Abstract
Objective
Determine the accuracy of the pre-transplant clinical diagnosis of heart disease in the UNOS database.
Background
Because survival on the heart transplant waitlist is dependent on underlying heart disease, a new allocation system will include type of heart disease. Accuracy of the pre-transplant clinical diagnosis and the effect of misclassification are unknown.
Methods
We included all adults transplanted at our center from January 2009 to December 2015. We compared the pre-transplant clinical diagnosis at listing to pathology of the explanted heart and determined the potential effect of misclassification with the proposed allocation system.
Results
334 patients had the following clinical diagnosis at listing: 148 dilated cardiomyopathy, 19 restrictive cardiomyopathy, 103 ischemic cardiomyopathy, 24 hypertrophic cardiomyopathy, 11 valvular disease, 16 congenital heart disease (CHD), and 13 “other.” Pathology of the explanted hearts revealed 82% concordance and 18% discordance (10% coding errors and 8% incorrect diagnosis). The most common incorrect diagnoses were sarcoidosis (66%), arrhythmogenic right ventricular dysplasia (60%), and other causes of predominately right-sided heart failure (33%). Among those misclassified, 40% were listed as UNOS status 2, 8% remained status 2 at transplantation and only sarcoidosis and CHD were potentially at a disadvantage with the new allocation.
Conclusion
There is high concordance between clinical and pathologic diagnosis except for sarcoidosis and genetic diseases. Few misclassifications result in disadvantaging patients based on the new allocation system but rare diseases like sarcoidosis remain problematic. To improve the UNOS database and enhance outcome research, pathology of the explanted hearts should be required post-transplantation.
Keywords: Dilated cardiomyopathy, heart failure, transplantation, sarcoidosis, restrictive cardiomyopathy
Using the Scientific Registry of Transplant Recipients (SRTR) database we have recently shown that survival on the national heart transplant waitlist depends on type of heart disease. (1) Patients with best survival have an ischemic, dilated non-ischemic, or hypertrophic cardiomyopathy; those with poor survival have restrictive cardiomyopathy, congenital heart disease, or prior transplantation. What remains unknown is whether the pre-transplant clinical diagnosis at listing is accurate. Misclassification of heart disease has become a critical issue because the heart transplant allocation system in the United States will soon change to include type of heart disease as a separate tier (Table 1). (2) Therefore, using histologic data from a single center, our objectives were to 1) determine the accuracy of pre-transplant clinical diagnosis of heart disease entered into the national database, 2) identify the heart diseases most likely to be incorrect, 3) determine whether cardiac biopsies (endomyocardial and apical core) prior to transplant improve accuracy of the pre-transplant clinical diagnosis, and 4) determine whether misclassification of heart disease would potentially disadvantage patients in the new UNOS allocation.
Table 1.
Current and Proposed Heart Transplant Allocation System
| Old Tiers | New Tiers | Heart Transplant Criteria* |
|---|---|---|
| Status 1A (most urgent) | Status 1 | ECMO Non-dischargeable BiVAD + RVAD MCS with life-threatening ventricular arrythmias |
| Status 2 | Ventricular tachycardia/fibrillation Intra-aortic balloon pump Percutaneous endovascular MCS Total artificial heart Dischargeable BiVAD/RVAD MCS with device failure |
|
| Status 3 | LVAD x 30 days High dose or >1 inotrope Status 1A Exceptions MCS with other complications |
|
| Status 1B | Status 4 | Stable LVAD Inotropes w/o monitoring Status 1B exceptions Hemodynamically significant CHD Hypertrophic cardiomyopathy + sxs or advanced signs Restrictive cardiomyopathy + sxs or advanced signs ICM with angina + ischemia Amyloidosis + sxs or advanced signs Retransplantation + CAV or advanced HF sxs |
| Status 5 | Combined organs | |
| Status 2 (less urgent) | Status 6 | All others |
CHD=congenital heart disease; ECMO= ICM=ischemic cardiomyopathy; MCS=mechanical circulatory support
Table is simplified. Actual criteria for new allocation system have more qualifying statements to define eligibility (2)
BIVAD=biventricular ventricular assist device, CAV= coronary allograft vasculopathy, CHD=congenital heart disease, ECMO=extracorporeal membrane oxygenation, HF=heart failure, ICM=ischemic cardiomyopathy, LVAD=left ventricular assist device, MCS= mechanical circulatory support, RVAD=right ventricular assist device, sxs= symptoms
Methods
Patient population
We included all adults 18 years and over who received heart transplantation at Cleveland Clinic between January 1, 2009 and December 31, 2015 including dual organ transplant recipients. Patients with prior heart transplantation were excluded because identifying heart disease by pathology depended on knowledge of past medical history. The study was a quality assurance and performance improvement (QAPI) project and approved by the institutional review board at our institution with a waiver of informed consent.
Study Design
The pre-transplant clinical diagnosis entered into the UNOS database was compared to the pathology of the explanted hearts to determine the degree of concordance. The pre-transplant clinical diagnosis of heart disease was classified into the following categories in the SRTR database: dilated cardiomyopathy (adriamycin, peripartum cardiomyopathy, familial, alcohol, myocarditis, viral, idiopathic, and other), ischemic cardiomyopathy, hypertrophic cardiomyopathy, restrictive cardiomyopathy (idiopathic, amyloid, endocardial fibrosis, sarcoidosis, radiation/chemotherapy induced heart disease, other), valvular, congenital heart disease, and other (cancer, arrhythmogenic right ventricular dysplasia, muscular dystrophy, other).
Pathology of the explanted heart was considered the gold standard in this study for type of heart disease. Two pathologists blinded to all clinical information reviewed the pathology reports of the explanted hearts and categorized them as the following: dilated cardiomyopathy (myocarditis, other), restrictive cardiomyopathy (amyloid, endocardial fibrosis, sarcoidosis, other restrictive cardiomyopathy), ischemic cardiomyopathy, hypertrophic cardiomyopathy, valvular heart disease (defined as valve replacement), congenital heart disease, and “other” (arrhythmogenic right ventricular dysplasia, cancer, and other). Degree of correlation between the pre-transplant clinical diagnosis entered into the national transplant database and the pathologic diagnosis of the explanted hearts were compared and labeled as either concordant or discordant (misclassified). Those that were discordant were further categorized as a coding error or incorrect diagnosis by clinician at time of listing based on data in the electronic medical records. Coding errors were listing diagnoses that differed from the last clinical assessment that was concordant with the pathological diagnosis. Incorrect diagnosis was defined as a pre-transplant clinical diagnosis in the electronic medical records entered by a cardiologist at time of listing that differed from the pathology of the explanted heart. If the pathologic report did not match either the data entered or the clinician’s pre-transplant diagnosis, the discordant information was labeled as incorrect diagnosis since further training for data entry would not yield more reliable data. Pathologic reports that described more than one disease (i.e. coronary artery disease and prosthetic valve) were labeled as both and deemed concordant with the pre-transplant clinical diagnosis if the pathology supported the type of heart disease clinical diagnosis entered into the national database.
Baseline characteristics were categorized by pre-transplant clinical diagnosis and taken from Cleveland Clinic data entered into the national transplant database with the exception of hemodynamics. Hemodynamics were obtained from medical chart review because we discovered in our previous SRTR heart transplant research that there was a high degree of missing values for several variables including right atrial pressure.
Cardiac biopsies prior to transplantation were reviewed to determine whether results of a myocardial sample would improve accuracy of the pre-transplant clinical diagnosis of heart disease. Cardiac biopsies included right ventricular endomyocardial tissue and cardiac tissue from an apical core obtained upon implantation of a ventricular assist device.
UNOS status at time of listing and time of transplant were compared to determine if misclassification of type of heart disease would potentially disadvantage patients after implementation of the new heart allocation system. Medical priority for transplantation was defined by UNOS status and documentation was supported by usage of inotropes, devices, intubation or exemptions at time of transplant.
Statistical Analysis
Continuous variables for baseline characteristics were expressed as median and 25th and 75th percentiles (Q1, Q3). Categorical variables were expressed as counts and percentages. Two-group comparisons were performed by Student t test or Kruskal test when appropriate. A p-value <0.05 was considered statistically significant. The statistical analyses were performed using the statistical package SPSS, version 17 (SPSS Inc).
Results
Study Population
Baseline characteristics of 334 patients (29% female) at the time of listing for heart transplant are shown in Table 2. This cohort included 148 patients with dilated cardiomyopathy (DCM), 19 with restrictive cardiomyopathy (RCM), 103 with ischemic cardiomyopathy (ICM), 24 with hypertrophic cardiomyopathy (HCM), 16 with congenital heart disease (CHD), 11 with valvular heart disease (VHD), and 13 “other” (including primary pulmonary hypertension/right ventricular hypertrophy, alpha 1 anti-trypsin deficiency and arrhythmogenic right ventricular dysplasia (ARVD)). There were 25 patients who underwent dual organ transplantation (17 heart and lung, 4 heart and liver and 4 heart and kidney). Median age of this cohort was 53 years and 82% were White.
Table 2.
Baseline Characteristics of Patients at the Time of Listing for Heart Transplant
| Type of Heart Disease Based on Pre-transplant Clinical Diagnosis in UNOS
|
|||||||||
|---|---|---|---|---|---|---|---|---|---|
| Total n (%)=334 | DCM n=148 (44%) n (%) or median (IQR) |
RCM n=19 (6%) n (%) or median (IQR) |
ICM n=103 (31%) n (%) or median (IQR) |
HCM n=24 (7%) n (%) or median (IQR) |
VHD n= 11 (3%) n (%) or median (IQR) |
CHD n=16 (5%) n (%) or median (IQR) |
Other n=13 (4%) n (%) or median (IQR) |
P value | |
|
| |||||||||
| Blood Type | |||||||||
| A | 143 (43) | 66 (45) | 9 (47) | 47 (46) | 8 (33) | 3 (27) | 7 (44) | 3 (23) | 0.001 |
| AB | 23 (7) | 10 (7) | 3 (16) | 10 (9) | 0 | 0 | 0 | 0 | 0.002 |
| B | 43 (13) | 18 (12) | 0 | 18 (18) | 2 (8) | 2 (18) | 0 | 3 (23) | 0.003 |
| O | 125 (37) | 54 (36) | 7 (37) | 28 (27) | 14 (59) | 6 (55) | 9 (56) | 7 (54) | 0.002 |
|
| |||||||||
| UNOS status | |||||||||
| 1A | 114 (34) | 60 (41) | 3 (16) | 34 (33) | 4 (17) | 4 (36) | 3 (19) | 6 (46) | 0.001 |
| 1B | 86 (26) | 43 (29) | 4 (21) | 22 (21) | 4 (17) | 6 (54) | 5 (31) | 2 (15) | 0.01 |
| 2 | 131 (39) | 43 (29) | 12 (63) | 46 (45) | 16 (66) | 1 (10) | 8 (50) | 5 (39) | 0.003 |
| 7 | 3 (1) | 2 (1) | 1 (1) | ||||||
|
| |||||||||
| Female | 96 (29) | 54 (36) | 3 (15) | 17 (16) | 5 (21) | 6 (55) | 5 (31) | 6 (46) | 0.003 |
|
| |||||||||
| Race | |||||||||
| White | 274 (82) | 112 (76) | 14 (75) | 90 (89) | 20 (82) | 11(100) | 14 (87) | 12 (91) | 0.001 |
| Black | 43 (12) | 30 (20) | 4 (20) | 6 (5) | 2 (9) | 0 | 0 | 1 (9) | 0.001 |
| Other | 18 (6) | 6 (4) | 1 (5) | 7 (6) | 2 (9) | 0 | 2 (13) | 0 | 0.002 |
|
| |||||||||
| Cigarette | 166 (49) | 64 (43) | 6 (32) | 73 (70) | 11 (46) | 0 | 7 (44) | 5 (38) | 0.001 |
|
| |||||||||
| HTN | 161 (48) | 69 (46) | 7 (37) | 65 (62) | 4 (17) | 3 (27) | 6 (38) | 7 (54) | 0.001 |
|
| |||||||||
| BMI | 26.3 (23,30) | 25.5 (22,30) | 25.3 (24,41) | 27.5 (24,30) | 25 (22,29) | 27.2 (23,30) | 26.1 (21,29) | 26.1 (27,35) | 0.001 |
|
| |||||||||
| Diabetes | 84 (25) | 27 (18) | 4 (21) | 46 (44) | 0 | 2 (18) | 4 (25) | 1 (8) | 0.0001 |
|
| |||||||||
| Creatinine mg/dl | 1.2 (0.9,1.5) | 1.1 (0.8,1.4) | 1 (1.2,1.2) | 1.4 (1,1.6) | 1.4 (1,1.6) | 1.1 (0.8,1.4) | 1.2 (0.9,1.5) | 1.2 (0.8,1.3) | 0.008 |
|
| |||||||||
| Age (Years) | 53 (46,63) | 55 (40,62) | 53 (52,64) | 61 (54,66) | 51 (38.5,63) | 57 (51,64) | 47 (32.5,58.5) | 42 (42,59) | 0.001 |
|
| |||||||||
| IABP | 20 (6) | 12 (8) | 0 | 7 (7) | 0 | 0 | 0 | 1 (8) | 0.002 |
|
| |||||||||
| ECMO | 10 (3) | 5 (3) | 0 | 1(1) | 0 | 0 | 1 (6) | 3 (8) | 0.007 |
|
| |||||||||
| Ventilator | 12 (4) | 6 (4) | 0 | 2 (2) | 0 | 0 | 1 (6) | 3 (8) | 0.005 |
|
| |||||||||
| ICD | 259 (77) | 124 (84) | 11 (58) | 87 (84) | 19 (80) | 7 (64) | 8 (50) | 3 (23) | 0.001 |
|
| |||||||||
| Inotrope | 97 (29) | 61 (41) | 3 (16) | 21 (20) | 1 (4) | 3 (27) | 3 (19) | 5 (38) | 0.003 |
|
| |||||||||
| MCS | |||||||||
| LVAD | 60 (18) | 30 (20) | 0 | 25 (23) | 2 (8) | 3 (27) | 0 | 0 | 0.001 |
| RVAD+LVAD | 4 (1) | 2 (1) | 0 | 1 (1) | 0 | 0 | 0 | 1 (7) | 0.003 |
| TAH | 3 (1) | 0 | 0 | 1 (1) | 0 | 0 | 0 | 1 (7) | 0.02 |
| None | 269 (80) | 116 (79) | 19 (100) | 77 (75) | 22 (92) | 8 (73) | 16 (100) | 11 (86) | 0.003 |
|
| |||||||||
| Hemodynamic | |||||||||
| RA, mmHg Median (IQR) | 10 (6,16) | 11 (5,16) | 10 (8,11) | 10 (6,13) | 12.5 (9,16) | *14 (10,19) | 14 (6,20) | *16 (12,19) | 0.04 |
|
| |||||||||
| PASP, mmHg Median (IQR) | 48 (35.59) | 50 (33,56) | 45 (40,44) | 45 (34,60) | 48 (36,52) | 55 (44,63) | 67 (25,79) | 34 (20,50) | 0.1 |
|
| |||||||||
| PADP, mmHg Median (IQR) | 24 (17,30) | 25 (17,30) | 25 (24,25) | 22 (16,30) | 23 (17,29) | 23 (8,25) | 27 (17,30) | 17 (14,26) | 0.5 |
|
| |||||||||
| mPA, mmHg Median (IQR) | 31 (22,40) | 33 (22,40) | 31 (29,29) | 30 (22,40) | 31.5 (22,37) | 34 (22,30) | 37 (12,35) | 20 (14,30) | 0.15 |
|
| |||||||||
| PCWP, mmHg Median (IQR) | 22 (14,28) | 22 (14,30) | 23 (24,24) | 22 (16,28) | 21 (16,27) | 26 (28,41) | 21 (17,48) | 17 (9,26) | 0.8 |
|
| |||||||||
| Fick CI Median (IQR) | 2 (1.6, 2.5) | 2.0 (2,4) | 2.2 (4,5) | 2.2 (2,5) | 2.2 (2,5) | *2.5 (3,5) | 2.0 (2,5) | *2.4 (3,6) | 0.02 |
|
| |||||||||
| Fick CO Median (IQR) | 4 (3,5) | 3.7 (1,4) | 4.5 (4,8) | 4.4 (4,7) | 4.5 (4,8) | 4.3 (4,7) | 4.1 (4,8) | 4.0 (4,6) | 0.01 |
Kruskal-Wallis equality-of-populations rank test was run to compare median between groups (Age, body mass index, and serum creatinine)
indicates significant difference between median of the group compare to reference group Mann Whiney Test was run to compare median between each group with reference group DCM= reference group
Abbreviations: CHD=congenital heart disease; CI=cardiac index; CO=cardiac output DCM=dilated cardiomyopathy; ECMO= extracorporeal membrane oxygenation; HCM=hypertrophic cardiomyopathy; HTN=hypertension; IABP=Intra-Aortic Balloon Pump; ICD=implantable cardioverter defibrillator ; ICM=ischemic cardiomyopathy; LVAD=left ventricular assist device; MCS=mechanical circulatory support; mPA=mean pulmonary artery pressure; PADP=pulmonary artery diastolic pressure; PASD=Pulmonary artery systolic pressure; PCWP=pulmonary capillary wedge pressure; RA=right atrial; RCM=restrictive cardiomyopathy; RVAD=right ventricular assist device; TAH=total artificial heart; VHD=valvular heart disease
Dilated cardiomyopathy patients represented the largest subgroup with a median age of 55 years. The majority were men and 21% had durable mechanical circulatory support (ventricular assist devices, and total artificial heart) at time of initial listing for heart transplantation. Ischemic cardiomyopathy represented the second largest subgroup with few women and an older median age compared to the other types of heart disease. Patients in this group also had higher body mass index, higher usage of tobacco and more hypertension compared to other groups. Restrictive cardiomyopathy represented only 6% of the cohort. Compared to all other subgroups, there were few women, more AB blood types, and no temporary or durable mechanical circulatory support. Hypertrophic cardiomyopathy represented 7% of the cohort with a median age of 51 years. They were mostly White, blood type O with no temporary intra-aortic balloon pump or extracorporeal membrane oxygenation and few durable mechanical circulatory support devices. Valvular cardiomyopathy patients represented 3% of the cohort with a median age of 57 years. All were white and none had temporary mechanical circulatory support. Congenital heart disease represented 5% of the cohort with a median age of 47 years. They had no durable mechanical support. “Other” cardiomyopathy represented 4% of the cohort and was the smallest subgroup with the youngest patients.
Accuracy of Pre-transplant Clinical Diagnosis
Pathologic evaluation of the explanted heart was 82% concordant with pre-transplant clinical diagnosis and 18% (n=60) discordant (10% coding errors and 8% incorrect diagnosis). The most common clinical incorrect diagnosis was sarcoidosis (66%), arrhythmogenic right ventricular dysplasia (60%), and other causes of predominately right-sided heart failure (33%).
Dilated cardiomyopathy was clinically incorrect in 3 of 148 patients (2%). Two were diagnosed as restrictive cardiomyopathy suspicious for sarcoidosis based on PET imaging and history of pulmonary sarcoidosis. Another patient with dilated cardiomyopathy was diagnosed as hypertrophic cardiomyopathy due to family history and prior diagnosis of hypertrophic cardiomyopathy. Ischemic cardiomyopathy was clinically incorrect in 6 of 103 patients (6%). Among these 6 patients, 2 patients had known coronary artery disease not suspected to be the cause of the cardiomyopathy. Two others had no significant coronary artery disease by cardiac catheterization less than 2 years prior to listing for transplantation, one was diagnosed with non-ischemic cardiomyopathy prior to referral and another patient was urgently transplanted and not evaluated for an ischemic etiology. All 6 ischemic cardiomyopathy patients who were diagnosed incorrectly had recurrent ventricular tachycardia.
Figure 1 shows pre transplant clinical diagnosis for patients with confirmed pathology for sarcoidosis, arrhythmogenic right ventricular dysplasia, hypertrophic cardiomyopathy, and other causes of predominately right-sided congestive heart failure. There were 13 cases of cardiac sarcoidosis and two cases were correctly identified clinically prior to transplant. Among the 11 other cases, 3 were coding errors, 7 were diagnosed incorrectly as dilated cardiomyopathy, and one diagnosed incorrectly as “other”. Among the incorrectly diagnosed patients, only one had non-invasive testing (PET and MRI) for sarcoidosis. Although the non-invasive testing was suggestive of sarcoidosis, the clinician diagnosed the patient as dilated cardiomyopathy instead of restrictive cardiomyopathy because the left ventricular dimensions by echocardiogram were 8.2 cm in diastole and 6.5 cm in systole. There were 5 patients with arrhythmogenic right ventricular dysplasia including 1 case that was correctly identified pre-transplantation. Among the other cases, one was miscoded, one incorrectly diagnosed as dilated cardiomyopathy, and two incorrectly diagnosed as hypertrophic cardiomyopathy. Among those incorrectly diagnosed, all had an ICD and none had cardiac MRI. There were 18 cases of hypertrophic cardiomyopathy including 10 correctly identified pre-transplantation. Among the others were 6 miscoded, and 2 incorrectly diagnosed as restrictive cardiomyopathy. Finally there were 9 cases of diseases predominately causing right-sided heart failure including 3 correctly identified prior to transplant, 3 miscoded and 3 others incorrectly diagnosed as dilated cardiomyopathy, restrictive cardiomyopathy and congenital heart disease.
Figure 1.
Pre-transplant Diagnosis for Histologic Confirmed Sarcoidosis, Arrhythmogenic Right Ventricular Dysplasia (ARVD), Right Sided Heart Failure (R-CHF), Hypertrophic cardiomyopathy (HCM)
There were 35 coding errors with dilated cardiomyopathy being the most common (37%, n=13) followed by hypertrophic cardiomyopathy (29%, n=10), restrictive cardiomyopathy (12%, n=4), valvular heart disease (11%, n=4) and “Other” (11%, n=4).
Cardiac Biopsies and Accuracy of the Pre-transplant Clinical Heart Disease Diagnosis
Of the 60 patients misclassified (coding errors and incorrect diagnosis), 26 of them had an endomyocardial biopsy or cardiac tissue from an apical core prior to transplantation (Table 3). Most cardiac biopsies (endomyocardial and apical core) showed myocyte hypertrophy with or without interstitial fibrosis. This histologic finding did not help differentiate type of heart disease because the pathologic diagnosis of the explanted whole hearts were hypertrophic cardiomyopathy, ischemic cardiomyopathy, restrictive cardiomyopathy (sarcoid), dilated cardiomyopathy, and “other”. The similar histologic biopsy findings were associated with 11 coding errors (5 hypertrophic cardiomyopathy, 5 dilated cardiomyopathy, 1 restrictive cardiomyopathy). Only one cardiac biopsy with myocyte hypertrophy and fibrosis had myocyte disarray distinguishing it as hypertrophic cardiomyopathy. Although cardiac biopsies led to some misclassification of heart disease, cardiac biopsy was helpful in identifying several cases of sarcoidosis prior to transplantation. Unfortunately, the data for sarcoidosis was not always available at time of listing and not always characteristic because some biopsy specimens did not have non-caseating granulomas present.
Table 3.
Correlation Between Cardiac Biopsy (Apical Core and Endomyocardial) and Pathology of the Explanted Heart Among Patients Misclassified in UNOS Database
| Pathologic Diagnosis of Explanted Hearts Post Transplant | Misclassification | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Cardiac Biopsy Report Pre-Transplant | DCM n= 6 |
RCM * n=6 |
ICM n=6 |
HCM n=6 |
VHD n= 0 |
CHD n=0 |
Other n=2 |
Coding Error | Incorrect Diagnosis |
| Myocyte hypertrophy +/− interstitial fibrosis | 6 | 2 | 4 | 5 | 2 |
5 DCM 5 HCM 1RCM |
7 DCM 1 RCM |
||
| Myocyte hypertrophy with myocyte disarray | 1 | RCM | |||||||
| Healed transmural infarct | 1 | HCM | |||||||
| Fibrocalcific tissue | 1 | DCM | |||||||
| Non-caseating granulomas | 4 |
HCM DCM DCM |
DCM | ||||||
RCM pathologic diagnosis post-transplantation all were sarcoidosis
Abbreviations: CHD=congenital heart disease; DCM=dilated cardiomyopathy; HCM=hypertrophic cardiomyopathy; ICM=ischemic cardiomyopathy; RCM=restrictive cardiomyopathy; VHD=valvular heart disease
Potentially Disadvantaged Patients by the New UNOS Heart Allocation System
Patients potentially disadvantaged with the new heart allocation system are patients misclassified who would be in tier 4 while waiting for transplantation, which corresponds to certain patients with advanced heart failure symptoms or signs in the current UNOS status 2. At listing, 24 of 60 patients (40%) misclassified by coding error or incorrect diagnosis were UNOS status 2 including 3 patients with cardiac sarcoidosis, 1 with arrhythmogenic right ventricular dysplasia, 2 with congenital heart disease, and 5 with hypertrophic cardiomyopathy (Table 4). Among the 24 patients listed as UNOS status 2, 12 were upgraded to UNOS status 1A, 7 to UNOS status 1B, and 5 remained UNOS status 2 at time of transplant. Among the 5 patients transplanted as UNOS status 2, two were misclassified as hypertrophic cardiomyopathy when pathology showed dilated cardiomyopathy and congenital heart disease, another was misclassified as dilated cardiomyopathy when pathology showed sarcoidosis, the fourth was misclassified as restrictive cardiomyopathy when pathology showed dilated cardiomyopathy, and the fifth patient was misclassified as “other” when pathology showed congenital heart disease.
Table 4.
UNOS Status at Listing and Heart Transplantation in Patients with Discordance between Clinical and Explanted Heart Pathologic Diagnoses
| Pathologic Diagnosis of Explanted Hearts Post Transplant
| ||||||||
|---|---|---|---|---|---|---|---|---|
| Total n (%)=60 |
DCM n=15 (25%) |
R-CHF n=7(12%) |
ICM n=13 (22%) |
HCM n=8 (13%) |
CHD n= 2 (3%) |
ARVD n=4 (7%) |
Sarcoidosis n=11 (18%) |
|
|
| ||||||||
| At listing time | ||||||||
|
| ||||||||
| UNOS status | ||||||||
| 1A | 22 (37) | 3 (20) | 3 (44) | 6 (47) | 2 (25) | 0 | 2 (50) | 6 (55) |
| 1B | 14 (24) | 5 (33) | 2 (28) | 3 (23) | 1 (12) | 0 | 1 (25) | 2 (18) |
| 2 | 24 (40) | 7 (47) | 2 (28) | 4 (30) | 5 (63) | 2 (100) | 1 (25) | 3 (27) |
|
| ||||||||
| At time of OHT | ||||||||
|
| ||||||||
| UNOS status | ||||||||
| 1A | 37 (62) | 8 (53) | 4 (58) | 8 (62) | 5 (62) | 0 | 3 (75) | 9 (82) |
| 1B | 18 (30) | 5 (33) | 3 (42) | 5 (38) | 3 (38) | 0 | 1 (25) | 1 (9) |
| 2 | 5 (8) | 2 (12) | 0 | 0 | 0 | 2 (100) | 0 | 1 (9) |
|
| ||||||||
| IABP | 3 (5) | 0 | 0 | 0 | 0 | 0 | 1 (25) | 2 (18) |
|
| ||||||||
| ECMO | 3 (5) | 0 | 1(14) | 1(7) | 0 | 0 | 0 | 1 (9) |
|
| ||||||||
| Inotrope | 26 (43) | 8 (54) | 5 (72) | 2 (15) | 3 (37) | 0 | 2 (50) | 6 (55) |
|
| ||||||||
| MCS | 22 (37) | 5 (33) | 0 | 7 (55) | 5 (62) | 0 | 1 (25) | 4 (36) |
|
| ||||||||
| Intubation | 3 (5) | 0 | 1 (14) | 1 (7) | 0 | 0 | 0 | 1 (9) |
|
| ||||||||
| Exemption | 9 (15) | 1 (6) | 3 (42) | 3 (23) | 0 | 0 | 1 (25) | 1 (9) |
Patients may have more than one indication for high priority
Abbreviations: ARVD= Arrhythmogenic Right Ventricular Dysplasia; CHD=congenital heart disease; DCM=dilated cardiomyopathy; ECMO= extracorporeal membrane oxygenation; HCM=hypertrophic cardiomyopathy; IABP=Intra-Aortic Balloon Pump; ICM=ischemic cardiomyopathy; MCS=durable mechanical circulatory support; RCM=restrictive cardiomyopathy; (R-CHF)=Right sided congestive heart failure; VHD=valvular heart disease
Table 4 also provides documentation for medical priority among those misclassified by coding error or incorrect diagnosis. Although most of the patients with cardiac sarcoidosis had ventricular arrhythmias, only 1 of 11 had an exemption for high priority because of recurrent ventricular arrhythmias. Four other sarcoidosis patients were supported by durable mechanical circulatory support, one by extracorporeal membrane oxygenation, and two by intra-aortic balloon pumps. All patients with arrhythmogenic right ventricular dysplasia were transplanted as high priority. One had an exemption for ventricular arrhythmias and three were supported with an intra-aortic balloon pump, inotropes, or durable mechanical circulatory support. All patients with hypertrophic cardiomyopathy patients were bridged to transplant with durable mechanical support or inotropes. Finally, congenital heart disease and other causes of predominately right-sided heart failure were the only misclassified patients who did not receive durable mechanical support.
Discussion
Principal Findings
In summary, we found a high correlation between clinical diagnosis of heart disease and pathologic confirmation among patients listed for heart transplantation at Cleveland Clinic. However, there were important number of coding errors (10%) and incorrect diagnosis (8%). Misinterpretation of biopsy reports prior to transplantation (endomyocardial biopsy or apical core biopsy from implantation of ventricular assist device) was one of the most frequent causes of coding errors performed by transplant personnel who reviewed electronic medical records. Incorrect diagnosis by clinicians occurred more often with sarcoidosis and genetic diseases like arrhythmogenic right ventricular dysplasia. These errors resulted in the wrong clinical classification in 18% of our cohort and recognition was only possible for most patients after pathologic examination of the explanted hearts post-transplantation.
Importance of Findings
Our study, to the best of our knowledge, is the first to assess the accuracy of the pre-transplant clinical diagnosis of underlying heart disease in the UNOS database. This is very important since the new heart transplant allocation scheme will include type of heart disease as a lower tier with preference given to patients deemed to be at a survival disadvantage such as restrictive cardiomyopathy, hypertrophic cardiomyopathy, complex congenital heart disease and re-transplantation (Table 1). In our study, 24 of 60 patients misclassified (40%) were UNOS status 2 at time of listing. Among those patients, 5 of them were transplanted as UNOS status 2 including one patient with sarcoidosis and another with congenital heart disease that would remain potentially at a disadvantage in the new allocation system. Although few patients appear to be at a disadvantage, this may have a larger significance for rare diseases like sarcoidosis that is difficult to diagnose and often misclassified as dilated cardiomyopathy. Misclassification is understandable given the continuum of cardiac sarcoidosis from a restrictive to a dilated cardiomyopathy. However, it still may be problematic if prognosis does not change with morphology or is different than other type of heart diseases with similar morphologic features.
The SRTR database is a national registry founded in 1987 to evaluate the clinical status of all patients awaiting transplantation in the United States and the status of the recipient after transplantation. Like any large database, it is subject to human error during data entry, which is minimized in UNOS by edit checks and internal verification when there are outliers. Data quality specialists resolve potential problems by reviewing the data and verifying when necessary discrepant data with the involved transplant center.(3) Despite these measures, our research demonstrates that coding errors for primary diagnosis of heart disease are possible and only identified when data entry of heart disease is compared to the clinical and pathologic diagnosis of the explanted heart.
Findings in Context to Other Studies
Incorrect diagnosis occurred in 8% of our cohort and has been reported in a few other studies.(4–6) A clinicopathologic analysis from China (4) reported 38% of their cohort (n=40) were incorrectly diagnosed as dilated cardiomyopathy when the pathologic diagnosis was arrythmogenic right ventricular dysplasia, giant cell myocarditis, and ischemic cardiomyopathy. In a study from Canada (5) comparing pathologic diagnosis of 296 explanted hearts with pre-transplant clinical diagnosis, 51 patients (17%) were misdiagnosed. All cardiac sarcoidosis (n=6), arrythmogenic right ventricular dysplasia (n=12), and iron toxicity-associated cardiomyopathy (n=1) were incorrectly diagnosed as dilated cardiomyopathy and 18% of hypertrophic cardiomyopathy (n=17) were misidentified as dilated cardiomyopathy. Ischemic cardiomyopathy was among the cardiac diseases least likely to be misdiagnosed (only 3% misdiagnosed). Finally, in a single center study from the United States (6) comparing 314 recipient hearts with pre-transplant clinical diagnosis, there were 42 patients (13%) incorrectly diagnosed. All patients with arrythmogenic right ventricular dysplasia (n=4), cardiac sarcoidosis (n=8), and non-compaction (n=3) were incorrectly diagnosed as dilated non-ischemic cardiomyopathy and 41% of hypertrophic cardiomyopathy (n=17) were incorrectly diagnosed. Ischemic cardiomyopathy was once again among the diseases most likely to be correctly identified (96%).
Sarcoidosis and genetic disorders are often difficult to diagnose without pathology. Histological confirmation is often not possible until time of transplant, especially for diseases with patchy infiltration such as sarcoidosis. For instance, Uemera et al. found that endomyocardial biopsy had a 19% sensitivity in a series of 26 patients with cardiac sarcoidosis.(7) Noninvasive testing such as cardiac MRI and PET may be diagnostic but are not required testing, not always possible, and, as demonstrated in our study, not 100% specific (i.e. two patients were misclassified because PET studies were suggestive of cardiac sarcoidosis when histology of the explanted hearts did not confirm this disease). For cardiac MRI, many patients are not eligible because of metallic devices like metal clips and older implantable cardioverter defibrillators. For PET scans, evaluation for sarcoidosis is not always covered by insurance. Consequently, sarcoidosis and genetic diseases like arrythmogenic right ventricular dysplasia remain a dilemma and a concern since they often are misclassified as non-ischemic cardiomyopathy.
Sarcoidosis has a priority level 4 in the new allocation system (2) but arrythmogenic right ventricular dysplasia is not included in this tier despite the subgroup “other” having one of the highest risk of mortality on the national heart transplant waitlist.(1) Does it matter? The subgroup “other” in the UNOS database is mainly composed of re-transplant. Arrthymogenic right ventricular dysplasia is a smaller cohort and associated with ventricular arrhythmias that are given high priority in the allocation system.(2) Is it possible to estimate the effect of excluding this subgroup as a high priority cardiac disease? In our cohort, there were five cases of histologically confirmed arthymogenic right ventricular dysplasia and four were at highest priority based on ventricular arrythmias or need for inotropes/devices. Only one patient with arrythmogenic right ventricular dysplasia was listed as UNOS status 2 and this patient was later transplanted at a higher priority. Although the potential number of arrythmogenic right ventricular dysplasia patients affected appears small, we do not know how many died on the waitlist and were misclassified because our analysis was limited to those who were transplanted.
Solutions to Findings
How can we improve entry of type of heart disease in the UNOS database? Given the difficulties in correctly identifying certain diseases clinically, we propose both a pre-transplant clinical diagnosis (working diagnosis) at time of listing and a pathologic diagnosis of heart disease at time of transplant similar to the policy for liver transplantation (8). We do not imply that UNOS should dictate clinical medical practice and do not advocate for any change in non-diagnostic or invasive testing. Transplant centers also should not be punished for misclassification given the challenge in establishing the clinical diagnosis for some diseases. Our intentions are only to improve database accuracy, because this remains a critical UNOS mission(9). UNOS data is used to understand the outcomes of certain subgroups and utilized to change heart transplant policy.(10) We also suggest some changes and clarification of existing categories in the UNOS database to prevent misclassification. For instance, right-sided heart failure is not a category in the UNOS database but is common with arrhythmogenic right ventricular dysplasia, Chagas disease, and pulmonary hypertension for those listed as dual organ candidates. We suggest that diseases causing predominately right-sided heart failure be categorized as “other” to include patients already classified in this category (i.e. arrhythmogenic right ventricular dysplasia), especially since coders at our center did not know how to handle this entity. Cardiac sarcoidosis remains a dilemma often with multi-focal ventricular arrythmias that are not amenable to ablation and remain difficult to treat. Given the difficulty in establishing the diagnosis, we simply encourage usage of the appropriate tier for their medical urgency including tier 2 for ventricular arrythmias. Valvular and congenital heart diseases need further clarification in the UNOS database to prevent miscoding for a few subgroups like bicuspid aortic valve, atrial septal defect and Ebstein anomaly. At our institution, clinicians labeled bicuspid aortic valve as “valvular disease” yet coders labeled it as “congenital heart disease” because the pathology was present since birth. Ebstein anomaly is another example of ambiguity. This type of heart disease should be classified as “congenital heart disease” and not “valvular disease” or “right-sided heart failure.” To prevent any confusion, all congenital heart disease and the definition of valvular disease should be clarified in the UNOS database. Finally, to better track whether two candidates with similar disease have been classified with the same type of heart disease, pathology of the explanted heart should be documented in the national database.
This verification of heart disease research began as a quality assurance and performance improvement (QAPI) project at our institution. Upon completion, we have identified many areas needing improvement that were not previously known to be a concern. We have learned that a cardiac biopsy with myocyte hypertrophy and fibrosis is non-specific and should not override the pre-transplant clinical diagnosis of the transplant staff. We have also learned that cardiac biopsy was most helpful in confirming cardiac sarcoidosis. Because apical core biopsies are available after implantation of a ventricular assist device that may occur after initial listing for transplantation, we need to update the UNOS database if the pre-transplant clinical diagnosis changes prior to transplantation. Finally, to improve accuracy when entering UNOS data, we intend to enhance communication between coders and staff physicians by establishing the pre-transplant cardiac diagnosis based on the SRTR criteria during our advanced heart failure weekly meeting with the entire team.
Limitations
The main limitation of our study is that it is a single center analysis that may not reflect the practice of other transplant centers. Cleveland Clinic heart transplant program is part of a large tertiary referral center and performs 43–65 heart transplantations/year including lung-heart transplantations. Members of our multi-disciplinary team have distinct roles and responsibilities with little crossover of duties. Across the nation transplant centers vary with regards to heart transplant volume, number of transplant specialists and supporting staff, and diversity of heart disease among patients awaiting transplantation. These differences may affect quality of data entry and accuracy of information. For instance, small transplant centers may not need coders to enter data into UNOS while large centers may require supporting staff to perform this duty creating potential for “coding error”. Tertiary care centers may also have a larger population referred with diseases more difficult to diagnose making them susceptible to incorrect diagnosis. Finally, those centers performing heart-lung transplantation may also be more susceptible for coding errors since right-sided heart failure due to pulmonary hypertension is not pre-specified as a category in the UNOS database.
Conclusions
The new allocation system is an important improvement from the coarse 3-tiered system currently being utilized. Our large single center study supports the fact that many patients will benefit. We found high concordance between type of heart disease entered into the UNOS database at listing and pathologic examination post-heart transplantation. Incorrect diagnosis accounted for 8% of our cohort and was more common among patient with sarcoidosis, arrhythmogenic right ventricular dysplasia, and other causes of predominately right-sided heart failure. Although few patients misclassified may be affected by the new allocation system, patients with rare diseases difficult to diagnose will need to be carefully monitored. To improve the accuracy of the UNOS database, we suggest that pathology of the recipient heart be required post-transplantation and definitions should be clear for each category of heart disease to prevent variance in data entry among centers. In addition, the UNOS database needs to include some new categories for heart disease, defining diseases with predominately right-sided heart failure, and decide whether they should be listed under “other” or as separate categories. Standardizing the definition of all categories of heart disease leading to transplantation will be important for the future and verification will be needed to improve accuracy of the UNOS database.
Clinical Perspectives.
The new heart transplant allocation system will include as a priority type of heart disease based on concern that some patients with cardiac diseases like restrictive cardiomyopathy and congenital heart disease remain at a disadvantage. The accuracy of the diagnosis entered into the UNOS database at listing remained unknown so we reviewed the histology of 334 patients transplanted and found 10% coding error and 8% incorrect diagnosis. Sarcoidosis and genetic diseases like arrhythmogenic right ventricular dysplasia were most likely to be incorrect. Patients with these rare diseases may remain at a disadvantage despite the higher tier for other cardiac diseases. Based on our analysis, we recommend histologic confirmation post-transplantation to improve the accuracy of the UNOS database and some changes in the classification of heart disease to reduce error.
Translational Outlook.
Future research will include correlation between pre-transplant clinical diagnosis and pathologic diagnosis for those who die on the heart transplant waitlist. This will be necessary to determine the full impact of including a tier for heart disease in the new allocation system. Challenges will include obtaining autopsies on all patients who die on the waitlist and limitations in clinical diagnostic testing for sarcoidosis and genetic diseases.
Acknowledgments
Financial Support: Funded by the National Heart, Lung and Blood Institute of the National Institute of Health under Award Number R56HL125420
Abbreviations List
- CHD
Congenital heart disease
- DCM
Dilated cardiomyopathy
- HCM
Hypertrophic cardiomyopathy
- ICM
Ischemic cardiomyopathy
- RCM
Restrictive cardiomyopathy
- SRTR
Scientific Registry of Transplant Recipients
- UNOS
United Network for Organ Shari
- VHD
Valvular heart disease
Footnotes
Relationship with Industry: No relationship with industry
Conflict of Interest Disclosures: None
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Hsich EM, Rogers JG, McNamara DM, et al. Does Survival on the Heart Transplant Waiting List Depend on the Underlying Heart Disease? JACC Heart Fail. 2016;4:689–97. doi: 10.1016/j.jchf.2016.03.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.OPTN/UNOS Thoracic Organ Transplantation Committee. [Last accessed on January 2017];Public Comment Proposal to Modify the Adult Heart Allocation System. Available at https://optn.transplant.hrsa.gov/media/1921/thoracic_adult_heart_allocation_modification_20160815.pdf.
- 3.Dickinson DM, Bryant PC, Williams MC, et al. Transplant data: sources, collection, and caveats. Am J Transplant. 2004;4(Suppl 9):13–26. doi: 10.1111/j.1600-6135.2004.00395.x. [DOI] [PubMed] [Google Scholar]
- 4.Lu FY, Song LF, Liu L, et al. Clinicopathologic analysis of dilated heart in cardiac transplant recipients. Zhonghua Bing Li Xue Za Zhi. 2007;36:796–800. [PubMed] [Google Scholar]
- 5.Roberts WC, Roberts CC, Ko JM, Filardo G, Capehart JE, Hall SA. Morphologic features of the recipient heart in patients having cardiac transplantation and analysis of the congruence or incongruence between the clinical and morphologic diagnoses. Medicine (Baltimore) 2014;93:211–35. doi: 10.1097/MD.0000000000000038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Luk A, Metawee M, Ahn E, Gustafsson F, Ross H, Butany J. Do clinical diagnoses correlate with pathological diagnoses in cardiac transplant patients? The importance of endomyocardial biopsy. Canadian Journal of Cardiology. 2009;25:e48–e54. doi: 10.1016/s0828-282x(09)70484-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Uemura A, Morimoto S, Hiramitsu S, Kato Y, Ito T, Hishida H. Histologic diagnostic rate of cardiac sarcoidosis: evaluation of endomyocardial biopsies. Am Heart J. 1999;138:299–302. doi: 10.1016/s0002-8703(99)70115-8. [DOI] [PubMed] [Google Scholar]
- 8. [Last accessed on April 9, 2017];Organ Procurement and Transplantation Network Policies. Available at https://optn.transplant.hrsa.gov/media/1200/optn_policies.pdf#nameddest=Policy_06.
- 9. [Last accessed on June 11, 2017];OPTN/UNOS Data Vision Statement. Available at https://optn.transplant.hrsa.gov/media/1991/dac_brief_201612.pdf.
- 10. [Last accessed on June 11, 2017];Scientific Registry of Transplant Recipients Mission, Vision, and Value. Available at http://www.srtr.org/about-srtr/mission-vision-and-values/