Discussion

In this single-centre retrospective cohort of paediatric catheterization laboratory patients, we observed three main findings. First, general anaesthesia and sedation were used in nearly equal proportions overall, but anaesthesia strategy differed markedly by procedure class: interventional procedures were predominantly performed under general anaesthesia, whereas diagnostic procedures were largely performed under sedation. Second, complications were recorded in 16.0% of patients, and the unadjusted complication rate was higher under general anaesthesia than under sedation. Third, when we addressed confounding by indication through stratification and multivariable adjustment within diagnostic catheterizations, general anaesthesia remained associated with higher odds of any recorded complication. These observations support the concept that comparisons between anaesthesia strategies in the catheterization laboratory should account for procedure type and baseline differences, rather than relying solely on pooled comparisons. ^{Reference Feltes, Bacha and Beekman2,Reference Tokel, Gümüş, Ayabakan, Varan and Erdoğan7,Reference Ogawa, Yamana and Noda8}

The overall pattern of anaesthesia use in our centre aligns with prior reports that sedation is widely utilized in paediatric catheterization laboratories, with substantial between-centre variation driven by institutional workflow, staffing, and case mix. ^{Reference Özcanoğlu and Özcan5} In our cohort, the strong linkage between general anaesthesia and interventional procedures is clinically intuitive: interventions such as device closure, balloon-based therapies, stent implantation, and valve implantation generally demand immobility, stable operating conditions, and reliable airway control. This tendency is also reflected in procedural anaesthesia reports for transcatheter closure in children, where general anaesthesia is commonly preferred for procedural conditions and airway/ventilation management. ^{Reference Karagöz, Ankay-Yılbaş, Kanbak, Tümer, Özkutlu and Karagöz4} As a result, crude comparisons of complications between anaesthesia types can be misleading because the anaesthesia choice often reflects the anticipated complexity and risk profile of the patient. ^{Reference Feltes, Bacha and Beekman2,Reference Tokel, Gümüş, Ayabakan, Varan and Erdoğan7} The greater use of general anaesthesia in older/heavier children in our cohort likely reflects institutional workflow and procedure mix (interventional procedures requiring immobility and often controlled ventilation), rather than preferential avoidance of sedation in infants. The greater use of sedation among patients with prior cardiac surgery in our cohort likely reflects follow-up diagnostic catheterization/angiography in this subgroup, whereas surgery-naïve patients more often undergo index interventional procedures, which were predominantly managed under general anaesthesia.

A recent multicentre registry analysis by Chenault et al. reported substantially lower rates of anaesthesia-related adverse events (0.8% overall; 0.4% high severity) in congenital cardiac catheterization and found that general anaesthesia was not associated with increased odds of high-severity anaesthesia-related adverse events after accounting for procedural complexity using the PREDIC3T risk category. ^{Reference Chenault, Rahman and Gauvreau9,Reference Quinn, Yeh and Gauvreau10} Our findings are not directly comparable because our primary endpoint was a broad composite of any recorded complication in retrospective clinical documentation, including both anaesthesia-related and procedure-related events, and we counted events when they were explicitly documented and/or required a clinical intervention. Differences in outcome definitions and capture, along with residual confounding by indication and case mix, may therefore explain why we observed higher overall event rates and an association with general anaesthesia in the diagnostic subgroup. In addition, because complications were abstracted from anaesthesia records as well as catheterization and post-procedure documentation, transient anaesthesia-related respiratory events such as bronchospasm may have been captured more completely than in catheterization registry-based studies, in which procedural adverse events may be preferentially recorded.

Respiratory events were prominent in the general anaesthesia group, particularly bronchospasm, which occurred only under general anaesthesia in our cohort. This is plausible given that airway instrumentation, airway irritation, and positive-pressure ventilation can increase the risk of bronchospasm and related respiratory events in susceptible children. Conversely, hypotension and desaturation occurred in both groups, with desaturation numerically more frequent under sedation in our cohort, emphasizing that sedation is not risk-free and requires vigilant monitoring and readiness to escalate support. Contemporary observational studies have similarly suggested that sedation may be associated with fewer high-severity adverse events than general anaesthesia in selected paediatric catheterization populations, although these studies also highlight the importance of careful patient selection and institutional expertise. ^{Reference Mikus, Welchowski, Schindler, Schneider, Mini and Vergnat6,Reference Ogawa, Yamana and Noda8}

A key methodological issue in this clinical question is confounding by indication. Since interventional cases were overwhelmingly managed with general anaesthesia, and only a small number of interventional procedures were performed under sedation, it is difficult to isolate the independent contribution of anaesthesia type to outcomes in the interventional setting. We therefore separated diagnostic and interventional procedures and, in the interventional group, used Firth penalized regression because the sedation subgroup had zero recorded events. ^{Reference Firth11,Reference Heinze and Schemper12} The resulting estimate was highly imprecise and should be interpreted as indicating limited information rather than evidence for or against a true association. This limitation is consistent with the broader literature in which adverse events are influenced by both procedural complexity and patient physiology, and robust comparisons require larger, more balanced samples or designs that better address selection effects. ^{Reference Feltes, Bacha and Beekman2,Reference Tokel, Gümüş, Ayabakan, Varan and Erdoğan7,Reference Ogawa, Yamana and Noda8}

From a clinical standpoint, our findings suggest two practical implications. First, diagnostic catheterizations performed under sedation in an experienced, well-monitored catheterization laboratory environment may be associated with a lower complication rate than comparable diagnostic procedures under general anaesthesia, even after adjustment for measured factors. Second, general anaesthesia remains central for many interventional procedures, but its benefits must be weighed against potential airway and respiratory events, underscoring the importance of pre-procedural planning, airway risk assessment, and complication preparedness in these cases. ^{Reference Feltes, Bacha and Beekman2,Reference Karagöz, Ankay-Yılbaş, Kanbak, Tümer, Özkutlu and Karagöz4} Rather than advocating a single default technique, our data support an individualized approach in which anaesthesia strategy is tailored to procedure type, anticipated duration, and patient risk profile. ^{Reference Feltes, Bacha and Beekman2}

This study has several limitations. The retrospective design limits causal inference and depends on the accuracy and completeness of documentation; minor events may be under-recorded, and the severity of complications could not be graded using a standardized adverse event taxonomy. Practice patterns and outcomes may differ in other centres with different staffing and sedation governance. Syndromic conditions, including Down syndrome, were present in a minority of encounters and may influence peri-procedural risk and anaesthetic planning. ^{Reference Delany, Gaydos and Romeo13} Airway management strategy (endotracheal tube versus supraglottic airway) may influence anaesthetic depth requirements, haemodynamics, and respiratory events, and could confound associations between anaesthetic technique and complications. Because the airway device was not consistently documented in a retrievable, structured format in this retrospective dataset, we could not evaluate device-specific risk. Finally, residual confounding is likely despite stratification and adjustment, and the interventional sedation subgroup was small, limiting the ability to draw firm conclusions for interventional procedures.

In conclusion, in this single-centre cohort of paediatric catheterization laboratory patients, the most robust comparative finding was within diagnostic catheterization/angiography, where general anaesthesia was associated with a higher unadjusted composite complication rate than sedation, and this association persisted within diagnostic procedures after adjustment for measured confounders. Comparisons in interventional procedures were limited by the small sedation subgroup and should be interpreted as descriptive rather than definitive. Future work with larger multicentre cohorts, standardized adverse event definitions, and risk-adjusted designs will be helpful to better define which patients and procedures derive the greatest net benefit from sedation versus general anaesthesia in the paediatric catheterization laboratory. ^{Reference Feltes, Bacha and Beekman2,Reference Tokel, Gümüş, Ayabakan, Varan and Erdoğan7,Reference Ogawa, Yamana and Noda8}