Hospitalization Rates and Costs of Care of Patients With Sickle-Cell Anemia in the State of Maryland in the Era of Hydroxyurea

Transcription

1 American Journal of Hematology 81: (2006) Hospitalization Rates and Costs of Care of Patients With Sickle-Cell Anemia in the State of Maryland in the Era of Hydroxyurea Sophie Lanzkron, 1 * Carlton Haywood Jr, 2 Jodi B. Segal, 1 and George J. Dover 3 1 Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland 2 Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland 3 Department of Pediatrics, Johns Hopkins University School of Medicine, Baltimore, Maryland The multicenter study of hydroxyurea (MSH) in sickle-cell anemia (SCA) demonstrated that patients treated with hydroxyurea (HU) had a 44% decrease in hospitalizations when compared with those taking placebo. A subsequent study looking at the cost-effectiveness of HU showed that decreased hospitalizations for painful crisis accounted for the majority of cost savings in those taking HU. The purpose of this study was to examine whether the expected decrease in hospital utilization occurred after the approval of HU in Maryland. We used data collected by the Maryland Health Services Cost Review Commission to obtain SCA discharge data for Maryland from FY1995 through FY2003. We also reviewed the inpatient and outpatient charts of all adults with SCA admitted to a large university hospital during Hospitalization rates for adults with SCA in Maryland have increased significantly since approval of HU. While the total costs of inpatient care in Maryland are estimated to have increased by 31% above inflation from 1995 to 2003, the costs of inpatient care for adult SCA patients has increased by almost 60% above inflation. By comparison, there has been no significant increase in the pediatric hospitalization rate. We found that 70% of patients in one hospital who were appropriate candidates for HU were not taking the medication. Hospital utilization among adults with SCA has increased significantly. There are likely many factors that have played a role in this increase. One factor that appears to be involved is the underutilization of HU. Am. J. Hematol. 81: , VC 2006 Wiley-Liss, Inc. Key words: sickle; costs; hydroxyurea INTRODUCTION The multicenter study of hydroxyurea for sicklecell anemia (MSH) was a randomized, placebo controlled study that demonstrated that patients with sickle-cell anemia (SCA) treated with hydroxyurea (HU) developed fewer complications than those who received placebo [1]. MSH demonstrated that in patients having three or more crises per year, HU decreased the number of crises by 44%, decreased the number of patients in whom chest syndrome developed and decreased the number of transfusions required. As a result of MSH, in 1998, HU became the first FDA approved medication for the treatment of patients with SCA. The current recommendations for the use of HU in patients with SCA are described in the most recent version of The Management of VC 2006 Wiley-Liss, Inc. Sickle-Cell Disease published by the National Heart, Lung, and Blood Institute (NHLBI) [2]. Contract grant sponsor: National Research Service Award, Minority Predoctoral Fellowship Program, NIH, NHLBI; Contract grant number: 1 F31 HL ; Contract grant sponsor: Johns Hopkins School of Medicine. *Correspondence to: Sophie Lanzkron, MD, Johns Hopkins University, 1830 E. Monument Street, Suite 7300, Baltimore, MD slanzkr@jhmi.edu Received for publication 16 November 2005; Accepted 17 May 2006 Published online 21 August 2006 in Wiley InterScience (www. interscience.wiley.com). DOI: /ajh.20703

2 928 Lanzkron et al. In 2000, Moore et al. examined the cost-effectiveness of HU based on the clinical effect that was demonstrated in the MSH [3]. Using the data from the MSH, the investigators applied cost estimates to each unit of resource utilization. Hospitalization for painful crisis accounted for the majority of costs in both arms of the study. The investigators concluded that the cost savings from reducing hospitalizations and other clinical care costs easily offset the increased costs of monitoring of HU. The investigators noted that if every eligible patient in the US were taking HU, at a savings of $5,210 per patient per year of medical care costs, this could potentially lead to a total cost savings of almost $26 million. Given this observation, we examined three measures of hospital utilization to determine whether there has been a decrease in hospital utilization for adults with SCA in Maryland since the FDA approval of HU. First, we examined trends in adult SCA hospitalization rates from fiscal year 1995 (FY1995) to fiscal year 2003 (FY2003). We also examined trends in the total hospital costs incurred by adult and pediatric SCA patients in Maryland, and we examined trends in the mean length of stay (LOS) for these patients over the time period. Finally, we examined in more detail the characteristics of adults with SCA hospitalized in one university hospital in FY2003. This provided a description of a patient population using HU outside of a clinical trial setting. METHODS Data Collection We used data collected by the Maryland Health Services Cost Review Commission (MHSCRC) to obtain SCA discharge data for the State from FY1995 through FY2003. The MHSCRC database collects annual statistics on the number of hospital discharges by hospital. Data collected include patient demographic characteristics, primary payer, secondary payer, principal diagnosis and other diagnoses (total of 15), length of stay, complexity, condition code, and charges. The database includes all hospitals in the State, except for Veterans Affairs facilities. We used International Classification of Disease-9-Clinical Modification codes (ICD-9-CM) codes recorded in the dataset to identify SCA inpatient data. We used the codes (sickle-cell Hgb SS without crisis) and (sickle-cell Hgb SS with crisis) in any of the 15 listed diagnoses to identify SCA admissions. We calculated the annual hospitalization rate as the number of discharges that included a diagnosis of SCA divided by the number of people estimated to be living with SCA in Maryland during the same year. This estimate was generated using 1990 and 2000 US census data [4] for an estimate of the number of African Americans in Maryland, and an estimated SCA prevalence of 1 in 600 African Americans [5]. To describe the total costs of hospitalized patients, we used the MHSCRC data and looked at total costs. This database breaks down costs as follows; routine charges (bed charges), operating room, pharmacy, radiology, laboratory, supplies, therapy, and other costs. To describe the characteristics of patients who were admitted to the university hospital in FY2003, we reviewed the inpatient and outpatient charts of all adults with sickle-cell disease (SCD) admitted in that year to Johns Hopkins Hospital (JHH), identified with the ICD-9-CM codes described earlier. We compared the data for SCA patients with fewer than three admissions to that of SCA patients with three or more admissions to determine whether there were differences between these two populations of patients. We used the indications described in the NHLBI management guide to identify individuals that would have been eligible for HU use [2]. Those indications include patients with hemoglobin SS or Sb 0 thal with frequent pain episodes, acute chest syndrome, severe vasoocclusive events, or severe symptomatic anemia. Statistical Methods For the statewide analyses, Poisson regression techniques were used to test the statistical significance of the trends in SCA hospitalization rates in Maryland over the time period. Negative binomial regression models were used as appropriate for the hospitalization rate analyses when suggested by the results of goodness-of-fit tests. Ordinary least squares (OLS) regression models were used to test the trends in the mean length of stay (LOS) for all Maryland SCA patients over the time period, and to test the trend in SCA expenditures as a percentage of total Maryland hospital expenditures for the time period. We compared changes in the total cost of hospital care for SCA patients in Maryland in 1995 with those in 2003 by converting 1995 SCA costs into 2003 dollars using an inflation factor of 1.12 that we obtained from the producer price index. We then divided the cost of care in 2003 by the converted 1995 figure. The statistical package used was STATA 8.2 (StataCorp, College Station, TX). For the JHH FY2003 patient comparisons, we used t-tests, Mann-Whitney U tests, and chi-square tests as appropriate to test differences in gender, age, mean LOS, and mean costs per hospitalization between the patients with less than three and the patients with three or more admissions in the year.

3 Costs of Care of Patients With SCA in Maryland 929 TABLE I. Discharge Data for Adults With Sickle-Cell Anemia in Maryland (MD): 1995 Through 2003 Year of patients with SCA (no.) Estimated no. of adults with SCA per person * ALOS ** Average cost per hospitalization per patient ($) *** Total hospital costs for adults with SCA ($) Percentage of total hospital expenditures for adults in MD **** ,584 1, ,525 8,751, ,740 1, ,983 10,411, ,725 1, ,792 9,991, , ,161 10,658, ,858 1, ,934 11,025, ,797 1, ,803 10,428, ,940 1, ,950 11,542, ,203 1, ,357 14,004, ,355 1, ,656 15,675, SCA, sickle-cell anemia; ALOS, average length of stay in days; Tests for trend: *P < 0.001; **P < 0.001; ***P ¼ 0.016; ****P ¼ RESULTS Maryland Hospital Discharge Data Admission data for adults with SCA in Maryland from FY1995 through FY2003 are presented in Table I. There was a statistically significant increase in the absolute number of discharges for adults with SCA over the time period studied. Additionally, the hospitalization rate for adult African Americans with SCA increased from in FY1995 to 1.29 in FY2003 (test for trend P < 0.001). The average length of hospital stay decreased from 6.16 days in FY1995 to 4.99 days in FY2003 (P < 0.001). The percentage of total statewide hospital costs due to adult SCA discharges increased from 0.25% in FY1995 to 0.30% in FY2003, though this increase was not statistically significant (P ¼ 0.200). The cost of care for adults with SCA in FY2003 was 60% higher than FY1995 costs even after adjusting for inflation (Table II). In comparison, total statewide adult hospital costs in FY2003 were 31% higher than FY1995 costs after adjusting for inflation. When we looked at the breakdown in costs, there was no one category that experienced any drastic change in its relative contribution to overall adult SCA costs (Fig. 1). We examined similar data for pediatric discharges and costs over the same time period. In children, there were no significant changes in hospitalization rates or mean LOS over the time period studied (Table III). There was a 40% increase in costs for pediatric SCA discharges in FY2003 compared to adjusted FY1995 costs, while total state-wide pediatric hospitalization costs in FY2003 were 20% higher than adjusted FY1995 costs (Table III). University Hospital Discharge Data Data from patients admitted in FY2003 are shown in Table IV. In FY2003, there were 74 patients with TABLE II. Increase in Hospitalization Costs in Maryland From 1995 to 2003: Adults and Children With SCA 1995 Costs a 2003 Costs Ratio of 2003 cost to adjusted 1995 cost Adults with SCA 9,802,012 15,675, All adults 3,948,840,390 5,178,470, Children with SCA 1,951,089 2,720, All children 375,622, ,037, SCA, sickle-cell anemia. a In 2003 dollars. SCA regularly treated as outpatients at the sicklecell center within this hospital. Fourteen (19%) were taking HU and six of these 14 patients were hospitalized in FY2003. In total, 52 of the 74 SCA patients were admitted with a total of 676 hospital days. Thirty-eight patients (73%) had fewer than three admissions, while 14 patients (27%) were admitted three or more times. This 27% of patients accounted for 60% of the total hospital days for all patients with SCA, and 55% of total SCA hospital charges. The mean age of those individuals admitted in FY2003 was 31.2 with a median of 27.5 years. The mean and median ages of adults admitted to this institution in FY1995 for SCA was similar. There was no statistical difference in the mean and median ages of those patients admitted less than three times, and those patients admitted three or more times. Of the 14 patients admitted three or more times, one was admitted for complications related to surgery, one was admitted with wound infection, and two were pregnant. These patients, therefore, were not candidates for HU. Of the 10 remaining patients, all were admitted with vasoocclusive crisis (VOC) and only three of these patients were on HU. Of the 38 patients admitted fewer than three times, 29 were admitted at least once with a VOC. Only three

4 930 Lanzkron et al. Fig. 1. Breakdown of total hospital costs 1995 compared to 2003 for adults with sickle-cell anemia. There was no one category that experienced any drastic change in its relative contribution to overall costs. TABLE III. Discharge Data for Children With SCA in Maryland: 1995 Through 2003 Year of children with SCA (no.) Estimated no. of children with SCA per child ALOS Total hospital costs for children with SCA ($) % Total hospital expenditures for children in Maryland ,742, ,935, ,898, ,538, ,225, ,193, ,635, ,342, ,720, SCA, sickle-cell anemia; ALOS, average length of stay in days. patients in this group were taking HU and they were admitted with VOC. There was no difference in mean costs per admission or in length of stay per admission between those admitted three times or more and those admitted fewer than three times. Using the NHLBI management guide criteria for HU use, 70% of the patients admitted with SCA would have been eligible for HU therapy, but were not taking it. The reasons that individuals were not taking HU varied, four of seven had no regular outpatient follow-up. Two had adherence issues related to side effects, and one patient started HU during FY2003. This patient had six admissions prior to starting HU and two after. DISCUSSION The number of adult SCA discharges in Maryland has significantly increased over the time period studied. The average length of stay for adults with SCA has significantly decreased over this same time period. While the total costs of inpatient care in Maryland are estimated to have increased by 31% above inflation from 1995 to 2003, the costs of inpatient care for adult SCA patients has increased by almost 60% above inflation. By comparison, there has been no significant increase in the number of pediatric discharges over the same time period. At an estimated increase of 40% above inflation from 1995 to 2003, the costs of inpatient care for pediatric SCA patients, while not as large of an increase as adult hospitalizations, is still nearly twice the increase estimated in the general pediatric population in Maryland. There are many potential explanations for the increase in hospital utilization. One explanation may be that because patients are now taking HU they are living longer and are admitted for other complications of their disease. This would be similar to the increase in hospital admissions that has been seen for people with congestive heart failure that is thought to be due, in part, to improvements in survival of patients after acute myocardial infarction [6,7]. Improved survival and an aging population is not likely the cause of increased hospital utilization in our university hospital setting, as the median age of patients did not change between FY1995 and FY2003. Another possibility is that fewer specialists familiar with the outpatient treatment of patients

5 Costs of Care of Patients With SCA in Maryland 931 TABLE IV. Characteristics of Adult Patients With SCA Admitted to an Urban University Hospital in 2003 All SCA patients 3 admissions/year <3 admissions/year Total patients (%) (27) 38 (73) Total hospital days (%) (60) 270 (40) Total Admits (%) (58) 50 (42) Total charges (%) (55) (45) Number on HU (%) 6 (12) 3 3 Mean age (median) 31.2 (27.5) 28.2 * (25.5) ** 32.3 * (30) ** SCA, sickle-cell anemia; HU, hydroxyurea P-values for differences in age. *P ¼ 0.12; **P ¼ with SCD are available, leaving patients to seek their care in the hospital setting. There is no published data that describes from which providers adults with SCA in Maryland receive their care. Although underutilization of HU alone cannot explain the increase in hospitalization, it would appear that HU has not had its expected impact. The reasons for this are unclear. It is possible that many eligible patients who should be receiving HU have chosen not to take it or were not prescribed HU due to concerns over side effects. It is also possible that the efficacy of HU seen in MSH is not representative of the effectiveness of HU when used in a community setting that includes patients who, for example, are using significant amounts of narcotics on a daily basis or find it difficult to comply with the required biweekly blood test monitoring. These patients would not have been eligible for MSH. Our data demonstrate that rather than HU showing a lack of efficacy in our eligible patients, those admitted to our university hospital were not taking the medication for a variety of reasons. Our data were similar to that published by Ballas and Lusardi, in which they reported that only 21 of 182 patients in their center were taking HU [8]. In another recent report by Zumberg et al., in a survey of both community and academic hematologist/ oncologists in Florida and North Carolina, almost half of the physicians prescribed HU to less than 10% of their patients [9]. They found that concerns over carcinogenicity, the effectiveness of HU and teratogenicity were barriers to prescribing HU. Several studies are currently examining the use of HU in younger people with SCD (i.e., pediatric hydroxyurea in SCA or BABY HUG study). If these studies confirm a benefit of HU in the pediatric patient population, understanding the barriers to HU utilization will be essential to use of this medication outside the setting of a clinical study. Reasons unrelated to the use of HU might also have contributed to increased hospitalization rates. Changes in the economy may have left more patients without medical coverage and therefore more patients were receiving their primary care through hospitals. Insurance companies may have changed patterns of reimbursements, pushing caregivers to discharge patients earlier in the course of a vasoocclusive crisis, thus leading to early readmissions for relief of pain. Other hospitals have seen the rate of hospitalization decrease over the last decade due to the development of day hospitals that treat patients with crises as outpatients [10]. In Maryland, no such day hospital exists. The experience in Maryland likely represents that of similar states that do not have day hospitals to treat patients. A limitation of our study is that in order to develop a hospitalization rate we had to estimate the number of individuals in Maryland who had SCA. The basis of this estimate was a prevalence of SCA of one in 600 African Americans. The accuracy of this estimate is unclear, but we assume that over the 9 years of this study the prevalence did not change significantly. Furthermore, we are modeling the relative change in hospitalization rates over time. While our estimate of the number of adults with SCA in Maryland will differ with different prevalence rates, the trend in hospitalization rates remains the same. Another limitation is that we could not track individual patients through the State database. We are therefore unable to assess the impact that improved survival of pediatric patients may have had. If more children survived and then were admitted as adults, over the 9 years we studied, this influx of patients may have contributed to an increase in hospitalization rates. If this were the case, however, we should have seen an increase in the rate of pediatric hospitalizations as well. The significant increases in costs are intriguing. We know that, in general, health care costs in the U.S. increase beyond the rate of inflation [11 13]. No one cost category drastically increased its relative contribution to overall adult SCA costs, so we conclude that the most likely explanation for the increase in cost of care for adult SCA patients is explained by many more adult SCA visits combined with the high rate of medical inflation. The differ-

6 932 Lanzkron et al. ence in the pediatric SCA population, compared with that in the general pediatric patients, is more difficult to understand, as no increase in hospitalization rates were seen in the pediatric SCA group. Further investigation into what drives the cost of care for patients with SCA is warranted. We aimed to understand the impact of hydroxyurea on health care costs outside the setting of a clinical study. To further understand hospital utilization in this patient population, an evaluation of data from across the US would help to determine whether there are regions of the country where improved delivery of care has led to a decrease in hospitalization rates. We cannot conclude that the availability of HU has reduced costs of care in Maryland over the past decade; studies aimed at understanding barriers to its use are vitally needed. REFERENCES 1. Charache S, Terrin ML, Moore RD, et al. Effect of hydroxyurea on the frequency of painful crises in sickle cell anemia. Investigators of the multicenter study of hydroxyurea in sickle cell anemia. N Engl J Med 1995;332: The Management of Sickle Cell Disease, 4th ed. Bethesda, MD: NHLBI; p 165. NHLBI No Moore RD, Charache S, Terrin ML, Barton FB, Ballas SK. Costeffectiveness of hydroxyurea in sickle cell anemia. Investigators of the multicenter study of hydroxyurea in sickle cell anemia. Am J Hematol 2000;64: American FactFinder. U.S. Census Bureau; American Fact- Finder Beutler E. The sickle cell diseases and related disorders. In: Beutler E, Lichtman M, Coller B, Kipps T, editors. Williams Hematology. New York: Mcgraw-Hill; p Berry C, Murdoch DR, McMurray JJV. Economics of chronic heart failure. Eur J Heart Fail 2001;3: Lowery SL, Massaro R, Yancy CW Jr. Advances in the management of acute and chronic decompensated heart failure. Lippincotts Case Manag 2004;9:S1 S Ballas SK, Lusardi M. Hospital readmission for adult acute sickle cell painful episodes: frequency, etiology, and prognostic significance. Am J Hematol 2005;79: Zumberg MS, Reddy S, Boyette RL, et al. Hydroxyurea therapy for sickle cell disease in community-based practices: a survey of Florida and North Carolina hematologists/oncologists. Am J Hematol 2005;79: Benjamin LJ, Swinson GI, Nagel RL. Sickle cell anemia day hospital: an approach for the management of uncomplicated painful crises. Blood 2000;95: Health costs still increasing far beyond inflation. Med Health 2005;59:1,6, Jaklevic MC. Runaway costs. Healthcare inflation at highest level in 7 years. Mod Healthc 2002;32:12, Evans M. Healthcare inflation plateaus. Some had hoped growth rate would be falling. Mod Healthc 2005;35:8 9.