Sepsis Awareness Month: Q&A with Dr Ben Morton

Effective sepsis treatment requires source control (for example, removal of an infected appendix), antimicrobials (to kill the pathogen), and organ support (e.g., fluids, oxygen, and sometimes critical care).

It is important to diagnose and treat sepsis quickly – every hour of delay increases the risk of death.

Diagnosis and treatment are resource intensive. There needs to be enough doctors and nurses with the correct skills to recognise a patient with potential sepsis. This can be a major problem in poorly resourced settings with critical shortages of healthcare workers.

After identifying sepsis in a patient, healthcare workers should administer antibiotics rapidly, whilst resuscitating the patient to support their organs.

Again, this can be very resource intensive with frequent monitoring required to assess response to treatment. After this initial fluid resuscitation, the patient with sepsis can be very vulnerable for several days, requiring frequent reassessment and treatment refinement by healthcare workers.

Together, this has been described as the “three delays model” for these basic pillars of sepsis management.

Whilst training and improved technologies may help make the process more efficient and effective, we must remain cognisant that the core rate-limiting factor for improved survival in sepsis globally is resource limitation.

Innovators must apply “health systems thinking” when developing prospective solutions, considering how these could and should be nested within existing healthcare pathways.

There are currently no clinical tests that can accurately diagnose sepsis. An accurate, cheap, and simple to use diagnostic test is urgently needed to guide clinical decision making for sepsis.

Many conditions not involving infection can mimic sepsis.

Antibiotics may be over-prescribed by doctors worried about the risks of delayed administration if the diagnosis is unclear. This can have negative consequences for patients and can also drive development of antibiotic resistant bacteria.

Healthcare workers must be confident in the test and its ability to effectively rule out sepsis before this culture of “just in case” antibiotic prescribing will change.

There are two key steps in antibiotic prescribing – initiation at the point of sepsis recognition and de-escalation approximately 48 hours later informed by clinical diagnostic tests.

Early administration of antibiotics is key in the initiation phase and improves the chance of survival. Therefore, any diagnostic must be simple to use with a rapid turnaround in results to guide clinical decision making.

Innovators should work with clinicians to think carefully about the strengths and weaknesses of their prospective technologies and where these could fit best within existing treatment pathways.

It may be that an infection diagnostic, whilst not suitable for the emergency antimicrobial administration phase, could be used elsewhere in the treatment pathway.

Innovators need to understand treatment pathways, and this may mean involving healthcare workers from the outset to identify clinical priorities and needs.

The diagnostic priority is an effective “rule out” test to support clinicians in the decision to not administer antibiotics where these are not required.

In my experience, context is everything when it comes to global health. Equitable partnership with clinicians and researchers is essential to drive effective improvements in healthcare.

Innovators should work with prospective partners to identify local priorities, build clinical and research capacity, and affect policy change.

I am involved with a group that has published a set of guidance on how to promote equitable partnership in international collaborations that may be useful in this respect (https://pubmed.ncbi.nlm.nih.gov/34647323/).

Globally, sepsis impacts most in Southern Africa where there is high prevalence of HIV-infection and other immunosuppressive conditions such as malnutrition. Any successful sepsis diagnostic must be tested in its intended use populations prior to rollout.