Impact case study database

The resistance of bacteria to antibiotics threatens our ability to treat infections. Approximately 30,000 women die each year as a result of severe infections when giving birth, with sepsis accounting for between 10% and 15% of maternal mortality and most deaths occurring in LMICs [see 5.1]. The UK Special Envoy on AMR has praised Salford’s partnership with Ugandan health workers in ‘driving amazing progress in Uganda and beyond’ in this area and confirmed that low-cost stewardship interventions of this kind can significantly decrease maternal mortality and ultimately reduce the impact of AMR [5.1].

4.1. Improved AMR surveillance and procurement planning for more effective treatment of AMR infections

Salford’s research, in collaboration with Ugandan partners, combined molecular and clinical data on antibiotic resistance profiles of bacteria isolated from infections at FPRRH [3.4]. This supported a new and comprehensive FPRRH antibiogram [5.1, Table 5.11, p. 99]. This is a tool which is well established in high-income countries, whereby collating all hospital Antimicrobial Susceptibility Testing (AST) data allows for evidence-based procurement of effective antibiotics for future empirical prescribing. These research activities and outputs have provided a strong evidence base to inform a 7% increase in procurement budget for antibiotics at FPRRH [5.1, Table 6.1, p. 111]. The 2020/21 Procurement Plan evidences a shift towards acquiring antibiotics for which lower levels of resistance have been detected in circulating bacterial strains [5.1, Table 6.1, p. 111].

Similarly, the FPRRH laboratory procurement policy has been modified in light of the microbiology testing results, with the Hospital Director noting that ‘without the correct access to laboratory supplies, we would be unable to detect undiagnosed MRSA, and are now including our Head of Microbiology in our processes to guide on planning and procurement of lab related supplies such as antibiotic disks’ [5.2] . Molecular evidence of inducible resistance to clindamycin [3.4] was reported at an infection control meeting at FPRRH; clindamycin is not currently used at the hospital, but any future use will be closely monitored despite clinical evidence of its efficacy [5.2].

Salford’s researchers have improved capacity for molecular AMR testing by training Ugandan early career researchers in DNA sequencing and bioinformatics analysis at Salford and repeating and validating the sequencing experiment in Uganda. This work involved collaboration with Makerere University, who had the necessary equipment to undertake state-of-the-art genome sequencing but lacked the training and consumables. Through research exchange visits, the team successfully sequenced the whole genomes of the S. aureus culture collection from FPRRH, showcasing this capacity for future sustainable utilisation of molecular methods to type AMR bacteria [3.4]. For example, this research has led to follow-up funding to establish a wider Ugandan sequencing infrastructure and associated supply chains towards infectious disease diagnostics and epidemiology (particularly COVID-19; Newton Fund (2020) EP/V029177/1 to Birtles, Goodhead, Ackers and Ugandan partners).

4.2. Sustainable improvements in laboratory, clinical and prescribing behaviours leading to reduced rates of maternal sepsis readmission

The Maternal Sepsis Action-Research Intervention (MSI) has ‘ generated the evidence base to support a new approach to the management of Maternal Sepsis’ [5.3] and led to a new policy brief being developed by the Pharmaceutical Society of Uganda, detailing a series of 21 policy changes arising from this research [5.4]. The MSI has resulted in a policy of midwives undertaking roles with greater responsibility, such as nurse or doctor roles, leading to closer collaboration with other hospital staff, and better opportunities for midwives to obtain training through mentoring and co-working, rather than expensive external opportunities. ‘Midwives are now taking the lead in managing patients; they used to wait for doctors to make decisions’ [5.1, p. 104]. Adoption of the model has resulted in midwives having an improved understanding of AMR issues and antibiotic stewardship. This has led to the reduction of inappropriate prescribing of antibiotics for non-surgical prophylaxis from 100% (February 2017) to just 20% (August 2018) [5.4, Table 1]. FPRRH’s Director confirms that *‘From an antimicrobial stewardship point of view the intervention and the research underpinning it has transformed Fort Port Regional Referral Hospital into an example of best practice’* and the hospital has now set up a Medicines Therapeutic Committee (MTC) to manage AMR as a result [5.2]. The establishment of MTCs is seen as critical by the Ugandan Ministry of Health to the operationalisation of the National Action Plan on Anti-Microbial Resistance in Uganda [5.2].

Salford’s research [3.5] has additionally led to midwives increasing their rates of wound-swabbing, and submission to laboratories for bacterial culture and sensitivity testing: FPPRH reported an increase of laboratory test orders from 0% to 95% in suspected sepsis cases from 2019 to 2020 [5.1, Table 5.2, p. 88]. This is also contributing to a comprehensive FPRRH antibiogram that enables, and provides confidence in, immediate empirical treatment whilst patients await test results [5.1, Table 5.11, p. 99]: ‘The sample size is now very adequate. The antibiogram will be good for the clinicians to guide prescribing and it will be good for the patients’ [5.1, p. 90]. The time taken from swabbing to results has led to a substantial reduction to appropriate antibiotic prescribing from 1 month to 3 days [5.1, p. 84] .

Engaging midwives (as key actors) with laboratory scientists, pharmacists and doctors has led to significant and sustained improvements in health workers’ compliance with hand hygiene guidelines: compliance with hand-washing guidelines has increased from ~18% in 2019 to ~67% in 2020, with hand gel/sanitizer use improving from ~33% to ~79% over the same period [5.1, Tables 4.2 and 4.3, pp. 63-64].

The Assistant Commissioner at the Ministry of Health in Uganda confirms that: ‘ The implementation of the Maternal Sepsis Intervention in FPRRH has had a truly transformative impact on maternal mortality rates at that hospital’ [5.5]. Fewer attending mothers have died from sepsis, [5.4, Figure 2], nor have been re-admitted [5.2]. The MSI has reduced the length of hospital stays for patients with suspected sepsis by an average of 6 days over an 18-month period [5.4, Figure 3]. The combination of reduced lengths of stay, a reduction of readmissions to zero, and fewer surgical referrals has resulted in **‘major cost savings’ in hospitals [5.2]. The Minister of Health for the Toro Kingdom (Western Uganda) has confirmed plans to scale up the approach: ‘ I am currently identifying ways of applying the Maternal Sepsis Intervention model to other public hospitals across Uganda’ [5.5].

4.3. Other contributions to policy making

Complex intervention research carried out by Salford’s researchers in partnership with Ugandan health workers [3.5] formed the basis for a policy brief prepared by the Pharmaceutical Society of Uganda in 2020 [see 4.2], which advocated for structural changes in pharmaceutical supply chains and the deployment of pharmacists in public hospitals in Uganda [5.4]. The brief was distributed to the Ugandan National Action Planning Group on Anti-Microbial Resistance as well as to the Ministry of Health in Uganda [5.4]. This has resulted in further policy change, including improvements in access to specialist clinical pharmacists, with specialist knowledge of drug combinations and complex conditions to guide evidence-based prescription of antibiotics [5.4, informed by the FPRRH hospital antibiogram; Policy Note 4E].

Tropical Health Education Trust (THET), a specialist global health organisation that trains and supports health workers through partnerships and enables people in LMICs to access essential healthcare, confirms that, as a result of this research [3.5], it has revised a number of policies attached to its future grants programmes [5.6]. The Grants Coordinator for THET highlighted the importance of strong links between the hospital and laboratory and noted that this is something ‘that will be assessed in future AMS programmes’ [5.6]. Also informed by Salford’s research, THET has revised the importance which needs to be placed on ensuring that basic levels of IPC are established at implementation sites before other AMS principles can be put in place [5.6].