Perioperative Prophylaxis in Pediatric Solid Organ Transplantation: A Baseline Stewardship Audit Using Routine Electronic Data

ABSTRACT

Background

Perioperative antibacterial prophylaxis in pediatric solid organ transplantation varies by center in regimen, add‐on agents, and stop dates. We used institutional protocols as a unified reference standard to establish baseline stewardship measures and targets for future quality‐improvement work.

Methods

We retrospectively evaluated 164 pediatric transplants from 2010 to 2021. Using anesthesia timing and medication‐administration data, we assessed on‐time dosing, regimen selection, MRSA‐ and VRE‐directed add‐on agents, double anaerobic coverage, Gram‐negative broadening during the protocol window, and discontinuation at organ‐specific stop targets. We also used transplant‐day anchoring and protocol‐defined windows to describe whether primary‐admission broad‐spectrum exposure represented protocol‐linked prophylaxis or later exposure. The audit measured protocol concordance and exposure patterns, not the appropriateness of deviations.

Results

Prophylaxis was given on time in 95.7% of procedures. Overall, 87.2% received the recommended initial regimen, 7.3% the protocol‐defined beta‐lactam allergy regimen, and 4.3% another broad Gram‐negative regimen. Only 61.0% stopped the main regimen by the intended cutoff; stop performance was best in kidney transplantation and poorest in intestinal and multivisceral transplantation. MRSA‐directed add‐ons clustered in intestinal and multivisceral procedures and often extended beyond the intended duration. A secondary descriptive analysis identified contiguous protocol‐linked continuation beyond the intended window in 8.6% of primary‐admission broad‐spectrum days of therapy.

Conclusions

Timing and initial regimen selection were strong. The main baseline process measure requiring prospective review was continuation of protocol‐linked antibiotics beyond cutoffs in abdominal transplantation. This audit shows how routine electronic data can be used to create reproducible measures and organ‐specific targets for future stewardship and quality‐improvement studies.

Using routine electronic data from 164 pediatric transplants, we created reproducible stewardship measures of perioperative prophylaxis practice across transplant populations and protocol pathways. Starting prophylaxis was reliable, but continuation beyond intended stop targets—especially in abdominal transplantation—identified the clearest target for future prospective review and quality‐improvement efforts.

Abbreviations

ARKS

Anesthesia Record Keeping System

CI

confidence interval

DOT

days of therapy

IgE

immunoglobulin E

IQR

interquartile range

LOS

length of stay

MAR

medication administration record

MRSA

methicillin‐resistant Staphylococcus aureus

NHSN

National Healthcare Safety Network

Ped_BSCA

pediatric broad‐spectrum antibacterial agents predominantly used for community‐acquired infections

Ped_BSHO

pediatric broad‐spectrum antibacterial agents predominantly used for hospital‐onset infections

Ped_GramPos

pediatric antibacterial agents predominantly used for resistant Gram‐positive infections (e.g., MRSA)

Ped_NSBL

pediatric narrow‐spectrum β‐lactam agents

Ped_XDR

pediatric antibacterial agents predominantly used for extensively drug‐resistant bacteria

POD

postoperative day

VRE

vancomycin‐resistant enterococci

1. Introduction

Perioperative antibacterial prophylaxis in pediatric solid organ transplantation helps prevent postoperative infection, but variation in practice can create measurable differences in antibiotic exposure. Standardized protocols can reduce that variation by clarifying the intended regimen, limiting extension of therapy beyond planned stop dates, and creating a framework for stewardship and safety monitoring [1, 2, 3].

Children undergoing transplantation receive intense immunosuppression, and some transplant admissions—particularly after intestinal and multivisceral procedures—extend for weeks. In that setting, even small differences in prophylactic breadth or stop dates can add many antibiotic‐days, increase the risks of toxicity, Clostridioides difficile infection, and resistance, and complicate later decisions about empiric escalation or de‐escalation [3, 4, 5, 6]. Because transplant admissions often extend for weeks, perioperative prophylaxis cannot be evaluated only at incision. Stewardship measurement also requires determining whether protocol‐linked exposure continues beyond intended stop targets during the admission.

Perioperative prophylaxis is especially well suited to protocol standardization because it depends on a small set of recurring decisions: which agents to use, when to give them, whether to add methicillin‐resistant Staphylococcus aureus (MRSA)‐ and vancomycin‐resistant enterococci (VRE)‐directed coverage, and when protocol‐linked antibiotics should be reassessed or discontinued. Future quality‐improvement efforts, however, require analytic methods that can reproducibly measure those stewardship targets across transplant populations.

When organ‐specific comparative evidence is limited, local protocols often define those recurring decisions and may evolve over time [2, 3, 7, 8, 9]. Multicenter pediatric data illustrate the resulting variation. In pediatric heart transplantation, postoperative prophylaxis may last only 1 day at some centers but commonly extends longer, and both agent selection and duration vary substantially across hospitals [10]. Uncertainty affects other organs as well, although not in the same way. In kidney transplantation, prophylaxis is typically brief (often ≤ 24 h) [11]. In liver transplantation, prophylaxis is usually framed around biliary and enteric contamination risk and often stops within 24–48 h, but varies among centers and may be longer in high‐risk patients [12, 13]. Intestinal and multivisceral transplantation carry the greatest early polymicrobial risk and the longest postoperative courses, yet evidence defining the optimal duration of prophylaxis remains limited [8, 14]. Consistent with that uncertainty, adult guidance from the American Society of Transplantation Infectious Diseases Community of Practice generally favors limiting prophylaxis in uncomplicated cases (≤ 72 h) while reserving longer courses, up to 7 days, for selected high‐risk situations such as infected mesh or fistula [7, 9].

At our center, pediatric infectious diseases, transplant services, and pharmacy update organ‐specific consensus protocols approximately every 2 years, including recommended regimens for patients with β‐lactam allergies. Because some operational details evolved during the study period, this audit used the most recently updated organ‐specific protocols as a prespecified unified reference standard rather than as year‐specific adjudication of contemporaneous practice.

Using these organ‐specific protocols as reference standards, we retrospectively evaluated baseline perioperative prophylaxis practice in pediatric solid organ transplantation. Our primary goal was to create reproducible electronic stewardship measures for major perioperative prophylaxis processes, including on‐time dosing, regimen selection, MRSA‐ or VRE‐directed add‐on use, double anaerobic coverage, Gram‐negative broadening during the prophylaxis window, and continuation beyond intended stop targets, rather than to retrospectively adjudicate the appropriateness of individual prescribing decisions. A secondary goal was to demonstrate that a small number of routinely available electronic data streams can be used to create reproducible measures for auditing perioperative prophylaxis practice, identifying stewardship targets, and supporting future prospective quality‐improvement studies [1, 15, 16].

2. Methods

2.1. Study Design and Setting

We conducted a retrospective audit of perioperative antibacterial prophylaxis in pediatric (age < 18 years) solid organ transplantation at Cleveland Clinic Children's, a tertiary‐care children's hospital, from September 11, 2010, through November 16, 2021. We compared observed practice with our most recent institutional organ‐specific perioperative prophylaxis protocols prespecified as the study reference standard (Table 1). Except for on‐time pre‐incision dosing, which followed general surgical prophylaxis guidance, regimen selection, stop cutoffs, and classification windows were defined by institutional protocol. The audit was designed to evaluate protocol concordance and exposure patterns, not to adjudicate the clinical appropriateness of individual deviations. No prospective stewardship intervention, audit‐and‐feedback program, or protocol‐directed quality‐improvement initiative was implemented as part of this study. Transplant candidates routinely underwent pretransplant infectious diseases evaluation, and transplant recipients were followed by infectious diseases clinicians during routine transplant care.

TABLE 1.

Perioperative prophylaxis regimens, stop targets, and audit windows for pediatric solid organ transplantation.

Transplant type	Standard perioperative regimen	Protocol‐defined regimen for IgE‐mediated β‐lactam allergy	Protocol‐defined MRSA‐ or VRE‐directed add‐on	Intended stop cutoff for main regimen	Intended stop cutoff for add‐on agents	Classification window used for the audit
Heart	Cefazolin; cefuroxime (legacy)	Vancomycin; clindamycin	MRSA: Vancomycin	Stop by 48 h post‐op	Stop by 48 h post‐op	POD 0–2
Liver	Piperacillin/tazobactam; ampicillin/sulbactam	Aztreonam + vancomycin	MRSA: Vancomycin; VRE: Daptomycin	Stop by 48 h post‐op	Stop by 48 h post‐op	POD 0–2
Kidney	Cefazolin	Vancomycin; clindamycin	MRSA: Vancomycin	Stop by 24 h post‐op	Stop by 24 h post‐op	POD 0–1
Intestinal	Piperacillin/tazobactam	Aztreonam + metronidazole + vancomycin	MRSA: Vancomycin; VRE: Daptomycin	Stop main regimen by POD 7	Stop add‐ons by 48 h post‐op	Main regimen: POD 0–7; Add‐ons: POD 0–2
Multivisceral	Piperacillin/tazobactam	Aztreonam + metronidazole + vancomycin	MRSA: Vancomycin; VRE: Daptomycin	Stop main regimen by POD 7	Stop add‐ons by 48 h post‐op	Main regimen: POD 0–7; Add‐ons: POD 0–2

Note: On‐time prophylaxis was defined as administration within 60 min before incision for most agents and within 120 min for vancomycin or fluoroquinolones. Protocol prophylaxis windows were defined in advance by postoperative day: kidney, POD 0–1; heart and liver, POD 0–2; and intestinal and multivisceral transplantation, POD 0–7 for the main regimen. Conditional MRSA‐ and VRE‐directed add‐on agents were evaluated through POD 2, or through the end of the organ‐specific prophylaxis window when that window was longer. For heart transplantation, cefuroxime reflects a legacy standard used in earlier study years.

Abbreviations: MRSA, methicillin‐resistant Staphylococcus aureus; POD, postoperative day; VRE, vancomycin‐resistant Enterococcus.

2.2. Cohort and Follow‐Up

The cohort included pediatric heart, liver, kidney, intestinal, and multivisceral transplant procedures. Each transplant procedure served as a separate observation. Follow‐up for each procedure began on the day of transplant, defined as postoperative day (POD) 0, and was linked to the primary transplant admission. For children who underwent more than one transplant during the study period, follow‐up for a given procedure ended at the next transplant procedure or POD 365, whichever came first, to avoid counting the same antibiotic exposure twice.

2.3. Data Sources

We linked three data sources: Peri‐incision prophylaxis administrations from the Anesthesia Record Keeping System (ARKS), inpatient antibacterial administrations from the medication administration record (MAR), and transplant procedure and admission data for the primary transplant hospitalization. Antibacterial exposure was summarized using 2024 National Healthcare Safety Network (NHSN) pediatric antimicrobial use categories, including pediatric broad‐spectrum antibacterial agents predominantly used for hospital‐onset infections (Ped_BSHO), pediatric antibacterial agents predominantly used for extensively drug‐resistant bacteria (Ped_XDR), and pediatric antibacterial agents predominantly used for resistant Gram‐positive infections (Ped_GramPos) [17]. We also tracked carbapenem exposure as a marker of broadening Gram‐negative therapy.

2.4. How Prophylaxis Was Assessed

For each transplant procedure, we assessed prophylaxis in six domains:

First, we assessed whether prophylaxis was given on time. On‐time administration was defined as at least one intravenous prophylactic dose given within 60 min before incision for most agents and within 120 min for vancomycin or fluoroquinolones, consistent with general surgical prophylaxis guidance [11].

Second, we classified the initial prophylaxis regimen given during the protocol‐defined prophylaxis window as one of three types: The recommended organ‐specific regimen, the protocol‐defined regimen for patients with IgE‐mediated β‐lactam allergy, or another broad Gram‐negative regimen not designated as protocol prophylaxis.

Third, we assessed whether the main prophylaxis regimen stopped by the protocol‐defined organ‐specific cutoff. These institutional classification windows were defined a priori by POD: Kidney, POD 0–1; heart and liver, POD 0–2; and intestinal and multivisceral transplantation, POD 0–7.

Fourth, we assessed double anaerobic coverage during the prophylaxis window, defined as any day on which a patient received metronidazole plus piperacillin‐tazobactam, ampicillin‐sulbactam, ertapenem, imipenem‐cilastatin, meropenem, or doripenem.

Fifth, we assessed the use of protocol‐defined Gram‐positive add‐on agents, specifically intravenous vancomycin for MRSA coverage and daptomycin for VRE coverage. For this audit, these add‐on agents were evaluated through POD 0–2, or through the end of the organ‐specific protocol window when that window was shorter, as in kidney transplantation. Because colonization status and clinical indication were not readily available, we measured use and duration rather than appropriateness. For this add‐on analysis, vancomycin used as part of a β‐lactam allergy regimen was not counted as an MRSA add‐on.

Sixth, we assessed whether Gram‐negative therapy broadened during the prophylaxis window. We grouped Gram‐negative antibiotics from narrower to broader categories using pediatric antimicrobial use groupings (Table S1): Ped_NSBL, Ped_BSCA, Ped_BSHO, carbapenem, and Ped_XDR. The starting category was assigned from peri‐incision agents when available or from the first Gram‐negative agent given within the prophylaxis window. We then recorded broadening during that window to Ped_BSHO, a carbapenem, or Ped_XDR.

2.5. Antibiotic Exposure and Days of Therapy

We measured antibiotic exposure as days of therapy (DOT), defined as one antibiotic agent administered on a given calendar day [17]. DOT was counted from transplant day (Day 0) through POD 365 or discharge from the transplant admission, whichever occurred first. MAR administrations were deduplicated to unique procedure‐by‐antibiotic‐by‐calendar‐day records within that primary admission.

Broad‐spectrum antibacterial exposure was defined using pediatric NHSN agent lists for Ped_BSHO and Ped_XDR [17]. Broad‐spectrum totals were calculated as Ped_BSHO plus Ped_XDR (Table S1). For agents considered broad‐spectrum only when given intravenously, including amikacin, aztreonam, and tobramycin, we required intravenous administration.

2.6. Continuation of Broad‐Spectrum Exposure Beyond Intended Stop Windows

For each procedure, we evaluated whether broad‐spectrum antibacterial exposure continued beyond the organ‐specific prophylaxis window. Broad‐spectrum exposure included pediatric NHSN broad‐spectrum hospital‐onset agents (Ped_BSHO) and Ped_XDR [17]. We first identified broad‐spectrum agents that were part of the recommended perioperative regimen or the protocol‐defined β‐lactam allergy regimen for that transplant type. For heart transplantation, cefuroxime was treated as an acceptable pre‐2019 alternative to cefazolin.

We then classified broad‐spectrum exposure as occurring within the intended prophylaxis window or after that window ended. To distinguish uninterrupted continuation from later reuse, post‐window use of the same agent was considered contiguous until a gap of three or more days without that agent occurred; subsequent use after such a gap was classified as later reuse. These categories were descriptive and were intended to support stewardship measurement and future quality‐improvement review rather than to adjudicate the appropriateness of later therapy.

2.7. Gram‐Positive Exposure Linked to Protocol Pathways

We separately evaluated Gram‐positive DOT during the primary admission, including all antibiotic‐days in the NHSN pediatric Gram‐positive category (Ped_GramPos) regardless of indication (Table S1). Within this total, we first identified Ped_GramPos DOT that occurred during the organ‐specific perioperative protocol window and classified those agent‐days by source: Protocol‐defined MRSA‐ or VRE‐directed add‐on agents, Gram‐positive agents included in the protocol‐defined regimen for patients with IgE‐mediated β‐lactam allergy, or other non‐protocol Gram‐positive exposure. We then evaluated Ped_GramPos DOT occurring after the perioperative protocol window. Post‐window Ped_GramPos exposure was partitioned descriptively into continued protocol‐linked exposure without a gap, later reuse after a gap, and other post‐window exposure.

2.8. Statistical Analysis

We summarized categorical measures as counts and percentages, using the number of procedures within each organ group as the denominator. Durations were summarized as medians with interquartile ranges (IQRs). DOT rates per 1000 LOS days were calculated using a transplant‐based timeline in which the day of transplant was defined as Day 0 for both antibiotic exposure and LOS; 95% confidence intervals were calculated with Poisson methods. Analyses were performed in Python 3.12 using Pandas, NumPy, and SciPy.

2.9. Ethics and Data Security

The study was approved by the Cleveland Clinic's Institutional Review Board. Data were accessed within the secure Enterprise Data Vault (Teradata Data Lab), clinical data were managed in REDCap, and analyses were performed on password‐protected enterprise workstations [18, 19].

3. Results

3.1. Cohort and Baseline Antibacterial Exposure During the Primary Admission

During the primary admission, total antibacterial use was 6325 DOT, or 896.3 per 1000 LOS days (95% CI, 874.3–918.6) (Table S2). Ped_BSHO accounted for 1829 DOT (259.2 per 1000 LOS days [95% CI, 247.4–271.3]). Ped_XDR exposure was uncommon overall (41 DOT; 5.8 per 1000 LOS days [95% CI, 4.2–7.9]) and occurred only in intestinal and multivisceral procedures. Broad‐spectrum exposure, defined as Ped_BSHO plus Ped_XDR, totaled 1870 DOT (265.0 per 1000 LOS days [95% CI, 253.1–277.3]). Ped_GramPos totaled 1008 DOT (142.8 per 1000 LOS days [95% CI, 134.2–151.9]).

3.2. Timing and Selection of the Initial Prophylaxis Regimen

Peri‐incision prophylaxis was given on time in 95.7% of procedures (157/164), with similarly high performance across organ groups (93.1%–100.0%; Table 2). Across the cohort, 87.2% of procedures received the recommended initial regimen, 7.3% received the protocol‐defined regimen for patients with IgE‐mediated β‐lactam allergy, and 4.3% received another broad Gram‐negative regimen not designated as protocol prophylaxis.

TABLE 2.

Cohort characteristics and perioperative prophylaxis performance, by transplant type.

Panel A: Cohort size and inpatient antibiotic exposure.
Transplant type	Procedures, n	Total hospital length of stay, d	Length of stay per transplant, median (IQR), d	Antibacterial DOT per transplant, median (IQR)
Heart	59	2604	30 (19–52)	17 (10–37)
Liver	58	1856	16 (12–24)	20 (11–28)
Kidney	29	299	9 (8–11)	9 (8–11)
Intestinal	8	862	100 (56–127)	129 (83–182)
Multivisceral	10	1436	138 (52–205)	168 (140–319)
Overall	164	7057	20 (13–47)	17 (9–34)

Panel B: Timing and initial regimen selection.
Transplant type	On‐time pre‐incision prophylaxis, n (%)	Recommended organ‐specific regimen, n (%)	IgE‐mediated β‐lactam allergy regimen, n (%)	Other broad Gram‐negative regimen not designated as protocol prophylaxis, n (%)
Heart	57 (96.6)	48 (81.4)	9 (15.3)	1 (1.7)
Liver	55 (94.8)	53 (91.4)	0 (0.0)	4 (6.9)
Kidney	27 (93.1)	28 (96.6)	1 (3.4)	0 (0.0)
Intestinal	8 (100.0)	7 (87.5)	1 (12.5)	0 (0.0)
Multivisceral	10 (100.0)	7 (70.0)	1 (10.0)	2 (20.0)
Overall	157 (95.7)	143 (87.2)	12 (7.3)	7 (4.3)

Panel C: Duration, discontinuation, and selected within‐window exposure features.
Transplant type	Initial prophylaxis duration, median (IQR), d	Initial prophylaxis regimen stopped by intended cutoff, n (%)	Metronidazole plus second anaerobic‐active agent, n (%)	Any carbapenem exposure during prophylaxis window, n (%)
Heart	3 (2–3)	44 (74.6)	0 (0.0)	0 (0.0)
Liver	4 (3–6)	21 (36.2)	2 (3.4)	3 (5.2)
Kidney	1 (1–1)	29 (100.0)	0 (0.0)	0 (0.0)
Intestinal	8 (8–16)	4 (50.0)	8 (100.0)	1 (12.5)
Multivisceral	10 (8–12)	2 (20.0)	8 (80.0)	4 (40.0)
Overall	3 (2–6)	100 (61.0)	18 (11.0)	8 (4.9)

Note: Panel A: LOS was measured from transplant day (POD 0) through discharge from the primary transplant admission, up to POD 365. Antibacterial DOT counts unique antibacterial agent‐days, deduplicated at procedure × agent × calendar day, during the primary transplant admission. Length of stay per transplant and antibacterial DOT per transplant are reported as median (IQR), in days. Panel B: Percentages are based on the number of procedures within each transplant type. On‐time prophylaxis was defined as administration within 60 min before incision for most agents and within 120 min for vancomycin or fluoroquinolones. “Recommended organ‐specific regimen” refers to the standard protocol‐defined regimen for that transplant type. “Other broad Gram‐negative regimen” refers to broad Gram‐negative exposure not designated as protocol prophylaxis. Panel C: Initial prophylaxis duration is reported as median (IQR), in days, and reflects the number of postoperative days on which the initial prophylaxis backbone was administered. “Stopped by intended cutoff” refers to discontinuation of the initial prophylaxis backbone by the organ‐specific protocol stop target. Metronidazole plus second anaerobic‐active agent indicates double anaerobic coverage during the protocol prophylaxis window. Carbapenem exposure was counted when any study‐defined carbapenem was administered during the protocol prophylaxis window. No antibacterial agents predominantly used for extensively drug‐resistant bacteria (Ped_XDR) were used during the prophylaxis window.

Abbreviations: DOT, days of therapy; IQR, interquartile range; LOS, length of stay; POD, postoperative day.

3.3. Continuation Beyond Intended Stop Targets and Gram‐Negative Broadening

Continuation beyond intended stop targets was the clearest process measure requiring prospective review (Table 2). Overall, 61.0% of procedures (100/164) stopped the main regimen by the intended organ‐specific cutoff.

Broadening of Gram‐negative therapy during the prophylaxis window was uncommon. Carbapenem exposure occurred in 4.9% of procedures (8/164), and no procedures received a Ped_XDR agent during the window. Kidney procedures began almost uniformly in the Ped_NSBL category (96.6%), whereas liver and abdominal transplant procedures more often began in the Ped_BSHO category (Table S3). Overall, therapy broadened to Ped_BSHO in 10.4% of procedures (17/164), most often in liver, kidney, and heart transplantation. Broadening to a carbapenem was uncommon overall (1.8%, 3/164) but was concentrated in multivisceral procedures. All 41 Ped_XDR DOT occurred later in the primary admission and were confined to intestinal and multivisceral procedures.

Double anaerobic coverage during the prophylaxis window occurred in 11.0% of procedures (18/164) and was concentrated in intestinal (100.0%) and multivisceral transplantation (80.0%) (Table 2).

3.4. Descriptive Partitioning of Broad‐Spectrum Exposure During the Primary Admission

Table S4 shows descriptive partitioning of primary‐admission broad‐spectrum DOT by protocol–regimen relationship and prophylaxis window. Contiguous protocol‐linked continuation beyond the intended prophylaxis window accounted for 8.6% of primary‐admission broad‐spectrum DOT.

3.5. MRSA‐ and VRE‐Directed Add‐On Agents

Use of MRSA‐directed add‐on agents was concentrated in the abdominal transplant groups (Table 3). MRSA add‐on agents were used in 75.0% of intestinal procedures (6/8) and 90.0% of multivisceral procedures (9/10), compared with 43.1% of liver procedures (25/58) and 22.0% of heart procedures (13/59). No kidney procedures received an MRSA add‐on. VRE‐directed add‐on agents were observed only in liver (1/58, 1.7%) and multivisceral procedures (3/10, 30.0%).

TABLE 3.

Perioperative Gram‐positive antibiotic exposure and protocol‐defined add‐on use.

Panel A: Use and duration of protocol‐defined MRSA‐ and VRE‐directed add‐on agents.
Transplant type	MRSA‐directed add‐on used, n (%)	MRSA‐directed add‐on beyond protocol‐defined duration, n (%)	MRSA‐directed add‐on DOT among users, median (IQR), d	VRE‐directed add‐on used, n (%)	VRE‐directed add‐on beyond protocol‐defined duration, n (%)	VRE‐directed add‐on DOT among users, median (IQR), d
Heart	13 (22.0)	0 (0.0)	2 (2–3)	0 (0.0)	0 (0.0)	—
Liver	25 (43.1)	0 (0.0)	2 (2–3)	1 (1.7)	0 (0.0)	1 (1–1)
Kidney	0 (0.0)	0 (0.0)	—	0 (0.0)	0 (0.0)	—
Intestinal	6 (75.0)	5 (62.5)	8 (6–8)	0 (0.0)	0 (0.0)	—
Multivisceral	9 (90.0)	7 (70.0)	6 (3–8)	3 (30.0)	3 (30.0)	8 (6–8)
Overall	53 (32.3)	12 (7.3)	3 (2–3)	4 (2.4)	3 (1.8)	6 (4–8)

Panel B: Sources of Gram‐positive DOT during the prophylaxis window.
Transplant type	Total protocol‐window Gram‐positive DOT	Protocol‐defined MRSA‐ or VRE‐ directed add‐on DOT, n (%)	IgE‐mediated β‐lactam allergy regimen DOT, n (%)	Other Gram‐positive DOT during protocol window, n (%)
Heart	56	29 (51.8)	22 (39.3)	5 (8.9)
Liver	60	58 (96.7)	0 (0.0)	2 (3.3)
Kidney	2	0 (0.0)	1 (50.0)	1 (50.0)
Intestinal	46	38 (82.6)	0 (0.0)	8 (17.4)
Multivisceral	77	68 (88.3)	0 (0.0)	9 (11.7)
Overall	241	193 (80.1)	23 (9.5)	25 (10.4)

Note: Gram‐positive refers to the National Healthcare Safety Network pediatric antimicrobial use category Ped_GramPos. Panel A: Percentages are based on the number of procedures within each transplant type. “Beyond protocol‐defined duration” indicates use extending past the protocol‐defined stop target for that add‐on agent. DOT among users summarizes exposure only among procedures that received that add‐on agent and is reported as median (IQR), in days. Panel B: Percentages are based on total Gram‐positive DOT during the protocol prophylaxis window for that transplant type. Protocol‐window Gram‐positive DOT includes exposure from protocol‐defined MRSA‐ or VRE‐directed add‐on agents, Gram‐positive agents used as part of the IgE‐mediated β‐lactam allergy regimen, and other Gram‐positive exposure during the protocol window.

Abbreviations: DOT, days of therapy; IQR, interquartile range; MRSA, methicillin‐resistant Staphylococcus aureus; VRE, vancomycin‐resistant Enterococcus.

Continuation beyond the protocol‐defined add‐on duration occurred in 12 procedures (7.3%) for MRSA‐directed agents and 3 procedures (1.8%) for VRE‐directed agents (Table 3). This pattern was concentrated in intestinal and multivisceral transplantation. MRSA‐directed add‐ons continued beyond the protocol‐defined duration in 5 of 8 intestinal procedures (62.5%) and 7 of 10 multivisceral procedures (70.0%), but in no heart, liver, or kidney procedures. Among add‐on users, median MRSA‐directed duration was 2 days (IQR, 2–3) in heart and liver transplantation, compared with 8 days (IQR, 6–8) in intestinal and 6 days (IQR, 3–8) in multivisceral transplantation; median VRE‐directed duration was 1 day in liver and 8 days (IQR, 6–8) in multivisceral transplantation.

3.6. Gram‐Positive Antibiotic Exposure During the Primary Admission

Gram‐positive exposure during the primary admission totaled 1008 agent‐days (Table S2). During the perioperative protocol window, Ped_GramPos exposure totaled 241 agent‐days (Table 3). Most came from protocol‐defined MRSA‐ or VRE‐directed add‐on agents (193/241, 80.1%), with smaller contributions from Gram‐positive agents included in the IgE‐mediated β‐lactam allergy regimen (23/241, 9.5%) and other non‐protocol exposure (25/241, 10.4%). In liver, intestinal, and multivisceral transplantation, perioperative Gram‐positive exposure was dominated by protocol‐defined add‐on agents, whereas heart transplantation showed larger contributions from the IgE‐mediated β‐lactam allergy pathway. Descriptive partitioning of post‐window Gram‐positive exposure patterns is provided in Table S5.

4. Discussion

In this retrospective audit of perioperative prophylaxis in pediatric solid organ transplantation, the primary finding was that the most reliable parts of prophylaxis were the start of therapy: Prophylaxis was almost always administered on time, and most procedures began with either the recommended organ‐specific regimen or the protocol‐defined alternative for patients with beta‐lactam allergy. The less reliable process measure was continuation of protocol‐linked antibiotics beyond the intended organ‐specific cutoff. The major stewardship opportunity is therefore not simply to stop antibiotics earlier, but to understand why they continue and to distinguish clinically justified continuation from modifiable prescribing patterns.

This distinction is important for prospective stewardship. A baseline audit does not need to determine whether every deviation was appropriate to identify measurable process targets. Instead, it defines where prospective review and balancing‐outcome monitoring should focus. In this cohort, those targets were stop‐date reliability, prolonged MRSA‐ or VRE‐directed add‐on use, double anaerobic coverage, and broadening of Gram‐negative therapy during the prophylaxis window.

The analysis of primary‐admission broad‐spectrum exposure should be interpreted as a descriptive measurement rather than clinical adjudication. Contiguous protocol‐linked continuation beyond the prophylaxis window represented the exposure pattern most directly connected to prolonged perioperative prescribing, accounting for 8.6% of broad‐spectrum DOT. Later reuse after a gap and non‐protocol broad‐spectrum exposure were retained to demonstrate reproducible electronic partitioning of exposure patterns, but those categories require case‐level clinical review before appropriateness can be determined.

Gram‐positive stewardship opportunities were concentrated in abdominal transplantation. Most perioperative Gram‐positive DOT reflected protocol‐defined MRSA‐ or VRE‐directed add‐on agents, and continuation beyond intended stop targets occurred primarily in intestinal and multivisceral transplantation. These findings suggest that abdominal transplant pathways may benefit from more standardized reassessment of add‐on duration and clearer stop expectations.

Perioperative stewardship also requires ongoing reassessment of the protocols themselves. Evidence supporting prophylaxis duration, add‐on agents, and spectrum selection remains limited for some pediatric transplant populations, particularly intestinal and multivisceral transplantation [7, 9, 14]. This study shows that only a small number of routinely available electronic data streams—anesthesia timing, medication‐administration records, transplant procedures, and admission data—are needed to create reproducible tools for auditing protocol performance over time and identifying areas that may benefit from prospective evaluation, standardization, or refinement as additional stewardship and clinical outcome data become available.

The main contribution of this work is therefore methodological as well as descriptive. By linking anesthesia timing, medication‐administration data, transplant procedures, and organ‐specific protocol windows, we created reproducible electronic measures for perioperative prophylaxis practice. These measures can support run‐chart monitoring, audit‐and‐feedback, and prospective evaluation of protocol revisions while preserving the ability to monitor safety outcomes such as surgical site infection, bloodstream infection, readmission, C. difficile infection, resistant organisms, and toxicity [1, 2, 3, 4, 15].

This study has limitations. It was conducted at a single center, and the intestinal and multivisceral groups were small. We used the most recently updated institutional protocols as a unified reference standard; therefore, the analysis should be interpreted as a baseline stewardship audit rather than year‐specific adjudication of contemporaneous protocol adherence. We did not determine the clinical reason for continued use or later reuse of antibiotics, nor did we assess the appropriateness of MRSA‐ or VRE‐directed add‐on agents because colonization and indication data were not readily available. Some deviations from protocol may have been clinically appropriate because of allergy history, colonization, donor or recipient infection risk, delayed closure, size mismatch, early complications, or evolving clinical concern. The analysis focused on the primary transplant admission and did not capture donor infections, readmissions, antifungal prophylaxis, antiviral prophylaxis, or later outcomes associated with the antibiotic exposures observed.

In summary, perioperative prophylaxis in pediatric solid organ transplantation was usually started on time and usually began with the expected regimen. The main baseline process measure requiring prospective review was continuation of protocol‐linked antibiotics beyond intended stop times, particularly in abdominal transplantation; future quality improvement work should prospectively evaluate which continuation episodes are clinically justified and which may represent modifiable prescribing patterns. This audit establishes a practical measurement framework for future prospective quality improvement studies focused on organ‐specific reassessment and stop targets, review of prolonged MRSA‐ and VRE‐directed add‐on use, double anaerobic coverage, and balancing outcomes.

Funding

Dr. Foster received support through a grant from the Cleveland Clinic Healthcare Delivery & Implementation Science Center to develop data extraction tools used in this project.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting Information: tid70266‐sup‐0001‐SuppMat.docx