Incidence, microbiological spectrum, and antimicrobial resistance patterns of surgical site infections at a tertiary care hospital in Pakistan: a retrospective observational study

Abstract

Background

Surgical site infections (SSIs) are among the most common healthcare-associated infections and contribute substantially to postoperative morbidity, prolonged hospitalization, and increased healthcare costs. Data on SSI epidemiology, microbiological profiles, and antimicrobial resistance patterns remain limited in many low- and middle-income countries, including Pakistan. This study aimed to determine the incidence of surgical site infections and characterize the associated microbiological spectrum and antimicrobial resistance patterns using surveillance data from a tertiary care hospital in Karachi, Pakistan.

Methods

This single-center retrospective cohort study was conducted at Aga Khan University Hospital, Karachi. Surgical procedures performed between January 2020 and December 2023 were identified from the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) database. Surgical site infections were defined according to standardized ACS NSQIP criteria. SSI incidence was calculated across nine surgical specialties including colorectal, neurosurgery, otolaryngology, orthopedic, cardiac, vascular, plastic, gynecologic, and general surgery. Microbiological and antimicrobial susceptibility data for culture-positive SSI cases were extracted from electronic medical records. Descriptive analyses were used to summarize pathogen distribution and antimicrobial resistance patterns.

Results

A total of 37,768 surgical procedures were analyzed, among which 557 SSIs were identified, corresponding to an overall incidence of 1.5% (95% CI: 1.36–1.61). Higher SSI occurrence was observed in colorectal surgery (6%), neurosurgery (3.1%), and otolaryngology procedures (2.7%), whereas lower rates were seen in vascular (0.1%), cardiac (0.4%), and orthopedic procedures (0.4%). Wound cultures were obtained in 247 of 557 SSI cases (44.3%), yielding 208 microbial isolates. Gram-negative bacteria predominated (52.9%), followed by Gram-positive cocci (41%) and Candida species (3.8%). The most frequently isolated pathogens included Escherichia coli, Pseudomonas aeruginosa, and Staphylococcus aureus. Among Gram-negative isolates tested, resistance was highest to amoxicillin–clavulanate (63.0%) and ciprofloxacin (40.8%), whereas lower resistance rates were observed for amikacin (6.5%) and imipenem (12.5%).

Conclusions

This study provides current data on the epidemiology and microbiology of surgical site infections in a tertiary care hospital in Pakistan. The predominance of Gram-negative pathogens and the observed antimicrobial resistance patterns highlight the importance of ongoing microbiological surveillance and integration of local antibiogram data to guide empiric antimicrobial therapy, optimize surgical prophylaxis, and strengthen infection prevention strategies.

Clinical trial

Not applicable.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-026-13312-8.

Introduction

Healthcare-associated infections (HAIs) remain a major global threat to patient safety and are a substantial cause of preventable morbidity and mortality worldwide. Among HAIs, surgical site infections (SSIs) represent one of the most common postoperative complications associated with prolonged hospitalization, increased antimicrobial use, and higher healthcare costs particularly in low- and middle-income countries (LMICs) [1, 2].

The incidence of surgical site infections (SSIs) varies across regions. HICs, reported SSI rates typically range from 1 to 5% whereas LMICs bear a substantially elevated burden, with pooled estimates of approximately 12% and ranges up to 24%, per World Health Organization data [3–5]. However, there is a paucity of data on SSI rates in developing countries such as Pakistan. Most studies had a small sample size, single-center studies, and varying methodologies with SSI rates varying from 9% to 34%. There is a lack of large-scale, standardized surveillance data that integrates clinical outcomes with microbiological and antimicrobial resistance pattern, particularly in the post-COVID era. This gap makes it difficult to develop locally relevant infection prevention strategies and empiric treatment guidelines [6, 7]. Beyond incidence rate, microbiological etiology of SSIs varies significantly between HICs and LMICs. In HICs, Gram-positive organisms such as Staphylococcus aureus and Staphylococcus species predominate while LMICs report a higher proportion of Gram-negative pathogens, including Escherichia coli, Klebsiella species, and Pseudomonas species. An additional challenge is the increasing burden of antimicrobial resistance. In many LMIC settings, Gram-negative organisms demonstrate high resistance to commonly used antibiotics, including third-generation cephalosporins and fluoroquinolones. Of particular concern are resistance phenotypes such as methicillin-resistant Staphylococcus aureus (MRSA), extended-spectrum beta-lactamase (ESBL)-producing Enterobacterales, and carbapenem-resistant organisms, all of which complicate treatment and limit available therapeutic options. SSIs further varies according to the type of surgery for instance clean surgeries (e.g., craniotomies, cesarean sections) typically yield Staphylococcus aureus and Staphylococcus epidermidis, whereas gastrointestinal and genitourinary surgeries more frequently yield Gram-negatives predominance [8–11]. The American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) is a standardized clinical registry that enables surveillance and benchmarking of surgical outcomes, including SSIs.

Identifying surgical procedures with higher SSI rates, predominant pathogens, and their antimicrobial susceptibility patterns is crucial for guiding the hospital’s antibiotic stewardship committee in optimizing prophylaxis and empiric therapy protocols. Furthermore, this information supports the development of targeted SSI prevention strategies Therefore, this study aimed to [1] estimate the incidence of surgical site infections [2], describe the microbiological profile of SSI pathogens, and [3] assess antimicrobial resistance patterns to support clinical decision-making and infection prevention efforts.

Methodology

Study design, setting, and duration

This retrospective cohort study was conducted at Aga Khan University Hospital (AKUH), a 750-bed tertiary care teaching hospital in Karachi, Pakistan, providing a broad range of surgical specialties, including neurosurgery, colorectal, plastic, otolaryngology, general, cardiac, orthopedic, gynecological, and vascular surgery. The hospital participates in the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP), a validated clinical registry that provides risk-adjusted outcome measures, including observed-to-expected ratios, derived using established logistic regression models, where 1.0 represents expected performance, values > 1 indicate excess SSIs, and values < 1 indicate fewer SSIs than expected [12, 13]. In this study, we utilized NSQIP-reported data for surveillance and descriptive analysis; however, no additional risk adjustment or independent statistical modeling was performed.

Data were obtained from the ACS NSQIP database and electronic medical records for the period January 2020 to December 2023. Data extraction and analysis were performed between January to December 2024. Patients underwent procedures across nine surgical specialties (neurosurgery, colorectal, plastic, otolaryngology, general surgery, cardiac, orthopedic, gynecology, vascular) at tertiary care hospital, Karachi, from January 2020–December 2023 were included. Analysis focused on specialties with NSQIP-calculated SSI odds ratios > 1.50, indicating > 50% elevated risk. The restriction was applied to focus the microbiological and antimicrobial resistance analysis on specialties with a higher surgical site infections.

Surgical procedure classification

Surgical procedures were grouped by specialty according to the operative site and nature of the procedure to allow meaningful comparison of surgical site infection (SSI) rates. General surgery included a range of gastrointestinal, hepatobiliary, and abdominal wall operations such as bowel resections, cholecystectomy, and hernia repairs. Colorectal procedures mainly comprised colectomies and proctectomies. Gynecological surgeries included hysterectomy, oophorectomy, and other pelvic procedures. Orthopedic surgeries covered fracture fixation, joint replacements (including hip and knee arthroplasties), and spinal procedures, including those involving implants. Neurosurgical procedures included craniotomies, tumor resections, and spinal decompression surgeries. Otolaryngology procedures consisted primarily of head and neck tumor resections and thyroid surgeries. Plastic surgery included reconstructive procedures, particularly flap-based surgeries. Cardiac surgery included coronary artery bypass grafting and valve replacement procedures. Vascular procedures included operations on peripheral veins and arteries, such as ligation and stripping of the long (greater) saphenous vein.

Study population and case selection

A total of 37,768 surgical procedures were performed during the study period, among which 557 SSIs were identified. Among these, 247 patients had microbiological cultures performed and were included in the microbiological analysis. Of these, 148 cases were culture-positive, yielding 208 microbial isolates for detailed microbiological and antimicrobial resistance evaluation. This focused approach was used to enable targeted assessment of microbiological patterns and resistance profiles in specialties with a higher SSI burden, where findings would have greater clinical relevance.

Microbiological methods

Wound cultures, were obtained at the discretion of the treating clinical team based on clinical suspicion including purulent discharge, wound breakdown, or systemic signs of infection. Blood culture data were extracted from clinical records where available; however, correlation with wound isolates or timing of sampling was not evaluated as part of the study. In routine practice, culture collection is encouraged prior to initiation or escalation of antibiotic therapy when feasible; however, some samples were obtained after antibiotic exposure.

Antimicrobial susceptibility testing was performed using automated systems Vitek 2 and/or standard disk diffusion methods, with interpretation according to Clinical and Laboratory Standards Institute (CLSI) guidelines applicable during the study period. Antimicrobial resistance proportions were calculated using organism-specific denominators, defined as the number of isolates tested against each antibiotic. Resistance rates are therefore presented as the proportion of non-susceptible isolates for a given organism–antibiotic combination, rather than as a proportion of total isolates.

Key resistance phenotypes were defined using standard criteria. Methicillin-resistant Staphylococcus aureus (MRSA) was identified based on resistance to cefoxitin or oxacillin. Extended-spectrum beta-lactamase (ESBL) production and carbapenem resistance among Gram-negative organisms were determined based on phenotypic susceptibility patterns in accordance with CLSI guidelines. Multidrug resistance was defined as non-susceptibility to at least one agent in three or more antimicrobial classes. A culture was considered positive if clinically significant organisms were isolated in the presence of compatible clinical features. In polymicrobial infections, all relevant isolates were included. Duplicate isolates from the same patient were excluded, retaining only the first isolate per organism per infection episode. Recurrent infections were considered separate only if documented as new clinical episodes after resolution of the initial infection.

SSI definition and surveillance

SSIs were defined according to standardized ACS NSQIP criteria, including superficial incisional, deep incisional, and organ/space infections, based on clinical findings, imaging, microbiological results, and physician documentation.

SSI surveillance was carried out in accordance with ACS NSQIP methodology. Case identification was based on a combination of inpatient monitoring and post-discharge surveillance. Post-discharge follow-up included review of outpatient clinic visits, emergency department presentations, hospital readmissions, reoperations, and microbiology reports within the postoperative period. Treating physician documentation and infection control records were also reviewed to ensure comprehensive case identification. Patients were followed for up to 30 days after surgery as per ACS NSQIP protocol.

Inclusion and exclusion criteria

Inclusion criteria

Patients with SSIs meeting ACS NSQIP definitions from specialties with an OR > 1.50.

Exclusion criteria

Cases excluded according to standard ACS NSQIP criteria [13].

Data analysis

Descriptive statistical analysis was performed. Continuous variables are presented as mean (SD) or median with interquartile range (IQR), as appropriate. Categorical variables are summarized as frequencies and percentages. As the study was primarily descriptive, no inferential statistical analysis or multivariable modeling was performed. As the study followed patients for a fixed 30-day postoperative period, cumulative incidence was used as the primary measure of SSI occurrence, rather than incidence density. SSI cumulative incidence was calculated as the proportion of surgical procedures complicated by surgical site infection within the 30-day postoperative follow-up period, in accordance with ACS NSQIP methodology.

Results

Surgical site infection incidence

A total of 37,768 surgical procedures were performed across nine specialties during the study period. Overall, 557 surgical site infections (SSIs) were identified, corresponding to an incidence of 1.5%.

SSI rates varied by procedure type, with the highest incidence observed in colorectal procedures (%), followed by neurosurgical (3.1%) and otolaryngology procedures (2.7%). Lower rates were seen in cardiac (0.4%), orthopedic (0.4%), and vascular procedures (0.1%) (Table 1).

Table 1.

SSI incidence rate of nine major specialties

S.no.	Specialties		Total surgeries	SSI cases	SSI Incidence (%)	95% CI
1	Colorectal	709		42	6	(4.3, 7.9)
2	Neurosurgery	3244		102	3.1	(2.5, 3.8)
3	Otolaryngology	6042		160	2.7	(2.2, 3.0)
4	General Surgery	7839		138	1.8	(1.4, 2.0)
5	Gynecology	5867		62	1.1	(0.8, 1.3)
6	Plastic	2103		12	0.6	(0.2, 0.9)
7	Cardiac	2314		10	0.4	(0.2, 0.7)
8	Orthopedic	7171		29	0.4	(0.2, 0.5)
9	Vascular	2479		2	0.1	(0.0, 0.2)
	Total	37,768		557	1.5	(1.3, 1.6)

Patient demographics and clinical characteristics

Among the 557 SSI cases, microbiological cultures were obtained in 247 patients. Within this subgroup, males accounted for 57.5%, and the median age was 50 years (IQR 39–60). Most procedures were elective (85.4%), and the majority of patients were functionally independent (87.0%). Common comorbidities included hypertension (39.7%) and diabetes mellitus (9.3%).Blood cultures, when obtained based on clinical indication, were positive in 59.9% of cases; these data were recorded as part of routine clinical documentation, and no attempt was made to correlate blood and wound culture isolates or their timing. Among wound cultures, 71.0% showed monomicrobial growth (Table 2).

Table 2.

Baseline demographics, clinical, and Laboratory characteristics of SSI Patient

	Total Patients = 247
	N (%)
Sex of Patient
Male	142 (57.5)
Female	105 (42.5)
Age Categories (Years)
18–29	19 (7.7)
30–39	45 (18.2)
40–49	50 (20.2)
50–59	66 (26.7)
60–69	47 (19.0)
70+	20 (8.1)
Age (Years)
Mean(SD)	49.94 ± 14.31
Median (IQR)	50.00 (39.00–60.00)
Specialties
Otolaryngology	61 (24.7)
General surgery	58 (23.5)
Neurosurgery	51 (20.6)
Gynecology	28 (11.3)
Colorectal	17 (6.9)
Orthopedic	16 (6.5)
Plastic	8 (3.2)
Cardiac	7 (2.8)
Vascular	1 (0.4)
Case Acuity
Elective	211 (85.4)
Emergency	4 (1.6)
Emergent	32 (13.0)
Sepsis (SIRS/Sepsis/Septic shock) (48 h)
SIRS	23 (9.3)
Sepsis	5 (2.0)
Diabetes Mellitus
Insulin-Dependent	23 (9.3)
No Diabetes	178 (72.1)
Non-Insulin-Dependent	46 (18.6)
Hypertension requiring medication
Yes	98 (39.7)
Current Smoker within 1 year
Yes	23 (9.3)
Blood Culture Result
Positive	148 (59.9)
Negative	99 (40.1)
Pus specimen
Number of Organisms in a wound Specimen
1	105 (70.9)
2	30 (20.3)
3	9 (6.1)
4	4 (2.7)
Type of microbial growth based on number of organisms
Monomicrobial	105 (71.0)
Polymicrobial	43 (29.0)

Microbiological spectrum

Of the 247 patients with cultures, 148 cases were culture-positive, yielding a total of 208 microbial isolates, indicating the presence of polymicrobial infections in some cases.

Gram-negative organisms accounted for 52.9% (n = 110) of isolates, followed by Gram-positive organisms (41%; n = 85) and fungal pathogens (3.8%; n = 8) (Fig. 1). Among Gram-negative organisms, Escherichia coli was the most frequent (61% of Gram-negative isolates; n = 67), followed by Klebsiella pneumoniae (24%; n = 26) and Pseudomonas aeruginosa. Among Gram-positive organisms, Staphylococcus aureus predominated (52%; n = 42), with Staphylococcus species and Enterococcus species also identified. Less commonly isolated organisms included non-fermenting Gram-negative bacilli, Corynebacterium species, and Candida species. The distribution of pathogens varied across procedure types. Escherichia coli was most frequent in colorectal procedures, whereas Pseudomonas aeruginosa and Staphylococcus aureus were more commonly observed across multiple surgical categories. A detailed breakdown of all microbial isolates across surgical procedures is provided in Supplementary Table S1, S2.

Fig. 1.

Distribution of microorganisms isolated from surgical site infections (n = 208)

Resistance profiles of SSI isolates

Resistance rates are presented using organism-specific denominators, reflecting the number of isolates tested for each antibiotic. Antimicrobial resistance rates were calculated based on the number of isolates tested for each antibiotic, as indicated by the denominator (N). High resistance was observed to commonly used antibiotics, including; amoxicillin–clavulanate 63.0%(35/56 isolates tested), co-trimoxazole 45.5% (51/112), ciprofloxacin: 40.8% (42/103). Moderate resistance was seen with: piperacillin–tazobactam: 19.3% (21/109), gentamicin 15.4% (21/136), and imipenem 12.5% (13/104). While, the lowest resistance rate was observed with amikacin 6.5% (7/108).

Discussion

This study provides an overview of surgical site infection (SSI) incidence, microbiological profile, and antimicrobial resistance patterns using standardized NSQIP-based surveillance in a tertiary care setting. The overall SSI incidence of 1.5% appears comparable to rates reported in high-income settings; however, this should be interpreted cautiously given potential differences in case mix, procedure types, and surveillance methodologies. Notably, previous surveillance studies at this tertiary care hospital reported SSI rate around 6.8-7% [14].

SSI occurrence varied notably across procedure types. Higher rates were observed in procedures involving the gastrointestinal tract, particularly colorectal surgeries, while lower rates were seen in predominantly clean procedures such as vascular and cardiac surgeries. This pattern is consistent with established evidence that procedure-related contamination risk is a major determinant of SSI. The cohort also included procedures involving prosthetic materials, such as orthopedic joint arthroplasties and abdominal wall repairs with mesh. As surveillance was limited to 30 days in accordance with NSQIP methodology, infections occurring beyond this period—particularly those related to implants—may not have been fully captured [15–18].

Microbiologically, Gram-negative organisms predominated, followed by Gram-positive cocci. Escherichia coli and Klebsiella pneumoniae were the most commonly identified Gram-negative pathogens, while Staphylococcus aureus was the leading Gram-positive organism. This distribution is in line with reports from similar settings, where Gram-negative pathogens are frequently associated with SSIs, especially in abdominal procedures. Polymicrobial infections were also observed, reflecting the complexity of SSI etiology.

The antimicrobial resistance patterns observed have important clinical implications. High resistance to commonly used agents such as amoxicillin–clavulanic acid limits their reliability for empiric therapy. In contrast, lower resistance rates to amikacin suggest retained activity in this setting. Carbapenem resistance, although less frequent, remains a concern and highlights the importance of preserving these agents. Susceptibility to last-line antibiotics such as colistin and tigecycline was largely maintained, consistent with their role as reserve therapies. Overall, these findings underscore the importance of using local microbiological data to guide empiric treatment and prophylaxis strategies [19].

The demographic and clinical characteristics of patients with SSI were broadly consistent with known risk factors. Comorbidities such as diabetes and hypertension were common and are well recognized to increase susceptibility to infection through impaired host defenses. These findings align with previously reported SSI risk profiles [16, 20].

Rather than attributing the observed SSI rates to specific institutional practices, which were not directly evaluated, this study highlights the role of structured surveillance in identifying procedure-specific risks and microbiological trends. Such data may help inform targeted infection prevention efforts and support antimicrobial stewardship initiatives tailored to local epidemiology [20].

Strengths and limitations

This study has several strengths. It utilizes standardized ACS NSQIP data, allowing consistent and validated identification of surgical site infections. The inclusion of multiple surgical specialties provides a broader perspective, and the integration of microbiological data with clinical outcomes adds practical relevance.

However, limitations should be considered. The retrospective single-center design may limit generalizability. Microbiological analysis was restricted to cases where cultures were obtained at clinician discretion, which may underestimate pathogen diversity. The focus on higher-risk procedures and culture-positive cases may introduce selection bias. In addition, follow-up was limited to 30 days in accordance with NSQIP protocol, and late-onset infections, particularly those associated with implants, may not have been captured. Incidence density was not calculated, which may be considered in future studies with more detailed time-to-event data. Finally, no multivariable analysis was performed to evaluate risk factors or clinical outcomes.

Future directions

Future studies should focus on multicenter surveillance to improve generalizability and explore procedure-specific prevention strategies, particularly in higher-risk procedures. Prospective evaluation of antimicrobial prophylaxis practices and regular updates of institutional antibiogram may further support optimized infection control and antimicrobial stewardship.

Conclusion

This study highlights the role of structured surgical site infection surveillance using risk-adjusted data in describing SSI burden and supporting infection prevention efforts in resource-limited healthcare settings. Procedure-specific analysis demonstrated variability in SSI occurrence across surgical specialties, underscoring the importance of targeted prevention strategies. The findings also emphasize the need for ongoing microbiological surveillance and the use of local antimicrobial resistance data to inform empiric therapy and support antimicrobial stewardship. As microbiological analysis was restricted to procedures with higher NSQIP risk-adjusted odds ratios (> 1.50), the results may not be generalizable to lower-risk procedures, and some degree of selection bias should be considered. Future studies should focus on prospective multicenter surveillance and evaluation of targeted interventions to further reduce SSI burden.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

SSI

Surgical site infection

HAI

Healthcare-associated infections

ACS NSQIP

American College of Surgeons National Surgical Quality Improvement Program

IRB

Institutional Review Board

WHO

World Health organization

GNR

Gram negative rods

GPC

Gram positive cocci

SD

Standard deviation

Author contributions

Conceptualization SI. Laboratory data collection: RA. data analysis: MN. Data interpretation: RA. Manuscript drafting: RA, SI. Revision and reviewed manuscript: RA, SI, HZ, RR, NV, FM. All authors read and approved the final manuscript.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethical approval

Ethical approval (2025-7233-35202) was obtained from the Aga Khan University Ethics Review Board (AKU-ERB), before data extraction. Patient confidentiality was ensured by de-identifying all records, and data stored securely with restricted access. Data security was assured by access to only principal investigator. All methods were performed in accordance with the relevant guidelines and regulations including the Declaration of Helsinki.

Waiver of informed consent

The AKU-ERB waived the requirement for informed consent due to the retrospective nature of the study using de-identified data from hospital electronic medical records and NSQIP database.

Consent to participate

NA.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.