variety of countries and at both at community and facility levels. A robust scale-up strategy will be required to support international scale-up to allow widespread and equitable access to the innovation. This will require appropriate resourcing to include commercial expertise and facilitate partnership between inventors, manufacturers and governments to maximise effective and efficient marketing, supply and distribution through appropriate channels and fair pricing.
Acknowledgements
John Hopkins Public Health and implementation department, JHPIEGO, contributed US$10,000 towards the development of the training videos and provided academic input and review during their development.
CRADLE-3 Trial Collaborative Group:
Natasha L Hezelgrave, King’s College London, UK
Umesh Charantimath, KLE University’s J N Medical College, India
Chandrappa C Karadiguddi, KLE University’s J N Medical College, India
Sphoorthi S Mastiholi, KLE University’s J N Medical College, India
Geetanjali M Mungarwadi, KLE University’s J N Medical College, India
Feiruz Surur, St Paul’s Millennium Medical School, Ethiopia
Lomi Yadeta, Maternity Worldwide, Ethiopia
Yonas Guchale, Maternity Worldwide, Ethiopia
Violet Mambo, University of Zimbabwe, Zimbabwe
Sebastian Chinkoyo, Ndola Teaching Hospital, Zambia
Thokozile Musonda, Ndola Teaching Hospital, Zambia
Christine Jere, Ndola Teaching Hospital, Zambia
Bellington Vwalika, University of Zambia, Zambia
Mercy Kopeka, University Teaching Hospital, Zambia
Martina Chima, University Teaching Hospital, Zambia
Josephine Miti, University Teaching Hospital, Zambia
Rebecca Best, Welbodi Partnership, Sierra Leone
Matthew Clarke, Welbodi Partnership, Sierra Leone
Jesse Kamara, Welbodi Partnership, Sierra Leone
Jeneba Conteh, Princess Christian Maternity Hospital, Sierra Leone
Patricia Sandi, Princess Christian Maternity Hospital, Sierra Leone
Margaret Sesay, Princess Christian Maternity Hospital, Sierra Leone
Fatmata Momodou, Princess Christian Maternity Hospital, Sierra Leone
Julius Wandabwa, Busitema University, Uganda
James Ditai, Sanyu Africa Research Institute, Uganda
Nathan Mackayi Odeke, Sanyu Africa Research Institute, Uganda
Annettee Nakimuli, Makerere University College of Health Sciences, Uganda
Josaphat Byamugisha, Makerere University College of Health Sciences, Uganda
Dorothy Namakula, Mulago National Referral Hospital, Uganda
Noela Kalyowa, Mulago National Referral Hospital, Uganda
Doreen Birungi, Mulago National Referral Hospital, Uganda
Emily Nakirijja, Mulago National Referral Hospital, Uganda
Carwyn Hill, Hope Health Action, Haiti
Grace Greene, Hope Health Action, Haiti
Adeline Vixama, Hope Health Action, Haiti
Paul Toussaint, Hospital Convention Baptiste d’Haiti, Haiti
Grace Makonyola, Maternity Worldwide, Malawi
Doreen Bukani, Maternity Worldwide, Malawi
Monice Kachinjika, Maternity Worldwide, Malawi
Jane Makwakwa, Maternity Worldwide, Malawi
Trial sponsor
King’s College London (contact: Professor Andrew Shennan)