International Journal of Infectious Diseases 97 (2020) 60–65Contents lists available at ScienceDirect
International Journal of Infectious Diseases
journal homepage: www.elsevier .com/ locate / i j id
Setting priorities for patient-centered surveillance of
drug-resistant infections
Elizabeth A. Ashleya,b,c,*  , Alistair b,c,d McLean , Francesca Chiarae, Nicholas f,g Feasey ,
Walter Jaokoh, Japheth A. Opintani, Sharon J. Peacockj, k Priscilla Rupali , Paul Turnerc,l
a Lao-Oxford-Mahosot Hospital-Wellcome Trust Research Unit, Microbiology Laboratory, Mahosot Hospital, Vientiane, Lao PDR
bMyanmar Oxford Clinical Research Unit, Yangon, Myanmar
cCentre for Tropical Medicine and Global Health, Nuffield Department of Medicine, University of Oxford, Oxford, UK
d Infectious Diseases Data Observatory (IDDO), Centre for Tropical Medicine and Global Health, University of Oxford, UK
eDrug-Resistant Infections Priority Programme, Wellcome, London, UK
fMalawi-Liverpool-Wellcome Trust Clinical Research Programme, Blantyre, Malawi
gDepartment of Clinical Sciences, Liverpool School of Tropical Medicine, Liverpool, UK
hDepartment of Medical Microbiology, University of Nairobi, Kenya
iDepartment of Medical Microbiology, University of Ghana, Accra, Ghana
jDepartment of Medicine, University of Cambridge, UK
kDepartment of Infectious Diseases, Christian Medical College, Vellore, Tamil Nadu, India
lCambodia Oxford Medical Research Unit, Angkor Hospital for Children, Siem Reap, Cambodia
A R T I C L E I N F O A B S T R A C T
Article history: Methods: A priority-setting process (PSP) was launched to define priorities for patient-centered
Received 28 April 2020 antimicrobial resistance (AMR) surveillance and research in low- and middle-income countries (LMICs).
Received in revised form 27 May 2020 A list of uncertainties related to AMR surveillance in human health was generated using an online survey
Accepted 28 May 2020 of stakeholders in LMICs, which asked for unanswered questions about diagnosis, treatment, or
prevention of antibiotic resistance.
Keywords: Results: A total of 445 respondents generated 1076 questions that were mapped to a final shortlist of 107
Antimicrobial resistance questions. The most common theme was the treatment of drug-resistant infections, followed by
surveillance
drug-resistant infections diagnosis, then prevention, and requests for local AMR data. The most asked question was a request for 
priority-setting local AMR data, revealing the lack of basic information in many LMICs to guide actions to tackle AMR. The
low- and middle-income countries steering group recommended three research areas to be prioritized for funding in the next five years:   
infection prevention and control in LMICs, improved electronic patient records, starting with laboratory
information management systems, and sustainable behavior change among doctors and other health
care professionals with a focus on diagnostic stewardship.
© 2020 The Authors. Published by Elsevier Ltd on behalf of International Society for Infectious Diseases.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-
nd/4.0/).1. Background & Rationale
Low- and middle-income countries (LMICs) are predicted to be
hit hardest by drug-resistant infections (DRI) (O’Neill, 2016). The
burden of infectious diseases is higher, laboratory infrastructure is
weaker, antibiotic use and quality are less regulated, and the
healthcare workforce is smaller per population and less sub-
specialized than in higher-income settings. There is evidence that
healthcare-associated infections, which are more likely to be drug-* Corresponding author. Microbiology Laboratory, Mahosot Hospital, Vientiane,
Laos.
E-mail address: liz@tropmedres.ac (E.A. Ashley).
https://doi.org/10.1016/j.ijid.2020.05.121
1201-9712/© 2020 The Authors. Published by Elsevier Ltd on behalf of International So
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).resistant, occur more frequently in LMICs (Allegranzi et al., 2011).
Currently, high-income countries approaches to tackling antimi-
crobial resistance (AMR) are being adapted in LMICs, but there is a
lack of evidence for how these should be prioritized, taking into
account differences in the DRI burden, the fact that human and
financial resources are much more limited, and that there may be
different competing health-priorities.
The Surveillance and Epidemiology of Drug-resistant Infections
Consortium (SEDRIC) is a global think-tank that brings together a
range of international experts to take action to tackle the gaps in
drug-resistant infection surveillance and epidemiology. Its mission
is to strengthen the availability of information needed to monitor
and track risks and to support the translation of surveillance data
into interventions, changes in policy, and application of moreciety for Infectious Diseases. This is an open access article under the CC BY-NC-ND
E.A. Ashley et al. / International Journal of Infectious Diseases 97 (2020) 60–65 61effective practices. The work described here was an activity of the
first SEDRIC working group. A priority-setting process (PSP) was
launched to define priorities for AMR surveillance and research to
benefit human health in LMICs. Priority-setting exercises are used
increasingly in health research (Sibbald et al., 2009). The rationale
for using this method was that in order to rank the multiple
competing priorities for tackling DRIs, stakeholders from LMICs
should first be consulted to generate a list of uncertainties, rather
than relying on expert opinion. The PSP described here was
adapted from The James Lind Alliance (2018) and the Child Health
and Nutrition Research Initiative (CHNRI) priority-setting partner-
ship methods (Rudan et al., 2008). Guiding principles were to have
a transparent process driven by stakeholders working in LMICs.
The aim of this priority-setting process was to identify priorities
for DRI surveillance or research in LMICs with the goal of reducing
DRIs spread, transmission, and associated morbidity and mortality.
The scope was restricted to human health, with the beneficiaries
from the research intended to be patients affected by AMR in
LMICs. Resistance to anti-tuberculous or antimalarial or antiretro-
viral drugs was not considered. The time-frame of interest in which
to see improvements when selecting priorities was set as the next
three to five years. The SEDRIC working group acted as the steering
group for the process. Members were selected based on their
knowledge and expertise of AMR, understanding of the patient and
clinician populations, their experience of working in LMICs, and
access to networks who could participate in the PSP.
2. Methods
The PSP was conducted in four steps (Box 1):
1) Survey. An online survey was piloted and then launched to
generate a list of uncertainties related to AMR surveillance in
human health in low-and middle-income countries. This was
designed using SurveyMonkey1 and translated from English
into French and Spanish (Supplementary Material). The main
survey question was, “Please write down questions about
diagnosis, treatment, or prevention of antibiotic resistance that,
if fully answered, could make a difference to your practice and to
the health and well-being of your patients.” Respondents could
enter up to three questions, and they were asked to be specific
and focus on questions they considered to be of high priority.Fig. 1. Steps in the priorThe target stakeholders were clinicians, pathologists, or
researchers involved in patient care in LMICs. Respondents’
anonymity was protected, but some demographic information
was collected: gender, age range, country working in, occupa-
tion, and grade. We disseminated the survey via professional
networks of the steering group using flyers, email, and Twitter.
Our target was to keep the survey open for two months or until
we had reached 500 respondents (whichever occurred first).
2) Categorization. Survey responses were collated by organizing
them into themes and mapping to a shortlist of questions. Initial
collation and classification of the questions into themes were
performed independently by two members of the Steering
Committee. Responses that fell out of the scope of the survey
were excluded. Theme headings were generated by the first
reviewer (EA). Where differences in theme allocation between
the two reviewers occurred, a third member of the steering
committee made the final decision on thematic categorization.
Once questions had been organized into themes, similar or
duplicate questions were combined and mapped to a shorter list
of questions by one member of the steering committee (EA). For
each theme, the long and short lists of questions were reviewed
by a second member of the steering group with questions added
or edited if the meaning of the original questions on the long list
was determined to have not been adequately captured.
3) Interim prioritization. The members of the Steering Group then
each ranked their top ten questions. These were combined to
come up with a ranking score, which took into account the
number of times a question was ranked by different steering
group members (top-ranked preference given 20 points; tenth
preference, 11 points).
4) Final prioritization. The final step was to identify three
questions to be recommended as priorities for further research
to the SEDRIC board according to the following defined criteria;
this was reached by consensus. Criteria for final prioritization
included: a) questions were within the scope of the PSP, b) they
identified a clear gap in knowledge regarding AMR surveillance
in LMICs, c) they had the potential to be answerable by research
within three to five years, d) they were amenable to a funding
call, e) there was minimal overlap with other initiatives.
Fig. 1ity-setting process.
62 E.A. Ashley et al. / International Journal of Infectious Diseases 97 (2020) 60–65
Table 1
Characteristics of 445 survey respondents
Characteristic n/N (%)
Gender
Female 186/433 (43%)
Male 242/433 (56%)
Rather not say 5/433 (1.2%)
Age
21-30 years 40/433 (9.2%)
31-40 years 150/433 (35%)
41-50 years 138/433 (32%)
51-60 years 63/433 (15%)
61 or older 36/433 (8.3%)
Rather not say 6/433 (1.4%)
Income group
High income 83/428 (19%)
Upper middle income 91/428 (21%)
Lower middle income 180/428 (42%)
Low income 74/428 (17%)
WHO Region
African (AFRO) 142/428 (33%)
South-East Asia (SEARO) 115/428 (27%)
European (EURO) 65/428 (15%)
Pan-America (PAHO) 48/428 (11%)
Western Pacific (WPRO) 47/428 (11%)
Eastern Mediterranean (EMRO) 11/428 (2.6%)
Language
English 386/445 (87%)
Spanish 32/445 (7.2%)
French 27/445 (6.1%)
Specialty: Medical (except Pediatrics) 113/433 (26%)
Specialty: Pathology (includes Microbiology) 85/433 (20%)
Specialty: Pediatrics 59/433 (14%)
Specialty: General practice 38/433 (8.8%)
Specialty: Surgical 14/433 (3.2%)
Grade
Consultant/Attending 218/433 (50%)
General Practitioner 57/433 (13%)
Trainee/Resident 40/433 (9.2%)
Other (please specify) 118/433 (27%)
Institution has prescribing guidelines
Yes 259/433 (60%)
No 128/433 (30%)
Don’t know 46/433 (11%)
Institution has laboratory
Yes 359/433 (83%)
No 59/433 (14%)
Don’t know 15/433 (3.5%)
n/N (%) are presented. Note: 12 survey respondents provided no personal
information but still submitted questions.3. Results
The survey was launched on 6th  December 2018 during the 59th
Colloquium of the Institute of Tropical Medicine, Antwerp, held in
Phnom Penh, Cambodia. Wider dissemination by the Steering
Group started on rd 3 January 2019 using a standard invitation sent
out by email or Twitter, and the survey closed two months later on
4th March 2019. 445 respondents generated 1076 questions.
Because the survey was disseminated via social media and
informal networks, we could not estimate the response rate.
While our sampling strategy could not ensure population
representativeness, the survey respondents’ characteristics con-
firm that the majority of them (80%) were from low- and middle-
income countries and were practicing clinicians or microbiologists
involved in patient care. The locations of the respondents are
shown in Fig. 2. The largest number of respondents came from
Myanmar (n = 44), followed by South Africa (31), UK (27), India
(22), and Ghana (16). Most respondents submitted their questions
in English (386/445 or 87%), followed by Spanish (n = 32) and
French (n = 27). Most were senior grade physicians or pathologists
with < 5% from surgical specialties. Respondent characteristics are
shown in Table 1.
Table 2
3.1. Categorization into themes-
After excluding 417 questions that fell outside the scope of the
survey (including blank or incomprehensible responses), the
remaining 659 questions were categorized into themes (including
one miscellaneous category). The most common theme was the
treatment of drug-resistant infections, followed by diagnosis, then
prevention, and then requests for local AMR data. All of the themes
are shown in Fig. 3. The 659 questions were mapped to a shortlist
of 107 questions as described in the methods. The top ten most
asked questions are shown in Fig. 4.
3.2. Results of interim prioritization-
Each member of the Steering Committee ranked their top ten
questions from the shortlist of 107 questions. These were
combined to come up with a top ten list of surveillance priorities
(Table 3).Fig. 2. Number of survey respondents from each country. 428/445 survey respondents provided this information.
E.A. Ashley et al. / International Journal of Infectious Diseases 97 (2020) 60–65 63
Table 2
Interim prioritization of top 10 questions
Ranking N respondents N steering group members who
scorea asking this Q ranked this Q in their top 5
1 Which IPC interventions should be prioritized in LMICs taking into account the context 94 37 6
(overcrowding, no isolation facilities, poor infrastructure, low availability of PPE, water and
sanitation issues) and limited financial resources
2 How can we bring about sustainable behavior change among doctors and other health care 65 14 4
professionals with regard to managing infections and prescribing antibiotics?
3 Can diagnosis of infections be improved and made more rapid, ideally available at the point of care? 59 24 3
4 What is the prevalence of AMR/ What are the antimicrobial susceptibility patterns of infections 51 64 3
encountered in my institution/country/region? Includes distinguishing community from hospital
acquired
5 What is the role of improved information patient management systems including electronic 51 2 3
prescribing, to tackle AMR?
6 What is the most appropriate methodology to assess the mortality/burden of AMR? 49 1 3
7 What is the role of combination antibiotic therapy to treat or prevent drug-resistant infections? 43 11 3
8 What is the potential impact of vaccines to reduce AMR? 42 5 3
9 Can more rapid methods be developed, e.g., an affordable POCT that can diagnose drug resistant 38 53 2
infections
10 Will improved diagnosis lead to appropriate antimicrobial use and a reduction in AMR? 36 7 2
a The ranking score was derived by combining the rankings of the individual steering group members (first preference is given 20 points; tenth preference 11 points).
Fig. 3. Number of questions asked by theme. 659/1076 questions could be classified into a theme. 417 questions were excluded.
Fig. 4. Top 10 number of mapped questions asked by survey respondents. Marker color reflects theme (see Fig. 3).
64 E.A. Ashley et al. / International Journal of Infectious Diseases 97 (2020) 60–653.3. Results of final prioritization-
To make the final prioritization, the Steering Group reviewed
the top 10 questions according to the predefined criteria. The final
prioritization of three questions for further research and prioriti-
zation for funding were:
1. Which IPC interventions should be prioritized in LMICs,
taking into account the context (overcrowding, no isolation
facilities, poor infrastructure, low availability of PPE, water,
and sanitation issues) and limited financial resources? This
was considered a high priority because there are significant
knowledge gaps, and it is an area where most approaches
implemented in high-income settings are not feasible in lower-
income settings. Case-definitions of HCAIs used in high-income
countries are difficult to operationalize in LMICs because they
frequently require more investigations such as radiology and
blood testing that add to healthcare costs. Adapted definitions
and surveillance methods may be needed. Some IPC inter-
ventions can be introduced without laboratory surveillance.
2. What is the role of improved information patient manage-
ment systems, including electronic prescribing, to tackle
AMR? The highest priority in LMICs was considered to be
laboratory information systems in the next 3-5 years with
electronic patient records showing potential as tools for AMR
and antimicrobial use (AMU) surveillance in the longer term.
3. How can we bring about sustainable behavior change among
doctors and other health care professionals concerning
managing infections and prescribing antibiotics? The steer-
ing group identified this as another high priority area, however,
the focus should be aligned with the objective of improving
AMR surveillance and data. Targeting diagnostic stewardship,
specifically changing clinician behavior when requesting blood
cultures, was recommended.
Development of diagnostics was not prioritized by the group
because several other stakeholders (commercial interests, non-
profit organizations, e.g., FIND, and academic organizations) are
working on diagnostic development, including improvements to
existing methods, novel markers, transcriptomic signatures, and
leapfrog technologies. Making a significant breakthrough within
three to five years was also judged to be unlikely.
Generating local data and network strengthening was also not
prioritized because providing guidance on AMR surveillance is one
of the normative functions of the WHO. Strengthening laboratory
AMR surveillance in LMICs is the primary goal of the UK aid
program, the Fleming Fund, and several countries have received
grants to build and strengthen their surveillance capacity.
4. Discussion
The results of our survey revealed that many clinicians in LMICs
lack AMR data from their local area with the top question (n = 64)
being a request for basic surveillance data at the local or country
level. Most responses were about the treatment of drug-resistant
infections (n = 133). This included questions about antibiotic
stewardship, but typically respondents were asking for guidance
on optimal treatment of drug-resistant infections in their patients.
There are global treatment guidelines for malaria, tuberculosis,
and HIV and for some specific syndromes requiring antibiotics,
published by the WHO, but as yet there are no comprehensive
treatment guidelines for drug-resistant bacterial or fungal
infections. Given the differing epidemiology of AMR between
institutions, regions, and nations, issuing a comprehensive global
treatment guideline would be challenging. However, authoritative
evidence reviews of the treatment of drug-resistant infectionswould be useful, and guidelines could also be used by countries to
facilitate access to second-line antimicrobial treatments.
Another common theme was a call for improved diagnostics.
The most common questions expressed a need for more rapid
diagnostic tests, ideally at the point of care, both to distinguish
bacterial from viral infections and to diagnose drug-resistant
infections more quickly. Laboratory diagnoses of infections in most
LMICs, where it exists, rely on conventional bacteriology and an
array of immunochromatographic tests that frequently lack
sensitivity or are not well-validated. There have been major
advances in the development of close to point-of-care molecular
diagnostics direct from patient samples in recent years, including a
small number of tests for drug resistance such as the Cepheid
GeneXpert1 testing platform (MRSA from superficial swabs, N.
gonorrhoeae and C. trachomatis from vaginal or urethral swabs,
respiratory virus PCRs from nose or throat swabs and carbapene-
mase genes from cultured isolates) and the BioFire1 FilmArray1
panels (meningitis/encephalitis, gastrointestinal infections, respi-
ratory infections, blood culture identification including AMR
genes). However, these are not accessible to most patients in
LMICs because of high unit cost (15-200USD/patient depending on
the number of targets). Furthermore, molecular diagnostic
methods do not always lead to a definitive diagnosis. Clinical
presentations with respiratory illness are one of the main drivers of
antibiotic prescription, but reliable etiological diagnosis of
bacterial pneumonia by nasopharyngeal sampling is still very
challenging, as demonstrated in the Pneumonia Etiology Research
for Child Health (PERCH) study. (Pneumonia Etiology Research for
Child Health Study, 2019)
Infection prevention and control remain relatively neglected in
LMICs. The most high profile global initiative is the “Save Lives:
Clean your Hands” campaign of the WHO. Uptake of good hand
hygiene practices varies in all countries but is very low in many
LMICs where there is poor access to soap and water or alcohol gel in
many areas. Other interventions practiced in high-income
countries, e.g., terminal cleaning of patient bed areas on discharge
from hospital, cleaning mattresses and curtains, patient isolation,
use of personal protective equipment, screening for carriage of
antimicrobial-resistant bacteria, HCAI surveillance, outbreak
investigation, policy development, hand hygiene, and other IPC
audits require a level of resources that is not available in most
LMICs. Survey respondents asked for guidance on which IPC
interventions should be prioritized in the LMIC context. The
steering committee considered this to be a high priority area for
research given the lack of evidence, the potential impact for
reducing AMR, and the fact that HCAI surveillance can be
performed without a laboratory. None of the questions mentioned
surveillance or prevention of healthcare-associated infections,
which suggests that awareness of these is still too low, especially
given the burden in LMICs (Allegranzi et al., 2011). Good case-
definitions adapted to LMIC contexts are needed.
Development of actions to promote behavior change to tackle
AMR falls under objective 1 of the global action plan on AMR
(Improve awareness and understanding of antimicrobial resistance
through effective communication, education, and training). Many
behavior-change initiatives targeting healthcare professionals
focus on antimicrobial prescribing/stewardship and hand hygiene.
Behavior change that could be prioritized to lead to improved
surveillance data is in the area of diagnostic stewardship, for
example, finding ways to achieve appropriate requesting of blood
cultures in patients, and avoiding selection bias.
Improving data management was mentioned by a low number
of survey respondents but was considered a high priority by the
steering group. In general, the survey respondents focused on
having access to data rather than the mechanisms needed to
acquire it. Linkage of laboratory and clinical data is an aspiration in
E.A. Ashley et al. / International Journal of Infectious Diseases 97 (2020) 60–65 65most LMICs at present and unlikely to be achieved in the short
term, with most hospitals still using paper-based systems. The
development of open access laboratory information management
systems is more likely to be achievable ahead of the widespread
implementation of electronic patient records.
Strengths of this PSP include canvassing opinion from stake-
holders based in LMICs- i.e., the clinicians and pathologists
involved in patient care, using open-ended questioning, and
translating the questionnaire into French and Spanish. Limitations
include the lack of respondents from surgical specialties, poor
uptake by non-Anglophone respondents, and the fact that the
survey was only available in three languages. A number of
responses fell outside the survey's scope and did not follow the
instructions, suggesting it was not well understood by all
participants despite pilot testing.
AMR is a complex and multidimensional topic. Increased
awareness and funding for AMR research mean there is now a lot of
activity on this subject, but there have been many different
approaches to the problem. Here we have given an example of one
approach to setting human health priorities in LMICs when
tackling AMR. We propose three high priorities for research:
infection prevention and control, improved information manage-
ment, and behavior change among healthcare providers.
Conflict of interest statement
None declared
Funding source
This work was supported by the Wellcome Trust
Ethical approval
The survey was discussed with the Oxford Tropical Ethics
Committee (OxTREC), who reviewed the content and advised it
was exempt from the ethics committee review.
Data sharing statement
Survey data are available as supplementary materialAcknowledgments
We would like to thank all of the survey respondents, especially
those who shared the survey with their networks. Thanks to Jean
Patel, who contributed to the survey design and initial dissemina-
tion, and to Alison Holmes, facilitators and participants to the
SEDRIC global meeting (“Mobilising epidemiology and surveillance
data to inform patient care”) in October 2019, who discussed some
of the topics generated by the PSP. Finally, thanks to the SEDRIC
Secretariat, especially Nick Vail and Jamie Nunn, for their support.
The work described here was an activity of the first SEDRIC
working group (Antimicrobial Resistance & Data) and was
supported by the Wellcome Trust.
Appendix A. Supplementary data
Supplementary material related to this article can be found, in
the online version, at https://doi.org/10.1016/j.ijid.2020.05.121.
References
The James Lind Alliance; 2018. Available from: http://www.jla.nihr.ac.uk/. [Accessed
5th July 2018].
Allegranzi B, Bagheri Nejad S, Combescure C, Graafmans W, Attar H, Donaldson L,
et al. Burden of endemic health-care-associated infection in developing
countries: systematic review and meta-analysis. Lancet (London, England)
2011;377(9761):228–41.
O’Neill J. Tackling Drug-resistant infections globally: final report and recommen-
dations.. 2016.
Pneumonia Etiology Research for Child Health Study G. Causes of severe pneumonia
requiring hospital admission in children without HIV infection from Africa and
Asia: the PERCH multi-country case-control study. Lancet (London, England)
2019;394(10200):757–79.
Rudan I, Gibson JL, Ameratunga S, El Arifeen S, Bhutta ZA, Black M, et al. Setting
priorities in global child health research investments: guidelines for
implementation of CHNRI method. Croatian medical journal 2008;49
(6):720–33.
Sibbald SL, Singer PA, Upshur R, Martin DK. Priority setting: what constitutes
success?. A conceptual framework for successful priority setting. BMC health
services research 2009;9:43.