Reducing surgical site infections in paediatric cardiac surgery in a resource-constrained setting: a quality improvement initiative

Abstract

Problem

A newly established congenital cardiac surgery programme in a low-resource setting identified a high early surgical site infection (SSI) rate of 23% during the first 3 months, reflecting variation associated with small case volumes and evolving systems.

Background

Global paediatric cardiac SSI rates range from 0.25% to 6%, but can exceed 20% in low and middle-income countries due to workflow variability, resource constraints and patient-related risk factors. Most SSI prevention bundles originate from high-income settings and require contextual adaptation. Quality improvement (QI) methods can strengthen reliability when tailored to local systems.

Methods

A learning health system approach was used, combining process mapping, root cause analysis, a key driver diagram and four sequential Plan Do Study Act cycles. Interventions targeted reliability across the perioperative pathway. The primary outcome was SSI rate per 100 surgeries. Process measures included training completion, preoperative risk screening, wound documentation and structured review of moderate and severe infections. Balancing measures included length of stay, mortality rate, case cancellation rate and readmissions. Interventions included strengthened sterilisation and risk screening practices, a nurse-led wound clinic with 30-day follow-up and training in aseptic technique and hand hygiene.

Results

Among 500 surgeries, 31 SSIs occurred. SSI rates decreased from 12 to 6 per 100 surgeries and reached 2 per 100 surgeries, before increasing to 6 then 5 per 100 surgeries in later phases. Process reliability improved across all measures, while balancing measures remained stable. Following sustainability actions, environmental cleaning compliance improved from 20% to 85% and hand hygiene from 45% to 90%.

Conclusion

QI interventions were associated with reduced SSI rates, although sustained benchmark performance was not achieved. Continued monitoring and behaviour-focused strategies will guide further improvement.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Surgical site infections are a major source of morbidity in paediatric cardiac surgery and occur at higher rates in low-income and middle-income countries. Most prevention strategies are derived from high-income settings and require adaptation to local workflows, resources and staffing. Quality improvement methods can reduce infection risk when tailored to context.

WHAT THIS STUDY ADDS

• This study demonstrates that a perioperative pathway approach, combining measurement systems, sequential improvement cycles, nurse-led wound care and structured tele-follow up, can reduce surgical site infection (SSI) rates in a new congenital cardiac programme. It shows improvement in key processes and highlights the role of internal audit and continuous monitoring in supporting reliability, even when gains are not sustained.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• This study provides a practical, low-cost model for SSI reduction in resource-limited congenital cardiac programmes. It reinforces the importance of continuous measurement, feedback and system-level interventions to support sustainable improvements in perioperative care.

Introduction

Surgical site infections (SSIs) in paediatric cardiac surgery contribute significantly to postoperative morbidity and mortality. Reported global rates range from 0.25 % to 6 %, but may reach up to 23% in low and middle-income countries (LMIC) due to contextual challenges such as inconsistent adherence to wound care practices, limited infection prevention resources and patient-related factors including malnutrition and cyanotic physiology.1,3 These risks have been well documented across studies from both high-income and low-income settings and are consistently associated with prolonged hospital stay, increased complications and higher mortality.⁴

Most SSI prevention strategies are developed in high-income settings and require adaptation before they can be implemented effectively in resource-constrained environments. A systematic review from LMICs shows that infection risk and cost burdens are higher when standard protocols are difficult to implement due to operational and resource limitations.⁵ Evidence-based quality improvement (QI) approaches such as structured process mapping, strengthened perioperative aseptic technique and improved postoperative wound care have shown benefit when tailored to local conditions.^{6 7} Recent paediatric cardiac surgery improvement work also demonstrates that multidisciplinary engagement, standardised care pathways and reliable wound surveillance can reduce infection rates when adapted to the realities of low-resource programmes.^{4 6}

Shortly after our congenital cardiac surgery programme began operating, early surveillance revealed an unexpectedly high SSI rate, indicating system-level gaps rather than isolated events. This prompted the development of a structured improvement initiative using principles of a Learning Health System (LHS), where continuous measurement, team-based review and iterative adaptation guide progressive strengthening of care processes.⁸ This report describes our QI journey to reduce SSIs in a resource-limited LMIC surgical programme through sequential interventions, contextual learning and progressive redesign of infection prevention practices.

Methods

Study setting and context

This QI initiative was conducted in a newly established public-sector congenital cardiac surgery programme in South Asia. The programme was initiated in 2022 and performed its first open-heart surgery in 2023. Between April 2023 and August 2025, 500 paediatric and congenital cardiac surgeries were performed, including 402 open-heart and 98 closed-heart procedures. The median patient age was 1.7 years and median weight was 12 kg.

The case mix included a broad spectrum of congenital heart disease (CHD), most commonly Tetralogy of Fallot (n=148), ventricular septal defect (n=121) and atrial septal defect (n=67), with smaller numbers of complex lesions. Surgical risk, using the Risk Adjustment for Congenital Heart Surgery (RACHS-1) system, was predominantly low to moderate, with most cases in categories 1–2 and fewer in categories 3–4; 2.8% were unassigned.

Care delivery followed a CHD continuum model, mapping the patient journey from diagnosis through intervention and follow-up to support consistent outcome tracking (online supplemental figure S1). The team included cardiac surgeons, paediatric cardiologists, anaesthetists, intensivists, perfusionists, operating room technicians and nursing staff with varying experience levels.

For benchmarking, the programme participated in the International Quality Improvement Collaborative (IQIC), which provided risk-adjusted data for congenital cardiac surgery programmes in LMICs.⁹ Internal dashboards were also used to track performance and identify gaps in infection prevention.

Study population and inclusion criteria

All consecutive patients undergoing paediatric and congenital cardiac surgery between April 2023 and August 2025 were included. Both open-heart and closed-heart procedures were considered, and no exclusion criteria were applied. SSIs were identified through in-hospital surveillance and 30-day postoperative follow-up. Infections were classified as Grade 1 (superficial), Grade 2 (deep incisional) or Grade 3 (organ or space) based on standard definitions, with classification assigned by the treating surgical team.

Study design and quality improvement approach

The project followed a structured improvement pathway grounded on LHS principles, emphasising continuous measurement, iterative testing and adaptation. Standard QI tools were applied alongside contextual assessment to guide practice changes. The overall process is summarised in online supplemental table S1 and figure 1.

Figure 1. Timeline of the SSI quality improvement (QI) initiative showing key programme milestones and sequential QI interventions from June 2023 to August 2025. COM-B, Capability–Opportunity–Motivation–Behavior; CTS, Cardiothoracic sciences; KDD, key driver diagram; PDSA, Plan Do Study Act; QI, quality improvement; RCA, root cause analysis; SSI, surgical site infection.

Theory of change

This QI initiative was guided by a context-specific theory of change, which defined the causal pathway linking interventions to outcomes across the congenital heart disease care continuum. The model assumed that improving surgical outcomes required coordinated changes in perioperative processes, supported by continuous measurement, feedback and adaptation within the local context. Perioperative infection prevention was addressed as a key component of the treatment phase (online supplemental figure S1). This approach aligns with theory of change frameworks for contextual clinical research in LMICs, which emphasise linking interventions to outcomes through defined pathways, supported by data systems, stakeholder engagement and iterative learning.¹⁰

Problem identification and team formation

Three months after the programme initiation, early surveillance identified an SSI rate of 23% (3 infections among 13 surgical surgeries). This was substantially higher than the IQIC benchmark of approximately 2.2% reported for comparable LMIC programmes. Given the small sample size and expected and early variability, this period was not considered a stable baseline but an early signal of system level gaps. The cohort consisted predominantly open-heart procedures and higher-risk patients, including those with low weight, cyanotic heart disease and malnutrition.

Contributing factors included both patient-related and system-level issues. Patient-related risks included nutritional compromise and case complexity. System-level gaps included variability in aseptic technique, inconsistent sterilisation practices, absence of standardised postoperative wound surveillance, gaps in documentation and lack of structured discharge counselling.

A multidisciplinary SSI improvement team was established including representatives from surgery, anaesthesia, intensive care, nursing, perfusion, infection control, patient safety and quality, and the data registry unit.¹¹ The team met weekly to review cases, identify workflow gaps and plan interventions.

Diagnostic assessment (process mapping and root cause analysis)

A detailed process map was developed to document each step from preoperative admission through postoperative discharge and follow-up. This was based on direct observations, review of standard operating procedures, case reviews and staff feedback (online supplemental figure S2).

At baseline, infection prevention practices were inconsistently implemented. Hand hygiene compliance was approximately 45%, and environmental cleaning adherence was around 20%. Discharge counselling and wound care documentation were variable, and no structured postoperative wound surveillance system existed. Follow-up relied primarily on patient-initiated visits, and early signs of infection were not consistently recognised or escalated.

Based on these findings, the postoperative pathway was redesigned to introduce standardised discharge counselling, structured wound assessment and early surveillance. Caregivers received uniform instructions on wound care and recognition of infection. A nurse-led surveillance system was initiated from the first day after discharge, with caregivers sharing wound images through a secure messaging platform. Suspected infections prompted early clinical evaluation and referral to a dedicated wound care service when required. Patients were stratified by clinical severity, with defined pathways for escalation to inpatient care or continued outpatient monitoring. Surveillance was maintained for 30 days postoperatively, with discharge on complete wound healing.

Root cause analysis was then performed using Fishbone and Five Whys methods (online supplemental figure S3). Contributing factors were categorised into human factors, materials and equipment, environmental conditions, procedural gaps and patient-related risks. These included inconsistent hand hygiene, variation in aseptic technique, reuse of selected consumables, gaps in documentation, limited caregiver understanding of wound care and patient factors such as malnutrition or delayed wound healing. These findings informed the design of subsequent intervention cycles.

Intervention design key driver diagram (KDD)

Insights from the diagnostic assessment were consolidated into a key driver diagram (figure 2), which defined the aim of reducing SSIs and identified the primary drivers of infection outcomes. These included staff competency in aseptic technique, reliability of sterilisation processes, postoperative wound-care quality, structured preoperative risk identification and ongoing monitoring through audits and feedback.

Figure 2. Key driver diagram. MDT, multidisciplinary meeting; RCA, root cause analysis; SMART, Specific–Measurable–Achievable–Relevant–Time-bound; SSI, surgical site infection.

Change ideas were developed for each driver, including strengthening aseptic training, improving sterilisation reliability, standardising wound care and follow-up, introducing preoperative risk assessment and embedding routine performance monitoring. These informed the design of sequential Plan Do Study Act (PDSA) cycles described below.

Outcome, process and balancing measures were defined prior to implementation. The primary outcome was SSI rate per 100 surgeries. Process measures included completion of aseptic competency assessments, documentation of preoperative risk screening, adherence to wound-care assessment and documentation standards, and completion of root cause analyses for all moderate or severe SSIs (online supplemental table S2). Balancing measures included length of stay, mortality, case cancellations and readmissions per 100 surgeries. Mortality was monitored to detect any adverse impact on overall patient outcomes, while length of stay was assessed to identify potential changes in discharge practices or care efficiency.

Sequential PDSA cycles

The improvement strategy was implemented through four sequential PDSA cycles, each addressing specific gaps identified during the diagnostic phase and informing subsequent interventions (online supplemental figure S4).

PDSA cycle 1— establishing measurement and basic reliability

The first cycle focused on establishing a reliable system for structured measurement and feedback. The QI team developed initial process metrics by consensus (online supplemental table S2), derived from infection logs, postoperative documentation reviews, bedside observations and workflow assessments. Standard operating procedures and documentation tools for infection prevention were developed and refined through iterative testing. Monthly tracking templates were introduced, with assigned staff for each metric. Run charts and dashboards were reviewed during weekly multidisciplinary meetings (MDT), identifying early gaps such as inconsistent documentation and inter shift variability. Documentation tools were simplified, and daily updates were introduced. Process performance review became a routine component of MDT meetings, establishing the measurement platform for subsequent interventions.

PDSA cycle 2—strengthening sterilisation practices and introducing preoperative risk screening

Following Cycle 1, SSI rates decreased to 0% by October 2023 but increased to 10% by February 2024 (2 infections among 20 surgeries), prompting focused review of perioperative infection control.

An internal audit identified inconsistent sterilisation practices, including reuse of arterial and venous cannulas during cardiopulmonary bypass. In collaboration with a high-volume congenital cardiac surgical centre, perioperative practices were standardised in partnership with the surgical team. Key measures included switching to single-use sterile consumables (including cardiopulmonary bypass cannulas and related disposables), strengthening preoperative skin preparation bundles, implementing nasal decolonisation when indicated, promoting early re-exploration for suspected wound complications and standardising postoperative wound-care pathways, including dressing practices and escalation protocols.

Procurement pathways were formalised, and compliance was monitored through operating room observations and perfusion supply logs. A structured SSI risk screening tool (online supplemental table S3) was introduced and integrated into preoperative planning. This was used during MDT discussions to identify patients at higher risk of infection, particularly those with malnutrition or higher surgical complexity, enabling targeted optimisation and closer preoperative monitoring. Case cancellations were monitored in relation to this tool to ensure that risk identification did not adversely affect surgical access. Postoperative care was stratified by SSI severity, with Grade 1 infections managed conservatively and Grades 2 and 3 requiring surgical review and escalation. Monthly SSI trends were reviewed, and adherence to new practices reinforced in MDT meetings.

PDSA cycle 3—strengthening postoperative wound care through a nurse-led wound clinic

The third cycle focused on improving postoperative wound care and earlier detection of SSIs, particularly those presenting after discharge. Case reviews indicated delayed recognition due to limited caregiver awareness and absence of structured follow-up. A clinical nurse-led wound clinic was established, integrated with cardiology and surgical follow-up. Standardised wound assessment and documentation protocols were introduced, and caregivers received structured predischarge education with demonstrations and written instructions.

A 30-day post-discharge follow-up system was implemented using a secure messaging platform, allowing caregivers to submit daily wound photographs and symptom updates. These were reviewed by the nursing team and escalated to the surgical or cardiology team when required. Monitoring included SSI incidence, severity and timing of detection, with data reviewed weekly.

PDSA cycle 4—reinforcing aseptic technique and hand hygiene

A subsequent rise in infections prompted a fourth cycle focused on strengthening aseptic technique and hand hygiene practices. Training was delivered across all shifts through bedside demonstrations and video-based refreshers aligned with WHO guidance.¹² A designated infection prevention nurse coordinated scheduling and ensured coverage for all shifts. Daily observational audits were carried out by trained nonclinical observers drawn from within the institution, without additional staffing. Staff requiring improvement received targeted bedside coaching.

To extend the impact of this cycle, aseptic technique and hand hygiene training were incorporated into the structured orientation programme for all new employees.

Sustainability actions

To support long-term reliability, an independent internal audit was conducted by the institutional infection control and quality teams to assess environmental cleaning, hand hygiene, sterilisation practices, visitor flow and preoperative preparation. The audit identified key gaps, including suboptimal hand hygiene compliance (overall ~56%, with 41% compliance before patient contact), inconsistent environmental cleaning and deficiencies in the availability and placement of hand hygiene resources (online supplemental table S4).

Process reliability gaps were identified, including discrepancies between documented checklist completion and actual task execution, reflecting weaknesses in adherence and monitoring. Variation in preoperative preparation and postoperative wound care was noted, including lack of supervised preoperative skin preparation and delayed recognition of wound complications.

Based on these findings, targeted sustainability interventions were implemented. Shift-based infection prevention champions were identified from existing nursing and operating room staff. They were tasked with reinforcing daily practices, supporting adherence during busy periods and escalating concerns when needed. Daily and weekly huddles during shift handovers were used to communicate priorities and address identified gaps. Infection prevention checklists were integrated into routine workflows, supported by visual reminders at key points of care.

To address supply-related lapses, unit level reserves of essential materials, including chlorhexidine, sterile dressings, gloves and hand hygiene products, were established and monitored. Ongoing reinforcement included monthly competency-based training, structured induction for new staff and regular dashboard reviews to guide feedback and coaching, in line with evidence that audit and feedback cycles improve reliability of infection prevention practices.^{13 14}

Ethical considerations and reporting standards

This project was approved by the Institutional Review Board and the institutional Patient Safety and Quality Unit as a QI initiative. Individual patient consent was not required, as the study involved analysis of routinely collected data and system level improvement activities. The manuscript was prepared in accordance with the Standards for QI reporting excellence (SQUIRE 2.0) guidelines.

Data analysis

SSI cases were recorded prospectively using a standardised template capturing patient characteristics, procedural details, infection grade, timing and management. The primary outcome was the SSI rate per 100 surgeries, calculated across five sequential time periods aligned with major intervention phases (April 2023 to August 2025). Primary outcome was SSI rate per 100 surgeries. For temporal analysis, monthly SSI counts were displayed using a statistical process control c-chart to assess variation over time (online supplemental figure S5).

Process measures were derived from monthly audit logs, direct observations and documentation reviews and were displayed as time series graphs in online supplemental figure S6.

Balancing measures were obtained from the programme’s routine surgical dashboard. These were reviewed in MDT meetings to monitor trends and identify potential unintended effects.

Results

From April 2023 to August 2025, the programme performed 500 paediatric and congenital cardiac surgeries and recorded 31 surgical site infections, giving an overall rate of 6.2 per 100 surgeries.

SSI rates were 12 per 100 surgeries during the baseline period (April 2023 to January 2024). Following the first two PDSA cycles implemented between August 2023 and March 2024, rates decreased to 6 per 100 surgeries (February to June 2024). After the fourth PDSA cycle in May 2024, rates further declined to 2 per 100 surgeries (July to November 2024).

A subsequent increase to 6 per 100 surgeries (December 2024 to April 2025) suggested reduced process reliability, prompting additional sustainability and reinforcement actions. Rates thereafter stabilised at 5 per 100 surgeries (May to August 2025). Monthly variation and temporal association with interventions are shown in figure 3 and online supplemental figure S5.

Figure 3. SSI rate per 100 surgeries across intervention phases. The blue line represents the surgical site infection (SSI) rate per 100 surgeries over sequential time periods. The orange dashed line indicates the overall mean SSI rate (6 per 100 surgeries). The grey dashed line represents the upper reference limit (UCL=17), and the yellow dashed line represents the lower reference limit (LCL = −5). Vertical arrows mark key phases of the quality improvement initiative, including problem identification, sequential PDSA cycles and sustainability actions. These annotations illustrate the temporal relationship between interventions and changes in SSI rates. PDSA, Plan Do Study Act.

Of the 31 infections, 17 were Grade 1, 6 were Grade 2 and 8 were Grade 3 infections. Two patients with Grade 3 SSI developed bacteraemia with sepsis, progressed to multi-organ dysfunction and died. Monthly severity patterns and case-level details are provided in online supplemental table S5.

Process reliability improved across all measured domains (online supplemental figure S6): training completion, increased from 40% at baseline to 80% by early 2024 and reached over 90% by mid-2025. Use of the preoperative SSI risk assessment tool increased from no uptake at baseline to full adoption by early 2024, with sustained use thereafter.

Wound care documentation improved from 0% at baseline to 100% adherence by early 2024 and was sustained throughout the study period. Completion of structured quality control reviews for moderate and severe infections increased from 0% at baseline to 50% by mid-2024 and reached complete adherence by mid-2025.

Balancing measures showed no evidence of unintended negative effects (online supplemental figure S7). Length of stay remained stable at 3–4 days throughout the study period. Mortality ranged from 3 to 16 per 100 surgeries and appeared to reflect case mix rather than the impact of QI interventions.

Case cancellation rates varied between 4 and 25 per 100 surgeries, with no sustained increase following implementation of infection prevention measures. Readmission rates ranged from 9 to 20 per 100 surgeries, with no temporal clustering associated with discharge practices or wound care practices.

Sustainability actions

Following the internal audit, targeted actions were implemented to address identified gaps. These included restricting entry to one parent, strengthening infection prevention practices, initiating nasal mupirocin for staff with Staphylococcus aureus colonisation and introducing chlorhexidine bathing as part of preoperative preparation.

Environmental cleaning compliance increased from 20% to 85%, and hand hygiene compliance improved from 45% to 90% (online supplemental table S5).

Discussion

SSI remains a major challenge in paediatric and congenital cardiac surgery, particularly in low-resource environments where workflow variability, supply constraints and staffing limitations affect adherence to infection prevention standards.¹⁵ At baseline, SSI rates at our centre were substantially higher than international benchmarks. Using an iterative QI approach,¹⁶ we implemented sequential interventions across the perioperative pathway and observed a reduction in SSI rates over time, although a statistically significant centreline shift was not demonstrated.

Continuous monitoring through IQIC and local dashboards enabled early identification of trends, highlighting the importance of real-time feedback in QI frameworks for congenital heart care in resource-limited environments.¹⁷

Our findings support existing evidence that SSI reduction in such settings requires coordinated system-level changes rather than isolated interventions. Process mapping of the perioperative pathway identified key areas of risk, including inconsistent sterilisation practices, variation in aseptic technique and gaps in postoperative wound monitoring. Similar studies from LMIC programmes have shown that pathway mapping can uncover workflow inefficiencies not detected through routine audit alone.¹⁸

A key intervention was introduction of a nurse-led wound care service integrated with cardiology and surgical follow-up. Caregivers submitted wound images through a structured messaging system, enabling timely review and escalation by the clinical team. This approach improved consistency in wound assessment and facilitated earlier recognition of complications using a low-cost telehealth model, consistent with prior reports demonstrating improved continuity and timely escalation.^{19 20}

Adaptation to externally informed practices adapted from a high-volume regional also contributed to early improvements. These included single-use arterial and venous cannulas, improved preoperative preparation and timely consideration for re-exploration when wound problems were suspected. Such collaboration provided practical insights into implementing reliable systems in comparable resource-constrained environments.²¹

Our sustainability strategies, including shift-based infection prevention champions, daily and weekly huddles, embedded checklists and visual cues, aimed to improve reliability across shifts and reduce reliance on individual performance. These approaches align with findings from Ethiopia’s Clean-Cut programme, where routine audit, feedback and team engagement supported sustained improvement despite low resources.²²

Despite these efforts, sustaining improvements remained challenging. Internal audit findings indicated that lapses in environmental cleaning, hand hygiene and compliance, preoperative bathing documentation and sterilisation storage practices were primarily due to operational reliability gaps rather than knowledge deficits. This reflects broader evidence that sustainability depends on consistent system, frontline ownership and reliable supply chains rather than protocols alone. Although formal Statistical process control(SPC) criteria for centreline shift were not met, SSI rates decreased from 12 to 6 to 2 infections per 100 surgeries. This trend suggests improvement, but continued monitoring is required to confirm sustained change. Similar patterns have been described in early-stage programmes, where measurable improvement often precedes statistical stability and longer observation is needed to detect special cause variation.⁴

Limitations

This work has several limitations. Baseline data before April 2023 were unavailable, as the programme was newly established and formal data collection systems were implemented after clinical services began. Early observations therefore likely reflect both a true signal and variation related to small case volumes during programme initiation.

The timing of interventions overlapped with the early data collection period, limiting clear separation of pre-intervention baseline from post-intervention effects. Some early audits relied on informal observations and may have introduced reporting bias. Behavioural change was inferred from process compliance rather than direct observation, which limits interpretation of practice variation.

This was a single-centre project with modest case numbers, which restricts generalisability. The preoperative risk assessment tool was evidence-informed but not formally validated. The nurse-led wound care model relied on an experienced nurse, which may challenge scalability in settings with fewer trained staff. The messaging-based follow-up system was cost-effective and acceptable to families, but it raises confidentiality considerations. Free care at our centre may also have improved follow-up adherence compared with settings where financial barriers.

Finally, the internal audit provided important insight into environmental and operational contributors, but was conducted at a single point, repeated audits would better assess reliability over time.

Future directions

As SSI rates have not yet reached benchmark levels, the next phase will focus on behavioural determinants of practice. The Capability–Opportunity–Motivation–Behavior (COM-B) model will be applied to identify barriers affecting staff capability, opportunity and motivation, to guide targeted interventions.²³

Conclusion

Reducing SSI in paediatric cardiac surgery in resource-limited settings requires coordinated system redesign, reliable processes and attention to the human and organisational factors. This study demonstrates that applying structured QI methods can strengthen infection prevention across the perioperative pathway. Although sustained benchmark performance was not achieved, the benchmark level was reached during the improvement period, indicating that the system can perform at the desired level under the appropriate conditions. Subsequent increases in infection rates highlight the need for continuous monitoring, timely feedback and adaptive responses. Framing this work within a LHS approach provides a practical model for similar centres seeking to achieve consistent and sustainable improvements in surgical quality and safety.

Data availability statement

Data are available upon reasonable request.