Page 21 - GHES-2-3

Global Health Economics and
Sustainability
AI in antibiotic prescribing in Nigeria

action plan for addressing this issue (Fanelli et al., 2020; an infection and its susceptibility to antimicrobial agents
Rabiu et al., 2022a; Yusha’u et al., 2010). Many studies directly from clinical samples, ideally within a timeframe of
concerning antibiotic resistance, especially with a focus around 30 min (Sulaiman et al., 2022; Tamma et al., 2023).
on different MDR genes, in the context of Nigeria have This advancement would significantly reduce the reliance on
been published. The major carbapenemases, including empirical treatment and facilitate adjustment to antimicrobial
but not limited to OXA-48 enzyme of class D, class A therapy before administering a second dose, ensuring more
KPC-type carbapenemases, and NDM-type metallo-beta- timely and suitable treatment. Consequently, there exists a
lactamases, all play a significant role in antibiotic resistance noteworthy gap in research, particularly in Nigeria, where AI
(Brownstein et al., 2023; Tamma et al., 2023). In addition, can play a pivotal role in optimizing antibiotic prescription
plasmid-mediated β-lactamases like blaTEM/SHV and practices (Vestesson et al., 2023).
blaCTX-M can inactivate last-resort antibiotics, limiting
treatment options for clinicians. This further compounds 5. Strategies and application of AI in
the severity of antibiotic resistance issue in Nigeria (Rabiu antibiotic prescription
et al., 2022c; Shitu et al., 2020; Yusha’u et al., 2010). Appropriately prescribing antimicrobial drugs presents
In Nigeria, antibiotics rank among the most frequently multifaceted implementation challenges, requiring the
prescribed drugs in hospitals (Jaafaru et al., 2022). selection of suitable treatment for the suspected pathogen.
However, the efficacy of antibiotics could be impacted by This process involves the regulation of antimicrobial
various factors, including inappropriate or unnecessary agent concentration, determination of the frequency of
prescriptions. Furthermore, improper usage of broad- administration, and identification of the appropriate route
spectrum antibiotics, whether at incorrect doses or for to ensure optimal levels of drug reaching the site of infection
infections that can be addressed with a specific, non- (Vestesson et al., 2023; Chang & Chen, 2022). A significant
antibiotic therapy, may pave the way for the emergence of challenge in prescribing antimicrobials is marked by the need
bacteria resistant to multiple drugs. The elevated resistance to continually adjust a patient’s treatment as new clinical
levels contribute to higher patient mortality rates, longer data emerge. However, the limitations of specialized health-
hospital stays, and greater health-care expenses (Fanelli care resources and the overwhelming volume of information
et al., 2020), all of which are dire associated consequences make manual surveillance impractical. As a result, there is
warranting an immediate counteracting plan and action. a growing reliance on automated decision support systems
With the rapid emergence of antibiotic resistance across to review antimicrobial prescriptions in hospitals. These
different classes of bacteria, there arises a crucial need systems often utilize rule-based algorithms derived from
to develop novel antibiotics and implement precision published and expert guidelines to identify inappropriate
in antibiotic prescription to effectively combat these prescriptions and prevent potential adverse events (Table 1)
escalating cases (Ali et al., 2022; Tamma et al., 2023). (Fanelli et al., 2020; Valderrama-Rios et al., 2023). This
shifts toward automated decision support systems reflects a
In general, antibiotics are isolated from a limited pragmatic response to the challenges posed by the intricacies
number of molecular scaffolds, and synthesized and of antimicrobial prescription, ensuring a more streamlined
optimized after undergoing a series of cycles. Given the and effective approach to healthcare delivery.
rising challenge of antibiotic resistance, there is a pressing
need to identify novel scaffolds. Emerging techniques for 6. AI for antimicrobial stewardship
scaffold discovery include exploring untapped microbial In recent years, considerable attention has been devoted
pockets for natural compounds and repurposing synthetic to creating and sustaining antimicrobial stewardship
molecular catalogues as potential antibiotics (Ali et al., programs customized to meet hospitals’ unique
2022; Rabiu et al., 2023). requirements. A crucial aspect of most of these initiatives
4. Antibiotic prescription involves reviewing antimicrobial prescriptions and
providing feedback to prescribers. This process evaluates
Antimicrobials are frequently prescribed based on empirical several critical parameters of prescribed antimicrobials,
evidence or data from surveillance cultures, if accessible. including their indication, dosage, route of administration,
In either case, the specific pathogen causing the infection and duration (Vestesson et al., 2023). While antibiotic
is typically unidentified. Unfortunately, knowledge of the stewardship programs have demonstrated effectiveness in
antimicrobial susceptibility of the causative pathogen typically numerous high-income countries, their success remains
becomes available only well after the initiation of antimicrobial unproven in low- and middle-income countries (LMICs), of
therapy. Ideally, rapid diagnostic methods should facilitate which Nigeria is a part. The advent of AI solutions presents
the swift identification of both the pathogen responsible for an innovative approach to addressing this challenge in

Volume 2 Issue 3 (2024) 3 https://doi.org/10.36922/ghes.2602