Abstract
Background:
There is an alarming increase in antimicrobial resistance (AMR) globally, complicating management of surgical infections, especially in low-resource settings. Of particular concern for surgeons are third generation cephalosporin-resistant and carbapenem-resistant Enterobacteriaceae.
Methods:
The published literature was searched to identify the scope and causative factors of emerging bacterial resistance in low- and middle-income countries (LMICs).
Results:
Antimicrobial resistance impacts economics, human development, health equity, health security, and food production. Factors that contribute to AMR include use of antibiotic agents in livestock, antibiotic agents in wastewater and sewage, poor sanitation, and overprescribing or unregulated use of antibiotic agents. Because the factors influencing AMR globally are multi-factorial, solutions must be addressed at multiple levels. In LMICs, these can occur through national initiatives, at the facility level, or at the community level with coordination engaging government agencies, the private sector, civil service, and professional groups.
Conclusions:
There is a growing recognition of the need for national AMR prevention programs. Meanwhile, infection prevention and control programs and antimicrobial stewardship remain cornerstones of management at the facility level.
Surgical infections include both infections necessitating surgical management and infections secondary to surgical procedures [1]. Management of surgical infections includes both operative interventions and antibiotic management. There is an alarming increase in antimicrobial resistance (AMR) globally, however, complicating management of surgical infections, especially in low-resource settings.
The potential for AMR has been recognized since the discovery of penicillin. Alexander Fleming recognized the inherent limitations of antibiotic agents and noted that penicillin was not a “magic elixir,” warning that bacteria could become resistant to penicillin [2]. In many ways, AMR is a natural evolutionary response to antimicrobial exposure [3]. Widespread use of antibiotic agents in both agriculture and medicine has created selection pressure toward bacteria with multi-drug resistance [4].
The AMR impacts economics, human development, health equity, health security, and food production [5]. An estimated 700,000 people worldwide die of resistant infections and, if not addressed, that could rise to 10 million a year by 2050, costing $100 trillion in global economic output [6]. In low- and middle-income countries (LMICs), AMR is associated with increased out-of-pocket health expenditure with a 10-point increase in percentage of out-of-pocket health expenditure associated with a 3.2% point increase in AMR [7].
Multi-Drug Resistant Bacterial Infections
Multi-drug resistance (MDR), commonly defined as resistance to three or more classes of antibiotic agents, is a growing global challenge (Table 1). At Kenyatta National Hospital, the largest public tertiary referral hospital in East and Central Africa, 88% of bacterial isolates from the medical ward were MDR, and 26% were extensively drug resistant [8]. In Hawassa, Ethiopia, 76% of isolates from the microbiology laboratory were MDR [9]. In patients with sepsis admitted to an intensive care unit (ICU) in Mongolia, 11% were extensively drug resistant, and 25% were MDR [10]. In those with diabetic foot ulcers in Kenya, 31% of Staphylococcus aureus and 40% of gram-negative isolates were MDR [11].
Bacterial Antimicrobial Resistance in Surgical Infections
The MDR is found not only in ill patients, but also in healthy individuals. In the Democratic Republic of the Congo, 21% of healthy children were carriers of pneumococcal isolates, with 43% of isolates resistant to three or more antibiotic agents [12].
The AMR is associated with worse clinical outcomes. In Thailand, the 30-day mortality rates in patients with MDR community-acquired bacteremia, health-care–associated bacteremia, and hospital acquired bacteremia were 35%, 49%, and 53%, respectively [13]. More than 19,000 extra deaths are caused by MDR bacteria in Thailand each year [13]. In an Egyptian pediatric ICU (PICU), infection with resistant organisms was associated with four times increased odds of death [14]. Infections in ICU patients in LMICs are particularly challenging because of the high rate of AMR with limited antibiotic options and problems with antibiotic stewardship and infection prevention and control practices [15].
Gram-negative pathogens
Gram-negative pathogens account for a substantial burden of surgical infections. Many intra-abdominal infections, such as perforated bowel and appendicitis, are secondary to gram-negative pathogens such as Escherichia coli and Klebsiella spp. [1]. Helicobacter pylori infection is associated with peptic ulcer disease whereas Salmonella typhi is the causative agent in typhoid intestinal perforation. While skin and soft tissue infections have been attributed predominantly to gram-positive pathogens, studies in LMICs commonly isolate gram-negative pathogens from these sites [11,16]. In a hospital in Mumbai, 66% of surgical site infections in clean and clean contaminated surgical procedures were from gram-negative pathogens [17].
Gram-negative pathogens such as Pseudomonas and Acinetobacter are also commonly isolated in hospital-acquired infections. The most common pathogens isolated from patients in a PICU in Egypt were Klebsiella (30.5%), Acinetobacter baumanii (22%), and Pseudomonas (17%) [14]. In Nigeria, more than 80% of nosocomial infections were because of gram-negative bacteria [18].
The AMR is common in gram-negative pathogens in LMICs. In Egypt, among 126 gram-negative isolates, 94% had a MDR phenotype [19]. Clarithromycin-resistant H. pylori is a high-priority pathogen because of the global prevalence of peptic ulcer disease [20]. In H. pylori strains in Brazil, 40% of strains were resistant to metronidazole, 19.5% to clarithromycin, and 10% to amoxicillin [21]. Fluoroquinolone-resistant S. typhii is a high-priority pathogen because it is a frequent cause of death in LMICs [20]. There are an estimated 18 million cases of typhoid fever each year with most occurring in central Africa and Asia [20]. More than one third of S. typhii infections in endemic areas have MDR with resistance to common antibiotic agents such as fluoroquinolones [22]. The percentage of S. typhii with MDR rises to more than 50% in sub-Saharan Africa [23].
Third-generation cephalosporin-resistant Enterobacteriaceae have been defined as pathogens of critical priority by the World Health Organization (WHO) because of their high community and health-care burden with limited treatment options [20]. Based on the ResistanceMap database, the global prevalence of third-generation cephalosporin resistance for E. coli and Klebsiella was 64.5% and 67%, respectively, and expected to rise to 77% and 58% by 2030 [24]. Third-generation cephalosporin resistance in Kenyatta National Hospital was 75% in E. coli isolates and 82% in K. pneumoniae isolates [8]. In Hawassa, Ethiopia, 67% and 87% of E. coli and Klebsiella isolates were resistant to ceftriaxone [9]. In Harare, Zimbabwe, resistance of E. coli to cephalosporins increased from 20% in 2012 to 35% in 2017 (p > 0.001) [25].
Extended-spectrum beta-lactamase (ESBL) production is one mechanism by which third-generation cephalosporin resistance is conferred, either chromosomally or plasmid transmission. In the Tigecycline Evaluation and Surveillance Trial (TEST) surveillance study, ESBL production was found in 16% of E. coli and 21% of Klebsiella isolates with the highest proportion of ESBL producers in Africa (40% of E. coli, 56% of Klebsiella spp.) and Latin America (23% of E. coli, 37% of Klebsiella spp.) [26]. In South Africa, 37% of patients had ESBL-producing bacteria on hospital admission, and 42% were positive for ESBL-producing bacteria after 48 h of hospital admission [27]. In Chad, 48% of Enterobacteriaceae isolates were ESBL producers with higher rates in inpatients (54%) compared with outpatients (34%) [28]. In Egypt, among 126 gram-negative isolates, 75% had at least one ESBL gene [19]. In a hospital in Mumbai, 64% of E. coli and Klebsiella isolates were ESBL producers [17].
Carbapenem resistance is often conferred and spread through carbapenemase enzymes. In a tertiary hospital in Mwanza, Tanzania, 35% of MDR gram-negative isolates were positive for one or more carbapenemase genes [29]. In a Pakistani PICU, a MDR gram-negative infection developed in 4% of patients with 50% of isolates resistant to carbapenems [30]. In Mulago Hospital, Kampala, 7.4% of P. aeruginosa isolates were carbapenem resistant [31].
Carbapenem resistant Enterobacteriaceae are pathogens of critical priority because of their high health-care burden [20]. Using data from the ResistanceMap database, the global prevalence for carbapenem resistance in E. coli and Klebsiella was 5.8% and 23%, respectively, and estimated to rise to 12% and 53% by 2030 [24]. In Togo, eight of 152 (5.2%) Enterobacteriaceae had reduced susceptibility to carbapenems [32].
Carbapenem-resistant A. baumanii has a high health-care burden because of its capacity to colonize environmental surfaces and spread within the hospital [20]. It is a leading cause of ventilator associated pneumonia and wound infections [20]. It is intrinsically resistant to several classes of antibiotic agents, and there are few treatment options available. At Kenyatta National Hospital in Nairobi, 85% of Acinetobacter isolates were MDR, with poor susceptibility to fluoroquinolones (13%–24%), penicillins (1%-27%), cephalosporins (0%-11%), tobramycin (37%), and meropenem (27%) [33]. In Mulago Hospital, Kampala, 2.7% of A. baumanii isolates were carbapenem resistant [31].
Gram-positive pathogens
The most common gram-positive pathogen in surgical infections is S. aureus. Methicillin-resistant S. aureus (MRSA) and vancomycin-resistant S. aureus (VRSA) are major causes of morbidity and death worldwide, commonly associated with skin and soft tissue infections and hospital-acquired infections [20]. In LMICs, MRSA is a serious threat because of the occurrence of MDR strains, absence of screening and surveillance platforms, and high misuse of antibiotic agents [34]. In the TEST surveillance study, the proportion of MRSA globally was 33% with the highest rates seen in Asia (46%) and North America (38%) [26]. In those with diabetic foot ulcers in Kenya, 31% of S. aureus isolates were MDR [11]. In Cameroon, MRSA accounted for 80% of S. aureus isolates and of these, 80% were resistant to vancomycin [35].
Factors Associated with Antimicrobial Resistance
The causes and factors associated with AMR in LMICs are multi-factorial (Table 2). Contributing factors include use of antibiotic agents in livestock, release of antibiotic agents in wastewater by drug manufacturers, lack of sanitation, poor filtration of wastewater from hospitals filtering into waterways, and overprescribing or unregulated use of antibiotic agents [36].
Factors Associated with Antimicrobial Resistance
Livestock and food production
Antimicrobial resistance is recognized as a One Health challenge, crosscutting across human, animal, and environmental domains [37]. Overall, there are limited data on AMR in livestock and food products in low-resource settings [37]. Food handlers, farmers, or animal caregivers may contaminate food products with MDR bacteria, and MDR bacteria have been detected in various food products including meat and fresh produce [4]. Most research focuses on poor-quality medical aspects of AMR, with less attention directed toward the poor-quality veterinary aspects of AMR [38]. In the United States, antibiotic agents are used commonly in food producing animals with the most common medically important antibiotic used being tetracycline [39]. In Tunisia, ESBL phenotype was identified in 29% of cefotaxime-resistant Enterobacteriaceae isolated from animal origins (feces, organs, milk) [40]. In Vietnam, there is widespread agricultural use of antibiotic agents, which contributes to AMR [41].
In the United States, approximately 42% of medically important antimicrobial agents approved for use in food producing animals were intended for use in cattle [39]. The MDR in meat products may be because of antibiotic exposure in veterinarian settings, antibiotic agents in feed, or antibiotic agents given to manage infections [4]. Of 432 samples from feedlots and feedlot cattle in South Africa, 289 Enterococcus spp. were isolated with 176 possessing vancomycin-resistant genes [42]. In Algeria, 23% of milk and dairy products were contaminated with S. aureus with 16% of those specimens being MRSA [43]. In Bangladesh, 31% of raw milk samples from producers, >60% from collectors, and 100% from chilling plants were positive for E. coli [44]. In Algeria, 19% of meat and meat products were positive for Salmonella, and 32% of these isolates were MDR [45].
In Nairobi, Kenya, 57% of chicken droppings contained E. coli and 12% contained Salmonella with resistance seen to commonly used antibiotic agents such as amoxicillin, cotrimoxazole, tetracycline, and streptomycin [46]. In Algeria, 46% of broiler breeder flocks had E. coli, of which 89% were MDR [47]. In Ethiopia, E. coli was isolated from 32.5% of chickens, and 78% of these isolates were MDR [48]. In South Africa, Enterococcus isolates were found in multiple poultry products and 37% of these isolates were MDR [49]. In Ghana, similar resistance plasmids in human, poultry and gecko isolates suggest transmission of ESBL bacteria and plasmids across these populations [50].
Wastewater
Improperly treated wastewater effluent may contribute to the spread of MDR in LMICs. In Tunisia, ESBL phenotype was identified in 71% of cefotaxime-resistant Enterobacteriaceae isolates from surface water and wastewater treatment plants [40]. In wastewater effluent in Durban, South Africa, MRSA strains were isolated from treated effluent [51]. Metagenomic analysis of untreated sewage found a relatively high proportion of resistance genes to sulfonamides and phenicols in Asia and Africa whereas there was a relatively high proportion of macrolide resistance genes in Europe and North America [52]. Rates of AMR gene levels in untreated sewage were highest in Africa and lowest in Oceania [52].
Economics and human development index
The human development index is strongly associated with AMR [52]. Up to 89% of observed variation in AMR levels can be explained through sanitation and general health data [52]. Crowding and homelessness have also been associated with AMR isolates in both hospital and community assessments [53]. Studies in high-income settings suggest an inverse correlation between income and AMR [53].
Antibiotic use and misuse
Availability, affordability, accessibility, and quality of antibiotic agents are contributing factors in AMR development. A situation analysis of antibiotic use in Vietnam found that increased accessibility of antibiotic agents as well as injudicious use have resulted in increased AMR [41].
Self-treatment with antibiotic agents is common in LMICs. Among children less than 2 years in households in Madagascar and Senegal, 37% had consumed antibiotic agents in the preceding three months with most having a prescription and purchase in a pharmacy [54]. About 81% of survey respondents in Saudi Arabia, Yemen, and Uzbekistan had used antibiotic agents in the previous three months, and half of survey respondents did not complete the full antibiotic course because they felt better [55]. In Ghana, self-treatment with antibiotic agents is also common [56].
In many countries, policies to regulate antibiotic use are insufficient and poorly enforced. Unregulated use of antibiotic agents in LMICs makes them easily accessible over the counter, contributing to risk of AMR from improper use [57]. The pharmaceutical law in 2005 made antibiotic agents prescription-only in Vietnam [41]. Antibiotic agents continue to be dispensed without a doctor's prescription, however, and pharmacies are not penalized for dispensing without a prescription [41]. Non-prescription antibiotic use was 48% in Saudi Arabia and 78% in Yemen and Uzbekistan with pharmacies the primary source of non-prescribed antibiotic agents [55]. In Ghana, antibiotic transactions without a prescription were common in rural medicine outlets, but the overall volume of antibiotic agents sold was greater in urban pharmacies [56].
In addition to easy availability outside of the hospital, antibiotic agents can be misused within the hospital, with broad and prolonged antibiotic coverage in low-risk patients. At a referral hospital in Tanzania, 81% of urology patients received antibiotic agents with 46% receiving ceftriaxone and 38% receiving ceftriaxone and metronidazole [58]. Most urology patients (86%) received antibiotic agents for five days regardless of diagnosis [58]. Despite this broad and prolonged antibiotic coverage, patients with resistant bacteria, compared with non-resistant bacteria, more often receive inadequate empiric antibiotic therapy [10].
Antibiotic development
In LMICs, the high rate of AMR is coupled with limited antibiotic options. To compound this problem, there are limited antibiotic alternatives currently in the pipeline [59]. Development of new antimicrobial drugs is not financially attractive [60]. Price control may be used to ensure affordability of antibiotic agents. This may inadvertently lead to restricted availability of antibiotic agents, however, as pharmaceutical companies cut back on production [61]. Currently, there are only six large pharmaceutical companies active in the field, compared with 18 in 1990 [60]. The number of antibiotic related patents has decreased to 5000, compared with 8600 over a decade ago, and approval of new antibiotic agents is one eighth of what is was 30 years ago [60].
Health-care provider awareness and knowledge
One potential contributing factor toward AMR in LMICs is antibiotic management by health-care professionals. Factors that can impact a health professional's antibiotic choice include accessibility and affordability, antibiotic sensitivity data, hygiene and sanitation, and interactions with pharmaceutical representatives [62]. Studies in LMICs have shown that overall awareness of AMR is high [62]. In Ethiopia, physicians were aware that lack of diagnostic tests contributes to antibiotic overuse, and pharmacists were aware of the links between use of broad-spectrum antibiotic agents and AMR [63].
Knowledge of local antibiotic resistance levels may be low, however [64]. In a tertiary referral hospital in Ethiopia, while there was general awareness on the importance of AMR, there was less knowledge on institutional resistance profiles for common bacteria [63]. In a multi-country study, rather than decrease antibiotic use, AMR awareness resulted in a higher likelihood of using next-line antibiotic agents [62]. Fear of treatment failure is a major driver of antibiotic prescription behavior [63].
Solutions
Because the factors influencing AMR globally are multi-factorial, the solutions should be addressed at multiple levels. Antibiotic stewardship in LMICs can occur through national initiatives, at the facility level, or at the community level [65].
National and international coordination engaging government agencies, the private sector, civil service, and professional groups is needed to combat AMR [66]. The WHO has developed a global action plan on AMR [67],and the WHO policy package on AMR defines critical actions that need to be taken by governments to stimulate change [68]. The AMR is a complex, multi-faceted problem that requires strong political commitment [66].
The AMR challenges faced in developing countries are different and unique compared with those from developed countries, and therefore locally relevant national action plans for AMR are needed [69]. For example, in 1998, the Nepalese Ministry of Health implemented a AMR surveillance program that has since expanded to 13 participating laboratories [70]. This AMR surveillance program has resulted in policy changes including choice of antimicrobial agents and vaccination strategies [70].
Key to addressing the challenge of AMR at the facility level are infection control practices and antibiotic stewardship programs. Infection control measures, while challenging, can be implemented in low resource settings with a resultant decrease in transmission of MDR bacterial infections [71].
Challenges to antimicrobial stewardship in LMICs include diagnostic challenges, knowledge and awareness, access to quality-assured antibiotic agents and health-care facilities [65]. Diagnostic challenges can be addressed by strengthening laboratory services. Similar to the WHO Model List of Essential Medicines, in 2018, the WHO released a Model List of Essential In Vitro Diagnostics that includes recommendations for culture and antibiotic sensitivity in clinical laboratories [72]. In very small hospitals in Brazil, the presence of a microbiology laboratory and infection control committee were associated with decreased overall and broad-spectrum antibiotic use [73].
Antimicrobial stewardship programs help enforce good antibiotic prescribing practices. Implementation of the Smart Use of Antibiotics Program at a neonatal ICU in Shanghai, China, reduced antibiotic use without affecting clinical care outcomes [74]. At a level 1 trauma center in India, despite a reduction in device-associated infections and surgical site infections and increased compliance with hand hygiene and preventive care bundles, antimicrobial consumption persisted, suggesting a need for further efforts targeted toward antimicrobial stewardship [75].
Communication and teamwork are key components of a successful antimicrobial stewardship program. Antimicrobial stewardship programs can be limited by professional boundaries and hierarchies with lack of health-care–workforce engagement as a key hindrance [76]. Challenges for antimicrobial stewardship in Ethiopia included high pharmacist turnover and poor communication between laboratory, pharmacists, and clinicians [63]. Involving pharmacists in different care settings has been shown to improve antibiotic use in both the hospital and community setting [77].
Conclusion
Antimicrobial resistance is a global challenge with multiple contributing factors. Coordinated efforts are needed to enact change, spanning international, national, facility, and community levels. There is a growing recognition of the need for national AMR prevention programs. Infection prevention and control and antibiotic stewardship programs are cornerstones of management at the facility level.
Footnotes
Funding Information
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Author Disclosure Statement
No competing financial interests exist.
