Intravenous fluid for adults with sepsis in sub-Saharan Africa: developing the design of an randomised controlled trial

Lead Research Organisation: Liverpool School of Tropical Medicine
Department Name: Clinical Sciences


Severe infections are a major global health problem, and are the leading cause of admission to hospital. The sources of infections are variable, but are commonly the lungs (pneumonia) and the gut (gastroenteritis). The body's usual protective responses should contain and destroy the bacteria or microbes causing infection. However, these responses can often become damaging: this is sepsis, and it is life threatening. During sepsis, a complex mix of abnormalities cause problems throughout the body, including with the immune system, breathing, blood circulation and kidneys. The flow of blood through vital organs can be compromised, causing further organ failure. Many patients die in the first few hours, before antibiotics have a chance to work. Improving early supportive care of patients with sepsis has been associated with significant improvements in patients' survival, but it is not certain which interventions work.
Much focus has been on early aggressive treatment with intravenous fluids, in order to raise low blood pressure and to improve blood supply to vital organs. However, fluids can cause "waterlogging" of the lungs (pulmonary oedema), of the kidneys (worsening kidney function), and other organs. This is particularly damaging where there are no intensive care facilities to "rescue" the patient. A large, well conducted trial in in sub-Saharan Africa showed that intravenous fluid given quickly for sepsis caused children to die. We do not know if the same could be expected in adults: current clinical guidelines for adults emphasise quick fluid "boluses" to try to bring blood pressure back to normal.

* What this development proposal would lead to
We believe that a randomised controlled trial of intravenous fluids in adults should answer the question: "In sub-Saharan Africa, which is the better treatment for sepsis: 1) quick, larger volume fluids given to try bring blood pressure back to normal or; 2) slower fluid given to prevent dehydration, but not designed to "correct" blood pressure.

* Objectives
Before this type of trial could start, we have to define carefully who should be enrolled, how many patients we could expect, and how they should be monitored during treatment. This development grant proposal will examine each of these areas, by systematically reviewing existing knowledge, and by collecting data from a representative sub-Saharan African hospital (Blantyre, Malawi).

* Why the trial is needed now and why in Malawi?
New guidelines for the diagnosis and management of sepsis have been published using data from US hospitals and Intensive Care Units. These used no data from Low Income settings, yet are promoted as internationally useful. Their generalisability to sub-Saharan Africa is poor, and the region must develop and act on clinical evidence which is relevant to areas with younger patient groups, high rates of HIV, and low healthcare spending.

* Stakeholders
Acute infection is a national research priority, and we will ensure that the Malawi Ministry of Health and the College of Medicine (University) have input into the design and chosen outcomes. We will also engage the public and clinical stakeholders to ensure that the trial will be acceptable and appropriate to the healthcare setting. We will build on expertise in Applied Health Research (CAHRD network), develop South-South collaborative networks, and enhance North-South support in trials design (Co-I: DM).

Technical Summary

Sepsis is a life threatening organ dysfunction caused by a dysregulated host response to infection. Globally, there are 31.5m cases per year; in-hospital mortality in Malawi is 22-50%. Supportive therapy is key to good outcomes. Intravenous (iv) fluids are used to support a cardiovascular system compromised by peripheral vasodilatation, insufficient cardiac output, and leakage of fluid to the interstitium. While fluids can raise blood pressure, they do not counter the circulatory dysfunction. In high incomes settings, aggressive intravenous fluid (and intensive care [ICU]) management is associated with reduced mortality (OR 0.64).

However, potentially beneficial adaptive responses to hypotension are abolished by high volumes of fluid, leading to acute respiratory distress syndrome, and worsening kidney function. This becomes apparent where ICU is not available. In LMIC, the only relevant trial in adults was stopped early due to high mortality. An uncontrolled Ugandan implementation study suggested benefit, but no dose-response effect above 1 litre in the first 6 hours (cf: 4.9l in a US study). Worryingly, the FEAST study from sub-Saharan Africa showed higher mortality in children treated with more aggressive fluid therapy.

A randomised clinical trial of iv fluid treatment strategy in adults is warranted, comparing conservative therapy versus a blood-pressure-target-driven approach. This development proposal will: 1) determine current practice and inform study protocols; 2) refine inclusion criteria and demonstrate feasibility of recruitment.

Using standard clinical pathways, and leveraging existing research structures, we will identify septic patients in the emergency department, intensively record their baseline and dynamic physiological responses and fluid administration over the first 6 hours, and record outcomes. A systematic review of fluid administration in research and clinical practice in LMIC will be performed.

Planned Impact

The major impact of this study will be the design of a definitive trial: a multicentre randomised controlled trial of intravenous fluids in sepsis relevant to sub-Saharan Africa would be the first of its kind, and as such would have widespread relevance at multiple levels: individual clinicians prescribing and patients receiving treatment; hospital administration in organising workflow for septic patients; governmental health policy, including cost-effectiveness data; international (WHO) policy and recommendations. A number of other immediate impacts should be expected to accrue from the development proposal.

* Clinical impact including public involvement
Clinical impact will be felt by engagement with medical training in the host hospital, and through incorporation of local nurses and clinicians into the research team. As an inherent part of the project, we will strengthen communication between primary care and secondary emergency facilities, using the existing channels and initiatives (such as the MLW "Chipatala robot" project for triage). Public engagement as detailed in "Engagement, communication and dissemination" will aim to promote early health-seeking behaviour for people with suspected infection. We will establish a dialogue with healthcare users to improve the integration of clinical research into the emergency department, and to maximise the acceptability of future large scale projects, particularly with respect to emergency consent.

* Enhancing research infrastructure impact
By focussing on research training in a supportive and well-established research environment, we will support Malawi to become a leading regional centre for investigation of acute medical care and sepsis in adults. The project impact will offer excellent value for money due to a close relationship with existing sepsis projects, and the clinical laboratory at MLW and CoM. Trials management advice is available from within our team, and will support the development of the local Clinical Research Support Unit.

* Policy and decision making impact
Existing connections with MLW partners strengthen decision making in health central government through knowledge synthesis and Health Economics expertise. We will make use of MLW commitment to developing an Evidence Informed Decision-making Centre (EvIDeNCe), and a national Evidence Informed Decision-making Network with key national stakeholders (EvIDeNt). These activities will allow early input of high-level stakeholders in a definitive trial. Internationally, our group is represented on the Global Intensive Care Working Group at the European Society of Intensive Care Medicine. This group have previously published guidelines on sepsis treatment in low-resource-settings; data from the project will inform future revisions of this guidance.
Sepsis is high on the international agenda, but due to its perception as a process rather than a disease, it is not visible or well represented on Global Burden of Disease estimates. We will raise this profile at international conferences, and in peer-reviewed journals.
Description Clinical Research Career Development Fellowship: "When is enough, enough? Physiological responses to fluid resuscitation in sub-Saharan African adults with sepsis"
Amount £813,000 (GBP)
Funding ID 211098 
Organisation Wellcome Trust 
Sector Charity/Non Profit
Country United Kingdom
Start 09/2018 
End 08/2021
Description NIHR Global Health Resaerch Group on Sepsis - the African Research Collaboration on Sepsis (ARCS)
Amount £2,000,000 (GBP)
Funding ID 17/63/42 
Organisation NIHR Evaluation, Trials and Studies Coordinating Centre (NETSCC) 
Sector Public
Country United Kingdom
Start 04/2018 
End 03/2021
Title Case descriptions of frequently-seen patient presentations in sepsis 
Description Collaboratively produced, these case descriptions define typical cases of patients presenting with sepsis. They will be used in conjunction with our systematic review to test the feasibility and applicability of fluids guidelines 
Type Of Material Improvements to research infrastructure 
Year Produced 2017 
Provided To Others? No  
Impact This has facilitated a more rounded view of sepsis guidelines, using a standardised set of patient presentations, we will be able to identify gaps in the given guidance, and point future guideline development groups to attend to them. 
Title Low cost fluid measurement device 
Description This is a bespoke electronic device to measure changes in intravenous fluid bag weight over time, recording this, and generating alarms where the delivery volume exceeds a preset value over time. It has been generated in collaboration with the University of Liverpool Electronic Engineering department. 
Type Of Material Physiological assessment or outcome measure 
Year Produced 2017 
Provided To Others? No  
Impact This tool is being feasibility tested at Aintree Hospital, Liverpool against a reference standard (pump). The potential implications are that resarch studies in resource limited settings could record minute-by-minute fluid administration at very low cost. The device is compatible will all fluid bags, allowing its use without special equipment, and being able to fit into widely different healthcare system circumstances. Circuit diagrams and instructions will be posted on an open-access site (GitHub) once beta testing is complete. 
Title Baobab-sourced model of outcome in severe illness 
Description Approximately 150,000 patient episodes attending a central hospital in Malawi with outcome data and entry physiological data. This is being used to understand the flow of patients, incidence of acute illness, and performance of triage systems in this population. 
Type Of Material Database/Collection of data 
Year Produced 2018 
Provided To Others? No  
Impact Manuscript in preparation. R shiny script available for local use during development. 
Title Example of Shiny presentation of physiological data to enhance triage decisions 
Description An openly available online tool to view the performance of early warning scores 
Type Of Material Data handling & control 
Year Produced 2019 
Provided To Others? Yes  
Impact Used for teaching purposes (both the presentation of information and the use of early warning scores) 
Description African Sepsis Alliance 
Organisation African Sepsis Alliance
Country Uganda 
Sector Charity/Non Profit 
PI Contribution PI has taken the position of treasurer in the newly formed, Africa based, non-governmental organisation, The PI has also contributed to the development of the constitution, and awareness of sepsis across Africa including presenting the capacity of research in Africa to alter outcomes in sepsis (African Federation of Critical Care Nurses).
Collaborator Contribution Provided a platform for discussion of sepsis research (as above). Ensured that sepsis research is well represented within the organisation with a specifically constituted sub-committee to allow co-ordination and dissemination of research ideas.
Impact Precipitated a research paper: Derivation and validation of a universal vital assessment (UVA) score: a tool for predicting mortality in adult hospitalised patients in sub-Saharan Africa Christopher C Moore, Riley Hazard, Kacie J Saulters, John Ainsworth, Susan A Adakun, Abdallah Amir, Ben Andrews, Mary Auma, Tim Baker, Patrick Banura, John A Crump, Martin P Grobusch, Michaëla A M Huson, Shevin T Jacob, Olamide D Jarrett, John Kellett, Shabir Lakhi, Albert Majwala, Martin Opio, Matthew P Rubach, Jamie Rylance, W Michael Scheld, John Schieffelin, Richard Ssekitoleko, India Wheeler, Laura E Barnes BMJ Global Health Jul 2017, 2 (2) e000344; DOI: 10.1136/bmjgh-2017-000344
Start Year 2017
Description Baobab Healthcare 
Organisation Baobab Healthcare Ltd
Country South Africa 
Sector Private 
PI Contribution None - pending results analysis
Collaborator Contribution We have engaged Baobab healthcare, who curate a database of patients presented to Queen Elizabeth Central hospital over the last 3 years. We are using this database (subject to an ethical review) to ensure data being collected within the project are representative of wider patient populations. Baobab have provided IT knowledge, including SQL search strategies.
Impact Results pending
Start Year 2018
Description Malawi Ministry of Health - Quality Management Unit partnership 
Organisation Government of Malawi
Country Malawi 
Sector Public 
PI Contribution We have reported results on prevalence of severe illness (and sepsis) in adults to the Quality Management Unit. This will inform the Malawi National Research Agenda, and ensures that acute illness is properly represented.
Collaborator Contribution Andrew Likaka (lead for the unit) has become an advisor on the African Research Collaboration on Sepsis, and has given direct and useful formative feedback during the set-up phase of this project. This includes the use of information technology in clinical trials, and the type of pilot programme required by the Ministry of Health to be useful for health policy.
Impact ARCS grant (see funding)
Start Year 2018
Description Neutrophil bead assay: Use in sepsis and concomitant HIV 
Organisation Liverpool School of Tropical Medicine
Country United Kingdom 
Sector Academic/University 
PI Contribution We have facilitated this project through common recruitment and use of staff to screen participants and enrol to the studies. We have acted as local (Malawi) lead for the project, including administrative support and ethical review.
Collaborator Contribution Research concept and assay have been developed by the collaborators. Training of local laboratory staff has been done through a site visit provided by the postdoctoral scientist form the project. SOPs have been contributed, and costs for the assay and sample collection are wholly funded by the collaborator.
Impact SOPs and assay QA procedures have been defined, and submitted for publication (under review). Ethics approval for sample collection has been given.
Start Year 2017
Description Deferred consent - qualitative study 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Public/other audiences
Results and Impact A series of formal group discussions with three groups have been undertaken in order to generate information about the acceptability of deferred consent for studies of critical illness (and sepsis): general public; healthcare practitioners; member of research ethics committees. Results are being formally analysed, and will be presented for publication. It has led to a requests for further training from the national ethics committee (including senior department of health staff). A reconsideration of national policy to deferred consent at ministry level has begun.
Year(s) Of Engagement Activity 2017,2018
Description Multidisciplinary critical care workshops 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Professional Practitioners
Results and Impact In collaboration with all of the hospital departments (at Queen Elizabeth Central Hospital, Blantyre), we have developed a series of interactive teaching sessions. Nurses, doctors, and untrained medical professionals have attended, with two objectives:1) improving interdepartmental working and referral e.g. obstetric patients with sepsis; 2) improving skills in recognition of critically unwell patients.
Year(s) Of Engagement Activity 2018
Description School visit (Blantyre, Malawi) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Schools
Results and Impact Supported multiple school visits to the Samala Moyo exhibition which hopes to demonstrate to school-aged Malawians that a career in clinical science is possible and rewarding. Participants and teachers report alot of enthusiasm after these events, with improved engagement with the curriculum.
Year(s) Of Engagement Activity 2018