Actual timing of the Fontan operation is variable. Our aim was to evaluate the impact of age at the time of Fontan operation on mortality and clinical outcome and characterize patients with worse outcomes. We conducted a retrospective, cross-sectional study on the Fontan operation using nationally representative databases from 2003 to 2016 and categorized the patient into 1 of 5 groups according to their age at the time of surgery: <2, 2, 3, 4, and ≥5 years. Survey-weighted logistic regression models were used to compare the outcomes of the different age groups. A total of 6,647 children underwent the Fontan completion procedure during the study period with median age 3 (interquartile range 2 to 4) years. The in-hospital mortality was 2.1%. In logistic regression models, in-hospital mortality, respiratory failure, acute kidney injury, chylothorax, arrhythmias, and sudden cardiac arrest were similar among the 5 age groups. Compared with children >2 years, those <2 years were less likely to be admitted for surgery on an elective basis (73.5% vs 90.4%, p <0.001), more likely to have chromosomal anomalies (2.7% vs 1.7%), and more likely to have repair of atrioventricular valves (8.5% vs 6.0%, p = 0.027). Mortality was higher in those with an underlying atrioventricular septal defect (AVSD) adjusted odds ratio 4.3 (2.4 to 7.9, p <0.001). Repair of AV valves was more common in the AVSD group compared with those in non-AVSD (14.3% vs 5.5%, p <0.001). In conclusion, age at Fontan completion does not adversely affect the in-hospital outcomes. Our focus should be on optimizing essential factors that are crucial for successful Fontan completion.
The Fontan procedure is a palliative surgery performed in children with a functional or anatomic single ventricle. This procedure is the most common and is the final stage of a series of palliative procedures performed to patients with any type of single ventricle physiology. In the United States, around 1,200 Fontan procedures are performed annually. Studies reporting outcomes from the pre-2000 era showed that Fontan completion surgery before the age of 4 was associated with worst outcomes in both short and long term. The present body of research suggested that the procedure performed at 2 to 4 years of age is associated with a better prognosis, as procedures performed at an older age was associated with worse postoperative hemodynamics. , In addition, surgery performed at an older age has also been reported to be associated with worsening of atrioventricular valve regurgitation and deterioration of the diastolic function. In a relatively recent multicenter study of 2,747 children undergoing Fontan surgery from 2000 to 2009, age and weight of the patients were not associated with outcomes; however, weight-for-age z score <−2 was associated with morbidity and mortality. A recent study with more than 3,000 Fontan surgeries reported the lowest risk of in-hospital mortality when surgery was performed at 3 years of age, but a comprehensive analysis and report for this comparison with the contemporary data is lacking in research. Here, we used a nationally representative pediatric hospitalization database of the United States, the Kids’ Inpatient Database (KID), to compare clinical outcomes among 6,647 children undergoing the Fontan palliative surgery based on age groups and also characterize the children with worse outcomes.
Methods
We conducted a retrospective, cross-sectional study of the Fontan completion procedure using the Kids’ Inpatient Database (KID) published for 2003, 2006, 2009, 2012, and 2016. KID database is available through the Healthcare Cost and Utilization Project (HCUP) developed by the Agency for Healthcare Research and Quality. The KID is published every 3 to 4 years, with 2016 being the most recent year of publication. The database contains administrative datasets of a nationally representative sample of pediatric hospitalization from 4,179 acute care hospitals in 44 states. The KID includes patient demographic information and International Classification of Disease, Ninth or Tenth Revision, Clinical Modification (ICD-9-CM and ICD-10-CM) discharge diagnoses and procedures. KID is a stratified database which includes discharge data for approximately 10% of uncomplicated newborn hospitalizations and 80% of acute hospitalizations in the United States. The KID is the largest publicly available all-payer health pediatric inpatient care database in the United States, including more than 3 million unweighted pediatric hospitalization records each year to estimate over 7 million hospitalizations.
The data use agreement with the Healthcare Research and Quality prohibits direct sharing of data, analytic methods, and study materials in the public domain. The deidentified nature of the datasets preclude a need for institutional review board approval.
We used ICD-9 (2003, 2006, 2009, and 2012 datasets) and ICD-10 (2016 dataset) diagnostic and procedural codes in the primary or secondary procedure code fields to identify children who underwent a Fontan completion procedure. Supplementary Table 1 listed diagnostic and procedural codes used in the study. Per HCUP, race includes both race and ethnicity; in the case that the source data supplied both, ethnicity takes precedence over race. For example, if a patient is coded as being both Black and Hispanic, then race is defined as Hispanic.
We divided the children who underwent the Fontan surgery into 5 groups according to their age at the time of surgery: <2, 2, 3, 4, and ≥5 years of age. The primary outcomes were in-hospital mortality and clinical complications including tachyarrhythmia, bradyarrhythmia, respiratory failure, acute kidney injury, cardiopulmonary compromise requiring extracorporeal membrane oxygenation (ECMO) support, and sudden cardiac arrest. Due to the reliability and complexity of retrieving AV valve repair with ICD-10 procedure codes, we only included AV valve repair data from 2003 to 2012 (coded with ICD-9 codes) and excluded 2016, which is coded with ICD-10 procedure codes. A complete list of ICD-9 and ICD-10 codes used for querying the aforementioned complications are available in Supplementary Table 1. Additionally, we analyzed the length of hospital stay and the use of home health care after hospital discharge.
Demographic covariates included patients age (which were categorized into 5 different groups), gender, race/ethnicity, year of the Fontan completion surgery, zip code of family income quartile, season of procedure performed (summer [April-October] vs winter [November to March]), presence of chromosomal anomalies, and types of single ventricles (tricuspid atresia, pulmonary atresia with intact ventricular septum, hypoplastic left heart syndrome, AV septal defects, and other single ventricle pathologies).
Sample weights were applied to patient-level discharge observations to generate a nationally representative estimate of US hospitalizations per recommendations from the HCUP Methods Series. Descriptive statistics were performed using the KID complex survey design. For descriptive statistics, median and interquartile intervals were reported for continuous variables. Chi-square tests were used for categoric variables. A median rank test was used to evaluate the differences in length of hospital stay.
We developed logistic regression models to estimate the difference in primary in-hospital clinical outcomes among age groups. All models were adjusted for the aforementioned covariates. To explore whether the outcomes varied as a result of the types of single ventricle physiology, we conducted a sensitivity analysis which was stratified by the single ventricle condition.
All statistical analyses were performed using Stata software version 15 (StataCorp LP, College Station, Texas), as well as R version 4.0.1 ( http://www.R-project.org ) in RStudio 1.2. All tests were 2-sided, and a p value <0.05 was considered statistically significant. Odds ratios (ORs) obtained in logistic regression models were all adjusted for covariates. Complex survey design of the KID was accounted for using the survey package. Figures were produced using the ggplot2 package.
Results
We identified 6,647 hospitalization records associated with Fontan completion during the study period from 2003, 2006, 2009, 2012, and 2016. 2,683 were girls (40.4%). Figure 1 shows the basic descriptive statistics.
The underlying cardiac anatomy of the patients undergoing Fontan completion were as follows: 2,324 with hypoplastic left heart syndrome, 488 with pulmonary atresia, 1,402 with tricuspid atresia, and 680 with AV septal defects. These 4 lesions comprised 73.6% of patients who underwent the Fontan completion surgery. Median age of surgery was 3 years, with an interquartile range (IQR) of 2 to 4 years.
The in-hospital mortality rate for the whole cohort was 2.1% (n = 139). The median length of stay was 10 (IQR 7 to 15) days. A total of 428 were discharged home with home health care services (6.4%). Common complications included respiratory failure in 4.2% (n = 282), acute kidney injury in 3.9% (n = 258), and chylothorax in 6.2% (n = 412) patients; whereas 1.8% (n = 117) needed ECMO during the ICU stay.
A majority of the patients (70%) underwent Fontan completion between 2 and 4 years of age ( Table 1 ). Although a majority of the surgeries were performed in summer, in the group of patients <2 years old, the percentage of Fontan completions in the summer (61.2%) was significantly lower compared with the other age groups (75.3%, 73.3%, 67.9%, and 67.7%, respectively, in the 2-, 3-, 4-, and ≥5-year-old groups [p <0.001]). A majority (90%) of the Fontan procedures were performed electively (based on a variable in the datasets that indicated elective hospitalizations), but these were least likely to occur in the <2-year-old group compared with the highest percentage in the 3-year-old group (73.4% vs 92.2%, p <0.001) ( Table 1 ). In terms of complications, the risk of cardiopulmonary failure leading to ECMO use was highest in children <2 years old (4.8%) compared with 1.2% at 3 years old, p <0.001. Similarly, sudden cardiac arrests were the highest in patients <2 years old (2.0% vs 0.9% at 3 years old). Age 3 was selected as a reference group because the median age of surgery was 3 years of age. In-hospital mortality was the lowest when surgery was performed at 3 years of age (1.3%, n = 21) and the highest when performed at <2 years of age (4.3%, n = 41), with a p <0.004 ( Table 1 ).
| Variable | Age(Years) | |||||
|---|---|---|---|---|---|---|
| <2 | 2 | 3 | 4 | ≥5 | p Value | |
| Total numbers | 956 | 2190 | 1634 | 829 | 1,039 | |
| Female | 384 (40.2%) | 821 (37.5%) | 666 (40.7%) | 363 (43.7%) | 449 (43.4%) | 0.046 |
| 0.003 | ||||||
| White | 493 (59.2%) | 1,046 (55.2%) | 716 (52.8%) | 363 (51.3%) | 387 (45.1%) | |
| Black | 113 (13.6%) | 190 (10.0%) | 169 (12.5%) | 102 (14.4%) | 115 (13.4%) | |
| Hispanic | 146 (17.6%) | 496 (26.1%) | 326 (24.0%) | 170 (24.1%) | 237 (27.6%) | |
| Others | 80 (9.6%) | 165 (8.7%) | 145 (10.7%) | 72 (10.2%) | 119 (13.9%) | |
| Year | 0.14 | |||||
| 2003 | 202 (21.1%) | 602 (20.0%) | 395 (15.1%) | 226 (13.9%) | 287 (19.1%) | |
| 2006 | 195 (20.4%) | 489 (22.3%) | 302 (18.5%) | 191 (23.0%) | 211 (20.3%) | |
| 2009 | 155 (16.2%) | 482 (22.0%) | 377 (23.1%) | 169 (20.4%) | 217 (20.9%) | |
| 2012 | 149 (15.6%) | 390 (17.8%) | 393 (24.0%) | 174 (21.0%) | 186 (17.9%) | |
| 2016 | 255 (26.7%) | 392 (17.9%) | 315 (19.3%) | 180 (21.7%) | 227 (21.8%) | |
| Zip code quartile (Family income) | 0.043 | |||||
| 0-25 th | 257 (27.6%) | 601 (28.2%) | 395 (24.8%) | 226 (28.2%) | 287 (28.9%) | |
| 26–50 th | 279 (30.0%) | 481 (22.6%) | 448 (28.1%) | 200 (25.0%) | 251 (25.3%) | |
| 51–75 th | 231 (24.8%) | 571 (26.8%) | 384 (24.1%) | 222 (27.8%) | 249 (25.1%) | |
| 76–100 th | 163 (17.5%) | 478 (22.4%) | 364 (22.9%) | 152 (19.0%) | 206 (20.8%) | |
| Summer | 585 (61.2%) | 1,648 (75.3%) | 1,198 (73.3%) | 563 (67.9%) | 704 (67.7%) | <0.001 |
| Elective admission | 697 (73.4%) | 1,972 (90.3%) | 1,503 (92.2%) | 735 (88.7%) | 921 (89.1%) | <0.001 |
| Chromosomal abnormalities | 25 (2.6%) | 24 (1.1%) | 32 (1.9%) | 23 (2.8%) | 19 (1.8%) | 0.05 |
| Complication | ||||||
| Chylothorax | 40 (4.1%) | 140 (6.4%) | 106 (6.5%) | 74 (8.9%) | 52 (5.0%) | 0.018 |
| Respiratory failure | 54 (5.7%) | 80 (3.6%) | 60 (3.7%) | 30 (3.6%) | 58 (5.5%) | 0.09 |
| AKI | 35 (3.7%) | 83 (3.8%) | 52 (3.1%) | 32 (3.9%) | 55 (5.3%) | 0.28 |
| ECMO | 46 (4.8%) | 36 (1.7%) | 20 (1.2%) | 3 (0.3%) | 12 (1.1%) | <0.001 |
| Tachycardia | 112 (11.7%) | 189 (8.6%) | 186 (11.4%) | 98 (11.8%) | 163 (15.7%) | <0.001 |
| Heart block | 53 (5.6%) | 122 (5.6%) | 71 (4.4%) | 44 (5.3%) | 61 (5.8%) | 0.7 |
| SCA | 19 (2.0%) | 21 (0.9%) | 15 (0.9%) | 3 (0.4%) | 15 (1.4%) | 0.08 |
| Cardiac lesion | ||||||
| Pulmonary atresia | 64 (6.7%) | 152 (6.9%) | 122 (7.5%) | 58 (6.9%) | 92 (8.8%) | 0.62 |
| Tricuspid atresia | 202 (21.1%) | 446 (20.3%) | 368 (22.5%) | 162 (19.6%) | 223 (21.5%) | 0.63 |
| HLHS | 336 (35.1%) | 873 (39.8%) | 575 (35.2%) | 304 (36.6%) | 237 (22.8%) | <0.001 |
| Other SV | 134 (14.0%) | 504 (23.0%) | 430 (26.3%) | 223 (26.8%) | 291 (27.9%) | <0.001 |
| AVSD | 86 (9.0%) | 195 (8.9%) | 167 (10.2%) | 98 (11.6%) | 134 (12.9%) | 0.06 |
| Mortality | 41 (4.3%) | 43 (1.96%) | 21 (1.31%) | 13 (1.56%) | 21 (2.0%) | 0.004 |
| LOS (median, IQR) days | 10 (7–16) | 9 (7–14) | 9 (7–12) | 10 (7–15) | 10 (7–15) | 0.14 |
| Discharged with home health | 125 (13.7%) | 137 (6.3%) | 92 (5.6%) | 27 (3.3%) | 47 (4.5%) | 0.005 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
