Patients who have undergone the Fontan procedure in later adolescence and adulthood represent a unique population at risk for significant morbidity and mortality. The optimal strategy for long-term management of such patients is unknown. The aim of this study was to evaluate outcomes of patients who had undergone Fontan surgery later in life, focusing on late survivorship, mode of death, and predictors of mortality. Eighty-eight patients were identified who had their initial Fontan operation from 1973 to 2007 at ≥15 years of age. A standardized tiered contact protocol was followed to capture the recent health status of each patient; the probability of survival was 83%, 71%, and 66% at 5, 10, and 15 years of follow-up, respectively. Despite focused efforts, the modes of death were not available in 48% of the patients. A prolonged intensive care unit stay at the time of operation was the single predictor of mortality (p = 0.0123). In conclusion, this investigation highlights the significant mortality that exists in patients who undergo a Fontan procedure later in life and the difficulties in achieving standardized medical follow-up for this high-risk group of patients.
Patients with functional single-ventricle circulation are among the most complex of those living with congenital heart disease (CHD). There are currently few and, at times, conflicting outcome data to assist in guiding their treatment. Guidelines for the care of adults with CHD have evolved largely based on extrapolations from limited cohort studies, and tracking adults with CHD remains problematic, complicating clinical and scientific review of outcomes. The objective of our investigation was to design, implement, and assess a simple protocol for locating patients who underwent the Fontan procedure at ≥15 years of age at our institution and, as feasible, to obtain basic outcome data characterizing early mortality, late survivorship, modes of death, and predictors of mortality in this cohort.
Methods
Patients were identified through departmental databases at Boston Children’s Hospital. All patients who underwent their initial Fontan operation from September 1973 to June 2007 were included if they met the following criteria: (1) they were ≥15 years of age at the time of initial Fontan and (2) they were residing within the United States at the time of initial Fontan. Exclusion criteria included Fontan conversions and revisions in patients who underwent initial Fontan surgery before 15 years of age. The study was conducted in accordance with the institutional Human Subjects Committee guidelines and was approved by the Institutional Review Board.
Retrospective review of medical records including available inpatient and outpatient notes, procedural reports, and laboratory investigations from providers practicing within or outside the core institutions (i.e., Boston Children’s Hospital and Brigham and Women’s Hospital) was performed for each patient. These data included predefined variables relating to pre- and postoperative patient functional status and postoperative complications and mortality. Patient place of residence was defined as instate (Massachusetts), New England (New Hampshire, Vermont, Maine, Connecticut, and Rhode Island), or other (all others). Two specific variables, body mass index (BMI) and New York Heart Association (NYHA) functional status, were collected at 3 time points: (1) before Fontan surgery, (2) ≤6 months after Fontan, and (3) >6 months after Fontan.
Patient diagnoses were divided into 3 groups: tricuspid atresia, double-inlet left ventricle, and “complex” in accord with previous similar investigations. Patients included in the diagnosis “complex” included all those with the anatomy unsuitable for a 2-ventricle repair who did not have tricuspid atresia or double-inlet left ventricle. Type of Fontan was classified into the following 3 categories based on operative reports: (1) right atrium to pulmonary artery anastomosis and right atrium to right ventricle anastomosis (classic/modified), (2) intra-atrial lateral tunnel (lateral tunnel), and (3) extracardiac conduit (extracardiac). Surgical era was categorized into 3 time periods: (1) January 1, 1973 to December 31, 1988, (2) January 1, 1989 to December 31, 1999, and (3) January 1, 2000 to December 31, 2007.
The status of each patient was investigated by a member of the research team using a tiered contact protocol. First, medical records from our core institutions were reviewed. Postoperative appointments with a cardiovascular surgeon, pediatric cardiologist, or adult congenital cardiologist associated with Boston Children’s Hospital were recorded. Postoperative follow-up appointment status was further categorized into patients having ≥2 appointments within a single discipline (i.e., cardiovascular surgeon, pediatric cardiologist, or adult congenital cardiologist) and patients having any postoperative appointments with a pediatric cardiologist or an adult congenital cardiologist at our core institutions. The date of the most recent cardiology visit was recorded.
A “possible lapse of care” was defined as no direct recorded contact with our adult congenital heart disease (ACHD) center within the last 3 years. In cases when patients had experienced a “possible lapse of care,” a specified protocol was followed: if the patient had no record of a visit after September 1, 2009, the patient’s last known physician was contacted. In the event that the patient’s status was still unknown, the Social Security Death Index was reviewed. If the patient’s mortality status was still not defined, an “opt-out” letter was sent to each patient’s last known address requesting that they send back notification if they did not wish to be contacted. If there was no response, an attempt was made to contact the patients at their last known phone number. For each patient, a standardized telephonic interview was performed, with attainment of verbal informed consent before any direct questioning. If all aforementioned attempts were unsuccessful, social networks (e.g., Facebook, peoplefinder.com ) were used in an attempt to find the patients.
The type of the most recent follow-up clinician was defined by the presence and review of medical records from both core institutions and clinicians practicing outside of core institutions, supplemented by data supplied by the patient interview. If a patient had a record of a medical encounter by a congenital cardiologist in the past 3 years at our core institutions or by a pediatric cardiologist practicing outside of our core institution, regardless of concomitant care by other noncongenital care providers outside of core institutions, care was categorized as delivered by a congenital cardiologist. If there was documentation of follow-up in the past 3 years by an internal medicine cardiologist practicing either within or outside of our core institutions, care was categorized as delivered by a “noncongenital cardiologist.” If no record of encounter with a cardiologist was noted, but documentation of follow-up in the past 3 years by other internal medicine or general practice clinicians was present, care was categorized as delivered by an “internal medicine or general practice” provider; if encounters were not noted with any of the aforementioned clinicians within the past 3 years, patients were classified as “presumed lost to all medical follow-up”.
Mortality was considered “early” if it occurred within 30 days of surgery or “late” if it occurred ≥30 days after the Fontan procedure. The cause of death was determined by review of available medical records and death certificates. Death was classified as secondary to heart failure if it complicated worsening heart failure, as defined by evidence of at least 1 of the following: orthopnea, nocturnal dyspnea, pulmonary edema, increasing peripheral edema or renal hypoperfusion related to cardiac dysfunction, or radiological signs of congestive heart failure. Sudden death was defined as death occurring within 1 hour of onset of acute symptoms. Remaining deaths were classified as deaths due to cerebrovascular accident, thromboembolism, sepsis, other, or “unknown” if the cause of death was unclear.
When appropriate, variables were divided into dichotomous values before tests of association based on accepted standards. Bivariate analysis was conducted using Pearson’s chi-square test. In the event of inadequate sample size, Fisher’s exact test was performed. The Friedman 2-way analysis of variance by ranks was used to detect differences in BMI and NYHA functional class at different time points. In the event of a significant difference, the Wilcoxon signed rank test was used to detect which conditions differed. p Values ≤0.05 were considered significant. Survival statistics were performed using the Kaplan-Meier Method. The day of the Fontan procedure was used as the starting point. Survival curves were compared using the log-rank test. All data management, graphs, and statistical analyses were performed using the R version 2.14.1 software (R Foundation for Statistical Computing, Vienna, Austria).
Results
A total of 88 patients who met inclusion criteria were identified; 43 (49%) were women. The oldest patient age at the time of Fontan surgery was 40 years. Most (n = 47) pre-Fontan diagnoses fell into the “complex” category. We identified 45 patients who lived instate at the time of the operation and an additional 22 patients lived in the other New England states. Table 1 lists the clinical characteristics of this cohort.
| Characteristic | Number (%) of Subjects |
|---|---|
| Gender | |
| Men | 45 (51) |
| Women | 43 (49) |
| Location | |
| Massachusetts | 45 (51) |
| New England | 22 (25) |
| Other | 21 (24) |
| Age (yrs) | |
| 15–19 | 45 (51) |
| 20–29 | 29 (33) |
| 30–39 | 13 (15) |
| 40+ | 1 (1) |
| BMI (kg/m 2 ) | |
| ≤18 | 33 (38) |
| 18–24.9 | 45 (51) |
| 25–29.9 | 6 (7) |
| ≥30 | 1 (1) |
| Unknown | 3 (3) |
| Cardiac diagnosis | |
| Tricuspid atresia | 25 (29) |
| Double-inlet left ventricle | 16 (18) |
| Complex | 47 (53) |
| Previous surgical procedure | |
| Pulmonary artery band | 7 (3) |
| Systemic-pulmonary arterial shunt | 62 (23) |
| Cavopulmonary anastomosis | 16 (10) |
| Preoperative NHYA class | |
| I | 2 (2) |
| II | 21 (24) |
| III | 58 (66) |
| IV | 2 (2) |
| Unknown | 5 (6) |
On an initial core institution chart review, 33 of 88 patients were classified as having “possible lapse of care.” Ultimately, we were able to determine the basic clinical status of all but 6 patients using a specific algorithm for follow-up data collection. The clinical status was determined by the chart review, including correspondence from outside centers in 65 patients. Of the remaining 17 patients, the status of 4 was verified by contacting them through their last known phone number. Using the Social Security Death Index, we were able to identify an additional 7 who had died. Querying social media resources, 6 were determined to be alive. The current clinical status of 6 patients remains unknown.
Information regarding the BMI of each patient was available for 85 of 88 patients preoperatively, 55 patients at follow-up ≤6 months from the operative date, and 67 patients at follow-up >6 months from the operative date. Data pertaining to NYHA functional class were available for 83 of 88 patients preoperatively, 43 patients at follow-up ≤6 months from the operative date, and 62 patients at >6 months from the operative date. In total, information regarding the BMI and NYHA class at all 3 time points was available for only 46 and 39 patients, respectively ( Table 2 ). BMI changed more dramatically at a later time compared with the changes reported in NYHA class, which improved within 6 months postoperatively. Both variables were significantly different preoperatively compared with follow-up >6 months from the operative date (p <0.001 and p <0.0001, respectively).
| Variable | Preoperative | <6 months | >6 months | p |
|---|---|---|---|---|
| BMI (kg/m 2 ) | n = 46 | <0.001 | ||
| <18 | 14 | 17 | 6 | |
| 18.6–24.9 | 28 | 27 | 23 | |
| 25–29.9 | 3 | 2 | 17 | |
| >30 | 1 | 0 | ||
| NYHA class | n = 39 | <0.0001 | ||
| I | 1 | 24 | 14 | |
| II | 12 | 14 | 19 | |
| III | 26 | 1 | 5 | |
| IV | 0 | 0 | 1 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
