Background:
We determined the prevalence of antimicrobial resistance (AMR) in polymicrobial pathogens in Pakistan.
Methods:
A total of 70,518 clinical samples were collected aseptically and confirmation of isolates and antibiogram were performed by the VITEK 2 system.
Results:
Of 70,518 samples, 441 (0.62%) were polymicrobial samples, with 882 (1.2%) polymicrobial pathogens with 689 (78.1%) Gram-negative rods (GNRs), 166 (18.8%) Gram-positive cocci and 27 (3.1%) Candida albicans. Among GNRs, 28.8% were Escherichia coli and 25.9% were Klebsiella pneumoniae. Majority, 15.1% of Pseudomonas aeruginosa and K. pneumoniae were found in combination. 30.1% of isolates were ESBL producers, 9.7% carbapenem-resistant organisms, 35.5% MRSA and 6.0% VRE. 100% of E. coli were resistant to ampicillin and 98% of K. pneumoniae were resistant to piperacillin.
Conclusion:
A high prevalence of AMR in polymicrobial pathogens was observed.
Keywords: Polymicrobial pathogens • Gram-negative rods • Gram-positive cocci • Candida albicans • ESBL • CRO • MRSA • VRE
Infections caused by one or more types of bacteria, viruses, fungi or parasites – known as polymicrobial infections – are a threat to health. These infections cause serious illness and are linked to high numbers of deaths, long hospital stays and high costs of treatment. Usually, polymicrobial infections are treated with combinations of antimicrobials. However, microbes becoming less susceptible to antimicrobials (known as antimicrobial resistance) is an increasing problem.
To find out how common resistance is in Pakistan, this study tested 70,518 clinical samples. Of these, 441 tested positive for polymicrobial infections. These included Candida albicans, Gram-positive cocci and Gram-negative rods infections. Many of these were resistant to widely used antibiotics such as penicillins, cephalosporins, quinolones and fluoroquinolones.
This study concluded that hospitals in Pakistan have a high prevalence of resistance and that better cleanliness practices should be put in place to combat this.
Global public health is being threatened by an increase in polymicrobial infections [1]. The presence of two or more bacteria (also known as polymicrobial infection) is becoming more common in clinical samples, with prevalence rates ranging from 6 to 32% [2]. The mortality rate for hospitalized patients diagnosed with polymicrobial illnesses is 21–63%, almost twice that of individuals diagnosed with monomicrobial infections. This leads to an increase in hospital stay duration, patient load, and healthcare expenses [3]. So far, our main tool against bacterial infections is antibiotics, which are the most frequently administered globally in both humans and animals [4].
One of the main drivers of antimicrobial resistance (AMR) is the excessive use of antibiotics. Around one to two decades ago, antibiotic usage worldwide increased by 39% and the daily recommended doses increased by 65% [5]. When it comes to treating chronic illnesses like cystic fibrosis, which are often affected by multiple pathogens, there is a lack of understanding about how commonly used antibiotics impact different types of microorganisms [6].
The most common risk factors that contribute to the development of polymicrobial infections are central venous catheters, administration of parenteral nutrition, gastrointestinal pathology (especially short intestine disorder), the use of broad-spectrum antimicrobials, and immunosuppression [7,8]. In addition, the majority of pathogens most frequently isolated as causes of polymicrobial infections are included in the World Health Organization (WHO) ‘critical pathogen list’ (carbapenem-resistant Enterobacter ales [CRE], extended-spectrum β-lactamase [ESBL]-producing Enterobacter ales, Acinetobacter baumannii and Pseudomonas aeruginosa) and the ‘high-priority pathogen list’ (methicillin-resistant Staphylococcus aureus [MRSA], Salmonella spp., and Enterobacter spp.) [9].
The rate of AMR in hospitals is increasing, and it is projected that 1.27 million people will die from otherwise harmless microbes by 2050 [10]. Furthermore, compared with developed countries, most deaths (around 85%) occur in low-to-middle income countries due to fragile healthcare systems, unhygienic conditions, and poor infection-control practices in clinical settings [10].
The battle against AMR in healthcare facilities is very challenging. Patients who are immunocompromised or hospitalized with major underlying diseases face greater treatment difficulties [11]. There is an increasing prevalence of multidrug-resistant (MDR) microorganisms, such as ESBL, CRE, MRSA, vancomycin-resistant Enterococcus (VRE), and drug-resistant Candida albicans [12,13].
The unavailability of reliable data in developing countries such as Pakistan makes it difficult to develop efficient methods to monitor and control AMR. Moreover, limited studies have been conducted to investigate the prevalence of polymicrobial infections in this country. As a result, we intend to conduct a thorough examination of antimicrobial susceptibility surveillance and the clinical characteristics of polymicrobial pathogens.
Ethical consideration: This retrospective cross-sectional and observational study was approved by the Ethical Committee at the Islamabad Diagnostic Centre (IDC) private limited, Islamabad. Patients who were being investigated were asked for their verbal or written consent to collect samples between June 2019 and July 2021. The patient’s information remained confidential.
Clinical Sample Collection: For this retrospective cross-sectional study, all the patient demographic isolate confirmations and polymicrobial pathogen antimicrobial susceptibility testing (AST) data were systematically collected from data management-software from 2019 to 2021 from the IDC laboratory, the first diagnostic facility providing an entire spectrum of imaging and laboratory services under one roof in Pakistan. The IDC laboratory has over 100 collection centers in over 30 Pakistan cities annually delivering more than 3 million tests on a 24/7 basis, as detailed in Supplementary Figure 1.
Clinical data of positive cultures were obtained from an electronic database containing patient reports at IDC Islamabad. Each case was given information on the patient’s age, sex, the year the sample was taken, the source of the specimen, the city the sample was taken from, the isolated pathogen in the positive culture and AST results, as per Clinical Laboratory Standard Institute (CLSI) guidelines.
A total of 70,518 clinical samples were collected from urine, blood, sputum, tracheal secretions, tissue, body fluids (CSF, pleural, ascetic, synovial), swabs (wound, ears, eyes, nose, throat, endotracheal tube) and indwelling catheters/tips. All samples were taken by the following aseptic techniques.
In short, blood culture samples were processed in an automatic Versa TREK™ blood culture system (Thermo Fisher Scientific, MA, USA) for up to 5 days. All the remaining clinical samples (mid-stream urine, pus, sputum and body fluids) were collected as per standard protocol. Furthermore, the pathogens were divided into groups based on the year the samples were taken, the age of the patients (≤5 years, 6–18 years, 19–45 years, 46–65 years and 65 years), and the sex of the patients.
Intermediate resistance was regarded as resistance in each scenario. Each bacterial strains susceptibility to each of the tested antimicrobials was evaluated and reported. The flowchart of the study is given in Figure 1.
Identification of the isolates: The clinical samples were cultured on blood agar, MacConkey agar, cystine–lactose–electrolyte-deficient agar and chocolate agar according to the samples. Plates were incubated at 37 ◦C overnight aerobically. The isolates werepreliminarily identified by colony morphology and Analytical Profile Index (API) identification kits (Bio Merieux, France). The isolates were further biochemically confirmed by an automatic VITEK 2 system (Bio Merieux, France).
Antimicrobial susceptibility testing: The VITEK-2 compact system was used to determine the AST of all isolated strains. All antibiotics used in this study were classified according to the WHO Aware classification system as Access, watch or Reserve. ‘Access’ antibiotics applied on isolates were ampicillin, amoxicillin-clavulanate, piperacillin-tazobactam, cefotaxime, amikacin, gentamicin, clindamycin and nitrofurantoin. ‘Watch’ group antibiotics included cefoxitin, cefepime, ceftazidime, ciprofloxacin, levofloxacin, moxifloxacin, vancomycin, teicoplanin, piperacillin-tazobactam, fosfomycin, tobramycin, imipenem, meropenem and ertapenem. A few ‘reserve’ antibiotics were applied against isolates in this study, namely linezolid, tigecycline, polymyxin B and colistin. For the AST of C. albicans, micafungin, voriconazole, itraconazole, fluconazole, amphotericin B and flucytosine were used. AST was interpreted according to CLSI guidelines 2019.
Phenotypic confirmation of MRSA: MRSA confirmation was done via CLSI 2019 guidelines using a cefoxitin (30 μg) disc. In short, 0.5 McFarland of S. aureus strains were streaked onto Mueller Hinton agar (MHA) plates and a cefoxitin disc was applied and the plates incubated overnight at 37 ◦C. A ≤22-mm zone of inhibition was considered MRSA positive.
Phenotypic detection of vancomycin-resistant Enterococcus: A minimum inhibitory concentration (MIC) measurement was performed for vancomycin against isolates of Enterococcus spp., using CLSI 2019. The MIC breakpoint for vancomycin resistance was ≥32 μg/ml for confirming VRE.
Phenotypic detection of ESBL: ESBL phenotyping was done by following the double-disc synergy test for Gram-negative isolates. The 0.5 McFarland standard inoculums of isolates were applied on the MHA plate by the lawned method. Amoxicillin/clavulanic acid discs and cephalosporins (ceftazidime, ceftriaxone and cefepime) were placed 15 mm apart from each other. MHA plates were incubated overnight at 37 ◦C. ESBL producers were identified if they followed the below criteria.
Phenotypic detection of carbapenem-resistant isolates: All suspected isolates were screened for carbapenem-resistant organism (CRO) phenotypes if they were resistant to meropenem and imipenem (MIC >1μg/ml) as per CLSI guidelines. Both antibiotics were routinely tested against all clinical isolates of Enterobacter ales in the laboratory.
Clinical characteristics: Between 2019 and 2021 a total of 70,518 clinical samples from different sources were processed, and among these 441 (0.62%) clinical samples were positive for polymicrobial isolates (882; 1.2%). Of 882 isolates, 689 (78.1%)were Gram-negative rods (GNRs), 166 (18.8%) Gram-positive cocci (GPC) and 27 (3.1%) were C. albicans. Most of the isolates were recovered from male patients (258; 58.5%) compared with female (183; 41.5%), and the male-to-female ratio was 2:1. Patients age ranged from 1 to 94 years, with the most isolates (136; 30.7%) recovered between the ages of 46–65 years, followed by 134 (30.4%) in the 19–45 years’ age group. Most of the isolates were recovered from the swab samples (222; 50.3%) followed by urine (96; 21.7%), sputum (42; 9.5%) and tips (37;8.4%). Geographically, 271 (62.5%) samples were collected from the Punjab province, followed by 103 (23.4%) from Khyber Pakhtunkhwa and 48 (10.9%) from Sindh province, as shown in Table 1.
Detection of clinical isolates in different samples: Among the GNRs (689; 78%), E. coli (199; 28.9%) was the most common pathogen, followed by K. pneumoniae(179; 25.9%), P. a e r u g i n o s a (165; 23.9%), P. mirabilis (53; 7.7%) and E. cloacae (31; 4.5%) were recovered. A total of 92 (46.2%) E. coli were recovered from pus samples, 59 (29.6%) from urine and 16 (8.1%) from sputum samples. Moreover, 94 (52.5%) K. pneumoniae samples were isolated from pus followed by 32 (18%) from urine and20 (11.3%) from sputum samples. Of the GPCs (166; 18.8%), the most often isolated pathogen was S. aureus(94; 56.2%) followed by E. faecalis (67; 40.3%) and S. pyogenes (5; 3%). S. aureus was mainly recovered from pus(69; 73.4%) and sputum samples (13; 14.0%). However, E. faecalis was isolated from urine (42; 63.6%) and pus samples (18; 27.3%). However, 27 (3.1%) C albicans were also recovered from polymicrobial infections. A total of 17 (62.9%) of C. albicans were recovered from urine and five (18.5%) from pus samples, as detailed in Table 2.
Age & gender distribution of polymicrobial isolates: Age and gender distribution revealed that patients fell under the age range of 1–94 years. Among these, 258(58.5%) were male and 183 (41.1%) were female. In male patients, 161 and 159 isolates were recovered between >40and ≤60 years and >20 and ≤40 years, respectively. However, in female patients 117 and 110 isolates were recovered from those >60 years and >40–≤60 years, respectively. Overall, 270/882 (30.6%) isolates were recovered from the age group >40–≤60 years, 265/882 (30.0%) isolates were from the age group >20–≤40 years and 258/882(29.2%) isolates from the age group >60. E. coli had 61/199 (30.6%) recovered between >40 and ≤60 years, K. pneumoniae had 57/179 (31.8%) from >20 to ≤40 years, P. a e r u g i n o s a had 60/165 (36.3%) from >60 years, and P. mirabilis had 23/53 (43.3%) from >40 to ≤60 years. In GPC, 43/94 (45.7%) of S. aureus were recovered.
Polymicrobial pathogens in clinical samples: In this study, the most common combinations were P. a e r u g i n o s a with K. pneumonia (66; 15.1%), E. coli (44; 10.3%),S. aureus (20; 4.6%) and P. mirabilis (10; 2.3%). Furthermore, E. coli was in combination with K. pneumoniae(378.5%), E. faecalis (34; 7.8%), S. aureus (27; 6.2%) and P. mirabilis (12; 2.9%). However, among GPC, K. pneumoniae were mainly found in combination with E. fecalis (19; 4.3%) and S. aureus (12; 2.7%), and P. mirabilis(14; 2.5%). S. aureus mainly coexisted with E. faecalis (13; 2.9%) and A. baumannii (5; 1.1%). C. albicans was detected most often in combination with E. coli (7; 1.6%), E. faecalis (7; 1.6%) and K. pneumoniae (5; 1.1%). The less frequent combination of S. aureus, P. aeruginosa, S. marcescens and E. cloacae was also observed (Supplementary Table 1 &Figure 2).
MIC of Gram-negative rods, Gram-positive cocci & C. albicans: The WHO recently classified antibiotics into three main groups: access, watch and reserve (Aware). The access group antibiotics belong to first- and second-line antibiotics, watch group antibiotics are only specific and limited to infective syndromes and more prone to antibiotic resistance, and reserve group antibiotics are only used as a‘ last resort’ and for life-threatening infections produced by MDR pathogens. Among GNRs, 100% of E. coli were resistant to ampicillin and co-amoxiclav, 83.4% to aztreonam, 76% to piperacillin, 54% to cephalosporins and
ciprofloxacin, 5% to carbapenems, and 1.5% to colistin. Furthermore, 98% of K. pneumoniae were resistant to piperacillin and 93% to aztreonam, followed by 62% to ceftriaxone and 58% to ciprofloxacin, 15.8% to carbapenem and 4.5% to colistin. P. a e r u g i n o s a is 100% resistant to aztreonam and 6.7% resistant to imipenem. Moreover, 100%of P. mirabilis is resistant to ampicillin and aztreonam, followed by 98% to co-amoxiclav, 90.5% to ciprofloxacin and84.6% to ceftazidime and cefepime (Table 4). Among GPCs, most of the S. aureus isolates tested were resistant to penicillin (100%), co-amoxiclav (95.6%), cephalosporins (64–70%), carbapenems (64–68%) and fluoroquinolones(88–90%) except for vancomycin and linezolid (0%). 96% of E. faecalis were resistant to ampicillin and co-amoxiclav, 66.7% to ciprofloxacin and levofloxacin, 56% to clindamycin, and 15.2% to vancomycin. Furthermore,
Prevalence of ESBL, CRO, MRSA & VRE in polymicrobial isolates: There was a high prevalence of ESBL, CRE, MRSA and VRE pathogens in polymicrobial isolates. Among GNRs(689; 78.1%), 266 (38.6%) were ESBL producers and 67 (9.7%) CRO. Among GPCs (166; 18.8%), 59 (35.5%)MRSA and 10 (6.0%) were VRE. Among ESBL-producing GNR, 96 (36.0%) were E. coli followed by 81 (30.4%)K. pneumoniae and 43 (16.1%) P. mirabilis. Among CRO, 28 (41.7%) were K. pneumoniae followed by 10 (14.9%)E. coli, and 11 (16.4%) P. aeruginosa. CROs (67/882) were mainly reported in urine samples (n = 27) followed by pus swabs (n = 23), sputum (n = 7), body fluids (n = 5), endotracheal tube tips (n = 4) and tissue samples(n = 1) (Supplementary Figure 2 &Table 6).
In both underdeveloped and wealthy nations, AMR has emerged as a major threat to public health [14]. To prevent the spread of infectious illnesses caused by microorganisms resistant to antibiotics, the WHO has urged researchers to keep studying AMR via monitoring [15]. When microorganisms become resistant to the drugs used to treat them, infections can become more difficult or even impossible to treat, leading to higher mortality rates. In these cases, even powerful last-resort antibiotics may be ineffective, leading to increased mortality rates [10]. The impact of AMR on mortality rates is particularly concerning in vulnerable populations, such as the elderly, newborns and individuals with weakened immune systems [16]. Reducing AMR is critical to improving global public health and preventing unnecessary deaths from infectious diseases.
Despite this urgent need to investigate AMR trends, only a handful of studies to date have reported resistance trends in polymicrobial pathogens in Pakistan. In this study, we have done a large-scale analysis of clinical data to fill the gap in AMR patterns and trends at a national level. We did a 3-year retrospective investigation of polymicrobial AMR pathogens that were isolated from clinical samples for that purpose.
While most of the previous studies were cross-sectional, presenting resistance at a single time point, our long-term analysis presents alterations in resistance patterns over 3 years. Few studies have revealed a rising tide of antibiotic resistance in polymicrobial illnesses in clinical settings in Pakistan, despite the urgent need to do so. In the present study, polymicrobial infections were 0.62% of clinical samples. Most of the isolates were recovered in the extreme age group, pus swabs samples and in urine. Geographically, most of the samples were collected from Punjab and Khyber Pakhtunkhwa. There are several studies also documenting the presence of bacteria at extreme ages and in urine and pus samples [17–19]. The prevalence of pathogens in extreme ages is mainly due to their compromised immune system and mainly in the Punjab region due to the increased population compared to other provinces.
Our data suggest that E. coli and K. pneumoniae were the predominant GNR pathogens, while S. aureus from GPC and C. albicans was also reported. E. coli were mainly recovered from pus and urine samples. Several studies previously reported the same data [20–22]. A study from Karachi also reported the prevalence of GNR associated with urinary tract infection patients [23]. Similarly, another study from Iraq also reported similar findings on the spread of E. coli and K. pneumoniae in urinary tract infection patients [24].
Polymicrobial infections can occur in a variety of settings, including hospital-acquired infections, community-acquired infections and infections in immunocompromised individuals. In the present study, the coexistence of polymicrobial infections was mainly observed between P. aeruginosa and K. pneumoniae, E. coli and S. aureus, while C. albicans was mostly detected alongside E. coli, E. faecalis and K. pneumoniae. In Pakistan there are less or no data available on the polymicrobial infections; however, studies from Lahore have documented the spread of polymicrobial infection in the pediatric patients [25]. Another study from Germany also depicted the presence of polymicrobial infections, with S. aureus and Enterococcus being the predominant pathogens [26]. The most common reason for polymicrobial infections might be poor wound care, inappropriate antibiotic usage, poor public hospital services and poor sanitation/waste-disposal conditions.
As per our knowledge, this study is the first of its kind in Pakistan and provides data on both demographic and temporal trends in polymicrobial infections of the resistant pattern.
AMR is a serious global health concern and a silent pandemic. Most of GNRs were highly resistant to β-lactams and β-lactam inhibitors, moderately resistant to quinolones and aminoglycosides, and the least resistant to colistin and carbapenems. C. albicans was resistant to micafungin, with moderate-to-low resistance to amphotericin B, voriconazole and fluconazole, while the least resistance was observed towards caspofungin and flucytosine. In total, 64.5% MRSA and 15.2% VRE was detected, 30% ESBL and 7.6% were CRE producers.
Furthermore, both GPC and GNRs were mostly resistant to commonly prescribed antibiotics including penicillins, cephalosporins, quinolones and aztreonam, and the most effective drugs remain carbapenems, vancomycin, colistin and tigecycline. We also observed the emergence of aztreonam resistance in GNR in our clinical settings, which may be due to excessive use of antibiotics or selective pressure. Most MDR strains were predominantly isolated from pus samples followed by urine, sputum and catheter tips in the present study. The MDR strains isolated from pus samples were mainly of nosocomial origin, such as P. aeruginosa, E. coli and A. baumannii.
Several studies have been published on AMR prevalence globally and the presence of MRSA, MDR, VRE, CRE and ESBL [27–30]. A study from India reported a high prevalence of CRE from clinical samples, with 59.5% of isolates belonging to the Enterobacterales family [31]. Another study from Pakistan also documented the prevalence of Gram-negative and Gram-positive bacteria in neonates with 42% neonatal mortality, with K. pneumoniae (39%) being the predominant pathogen [25]. The main triggers for AMR are access to and excess antimicrobial use, a broad range of antibiotics and excessive use of antibiotics in livestock [32].
Improving infection-control practices and antibiotic prescription strategies is suggested. Government/public management should actively take action towards implementing strategies to limit the spread of antibiotics resistant to infectious diseases. Effective treatment of polymicrobial infections requires a careful and tailored approach based on the specific pathogens involved, their susceptibilities to antimicrobial agents and the patients’ clinical circumstances [33]. It is important to prevent and control polymicrobial infections through appropriate infection-control measures and antimicrobial stewardship practices [34].
This study concluded that P. aeruginosa coexists with K. pneumoniae and E. coli coexisted with P. aeruginosa in polymicrobial infections in Pakistan. E. coli, K. pneumoniae, P. aeruginosa, S. aureus and C. albicans were found significantly in pus swabs, urine and sputum samples. Resistance to β-lactams and β-lactam inhibitors was high, fluoroquinolones and aminoglycosides was moderate, and colistin was the least resistant. MRSA, VRE, ESBL and CRE were also prevalent in polymicrobial samples.
Therefore, appropriate policies and the encouragement of infection prevention, control practices and antimicrobial stewardship measures must be taken in clinical settings. The healthcare system must be completely strengthened in order to execute the National Action Plan (NAP) on AMR, address ongoing improper antibiotic usage, and lower Pakistan’s high AMR rates. Therefore, the Ministry of Health, Pakistan must effectively implement the NAP on a national scale.
