Excess Bed Days and Hospitalization Costs Associated With 30‐Day Complications Following Catheter Ablation of Atrial Fibrillation Authors: Linh Ngo, Russell Denman, Karen Hay, Billingsley Kaambwa, Anand Ganesan, Isuru Ranasinghe Published: 2023 ABSTRACT BackgroundThe incidence and type of complications following catheter ablation of atrial fibrillation have been extensively examined, but the impact associated with these complications on the length of stay and hospitalization costs is unknown.Methods and ResultsThis cohort study included 20 117 adult patients (mean age 62.6±11.4 years, 30.3% women, median length of stay 1 day [interquartile range 1–2 days]) undergoing atrial fibrillation ablation in financial years 2011 to 2017 in Australia with available cost data from the National Hospital Cost Data Collection, which determines government reimbursement of health services provided. The primary outcome was the costs associated with complications occurring up to 30 days postdischarge adjusted for inflation to 2021 Australian dollars. We used generalized linear models to estimate the increase in length of stay and cost associated with complications, adjusting for patient characteristics. Within 30 days of hospital discharge, 1151 (5.72%) patients experienced a complication with bleeding (3.35%) and pericardial effusion (0.75%) being the most common. On average, the occurrence of a complication was associated with an adjusted 3.3 (95% CI, 3.1–3.6) excess bed days of hospital care (totaling 3851 days), and a $7812 (95% CI, $6754–$8870) increase in hospitalization cost (totaling $9.0 million). Most of the total excess cost was attributable to bleeding ($3.8 million, 41.9% of total excess cost) and pericardial effusion ($1.6 million, 18.2%).ConclusionsComplications following atrial fibrillation ablation were associated with significant increase in length of stay and hospitalization costs, most of which were attributable to bleeding and pericardial effusion. Strategies to improve procedural safety and reduce health care costs should focus on these complications. Nonstandard Abbreviation and Acronym intracardiac echocardiography Clinical PerspectiveWhat Is New? Although the incidence of complications after atrial fibrillation ablation is well described, the cost and resource burden on hospitals is poorly understood.We found that complications following catheter ablation were associated with an average 3.3 extra days in hospital and an average excess hospitalization cost of $7812 for each patient who experienced a complication.Most of the excess cost was attributable to bleeding and pericardial effusion. What Are the Clinical Implications? Efforts to improve care may benefit from focusing on the most common and most costly events like bleeding or pericardial effusion.Future cost‐effectiveness analysis should consider both the incidence and cost of procedural complications, and our study provides insightful data that could be used as inputs for such models. Clinical Perspective What Is New? Although the incidence of complications after atrial fibrillation ablation is well described, the cost and resource burden on hospitals is poorly understood.We found that complications following catheter ablation were associated with an average 3.3 extra days in hospital and an average excess hospitalization cost of $7812 for each patient who experienced a complication.Most of the excess cost was attributable to bleeding and pericardial effusion. Although the incidence of complications after atrial fibrillation ablation is well described, the cost and resource burden on hospitals is poorly understood. Although the incidence of complications after atrial fibrillation ablation is well described, the cost and resource burden on hospitals is poorly understood. We found that complications following catheter ablation were associated with an average 3.3 extra days in hospital and an average excess hospitalization cost of $7812 for each patient who experienced a complication. We found that complications following catheter ablation were associated with an average 3.3 extra days in hospital and an average excess hospitalization cost of $7812 for each patient who experienced a complication. Most of the excess cost was attributable to bleeding and pericardial effusion. Most of the excess cost was attributable to bleeding and pericardial effusion. What Are the Clinical Implications? Efforts to improve care may benefit from focusing on the most common and most costly events like bleeding or pericardial effusion.Future cost‐effectiveness analysis should consider both the incidence and cost of procedural complications, and our study provides insightful data that could be used as inputs for such models. Efforts to improve care may benefit from focusing on the most common and most costly events like bleeding or pericardial effusion. Efforts to improve care may benefit from focusing on the most common and most costly events like bleeding or pericardial effusion. Future cost‐effectiveness analysis should consider both the incidence and cost of procedural complications, and our study provides insightful data that could be used as inputs for such models. Future cost‐effectiveness analysis should consider both the incidence and cost of procedural complications, and our study provides insightful data that could be used as inputs for such models. Catheter ablation is a well‐established therapeutic option for the treatment of atrial fibrillation (AF). 1 This procedure has been shown to be superior to medical therapy in restoring normal sinus rhythm with a potential survival benefit in patients with comorbid heart failure. 1 , 2 Despite these well‐proven benefits, catheter ablation is associated with a 2.9% to 7.2% risk of early complications, including serious adverse events such as stroke, pericardial effusion, and major bleeding. 3 , 4 , 5 , 6 Understanding and minimizing these procedural complications are essential to improve procedural safety and patient outcomes. Although numerous studies have examined the incidence and types of complications after AF ablation, 3 , 4 , 5 , 6 to our knowledge, the impact of complications on the health care system has never been examined. This information is critical for hospitals and policymakers to establish strategic approaches to improve care quality and reduce patient harm by allocating resources to reduce the more costly complications. Such information is also essential to better evaluate the cost‐effectiveness of AF ablation given that most prior models account for the incidence of complications but do not consider their costs. 7 , 8 , 9 Despite these considerable needs, no data are currently available about the excess in length of hospital stay and hospitalization costs attributable to these complications. Although cost‐analysis of procedural complications has been performed for other procedures such as implantable cardiac defibrillators implantation, 10 , 11 , 12 a similar analysis has not been conducted for AF ablation. In this study, we estimated the burden on the health care system associated with 30‐day complications following AF ablations using national data from Australia during the financial years of 2011 to 2017. Specifically, we quantified the number of additional days in hospital (excess bed days) and the excess in hospitalization cost associated with the occurrence of 1 or more procedural complications. Methods The data underlying this article were provided by a third party including the Data Custodian Units of each state and territory in Australia under ethics approval. Data may be accessed upon request to the third party. Data Source Admitted Patient Collection We used hospitalization data recorded in the Admitted Patient Collection, which includes all inpatient and same‐day admissions from all states and territories, to identify patients undergoing catheter ablation of AF in Australia. Data are not released to researchers from private hospitals in South Australia, Northern Territory, and Tasmania, although collectively these regions account for <10% of the Australian population. 13 A standardized set of variables is recorded for each hospital encounter, including patient demographic characteristics, primary and up to 50 secondary diagnoses coded using the International Classification of Diseases, Tenth Revision, Australian Modification (ICD‐10‐AM); up to 50 procedures coded using the Australian Classification of Health Interventions; and the patient's status at discharge. The overall accuracy of diagnosis and procedure coding exceeds 85% when compared with medical records 14 with high accuracy also being reported for the coding of AF in administrative data. 15 National Hospital Cost Data Collection The Independent Health and Aged Care Pricing Authority collects the cost of hospitalizations from a sample of >400 public and private sector hospitals and determines the average cost for each hospital admission by grouping patients with similar conditions and resources consumption into 1 Australian refined diagnosis‐related group (DRG). 16 The average cost per DRG for the private sector is not published. Instead, a cost weight ratio is published, which indicates the relative ratio of cost for the same DRG in the private sector compared with the public sector. The average cost in the private sector can then be estimated by multiplying the cost weight ratio by the average cost for the same DRG in the public sector in the same fiscal year. Study Cohort We have previously described methods to identify AF ablation using coded hospitalization data. 17 , 18 In short, we included Australian residents aged >18 years who were hospitalized with a primary diagnosis of AF and a procedure code for catheter ablation from July 1, 2011, to June 30, 2017. We excluded patients who (1) had secondary diagnoses of other arrhythmias (to ensure that the ablation was solely for AF); (2) had a procedure code indicating a cardiac device implantation (pacemaker or implantable cardioverter‐defibrillator) or a diagnosis code indicating the presence of a cardiac device (to avoid including patients undergoing atrioventricular nodal ablation for AF rate control); (3) had a procedure code for open ablation; (4) were discharged against medical advice as adverse events may reflect patient's choice rather than care quality; and (5) had nonelective (ie, unplanned) hospitalizations to avoid including adverse events and costs due to the acute illness. Table S1 lists all diagnoses and procedure codes used for patient selection. Identifying AF ablation using a primary diagnosis of AF combined with a procedure code of catheter ablation has previously been shown to have 100% specificity and 87.3% sensitivity. 19 Procedure‐Related Complications Any procedural complications occurring during the hospital stay or up to 30 days postdischarge were identified using definitions we have previously published. 17 , 18 Specifically, procedural complications included (1) death; (2) cardiorespiratory failure and shock; (3) stroke or transient ischemic attack; (4) pericardial effusion; (5) hemothorax/pneumothorax; (6) any bleeding (including postprocedural hemorrhage or hematoma formation, bleeding from other sites [gastrointestinal, respiratory and urinary tract], and bleeding requiring blood transfusion); (7) vascular injury (including artery aneurysm, arteriovenous fistula, or injury that required intervention or surgery); (8) infection (sepsis, pneumonia, or endocarditis); (9) pericarditis; (10) acute myocardial infarction; (11) venous thromboembolism (pulmonary embolism or deep vein thrombosis); (12) acute kidney injury; (13) complications requiring cardiac surgery; and (14) complete atrioventricular block. We used the secondary diagnoses and procedure codes of the index hospitalizations to identify complications occurring during the hospital stay. Postdischarge complications were defined as deaths occurring after the patient left the hospital or any readmission with 1 of the aforementioned complications coded as the primary diagnosis. Postdischarge outcomes were captured using the linkages between health records and registries of deaths established by the designated Data Linkage Unit in each state and territory with a reported accuracy of >99%. 20 Table S2 outlines all diagnoses and procedure codes used to define these complications. Length of Hospital Stay and Hospitalization Cost The total length of stay (LOS) for each patient undergoing AF ablation was estimated by summing the index LOS (the LOS of the index hospitalization) and postdischarge LOS (the total LOS of 1 or more rehospitalizations for complications). Postdischarge LOS was considered 0 if the patient did not experience any complications that required rehospitalization. The total hospitalization cost related to AF ablation for each patient was the sum of in‐hospital cost (the cost of the index hospitalization) and postdischarge cost (the cost of 1 or more readmissions due to procedural complications). No postdischarge cost was assigned for patients who were not rehospitalized for complications after discharge. The cost of each hospitalization was estimated by merging the Admitted Patient Collection and National Hospital Cost Data Collection data sets based on the DRG code, financial year, and hospital sector. Data were not available for the private sector in the financial year of 2014 to 2015 and were estimated by averaging the cost of the same DRG code in the previous and subsequent fiscal years. The estimated cost for the DRG codes F24A and F24B was available only in the financial year of 2017 to 2018, which was used to estimate the cost associated with these DRG codes in previous years by using the Consumer Pricing Index. 21 Costs could not be estimated for records with missing or unmatched DRG and these records were excluded. All costs were then adjusted for inflation to 2021 Australian Dollar using Consumer Pricing Index. 21 Statistical Analysis We presented data as frequencies and percentages for categorical variables and mean±SD or median and interquartile range for continuous variables. Differences between groups were compared using χ2 or Fisher's exact test for categorical variables and Student's t test or Mann–Whitney U test for continuous variables. To calculate the incidence of complications, patients who experienced multiple events were counted only once. For estimation of costs, multiple postdischarge complications for the same patient were included. LOS and cost are skewed data and therefore, we used generalized linear models to estimate the excess bed days and excess cost associated with complications adjusting for differences in baseline characteristics between patients with and without complications. 22 , 23 , 24 , 25 , 26 , 27 Specifically, we used multivariable negative binomial regression for LOS (count data) and a generalized linear model with a gamma distribution and log link for cost (continuous data) to estimate the average marginal effect associated with complications on the respective outcomes per patient experiencing a complication. 22 , 23 Candidate variables considered for these models included patient's age and sex, history of ablation in the previous year, ablation of both atria, ablation at a private hospital, presenting region, and a range of cardiac and noncardiac comorbidities. We extracted comorbidities by using the Condition Categories system, which groups all secondary diagnoses of the index admissions with all diagnoses of prior hospitalizations in the preceding year into 180 meaningful conditions. 22 Variables eligible for the multivariable models were those that had a rate ≥0.5% (for binary variables) and a P value ≤0.250 in univariate analysis (for all variables). Backward elimination was then applied until only variables with a P value <0.05 remained in the model. Results are reported as incidence rate ratio (and corresponding 95% CI) for LOS or cost ratio (and 95% CI) for hospitalization cost. The incidence rate ratio and cost ratios reflect the ratio between the LOS or cost of a patient with complications and that of a patient with a similar profile who did not experience complications. Separate models were built for in‐hospital and total LOS, as well as in‐hospital and total hospitalization cost (refer to Tables S3 through S6 for more details). The total excess bed days and excess cost were the product of average excess bed days and cost per patient experiencing a complication multiplied by the number of patients having complications. For postdischarge complications, we estimated the average excess bed days and hospitalization cost by dividing the total LOS and cost of 1 or more rehospitalization for complications by the number of patients rehospitalized for complications. All analyses were performed using Stata version 16.0, and a 2‐tailed P value <0.05 was considered statistically significant. Ethical approval was granted by respective Human Research Ethics Committees of each state and territory in Australia. A waiver of informed consent was provided for the use of deidentified data. Results We identified 28 791 hospitalizations with a primary diagnosis of AF and a procedure code of catheter ablation meeting inclusion criteria. Of these, 8614 were excluded, mostly due to having current or past cardiac device implantation (n=3678) or lacking or having an unmatched DRG code (n=2355). The final study cohort included 20 117 patients (Figure 1). Patient selection flow diagram. DRG indicates diagnosis‐related group. Study Cohort The mean age was 62.6 years with 46.2% aged ≥65 years (Table 1). More than two‐thirds of patients (69.7%) were men. The median LOS for an AF ablation was 1 day (interquartile range 1–2 days), and the median hospitalization cost was $6907 (interquartile range $6079–$9336). Most (72.0%) AF ablations were performed in private sector hospitals. Rates of comorbidities were generally low, with hypertension (11.6%), coronary artery disease (9.5%), and diabetes (10.5%) being the most common. Table 1Baseline Characteristics of the Study CohortVariablesOverall (N=20 117) n (%)No complications (N=18 966) n (%)With complications (N=1151) n (%) P valuePatient demographicsAge, y (mean±SD)62.6±11.462.5±11.464.5±10.8<0.01Age group, y18–34385 (1.9)372 (2.0)13 (1.1)<0.0135–492106 (10.5)2012 (10.6)94 (8.2)50–648342 (41.5)7908 (41.7)434 (37.7)65–798260 (41.1)7728 (40.8)432 (46.2)≥ 801024 (5.1)946 (5.0)78 (6.8)Female sex6092 (30.3)5659 (29.8)433 (37.6)<0.01Median length of stay (IQR), d1 (1–2)1 (1–2)2 (1–4)<0.01Median hospitalization cost (IQR) (in 2021 Australian dollars)$6907 (6079–9336)$6907 (6079–9327)$13 020 (7232–16 165)<0.01RegionAustralian Capital Territory/New South Wales4210 (20.9)3946 (20.8)264 (22.9)<0.01Queensland5027 (25.0)4831 (25.5)196 (17.0)South Australia1400 (7.0)1279 (6.7)121 (10.5)Victoria5907 (29.4)5597 (29.5)310 (26.90Western Australia3573 (17.8)3313 (17.5)260 (22.6)Private hospital14 486 (72.0)13 790 (72.7)696 (60.5)<0.01CHA2DS2‐VASc score* 1 (0–2)1 (0–1)1 (0–2)<0.01Cardiovascular historyHypertension2333 (11.6)2131 (11.2)202 (17.6)<0.01Heart failure1701 (8.5)1579 (8.3)122 (10.6)0.01Valvular and rheumatic heart disease714 (3.6)660 (3.5)54 (4.7)0.03Coronary artery disease1912 (9.5)1754 (9.3)158 (13.7)<0.01Vascular disease310 (1.5)287 (1.5)23 (2.0)0.20Prior atrial fibrillation hospitalizations12 710 (63.2)11 968 (63.1)742 (64.5)0.35Other comorbiditiesDiabetes mellitus2110 (10.5)1990 (10.5)120 (10.4)0.94Chronic lung diseases327 (1.6)292 (1.5)35 (3.0)<0.01Chronic kidney disease596 (3.0)535 (2.8)61 (5.3)<0.01Prior stroke/transient ischemic attack277 (1.4)261 (1.4)16 (1.4)0.97Hematological disorders877 (4.4)749 (4.0)128 (11.1)<0.01Pneumonia407 (2.0)338 (1.8)69 (6.0)<0.01Musculoskeletal and connective tissue disorders1490 (7.4)1364 (7.2)126 (11.0)<0.01Dementia34 (0.2)30 (0.2)4 (0.4)0.13Major cancer118 (0.6)101 (0.5)17 (1.5)<0.01End‐stage liver disease27 (0.1)25 (0.1)2 (0.2)0.67Drug or alcohol abuse, psychosis, or dependence298 (1.5)278 (1.5)20 (1.7)0.46Psychiatric disorders252 (1.3)234 (1.2)18 (1.6)0.33Neurologic disorders and paralysis207 (1.0)191 (1.0)16 (1.4)0.21Urinary tract disorders and incontinence824 (4.1)743 (3.9)81 (7.0)<0.01IQR indicates interquartile range.*CHA2DS2‐VASc score is a score used to evaluate risk of experiencing thromboembolic events of patients with atrial fibrillation in which a point each is given for the presence of congestive heart failure (C), hypertension (H), age 65–74 years old (A), diabetes (D), vascular disease (VASc) and female sex and 2 points each are given for age>=75 years old and history of stroke (S). The total score ranges from 0 to 9, with a higher score indicating a higher risk. 1 Baseline Characteristics of the Study Cohort IQR indicates interquartile range. CHA2DS2‐VASc score is a score used to evaluate risk of experiencing thromboembolic events of patients with atrial fibrillation in which a point each is given for the presence of congestive heart failure (C), hypertension (H), age 65–74 years old (A), diabetes (D), vascular disease (VASc) and female sex and 2 points each are given for age>=75 years old and history of stroke (S). The total score ranges from 0 to 9, with a higher score indicating a higher risk. 1 Complications Associated With Catheter Ablation of AF Overall, 1151 patients (5.72%) experienced 1223 complications (Table 2). Of these, 899 patients (4.47%) experienced at least 1 complication during the hospital stay. A further 300 patients (1.49%) either died or were rehospitalized for 334 complications within 30 days of discharge. Death (n=15, 0.07%), stroke or transient ischemic attack (n=48, 0.24%), or complications requiring cardiac surgery (n=20, 0.10%) occurred infrequently, whereas nearly 70% of all complications were attributable to bleeding (n=673, 3.35%) and pericardial effusion (n=151, 0.75%). Among patients with bleeding events, 525 (78.0% of all bleedings) experienced postprocedural hemorrhage or hematoma, 113 (16.8%) had bleeding from other sites (pulmonary, gastrointestinal, or urinary tract), and 101 (15.0%) required a blood transfusion. In patients with a pericardial effusion, 87 (57.6%) required a drainage. Table 2Incidence of Complications Following Catheter Ablation of AFComplicationsOverallIn‐hospitalPostdischargeNo. of patients n (%)No. of events* No. of patients n (%)No. of patients n (%)No. of events* Any complications1151 (5.72)1223899 (4.47)300 (1.49)334Death15 (0.07)154 (0.02)11 (0.05)11Cardiopulmonary failure and shock40 (0.20)4038 (0.19)2 (0.01)2Stroke/transient ischemic attack48 (0.24)5222 (0.11)26 (0.12)30Pericardial effusion151 (0.75)154135 (0.67)18 (0.09)19Pericardiocentesis87 (0.43)8776 (0.38)11 (0.05)11Hemothorax/pneumothorax29 (0.14)3021 (0.10)9 (0.04)9Bleeding673 (3.35)713551 (2.74)142 (0.71)162Postprocedural hemorrhage or hematoma525 (2.61)541464 (2.31)71 (0.35)77Bleeding from other sites113 (0.56)12084 (0.42)31 (0.15)36Bleeding requiring blood transfusion101 (0.50)11359 (0.29)47 (0.23)54Vascular injury92 (0.46)9769 (0.34)25 (0.12)28Postprocedural infection95 (0.47)10045 (0.22)50 (0.25)55Pericarditis57 (0.28)5846 (0.23)12 (0.06)12Procedure‐related acute myocardial infarction20 (0.10)205 (0.02)15 (0.07)15Venous thromboembolism14 (0.07)156 (0.03)8 (0.04)9Acute kidney injury52 (0.26)5248 (0.24)4 (0.02)4Complications requiring cardiac surgery20 (0.10)2013 (0.06)7 (0.03)7Complete atrioventricular block44 (0.22)4443 (0.21)1 (0.0)1*As one patient could experience multiple rehospitalizations for nonfatal events, the number of events might be greater than the number of patients. For patients with multiple complications during the index hospital stay, the cost of the index hospitalization was not changed and therefore, we did not report the number of in‐hospital events. Incidence of Complications Following Catheter Ablation of AF As one patient could experience multiple rehospitalizations for nonfatal events, the number of events might be greater than the number of patients. For patients with multiple complications during the index hospital stay, the cost of the index hospitalization was not changed and therefore, we did not report the number of in‐hospital events. Excess Bed Days Associated With Complications The occurrence of any complications during the hospital stay was associated with more than doubling of the average in‐hospital LOS (incidence rate ratio, 2.23 [95% CI, 2.13–2.33], Table S3), or 1.8 extra days in the hospital in absolute terms, after adjusting for differences in baseline characteristics between patients with and without complications. All individual complications were associated with significantly prolonged index LOS (Figure S1) except for complete heart block. Complications requiring cardiac surgery (5.7 days) and acute myocardial infarction (5.1 days) had the highest average increase per patient in the index LOS (Figure S2). When the complication occurred after discharge, it was associated with an average of 7.1 additional days in hospital, with those requiring cardiac surgery (35.0 days) and acute kidney injury (13.5 days) having the longest average postdischarge LOS (Figure S2). In total, procedural complications occurring within 30 days of discharge were associated with an almost 3‐fold increase in the total LOS (incidence rate ratio, 2.91 [95% CI, 2.77–3.04], Table S4 and Figure S3). In absolute terms, this equated to an average of 3.3 extra hospital days for each patient experiencing a complication (Figure 2), totaling 3851 excess bed days. Average increase in bed days associated with procedural complications per patient experiencing a complication within 30 days of discharge. CRF indicates cardio‐respiratory failure; LOS, length of stay; and TIA, transient ischemic attack. Excess Hospitalization Costs Associated With Complications The occurrence of an in‐hospital complication was associated with a 70% increase in the cost of the index hospitalization (cost ratio, 1.70 [95% CI, 1.61–1.79], Table S5). This corresponds to an average $5692 (95% CI, $4965–$6418) increase in cost per patient who experienced an in‐hospital complication, totaling $5.1 million (Table 3). Of the 300 patients experiencing postdischarge complications, 6 died in the community without a rehospitalization for complication. The remaining 294 patients experienced 328 readmissions for a complication, and the average cost per patient readmitted for a complication was $13 625, with a total cost of $4.0 million. Table 3Adjusted Excess in Hospitalization Costs Associated With Complications Occurring During the Hospital Stay, and Up to 30 Days PostdischargeIn‐hospital complicationsPostdischarge complications30‐day complicationsComplicationsAverage incremental cost per patient and 95% CI*Total cost of all complicationsAverage cost per patientTotal costAverage incremental cost per patient and 95% CITotal cost of all complicationsAny complications$5692 (4965 to 6418)$5 116 942$13 625$4 005 730$7812 (6754 to 8870)$8 991 515Death$22 262 (−6800 to 51 325)$89 049NANA$15 986 (418 to 31 554)$239 793Cardiopulmonary failure and shock$14 640 (6513 to 22 766)$556 313$26 052$52 104$15 030 (6409 to 23 651)$601 202Stroke/transient ischemic attack$9820 (5405 to 14 235)$216 047$14 484$391 081$15 179 (7272 to 23 086)$728 598Pericardial effusion$7086 (5127 to 9045)$956 597$23 791$428 241$10 862 (6142 to 15 583)$1 640 207Pericardiocentesis$9922 (6718 to 13 125)$754 039$34 161$375 772$12 364 (8272 to 16 456)$1 075 676Hemothorax/pneumothorax$20 905 (11 027 to 30 783)$439 008$23 437$210 934$24 496 (14 004 to 34 988)$710 379Bleeding$4250 (3390 to 5110)$2 341 670$9426$1 338 430$5591 (4511 to 6672)$3 763 057Postprocedural hemorrhage or hematoma$4127 (3190 to 5064)$1 914 862$5143$365 128$4606 (3606 to 5606)$2 418 031Bleeding from other sites$5843 (3523 to 8164)$490 827$5952$184 514$5753 (4158 to 7348)$650 104Bleeding requiring blood transfusion$13 238 (8965 to 17 510)$781 025$17 543$824 513$15 365 (10 463 to 20 266)$1 551 857Vascular injury$16 015 (11 348 to 20 681)$1 105 028$19 164$479 090$14 441 (10 756 to 18 126)$1 328 557Postprocedural infection$16 814 (9395 to 24 234)$756 635$12 202$610 100$15 068 (10 617 to 19 520)$1 431 494Pericarditis$6867 (4675 to 9060)$315 903$7801$93 612$7728 (5111 to 10 345)$440 482Procedure‐related acute myocardial infarction$17 175 (−3535 to 37 886)$85 877$11 818$177 273$14 820 (6322 to 23 317)$296 394Venous thromboembolism$4246 (417 to 8075)$25 476$8910$71 282$11 211 (2786 to 19 637)$156 958Acute kidney injury$10 331 (6787 to 13 875)$495 900$10 851$43 402$9754 (6573 to 12 935)$507 216Complications requiring cardiac surgery$54 988 (47 133 to 62 844)$714 844$122 809$859 662$75 669 (44 520 to 106 818)$1 513 381Complete atrioventricular block$1339 (−311 to 2989)$57 588$7973$7973$1210 (−369 to 2789)$53 252NA indicates not applicable. Adjusted Excess in Hospitalization Costs Associated With Complications Occurring During the Hospital Stay, and Up to 30 Days Postdischarge NA indicates not applicable. Taken together, the occurrence of any complications up to 30 days postdischarge was associated with a near doubling of the total hospitalization cost (cost ratio, 1.95 [95% CI, 1.83–2.09], Table S6), corresponding to a $7812 (95% CI, $6754–$8870) increase in hospitalization cost per patient who experienced a complication. Overall, this constituted $9.0 million in excess hospitalization costs. When individual complications were considered, those requiring cardiac surgery were the costliest, with an average $75 669 (95% CI, $44 520–$106 818) increase in cost, or a total cost of $1.5 million, contributing to 16.8% of the total excess cost. A bleeding event was associated with an increase of only $5591 per patient (95% CI, $4511–$6672), but owing to their high incidence, all bleeding events were associated with a total cost of $3.8 million, or 41.9% of the total excess cost. Other complications with significant excess cost included pericardial effusion ($1.6 million, 18.2% of total excess cost) and infection ($1.4 million, 15.9%). Conditions significantly increasing hospitalization cost included congestive heart failure (cost ratio, 1.09 [95% CI, 1.06–1.12]) and diabetes (cost ratio, 1.03 [95% CI, 1.00–1.05]), whereas chronic kidney disease was associated with slightly lower cost (cost ratio, 0.95 [95% CI, 0.91–0.99]). Impact of Records With Unavailable Cost Data Overall, 2355 records had a missing or unmatched DRG and were excluded from the primary analysis. However, between patients with and without available cost data, there were only modest differences in baseline characteristics (Table S7), and incidence of overall, in‐hospital, and postdischarge complications was comparable except for a modest difference in the incidence of pericardial effusion (Table S8). Discussion In this nationwide study in Australia, we found that 30‐day complications following elective AF ablations were associated with substantial health care resource use, averaging 3.3 extra days of hospital care and $7812 in excess hospitalization costs. The total excess bed days and cost associated with procedural complications were 3851 days and $9.0 million in health care costs, respectively. More than half of this excess cost was attributable to bleeding and pericardial effusion. Collectively, these findings provide important insights into the health care resource burden associated with complications following AF ablations. They also provide guidance to strategies aiming at reducing procedural complications of AF ablation and have important implications for future cost‐effectiveness analysis of AF ablation. The burden on health care resource use associated with complications following AF ablation has not been evaluated before, although extensive research on the incidence and types of procedural complications has been carried out. 3 , 4 , 5 , 6 We found that procedural complications were not uncommon, occurring in 1 out of 17 patients undergoing AF ablation, although this study cohort was relatively young and had low rates of comorbidities. This rate is consistent with previous publications, which reported a rate of complications ranging from 3.4% to 7.2%. 3 , 4 , 5 , 6 , 28 , 29 More important, we found that procedural complications were associated with almost a 3‐fold increase in the LOS and doubling of hospitalization cost. We also observed a significant increase in LOS and hospitalization costs associated with complications that are often anecdotally deemed not to be a “major” event in previous studies, 5 , 6 such as pericardial effusion not requiring drainage or bleeding without the need for blood transfusion. Nevertheless, our observation suggests that although these patients did not require intervention for their complications, they likely stayed in hospital for observation, further investigation, or medical treatment, all of which would prolong LOS and increase the cost. Although often deemed clinically “minor” events, these complications still appeared to have a significant burden on health care resources. Collectively, our findings show that complications following catheter ablation of AF are not only common but exceedingly costly. This represents significant burden given that the study cohort was relatively young and had low risk profile. Moreover, to fully capture the impact of procedural complications on patient care and hospital resources, future studies of AF ablation safety should consider all clinically meaningful events. Comparisons with studies in other countries are challenging due to the fundamental differences in the funding mechanisms of different health care systems and different time points. Nevertheless, compared with the data reported from United States, the hospitalization cost in Australia is much lower (6907 Australian dollars in 2021 versus 24 188–$98 436 US in 2014). 30 , 31 Besides the inherent differences between the 2 health care systems, this observation could be explained partly because the US studies include in‐patient procedures only, which are usually performed on sicker patients who may require more complex, costly treatment, and longer stay. 30 , 31 For example, the LOS reported in the US study is 2.4 days, which is not typical for AF ablation, a procedure that usually requires day‐only admission or overnight stay. Indeed, Bollmann et al reported that a significant portion of AF ablations is being performed in the outpatient setting (accounting for 37% to more than 90% of all AF ablations). 32 To our knowledge, there are no prior data available regarding the costs associated with complications of AF ablation before this study. Nevertheless, few prior publications have reported on the high cost of complications related to cardiac devices. For example, Reynold et al reported a 3.4‐day increase in the LOS of the index hospitalization (median LOS was 2.0 days) and a $US 7251 increase in hospitalization cost associated with any procedural complication following a defibrillator implantation. 10 Similarly, researchers in Australia and France also found cardiac device‐related infections to be associated with very high costs ($19 043 per admission and €23 234 per de novo infected patient in Australia and France, respectively). 11 , 12 Besides extending the knowledge about AF ablation safety, our study also has important implications for clinicians and policymakers seeking to establish targeted strategies to reduce procedural complications. Specifically, our results suggest that efforts to improve care may benefit from focusing on the most common and most costly events. Although complications requiring cardiac surgery had the highest per patient cost, these occurred infrequently. Conversely, a single bleeding event or pericardial effusion did not incur a high per patient cost yet contributed to approximately 50% of all excess cost because these events were far more common. More important, there are also effective strategies to reduce these complications. Specifically, cardiac perforation and subsequent pericardial effusion can be reduced by better visualization of the left atrium, either with multiple fluoroscopic views, the guidance of transesophageal echocardiography, or intracardiac echocardiography (ICE). 33 The use of ICE, for example, is associated with a halved complication rate (hazard ratio of 0.48, 95% CI 0.44–0.51) and a significantly shortened LOS (2.1 days versus 4 days) 31 when compared with the no‐ICE group. It is also notable the ICE usage maybe suboptimal given that only 13% of guidelines‐writing members outside of North America report using ICE. 34 Similarly, the use of novel anticoagulants, compared with vitamin K antagonists, is associated with a halved rate of major bleeding (0.8% versus 1.5%) and a comparable risk of thromboembolic events. 35 Although some interventions may be expensive, especially ICE, they may have an incremental cost‐effectiveness ratio by reducing the incidence of complications, shortening fluoroscopy and dwell time, and length of hospital stay. 31 , 36 Another cheaper and more readily available option is vascular ultrasound, which has been shown to be associated with a nearly 70% reduction in risk of vascular injury (odds ratio, 0.32 [95% CI, 0.21–0.49]) compared with the no‐ultrasound group. 37 Although ICE and ultrasound‐guided vascular access have become standard of care in many high‐volume centers, a higher uptake of these measures, and other strategies such as clinical audits and safety checklists, which have been well proven to minimize adverse events, 38 , 39 should be considered to reduce procedural complications and their associated costs. These findings also have implications for evaluating the cost‐effectiveness of AF ablation. Existing cost‐effectiveness analysis mostly factors in the incidence but not the cost of complications. 7 , 8 , 9 A cost‐effectiveness analysis that fails to incorporate the cost of complications will likely overestimate the cost‐effectiveness of catheter ablation, although the authors may have paid tremendous attention to other details in the model. 7 , 8 , 9 A comprehensive cost‐effectiveness analysis will greatly facilitate policymakers in allocating resources more appropriately. Therefore, we believe that future cost‐effectiveness models should consider both the incidence and cost of procedural complications, and our study provides insightful data that could be used as input for such models. The following limitations should be considered when interpreting our results. Routinely collected administrative data are generally not as granular as data purposefully collected for research. Nevertheless, reasonable accuracy has been reported for coded data in Australia when compared with medical records. 14 , 15 Good accuracy has also been reported for identifying AF ablation in administrative data, 19 as well as linkages of health records. 20 Although hospitalization costs could not be estimated for patients with missing or unmatched DRG, the patient characteristics and the incidence of complications among these patients was largely comparable to that of those with available cost data. Therefore, our findings are likely to be generalizable to all patients undergoing AF ablation in Australia. Furthermore, some variables are not routinely collected in administrative data and, therefore, not adjusted for in our cost models. These include blood test results, type of AF, ablation energy or lesions created, imaging results, medications provided, and socioeconomic status. Some complications were not reported on, including atrioesophageal fistula, phrenic nerve injury, and pulmonary vein stenosis, due to the lack of specific diagnosis and procedure codes to identify them. However, these are rare complications that either rarely require treatment (phrenic nerve injury and pulmonary vein stenosis) 40 , 41 or could be captured under other categories (atrioesophageal fistula usually presents as sepsis or stroke). 42 All hospitalization costs in this study were derived from the cost estimated from a sample of hospitals, which reflects the average costs that government reimburses all hospitals for services provided. Actual costs incurred by hospitals may vary, especially in private hospitals, where patients may be charged additional out‐of‐pocket expenses that are not included in DRG‐based estimates. We were also not able to calculate how much the actual costs were under‐ or overestimated, which is a source of potential measurement error. Our analysis did not include the indirect cost of care, such as visits to a general practitioner clinic or emergency department, prescribed medications, or workdays lost. Lastly, we could not estimate how much of the excess cost was avoidable, mostly due to the lack of studies of the cost‐effectiveness of interventions such as ICE or ultrasound‐guided vascular access. Nevertheless, we have shown that the procedural complication rate is reducing over the study period 18 and there are significant differences in complication rates among individual hospitals, 43 suggesting that there are opportunities to improve procedural safety. There are also interventions that have been shown to significantly reduce procedural complications, as we have alluded to earlier. 31 , 37 Furthermore, a study examining the preventability of hospital complications more broadly shows that about half of all hospital‐related adverse events and complications are preventable. 44 Despite these limitations, our study provides important insights into the economic burden associated with complications following AF ablation. Conclusions Complications following catheter ablation of AF were on average associated with an extra 3.3 bed days in hospital and a $7812 increase in hospitalization costs per each patient experiencing a complication. Efforts to reduce complications should consider targeting bleeding and pericardial effusion, which were the most common complications and associated with the highest total excess costs. Sources of Funding This work was supported by the National Heart Foundation of Australia (ID 101186). Disclosures Dr Ngo was a recipient of The Hospital Research Foundation Postgraduate Scholarship and a Research Training Program Scholarship from The University of Queensland. Drs Ranasinghe and Ganesan were recipients of the National Heart Foundation of Australia Future Leader Fellowships (IDs 101 186 and 101 188, respectively). Dr Denman has received speaking honoraria from Medtronic. The remaining authors have no disclosures to report. Supporting information Tables S1–S8 Figures S1–S3 Click here for additional data file.