Inpatient Costs and Charges for Surgical Treatment of Hypoplastic Left Heart Syndrome

Abstract

OBJECTIVE:

Hypoplastic left heart syndrome (HLHS) is one of the most serious congenital cardiac anomalies. Typically, it is managed with a series of 3 palliative operations or cardiac transplantation. Our goal was to quantify the inpatient resource burden of HLHS across multiple academic medical centers.

METHODS:

The University HealthSystem Consortium is an alliance of 101 academic medical centers and 178 affiliated hospitals that share diagnostic, procedural, and financial data on all discharges. We examined inpatient resource use by patients with HLHS who underwent a staged palliative procedure or cardiac transplantation between 1998 and 2007.

RESULTS:

Among 1941 neonates, stage 1 palliation (Norwood or Sano procedure) had a median length of stay (LOS) of 25 days and charges of $214 680. Stage 2 and stage 3 palliation (Glenn and Fontan procedures, respectively) had median LOS and charges of 8 days and $82 174 and 11 days and $79 549, respectively. Primary neonatal transplantation had an LOS of 87 days and charges of $582 920, and rescue transplantation required 36 days and $411 121. The median inpatient wait time for primary and rescue transplants was 42 and 6 days, respectively. Between 1998 and 2007, the LOS for stage 1 palliation increased from 16 to 28 days and inflation-adjusted charges increased from $122 309 to $280 909, largely because of increasing survival rates (57% in 1998 and 83% in 2007).

CONCLUSIONS:

Patients with HLHS demand considerable inpatient resources, whether treated with the Norwood-Glenn-Fontan procedure pathway or cardiac transplantation. Improved survival rates have led to increased hospital stays and costs.

WHAT'S KNOWN ON THIS SUBJECT:

Hypoplastic left heart syndrome (HLHS) is a serious congenital heart defect that requires 3 staged palliative operations or cardiac transplantation. Mortality rates have improved significantly over the past decade, but patients continue to consume extensive inpatient hospital resources.

WHAT THIS STUDY ADDS:

Improved survival rates for patients born with HLHS have led to increased hospital stays and costs. Whether treated with 3 staged palliative operations or transplantation, patients require resources similar to those needed for prematurity or other major birth defects.

Hypoplastic left heart syndrome (HLHS) affects 1 in 5000 live births ¹ and is defined by underdevelopment of the left ventricle, aortic and mitral stenosis or atresia, and hypoplasia of the ascending aorta. ² HLHS is considered one of the most serious congenital heart defects and, without intervention, the defect is typically fatal in the first week of life. ²

In the 1980s, staged palliative repair and cardiac transplantation arose as treatment options for neonates born with HLHS. Staged palliative repair involves 3 cardiac procedures that typically are performed before school age. Stage 1 palliation (S1P), which is performed in the first few days of life, involves the creation of a neo-aorta (Norwood procedure ³ ) and shunt placement to provide pulmonary blood flow (Blalock-Taussig shunt or Sano procedure ⁴ ). Stage 2 palliation (S2P), which is performed between 4 and 6 months of age, involves formation of a superior vena cava-pulmonary artery connection (Glenn ⁵ or hemi-Fontan ⁶ procedure). Stage 3 palliation (S3P), which is performed at 2 to 4 years of age, involves formation of a total cava-pulmonary artery connection (Fontan procedure ⁷ ). Cardiac transplantation most commonly is used as “rescue” therapy if these procedures do not yield acceptable results, but it also has been used as primary treatment, as an alternative to the staged palliation pathway.

Outcomes have improved greatly over the past 3 decades, which has allowed many patients to survive into adulthood. ^{8 – 11} These improvements have led to more patients undergoing surgical palliation and fewer patients being offered or requesting comfort care. ^{12 – 15} We suspect that this change has caused increases in the resources and dollars spent on children born with HLHS.

There have been analyses of the costs of initial hospitalization for patients with HLHS ^16,17 and single-center costs, ^18,19 but we are not aware of any multicenter studies analyzing the costs of the staged palliation pathway and cardiac transplantation. Our goal was to determine resource use by patients with HLHS across multiple academic medical centers throughout the United States in the current era.

METHODS

The University HealthSystem Consortium (UHC) is an alliance of 101 academic medical centers and 178 affiliated hospitals sharing information, including demographic, diagnostic, procedural, and financial data for each hospital admission. Encoded hospital identifiers were provided for individual institutions.

The database was queried for data on all patients who were discharged between 1998 and 2007 with the diagnosis of HLHS. Patients were deemed to have HLHS on the basis of International Classification of Diseases, Ninth Revision, Clinical Modification, code 746.7 being included as any one of their coded diagnoses. Procedures during each hospitalization were determined by using the listed procedure codes, as described previously. ⁸ Patient mortality rates were determined on the basis of survival to discharge.

Patients undergoing transplantation were divided into 2 categories, namely, primary transplantation and rescue transplantation. Patients were considered to have undergone primary transplantation if they were admitted to the hospital at younger than 30 days and the only cardiac procedure coded was cardiac transplantation (code 37.5). Patients were considered to have undergone rescue transplantation if they were admitted when they were older than 30 days and had a procedure code for cardiac transplantation.

To determine whether average daily hospital costs were related to surgical volume, average costs for small-, medium-, and large-volume institutions were compared through 1-way analysis of variance (SPSS 18 [SPSS Inc, Chicago, IL]). Institutions that performed <20 procedures during the studied time period were considered small volume, those that performed 21 to 64 procedures were considered medium volume, and those that performed >64 procedures were considered large volume, as in our study of surgical mortality rates. ⁸

Hospital length of stay (LOS) and hospital charges were reported by each hospital. Hospital costs are estimated by the UHC by using Medicare cost/charge ratios and further adjustments in labor costs according to geographic area. LOS data were available for all patients, whereas charge and cost data were available for most. We extracted this information for each hospitalization during which patients underwent S1P, S2P, S3P, or cardiac transplantation. We did not include admissions that did not include 1 of the aforementioned procedures; however, it is likely that hospitalizations before or after the surgical admissions were related to the surgical procedures, which would result in underestimation of the full inpatient costs of those procedures. To adjust for inflation, dollar amounts were converted to 2007 equivalents by using the Consumer Price Index inflation calculator from the US Bureau of Labor Statistics. ²⁰

RESULTS

Between 1998 and 2007, there were 9197 discharges of 5416 patients with HLHS who were admitted to a UHC-affiliated hospital. There were 3681 admissions of patients younger than 1 month, 2749 of patients 1 month to 1 year of age, and 2750 of patients older than 1 year. Data were available for 1941 neonates undergoing S1P, 936 undergoing S2P, and 839 undergoing S3P. Primary cardiac transplantation was performed for 28 neonates, and rescue cardiac transplantation was performed for 62 patients.

The LOS, costs, and charges were greatest for S1P (Tables 1 and 2). During the decade studied, the median LOS, costs, and charges were 25 days, $99 070, and $214 680, respectively, for S1P. From 1998 to 2007, the LOS for S1P increased from 16 days to 28 days and inflation-adjusted costs increased from $65 041 to $113 827. Costs and charges increased progressively between 1998 and 2007 (Fig 1). This was likely attributable to increasing survival rates (rates were 57% in 1998 and 82% in 2007), because the median LOS for survivors was 28 days, compared with 13 days for patients who died. The median LOS for the 556 neonates who died after surgery was only 13 days, compared with 28 days for the 1265 neonates who were discharged.

TABLE 1.

LOS for HLHS

	n	LOS, Median (Range), d
S1P (Norwood or Sano procedure)	1941	25 (0–312)
S2P (Glenn procedure)	936	8 (1–326)
S3P (Fontan procedure)	840	11 (0–373)
Primary transplantation	28	87 (16–285)
Rescue transplantation at >30 d	62	36 (2–255)

TABLE 2.

Costs and Charges of 3-Stage Palliation and Cardiac Transplantation for HLHS

	Median	Mean	Range
Costs, $
S1P (Norwood or Sano procedure)	99 070	135 278	6499–2 025 387
S2P (Glenn procedure)	35 674	65 962	1252–992 958
S3P (Fontan procedure)	36 928	61 058	2712–1 143 363
Primary transplantation	289 292	361 724	80 854–1 118 266
Rescue transplantation at >30 d	222 509	283 398	16 181–1 064 411
Charges, $
S1P (Norwood or Sano procedure)	214 680	335 457	9932–6 856 341
S2P (Glenn procedure)	82 174	164 827	3340–2 401 738
S3P (Fontan procedure)	79 549	146 186	7033–4 407 888
Primary transplantation	582 920	827 574	182 058–4 024 679
Rescue transplantation at >30 d	411 121	683 566	26 720–3 781 342

FIGURE 1.

Mean costs and charges of S1P between 1998 and 2007.

The estimated costs for hospitalization for S1P remained constant at $4065 per day from 1998 to 2007, but charges increased from $7644 to $10 032 per day. Average daily costs were similar for small-, medium-, and large-volume institutions ($5148 ± 2382, $4889 ± 2976, and $4995 ± 2507, respectively; P > .05). Several institutions exhibited marked variability of costs because of very high resource utilization for a few patients with very short hospital stays.

The LOS, costs, and charges for S2P (Glenn procedure) and S3P (Fontan procedure) were similar and were approximately one-third of those for S1P (Tables 1 and 2). The cumulative median LOS, costs, and charges for 3-stage palliation were 44 days, $171 672, and $376 403, respectively.

Cardiac transplantation was associated with the greatest use of resources (Tables 1 and 2). Primary transplantation as an alternative to S1P had LOS, costs, and charges of 87 days, $289 292, and $582 920, respectively. Rescue transplantation was associated with a shorter hospital stay (36 days), costs were 23% lower ($222 509), and charges were 29% lower ($411 121). The median number of inpatient days before a heart became available was 42 days (range: 0–164 days) for primary transplantation and 6 days (range: 0–145 days) for rescue transplantation.

The average costs for the procedures are skewed by the very long hospitalizations of a few patients. For example, the median costs and charges for S1P were $99 070 and $214 680, respectively, but the averages were $135 278 and $335 457. Charges exceeded $1 million for 3% of the procedures in the palliative pathway, 21% of primary transplants, and 25% of rescue transplants.

DISCUSSION

By using a large multicenter database, we determined the LOS, costs, and charges for surgical procedures for HLHS across the United States during a recent 10-year period. We demonstrated the significant inpatient financial and resource demands made by patients with HLHS, whether treated with staged palliative operations or transplantation. As expected, S1P demanded more resources than S2P or S3P. Interestingly, improved survival rates, measured as survival to discharge, for S1P over the decade resulted in increased LOS, costs, and charges. Cardiac transplantation (primary or rescue) was more costly than any of the 3 palliative operations. As noted by Gajarski et al, ¹⁸ however, palliative surgery followed by transplantation was the most costly.

Our previous analysis of this database demonstrated a decrease in mortality rates for S1P, from 43% in 1998 to 18% in 2007. ⁸ Others also demonstrated mortality rates ranging from 7% to 26%, ^{21 – 23} with significant improvement over the past decades. ^10,11 The recent report from the Pediatric Heart Network showed either death or a need for transplantation for 26% and 36% of survivors of S1P with the Sano or Norwood technique, respectively. ²⁴ Rates of survival after S2P and S3P were significantly better than that after S1P, and mortality rates ranged from 2% to 5% for S2P ^8,25,26 and from 2% to 4% for S3P. ^{8,27 – 29}

The decrease in mortality rates for S1P has come with significant increases in financial and resource costs. Despite adjustment for inflation, the costs, charges, and LOS after S1P were considerably greater in 2007 than in 1998. This increase is largely attributable to the increase in LOS produced by the increased survival rate. Interestingly, between 1998 and 2007 the charges per day increased for each stage, whereas the estimated costs per day remained constant (Table 3). Hospital costs seemed to be unrelated to institutional surgical volume.

TABLE 3.

Costs and Charges for HLHS in 1997 and 2008

	S1P		S2P		S3P
	1998	2007	1998	2007	1998	2007
LOS, d	16	28	7	9	10	11
Costs, median, $	65 041	113 827	25 183	44 067	27 286	39 611
Charges, median, $	122 309	280 909	57 527	112 197	42 026	97 165
Costs per d, $	4065	4065	3598	4896	2729	3601
Charges per d, $	7644	10 032	8218	12 466	4203	8833
Cost/charge ratio	0.53	0.41	0.44	0.39	0.65	0.41
Mortality rate, % 8	43	18	5	3	3	3

Although the resource utilization for surgical management of HLHS is considerable, it is comparable to that for other major congenital malformations, as well as prematurity ^30,31 (Table 4). It also is similar to that of highly technical interventions such as extracorporeal membrane oxygenation and ventricular assist device placement. ^{32 – 34}

TABLE 4.

Costs for HLHS Compared With Other Pediatric Diseases

	LOS, Mean, d	Hospital Costs, Mean, $
S1P (Norwood or Sano procedure)	34	135 278
S2P (Glenn procedure)	17	65 962
S3P (Fontan procedure)	17	61 058
Extracorporeal membrane oxygenation		156 324 32
Ventricular assist device implantation	29–50	174 743–624 798 33,34
Premature infant, 25 wk	87–98	150 000–202 700 30,31
Premature infant, 28 wk	52–57	65 000–119 600 30,31
Omphalocele	32	141 724 17
Gastroschesis	41–47	123 200–155 629 17,35
Congenital diaphragmatic hernia	25	162 700 17
Myelomenigocele	12–15	49 602–65 342 17,41,42

The data presented include only the inpatient hospital costs and charges for the major surgical procedures used in the management of HLHS and thus represent only a fraction of the total costs of this malformation. Inpatient physician charges were not included in our study and might be expected to add 15% to these costs. ³⁵ Additional hospitalizations are frequent; it should be noted that our database included 9197 admissions for patients with HLHS but only 3806 admissions involved the 5 surgical procedures analyzed here. Survivors of the surgical procedures face a future of lifelong outpatient visits, cardiac catheterizations, imaging studies, and medication use. A portion of those on the 3-stage palliative pathway will ultimately require cardiac transplantation along with the requisite repeated myocardial biopsies and lifelong immunosuppression. After completion of the Fontan procedure, patients remain with long-term complications and have a 10% to 13% chance of death or transplantation within 10 years. ^36,37 Despite the perioperative and long-term complications after the Fontan procedure, however, the majority of patients are able to lead relatively normal lives and are satisfied with their quality of life. ^{38 – 40}

HLHS is an example of a disease entity that is changing the financial landscape of health care in the United States. Before the middle 1980s, children born with HLHS had minimal health care resource use because the only available option was comfort care and most patients died shortly after birth. With the success of surgical palliation, comfort care is offered less often and most parents now are counseled that surgery is a feasible option. ^{12 – 15} As the operative and perioperative care of patients with HLHS continue to improve and mortality rates improve, we predict that costs, charges, and hospital stays will continue to increase.

There are several limitations to this study. The UHC database is an administrative database with the inherent limitations of such data sources. We were able to analyze only single hospitalizations, and we were unable to track patients across subsequent hospitalizations or as outpatients. Data on survival to discharge were the only outcome data available; therefore, we were unable to assess quality of life or rates of survival after discharge. Actual reimbursement rates were not available and undoubtedly vary widely depending on third-party payer agreements. Cost-effectiveness calculations were not attempted because of the long-term uncertainty of survival for patients with a single right ventricle and the frequent neurodevelopmental deficits found among patients who have undergone staged palliative operations. In addition, the considerable related inpatient, outpatient, and physician costs were not available for this population, which indicates that these results represent a conservative estimate of the total costs of treating HLHS.

CONCLUSIONS

We present the results of a large multicenter study that examined the inpatient costs of HLHS in the current era. Improved survival rates have led to increased hospital stays and costs. Whether treated with 3-stage palliation or transplantation, patients with HLHS demand extensive and expensive inpatient resources, similar to those for prematurity or other major birth defects.