Relationship between Antibiotic Consumption and Resistance: A Systematic Review

Background Unreserved use of antibiotics exerted selective pressure on susceptible bacteria, resulting in the survival of resistant strains. Despite this, the relationship between antibiotic resistance (ABR) and antibiotic consumption (ABC) is rarely studied. This systematic review aims to review the relationship between ABC and ABR from 2016 to 2022. Methods Articles published over 7 years (2016–2022) were searched from December 23 to 31, 2022. The search strategy was developed by using keywords for ABC and ABR. From 3367 articles, 58 eligible articles were included in the final review. Results The pooled ABC was 948017.9 DPDs and 4108.6 DIDs where over 70% of antibiotics were from the Watch and Reserve category based on the WHO AWaRe classification. The average pooled prevalence of ABR was 38.4%. Enterococcus faecium (59.4%), A. baumannii (52.6%), and P. aeruginosa (48.6%) were the most common antibiotic-resistant bacteria. Cephalosporins (76.8%), penicillin (58.3%), and aminoglycosides (52%) were commonly involved antibiotics in ABR. The positive correlation between ABR and consumption accounted for 311 (81%). The correlation between ABR P. aeruginosa and ABC accounted for 87 (22.7%), followed by 78 (20.3%) and 77 (20.1%) for ABR E. coli and K. pneumoniae with ABCs, respectively. Consumption of carbapenems and fluoroquinolones was most commonly correlated with resistance rates of P. aeruginosa, K. pneumoniae, E. coli, and A. baumannii. Conclusion There is a positive correlation between ABC and the rate of ABR. The review also revealed a cross-resistance between the consumption of different antibiotics and ABR. Optimizing antibiotic therapy and reducing unnecessary ABC will prevent the emergence and spread of ABR. Thus, advocating the implementation of stewardship programs plays a pivotal role in containing ABR.


Introduction
Te discovery of antibiotics is the most signifcant achievement in the twentieth century [1].It changed medical practice and signifcantly decreased morbidity and mortality associated with bacterial infections.In recent years, the emergence and spread of antibiotic-resistant pathogens on the one hand and decreased invention of new antibiotics on the other hand challenged healthcare [2,3].High rates of resistance against frequently used antibiotics to treat infections have been observed worldwide, resulting in running out of efective antibiotics to treat common infections [4].Due to this, access to antibiotics remains a critical issue globally [3,5].According to the Centers for Disease Control and Prevention (CDC), there were 2,868,700 infections due to resistant pathogens and 35,900 deaths from antibioticresistant bacterial infections each year [6].
ABR is a natural phenomenon augmented by human actions such as the inappropriate use of antibiotics [7].Increased utilization of antibiotics results in an increased frequency of inappropriate antibiotic use [8].Given the association between antibiotic use and the selection of resistant pathogens, inappropriate use of antibiotics is often used as a surrogate marker for the avoidable ABR [9].Tus, the increase in bacterial resistance is contributed by selection pressure on antibiotics as a result of use, overuse, and misuse [10][11][12], and total consumption of antibiotics is the critical factor in selecting resistance [11].Tere were individual studies that confrmed the correlation between ABC and ABR patterns [13][14][15][16][17]. Antibiotic stewardship programs (ASPs) are usually aimed at reducing overall ABC, and thus, preventing and reversing resistance [1,11].Tus, it is recommended to monitor antibiotic prescribing to improve the quality of antibiotic use and to reduce ABR [18].Tus, this review aims to determine the relationship between ABC and rates of ABR based on articles published globally from 2016 to 2022.

Methodology
2.1.Study and Data Collection.Eligible articles were identifed by the search strategy developed by using keywords for antibiotic resistance and ABC.Mainly the PubMed (Medline) database was used to search for eligible articles.Additional articles were identifed by searching from Google Scholar.Searches were performed for 7 years, and it was performed from December 23 to 31, 2022.Search terms used included a combination of keywords like "drug resistance," "antimicrobial resistance," "bacterial resistance," "antibiotic resistance," "antibiotic use," "ABC," "antimicrobial use," and "antimicrobial consumption."Articles were initially reviewed based on title and abstracts, and then, the whole document was read to select eligible articles for review.Articles that had determined the correlation between ABR and ABC using correlation coefcients and association at a P value of 0.05 were included in the review.Tus, articles relating to nonhuman infections, previous systematic reviews, commentaries, editorial letters, and available only in the abstracts were excluded from the review (Figure 1).

General Description of Articles Included in the Review.
Overall, 58 articles  published globally from 2016 to 2022 were included in the systematic review.Te summary of articles included, including the correlation between bacterial resistance and ABC, is summarized and annexed in Annex 1.

Relationship between ABC and ABR.
Here, the relationship between ABC and ABR is described for each bacterium based on the correlation coefcient reported by articles .However, it has to be noted that all studies did not provide detailed information and similar reports.
A strong correlation was observed between the consumption of aminoglycosides and the rate of P. aeruginosa resistance to amikacin and gentamicin, gentamicin with the rate of aminoglycoside-resistant P. aeruginosa [34], and levofoxacin with the resistance rate of P. aeruginosa to levofoxacin [42].A moderate correlation was found between the consumption of extended-spectrum antibiotics and cephalosporin with resistant P. aeruginosa [37], piperacillin/tazobactam with the incidence of combined-resistant P. aeruginosa, and amikacin with aminoglycoside-resistant P. aeruginosa [34].Moderate correlations were found between the consumption of carbapenem and the rate of carbapenem-resistant P. aeruginosa [37,57], carbapenem with imipenem and meropenem-resistant P. aeruginosa [30], and amikacin, gentamicin, and levofoxacin with resistant P. aeruginosa to respective antibiotics [72].
Some studies reported a decreased resistance rate of P. aeruginosa from the consumption of carbapenems.Tere was a very strong negative correlation between consumption of carbapenem and resistance rates in P. aeruginosa [20,57], a strong negative correlation between carbapenem and the rate of imipenem resistance in P. aeruginosa [25], and a moderate negative correlation between carbapenem and the rate of carbapenem-resistant P. aeruginosa [37], and amikacin was very strongly correlated with decreased rates of multidrug-resistant (MDR) P. aeruginosa [57].
Tere was also cross-resistance between ABC and resistance to another antibiotic, which was also common.Tere was a very strong correlation between consumption of ceftazidime and meropenem-resistant P. aeruginosa [69].Tere is a very strong correlation between the consumption of carbapenems and the incidence density of P. aeruginosa resistance to third-generation cephalosporins and aminoglycosides [22] and aminoglycosides with the rate of imipenem resistance in P. aeruginosa [25].Tere was a very   ESBLs: extended-spectrum beta-lactam/beta lactamase-inhibitor; TXS: trimethoprim sulfamethoxazole; 3GC: third-generation cephalosporins.Canadian Journal of Infectious Diseases and Medical Microbiology strong cross-correlation between consumption of amikacin and rates of imipenem resistance strains of P. aeruginosa and MDR strains of P. aeruginosa [57] and aminoglycosides with rates of P. aeruginosa resistance to tazobactam-piperacillin [39].
Similar to P. aeruginosa, there were records on the relationship between the rate of antibiotic-resistant A. baumannii and ABC.Consumption of carbapenems was very strongly correlated with imipenem resistance in A. baumannii [25] and with carbapenem resistance in A. baumannii [60,68].Consumption of carbapenems was strongly correlated with the rate of carbapenem-resistant A. baumannii [33,37,52,77] and carbapenems and imipenem with imipenem-resistant A. baumannii [55,61], but meropenem was moderately correlated with the rate of meropenem-resistant A. baumannii [32].
Likewise, consumption of cefepime was very strongly correlated with cefepime-resistant A. baumannii [33] and tigecycline with the incidence density of Acinetobacter [52].Te consumption of piperacillin/tazobactam and extendedspectrum cephalosporins was strongly correlated with piperacillin/tazobactam-resistant and extended-spectrum cephalosporin-resistant A. baumannii retrospectively [37].Tere was also a strong correlation between the consumption of fuoroquinolones and ciprofoxacin-resistant A. baumannii [62].A moderate correlation was found between the consumption of cefepime and ciprofoxacin with cefepime and ciprofoxacin-resistant A. baumannii, respectively [32], gentamicin with gentamicin-resistant A. baumannii [58], and fosfomycin with the rate of fosfomycin resistance in A. baumannii [55].However, there were strong negative correlations between the consumption of aminoglycosides and an incidence density of aminoglycoside-resistant Acinetobacter spp.[52] and a moderate negative correlation between ciprofoxacin and ciprofoxacin-resistant A. baumannii [37].
Cross-resistance was also reported for diferent antibiotics and resistant A. baumannii.Tere was a very strong correlation between consumption of ceftazidime and both imipenem and meropenem-resistant A. baumannii [69], cephalosporin/beta-lactamase inhibitor combinations [60,77], and tetracyclines [60] with carbapenem-resistant A. baumannii.Tere was a strong correlation between consumption of cephalosporins, cephalosporin/betalactamase inhibitor combinations, and other beta-lactam/ beta-lactamase inhibitor combinations with carbapenemresistant A. baumannii [77], but a moderate correlation was evident between cephalosporin carbapenem-resistant A. baumannii [77] and glycoprotein with carbapenemresistant A. baumannii [68], and monobactams were moderately correlated with rates of carbapenem-resistant A. baumannii [77], but amikacin was very strongly and negatively correlated with rates of MDR strain A. baumannii, rates of imipenem-resistant strain A. baumannii, and rates of resistant meropenem strain A. baumannii [57].A negative strong cross-correlation was also reported between the consumption of glycopeptides and amikacin-resistant A. baumannii [62] and aminoglycosides, quinolones, and oxacephems with imipenem-resistant A. baumannii [25], whereas ciprofoxacin was strongly and negatively correlated with both imipenem and meropenem-resistant A. baumannii [69].
Similarly, there was a very strong correlation between the consumption of imipenem/cilastatin and the resistance rate of K. pneumoniae to imipenem/cilastatin and fuoroquinolones with the prevalence of ciprofoxacin-resistant K. pneumoniae [43], but there were very strong negative correlations between imipenem and imipenem-and meropenem-resistant K. pneumoniae [69].Tere was a strong correlation between the consumption of carbapenems and carbapenem-resistant K. pneumoniae [77] and extended-spectrum cephalosporins with extended-spectrum cephalosporin-resistant K. pneumoniae [69].Colistin resistance was strongly correlated with the consumption of colistin [39].Consumption of carbapenems was moderately correlated with doripenem-resistant, ertapenem-resistant, and meropenem-resistant K. pneumoniae [39] and carbapenemresistant K. pneumoniae [44].Consumption of cephalosporins was also moderately correlated with the rate of cephalosporinresistant K. pneumoniae [44] and fuoroquinolones with ciprofoxacin-resistant K. pneumoniae [61].
A very strong cross-correlation was observed between the consumption of piperacillin/tazobactam and aminoglycosideresistant K. pneumoniae [22], meropenem-resistant and imipenem-resistant K. pneumoniae [69], fuoroquinolones with trimethoprim/sulfamethoxazole-resistant K. pneumoniae [22], and the beta-lactam/beta-lactamase inhibitor combinations with carbapenem-resistant K. pneumoniae [22].Consumption of sulfonamides was also very strongly correlated with the rate of ceftazidime-resistant K. pneumoniae and ceftazidime with quinolone-resistant K. pneumoniae [25].Tere was a strong correlation between colistin-resistant K. pneumoniae and imidazoles, carbapenems, and colistin [39], doripenemresistant K. pneumoniae with the consumption of imidazoles, ertapenem-resistant K. pneumoniae with glycopeptides, and meropenem-resistant K. pneumoniae with glycopeptides [39], and ceftriaxone was moderately correlated with the rate of carbapenem-resistant K. pneumoniae [35] and carbapenems with K. pneumoniae resistant to third-generation cephalosporin [63].A moderate cross-correlation was observed 10 Canadian Journal of Infectious Diseases and Medical Microbiology between the consumption of glycopeptides and carbapenemresistant K. pneumoniae [68].
Antibiotic-resistant E. coli was also correlated with the consumption of several antibiotics.Tere was a very strong correlation between ciprofoxacin consumption with fuoroquinolone-resistant E. coli [65], and a strong correlation was found between fuoroquinolone and fuoroquinolone-resistant E. coli [48,65], fuoroquinolones with the resistance rate of E. coli to levofoxacin and resistance rate of E. coli to ciprofoxacin [42], and ciprofoxacin with ciprofoxacin-resistant E. coli [37], but there was a strong negative correlation between fuoroquinolone and ciprofoxacin-resistant E. coli [62].Tere was also a very strong correlation between cefotaxime and cefotaximeresistant E. coli and a strong correlation between cefoxitin and cefoxitin-resistant E. coli [65], gentamicin with resistant E. coli [33], extended-spectrum cephalosporin with resistant E. coli [37], and imipenem/cilastatin with imipenem/cilastatin-resistant E. coli [62].
Antibiotic-resistant N. gonorrhoeae was reported in two studies [23,49].Consumption of ceftriaxone was strongly correlated with resistant N. gonorrhoeae, but consumption of ciprofoxacin was moderately correlated [23].Macrolide consumption and cefxime resistance in N. gonorrhoeae and consumption of cephalosporin, macrolide, and quinolone were strongly correlated with ciprofoxacin resistance in N. gonorrhoeae [49].Consumption of quinolones and cefotaxime was positively associated with ciprofoxacin-resistant and cefotaxime-resistant N. gonorrhoeae, respectively [74].
Consumption of aminoglycosides, quinolones, and carbapenem was strongly correlated with amikacin-resistant E. cloacae [25].But carbapenem-resistant E. cloacae showed signifcant, very strong negative correlations with the usage of penicillin/beta-lactamase inhibitor I combinations, betalactam/beta-lactamase inhibitor combinations, meropenem, and carbapenems, but a very strong positive correlation was shown with total cephalosporin use [60].Consumption of piperacillin/tazobactam was very strongly correlated with resistant C. difcile, but it was very strongly but negatively correlated with vancomycin [31] and ESBL-positive Enterobacteriaceae rates [33].A positive association was also reported from the consumption of macrolides and clarithromycin-resistant H. pylori [50,75] and consumption of quinolones with levofoxacin-resistant H. pylori [75].

Canadian Journal of Infectious Diseases and Medical Microbiology
Tere was a very strong negative correlation between the consumption of fuoroquinolone with ciprofoxacinresistant E. faecalis [62] and third-generation cephalosporins and Hlr-gentamicin-resistant E. faecalis [22] and strong negative correlation between the consumption of fuoroquinolone and ciprofoxacin-resistant S. aureus, glycopeptides, and gentamicin-resistant S. aureus, glycopeptides, and gentamicin-resistant CoN Staphylococcus; glycopeptides with amikacin-resistant E. faecium and gentamicin-resistant E. faecalis [62] and piperacillin/tazobactam and vancomycin with oxacillin-resistant S. aureus [33].A moderate correlation was shown between the consumption of gentamicin and gentamicin-resistant E. faecalis [48] and fuoroquinolone with ciprofoxacin-resistant CoN Staphylococcus [62].Te consumption of amoxicillin/clavulanate and oxacillin resistance in S. aureus [33] and imipenem/cilastatin and imipenem/cilastatin-resistant MRSA were very strongly and negatively correlated [42].Consumption of glycopeptides and aminoglycosides was negatively correlated with MRSA [64].

Discussion
ABR is identifed as one of the top threats to public health and challenges progress in health care, food production, and life expectancy [6,78].Due to overuse and misuse of antibiotics, bacteria have become increasingly resistant to various antibiotics [1].A recent point prevalence survey at the global level reported about 136 million per year hospitalassociated infections resistant to antibiotics caused by highpriority pathogens (E.coli, Acinetobacter spp., Klebsiella spp., S. aureus, Enterobacter spp., and Pseudomonas spp.) [79].Te current review revealed 38.4% of the average pooled prevalence of antibiotic resistance to bacteria, with E. faecium (59.4%), A. baumannii (52.6%),P. aeruginosa (48.5%), coagulase-negative Staphylococcus (43.7%),Enterobacter (46.1%), and P. mirabilis (48.5%).Tis complements the global concerns on the ABR [1,6,78,79].Te burden of the ESKAPE and antibiotic resistance pattern calls for surveillance of antibiotic resistance targeting these pathogens by incorporating them into the infection control policy [80,81].It requires preventing infections as priority, slowing the development of ABR through better antibiotic use, and stopping the spread of ABR when it develops [78].
According to a recent report from a global point prevalence survey, high use of antibiotics (62.0%) and prolonged surgical antibiotic prophylaxis (60.9%) were the most common problems [82].Tus, overuse and inappropriate use of antibiotics are recognized as a signifcant driver of the emergence of resistant strains of bacteria as the increased exposure to antibiotics encourages the development of ABR [83][84][85].Increased ABC of broad-spectrum and last-resort antibiotics is a serious concern for public health [84][85][86].Inappropriate consumption of antibiotics has also been shown to contribute to the occurrence of MDR organisms [86].Te current review  has clearly shown the relationship between ABC and ABR.It revealed more than 70% consumption of antibiotics from the Watch and Reserve categories based on the WHO AWaRe classifcation, among which consumption of carbapenems and fuoroquinolones were the most frequently involved in ABR including cross-resistance with consumption of other antibiotics.Tis was in line with a recent report from the global level which concluded, "a large proportion of prescriptions for key Watch antibiotics were issued for indications other than those for which they were included in the Essential Medicine List" [82].Tus, a direct epidemiological relationship between ABC and the emergence and dissemination of resistant bacterial strains needs to be worked on for further understanding.
Similarly, the CDC classifed carbapenem-resistant Acinetobacter, C. difcile, carbapenem-resistant Enterobacteriaceae, and drug-resistant N. gonorrhoeae as urgent threats and drug-resistant Campylobacter, extended-spectrum beta-lactamase-producing Enterobacteriaceae, vancomycinresistant Enterococci, MDR P. aeruginosa, drug-resistant nontyphoidal Salmonella, drug-resistant Salmonella serotype Typhi, drug-resistant Shigella, methicillin-resistant S. aureus, drug-resistant S. pneumoniae, and drug-resistant Tuberculosis as serious threats to health care [6].As can be noted, the current review strengthens the need to approach pathogens as a major issue problem related to resistance against antimicrobial agents.
Antibiotics with "high resistance potential," even with limited use, have been associated with the emergence of MDR Gram-negative bacteria, e.g., imipenem, ceftazidime, gentamicin/tobramycin, and ciprofoxacin [89].ASP is known to result in decreased ABC and ABR [90].Due to a strong association between ABC and the development of antibiotic resistance [83], reducing the need for and inappropriate consumption of antibiotics through the implementation of appropriate ASPs can help delay the emergence and spread of ABR [90,91].Prudent use of antibiotics has to be a pillar in the fght against ABR [82].Tis requires aligning practice with evidence which requires changes in prescribing behavior based on the implementation of efective strategies to modify prescribing practices by aligning them with evidence-based recommendations for diagnosis and management [92].Tus, the initial step in fght against ABR in healthcare institutions has to be to measure the situation of antimicrobial use and resistance in their setting to raise awareness on areas of improvement of local prescribing behaviors [82].Te current review will help to understand the relationship between ABC and resistance and justify the need for universal funding and urgency for implementation of ASPs taking into account the local context for sustainable behavioral change in physicians' antibiotic prescribing practices.
Te review will serve as an important basis to better understand the issue and will inform policies, regulations, and interventions to optimize antibiotic use while reducing ABR.However, it has to be noted that the review does have limitations.Te review includes literature in the English language and was not evaluated by independent evaluators.Because of the multiple numbers of antibiotics and microorganisms tested, risk assessment for bias and quality was not considered.Since most of the studies were done retrospectively, they may not necessarily show the true relationship between ABC and the rate of resistance.Tere are no data on the relationship between the consumption of antibiotics and the respective resistance rates in all parts of the world.
In conclusion, there was a strong relationship between antibiotic consumption and antibiotic resistance.Te review also revealed a signifcant cross-resistance among diferent antibiotics.Most correlations were positively related, but a minority was negatively correlated, indicating the protective nature of antibiotic consumption for the development of resistance.Tere was a very strong correlation between fuoroquinolones, carbapenems, aminoglycosides, and penicillin consumption and respective resistance rates, but colistin, imipenem/cilastatin, and fuoroquinolones consumption was strongly correlated with the resistance rate of K. pneumoniae.Carbapenems were the most commonly used and strongly correlated with the rate of resistance for A. baumannii.Strict use of antibiotics is thus crucial to minimize the risk of emerging resistant organisms.ASP will help to optimize antibiotic therapy while reducing the amount of ABC to prevent the development and rate of ABR.Systems to assess institutional antibiotic utilization and the relationship between the trend of antibiotic use and rates of antibiotic resistance are strongly needed in health facilities to reduce resistance.Furthermore, special emphasis should be paid to antibiotics in the Watch and Reserve category of the WHO AWaRe classifcations such as carbapenems, fuoroquinolones, macrolides, glycopeptides, and second-fourth generation cephalosporins.