Hospital Costs for Acute Myocardial Infarction Patients Receiving

Document technical information

Format pdf
Size 740.5 kB
First found May 22, 2018

Document content analysis

Category Also themed
Language
English
Type
not defined
Concepts
no text concepts found

Persons

Donald Berwick
Donald Berwick

wikipedia, lookup

Organizations

Places

Transcript

Regis University
ePublications at Regis University
All Regis University Theses
Fall 2008
Hospital Costs for Acute Myocardial Infarction
Patients Receiving Perfect Compliance of
Evidence-Based Care Bundle
Jill S. McCormick
Regis University
Follow this and additional works at: http://epublications.regis.edu/theses
Part of the Medicine and Health Sciences Commons
Recommended Citation
McCormick, Jill S., "Hospital Costs for Acute Myocardial Infarction Patients Receiving Perfect Compliance of Evidence-Based Care
Bundle" (2008). All Regis University Theses. Paper 521.
This Thesis - Open Access is brought to you for free and open access by ePublications at Regis University. It has been accepted for inclusion in All Regis
University Theses by an authorized administrator of ePublications at Regis University. For more information, please contact [email protected]
Regis University
Rueckert-Hartman College for Health Professions
Final Project/Thesis
Disclaimer
Use of the materials available in the Regis University Thesis Collection
(“Collection”) is limited and restricted to those users who agree to comply with
the following terms of use. Regis University reserves the right to deny access to
the Collection to any person who violates these terms of use or who seeks to or
does alter, avoid or supersede the functional conditions, restrictions and
limitations of the Collection.
The site may be used only for lawful purposes. The user is solely responsible for
knowing and adhering to any and all applicable laws, rules, and regulations
relating or pertaining to use of the Collection.
All content in this Collection is owned by and subject to the exclusive control of
Regis University and the authors of the materials. It is available only for research
purposes and may not be used in violation of copyright laws or for unlawful
purposes. The materials may not be downloaded in whole or in part without
permission of the copyright holder or as otherwise authorized in the “fair use”
standards of the U.S. copyright laws and regulations.
HOSPITAL COSTS FOR ACUTE MYOCARDIAL INFARCTION PATIENTS RECEIVING
PERFECT COMPLIANCE OF EVIDENCE-BASED CARE BUNDLE
by
Jill S. McCormick
A Master’s Thesis Presented in Partial Fulfillment
Of the Requirements for the Degree
Master of Science, Health Service Administration
Regis University
December, 2008
FINAL APPROVAL OF MASTER’S PROJECT
HSA696 MASTER’S PROJECT
I have READ AND ACCEPTED
the Master’s Project by:
Jill S. McCormick
Hospital costs for acute myocardial infarction patients receiving perfect compliance of
evidence-based care bundle
Submitted in partial fulfillment of
requirements for the
Master of Science in Health Services Administration
degree at
Regis University
Primary Research Advisor: Michael Cahill MS
Date: December, 2008
Abstract
An estimated 565,000 new myocardial infarctions and 300,000 recurrent myocardial infarctions
will occur each year (AHA, 2006). This study sought to find if there was a difference in hospital
costs between those acute myocardial infarction patients that received 100% of eligible core
measures (evidence-based care bundle) and those that did not. There is limited research on
actual hospital costs (vs. charge data) for acute myocardial infarction evidence-based treatment
in the United States. The results of the study did not show any statistically significant difference
in hospital costs between those patients that received 100% of core measures and those that did
not. Hospital costs were mostly driven by length of stay, APR-DRG severity and gender. The
study did evidence a statistically significant difference in hospital costs between men and women
that could not be explained by length of stay, age, race, APR-DRG severity or mortality. As
more quality data is publicly reported and the Centers for Medicare and Medicaid Services
places more finances behind reaching performance indicators evidence-based core measures will
come under greater review.
1
Table of Contents
CHAPTER 1: INTRODUCTION ........................................................................................................................ 1
MEDICARE ......................................................................................................................................................... 2
CORE MEASURES-EVIDENCE-BASED CARE BUNDLE .............................................................................................. 3
INCREASED PRESSURE FROM EXTERNAL FORCES .................................................................................................. 5
COMPOSITE SCORE VS. “PERFECT” SCORE ........................................................................................................... 6
CHAPTER 2: LITERATURE REVIEW ............................................................................................................. 8
MORTALITY ....................................................................................................................................................... 9
COMBINATION THERAPY AND MORTALITY ........................................................................................................... 9
AGE DIFFERENCES ............................................................................................................................................ 12
GENDER DIFFERENCES...................................................................................................................................... 13
COSTS .............................................................................................................................................................. 14
INDICATORS FOR COST ..................................................................................................................................... 14
COSTS AND QUALITY MEASURES ....................................................................................................................... 15
CHAPTER 3: METHODOLOGY ..................................................................................................................... 16
SAMPLING STRATEGY ....................................................................................................................................... 16
INSTRUMENTATION........................................................................................................................................... 20
DATA COLLECTION .......................................................................................................................................... 21
STATISTICAL ANALYSIS .................................................................................................................................... 22
CHAPTER 4: RESULTS ................................................................................................................................... 25
DESCRIPTIVE STATISTICS .................................................................................................................................. 25
STATISTICAL ANALYSIS .................................................................................................................................... 29
ST-elevated (STEMI) and Non-ST-elevated (NSTEMI) ................................................................................. 29
Race............................................................................................................................................................ 30
Gender ........................................................................................................................................................ 30
Length of Stay ............................................................................................................................................. 32
Age ............................................................................................................................................................. 34
Mortality ..................................................................................................................................................... 36
APR-DRG Severity ...................................................................................................................................... 37
Perfect score ............................................................................................................................................... 40
Interaction Effects ....................................................................................................................................... 43
CHAPTER 5: DISCUSSION.............................................................................................................................. 44
REFERENCES ................................................................................................................................................... 49
APPENDIX A ..................................................................................................................................................... 58
Chapter 1: Introduction
The American Heart Association estimates the prevalence of 7,200,000 acute myocardial
infarctions in the United States (American Heart Association, 2006). An estimated 565,000 new
myocardial infarctions and 300,000 recurrent myocardial infarctions will occur each year (AHA,
2006). Heart disease, in which acute myocardial infarction is included, is the number one killer
in the United States claiming 654,094 lives in 2004 (Levit, Ryan, Elixhauser, Stranges, Kassed,
and Coffey, 2007). Sixteen percent of all hospital stays resulted from circulatory conditions
(including coronary artery disease, congestive heart failure, heart attack and irregular heartbeat)
(Levit, et al., 2007).
In 2006, direct medical costs of cardiovascular disease totaled $257.6 billion.
Hospitalizations related to heart conditions comprise six of the twenty highest costing conditions
for hospitals, making up 17% of all community hospital costs in 2005 (Levit et al., 2007). As a
primary diagnosis, acute myocardial infarction represented 1.7% of all discharges in 2005 and
was the ninth most frequent principal diagnosis for inpatient stays (Levit et al.). Acute
myocardial infarction, as a principal diagnosis, ranked second in highest aggregate costs in 1997,
2004, and 2005, with total inflation-adjusted hospital costs of $8.7 billion, $11.6 billion and
$10.9 billion, respectively. (Levit et al., p49). Between 1997 and 2005, the aggregate costs for
stays in community hospitals had an average annual increase, after inflationary adjustment, of
5.1% per year for eight years (Levit et al.).
Tran (2004) points out the numerous studies of the quality of acute myocardial infarction
care that have illustrated the underutilization of evidence-based treatment with proven efficacy,
even after controlling for contraindications to the therapy. While increases in utilization of
evidence-based cardioprotective medications (aspirin, β-blockers, angiotensin-converting
2
enzyme (ACE) inhibitors, lipid lowering agents and combinations thereof) have occurred over
the past decade, there is still a great deal of opportunity (Spencer 2005). As the number of
elderly increases sharply, the cost and quality of care for these individuals will be continuously
scrutinized. The purpose of this study is to determine if there is a difference in hospital costs
(total hospital costs, direct costs, indirect costs, fixed costs and variable costs) for those acute
myocardial infarction patients that receive 100% of eligible core measures, “perfect score” and
those that do not. Hospital costs do not include physician costs. The research question is: Is there
a difference in hospital costs in those acute myocardial infarction patients that receive 100% of
eligible core measures, “perfect score” and those that do not receive 100% of eligible core
measures? The null hypothesis is: there is no difference in hospital costs for those acute
myocardial infarction patients that receive 100% of eligible core measures, “perfect score” and
those that do not receive 100% of eligible core measures.
Medicare
In 2005, the largest group of all acute myocardial infarction patients was between 65 and
84 years of age, 45.23% (Levit et al., 2007). According to the 2000 Census, there are nearly 35
million (34,991,753) people 65 years or older, representing 12.4% of the total United States
population. According to the United States Census projections, the population of 65-84 year olds
will increase 38.8% from 2010 to 2020 and the greater than 85 year old population will increase
18.7%. From 2020 to 2030 these groups are projected to increase 30.6% and 32.1%, respectively
(US Census 2000). Medicare is a social health insurance for people over the age of 65, people of
any age with End Stage Renal Disease (ESRD) and people under the age of 65 with specific
disabilities (Hoffman, 2005). Medicare enrollment has increased nearly 125% since its inception
in 1965. Medicare covers 95% of our nation’s aged population, as well as the disabled. In 2004,
3
Part A (hospital expenses and specific other medical care) covered approximately 41 million
beneficiaries with benefit payments of $167.6 billion (Hoffman, 2005). Medicare spending is
projected to increase by nearly $425 billion between 2008 and 2017 (Keehan, Sisko, Truffer,
Smith, Cowan, Poisal, et al., 2008). Only 12% of the population is currently over 65 years old,
yet they make up 34% of all hospitalizations (Levit et al., 2007). 574 stays for every 1,000 over
85 adults took place in 2005 (Levit et al.).
Core Measures-evidence-based care bundle
The Medicare Modernization Act of 1997 required a Medicare Health Quality
component. The resulting Medicare Health Quality Demonstration Project goals included:
“improve safety; enhance quality of care by increasing efficiency; and reduce scientific
uncertainty and unwarranted variation in medical practice that results in both lower quality and
higher costs.” (Mason, 2005, p.2). As a result of these demonstration projects and scientific
review, evidence-based core measures for the treatment of acute myocardial infarction were
developed. Table 1 details these core quality measures.
4
Table 1
NQF-Endorsed Voluntary Consensus Standards for Hospital Care. Acute Myocardial Infarction
Measure Information Form (AMI evidence-based care bundle)
Core Measure
AMI-1: Aspirin at Arrival
Description
AMI patients without aspirin contraindications who received aspirin
within 24 hours before or after hospital arrival.
AMI-2: Aspirin prescribed at
AMI patients without aspirin contraindications who are prescribed
discharge
aspirin at hospital discharge.
AMI-3: ACEI or ARB for left
Angiotensin converting enzyme inhibitor (ACEI) or angiotensin
ventricular systolic dysfunction
receptor blocker (ARB) for LVSD.
(LVSD)
AMI-4: Adult Smoking
AMI patients with a history of smoking cigarettes, who are given
Cessation Advice/Counseling
smoking cessation advice or counseling during hospital stay. For the
purposes of this measure, a smoker is defined as someone who has
smoked cigarettes anytime during the year prior to hospital arrival.
AMI-5: Beta Blocker Prescribed AMI patients without beta blocker contraindications who are
at Discharge
prescribed a beta blocker at hospital discharge.
AMI-6: Beta Blocker at Arrival
AMI patients without beta blocker contraindications who received a
beta blocker within 24 hours after hospital arrival.
AMI-7: Median Time to
Median time from arrival to administration of fibrinolytic agent in
Fibrinolysis
patients with ST-segment elevation or left bundle branch block
(LBBB) on the electrocardiogram (ECG) performed closest to hospital
arrival time.
5
Table 1 (continued).
Core Measure
Description
AMI-7a: Fibrinolytic Therapy
AMI patients receiving fibrinolytic therapy during the
Received within 30 Minutes of
hospital stay and having a time from hospital arrival to
Hospital Arrival
fibrinolysis of 30 minutes or less.
AMI-8: Median Time to Primary PCI
Median time from arrival to percutaneous coronary
intervention (PCI) in patients with ST-segment elevation or
left bundle branch block (LBBB) on the electrocardiogram
(ECG) performed closest to hospital arrival time.
AMI-8a: Primary PCI Received
AMI patients receiving percutaneous coronary intervention
within 90 Minutes of Hospital
(PCI) during the hospital stay with a time from hospital
Arrival
arrival to PCI of 90 minutes or less
Adapted from: “Specifications Manual for National Hospital Quality Measures (Acute Myocardial Infarction) by
Centers for Medicare and Medicaid Services and Joint Commission of Accreditations on Healthcare Organizations,
Version 2.0 (2006, July).
Increased pressure from external forces
Centers for Medicare and Medicaid Services (CMS), The Joint Commission, and Institute of
Health Improvement (IHI) 100,000 lives campaign have brought questions about quality core
measures to a higher status among administrators and is forcing them to review and present data
in a public forum. Furthermore, The American College of Cardiology and the American Heart
Association have released evidence-based guidelines for the management of patients with AMI,
thus increasing visibility from the physician side. The Medicare Prescription Drug, Improvement
and Modernization Act of 2003 requires hospitals to submit data on AMI measures or they will
6
receive a 0.4% reduction in annual payment update from CMS for FY2005, 2006, and 2007
(CMS, 2008).
Donald Berwick, M.D., IHI President and CEO notes, “[t]he average care system is just
pumping out scrap at a very high rate. That’s where conventional ROI thinking ought to work…
there’s big money in getting it right” (Carpenter, 2006, p.26).
Composite Score vs. “Perfect” Score
There are two different measurements to determine adherence to the Centers for
Medicare and Medicaid Services guidelines. The composite score is most often referred to and is
utilized in most publicly reported data. The composite score focuses on the number of times an
intervention takes place divided by the total number of opportunities to complete the
intervention.
Number of total interventions
Number of possible interventions
= Composite Score
Dr. Steven Corwin (2006), Cardiologist, Executive Vice-President and Chief Operating
Officer of New York Presbyterian Hospital spoke of the Composite Score:
The measurements look at each medicine individually. So results will show whether a
patient received that medicine, which is a good measurement of adherence to accepted
medical practices. A good measurement of quality, however, would look at the percent of
patients who received every medicine that they should have received. (p.20)
The Appropriate Care Score (ACS) “Perfect Score” is a measure of the number of times
patients received all the care for which they were eligible.
7
Appropriate Care Score “Perfect Score”=
Total number of patients that received all the care for which they were eligible
Total number of patients eligible for the focus area
Eligible for denotes no contraindications. The “Perfect Score” is the total number of patients that
received all the care they were eligible for divided by the total number of patients eligible for the
focus area (Premier, n.d.).
The efficacy of individual core measure therapies has been determined through many
years of research and several studies (Antiplatelet Trialist’s Collaboration, 1994; Brodie,
Stuckey, Wall, et al., 1998; Flather, Yusuf, Kober, 2000; French, 2000; Jencks, Cuerdon,
Burwen, et al., 2000; Krumholz, Anderson, Brooks, et al., 2006; Krumholz, Radford, Wang, et
al., 1998). As the study is reviewing the perfect score (100% implementation of core measures
vs. not 100% implementation of core measures) the literature review will review combination
therapy, not individual therapies.
The combination of increased external pressures to report quality outcomes and
healthcare consuming a larger share of the economy will lead to, “policymakers, insurers, and
the public [facing] increasingly difficult decisions about the way that healthcare is delivered and
paid for” (Keehan, et al., 2008, w154).
8
Chapter 2: Literature Review
The purpose of this study is to determine if there is a difference in hospital costs for those
acute myocardial infarction patients that receive 100% of eligible core measures, “perfect score”
and those that do not. The research question is: Is there a difference in hospital costs in those
acute myocardial infarction patients that receive 100% of eligible core measures, “perfect score”
and those that do not receive 100% of eligible core measures? The literature review will review
evidence-based combination cardioprotective therapies, hospital costs, and quality for treatment
of acute myocardial infarction.
Medicare is the nation’s largest purchaser of healthcare (Abelson, 2003). Historically,
high quality and improved effectiveness have not been rewarded. In the 2004 Annual Report on
the status of the Social Security and Medicare Programs indicated that the Medicare Hospital
Insurance (MHI) trust fund had significantly deteriorated and is expected to continue the drastic
deterioration after 2010 as Baby Boomers begin to retire. Estimates at that time predicted
complete depletion by 2019 (Cleverly, 2004). Current estimates have Medicare funds “lasting”
slightly longer than previously predicted, yet the introduction of Medicare Part D and other
increases have led to projected increases in Medicare spending, 2008-2017 of nearly one quarter
of a trillion dollars (Keehan, et al., 2008).
Cleverly comments that each provider [hospital] must provide high quality products and
services efficiently and at a reasonable cost. Those that do not will see decreased financial
performance. “While healthcare has some unique characteristics, it is not immune to basic
economic forces” (Cleverley, 2004, p52). The current Medicare reimbursement environment is
not aligned with acute myocardial infarction quality performance indicators. The literature
suggests that there is a threshold of payment for increased quality of care in healthcare
9
organizations (Weech-Maldonado, 2003). While there is a trend to move towards more pay-forperformance mechanisms, it will not reduce the need to review costs.
Mortality
Mortality of patients varies widely in lower and higher quality performing organizations.
Compliance with acute myocardial infarction guidelines has found to lower inpatient mortality
(Szekendi, 2003). Szekendi (2003) found that “patients treated in facilities in the highest quartile
had an average in-hospital mortality rate of 8.3%, while patients treated at hospitals in the lowest
quartile had an average in-hospital mortality rate of 15.3%” (p. 359). Peterson et al. (2006)
presented the first comprehensive study illustrating that acute myocardial infarction mortality
rates were lower in hospitals that followed the AHA College of Cardiology (ACC/AHA)
Guidelines for the management of patients with acute myocardial infarction. The Centers for
Medicare and Medicaid Services (CMS) guidelines are closely aligned with those of the
American College of Cardiology.
Combination Therapy and mortality
Combination cardioprotective medications have been shown to be associated with
reduced risk of death. As illustrated in Table 2, 31% reduction in mortality risk was attained with
aspirin use after adjusting for covariates. 56% improved survival was achieved with the aspirin
and beta blocker group and 45% improved survival in the group with aspirin with beta blockers
and ACE (angiotensin converting enzyme)-inhibitors when compared to those not prescribed any
cardioprotective medications at discharge (Krause et al., 2004). More interestingly, the greatest
survival advantage was made by those patients with the most advanced renal dysfunction.
10
Table 2
Overall unadjusted and adjusted hazard ratios (HR) and 95% confidence interval for mortality
after hospital discharge for acute myocardial infarction by cardioprotective medication group.
Unadjusted HR (95% CI)
No medications**
Adjusted HR* (95%)
1.00
1.00
Aspirin alone
0.79 (0.56, 1.12)
0.69 (0.48, 0.99)
Aspirin and β-blockers
0.39 (0.26, 0.57)
0.44 (0.30, 0.65)
Aspirin, β-blockers, and ACE
inhibitors
0.60 (0.41, 0.87)
0.55 (0.37, 0.81)
Adapted from “Combination therapy improves survival after acute myocardial infarction in the elderly with chronic
kidney disease,” by Krause, M.W., Massing, M., Kshirsagar, A., Rosamond, W., & Simpson, R., 2004, Renal
Failure, 26, p.720.
*Adjusted for age, race, gender, history of diabetes mellitus, hypertension, congestive heart failure, anterior MI
location and level of chronic kidney disease.
** No aspirin, β-blockers, or ACE-inhibitors at hospital discharge.
Danchin, et al.’s (2005) review of nationwide French cardiac registry data evidences the
increased survival of patients receiving triple therapy, combination of anti-platelet agents, βblockers, and statins. The one-year survival was 97% in patients that received triple therapy and
88% in those who did not (p<.0001). Of note, in all quartiles, combination therapy was
associated with lower mortality. As with the renal compromised patients, the group with the
highest risk score and highest mortality also gained the most from triple therapy. However, this
group was the least likely to receive triple therapy. The benefit of triple therapy was evidenced in
ST-elevation myocardial infarction and non-ST-elevation myocardial infarction patients.
11
The Controlled Abciximab and Device Investigation to Lower Late Angioplasty
Complications (CADILLAC) trial found a significantly higher 30-day mortality rate, posthospital discharge, who did not receive aspirin therapy (6.2% vs. 0.3%, p <0.0002) (Kandarzi,
2004). Patients who did not receive a prescription for aspirin at discharge still had a higher
mortality rate at one year (12.4% vs. 2.3%, p <0.0001) and consequently, saw as a result, a
significant increase in composite occurrence of major adverse cardiac events take place.
Moreover, those patients that did not receive aspirin at discharge had a greater need for repeat
targeted vessel revascularization. Kardanzi, et al. (2004) note “our [CADILLAC] results support
this paradoxic quality of care for patients who have AMI and are at greater risk; despite the
protocol-specified administration of aspirin, greater baseline clinical risk, and less procedural
success among patients who did not receive aspirin at discharge, 67 patients did not receive such
treatment.” (p. 1033).
While studies have illustrated increases in usage of combination therapy, there were still
areas in need of improvement (Spencer et al., 2005). Spencer et al. (2005) studied a sample of
5965 adult men and women of all ages discharged after AMI from all greater Worcester hospitals
between 1990 and 2001. The study reviewed the usage of angiotensin-converting enzyme (ACE)
inhibitors, aspirin, β-blockers, and lipid-lowering agents, cardiac medications with proven
efficacy in managing AMI patients and found an increase from 12.9% to 74.0% of hospital
survivors who received three or more cardiac medications, but more work is needed. A Swiss
study of nearly 12,000 patients with acute coronary syndrome (ACS) found increased
underutilization of combination cardioprotective therapy even after accounting for
contraindications and controlling for comorbidities (Schoenenberger, Radovanovic, Stauffer,
Windecker, Urban, Eberli, et al. 2008).
12
Yusuf (2002) has suggested that two-thirds to three-quarters of future vascular events could be
prevented through the effective usage of combination cardioprotective therapy for high risk
patients.
Age differences
Even after recommendations from The American College of Cardiology (ACC) and the
American Heart Association (AHA), studies in multiple countries have indicated a difference in
implementation of guideline-recommended AMI therapies between younger and older patients.
In 2004, a Canadian study utilizing the Canadian Cardiovascular Research Team
(CCORT)/Canadian Cardiovascular Society (CCS) Quality Indicators for AMI Care, found that
the odds ratio of ideal (no contraindications) 65 year old or older patients receiving evidencebased AMI therapies was less than half of 65 year old or younger patients with AMI, except
ACEIs at discharge, suggesting less than optimal implementation (Tran et al., 2004). The
difference was most pronounced in the oldest patients. Adjustments for common
contraindications, as per practice guidelines and applicable literature, did not alter this finding.
Figure 1 details the comparison of the percentage of ideal patients, those without common
contraindications or exclusion criteria, to the benchmark values.
13
Figure 1. Comparison of proportion of ideal patients who received treatment relative to
benchmark values.
*Benchmark values are defined as 90% or greater of ideal patients receiving aspirin within six hours of arrival and
aspirin at discharge. The benchmark values are 85% or more of ideal patients receiving thrombolytics within or less
than 30 minutes of arrival, ideal patients receiving β-blocker within 12 hours of admission, ideal patients receiving
β-blocker at discharge, ideal patients being prescribed an ACEI at discharge, and ideal patients having a lipid sample
obtained within 24 hours of admission. The benchmark value is 70% or more of ideal patients having a statin
prescribed at hospital discharge.
From “Effect of age on the use of evidence-based therapies for acute myocardial infarction,” by C.T.T. Tran, A.
Laupacis, M.M. Mamdani, and J.C. Tu, 2004, American Heart Journal,148(5), p.838.
Kardanzi (2004) emphasizes that risks or concerns about contraindications do not seem to
factor into treatment practice as similar proportions of eligible and ideal elderly patients receive
medications. Some physicians appear to be unaware of the medical evidence available detailing
the benefit of aspirin, β-blockers, thrombolytics, ACEIs, lipid measurements, and statin treatment
in the elderly population. They hypothesize that while physicians may have knowledge of the
reported therapeutics benefits they do not believe there is an actual benefit on patient outcomes.
Moreover, perhaps concerns over poly-pharmacy are greater than the perceived benefits for the
elderly patient should he receive all for treatments for which he was eligible.
Gender Differences
Significant differences in adherence to acute myocardial infarction guidelines have been
found by gender. Spencer (2005) found the female sex to be independently associated with the
14
underuse of combination medical therapy. While the study was limited by the lack of ability to
determine eligibility for treatments, the association is of note. Correa-de-Araujo, et al. (2006)
found significant differences between non-Hispanic white males and females in aspirin upon
arrival, aspirin at discharge, β-blocker at arrival and β-blocker at discharge. Significant
differences, between the same groups, were also found in treatment administration for AMI
patients with diabetes and those AMI patients with hypertension/ESRD. These significant
differences were not found between genders of other races or ethnicities. Correa-de-Araujo, et
al.’s (2006) study excluded those patients that were not eligible to receive treatment, expanding
on Spencer’s (2005) findings.
Costs
Figures on the overall hospital costs for the long term treatment of acute myocardial
infarction patients in the United States is very limited. Furthermore, comparing costs amongst
groups within the AMI population are further limited by differing reimbursement systems and
the complexity and confidentiality of contracts between insurers and hospitals. Eisenstein, et al.
(2001) attempted to calculate, through models not actual costs incurred, long term economic
outcomes for coronary artery disease [of which acute myocardial infarction is included]. They
found that while the total acute costs for unstable angina (NSTEMI) patients was less than for
STEMI patients ($21,957 vs. $24,956) the post acute costs for unstable angina were greater than
those for STEMI patients ($27,787 vs. $22,421) (Eisenstein et al., 2001).
Indicators for Cost
Several studies have shown length of stay in hospitals to be highly correlated with total
hospital costs (Kauf, Velasquez, Crosslin, Weaver, Diaz, et al., 2006; Krumholz, Chen, Murillo,
Cohen, and Radford, M., 1998; Mahon, McCann, Rahallaigh, Codd and O’Sullivan, 2008). In
15
addition to length of stay, Brampkamp, et al., (2007) found the strongest predictors of higher
AMI hospital costs to be gender, cerebrovascular disease and diabetes. Polverejan, et al. (2003)
found cardiac procedures, ejection fraction, and age at admission to be significant predictors of
higher costs. Studies differ in finding higher AMI treatment costs for men and women
(Brampkamp, et.al, 2007; Polverejan, Gardiner, Bradley, Holmes-Rovner, & Rovner, 2003).
Furthermore, advanced age has been found to be associated with lower costs. Udvarhelyi &
Gatsonis, (1992), found that older patients are less eligible for various cardiac treatments.
Costs and quality measures
Minimal studies have examined the difference in hospital costs amongst patients that
received acute myocardial infarction evidence-based guidelines. Krumholz, et al. did find that
those that received evidence-based guidelines (i.e. thrombolysis, aspirin, and β-blockers) were
significantly more costly than those that did not receive such measures (1998). Those patients
were also found to have lower in-patient mortality and were referred for more cardiac procedures
(Krumholz, et al., 1998).
This retrospective study seeks to further understanding of actual hospital costs and the
perfect compliance with the acute myocardial infarction evidenced-based care bundle.
16
Chapter 3: Methodology
The purpose of this study was to identify if there was a difference in hospital costs
between those acute myocardial infarction patients that received 100% of eligible core measures,
a “perfect score”, and those that did not receive 100% of eligible core measures.
A quantitative study was conducted through the utilization of retrospective, secondary
data on 440 acute myocardial infarction patients that were treated at a community-based hospital
in the Western United States.
Sampling Strategy
The participants in the study included those as defined by the Centers for Medicare and
Medicaid Services Acute Myocardial Infarction Core Measure algorithm (Appendix A and B).
The participants in the study included those acute myocardial infarction patients over 18 years of
age, on date of admission, that were randomly sampled by the organization’s Core Measure
reporting vendor, Premier. Premier collects data from member hospitals throughout the country
and “houses the nation’s largest detailed clinical and financial database, housing information on
more than 130 million patient discharges.”(Premier, n.d.). Premier randomly selects 10% of
patients discharged with (ICD-9-CM 410.x principal diagnosis code) up to a maximum of 26
patients per month. Given the hospital’s large AMI population, 26 AMI patient charts were
abstracted per month. If it was discovered that a patient is to receive comfort care only, another
randomly sampled AMI patient was requested by the data collector and supplied by Premier for
abstraction. Trained data collectors within the healthcare system used standardized definitions to
abstract the data. Variables included demographics, treatments administered, associated major
contraindications to evidence-based therapies, discharge recommendations, and interventions.
These patients were discharged from the hospital from August 1, 2006 to December 31, 2007.
17
The sample included male and female patients with an International Classification of DiseasesCM-9 Principal Diagnosis Code of AMI, (410.x.x) (International Classification of Diseases,
2003). As defined by CMS, the population for the measure set only included patients admitted to
the hospital for inpatient acute care. Furthermore, the patients came directly to the hospital, that
is, not transferred from another facility or transferred out to another facility. It included those
patients transported by ambulance or walk-ins.
Exclusion criteria
Those patients that were not eligible for any of the core measures were removed from the
sample. Those patients that were transferred to another acute care facility or federal hospital
were removed because it was unclear if the patient received additional core measures and
mortality was not determined. Patients discharged to hospice and patients with comfort care
measures only as documented by a physician, nurse practitioner, or physician assistant were also
excluded from the population. Patients that were treated outside of an inpatient environment
were also excluded from this population as defined by the Centers for Medicare and Medicaid
Services core measure eligibility. After implementing the exclusion criteria the sample
population included 382 acute myocardial infarction patients.
Participation
There were 382 acute myocardial infarction patients in the study. Of these, 304 patients
did not have ST-segment elevation or left bundle branch block (LBBB). ST-segment elevation or
left bundle branch block (LBBB) is defined from the initial ECG interpretation performed closest
to hospital arrival. ST-segment elevation or a left bundle branch block (LBBB) (as defined by
Hospital Quality Measures Specification Manual) is outlined.
18
The normal ECG is composed of a P wave (atrial depolarization), Q, R, and S
waves (QRS complex, ventricular depolarization), and a T wave (ventricular
repolarization). The ST-segment, the segment between the QRS complex and the T wave,
may be elevated when myocardial injury (AMI) occurs. Between the atria and the
ventricles, the conduction system divides electrical impulses into right and left bundle
branches. A bundle branch block (BBB) results from impaired conduction in one branch,
which in turn results in abnormal ventricular depolarization. In LBBB, left ventricular
depolarization is delayed, resulting in a characteristic widening of the QRS complex on
the ECG. LBBB may be an electrocardiographic manifestation of an AMI (CMS, 2006).
Patients were placed in these categories as STEMI or left bundle branch block (LBBB)
are eligible for percutaneous coronary intervention and those with NSTEMI or left bundle branch
block were not eligible for percutaneous coronary intervention. Patients were separated into two
race categories: Caucasian, and non-Caucasian due to minimal population diversity in the region.
Length of stay was categorized based upon average length of stay for acute myocardial infarction
(principal diagnosis 410.x.x.) of over 450 hospitals in the United States (Premier, 2008). Length
of stay is defined as the time period the patient has been in the hospital for their inpatient stay. A
day for an inpatient is based on the patient being in the hospital at midnight. The population was
separated into five age categories: less than 50 years old, 50 years old and less than 60 years old,
60 years old and less than 70 years old, 70 years old and less than 80 years old, and greater than
80 years old. These age categories were determined by assessing the age of the nationwide AMI
population and separating it into manageable time periods.
Inpatient mortality was divided into two categories: alive at discharge or expired during
hospital stay. This is the inpatient mortality description for set measure id number: AMI-9 (CMS,
19
2008). The Perfect score population was divided into groups: acute myocardial infarction
patients that received 100% of eligible core measures, “perfect score” and those that did not
receive 100% of eligible core measures. The “perfect score”, also known as the "appropriate care
score", is a measure of the number of times patients received all the care for which they were
eligible. Eligible for denotes no contraindications.
The population was divided into two gender groups: female and male. The population
was separated by APR-DRG severity groups (1, 2, 3, and 4) to account for some of the comorbidities which may influence costs. APR-DRG Severity Grouper Methodology as defined in
the Premier database:
Patients in Clinical Advisor are grouped with 3M™’s APR-DRG grouper. APR-DRGs™
integrate the Medicare DRGs, New York AP-DRGs, NACHRI DRGs and Yale
Complication and Comorbidity Refinements into a comprehensive DRG system. They
attempt to explain most severity of illness within the severity of illness levels of the base
DRGs. The APR-DRG grouper categorizes patients into similar disease categories and
then stratifies them into four subclasses for severity of illness and four subclasses for risk
of mortality. There are 316 base APR-DRGs in version 20.0. The subdivision of each of
the 316 APR-DRGs into four severity of illness subclasses, combined with two error
APR-DRGs (955, 956), which are not subdivided, results in 1,258 APR-DRGs. Severity
of illness-adjusted data focuses on explaining differences in length of stay, resource
utilization or costs by adjusting for the interaction of diagnoses, procedures, and age.
Resource use and outcomes are similar for patients in each severity of illness level,
providing more accurate comparisons. Patients fall into a base APR-DRG according to
the following variables: • Age • Procedure • Principal Diagnosis. They are further
20
classified into one of four severity of illness levels based on the following variables: •
Base APR-DRG • Age • Non-operating room procedures • Additional diagnosis •
Combinations of all the above. (Premier, Inc., 2008, p. 137, 2-3)
Instrumentation
There was no formal instrumentation to be used since the data was collected as
retrospective, secondary data, which was derived from two pre-existing electronic applications,
Premier and Trendstar®. Trendstar® was an electronic decision support solution from
McKesson Corporation (Healthcare Management Insight, n.d.). Trendstar® housed the hospital’s
financial information. Premier’s Quality Measures Reporter solution tracks performance
compared to national benchmarks such as Centers for Medicaid and Medicare Services Scope of
Work, The Joint Commission, Leapfrog, other hospitals submitting data to Premier, etc. It
minimizes complexity of abstraction and enables one to immediately correct abstraction errors.
The Premier database and Trendstar® solution were selected over ad hoc reports from existing
clinical applications (ex. MEDITECH and Epic) for economy of time and more extensive ability
to mine the data.
The data used for this research project was from the Premier database and the Trendstar®
database. Approval was granted from the Directors of Quality Decision Support and Financial
Decision Support. As the data included Patient ID (visit number) information, an identifying
factor under HIPAA, approval was sought from the hospital Institutional Review Board and
Regis University Institutional Review Board prior to collecting the data. The data was from 17
month period (August 1, 2006-December 31, 2007). Approval was received from the hospital
Institutional Review Board on October 13, 2008. Approval was received from the University
Review Board on November 1, 2008.
21
The hospital in the study was a non-teaching, trauma-designated, community-based
facility located in the Western United States. The hospital had open heart and cardiac
catheterization capabilities. Eighty-six percent of U.S. hospitals are community-based making
the usage of hospital data more generalizeable. The number of discharges for this group,
community-based hospitals, is increasing an average of 1.5% per year (Levit et al., 2007).
Data Collection
Administrative data reports and financial data reports were run out of Premier and
Trendstar®, respectively. The two reports were combined in an Excel document utilizing the
patient ID (visit number) as a unique identifier. The reports did not include names, medical
record numbers or addresses to minimize the Protected Health Information. The administrative
reports were run by a member of the organization’s Quality Decision Support team. Permission
for running said report was received from the director of the department. This data was
transmitted over a secure intranet and stored on a secure network location with access given only
to members of the Quality Improvement team. The financial reports were run by a member of the
organization’s Financial Decision Support team. Permission for running said report was received
from the director of the department. This data was transmitted over a secure intranet and stored
on a secure network location with access given only to members of the Quality Improvement and
Financial Decision Support team. Patient identifier information was minimal with names,
medical record numbers and addresses removed; therefore an individual had to have access
(password protected) to the clinical database to learn additional information. Names and
addresses were not included in the administrative database. Furthermore, an individual had to
have access (password protected and tracked) to the financial database to learn additional patient
22
identifier information. The patient identifier information was not revealed in the analysis and
therefore, unavailable to providers or others who may influence future care for an individual.
Statistical Analysis
The independent variable in the study was the receipt of 100% of eligible core measures
“perfect score”. The measures were 100% or not 100%. The dependent variables in the study
were hospital costs: total hospital costs, variable costs, fixed costs, direct costs and indirect costs.
An alpha level of 0.05 was used on all statistical tests. Descriptive statistics were run on the data
to provide information. Potential influencing variables were tested to identify whether or not
there was a significant difference in total costs, variable costs, fixed costs, direct costs and
indirect costs between the different levels for each variable.
Independent sample t-tests were completed to assess if there was a significant difference
in total hospital costs, variable costs, fixed costs, direct costs, indirect costs between STEMI and
NSTEMI patients.
If there was a very minimal non-Caucasian sample size no statistical tests would be
completed to assess if there was a significant difference in total hospital costs, variable costs,
fixed costs, direct costs, and indirect costs between Caucasian and non-Caucasian patients.
Independent sample t-tests were completed to assess if there was significant difference in
total hospital cost, variable costs, fixed costs, direct costs, and indirect costs between males and
females. If there was a significant difference, then gender was identified as a covariate.
ANCOVA would be run to normalize the effect of the covariate if gender was identified as a
covariate.
Independent sample t-tests were completed to assess if there was significant difference in
total hospital cost, variable costs, fixed costs, direct costs, and average indirect costs between
23
length of stay five days or less and length of stay greater than five days. If there was a significant
difference and length of stay was identified as a covariate. A Pearson correlation test would be
completed to assess if there was a relationship between length of stay and total hospital cost,
variable costs, fixed costs, direct costs, and indirect costs.
ANOVA was completed to assess if there is a significant difference in hospital costs,
variable costs, fixed costs, direct costs, and indirect costs between age groups (less than 50 years
old, 60-70 years old, 70-80 years old, 80 years and older). If a significant difference was found
between the age groups and total hospital cost, average variable costs, fixed costs, direct costs,
and indirect costs age was identified as a covariate. If age was identified as a covariate ANOVA
Scheffe post-hoc test would be completed to determine between which age groups there was a
significant difference. If differences were found between a few groups interval data would be
used to assess if there was a relationship between age and costs on a continual basis using
Pearson correlation.
Independent samples t-test was completed to assess if there was a significant difference
between total hospital costs, variable costs, fixed costs, direct costs, and indirect costs and
inpatient mortality. If there was significant difference between total hospital costs, variable costs,
fixed costs, direct costs, and indirect costs and inpatient mortality, inpatient mortality would be
identified as a covariate. ANCOVA would be run to normalize the effect of the covariate if
inpatient mortality was identified as a covariate.
ANOVA was completed to assess if there is a significant difference between total
hospital costs, variable costs, fixed costs, average direct costs, and indirect costs and APR-DRG
severity levels (1, 2, 3, and 4). If there was a significant difference and APR-DRG severity
24
would be identified as a covariate. If APR-DRG severity was identified as a covariate ANOVA
Post hoc Scheffe test would be run to determine the difference between groups.
Independent sample t-tests were completed to assess if there was a significant difference
between total hospital costs, variable costs, fixed costs, direct costs, and indirect costs and those
patients that received 100% of eligible core measures, “perfect score” and those that did not.
The methodology and statistical analysis conducted sought to answer the null hypothesis:
there was no difference in hospital costs for those acute myocardial infarction patients that
receive 100% of eligible core measures and those that do not receive 100% of eligible core
measures.
25
Chapter 4: Results
The purpose of this study was to evaluate whether there was a significant difference in
hospital costs between those acute myocardial infarction patients that received 100% of eligible
core measures, “perfect score”, and those that did not. There were a total of 382 acute
myocardial infarction patients in the study sample size. Seven potential covariates were
analyzed in this study prior to the assessment of the “perfect score”.
Descriptive Statistics
The descriptive statistics for those seven potential covariates are shown in Tables 3, 4, 5, 6, 7, 8,
9 and 10.
Table 3
Number of AMI patients treated at facility per STEMI or NSTEMI status
Type
n
Percent
STEMI
78
20.4
NSTEMI
304
79.6
Total Patients
382
100
Table 4
Number of AMI patients treated at the facility per Race
Race
n
Percent
Caucasian
370
96.9
Non-Caucasian
4
1.0
Undetermined Race
8
2.1
382
100
Total Patients
26
Table 5
Number of AMI Patients treated at the facility by Gender
Gender
n
Percent
Female
150
39.3
Male
232
60.7
Total Patients
382
100
Table 6
Number of AMI patients treated at the facility by Length of Stay (LOS)
Length of Stay
n
Percent
5 days of less
317
83.0
Greater than 5 days
65
17.0
Total Patients
382
100
Length of Stay
The majority, 83.0% of AMI patients had a length of stay five days or less. The mean
length of stay was 3.96 days with a standard deviation of 4.15 days.
27
Table 7
Number of AMI patients treated at the facility by Age
Age
n
Percent
<50
44
11.5
50 to <60
78
20.4
60 to <70
91
23.8
70 to <80
80
20.9
≥ 80
89
23.3
Total Patients
382
100
Age
The mean age of patients at discharge was 67.41 years old with a standard deviation of
14.29 years.
Table 8
Number of AMI patients treated at the facility per discharge status
Discharge Status
n
Percent
Alive
378
99.0
Dead
4
1.0
Total Patients
382
100
28
Mortality
The overwhelming number of patients in the sample survived the hospital stay, 99.0%.
Only 1.0% died during the inpatient stay.
Table 9
Number of AMI patients treated at the facility per APR-DRG severity
APR-DRG Severity
n
Percent
APR-DRG Severity-1
115
30.1
APR-DRG Severity-2
161
42.1
APR-DRG Severity-3
80
20.9
APR-DRG Severity-4
26
6.8
Total Patients
382
100
Table 10
Number of AMI treated at facility per “Perfect Score” status
Perfect Score Status
n
Percent
<100% eligible core measures
36
9.4
100% eligible core measures
346
90.6
Total Patients
382
100
Perfect Score
Of those that did not receive 100% of eligible core measures, 8 (22.2%) were female and
28 (77.78%) were male. Of those that did receive 100% of eligible core measures, 142 (41.0%)
were female and 204 (59.0%) were male. Ninety-four and seven tenths percent of females in the
29
sample and 87.9% of men in the sample received 100% of eligible core measures. The average
length of stay for those patients that did not receive 100% of eligible core measures was 4.17
days. The average length of stay for those patients that received 100% of eligible core measures
was 3.93 days.
Table 11
Hospital Costs for AMI patients by cost type
Hospital Cost
Total Costs
Range
Minimum
Maximum
Mean
SD
116,719.31
1,596.10
118,315.41
15,839.23
13,494.53
Variable Costs
72,350.60
608.54
72,959.14
7,362.47
6,924.36
Fixed Costs
51,770.91
987.56
52,758.47
8476.76
6,896.83
Direct Costs
89,849.12
959.70
90,808.82
10,843.33
9,362.63
Indirect Costs
28,937.96
636.40
29,574.36
4,995.90
4,333.12
Statistical Analysis
An alpha level of 0.05 was used for all statistical tests.
ST-elevated (STEMI) and Non-ST-elevated (NSTEMI)
An independent samples t-test was completed to compare total hospital cost, variable
costs, fixed costs, direct costs, and indirect costs between STEMI and NSTEMI patients. Table
12 shows there was no statistically significant difference between total hospital costs, variable
costs, fixed costs, direct costs, or indirect costs between STEMI and NSTEMI patients.
30
Table 12
Difference between STEMI and NSTEMI t-tests
t
p
95% Confidence interval
of the mean difference
Total Costs
.850
.396
(-1414.70, 3558.52)
Variable costs
.548
.584
(-914.55, 1619.65)
Fixed costs
1.091
.277
(-581.12, 2019.85)
Direct costs
.540
.590
(-1235.45, 2167.46)
1.440
.151
(-223.88,1435.68)
Indirect costs
Race
There was a very minimal non-Caucasian sample size, 4, and as a result, no statistical
tests were completed to assess if there was a significant difference in total hospital costs, variable
costs, fixed costs, direct costs, and indirect costs between Caucasian and non-Caucasian patients.
Gender
An independent sample t-test was completed to determine if there was a difference in
total hospital costs between male and female patients. Male patients had significantly higher total
hospital costs than female patients, t(370.37) = 3.92, p < .001. An independent sample t-test was
completed to determine if there was a difference in variable costs between male and female
patients. Male patients had significantly higher variable costs than female patients, t(376.17) =
4.03, p = <.001. An independent sample t-test was completed to determine if there was a
difference in fixed costs between male and female patients. There was a significant difference in
fixed costs between male and female patients, t(364.74) = 3.64, p < .001. An independent
sample t-test was completed to determine if there was a difference in direct costs between male
31
and female patients. There was a significant difference in direct costs between male and female
patients, t(370.58) = 4.01, p < .001. An independent sample t-test was completed to determine if
there was a difference in indirect costs between male and female patients. There was a significant
difference in indirect costs between male and female patients, t(372.10) = 3.56, p < .001.
Table 13
Difference between gender t tests
t
p
Total cost
3.921
95% Confidence
Interval of the mean
difference
<0.001 2565.56, 7726.86
Variable cost
4.029
<0.001 1372.46, 3988.45
Fixed cost
3.637
<0.001 1132.52, 3798.99
Direct cost
4.010
<0.001 1858.71, 5435.60
Indirect cost
3.557
<0.001 670.30, 2327.81
An independent samples t-test was completed to compare length of stay between male
and female patients. There is no statistically significant difference in length of stay between male
and female patients, t(380) = 1.12, p =0.263. A Chi square test was completed to determine if
there was an association between APR-DRG severity levels and gender. An association was
found between APR-DRG severity level and gender, χ² (1) = 5.24, p = 0.022. APR-DRG
severity-1: A greater percentage of men were in APR-DRG severity-1 (36.6%) than women
(20.0%). APR-DRG severity-2: A greater percentage of women (48.7%) were in APR-DRG
severity-2 than men (37.9%). APR-DRG severity-3: A greater percentage of women (25.3%)
than men (18.1%) were in APR-DRG severity-3. APR-DRG severity-4: Percentages of men
(7.3%) and women (6.0%) were very similar in APR-DRG severity 4.
32
Length of Stay
An independent samples t-test was completed to compare total hospital costs between
patients with length of stay five days or less and length of stay greater than five days. There was
a statistically significant difference between total hospital costs for patients with length of stay
five days or less and those with a length of stay greater than five days, t(65.49) = -8.90, p < .001.
An independent samples t-test was completed to compare variable costs between patients with
length of stay five days or less and length of stay greater than five days. There was a statistically
significant difference between variable costs for patients with length of stay five days or less and
those with a length of stay greater than five days, t(66.23) = -7.77, p < .001. An independent
samples t-test was completed to compare fixed costs between patients with length of stay five
days or less and length of stay greater than five days. There was a statistically significant
difference between fixed costs for patients with length of stay five days or less and those with a
length of stay greater than five days, t(65.09) = -9.61, p < .001. An independent samples t-test
was completed to compare direct costs between patients with length of stay five days or less and
length of stay greater than five days. There was a statistically significant difference between
direct costs for patients with length of stay five days or less and those with a length of stay
greater than five days, t(65.86) = -8.18, p < .001. An independent samples t-test was completed
to compare indirect costs between patients with length of stay five days or less and length of stay
greater than five days. There was a statistically significant difference between indirect costs for
patients with length of stay five days or less and those with a length of stay greater than five
days, t(65.01) = -10.10, p < .001. Those patients that had lengths of stay greater than five days
had significantly higher costs (total hospital costs, variable, fixed, indirect and direct) than those
patients that had lengths of stay five days or less. Length of stay was identified as a covariate.
33
Table 14
Differences between lengths of stay t tests
t
-8.901
95% Confidence
interval of the mean
difference
<0.001 -29402.57, -18627
Variable cost
-7.77
<0.001 -13879.70, -8206.91
Fixed cost
-9.61
<0.001 -15667.44, -10275.91
Direct cost
-8.18
<0.001 -19459.47, -11822.75
Indirect cost
-10.10
<0.001 -10029.32, -6718.41
Total cost
p
A Pearson correlation test was completed to assess if there was a relationship between
length of stay and total hospital costs. A strong significant relationship was found between length
of stay and total hospital costs, r = 0.84, p < .001. A Pearson correlation test was completed to
assess if there was a relationship between length of stay and variable costs. A strong significant
relationship was found between length of stay and variable costs, r = 0.76, p < .001. A Pearson
correlation test was completed to assess if there was a relationship between length of stay and
fixed costs. A strong significant relationship was found between length of stay and fixed costs, r
= 0.88, p < .001. A Pearson correlation test was completed to assess if there was a relationship
between length of stay and direct costs. A strong significant relationship was found between
length of stay and direct costs, r = 0.81, p < .001. A Pearson correlation test was completed to
assess if there was a relationship between length of stay and indirect costs. A strong significant
relationship was found between length of stay and indirect costs, r = 0.87, p < .001.
34
Age
ANOVA was completed to assess if there was a significant difference in total hospital
costs between age groups (less than 50 years old, 50 years old to less than 60 years old, 60 years
old to less than 70 years old, 70 years old to less than 80 years old, and 80 years old and older).
A significant difference was found in total hospital costs between age groups, F (4,381) = 3.99, p
= 0.003. ANOVA was completed to assess if there was a significant difference in variable costs
between age groups (less than 50 years old, 50 years old to less than 60 years old, 60 years old to
less than 70 years old, 70 years old to less than 80 years old, and 80 years old and older). A
significant difference was found in variable costs between age groups, F (4,381) = 3.26, p =
0.012. ANOVA was completed to assess if there was a significant difference in fixed costs
between age groups (less than 50 years old, 50 years old to less than 60 years old, 60 years old to
less than 70 years old, 70 years old to less than 80 years old, and 80 years old and older). A
significant difference was found in fixed costs between age groups, F (4,381) = 4.53, p = 0.001.
ANOVA was completed to assess if there was a significant difference in direct costs between age
groups (less than 50 years old, 50 years old to less than 60 years old, 60 years old to less than 70
years old, 70 years old to less than 80 years old, and 80 years old and older). A significant
difference was found in direct costs between age groups, F (4,381) = 3.85, p = 0.004. ANOVA
was completed to assess if there was a significant difference in indirect costs between age groups
(less than 50 years old, 50 years old to less than 60 years old, 60 years old to less than 70 years
old, 70 years old to less than 80 years old, and 80 years old and older). A significant difference
was found in indirect costs between age groups, F (4,381) = 4.12, p = 0.003. Age was identified
as a covariate.
35
An ANOVA Scheffe post-hoc test was completed to determine between which age
groups there was a significant difference in total hospital costs. There was a significant
difference in total costs between 60 year old to less than70 age group and the 80 year old or older
age group, α = 0.011, (1082.57, 13346.41), with the former group having higher total costs. No
significant differences in total costs were found between the other groups. An ANOVA Scheffe
post-hoc test was completed to determine between which age groups there was a significant
difference in variable costs. There was a significant difference in variable costs between 60 year
old to less than70 age group and the 80 year old or older age group, α = 0.031, (192.03,
6508.41), with the former group having higher costs. No significant differences in variable costs
were found between the other groups. An ANOVA Scheffe post-hoc test was completed to
determine between which age groups there was a significant difference in fixed costs. There was
a significant difference in fixed costs between 60 year old to less than70 age group and the 80
year old or older age group, α = 0.006, (738.78, 6989.77), with the former group having higher
costs. No significant differences in fixed costs were found between the other groups. An
ANOVA Scheffe post-hoc test was completed to determine between which age groups there was
a significant difference in direct costs. There was a significant difference in direct costs between
60 year old to less than70 age group and the 80 year old or older age group, α = 0.012, (706.31,
9221.18), with the former group having higher costs. No significant differences in direct costs
were found between the other groups. An ANOVA Scheffe post-hoc test was completed to
determine between which age groups there was a significant difference in indirect costs. There
was a significant difference in indirect costs between 60 year old to less than70 age group and
the 80 year old or older age group, α = 0.015, (283.10, 4218.39), with the former group having
higher costs. No significant differences in indirect costs were found between the other groups.
36
Because there was only a difference between two of the five groups interval data was
used in a Pearson correlation to assess if there was a relationship between age and total cost on a
continual basis. No significant relationship was found between age and total costs, r = -0.04, p =
0.442. A Pearson correlation was completed to assess if there was a relationship between age and
variable cost on a continual basis. No significant relationship was found between age and
variable costs, r = -0.04, p = 0.454. A Pearson correlation was completed to assess if there was a
relationship between age and fixed cost on a continual basis. No significant relationship was
found between age and fixed costs, r = -0.04, p = 0.452. A Pearson correlation was completed to
assess if there was a relationship between age and direct cost on a continual basis. No significant
relationship was found between age and direct costs, r = -0.044, p = 0.393. A Pearson correlation
was completed to assess if there was a relationship between age and indirect cost on a continual
basis. No significant relationship was found between age and indirect costs, r = -0.03, p = 0.583.
Mortality
An independent sample t-test was completed to determine if there was a difference in
total hospital costs, variable costs, fixed costs, direct costs, and indirect costs between patients
alive at discharge and those that died during the inpatient stay. As illustrated in Table 15, there
was no significant difference between total hospital costs, variable costs, fixed costs, direct costs,
or indirect costs between those that died during the inpatient stay.
37
Table 15
Differences between mortality (discharge status) t tests
t
p
Total cost
.488
95% Confidence
interval for difference
of means
.626 -10033.70, 16666.32
Variable cost
.380
.704 -5526.15, 8175.94
Fixed cost
.574
.566 -4830.74, 8813.56
Direct cost
.409
.683 -7336.34, 11190.10
Indirect cost
.637
.524 -2896.34, 5675.19
APR-DRG Severity
ANOVA was completed to assess if there was a significant difference in total hospital
costs between APR-DRG severity categories. A significant difference was found in total hospital
costs between APR-DRG severity categories, F (3,381) = 27.68, p < .001. ANOVA was
completed to assess if there was a significant difference in variable costs between APR-DRG
severity categories. A significant difference was found in variable costs between APR-DRG
severity categories, F (3,381) = 22.47, p < .001. ANOVA was completed to assess if there was a
significant difference in fixed costs between APR-DRG severity categories. A significant
difference was found in fixed costs between APR-DRG severity categories, F (3,381) = 30.71,p
< .001. ANOVA was completed to assess if there was a significant difference in direct costs
between APR-DRG severity categories. A significant difference was found in direct costs
between APR-DRG severity categories, F (3,381) = 26.30, p < .001. ANOVA was completed to
assess if there was a significant difference in indirect costs between APR-DRG severity
38
categories. A significant difference was found in indirect costs between APR-DRG severity
categories, F (3,381) = 28.71, p < .001.
An ANOVA Scheffe post-hoc test was completed to determine between which APRDRG severity groups there was a significant difference in total hospital costs. There was a
significant difference in total hospital costs between APR-DRG severity levels. APR-DRG
severity-1 total hospital costs were significantly lower than APR-DRG severity-3 total hospital
costs, α = 0.005, (-11491.51,-1461.09) and significantly lower than APR-DRG severity-4 total
hospital costs, α = 0.000, (-30207.34, -15246.02). APR-DRG severity-2 total hospital costs were
significantly lower than APR-DRG severity-3 total hospital costs, α = 0.025, (-9873.70, -449.46)
and significantly lower than APR-DRG severity-4 total hospital costs, α = 0.000, (-28692, 14131.03). APR-DRG severity-3 total hospital costs were significantly lower than APR-DRG
severity-4 total hospital costs, α = 0.000, (-24026.94, -8476.83).
An ANOVA Scheffe post-hoc test was completed to determine between which APRDRG severity groups there was a significant difference in variable costs. There was a significant
difference in variable costs between some APR-DRG severity levels. APR-DRG severity-1
variable costs were significantly lower than APR-DRG severity-4 variable costs, α = 0.000, (14262.97, -6816.38). APR-DRG severity-2 variable costs were significantly lower than APRDRG severity-4 variable costs, α = 0.000, (-14228.78,-6626.72). APR-DRG severity-3 variable
costs were significantly lower than APR-DRG severity-4 variable costs, α = 0.000, (-12359.60, 4240.05).
An ANOVA Scheffe post-hoc test was completed to determine between which APRDRG severity groups there was a significant difference in fixed costs. There was a significant
difference in fixed costs between some APR-DRG severity levels. APR-DRG severity-1 fixed
39
costs were significantly lower than APR-DRG severity-3 fixed costs, α = 0.000, (-6592.78, 1516.12). APR-DRG severity-1 fixed costs were significantly lower than APR-DRG severity-4
fixed costs, α = 0.000, (-15791.17, -8218.85). APR-DRG severity-2 fixed costs were
significantly lower than APR-DRG severity-3 fixed costs, α = 0.006, (-5418.59, -648.73). APRDRG severity-2 fixed costs were significantly lower than APR-DRG severity-4 fixed costs, α =
0.000, (-14669.29, -7299.14). APR-DRG severity-3 fixed costs were significantly lower than
APR-DRG severity-4 fixed costs, α = 0.000, (-11886.48, -4014.64).
An ANOVA Scheffe post-hoc test was completed to determine between which APRDRG severity groups there was a significant difference in direct costs. There was a significant
difference in direct costs between some APR-DRG severity levels. APR-DRG severity-1 direct
costs were significantly lower than APR-DRG severity-3 direct costs, α = 0.035, (-7168.14, 177.43). APR-DRG severity-1 direct costs were significantly lower than APR-DRG severity-4
direct costs, α = 0.000, (-20716.34, -10289.04). APR-DRG severity-2 direct costs were
significantly lower than APR-DRG severity-4 direct costs, α = 0.000, (-20038.56, -9889.65).
APR-DRG severity-3 direct costs were significantly lower than APR-DRG severity-4 direct
costs, α = 0.000, (-17249.77, -6410.03).
An ANOVA Scheffe post-hoc test was completed to determine between which APRDRG severity groups there was a significant difference in indirect costs. There was a significant
difference in indirect costs between some APR-DRG severity levels. APR-DRG severity-1
indirect costs were significantly lower than APR-DRG severity-3 indirect costs, α = 0.000, (4408.54, -1198.48). APR-DRG severity-1 indirect costs were significantly lower than APR-DRG
severity-4 indirect costs, α = 0.000, (-9618.04, -4829.94). APR-DRG severity-2 indirect costs
were significantly lower than APR-DRG severity-3 indirect costs, α = 0.003, (-3535.40, -
40
519.35). APR-DRG severity-2 indirect costs were significantly lower than APR-DRG severity-4
indirect costs, α = 0.000, (-8777.99, -4117.73). APR-DRG severity-3 indirect costs were
significantly lower than APR-DRG severity-4 indirect costs, α = 0.000, (-6909.23, -1931.74).
APR-DRG severity level was identified as a covariate.
Perfect score
An independent sample t-test was completed to determine if there were significant
differences in total hospital costs, variable costs, fixed costs, direct costs and indirect costs
between those patients that received 100% of eligible core measures, “perfect score” and those
that did not receive 100% of eligible core measures. Table 16 shows the results of the t-tests,
there was no significant difference in total hospital costs, variable costs, fixed costs, direct costs
or indirect costs between those patients that received 100% of eligible core measures, “perfect
score” and those that did not receive 100% of eligible core measures.
Table 16
Differences between Perfect score and non-Perfect score
t
p
Total cost
.625
95% Confidence
Interval of the mean
difference
.532 -3171.03, 6129.57
Variable cost
.506
.613 -1772.32, 3000.89
Fixed cost
.716
.475 -1511.33, 3241.30
Direct cost
.522
.602 -2370.64, 4083.21
Indirect cost
.821
.412 -869.68, 2115.65
ANCOVA was run to normalize the effects of the covariates length of stay, gender and
APR-DRG severity on hospital costs. Length of stay and gender were shown to have a
41
statistically significant positive effect on the perfect score’s relationship with total cost. The
ANCOVA results are shown in Tables 17, 18, 19, 20 and 21.
Table 17
Total Cost ANCOVA Results
F
Length of Stay
p
211.097
<0.001
18.664
<0.001
APR-DRG Severity
3.453
0.064
Perfect Score
0.005
0.943
Gender
n = 382. r² = .477
Table 18
Variable Cost ANCOVA Results
F
p
Length of Stay
147.37
<0.001
Gender
16.800
<0.001
APR-DRG Severity
1.693
0.194
Perfect Score
0.006
0.938
n = 382 r² =.388
42
Table 19
Fixed Cost ANCOVA Results
F
Length of Stay
p
257.510
<0.001
17.750
<0.001
APR-DRG Severity
5.474
0.020
Perfect Score
0.056
0.813
Gender
n = 382 r² = .527
Table 20
Direct Cost ANCOVA Results
F
Length of Stay
p
168.542
<0.001
18.279
<0.001
APR-DRG Severity
2.857
<0.001
Perfect Score
0.004
0.948
Gender
n = 382 r² = .425
43
Table 21
Indirect Cost ANCOVA Results
F
Length of Stay
p
293.118
<0.001
16.604
<0.001
APR-DRG Severity
4.481
0.035
Perfect Score
0.162
0.687
Gender
n = 382 r² = .552
Interaction Effects
There was a significant interaction effect for gender and length of stay on total costs, F =
7.71, p = .006. There was a significant interaction effect for gender and length of stay on
variable cost, F = 6.572, p = .011. There was a significant interaction effect for gender and length
of stay on fixed cost, F = 7.70, p = .006. There was a significant interaction effect for gender and
length of stay on direct cost, F = 6.52, p = .011. There was a significant interaction effect for
gender and length of stay on indirect cost, F = 9.55, p = .002.
The null hypothesis: there is no difference in hospital costs between those acute
myocardial infarction patients that received 100% of eligible core measures and those that did
not receive 100% of eligible core measures was not rejected.
44
Chapter 5: Discussion
The results of the study indicated that there was no statistically significant difference in
hospital costs (total hospital costs, variable, fixed, direct or indirect costs) between those patients
who received 100% of eligible core measures, “perfect score” and those that did not. Adjusting
for the covariates, length of stay, gender and APR-DRG severity made the results less
statistically significant.
There was not a significant difference between STEMI and NSTEMI patients’ hospital
costs (total hospital costs, variable, fixed, direct or indirect). This finding contradicts previous
studies which found STEMI was associated with increased costs (Kauf, et al., 2006; Krumholz,
et al. 1998). STEMI patients had shorter lengths of stay than NSTEMI patients. This aligns with
findings that procedural costs (Tiemann, 2008) are greater for STEMI than NSTEMI, however,
the shorter length of stay offsets this increase in procedural cost in this study. This finding calls
into question if more aggressive treatment on NSTEMI patients would lead to decreased lengths
of stay and reduced hospital costs.
Contrary to previous studies (Polverejan, et al. 2003), hospital costs were not
significantly different between age groups. This finding questions the literature findings that
older patients do not receive as many cardiac interventions.
Chi-square confirmed an association of receipt of 100% of eligible core measures with
gender, χ² (1) = 4.83, p = 0.028. A higher percentage of women, 94.7%, received 100% of
eligible core measures than men, 87.9%. This finding contradicts the literature suggesting
women receive less evidence-based interventions than men (Correa-de-Aruajo, et al., 2006).
Men had a significantly higher cost (total cost, variable, fixed, direct, and indirect) than
women. Of note, there was no significant difference in hospital cost between men and women,
45
thus eliminating length of stay as the reason for the significantly higher cost. Moreover, men
were found to have lower severity with 36.6% being APR-DRG severity 1 vs. 20.0% of women
in APR-DRG severity 1. Women also had greater severity with 74.0% in APR-DRG severity 2
and 3 as compared to men, 56.0%. This finding indicates that it is not APR-DRG severity
differences between men and women that is accounting for the difference in hospital costs.
Overall, the vast majority of the acute myocardial infarction patients received 100% of
eligible core measures, “perfect score”, 90.6%. In addition, very few patients in the sample died
during their inpatient stay, 1.0% (n = 4).
Significance to healthcare administration
Many suggest that providing quality care costs more money. This study evidences that is
not the case. Hospitalizations related to heart conditions comprise six of the twenty highest
costing conditions for hospitals, making up 17% of all community hospital costs in 2005 (Levit
et al., 2007). As a primary diagnosis, acute myocardial infarction represented 1.7% of all
discharges in 2005 and was the ninth most frequent principal diagnosis for inpatient stays (Levit
et al.). Acute myocardial infarction, as a principal diagnosis, ranked second in highest aggregate
costs in 1997, 2004, and 2005, with total inflation-adjusted hospital costs of $8.7 billion, $11.6
billion and $10.9 billion, respectively. (Levit et al., p49).
As more quality measure data is being required to be publicly reported and more
reimbursement dollars are tied to the outcomes, the actual hospital costs are of greater
importance in conducting cost benefit analysis of implementing evidence-based interventions.
Furthermore, this information can be a foundation on which more quality of life data can be
integrated.
46
Strengths
The study had a sample size of 382 patients from a community-based facility in the
Western United States. This study utilized actual hospital costs as hospital charges are not a
reliable measure of resource utilization and consumption. The study reviewed seven potential
covariates for their influence on hospital costs. Lastly, the study only included those measures
for which individuals were eligible, a stronger indicator of quality care.
Limitations
The study exhibits several potential limitations. First, it is possible that unmeasured
factors may have influenced hospital costs, despite accounting for multiple covariates. The
sample represented patients from a single community-based facility in the Western United
States. Regional differences in acute myocardial infarction treatment practices cannot be
accounted for in this study. The sample population did not encompass enough non-Caucasian
patients to determine if there was a significant difference in hospital costs between Caucasian
and non-Caucasian patients. With only 1% of the study patients dying during their inpatient stay,
the statistical analysis on inpatient mortality was incomplete. Furthermore, the high rate of
“perfect score”, 90.6% of the study population made the comparison group, those that did not
receive 100% of eligible core measures, only 9.4%. The study considered 100% or less than
100% of eligible core measures. This equation did not take into account what measures
individuals did or did not receive. Moreover, co-morbidities making a patient ineligible to
receive the measure may have required additional procedures or interventions.
Recommendations for further study
Given the single facility of the study, further study of more hospitals of varying types
(community, teaching, critical access, non-for-profit, for-profit) should be pursued. Furthermore,
47
given regional differences in healthcare practices and costs (Sirovich, Gottlieb, Welch, and
Fisher, 2006) more hospitals throughout the country should be studied. The study evidences
significantly higher hospital costs for male patients than for female patients. These differences
are not accounted for by length of stay, APR-DRG severity, or receipt of 100% of eligible core
measures. Since the study only analyzed costs associated with those the individual was eligible
for, further review of individual treatments received by the men and women should be reviewed.
Lastly, since many of the evidenced-based care bundle interventions are predominantly for
secondary prevention, more research on long term hospital costs and outpatient costs after initial
acute myocardial infarction will be critical to understanding the full financial impact of
providing quality acute myocardial infarction care. Eisenstein, et al. (2001) found that total
inpatient costs for unstable angina (NSTEMI) and STEMI to be very similar. However, post
acute costs were higher for those unstable angina patients than for STEMI patients. Overall, most
ACS costs are incurred in the post acute phase further emphasizing the need for implementation
of cardioprotective therapies for secondary prevention.
Conclusion
This study illustrates that providing the highest quality of care for acute myocardial
infarction patients, the “perfect score” is not more costly. Heart disease, in which acute
myocardial infarction is included, is the number one killer in the United States claiming 654,094
lives in 2004 (Levit, Ryan, Elixhauser, Stranges, Kassed, and Coffey, 2007). As Medicare
spending is projected to increase by $425 billion between 2008 and 2017, the economic
pressures to contain cost while providing high quality care will come under greater scrutiny. Of
particular interest, are the treatment practices among various groups of patients (i.e. males and
females, different races, payers, age, etc.). Hospitals need to further examine their evidence-
48
based care practices, the outcomes and costs in order to meet the needs of their patients and
payers.
49
References
Abelson, R. (2003). Hospitals Say They're Penalized by Medicare for Improving Care. (Cover
story). New York Times, 153(52688), A1-C2. Retrieved Saturday, January 06, 2007 from
the Academic Search Premier database.
American Heart Association (2006). Heart Disease and Stroke Statistics-2006 Update. Dallas,
Texas. American Heart Association. Retrieved October 23, 2008, from
http://www.dsf.health.state.pa.us/health/lib/health/epht/heart_disease_and_stroke_stats2006_update.pdf
Antiplatelet Trialist’s Collaboration (1994). Collaborative overview of randomized trials of
antiplatelet therapy-I: prevention of death, myocardial infarction, and stroke by prolonged
antiplatelet therapy in various categories of patients. British Medical Journal, 308: 81106.
Bramkamp, M., Radovanovic, D., Erne, P., & Szucs, T. (2007, October). Determinants of costs
and the length of stay in acute coronary syndromes: a real life analysis of more than
10,000 patients. Cardiovascular Drugs and Therapy / Sponsored By The International
Society of Cardiovascular Pharmacotherapy, 21(5), 389-398. Retrieved November 28,
2008, from MEDLINE database.
Brodie, B., Stuckey, T., Wall, T., Kissling, G., Hansen, C., Muncy, D., et al. (1998, November).
Importance of time to reperfusion for 30-day and late survival and recovery of left
ventricular function after primary angioplasty for acute myocardial infarction. Journal of
The American College of Cardiology, 32(5), 1312-1319. Retrieved November 23, 2008,
from MEDLINE database.
50
Carpenter, D. (2006, June). Attention investors. The Q-revolution is spreading. Trustee: The
Journal for Hospital Governing Boards, 59(6), 22. Retrieved May 18, 2008, from
MEDLINE database.
Centers for Medicare and Medicaid Services (2008). Hospital Quality Initiative Overview:
Centers for Medicaid and Medicare Services. Retrieved November 12, 2008, from
http://www.cms.hhs.gov/HospitalQualityInits/Downloads/Hospitaloverview.pdf.
Centers for Medicare and Medicaid Services and Joint Commission of Accreditations of
Healthcare Organizations (2006, July). Specifications Manual for National Hospital
Quality Measures, version 2.0. Retrieved October 23, 2008, from
http://qualitynet.org/dcs/ContentServer?cid=1142976368240&pagename=QnetPublic%2
FPage%2FQnetTier3&c=Page
Cleverley, W. (2004). Looking ahead to Medicare 2005. Healthcare Financial Management:
Journal of The Healthcare Financial Management Association, 58(12), 50-55. Retrieved
Sunday, January 14, 2007 from the MEDLINE database.
Correa-de-Araujo, R., Stevens, B., Moy, E., Nilasena, D., Chesley, F., & McDermott, K. (2006,
March). Gender differences across racial and ethnic groups in the quality of care for acute
myocardial infarction and heart failure associated with comorbidities. Women's Health
Issues, 16(2), 44-55. Retrieved October 23, 2008, from Academic Search Premier.
Corwin, S. (2006, May 15). Quality over Quantity. Many measures fail to provide accurate
reading of a provider’s performance. Modern Healthcare, 36, 20-20. Retrieved May 18,
2008, from Academic Search Premier database.
51
Danchin, N., Cambou, J.-P., Hanania, G., Kadri, Z., Genès, N., Lablanche, J.-M. et al. (2005).
Impact of combined secondary prevention therapy after myocardial infarction: Data from
a nationwide French registry. American Heart Journal, 150, 1147-53. Retrieved June 29,
2008, from MEDLINE database.
Eisenstein, E., Shaw, L., Anstrom, K., Nelson, C., Hakim, Z., Hasselblad, V., et al. (2001,
August). Assessing the clinical and economic burden of coronary artery disease: 19861998. Medical Care, 39(8), 824-835. Retrieved October 23, 2008, from MEDLINE
database.
Flather, M., Yusuf, S., Kober, L., Pfeffer, M., Hall, A., Murray, G., et al. (2000, May 6). Longterm ACE-inhibitor therapy in patients with heart failure or left-ventricular dysfunction: a
systematic overview of data from individual patients. Lancet, 355(9215), 1575. Retrieved
November 23, 2008, from Academic Search Premier database.
French, W. (2000, March 9). Trends in Acute Myocardial Infarction Management: Use of the
National Registry of Myocardial Infarction. American Journal of Cardiology, 85(5A),
5B. Retrieved November 23, 2008, from Academic Search Premier database.
Healthcare Management Insight (n.d.). Trendstar®: McKesson Corporation. Retrieved October
23, 2008, from
http://www.healthcaremanagementinsight.com/HMIProduct/TRENDSTAR.html
52
Hoffman, Jr., E.D, Klees, B.S., and Curtis, C.A. (2005). Brief summaries of Medicare and
Medicaid Title XVIII and Title XIX of The Social Security Act as of November 1, 2005.
Office of the Actuary Centers for Medicare & Medicaid Services Department of Health
and Human Services. P 3. Retrieved October 15, 2006, from:
http://www.cms.hhs.gov/MedicareProgramRatesStats/downloads/MedicareMedicaidSum
maries2005.pdf
International Classification of Diseases, Ninth Revision, Clinical Modification (2003). 6 th ed.
Washington, DC: US Dept of Health and Human Services, Centers for Disease Control
and Prevention, Centers for Medicare and Medicaid Services.
Jencks, S., Cuerdon, T., Burwen, D., Fleming, B., Houck, P., Kussmaul, A., et al. (2000, October
4). Quality of Medical Care Delivered to Medicare Beneficiaries. JAMA: Journal of the
American Medical Association, 284(13), 1670. Retrieved November 23, 2008, from
Academic Search Premier database.
Kandzari, D., Tcheng, J., Grines, C., Cox, D., Stuckey, T. Griffin, J. et al. (2004). Influence of
admission and discharge aspirin use on survival after primary coronary angioplasty for
acute myocardial infarction. The American Journal of Cardiology, 94 (8), 1029-1033.
Retrieved June 29, 2008, doi:10.1016/j.amjcard.2004.06.060
Kauf, T., Velazquez, E., Crosslin, D., Weaver, W., Diaz, R., Granger, C., et al. (2006, January).
The cost of acute myocardial infarction in the new millennium: evidence from a
multinational registry. American Heart Journal, 151(1), 206-212. Retrieved October 23,
2008, from MEDLINE database.
53
Keehan, S., Sisko, A., Truffer, C., Smith, S., Cowan, C., Poisal, J., et al. (2008, May 2). Health
Spending Projections Through 2017: The Baby-Boom Generation Is Coming To
Medicare. Health Affairs, 27(1/2), w145-w155. Retrieved November 22, 2008,
doi:10.1377/hlthaff.27.2.w145
Krause, M.W., Massing, M., Kshirsagar, A., Rosamond, W., & Simpson, R. (2004).
Combination therapy improves survival after acute myocardial infarction in the elderly
with chronic kidney disease. Renal Failure, 26 (6), 715-725. Retrieved June 29, 2008,
doi:10.1081/JDI-200037110
Krumholz, H., Anderson, J., Brooks, N., Fesmire, F., Lambrew, C., Landrum, M., et al. (2006,
February 7). ACC/AHA clinical performance measures for adults with ST-elevation and
non-ST-elevation myocardial infarction: a report of the American College of
Cardiology/American Heart Association Task Force on Performance Measures (Writing
Committee to Develop Performance Measures on ST-Elevation and Non-ST-Elevation
Myocardial Infarction). Circulation, 113(5), 732-761. Retrieved November 23, 2008,
from MEDLINE database.
Krumholz, H., Chen, J., Murillo, J., Cohen, D., & Radford, M. (1998, March). Clinical correlates
of in-hospital costs for acute myocardial infarction in patients 65 years of age and older.
American Heart Journal, 135(3), 523-531. Retrieved October 23, 2008, from MEDLINE
database.
54
Krumholz, H., Radford, M., Wang, Y., Chen, J., Heiat, A., & Marciniak, T. (1998, August 19).
National use and effectiveness of beta-blockers for the treatment of elderly patients after
acute myocardial infarction. JAMA: Journal of the American Medical Association,
280(7), 623. Retrieved November 23, 2008, from Academic Search Premier database.
Levit, K, Ryan K, Elixhauser, A, Stranges, E, Kassed C, Coffey R. HCUP Facts and Figures:
Statistics on Hospital-based Care in the United States in 2005. Rockville, MD: Agency
for Healthcare Research and Quality, 2007. http://www.hcup-us.ahrq.gov/reports.jsp
retrieved Sunday, May 18, 2008.
Mahon, N., Rahallaigh, P., O'Sullivan, J., Codd, M., McCann, H., & Sugrue, D. (2000, June).
Hospital cost of acute myocardial infarction in the thrombolytic era. Irish Medical
Journal, 93(4), 122-122. Retrieved October 23, 2008, from MEDLINE database.
Mason, C. (2005). Medicare Health Care Quality Demonstration: Medicare Modernization Act
Section 646. Centers for Medicaid and Medicare Services. Retrieved January 04, 2007,
from
http://www.cms.hhs.gov/DemoProjectsEvalRpts/downloads/MMA646_Request_for_Info
rmation.pdf
Peterson, E., Roe, M., Mulgund, J., DeLong, E., Lytle, B., Brindis, R., et al. (2006, April 26).
Association Between Hospital Process Performance and Outcomes Among Patients With
Acute Coronary Syndromes. JAMA: Journal of the American Medical Association,
295(16), 1912-1920. Retrieved June 13, 2008, from Academic Search Premier database.
55
Polverejan, E., Gardiner, J., Bradley, C., Holmes-Rovner, M., & Rovner, D. (2003, November).
Estimating mean hospital cost as a function of length of stay and patient characteristics.
Health Economics, 12(11), 935-947. Retrieved November 28, 2008, from MEDLINE
database.
Premier, Inc. (2008, November). [Average length of stay for ICD-9 410.xx principal diagnosisall hospitals in Premier database discharges August 2006 to December 31, 2007.]
Unpublished raw data.
Premier, Inc. (2008, May). GrouperMethology. Clinical Advisor Reference Manual:
Methodologies. Retrieved November 23, 2008, from
https://informaticsbeta.premierinc.com/microstrategy/asp/Main.aspx?pg=subProduct&slic
e=1&parAppId=&prCrumbId=0#
Premier, Inc. (n.d.). Mission/Vision: What Premier does. Retrieved December 11, 2008, from
http://www.premierinc.com/about/mission/index.jsp
Premier, Inc. (n.d.). Quality Measures Reporter. Retrieved October 23, 2008, from
http://www.premierinc.com/quality-safety/tools-services/performancesuite/QMR_21907v2.pdf
Schoenenberger, A., Radovanovic, D., Stauffer, J., Windecker, S., Urban, P., Eberli, F., et al.
(2008, March). Age-Related Differences in the Use of Guideline-Recommended Medical
and Interventional Therapies for Acute Coronary Syndromes: A Cohort Study. Journal of
the American Geriatrics Society, 56(3), 510-516. Retrieved November 12, 2008 from
MEDLINE database.
56
Sirovich, B.E., Gottlieb, D.J., Welch, H.G., and Fisher, E.S. (2006). Regional variations in
health care intensity and physician perceptions of quality of care. Annals of Internal
Medicine, 144(9), 641-649.
Spencer, F., Lessard, D., Yarzebski, J. Gore, J. & Goldberg, R. (2005). Decade-long changes in
the use of combination evidence-based medical therapy at discharge for patients
surviving acute myocardial infarction. American Heart Journal, 150, 838-44. Retrieved
June 29, 2008, from MEDLINE database.
Szekendi, M. (2003). Compliance With Acute MI Guidelines Lowers Inpatient Mortality.
Journal of Cardiovascular Nursing, 18(5), 356-359. Retrieved Thursday, January 04,
2007 from the Academic Search Premier database.
Tiemann, O. (2008, January). Variations in hospitalisation costs for acute myocardial infarction a comparison across Europe. Health Economics, 17(1 Suppl), S33-45. Retrieved October
23, 2008, from MEDLINE database.
Tran, C.T.T., Laupacis, A., Mamdani, M., & Tu, J.V. (2004). Effect of age on the use of
evidence-based therapies for acute myocardial infarction. American Heart Journal, 148,
834-41. Retrieved June 29, 2008, from MEDLINE database.
Udvarhelyi, I., & Gatsonis, C. (1992, November 11). Acute myocardial infarction in the
Medicare population. JAMA: Journal of the American Medical Association, 268(18),
2530. Retrieved November 28, 2008, from Academic Search Premier database.
U.S. Census Bureau, Population Division, Special Populations Branch (2000). Population 65
Years and Over, for States and for Puerto Rico: 1990 and 2000. Retrieved October 15,
2006 from: http://www.census.gov/population/www/cen2000/phc-t13.html
57
Weech-Maldonado, R., Neff, G., & Mor, V. (2003). The relationship between quality of care and
financial performance in nursing homes. Journal Of Health Care Finance, 29(3 (Print)),
48-60. Retrieved Thursday, January 04, 2007 from the MEDLINE database.
Yusuf, S. (2002, July 6). Two decades of progress in preventing vascular disease. Lancet,
360(9326), 2. Retrieved November 13, 2008, from Academic Search Premier database.
58
Appendix A
59
Appendix B
×

Report this document