Seasonal and hospital settings variations in antimicrobial resistance among clinical isolates from cardiac patients: insights from a 7-Year study

Abstract

Background

Antimicrobial resistance (AMR) is a growing global health challenge, particularly in specialized tertiary care settings. Despite its significance, data on the effects of seasonal variations, demographics, and healthcare settings on AMR in Pakistan are scarce.

Aim

This study aimed to evaluate the prevalence and resistance patterns of AMR and identify the key contributing factors at a cardiac hospital in Faisalabad from 2012 to 2019 to inform targeted infection control strategies.

Methodology

This retrospective cross-sectional analysis of 3,035 patient records adhered to STROBE guidelines. AMR profiles, including the multiple antibiotic resistance (MAR) index and antibiotic resistance genes (ARGs), were examined in relation to seasonality, hospital settings, and patient demographics.

Results

This study identified Staphylococcus aureus, Klebsiella pneumoniae, and Escherichia coli as the most prevalent pathogens, with S. aureus and K. pneumoniae classified among the ESKAPE group (Enterococcus faecium, S. aureus, K. pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter spp.). Isolation rates were higher in inpatients (IPs) than outpatients (OPs), with peak occurrences in autumn and winter among IPs and in spring among OPs. Age and sex significantly influenced pathogen isolation rates. The multiple antibiotic resistance (MAR) index was highest for S. aureus isolates between 2017 and 2019, which showed complete resistance to vancomycin and oxacillin. Key resistance genes mecA, vanA, tetM, and aph(3') were frequently co-detected in S. aureus. Polymyxin B and colistin remained the most effective antibiotics against multidrug-resistant (MDR) strains.

Conclusion

These findings highlight the critical need for year-round AMR surveillance, with an enhanced focus during seasonal peaks, particularly for high-risk IPs in winter and autumn. Implementing localized antimicrobial stewardship programs (ASPs) and robust infection control measures is essential to reduce the AMR burden and curb its spread in inpatient settings, especially in resource-constrained healthcare systems, such as Pakistan.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-025-11171-3.

Introduction

Antimicrobial resistance (AMR) is a global health crisis that reduces the effectiveness of antimicrobial agents, resulting in treatment failure [1]. Resistant pathogens employ molecular mechanisms to evade drugs, leading to invasive infections with poor clinical outcomes and contributing to an estimated 4.95 million deaths annually [1–4].

Patients with cardiovascular disease (CVD) are particularly vulnerable to AMR owing to prolonged hospital stays, invasive procedures, and frequent exposure to broad-spectrum antibiotics [5–7]. The confluence of these factors predisposes patients with CVD to hospital-acquired infections (HAIs), which are often resistant to first-line treatment, further complicating clinical outcomes.

In South Asia, where the prevalence of CVD is disproportionately high, countries such as Pakistan face a dual health challenge: the rising burden of CVDs and the escalating AMR crisis [8–10]. Pakistan, a low- to middle-income country (LMIC), accounts for a significant share of the global AMR-related mortality [11, 12]. Resistance rates among Escherichia coli and Staphylococcus aureus, the leading causes of HAIs, exceed 90% for critical antibiotics, such as fluoroquinolones, aminoglycosides, and β-lactams. In contrast, resistance levels in high-income countries, such as the United States and Canada, have been reported to be below 50% for similar pathogens [13]. The lack of a comprehensive AMR surveillance system in Pakistan exacerbates this issue and impedes evidence-based interventions [14, 15].

Among the bacterial pathogens of concern, methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant Staphylococcus aureus (VRSA) present significant therapeutic challenges [16]. The emergence of VRSA is particularly alarming, as vancomycin remains the cornerstone for treating serious gram-positive infections, including those in vulnerable patients with CVD [17–19]. Hospital-acquired urinary tract infections (HA-UTIs), frequently caused by VRSA, are common complications in patients with CVD, leading to extended hospital stays, higher treatment costs, and increased mortality [20]. Despite the clinical importance of these infections, data on the burden, resistance trends, and drivers of VRSA in Pakistan are limited [21].

Seasonal variations, patient demographics, and healthcare setting dynamics (e.g., inpatient (IP) versus outpatient (OP)) have been shown to influence AMR patterns globally; however, these factors are underexplored in Pakistan [22].

The absence of context-specific data has hampered the development of targeted infection control policies and stewardship programs [23]. Understanding the epidemiology and resistance profiles of AMR pathogens, particularly in specialized healthcare settings, such as cardiac hospitals, is crucial.

This study aimed to fill these gaps by investigating the prevalence, resistance patterns, and factors associated with AMR at a specialized cardiac hospital in Faisalabad, Pakistan. By analyzing the influence of sociodemographic variables, seasonal trends, and healthcare practices, this study provides actionable insights into infection prevention and control strategies. These findings contribute to strengthening AMR surveillance and stewardship efforts in Pakistan and worldwide.

Materials and methods

Study design and methodology for analyzing antimicrobial resistance data

This retrospective cross-sectional study was conducted in accordance with the STROBE guidelines (Fig. 1), analyzing AMR patterns among cardiac patients from 2011 to 2019, focusing on seasonal variations in both IP and OP hospital settings. Data were obtained from the Microbiology Laboratory of a 202-bed tertiary care cardiac hospital in Faisalabad, Pakistan, which serves three neighboring districts [24, 25].

Fig. 1.

Study design and framework for the retrospective cross-sectional analysis (2012–2019). The schematic illustrates the study design, highlighting the key steps of the retrospective cross-sectional analysis conducted in accordance with the STROBE guidelines. Data collection spanned from 2012 to 2019, encompassing patient demographics, bacterial isolate identification, antibiotic susceptibility testing, and temporal and seasonal analyses of antimicrobial resistance trends

Bacterial identification was performed using standard biochemical tests and API kits (bioMérieux, Marcy L'Etoile, France), and antimicrobial susceptibility testing was conducted according to the Clinical and Laboratory Standards Institute (CLSI) guidelines using disc diffusion [26]. Quality control was ensured by testing the reference strains alongside the clinical isolates, including E. coli ATCC 25922, S. aureus ATCC 25923, P. aeruginosa ATCC 27853, E. faecalis ATCC 29212, K. pneumoniae ATCC 700603, and A. baumannii ATCC 19606. These strains were used to validate the accuracy of susceptibility testing and to ensure adherence to CLSI performance standards.

Moreover, the antibiotic panel included beta-lactam/beta-lactamase inhibitor combinations (amoxicillin-clavulanic acid 20/10 µg, ampicillin-sulbactam 10/10 µg, piperacillin-tazobactam 30/6 µg, piperacillin 30 µg, oxacillin 5 µg), cephalosporins (ceftriaxone 30 µg, cefoxitin 30 µg, ceftazidime 30 µg, cefepime 30 µg, cefotaxime 30 µg), carbapenems (meropenem 10 µg, imipenem 10 µg, doripenem 10 µg), monobactams (aztreonam 30 µg), fluoroquinolones (ciprofloxacin 5 µg, levofloxacin 5 µg), aminoglycosides (gentamicin 10 µg, amikacin 30 µg, tobramycin 10 µg), tetracyclines (doxycycline 30 µg), glycylcyclines (tigecycline 15 µg), macrolides (azithromycin 15 µg, erythromycin 15 µg), glycopeptides (vancomycin 30 µg), oxazolidinones (linezolid 30 µg), sulfonamides and folate pathway inhibitors (trimethoprim-sulfamethoxazole 1.25/23.75 µg), and polymyxins (colistin 10 µg, polymyxin B 10 µg).

The isolates were categorized as sensitive, intermediate, or resistant based on the CLSI breakpoints. For data interpretation, isolates classified as intermediate were considered to be resistant.

Data extracted from electronic records included patient demographics (age, sex, and background), year, and season of sample collection, defined as follows: spring (March to May), summer (June to August), autumn (September to November), and winter (December to February). Additional variables included specimen type, hospital setting, isolated pathogens, and antibiotic susceptibility profiles, which were categorized as susceptible, intermediate, or resistant.

The inclusion criteria were complete patient demographics, culture data, and bacterial isolates with a minimum occurrence of 100 cases.

The exclusion criteria were incomplete records, mixed growth, low-frequency isolates, fungal cultures, and samples from healthcare providers or other environmental sources.

Of the 3,612 cases, records with missing age (n = 158), sex (n = 127), incomplete pathogen identification (n = 198), and non-patient samples (n = 74) were excluded from the study. Pathogens with fewer than 100 occurrences (n = 20) were excluded from the analysis. Intermediate resistance was classified as resistance, and the AMR data were analyzed individually across the study period. The overall study design and methodology are shown in Supplementary Fig. 1.

Multiple antibiotic resistance index

Antibiograms were used to calculate the multiple antibiotic resistance (MAR) index, an effective method for determining resistance patterns. The MAR index for a specific isolate is defined as the ratio of the number of antibiotics to which the isolate exhibits resistance to the total number of antibiotics tested against that bacterial isolate [27].

Laboratory methods for antibiotic resistance gene detection

The isolates with the highest MAR indices were selected for the detection of specific antibiotic resistance genes (ARGs). The isolates were streaked onto Blood Agar and Mannitol Salt Agar (Oxoid, Basingstoke, Hampshire, UK) to obtain pure colonies. The isolates were subsequently confirmed using a combination of biochemical assays including catalase (bioMérieux, Marcy L'Etoile, France), coagulase (Remel™ Coagulase Plasma, Thermo Fisher Scientific, Heysham, UK), latex agglutination tests (Slidex Staph Plus, bioMérieux, Marcy L'Etoile, France), and 16S rRNA gene amplification.

Genomic DNA was extracted using a commercial DNA extraction kit (Thermo Fisher Scientific, Heysham, UK), according to the manufacturer’s protocol. The quality and purity of the extracted DNA were evaluated using a NanoDrop spectrophotometer (Thermo Fisher Scientific). This rigorous methodology ensured the reliability and reproducibility of subsequent molecular analyses.

ARGs associated with specific classes of resistance were targeted using singleplex PCR with specific primers (Supplementary Table 1). Each reaction consisted of 12.5 μL of AccuPrime master mix (Invitrogen, Waltham, Massachusetts, USA), 1 μL (10 pmol) of the forward primer, 1 μL (10 pmol) of the reverse primer, 2 μL of DNA extract, and the final volume was adjusted to 25 μL with sterile H₂O, followed by amplification using an Applied Biosystems thermocycler (Thermo Fisher Scientific, Heysham, UK) with the following program: initial denaturation at 94 °C for 4 min, followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at temperatures specified in Supplementary Table 1, 45–55 s, extension at 72 °C for 55 s to 1 min, and a final extension step at 72 °C for 5 min. For analysis, the PCR products were subjected to electrophoresis, stained with 10 µM ethidium bromide, and visualized using UV transillumination with a Gel Doc EZ Imager (Bio-Rad Laboratories, Hercules, California, USA). The bacterial strains ATCC 29212, ATCC 51299, ATCC 43300, ATCC 35218, ATCC 700699, NCTC 13801, and NCTC 10418 were used as reference and positive controls to validate ARGs detection (Supplementary Table 1).

Table 1.

Temporal and demographic associations of isolation rates. This table highlights the relationships between isolation rates over time and across various demographic factors, providing insights into trends and patterns relevant to antimicrobial resistance surveillance

Factors	Total Specimens (N = numbers)	Isolates (n = numbers)	Isolates Frequency (n = numbers), %			R2 value
			Outpatients	Inpatients
Age groups (years)
0–22	416	111	(19), 17.1		(92), 82.9	0.99*
23–50	437	320	(79), 24.6		(241), 75.4
51–73	2182	1444	(462), 31.9		(982), 68.1
Gender
Female	1081	1000	(110), 11		(890), 89	p = 0.032*
Male	1954	875	(375), 42.8		(500), 57.2
Year of isolation
2012	173	121	(52), 42.9		(69), 57.1	0.97*
2013	240	134	(55), 41.0		(79), 59.0
2014	291	155	(71), 45.8		(84), 54.2
2015	331	198	(93), 46.9		(105), 53.1
2016	400	230	(99), 43.0		(131), 57.0
2017	422	290	(110), 37.9		(180), 62.1
2018	520	310	(115), 37.0		(195), 63.0
2019	658	437	(127), 29.0		(310), 71.0
Season
Spring	297	177	(65), 36.7		(112), 63.3	0.95*
Summer	410	288	(102), 35.4		(186), 64.5
Autumn	661	590	(149), 25.2		(441), 74.7
Winter	1667	820	(230), 28.0		(590), 72.0
Specimen type
Sputum	21	5	(1), 20.0		(4), 80.0	0.99*
Fluids and effusions	29	13	(5), 38.4		(8), 61.5
Wound and pus	80	60	(17), 28.3		(43), 71.6
Catheter tips	230	110	(37), 33.6		(73), 66.4
Urine	583	400	(105), 26.2		(295), 73.8
Blood	2092	1287	(367), 28.5		(920), 71.4
Total (Each group) (n = numbers)	3035	1875	(Outpatients + Inpatients = 1875)

Statistical analysis and interpretation of data

To illustrate the study design, spatial data for the study area were sourced from an online repository at https://www.diva-gis.org. A geographic information system (GIS) map was generated using the sf library in RStudio (version 4.3.2; RStudio, Boston, MA, USA). The study design was further visualized and explained using cartographic tools available at Biorender (https://www.biorender.com).

Sociodemographic data were entered into Microsoft Access XP (version 2404; Microsoft Inc., Redmond, WA, USA) and exported for analysis. The data were imported into the SPSS statistical software (version 16.0; SPSS Inc., Chicago, IL, USA). Descriptive statistics, such as frequencies and percentages of categorical variables, were calculated. Categorical data were compared using the chi-square test. Statistical significance was determined using two-sided p-values, with a p-value < 0.05 considered significant.

To assess the correlation matrix coefficients between demographic variables and isolation rates across hospital settings, a linear regression model was applied using RStudio (version 4.3.2; RStudio Software Inc., Boston, MA, USA).

Logistic regression analysis was conducted to identify the independent predictors of the risk factors associated with AMR. The final independent variables of the multivariate model were selected using a stepwise forward approach, with univariate analysis as the initial step to identify relevant variables for inclusion. In the final model, we assessed collinearity and interactions between variables by calculating adjusted odds ratios (aOR) and 95% confidence intervals (CI) for the independent risk factors. The robustness of the model was evaluated using the Hosmer–Lemeshow goodness-of-fit test (SPSS version 16.0, SPSS Inc., Chicago, IL, USA).

The Cochran-Armitage trend test was performed using the “DescTools” package in RStudio (Version 4.3.2, RStudio Software Inc., Boston, MA, USA), categorizing resistance data by annual time intervals, and calculating the test statistics to identify significant linear trends in AMR rates.

ARGs patterns were analyzed by clustering with proximity enhancement, and the results were visualized on a heat map. A complete-linkage hierarchical clustering heat map of the antibiotic resistance profiles of the isolates was generated using XLSTAT software (version 2025.1; (Addinsoft, New York, USA).

To investigate the co-occurrence of ARGs, a correlation matrix was computed using the “cor” function in RStudio (version 4.3.2; RStudio, Boston, MA, USA). The correlation results were visualized using the Corrplot package in R.

Results

Temporal trends and demographic correlations in clinical isolation rates in a hospital setting

A total of 3,035 specimens were analyzed, yielding 1,875 isolates (61.7%) with notable variations across demographic factors and specimen types. The highest isolation rate was observed in the 51–73 years age group, where 68.1% of isolates were from inpatients (IPs) and 31.9% were from outpatients (Ops), demonstrating a strong age-related correlation (R² = 0.99) (p < 0.05). Males exhibited a significantly higher OP isolation rate (42.8%) than females (11%), with a statistically significant sex difference (χ² = 246.9, p < 0.05).

Over time, OP isolation rates decreased from 46.9% in 2015 to 29.0% in 2019, with a strong correlation between the year and isolation frequency (R² = 0.97) (p < 0.05).

The seasonal analysis revealed that IPs had the highest isolation rates in autumn (74.7%) and winter (72.0%), whereas OPs had higher rates in spring (36.7%) and summer (35.4%), with a marked decrease in autumn (10.2%). This seasonal variation was strongly correlated with the isolation frequency (R² = 0.95) (p < 0.05).

The specimen type significantly influenced the isolation rate. Blood samples (68.6%) were predominant across all specimen types (71.6% of IPs), followed by urine samples (21.3%), which showed a higher isolation rate among IPs (73.8%). The correlation between the specimen type and isolation frequency was highly significant (R² = 0.99) (p < 0.05).

These findings revealed significant associations between isolation rates and key demographic variables, including age, sex, seasonality, and year of isolation, offering valuable epidemiological insights into AMR patterns (Table 1).

Prevalence and characterization of isolates in a hospital setting

To elucidate the spectrum of bacterial isolates responsible for infections in the study population, a comprehensive prevalence analysis was conducted. The most prevalent bacterial isolate among cardiac patients was Staphylococcus aureus (27.84%, n = 522), followed by Klebsiella pneumoniae (22.4%, n = 420), and Escherichia coli (20.5%, n = 386). Other significant isolates included Acinetobacter baumannii (15.6%, n = 294), Pseudomonas aeruginosa (7.25%, n = 136), and Enterococcus faecalis (6.2%, n = 117) (Fig. 2A). These findings underscore the clinical significance of S. aureus infections in this patient cohort.

Fig. 2.

Prevalence and distribution of bacterial isolates in a hospital setting.A Bar graph depicting the overall prevalence of bacterial isolates linked to healthcare-associated infections (HAIs) related to ESKAPE pathogens. B Box plot comparing the total number of bacterial isolates recovered from inpatients and outpatients, highlighting the variations between the patient groups. C Horizontal bar graph showing the relative prevalence of ESKAPE pathogens in inpatients and outpatients. This figure highlights the distribution and variability of major bacterial isolates across patient categories and hospital settings, emphasizing the significance of ESKAPE pathogens in the hospital studied. Key: % Prevalence = (number of isolates/total number of isolates) × 100

To further evaluate the burden of bacterial isolates in hospital settings, linear regression analysis (R² = 0.0467, p = 0.0237) revealed a significantly higher isolation rate among IPs than that among OPs (Fig. 2B). The predominant isolates in the IPs group were S. aureus (83%, n = 433) and K. pneumoniae (80.2%, n = 337), whereas A. baumannii (35.7%, n = 105) and P. aeruginosa (33.8%, n = 46) were more prevalent in the OPs (t = 3.201, p < 0.05) (Fig. 2C). These results highlight the differential bacterial burden across hospital settings.

Seasonal and annual trends of bacterial isolates were assessed using ANOVA with Tukey’s Honestly Significant Difference (HSD) post-hoc test. Among the IPs, S. aureus showed significant seasonal variation, with higher isolation rates in autumn (F = 9.8, p < 0.05), summer (F = 21.08, p < 0.05), and winter (F = 28.63, p < 0.05), peaking in autumn of 2019. Similarly, K. pneumoniae exhibited significant variation across all seasons, peaking in winter 2019.

Among the OPs, A. baumannii did not show significant variation in winter (F = 2.06, p > 0.05) but reached its highest isolation rate in winter 2019. P. aeruginosa showed significant variation in autumn (F = 18.67, p < 0.05), with its peak occurring in the winter of 2019. These data suggest notable seasonal fluctuations in the bacterial prevalence across hospital settings (Fig. 3). The prevalence of bacterial isolates across different demographic variables is presented in Supplementary Table 2.

Fig. 3.

Seasonal and annual trends in bacterial isolate prevalence. This figure depicts seasonal and annual variations in the prevalence of bacterial isolates among inpatients and outpatients, highlighting their temporal dynamics and potential implications for patient care and infection management strategies. Key: % Prevalence = (Number of isolates/Total number of isolates) × 100, Inpatients (IPs), Outpatients (OPs)

Seasonal determinants and variation in antibiotic resistance patterns among predominant hospital isolates

Using the Cochran-Armitage trend test and a multivariate regression model, we evaluated seasonal determinants and annual resistance trends among the predominant hospital isolates. Our findings indicated significant seasonal associations, with IPs isolates displaying higher AMR rates (65.3%) than OPs isolates (29.2%) (p < 0.05), particularly during the winter season (aOR = 2.50, 95% CI: 1.20–4.98, p < 0.05), which accounted for 98% of the observed variance in AMR rates (R² = 0.98).

S. aureus: In IPs isolates, resistance peaked during winter for most antibiotics (R² = 0.91). Vancomycin (VA), tigecycline (TGC), and sulfamethoxazole/trimethoprim (SXT) exhibited the highest sensitivity (18.8%), whereas resistance rates were highest for amoxicillin/clavulanate (AMC) and ciprofloxacin (CIP) (97.5%). Among the OPs isolates, resistance displayed dual seasonal peaks in winter and summer, with consistent winter trends from 2012 to 2019 and increased summer resistance from 2017 to 2019 (Z = 2.56, p < 0.05) (R² = 0.91), especially against ceftriaxone (CRO), gentamicin (GN), CIP, azithromycin (AZM), and AMC (100%). VA, SXT, TGC, and linezolid (LZD) retained sensitivity (100%) in OPs isolates, whereas AMC, AZM, and CIP showed the highest resistance rates (100%) (Fig. 4).

Fig. 4.

Antimicrobial resistance patterns among predominant isolates. This figure presents radar graphs depicting the antimicrobial resistance (AMR) percentages of the predominant bacterial isolates analyzed across seasonal and annual trends to assess temporal variations. Susceptibility testing was conducted following the Clinical and Laboratory Standards Institute (CLSI) guidelines using Oxoid discs (Basingstoke Hampshire, UK). The antibiotics tested are grouped as follows: beta-lactams, including amoxiclav (AMC, 40 µg), ceftriaxone (CRO, 30 µg), cefoxitin (FOX, 30 µg), oxacillin (OX, 30 µg), ceftazidime (CAZ, 30 µg), cefepime (FEP, 30 µg), ampicillin/sulbactam (SAM, 40 µg), piperacillin/tazobactam (TZP, 100 µg), cefotaxime (CTX, 30 µg), aztreonam (ATM, 30 µg), imipenem (IMP, 10 µg), doripenem (DOR, 10 µg), meropenem (MEM, 10 µg), and piperacillin (PRL, 30 µg); macrolides, including azithromycin (AZM, 30 µg) and erythromycin (ETM, 15 µg); quinolones, including ciprofloxacin (CIP, 5 µg) and levofloxacin (LEV, 5 µg); aminoglycosides, including gentamicin (CN, 10 µg), tobramycin (TOB, 10 µg), and amikacin (AK, 30 µg); glycopeptides, including vancomycin (VA, 30 µg) and teicoplanin (TGC, 30 µg); oxazolidinones, including linezolid (LZD, 30 µg); and others, including trimethoprim/sulfamethoxazole (SXT, 25 µg), polymyxin B (PB, 10 µg), and doxycycline (DOX, 30 µg). This comprehensive analysis provides insights into the seasonal and temporal dynamics of resistance patterns and offers valuable information for understanding the AMR trends. Key: Inpatients (IPs), outpatients (OPs), and AMR percentages were calculated as follows: (number of resistant isolates/total isolates) × 100

K. pneumoniae: Winter was associated with resistance peaks among IPs isolates, with notable annual fluctuations (R² = 0.90). Resistance to ceftazidime (CAZ) and cefepime (FEP) was the highest (100%) in 2012 but gradually declined in 2015 (71%) (Z = −3.20, p < 0.05) and remained elevated for CIP and CTX (29%) (Z = 0.95, p > 0.05), whereas amikacin (AK) (50%) and gentamicin (CN) (100%) exhibited increasing resistance trends (Z = 1.35, p > 0.05). From 2016 to 2019, the resistance levels stabilized for CIP (38.8–75%) (Z = 3.25, p < 0.05) and CTX (35.2%–64.5%) (Z = 2.89, p < 0.05), with intermittent carbapenem resistance noted for ertapenem (ETP) (46.3%) (Z = 0.15, p > 0.05) and meropenem (MEM) (55.4%) (Z = 0.28, p > 0.05). Colistin (CT) and polymyxin B (PB) consistently demonstrated sensitivity (100%), whereas TGC, piperacillin/tazobactam (TZP), ceftazidime (CAZ), and ampicillin/sulbactam (SAM) showed high resistance rates (100%). Among the OP isolates, AMR exhibited a notable increase during the spring seasons from 2012 to 2016, particularly for SAM, TZP, CTX, and MEM, with resistance rates approaching 100% (Z = 2.3, p < 0.05) (R² = 0.52). A temporary reduction in resistance was observed in 2017 for SAM, CAZ, aztreonam (ATM), FEP, MEM, and LEV, showing a 50% decrease that extended to 2018. This was followed by a modest resurgence in the resistance levels in 2019. By the end of the observation period, the resistance rates for SAM, ATM, and LEV increased to 66.6% (Z = 2.5, p < 0.05), CAZ and FEP demonstrated complete resistance at 100% (Z = 4.1, p < 0.05), and MEM displayed a resistance rate of 33.3% (Z = −0.5, p > 0.623) (Fig. 4).

A. baumannii: Autumn peaks in resistance were observed in IP isolates from 2012 to 2019, with high resistance rates for SAM, CAZ, and FEP in 2012 (100%), extending to the TGC in 2013 (100%) (R² = 0.69). Resistance declined in 2014 (100%) and remained low until a resurgence in 2018 for SAM, CAZ, FEP, IPM, DOX, and MEM (100%) (Z = −2.68, p < 0.05). By 2019, the resistance levels for SAM, CAZ, CIP, and LEV had remained elevated, whereas CT and PB remained consistently susceptible (100%) (Z = 2.35, p < 0.05). The resistance spiked during spring from 2015 to 2019, with resistance levels near 100% for FEP (Z = 3.45, p < 0.05), MEM (Z = 3.89, p < 0.05), and CIP (Z = 4.12, p < 0.05) (R² = 0.57). CT and PB consistently maintained a 100% sensitivity (Fig. 4).

P. aeruginosa: Resistance among IPs isolates peaked in summer, with near-total resistance to CAZ, FEP, and MEM (100%), except for slight reductions in 2015 and 2017 (50%) (Z = −0.156, p > 0.05) (R² = 0.52). The CT and PB remained sensitive throughout the study period (100%). OPs isolates exhibited peak resistance during winter, with elevated rates of piperacillin (PRL), piperacillin/tazobactam (TZP), ceftazidime (CAZ), cefepime (FEP), aztreonam (ATM), IPM, MEM, doripenem (DOR), CN, tobramycin (TOB), AK, CIP, and LEV (100%) (Z = 3.12, p < 0.05) (R² = 0.91). Temporary reductions in resistance were noted in 2015 (50%), 2016 (66.6%), and 2019 (40%) (Z = −0.196, p > 0.05), whereas CT consistently showed effectiveness in these isolates (100%) (Fig. 4).

These results suggest seasonal peaks in AMR, particularly in the winter, highlighting the need for targeted antimicrobial stewardship and tailored interventions during high-risk periods. The AMR patterns of the other bacterial isolates are shown in Supplementary Fig. 2.

Trends in multiple antibiotic resistance among predominant hospital isolates: annual variations and peak seasonal patterns

The MAR index was calculated and analyzed across years and hospital settings to investigate the resistance dynamics associated with the peak AMR seasons among the predominant hospital isolates. Comparative analysis revealed distinct patterns in MAR burden.

Among all the isolates, S. aureus exhibited the highest MAR index (0.9–1.0), predominantly in IP settings, with a pronounced peak during winter in 2017, 2018, and 2019. P. aeruginosa followed closely, showing elevated MAR values (0.8–0.89 and 0.9–1.0) in both IP and OP settings, with seasonal peaks observed in summer (2018) and winter (2016–2018) (Table 2).

Table 2.

Trends in MAR Index among predominant hospital isolates. During peak AMR seasons, the predominant isolates were analyzed to calculate the MAR index. This analysis highlighted the annual variations and peak AMR seasons. S. aureus and P. aeruginosa exhibited the highest MAR index, with notable seasonal peaks in winter and summer, emphasizing the importance of targeted surveillance during high-risk periods. MAR index values were categorized as follows: 0.2–0.5, multidrug resistance (MDR); 0.5– < 1.0, extensive drug resistance (XDR); and 1.0, pan-drug resistance (PDR). Note: In S. aureus, resistance rates peaked during winter for both inpatients and outpatients. K. pneumoniae exhibited winter peaks in inpatients and spring peaks in outpatients. For A. baumannii, the highest AMR rates were observed in autumn among inpatients and in spring among outpatients. Similarly, P. aeruginosa demonstrated summer peaks in inpatients and winter peaks in outpatients, highlighting the pathogen-specific seasonal resistance patterns. Key: MAR, multiple antibiotic resistance. AMR: antimicrobial resistance. N denotes the total number of isolates, and n represents the number of isolates in a specific subset or group

Hospital Settings	Years	Total, (N)				MAR Index, (n)
						0.5–0.69				0.7–0.79
		S. aureus	K. pneumoniae	A. baumanii	P. aeruginosa	S. aureus	K. pneumoniae	A. baumanii	P. aeruginosa	S. aureus	K. pneumoniae
Inpatients	2012	5	2	1	1	-	2	1	1	-	-
	2013	17	9	1	1	-	-	-	-	-	-
	2014	18	13	2	1	-	-	-	-	-	-
	2015	19	14	3	2	-	-	-	2	-	-
	2016	21	16	7	2	-	16	-	2	-	-
	2017	25	17	7	2	-	17	-	2	-	-
	2018	26	18	8	2	-	-	8	-	-	-
	2019	27	48	9	8	-	48	-	-	-	-
Outpatients	2012	1	1	0	1	1	1	-	-	-	-
	2013	2	1	0	1	2	-	-	-	-	-
	2014	3	1	0	2	3	-	-	-	-	-
	2015	4	1	1	2	4	-	1	-	-	-
	2016	5	1	1	3	5	1	1	-	-	-
	2017	6	1	1	3	6	1	-	-	-	-
	2018	8	2	1	4	8	2	1	-	-	-
	2019	11	3	1	10	11	3	-	-	-	-

Hospital Settings
			0.8–0.89				0.9–1
	A. baumanii	P. aeruginosa	S. aureus	K. pneumoniae	A. baumanii	P. aeruginosa	S. aureus	K. pneumoniae	A. baumanii	P. aeruginosa
Inpatients	-	-	-	-	-	-	-	2	-	-
	-	1	-	-	-	-	17		-	-
	-	-	-	-	-	-	18	-	-	1
	-	-	-	-	-	-	19	-	-	-
	-	-	-	-	-	-	21	16	-	-
	-	-	-	-	-	-	25	17	-	-
	-	-	-	-	-	-	26	-	-	2
	9	8	-	-	-	-	27	48	-	-
Outpatients	-	1	-	-	-	-	-	-	-	-
	-	1	-	-	-	-	-	-	-	-
	-	2	-	-	-	-	-	-	-	-
	-	2	-	-	-	-	-	-	-	-
	-	-	-	-	-	-	-	-	-	-
	1	-	-	-	-	-	-	-	-	-
	-	-	-	-	-	-	-	-	-	-
	1	10	-	-	-	-	-	-	-	-

These findings highlight the disproportionate contribution of these pathogens to the hospital AMR burden, emphasizing the urgent need for targeted surveillance and robust infection control strategies during high-risk periods of S. aureus infection.

Association of antibiotic resistance genes with high multiple antibiotic resistance index isolates: a laboratory-based study

To assess the burden and temporal trends of AMR among S. aureus isolates with a high MAR index, the prevalence of key ARGs was evaluated using singleplex PCR. Notably, mecA, vanA, and tetM demonstrated consistently high prevalence (> 70%) during the peak MAR index years (2017–2019), whereas mecB, cfr, and dfrA1 were detected at lower frequencies (< 30%). The ermB, ermC, gyrA, gyrB, sul1, and sul2 genes were moderately persistent (50–70%) (Fig. 5A). These findings underscore the persistent threat posed by critical ARGs, particularly mecA and vanA, which confirm the presence of MRSA and VRSA, respectively. This highlights the urgent need for enhanced AMR surveillance and targeted interventions.

Fig. 5.

Molecular profiling of antibiotic resistance genes. A Prevalence of ARGs detected via singleplex PCR in the highest MAR index harboring S. aureus isolates from 2017 to 2019, with vanA and mecA being identified as the predominant resistance genes. B Hierarchical clustering (complete linkage) illustrates the similarity patterns among various ARGs, providing insights into the structure and dynamics of the resistome. C Correlation matrix analysis revealed positive associations among vanA, mecA, tetM, and aph(3'), suggesting co-occurrence and potential horizontal dissemination during this period. Key: Antibiotic resistance genes (ARGs), multiple antibiotic resistance (MAR), beta-lactams: mecA, mecB, aminoglycosides: aac(6'), aph(3') macrolides: ermB, ermC, quinolones: gyrA, gyrB, parC, oxazolidinones: cfr, glycopeptides: vanA, vanB, tetracyclines: tetA, tetM, sulfonamides, and trimethoprim: sul1, sul2, dfrA1

Hierarchical clustering using a complete-linkage approach revealed complete similarity among vanA, mecA, tetM, and aph (3'), while other ARGs formed clusters with varying degrees of similarity (Fig. 5B). This indicates that these ARGs play a prominent role in driving resistance and facilitating their dissemination.

A correlation matrix further confirmed the strong associations among vanA, mecA, tetM, and aph (3') (r = 1.0, p < 0.05), suggesting their co-occurrence and collective dissemination (Fig. 5C). These results highlight the complexity of the resistance profile of S. aureus, emphasizing the intricate interactions among ARGs that shape resistance patterns.

Discussion

This study provides a comprehensive analysis of AMR trends, bacterial prevalence, and seasonal dynamics in a specialized cardiac hospital in Pakistan, emphasizing the differential pathogen burden across the IP and OP settings. As the first investigation in this region, our findings underscore the importance of local epidemiological data for understanding seasonal AMR fluctuations and guiding targeted infection control strategies in high-risk healthcare environments.

Higher pathogen isolation rates in the 51–73-year age group, along with significant sex-based differences, align with the established evidence of the increased susceptibility of older patients to HAIs [28, 29]. Patients with CVDs are particularly vulnerable to immunosuppression and prolonged hospital stays, predisposing them to HAIs [30, 31].

ESKAPE (Enterococcus faecium, S. aureus, K. pneumoniae, A. baumannii, P. aeruginosa, and Enterobacter species) pathogens have emerged as the predominant contributors to HAIs, imposing substantial morbidity and mortality burdens [32]. In this study, key ESKAPE pathogens were identified, with the exception of E. faecium and Enterobacter spp., with S. aureus being the most prevalent isolate, followed by K. pneumoniae. This contrasts with prior studies, in which K. pneumoniae was the dominant pathogen [33].

Pathogen distribution varies significantly among different hospital settings, influencing infection control strategies. S. aureus (83%) and K. pneumoniae (80.2%) predominated in IPs, consistent with their strong association with HAIs [34]. In contrast, A. baumannii (35.7%) and P. aeruginosa (33.8%) were more prevalent in OPs, which is consistent with reports of their dominance in community-acquired infections, particularly respiratory infections [35]. The higher pathogenic burden in IPs highlights the impact of prolonged hospital stays, invasive procedures, and antimicrobial selection pressure, necessitating stringent surveillance and targeted interventions [36].

Seasonal variations significantly influenced HAIs, with S. aureus and K. pneumoniae exhibiting peak prevalence in the autumn and winter, respectively. Increased HAIs rates during colder months are linked to shifts in host immunity, higher hospital admissions, and prolonged indoor exposure [37]. While environmental parameters have not been directly assessed, prior studies suggest that humidity, temperature fluctuations, and air quality influence bacterial survival and transmission [38]. Understanding these seasonal patterns is critical for optimizing infection control strategies, resource allocation, and antimicrobial stewardship.

AMR rates were significantly higher in IP isolates (65.3%) than in OP isolates (29.2%), with winter emerging as the strongest predictor (aOR = 2.50, p < 0.05). S. aureus resistance peaked in winter (R² = 0.91), particularly against AMC and CIP (97.5%), while K. pneumoniae exhibited sustained resistance to CIP, CTX, and aminoglycosides. These trends align with increased respiratory and bloodstream infections, likely driven by seasonal fluctuations in temperature and humidity, which enhance pathogen persistence and transmission [22]. Winter surges in respiratory illnesses lead to increased antibiotic use, particularly β-lactams and fluoroquinolones, thereby accelerating the development of resistance [22].

A major driver of winter AMR is the surge in empirical antibiotic use for respiratory infections, which exerts a selective pressure that accelerates the evolution of resistance [39]. The increased antibiotic prescription rates during winter reinforce the link between seasonal infection patterns and AMR [40]. The strong correlation between resistance trends and seasonal peaks (R² = 0.98) underscores the need for adaptive interventions, including enhanced hand hygiene, environmental decontamination, and antimicrobial stewardship, particularly during winter.

The pronounced seasonal variations in AMR observed in this study have direct implications for empirical antibiotic prescribing, particularly during high-transmission periods. The winter resistance peaks most notably among S. aureus and K. pneumoniae underscore the need for seasonally tailored prescribing strategies. During colder months, clinicians should exercise greater caution with β-lactams and fluoroquinolones, favoring therapy guided by real-time local antibiograms. Where appropriate, alternative or combination regimens may enhance therapeutic outcomes while reducing the selection pressure. Incorporating seasonal resistance surveillance into stewardship programs will enable the formulation of dynamic and seasonally adjusted treatment protocols. Regularly updating hospital formularies and resistance profiles in accordance with seasonal trends will support evidence-based clinical decisions, reduce inappropriate prescribing, and ultimately curb resistance proliferation.

The seasonal dynamics of AMR, as reflected by the MAR index, revealed pronounced temporal variability between S. aureus and P. aeruginosa. S. aureus exhibited a MAR index of 0.9–1.0 among IPs, peaking in winter (2017–2019), coinciding with heightened antimicrobial use [41]. P. aeruginosa resistance peaked in both summer and winter, suggesting a complex interplay between antibiotic consumption, microbial activity, and environmental conditions [42]. Elevated AMR rates during warmer months may stem from increased microbial proliferation, whereas winter resistance surges align with intensified antibiotic use in response to seasonal infection spikes [22]. Targeted and seasonally adaptive interventions, including continuous surveillance and climate-informed treatment guidelines, are essential for mitigating seasonal AMR escalation.

The high MAR index in S. aureus correlates with the prevalence of key ARGs, including mecA, vanA, and tetM, along with resistance determinants for β-lactams and aminoglycosides. The strong co-occurrence of these ARGs (r = 1.0, p < 0.05) in high-MAR isolates suggests a complex resistance network, likely shaped by seasonal antibiotic selection pressures. Understanding these genetic signatures is crucial for refining intervention strategies because horizontal gene transfer (HGT) may drive sustained resistance surges during high-risk periods.

The detection of mecA, vanA, tetM, and aph(3) highlights the genetic basis of AMR in this setting. mecA encodes penicillin-binding protein 2a (PBP2a), which confers MRSA resistance by reducing β-lactam affinity [43]. Similarly, vanA mediates VRSA by modifying peptidoglycan precursors, thereby preventing antibiotic binding [44]. tetM, often plasmid- or transposon-borne, provides tetracycline resistance via ribosomal protection, whereas aph(3') encodes aminoglycoside-modifying enzymes that inactivate antibiotics [45]. The co-occurrence of these genes within the same isolates suggests that plasmid-mediated HGT plays a major role in the dissemination of resistance, particularly in hospital settings with a high transmission potential.

Although this study focused on ARGs, efflux pump-mediated resistance was not analyzed, although its role in multidrug resistance (MDR) has been well established. Efflux pumps, such as norA in S. aureus, actively extrude antibiotics, contributing to treatment failure [46, 47]. Future studies should incorporate efflux pump analysis to provide a more comprehensive understanding of AMR mechanisms in high-MAR pathogens.

The emergence of MRSA and VRSA presents a significant clinical challenge, driven by the acquisition of mecA (80–90%) and vanA (70–80%). The high MAR index (0.9–1) in S. aureus underscores the growing threat of these strains to HAIs. While our findings align with global MRSA prevalence trends, notable deviations in VRSA patterns warrant further investigation of alternative genetic determinants [48]. The increasing prevalence of these strains necessitates enhanced surveillance, targeted infection control strategies, and the development of alternative antimicrobial agents or combination therapies.

Consistent with the trends in low- and middle-income countries (LMICs), the high burden of MRSA and VRSA likely reflects inadequate antibiotic regulation and widespread misuse [49]. According to World Health Organization (WHO) reports, the prevalence of MRSA in South Asia (45%) is nearly double that of North America (23%) and Europe (18%), highlighting the consequences of unregulated antimicrobial use [50]. High-income countries have successfully reduced resistance rates through robust antimicrobial stewardship programs (ASPs), thereby providing a model for LMICs. The observed prevalence of vanA-positive VRSA (~ 70–80%) in our study was significantly higher than that reported in Bangladesh (8%) and Nepal (11%), suggesting an intensified selection pressure in this setting [51, 52].

The high prevalence of mecA and vanA in 2017–2019 underscores their critical role in the dissemination of resistance. Strengthening ASPs, implementing rapid molecular diagnostics, and optimizing treatment regimens are crucial steps for mitigating the burden of MRSA and VRSA infections.

These findings underscore the need for a data-driven approach that integrates seasonal trends, molecular diagnostics, and ASPs to mitigate AMR in high-risk healthcare settings. Implementing seasonally adaptive prescribing protocols guided by real-time resistance surveillance is a crucial step towards curbing resistance escalation and improving patient outcomes in hospitals. Despite these important insights, certain limitations must be acknowledged to fully contextualize the findings and their implications.

Limitations

We acknowledge that the single-center nature of our study may introduce regional bias and limit the generalizability of our findings to other healthcare settings with different patient demographics, prescribing patterns, and infection control practices. While our study provides valuable insights into seasonal AMR dynamics in a specialized cardiac hospital, multicenter investigations across diverse geographical and institutional contexts are essential to validate these observations and enhance their broader applicability to other settings. Additionally, although this study provides valuable insights, the absence of comprehensive genomic analysis, plasmid profiling, and environmental assessments limits our understanding of the specific resistance mechanisms and their drivers.

The seven-year dataset, though extensive, may not fully capture long-term resistance trends, underscoring the need for continuous surveillance. Moreover, the lack of clinical outcome data restricts the translational impact of our findings; future studies integrating clinical outcomes with resistance patterns would provide a more comprehensive framework for ASPs and infection control.

Conclusion

This study highlights the substantial burden of AMR in cardiac care, revealing distinct transmission dynamics among S. aureus, K. pneumoniae, A. baumannii, and P. aeruginosa across hospital settings. Seasonal fluctuations in AMR, particularly winter peaks in S. aureus and K. pneumoniae, suggest a potential role for environmental and clinical factors in shaping resistance trends.

The high MAR index in S. aureus, along with the strong co-occurrence of mecA, vanA, tetM, and aph(3), underscores the complexity of resistance mechanisms, which are likely influenced by HGT. Future research integrating whole-genome sequencing (WGS) and longitudinal genomic surveillance is critical for characterizing resistance determinants and monitoring their evolution.

Additionally, investigating environmental reservoirs and hospital-specific practices will provide a more comprehensive understanding of the persistence of AMR. Targeted infection control strategies coupled with improved ASPs are essential to mitigate resistance escalation and improve patient outcomes in high-risk healthcare settings.

Authors’ contributions

Muhammad Umer Asghar: Conceptualization, Data curation, formal analysis, Investigation, Methodology, Resources, Software, Validation, Visualization, Writing – original draft, writing – review, and editing. Arsalan Haseeb Zaidi: Supervision, Validation. Muhammad Tariq: Supervision, Validation. Noor Ul Ain: Investigation, Writing, review, and editing.

Funding

This study did not receive any specific grants from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

All data analyzed in this study were obtained from the corresponding author upon request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the ethical principles of the Declaration of Helsinki. The direct involvement of patients was not included because only medical data were used. This study was approved by the Institutional Ethical Review Committee of the Faisalabad Institute of Cardiology (FIC) (approval no. 17–2019/DME/FIC/FSD). As this study involved retrospective medical data, the requirement for informed consent was waived by the Institutional Ethical Review Committee of the FIC.

Consent for publication

All the authors have reviewed and approved the final manuscript.

Competing interests

The authors declare no competing interests.