Association of Intraoperative Ketorolac Administration with Pediatric Anterior Cruciate Ligament Reconstruction Outcomes: A Retrospective Cohort Study Authors: Muaaz Wajahath, Ahmad Hassan, Amir Mahdi Ghafarian, Malik Al-Jamal, Abdul-Lateef Shafu, Haytham Al-Qasmi, Walid Yasir Published: 2026 ABSTRACT BackgroundPediatric anterior cruciate ligament reconstruction (ACLR) is increasingly common. Optimizing pain management while minimizing opioid exposure is a priority. Intraoperative ketorolac, a nonsteroidal anti-inflammatory drug (NSAID), is used for analgesia, but its impact on healing and surgical outcomes remains controversial. This study evaluates the effects of intraoperative ketorolac on postoperative outcomes in pediatric ACLR patients (< 18 years).MethodsWe conducted a retrospective cohort study using the TriNetX national clinical database (102 healthcare organizations). We identified pediatric patients who underwent arthroscopic ACLR (Current Procedural Terminology [CPT] 29888) between 2005 and 2025. Two groups were defined: those who received intraoperative ketorolac (15 or 30 mg intravenous [IV]) and those who did not. Propensity score matching (1 : 1) was performed on baseline covariates (age, sex, race/ethnicity, and comorbidities) to balance the cohorts. Outcomes were assessed at 90 days (opioid utilization and complications) and at 2 years (ACL revision surgery, knee stiffness, and mechanical complications).ResultsA total of 3,409 patients met inclusion criteria; 1,443 pairs were successfully matched. At 90 days, the ketorolac group had a slightly lower mean number of opioid prescriptions (1.20 ± 0.44 vs. 1.43 ± 1.26, p = 0.039). Rates of muscle atrophy, wound disruption, and infection were similar (~0.8% in both groups, p = 1.0). The incidence of an opioid dependence diagnosis was low but significantly higher in the ketorolac group (1.0% vs. 0.3%; risk ratio [RR], 4.0; p < 0.001). At 2-year follow-up, most complication rates were similar between groups (e.g., mechanical complications 0.7% each; knee stiffness 12.5% vs. 10.9%, p = 0.18). However, the ketorolac cohort had a significantly higher rate of repeat ACLR (6.0% vs. 2.6%; RR, 2.29; 95% CI, 1.58–3.33; p < 0.001).ConclusionsIntraoperative ketorolac administration was associated with reduced postoperative opioid requirements and no increase in short-term complications. However, a higher incidence of ACL revision surgery was observed at 2 years, representing an association that does not establish causation. Pediatric and adolescent anterior cruciate ligament (ACL) injuries have increased in frequency over the past 2 decades, in part due to higher youth sports participation, early sports specialization, and improved recognition of ACL tears in young athletes.1) ACL reconstruction (ACLR) in skeletally immature patients poses unique challenges but generally yields favorable functional outcomes with modern techniques. Nevertheless, young patients are at substantial risk for subsequent knee injuries. Epidemiologic data indicate that within 2 years of primary ACLR, approximately 3%–5% of pediatric patients require revision of the ipsilateral graft and a similar proportion sustain contralateral ACL tears.12) Female athletes tend to have higher rates of contralateral ACL injury, whereas males have slightly higher graft rupture.12) Aside from re-injury, other complications such as arthrofibrosis and infection, though uncommon, can occur in pediatric ACLR. Large series report overall surgical complication rates around 10%, with knee stiffness (~3%–5%) and deep infection (~1%–2%) being the most prevalent issues.3) Preventing these adverse outcomes is critical to optimize long-term knee health in this young population. Effective postoperative pain management is also a priority in pediatric orthopedic surgery. Concern over opioid misuse has prompted efforts to improve opioid stewardship through multimodal analgesia.14) Nonsteroidal anti-inflammatory drugs (NSAIDs) like ketorolac, a nonselective cyclooxygenase (COX) inhibitor, are commonly used adjuncts for pain control due to their opioid-sparing effects.567) In children, ketorolac has demonstrated significant analgesic benefits, such as reduced pain scores and lower opioid requirements after orthopedic procedures.57) For example, previous research in pediatric fracture surgery found that adding ketorolac led to reductions in length of stay, postoperative pain ratings, and opioid dosing, with no increase in short-term complications.8) These findings align with the general notion that NSAIDs can improve postoperative comfort and facilitate recovery.9) Given the current emphasis on opioid stewardship and the risks of opioid exposure in adolescents, any safe reduction in opioid requirements is clinically meaningful. Notably, although most pediatric patients do not require long-term opioids after ACLR, recent analyses have found that approximately 4% to 7% of opioid-naive adolescents exhibit new persistent opioid use after common surgeries.410) Reducing initial opioid exposure through multimodal analgesia (including NSAIDs) is therefore an important strategy to mitigate the pathway to chronic opioid use. Despite its analgesic advantages, the use of ketorolac in orthopedic surgery has been approached with caution due to concerns about its effects on tissue healing. NSAIDs inhibit COX and prostaglandin production, pathways integral to inflammation and bone remodeling.511) Animal studies have suggested that selective COX-2 inhibitors may impair soft tissue, tendon-to-bone, and fracture healing.712) NSAIDs pose a theoretical risk for increased rates of graft failure or meniscal repair failure, as ACL graft incorporation and meniscal repair healing are thought to rely on post-injury inflammation; however, this has not been definitively demonstrated in human clinical studies.57) Clinical results present mixed findings. Some early studies raised concern that NSAID use might negatively affect outcomes of ligament or meniscus surgery, but more recent clinical data have not borne out a large detrimental effect.7) Multiple clinical studies in adult and mixed-age cohorts have reported no significant association between perioperative NSAID use and ACL graft failure or meniscal repair failure, suggesting that the concern raised by preclinical data may not translate directly to clinical practice.5) For example, Soreide et al.13) reported that limited perioperative NSAID use did not compromise graft survival after ACLR. Similarly, an investigation of ketorolac during meniscal repair (often concomitant with ACLR) showed no increase in meniscal repair failure or reoperation with ketorolac vs. no NSAID.14) Further, a randomized trial by Ge et al.15) found that postoperative NSAID (celecoxib) use yielded equivalent knee stability and patientreported outcomes compared to an opioid-based regimen after ACLR. These clinical studies, mostly in adults or mixed-age cohorts, suggest that short-term NSAID use may be safe with respect to soft-tissue healing in ACL surgery. However, high-quality evidence in pediatric patients is limited. A 2018 Cochrane review noted insufficient data to determine whether ketorolac provides a clear benefit or harm in children.7) Importantly, children and adolescents might have different biological responses, and the impact of NSAIDs on an immature musculoskeletal system, with unfused epiphyseal growth plates, is not fully understood. Given the ongoing debate and the notable lack of large-scale pediatric-specific data regarding single-dose perioperative NSAID exposure, we sought to investigate the association of intraoperative ketorolac with ACLR outcomes in patients under 18. Our aim was to provide evidence to guide safe analgesic practices in pediatric ACLR and to detect any sign of adverse effects of ketorolac on surgical success. We hypothesized that ketorolac administration would be associated with reduced postoperative opioid use and equivalent rates of complications. METHODS Ethical review and approval were waived for this study due to the use of de-identified, aggregate data from the TriNetX Research Network. As the data contains no protected health information, the study is exempt from institutional review board. Patient consent was waived due to the retrospective design of the study and the use of de-identified data. Study Design and Data Source This study is a retrospective cohort analysis conducted using the TriNetX Research Network, a federated database of de-identified electronic health records from more than 100 healthcare organizations across the United States. This platform allows real-time aggregation of clinical data, including diagnoses, procedures, medications, and outcomes. Data for this analysis were extracted on February 15, 2025. The study was exempt from Institutional Review Board review as it used only de-identified, aggregate data. Patient Selection We identified pediatric patients (< 18 years old) who underwent primary ACLR between January 1, 2005, and January 1, 2025. ACLR was defined by the presence of Current Procedural Terminology (CPT) code 29888 (arthroscopically aided ACL repair/augmentation or reconstruction) in the patient's record. Within this cohort, patients were stratified into 2 groups based on exposure to intraoperative ketorolac. The ketorolac group consisted of those who received an injectable form of ketorolac (15 mg or 30 mg dose, RxNorm code 35827 for ketorolac tromethamine) on the day of surgery. Patients who did not receive any form of ketorolac during the surgical encounter comprised the comparison group. We excluded patients if ketorolac was given only orally or outside the immediate perioperative period, or if they had incomplete records of the index surgery. By focusing on intraoperative intravenous (IV)/intramuscular (IM) ketorolac administration, we aimed to standardize the exposure to a single dose given by the anesthesia or surgical team during ACLR. Outcomes and Follow-up We evaluated postoperative outcomes at 2 clinically relevant follow-up intervals: short-term (within 90 days post-surgery) and longer-term (up to 2 years post-surgery). Short-term outcomes were chosen to capture perioperative recovery, including pain management efficacy and early complications. These 90-day outcomes were (1) opioid analgesia utilization, measured by postoperative opioid prescription fills, identified using the Veterans Health Administration (VA) National Drug File class CN101 (opioid analgesics) as mapped within the National Library of Medicine (NLM) RxNorm terminology; we recorded the number of opioid prescriptions filled per patient in the 90-day period; (2) opioid dependence diagnosis, defined by International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) code F11.2 (indicating opioid dependence or abuse) within 90 days; (3) muscle atrophy, defined by new diagnosis codes for muscle wasting/atrophy of lower limb (International Classification of Diseases, Tenth Revision [ICD-10] M62.5x) which could indicate disuse atrophy post-surgery; (4) wound disruption (ICD-10 T81.3, dehiscence of operative wound); and (5) wound infection (ICD-10 T81.4, infection of surgical wound). These diagnoses were used as proxies for significant early complications such as poor wound healing or infection requiring treatment. Long-term outcomes (assessed up to 2 years postoperatively) were selected to evaluate graft integrity and knee function. These 2-year outcomes included (1) mechanical complications related to internal fixation or implant (ICD-10 T84.4, which in this context could capture hardware issues such as screw loosening or breakage); (2) implant-related infection (ICD-10 T84.7, deep infection involving the graft or hardware after the immediate postoperative period); (3) knee stiffness (arthrofibrosis), identified by ICD-10 M25.66 (ankylosis/stiffness of knee joint); and (4) repeat ACLR (hereafter “repeat ACLR”), defined by any subsequent occurrence of CPT 29888 within the follow-up window. Because CPT 29888 does not encode laterality in the available data, this outcome is a composite that captures both ipsilateral revision ACLR (graft failure of the index knee) and contralateral primary ACLR (a new injury to the previously uninjured knee), which cannot be distinguished from one another. All subsequent references to “repeat ACLR” in this manuscript refer to this composite outcome and should not be interpreted as a measure of ipsilateral graft failure alone; and (5) knee manipulation under anesthesia (MUA) for postoperative motion loss (CPT 27570). These outcomes were tracked via diagnosis and procedure codes within the available data. Each patient was followed for the maximum available duration up to 2 years; patients without records beyond the index surgery were assumed lost to follow-up for long-term outcomes and were not counted as having the event. Outcome incidence was determined at the cohort level (proportion of patients experiencing the event at least once in the follow-up window). Statistical Analysis All analyses were performed within the TriNetX analytics platform, which provides built-in statistical tools. We first summarized baseline demographic and clinical characteristics of the ketorolac and no-ketorolac groups. Continuous variables (e.g., age) were reported as means with standard deviations (SDs) and compared using independent-samples t-tests. For all categorical outcomes including longer-term outcomes such as ACL revision surgery, cumulative incidence was compared between matched groups using risk ratios (RRs) with 95% CIs at the defined follow-up windows. To control for potential confounding, we utilized propensity score matching to create comparable cohorts. A logistic regression model was used to estimate the propensity of receiving intraoperative ketorolac, incorporating the following covariates: age at surgery, sex, race, ethnicity, baseline body mass index when available, and relevant baseline comorbidities, including diabetes mellitus, hypertensive disease, obesity, chronic kidney disease, chronic liver disease, and any recorded preoperative opioid use. One-to-one matching without replacement was performed using a greedy nearest-neighbor algorithm, with a caliper width of 0.1 pooled SD of the logit of the propensity score. Matching quality was assessed by examining standardized mean differences (SMDs) for each covariate, with SMD < 0.1 indicating adequate balance between groups. After matching, outcomes were compared with the matched sample. For categorical outcomes, we calculated absolute risk differences and RRs with 95% CIs. Chi-square or Fisher’s exact tests were used for matched categorical comparisons as appropriate. RESULTS Patient Characteristics A total of 3,409 pediatric patients underwent ACLR in the study period and met the inclusion criteria (Fig. 1). Among these, 1,624 patients (47.6%) received intraoperative ketorolac (15 or 30 mg IV) during the ACLR, while 1,785 patients (52.4%) did not receive ketorolac. Baseline comorbidities (e.g., obesity, diabetes, hypertension, chronic kidney disease, liver disease) were rare in this young population and showed no significant differences between groups. After 1:1 propensity matching, 1,443 patients from each group were successfully matched, representing 89% of the ketorolac cohort. No significant differences remained in age, sex, race/ethnicity, or comorbidity prevalence after matching (all p ≥ 0.15) (Table 1). The SMDs for all covariates were below 0.1, indicating successful mitigation of observable confounders. Subsequent outcome comparisons were conducted on this matched cohort (n = 2,886 total patients). Ninety-Day Outcomes Within 90 days after surgery, patients who received intraoperative ketorolac filled fewer opioid prescriptions on average than those who did not. The mean number of opioid prescription fills in the ketorolac group was 1.20 ± 0.44, compared to 1.43 ± 1.26 in the no-ketorolac group, suggesting a small but statistically significant reduction in opioid requirements when ketorolac was used (p = 0.039) (Table 2). Notably, the incidence of opioid dependence diagnoses within 90 days was low in both cohorts, but was higher in the ketorolac group (1.0% vs. 0.3% in the no-ketorolac group). While rare in absolute terms, this difference was statistically significant (RR, 4.00; 95% CI, 2.34–6.84; p < 0.001) (Table 2). Early postoperative complications within 90 days were uncommon and did not significantly differ between the 2 groups (Table 2). The rate of muscle atrophy diagnoses and wound complications was statistically identical at 0.8% in both groups. We additionally examined other 90-day adverse events and healthcare utilization as safety endpoints. Anemia requiring care occurred in 0.7% of patients in both groups (no difference, p = 1.0) (Table 3). No cases of postoperative myocardial infarction (MI), stroke, or pulmonary embolism were recorded in either group within 90 days (0% incidence) (Table 3). Although there were slightly fewer deep vein thromboses in the ketorolac group (0% vs. 0.7%), this difference was not statistically significant (p = 0.31). There was no difference in 90-day hospital readmissions (0.7% vs. 0.7%) or postoperative pneumonia (0% in both). Emergency department (ED) visits within 90 days occurred in 1.7% of ketorolac patients and 2.0% of non-ketorolac patients, a difference that was not significant (RR, 0.85; 95% CI, 0.49–1.48; p = 0.57). Two-Year Outcomes Longer-term outcomes were assessed in 1,443 matched patients per group, with a median follow-up of approximately 2 years. The incidence of most complications by 2 years was low and did not differ significantly between the ketorolac-exposed and unexposed cohorts (Table 4). Mechanical complications related to the ACLR (e.g., hardware or graft failure aside from graft rupture) were infrequent, occurring in 0.7% of patients in each group (RR, 1.00; p = 1.00). Deep graft or hardware infections were also rare, noted in 0.7% of patients in both groups (RR, 1.00; p = 1.00). Knee stiffness within 2 years was coded in 12.5% of ketorolac patients versus 10.9% of patients without ketorolac, a difference that was not statistically significant (RR, 1.15; 95% CI, 0.94–1.40; p = 0.182). Likewise, MUA for postoperative knee stiffness was very uncommon in both cohorts (0.8% vs. 1.0%), with no significant difference between groups (p = 0.56). In contrast, the rate of ACL revision surgery was significantly higher in the ketorolac group by 2 years postoperatively. A repeat ACLR was performed in 6.0% of patients who received intraoperative ketorolac, compared to 2.6% of patients in the no-ketorolac group (RR, 2.29; 95% CI, 1.58–3.33; p < 0.001) (Table 4). DISCUSSION In this large retrospective cohort of pediatric ACLR patients, we found that intraoperative ketorolac administration was associated with both potential risks and benefits. The key findings emphasize ketorolac’s role as part of a multimodal analgesia approach, as evidenced by its modest reduction in opioid prescription utilization postoperatively without increasing short-term complication rates, and a higher rate of subsequent ACLR within 2 years. This is one of the first studies to specifically examine the impact of intraoperative NSAID use on pediatric ACLR outcomes on a large scale. These results provide new insight into the balance of analgesic benefits versus healing implications of ketorolac in a young athletic population. Our findings support the role of ketorolac in improving early postoperative pain management in pediatric ACLR. Patients who received intraoperative ketorolac required significantly fewer opioid prescriptions in the 3 months following surgery. Although the absolute difference was small (~0.2 fewer prescriptions on average), and baseline opioid utilization was low in both groups, this modest reduction may still carry some clinical relevance in the context of opioid stewardship efforts in adolescents. The reduction in opioid use aligns with prior pediatric studies supporting the opioid-sparing effects of ketorolac. For instance, Nelson et al.16) reported that adding ketorolac in children undergoing orthopedic surgery (supracondylar humerus pinning) led to lower pain scores and a reduction in opioid doses during hospitalization. Similarly, Adams et al.8) found that multimodal analgesia including ketorolac significantly decreased postoperative morphine requirements and length of stay in pediatric fracture cases, without increasing short-term complications. Our study extends these observations to pediatric ACLR, suggesting that intraoperative ketorolac may confer sufficient analgesic benefits to decrease the need for opioids during recovery. Given the current emphasis on opioid stewardship and the risks for opioid exposure in adolescents, any safe reduction in opioid requirements is clinically meaningful. Our data did not indicate ketorolac compromises pain control to a degree that increased other healthcare encounters, supported by the ED visit rate not being elevated. Thus, the analgesia benefit appears to come without a trade-off in unmanaged pain. Our data also found that intraoperative ketorolac did not lead to higher rates of acute postoperative complications. Concerns about NSAIDs causing bleeding or wound-healing problems were not observed. The ketorolac group also had identical rates of wound dehiscence and superficial infection as the control group (each ~0.8%). This complements previous reports of ketorolac not significantly increasing surgical site bleeding or infection risk in orthopedic procedures.817) In pediatric fracture surgery and spinal fusions, studies have shown no difference in transfusion requirements or wound complications with perioperative ketorolac use.17) Our analysis similarly detected no sign of increased hemorrhage, even though ketorolac can transiently affect platelet function. These findings should reassure clinicians that a single intraoperative dose of ketorolac is safe from a short-term surgical standpoint in pediatric ACLR. This is consistent with a growing body of literature in pediatric orthopedics indicating that ketorolac is a safe adjunct for pain control across multiple settings.171819) A prior review noted that while NSAIDs have known side effects (renal impairment and gastrointestinal irritation), short courses or single doses in healthy children are seldom associated with serious adverse events.7) Our real-world data did not reveal any signal of acute kidney injury or other systemic complications (no difference in rate of acute MI, stroke, or renal events), affirming the low risk of serious adverse events accompanying ketorolac usage. One unexpected result was the slightly higher incidence of opioid dependence diagnosis in the ketorolac group compared to controls (1.0% vs. 0.3%); however, the absolute incidence was very low in both groups and should be interpreted with caution. The database relied on coding of an "opioid dependence" ICD-10 diagnosis, which may not reliably capture true opioid use disorder, especially early after surgery. Other studies have not suggested that ketorolac would increase the risk of persistent opioid use, with some research suggesting ketorolac decreases persistent opioid use.15) Therefore, we do not believe our data indicates a causal relationship whereby ketorolac predisposes adolescents to opioid dependence. This finding is more likely a statistical anomaly, or related to patient-specific factors not fully captured in matching. This outcome should be considered exploratory given the rarity of events and the imprecision of administrative coding in capturing true opioid use disorder, and should not be interpreted as a primary finding of this study. The most noteworthy finding of caution in our study was the higher rate of repeat ACLR in patients who received intraoperative ketorolac. By 2 years postoperatively, 6.0% of the ketorolac group had undergone a repeat ACLR procedure, compared to 2.6% of those who did not receive ketorolac. This finding must be interpreted in light of a constraint of the outcome definition: a repeat CPT 29888 event captures both ipsilateral graft revision and contralateral primary ACLR, which our data cannot distinguish. DeFrancesco et al.20) reported a 4-year cumulative contralateral ACL rupture rate of 12% in 996 pediatric patients, alongside an ipsilateral graft rupture rate of 19.7%. Grassi et al.,21) in long-term follow-up of patients younger than 18 at index surgery, found that contralateral reconstructions came to exceed ipsilateral revisions by a factor of 2.4 to 3.6 beyond 6 years. The observed association between ketorolac and a higher rate of repeat ACLR could therefore reflect impaired ipsilateral graft incorporation, an increased contralateral injury rate, or confounding unrelated to either mechanism. A mechanistic interpretation framed exclusively in terms of graft-healing biology is stronger than what our outcome definition can support. One important distinction is that our study specifically involves a pediatric population. Younger patients have higher baseline graft failure rates than adults, often related to returning to high-risk sports and possible biologic and lifestyle factors, which could make any deleterious effect of NSAIDs easier to detect. Thus, if an NSAID does slightly increase graft failure risk, that effect might only become evident in a large sample of high-risk individuals such as adolescent athletes. We attempted to match on known confounders, but certain factors were not available in the database. For example, graft choice (allograft vs. autograft) is a known determinant of ACL revision risk in adolescents. Previous research has found allografts are associated with higher failure rates in young patients.2) Further, differences in postoperative rehabilitation aggressiveness or return-to-sport timing could influence retear risk and might coincidentally correlate with ketorolac use (for instance, institutions that use ketorolac routinely might also encourage earlier return to sport). Our dataset lacked granular details on rehabilitation protocols or graft type, making it difficult to fully account for these factors. Furthermore, it is biologically unclear how a single intraoperative dose of ketorolac could plausibly drive graft failure outcomes observed up to 2 years postoperatively. This temporal disconnect raises the strong possibility that unmeasured confounders, such as graft type, surgical technique, institution-level practices, or patient activity level, may account for the observed difference in revision rates, rather than ketorolac exposure itself. Moreover, ketorolac administration may reflect institutional multimodal analgesia practices rather than an isolated exposure, meaning that unmeasured differences in perioperative care between institutions could contribute to the observed association. The biologic plausibility of an NSAID effect on graft survival is best evaluated with consideration of prior studies evaluating the effects of NSAIDs on ACL graft healing, tendon-to-bone healing, ligamentization, and soft-tissue incorporation. Preclinical evidence has long indicated that NSAIDs, through COX inhibition, can impair the early inflammatory phase of healing necessary for tissue repair.22) Tendon graft incorporation following ACLR proceeds through vascular ingrowth at the graft-tunnel interface, mineralization, and the development of Sharpey-like fibers between the graft and bone tunnel walls.23) Dimmen et al.24) studied tendon healing in a tibial bone tunnel in 60 rats and reported that 7 days of either parecoxib (COX-2 selective) or indomethacin (nonselective) significantly reduced pull-out strength, energy absorption, and stiffness at 14 days, with the largest effect in the parecoxib group. In a rabbit ACLR model, Sauerschnig et al.25) found that postoperative selective COX-2 inhibition significantly lowered bone density and new bone formation at the tunnel aperture at 3 weeks and produced a 37% reduction in graft mechanical stability at 6 weeks compared with controls. Bektaş et al.26) reported similar effects in a rat rotator cuff repair model: both selective (celecoxib) and nonselective (naproxen) NSAID exposure produced significantly worse histologic tendon-to-bone healing than acetaminophen, though biomechanical differences did not reach significance at 28 days. Ligamentization, the intra-articular remodeling phase of graft maturation, is a process distinct from tunnel healing and has not been directly studied under NSAID exposure. The process depends in part on prostaglandin E2-mediated macrophage and fibroblast activity, which COX-2 inhibition suppresses; whether this translates to a clinically meaningful effect at standard NSAID doses is unknown.27) Prior clinical work on NSAIDs after ACLR has been conducted in adult populations and has not demonstrated increased graft failure.1213) No study specifically evaluating NSAID exposure against ACLR outcomes in a pediatric population existed prior to ours. Pediatric patients carry higher baseline rates of both ipsilateral graft failure and contralateral injury and have an immature musculoskeletal system, which may make them more biologically vulnerable to NSAID effects on tissue healing than the adult populations in which prior clinical work has been conducted. Pereira et al.'s observation28) that postoperative ketorolac was associated with an approximately 2- to 3-fold increase in long-bone fracture nonunion in pediatric patients has been cited in commentary on pediatric NSAID exposure; however, endochondral fracture repair and ACL graft incorporation are biologically distinct, and that finding should not be invoked as direct mechanistic support for ours. Definitive evidence for an effect of single-dose intraoperative ketorolac on ACL graft healing in humans does not exist. Clinicians need to weigh the therapeutic advantages of ketorolac against potential risks to healing in this young population. For pediatric ACLR, an individualized approach is warranted. Currently, there are no formal guidelines on NSAID use specifically in pediatric ACL surgery. Our findings contribute pediatric-specific data to inform such practice patterns. Several limitations of this study should be acknowledged. First, the retrospective design and use of an administrative database introduce inherent biases. We relied on coding (ICD/CPT) to define exposures and outcomes, which may be subject to misclassification. Our outcomes were based on diagnosis codes, which might vary in how clinicians apply them. In particular, ACL revision surgery was defined by repeat occurrence of CPT code 29888, which does not distinguish between ipsilateral revision and contralateral primary reconstruction, nor does it confirm that the repeat procedure was performed for graft failure specifically. This may introduce outcome misclassification bias and could lead to an overestimation of the true revision rate. Additionally, ketorolac exposure was defined solely by the presence of an intraoperative IV/IM administration record, without detail on timing of administration, weight-based dosing, or whether its use reflects a broader institutional multimodal analgesia protocol. As such, ketorolac may serve as a surrogate marker for wider perioperative care patterns rather than an isolated pharmacologic exposure, which limits causal inference. Second, despite propensity matching, there may be residual confounding. Important factors such as graft type, surgical technique (for example, physeal-sparing vs. transphyseal), concomitant meniscal procedures, return-to-sport protocols, and other intraoperative variables were not available in the dataset. Pediatric ACLR is performed using a range of surgical techniques, including extra-articular physeal-sparing reconstruction, all-epiphyseal techniques, partial transphyseal techniques, and complete transphyseal reconstruction, with selection driven by skeletal maturity. These techniques differ in reported graft failure rates, and surgical technique is therefore a meaningful source of variation in pediatric ACLR outcomes. CPT 29888 does not distinguish among these techniques in administrative data, and stratified analyses by surgical technique were therefore not feasible in this dataset. Third, the analysis of opioid usage was limited to prescription fill data and diagnosis of dependence, which can be imprecise measures of true consumption and addiction. Therefore, our opioid-related findings should be interpreted as general trends rather than exact measures of opioid consumption behavior. Future studies should prospectively evaluate the effect of perioperative NSAID use on ACL graft outcomes in pediatric patients, with attention to graft type, surgical technique, and rehabilitation protocols. Translational research examining the biological effect of ketorolac on graft incorporation, as well as comparative studies of alternative non-opioid analgesic strategies in this population, would further inform clinical practice. Lastly, patients without subsequent records were assumed to be event-free rather than formally censored at last contact, which may introduce informative censoring and outcome ascertainment bias if lost-to-follow-up patients differ systematically from those with continued records. Median follow-up duration per group was not available for this analysis, and neither a time-to-event analysis nor a sensitivity analysis restricted to patients with documented follow-up beyond 1 year was performed. Future studies should address this limitation through prospective designs with complete follow-up ascertainment. Despite these limitations, the study has several strengths. It leverages a very large multicenter sample, enhancing generalizability across different hospital settings and geographic regions. By using propensity score matching, we improved the comparability of the 2 cohorts on observable characteristics, and the balance achieved was excellent for key confounders. The follow-up duration of 2 years is typical of other pediatric ACL studies and covers the high-risk period for graft re-injury in pediatric patients.2930) Additionally, we examined a comprehensive set of outcomes, ranging from immediate surgical complications to longer-term reoperations and systemic safety events, providing a broad view of ketorolac’s impact. Finally, all data were obtained from real-world clinical practice, which can complement findings from smaller prospective studies or trials. In summary, this retrospective analysis of over 2,800 matched pediatric ACLR cases found that intraoperative ketorolac was associated with reduced postoperative opioid requirements and no increase in early postoperative complications, supporting its short-term safety and utility as part of a multimodal analgesia strategy. However, we observed a statistically significant association between ketorolac use and a higher rate of ACL revision surgery at 2 years, which should be interpreted as an observed association only and does not establish causation. The absolute risk increase was modest, and unmeasured confounders may account for this finding.