HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Richard L. Prager Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Longo, K. M. Articles by Prager, R. L. Search for Related Content PubMed PubMed Citation Articles by Longo, K. M. Articles by Prager, R. L.

Ann Thorac Surg 1998;66:740-746
© 1998 The Society of Thoracic Surgeons

---

Original articles: Cardiovascular

Preoperative predictors of cost in medicare-age patients undergoing coronary artery bypass grafting

^a St. Joseph Mercy Hospital Ann Arbor, Michigan, USA
^b University of Michigan School of Public Health, Ann Arbor, Michigan, USA

Address reprint requests to Dr Prager, 5325 Elliott Dr, Suite 102, PO Box 972, Ann Arbor, MI 48106-0972

Presented at the Forty-fourth Annual Meeting of the Southern Thoracic Surgical Association, Naples, FL, Nov 6–8, 1997.

	Abstract

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
Background. Identification of preoperative factors that contribute to the cost of coronary artery bypass grafting could aid in predicting the procedure’s expense. In this study, 30 sociodemographic and clinical preoperative factors were examined with "survival analysis" techniques to determine characteristics related to total hospital cost.

Methods. Characteristics of all patients age 65 or older undergoing isolated coronary artery bypass grafting from July 1993 to April 1995 (n = 757) were recorded. Software was developed within the hospital’s Transitions Systems, Inc, database to calculate the outcome variable of total cost. Nonparametric methods were used for the univariate analysis of the data, and the Cox proportional hazards model was used for the multivariable analysis, censoring 25 patients who died in the hospital.

Results. Median hospital cost from the day of the operation until discharge was $15,198. Median length of stay after the operation was 6 days. Multivariable analysis revealed that age, preoperative renal failure, history of cerebrovascular accident, low ejection fraction, and surgical urgency were independent predictors of total cost.

Conclusions. This study, using an accurate representation of true hospital cost and a modeling technique that accounts for the confounding effect of in-hospital death on cost, provides a template for analysis of cost in other patient groups.

	Introduction

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
Coronary artery bypass grafting (CABG) is a well-accepted therapy for ischemic heart disease providing symptomatic palliation, myocardial preservation, increased functional ability, and improved survival when used on appropriately selected patients [1]. These benefits can be costly, however, because CABG is the most expensive operative procedure that is routinely reimbursed by Medicare [2].

As the proportion of elderly patients in the US population increases, more older patients are being referred for CABG. It is well known that the cost of CABG increases with age, but data regarding the cost-effective use of this procedure in the older population are incomplete [3–5]. It is not known whether some subsets of older patients are more or less expensive than others. If characteristics related to increased cost could be identified in these patients, specific preoperative interventions to decrease cost may be possible. For example, if patients with renal failure are found to be more costly, and the reasons for this are explored, preoperative care decisions might be altered to include a nephrology consultation in an effort to decrease cost. Further understanding of patient characteristics that contribute to increased costs for CABG may aid cost-effective analyses of this procedure.

Several studies have described important predictors of cost of CABG but have been limited by the use of charge data, or length of stay, as proxy measures for actual cost [6–9]. Actual cost is a more preferable outcome than charge data because it represents resource use for supplies, equipment, personnel, and utilities rather than profit margins, which are more likely to vary by institution [10].

A few preoperative characteristics—age, ejection fraction, extent of coronary artery disease, and congestive heart failure—have been identified as significant predictors of cost in more than one study. These predictors were often derived using proxy measures for cost such as charges and have explained little of the variance in cost [6–9]. Many preoperative factors relating to cost have yet to be identified. In this study other preoperative factors were examined to see whether any more of the variance in true cost of CABG can be explained in older patients.

	Material and methods

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
A computerized cardiovascular surgery registry was used to assemble a retrospective cohort of all patients, age 65 years or older, who were discharged after isolated CABG at St. Joseph Mercy Hospital between July 1993 and April 1995 (n = 770). St. Joseph Mercy Hospital is a 560-bed university-affiliated tertiary care teaching hospital with a group of 5 cardiovascular surgeons who perform more than 1,000 open heart procedures each year. The research protocol was approved by the Clinical Research Committee of the institution.

Seven clinical nurse specialists were responsible for the collection of preoperative data by abstracting the office and hospital records at the time of patient admission. All preoperative information was entered into a customized Summit Medical database that was created in 1984, and later modified to conform to The Society of Thoracic Surgeons definitions (Summit Medical Systems Inc, Minneapolis, MN).

A number of demographic, socioeconomic, and clinical variables were selected for analysis and defined using the Definitions of Terms of The Society of Thoracic Surgeons (STS) National Cardiac Surgery Database [11]. Sociodemographic characteristics such as age, sex, race, marital status, employment status, and preoperative and postoperative living arrangements were recorded. Patients who were widowed or separated were classified as unmarried. Patients living with a partner were classified as married. Patients who were semiretired or employed part-time were classified as employed. Living arrangement was defined as "other group setting" if it was a nursing home, retirement community, foster home, prison, or with friends.

Clinical variables included weight, height, angina type, preoperative cardiac arrest, preoperative cardiogenic shock, preoperative intraaortic balloon pump, renal failure (RF), preoperative congestive heart failure, history of cerebrovascular accident (CVA), history of diabetes mellitus, current or past tobacco use, hypertension, preoperative inotropic medication, preoperative intravenous nitrate, previous percutaneous transluminal coronary angioplasty, previous myocardial infarction, ejection fraction (EF), left main coronary artery disease greater than 50% stenosis, number of diseased coronary arteries, New York Heart Association (NYHA) functional class, initial or reoperative CABG, urgent or emergent operative priority at operation, and surgeon of record. All clinical variables were measured as defined in the STS database with the following modifications: a body mass index (kg/m²) of greater than 27 kg/m² was chosen to correspond to being at least 20% overweight for both sexes [12]; RF was defined as a documented serum creatinine level of greater than or equal to 1.8 mg/dL before the operation; angina was classified as none, exertional, at rest, or atypical; a patient was noted to have had a preoperative cardiac arrest if one occurred just before the admission or in the hospital before operation; preoperative congestive heart failure was determined to be present if the patient was on active medical treatment for congestive heart failure at the time of the operation; history of myocardial infarction was determined by electrocardiographic findings or patient history; elective operative priority was defined as a case that was scheduled at least 1 day in advance in which the patient had a stable cardiac status; urgent operative priority included patients who were admitted to the hospital for a cardiac problem and who were not allowed to be discharged before their CABG; emergent operative priority was defined as patients who required an emergent operation for unrelenting cardiac compromise, with or without hemodynamic instability, who were not responsive to any form of therapy except a cardiac operation.

The outcome variable of total cost was obtained through the St. Joseph Mercy Hospital Clinical Information System. Total cost was calculated using software developed by Transition Systems Incorporated (Boston, MA). Both variable and fixed direct costs, such as labor, occupancy expense, pharmacy, medical supplies, and laboratory testing, were calculated for each product or service using a relative value unit developed for each hospital department. A share of the hospital’s indirect costs, such as plant, administration, maintenance, and utilities, was then allocated to each department’s direct cost to obtain total cost. Costs related to the emergency department, and cardiac catheterization performed on or before the day of the operation, were excluded from the analysis. Physician professional fees were also excluded. Data for 13 patients (2%) could not be used because cost information was missing or inaccurate. Alive or dead status at discharge was obtained through the inpatient computerized database, and cross-referenced with the office database.

All analyses were performed using the SAS System for Windows, release 6.11 (SAS Institute, Cary, NC). Two-tailed tests of significance were performed, with statistical significance corresponding to a p value less than 0.05. The distribution of total cost was examined and found to be highly positively skewed, so nonparametric methods were used for univariate analysis. The Wilcoxon rank-sum test was performed on variables with two levels, the Kruskal-Wallis test was performed for variables with more than two levels, and the Spearman rank correlation coefficient was used for continuous variables [13]. Potential confounders were assessed in univariate analysis using the Mantel-Haenszel ² test [14].

The Cox proportional hazards model was used for the multivariable analysis in this study, censoring the costs of patients who died in the hospital [15]. All preoperative predictors with a p value of 0.20 or less were included in the initial model, along with sex, which was considered important a priori. Predictors were removed individually in a backward stepwise fashion. The significance level required for a variable to remain in the model was p less than or equal to 0.05. The proportional hazards assumption was examined graphically for all variables in the final model. Cost-dependent interaction terms were examined for variables when the proportional hazards assumption was in question. The linearity assumption for independent variables was tested using quadratic functions and spline functions when appropriate. Two-way interactions between all variables in the final model were evaluated at an of 0.01. Deviance residuals were plotted against each covariate, and no patterns were discerned [16]. Median costs were determined by finding the value associated with a cost probability of 50%.

	Results

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
The sociodemographic characteristics of the study population are outlined in Table 1. A total of 757 patients were included in the final analysis, 68% of whom were men. Mean age was 72.7 ± 5 years (±SE). The patients were predominately white (98%), married (72%), and retired (66%). Most patients lived with a spouse or other family member (79%) before the operation, and more planned to live with a family member after the operation (91%). Median length of stay after the operation was 6 days.

View larger version (14K): [in this window] [in a new window]   Fig 1. Kaplan-Meier function for 757 coronary artery bypass grafting patients. The horizontal axis indicates the cost in dollars, and the vertical axis indicates the proportion of patients at or above that level of cost.

Numerous clinical factors were found to be significantly related to cost (Table 3). As might be expected, RF, hypertension, and history of myocardial infarction or CVA were associated with a higher median cost. Indicators of poor cardiac status, such as having had a preoperative cardiac arrest or preoperative intravenous nitrate, lower EF%, and higher NYHA functional class, were also associated with increased cost. Patients having a repeat bypass, and those undergoing urgent or emergent operation, had significantly higher costs than the comparison groups. However, being in cardiogenic shock preoperatively was associated with a lower median cost, likely related to the high in-hospital mortality in this group. All other clinical factors, including operating surgeon, were not associated with higher cost.

Table 5 illustrates the median costs for a patient with the cohort mean age of 72 years in the presence or absence of various preoperative factors. The first row in the table reflects the baseline cost for this patient with no high-risk characteristics. The subsequent rows demonstrate the effect that adding one or more risk factors has on the total cost.

	Comment

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

Few studies have examined predictors of actual hospital costs in older patients undergoing isolated CABG operation using Cox proportional hazards analysis. Survival analysis has several advantages for analyzing cost data as compared with other types of multivariate analysis [9]. The distribution of cost data can be skewed and the Cox model requires no assumptions regarding the distribution of the dependent variable or its variance. Survival analysis techniques also enable the examination of the distribution of costs, rather than the average cost. The ability to censor the costs for patients who die in the hospital addresses the confounding effect of in-hospital mortality on cost. Because predictors of cost also predict mortality, and because mortality before discharge is associated with lower cost, it is inappropriate to examine predictors of cost without accounting for mortality. Finally, the significant cost-dependent interaction effects observed in this cohort demonstrate the need to test for nonproportional hazards to ensure the most accurate model for cost outcomes. The R² term is used in linear regression as an indication of how much of the variance of the dependent variable can be explained by the independent variable. In the study by Mauldin and associates [7] an R² of 19% was reported for their linear regression of clinical preoperative predictors of cost. In general, studies that use Cox proportional hazards regression do not report the equivalent of the R². We used a method described by Allison [18] to calculate an R² of 16% for this model. This value would suggest that there is a large amount of variation in cost that is not explained by these covariates.

This study had several limitations. As with most retrospective cohort studies, the validity of the findings are dependent on the accuracy of the data collection in the clinical and administrative databases used to determine the predictors and the outcome variable. The potential problems of data quality in STS registries and in administrative databases have been recognized in the literature [19, 20].

Treating deaths as censored observations may introduce the problem of "informative censoring" into the analysis [16]. In an attempt to analyze the effect of the 25 censored patients on these results, a sensitivity analysis was performed by running the final model on the subset of patients who were alive at discharge, and comparing the results to those from the entire cohort. The magnitude, direction, and significance of the main effects in the model did not change appreciably, aside from the p value of the coefficient for previous stroke, which changed from 0.04 to 0.14. The predictions of median cost in Table 5 were very similar except at very high costs. Although the technique of treating the costs of patients who die in the hospital as censored probably did not affect this analysis in a meaningful way, the question of how to treat these patients’ costs remains an area of concern. Other approaches to this problem have been described recently, and might be adapted for this type of analysis in the future [21].

Care is needed when attempting to generalize these findings to other settings. This study was performed at a large community teaching hospital, so the costs may not reflect those in different types of institutions. It should be emphasized that these results only apply to patients who survive CABG operation because of the censoring technique used. Also, as this was not a racially diverse group of patients, the results may not apply to more diverse populations.

The cost of CABG operation is, and will remain, an important issue as the number of older patients undergoing this procedure increases, and managed care takes a firm hold of the health care sector. Age alone is associated with a small but significant increase in cost. Other conditions may have a greater impact on total cost. The additional predictors described here—EF, operative priority, history of CVA, and RF—may not be modifiable in the short term to decrease cost in these patients, but more careful preoperative planning to address some of these factors may result in decreased hospital cost.

Information such as this, combined with further data on direct hospital costs, outpatient postoperative costs, survival, and quality of life, can aid health care professionals and administrators in planning for future resource allocation. Cost-benefit analyses usually assume one cost per procedure per patient. This study suggests that the cost-effectiveness of CABG may differ among subsets of elderly patients. The results of this study are meant to be used as a tool, when combined with other information, to achieve cost-effective health care planning in a system with finite resources. Caution should be used before applying these data to individual patients, as these findings provide information on only part of the larger picture of cost-effective resource allocation for the older patient undergoing CABG operation.

	Acknowledgments

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
This study was supported in part by a grant from the Blue Cross Blue Shield of Michigan Foundation. We thank Jaelene Williams, RN, for her support in data collection, Laura Kassler and Myra Daoud, MS, for their assistance with the TSI system, and Brenda Gillespie, PhD, and Robert Strawderman, ScD, for their advice on the analysis. The surgeons and staff of Cardiovascular and Thoracic Surgeons of Ann Arbor, PC, are to be thanked for their cooperation in this study.

	References

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 

1. Braunwald E, Mark DB, Jones RH, et al. Unstable angina: diagnosis and management. Clinical Practice Guideline Number 10. Rockville: AHCPR Publication No. 94-0602: Agency for Health Care Policy and Research and the National Heart, Lung, and Blood Institute, Public Health Service, US Department of Health and Human Services, March 1994:99.
2. Health Care Financing Review. Medicare and Medicaid Statistical Supplement, 1996. Health Care Financing Administration Publication No. 03386. Baltimore: Health Care Financing Administration, US Department of Health and Human Services, October 1996:54.
3. Krueger H., Goncalves J.L., Caruth F.M., Hayden R.I. Coronary artery bypass grafting: how much does it cost?. Can Med Assoc J 1992;146:163-168.[Abstract]
4. Peigh P.S., Swartz M.T., Vaca K.J., Lohmann D.P., Naunheim K.S. Effect of advancing age on cost and outcome of coronary artery bypass grafting. Ann Thorac Surg 1994;58:1362-1366.[Abstract]
5. Roberts A.J., Woodhall D.D., Conti C.R., et al. Mortality, morbidity and cost-accounting related to coronary artery bypass graft surgery in the elderly. Ann Thorac Surg 1985;39:426-432.[Abstract]
6. Smith L.R., Milano C.A., Molter B.S., Elberry J.R., Sabiston D.C., Jr, Smith P.K. Preoperative determinants of postoperative costs associated with coronary artery bypass graft surgery. Circulation 1994;90(Suppl 2):124-128.
7. Mauldin P.D., Weintraub W.S., Becker E.R. Predicting hospital costs for first-time coronary artery bypass grafting from preoperative and postoperative variables. Am J Cardiol 1994;74:772-775.[Medline]
8. Lahey S.J., Borlase B.C., Lavin P.T., Levitsky S. Preoperative risk factors that predict hospital length of stay in coronary artery bypass patients >60 years old. Circulation 1992;86(Suppl 2):181-185.
9. Dudley R.A., Harrell F.E., Jr, Smith L.R., et al. Comparison of analytic models for estimating the effect of clinical factors on the cost of coronary artery bypass graft surgery. J Clin Epidemiol 1993;46:261-271.[Medline]
10. Finkler S.A. The distinction between cost and charges. Ann Intern Med 1982;96:102-109.
11. The Society of Thoracic Surgeons National Cardiac Surgery Database. Manual for data managers. Minneapolis: Summit Medical Systems, Inc, 1995.
12. Williamson D.F. Descriptive epidemiology of body weight and weight change in U.S. adults. Ann Intern Med 1993;119:646-649.[Abstract/Free Full Text]
13. Snedecor G.W., Cochran W.G. Statistical methods, 7th ed. Ames: Iowa State University Press, 1980.
14. Mantel N., Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Am Stat Assoc 1959;22:719-748.
15. Cox DR, Oakes D. Analysis of survival data 1984. London: Chapman and Hall.
16. Collett D. Modeling survival data in medical research. London: Chapman and Hall, 1994.
17. Jones R.H., Hannan E.L., Hammermeister K.E., et al. Identification of preoperative variables needed for risk adjustment of short-term mortality after coronary artery bypass graft surgery. J Am Coll Cardiol 1996;28:1478-1487.[Abstract]
18. Allison P.D. Survival analysis using the SAS system: a practical guide. Cary, NC: SAS Institute Inc, 1995:248.
19. Grover F.L., Shroyer A.L.W., Edwards F.H., et al. Data quality review program: The Society of Thoracic Surgeons Adult Cardiac National Database. Ann Thorac Surg 1996;62:1229-1231.[Free Full Text]
20. Weinberger M, Hui S, Laine C, eds. Measuring quality, outcomes and cost of care using large databases. Ann Intern Med 1997;127:665–74.
21. Lin D.Y., Feuer E.J., Etzioni R., Wax Y. Estimating medical costs from incomplete follow-up data. Biometrics 1997;53:419-434.[Medline]

This article has been cited by other articles:

				E. Ott, C. D. Mazer, I. C. Tudor, L. Shore-Lesserson, S. A. Snyder-Ramos, B. A. Finegan, P. Mohnle, C. B. Hantler, B. W. Bottiger, R. D. Latimer, et al. Coronary artery bypass graft surgery care globalization: The impact of national care on fatal and nonfatal outcome J. Thorac. Cardiovasc. Surg., May 1, 2007; 133(5): 1242 - 1251. [Abstract] [Full Text] [PDF]

				L. V. Doering, A. W. McGuire, and D. Rourke Recovering From Cardiac Surgery: What Patients Want You To Know Am. J. Crit. Care., July 1, 2002; 11(4): 333 - 343. [Abstract] [Full Text] [PDF]

				J. T. Cope, A. K. Kaza, C. C. Reade, K. S. Shockey, J. A. Kern, C. G. Tribble, and I. L. Kron A cost comparison of heart transplantation versus alternative operations for cardiomyopathy Ann. Thorac. Surg., October 1, 2001; 72(4): 1298 - 1305. [Abstract] [Full Text] [PDF]

				D. M. Shahian, G. J. Heatley, and G. A. Westcott Relationship of hospital size, case volume, and cost for coronary artery bypass surgery: Analysis of 12,774 patients operated on in Massachusetts during fiscal years 1995 and 1996 J. Thorac. Cardiovasc. Surg., July 1, 2001; 122(1): 53 - 64. [Abstract] [Full Text] [PDF]

				S. Subramanian, C.-F. Liu, J. Cromwell, and S. Thestrup-Nielsen Preoperative Correlates of the Cost of Coronary Artery Bypass Graft Surgery: Comparison of Results From Three Hospitals American Journal of Medical Quality, May 1, 2001; 16(3): 87 - 92. [Abstract] [PDF]

				L. V. Doering, F. Esmailian, and H. Laks Perioperative Predictors of ICU and Hospital Costs in Coronary Artery Bypass Graft Surgery Chest, September 1, 2000; 118(3): 736 - 743. [Abstract] [Full Text] [PDF]

				T. B. Ferguson Jr, S. W. Dziuban Jr, F. H. Edwards, M. C. Eiken, A. L. W. Shroyer, P. C. Pairolero, R. P. Anderson, and F. L. Grover The STS National Database: current changes and challenges for the new millennium Ann. Thorac. Surg., March 1, 2000; 69(3): 680 - 691. [Abstract] [Full Text] [PDF]

				F. L. Grover The Society of Thoracic Surgeons National Database: current status and future directions Ann. Thorac. Surg., August 1, 1999; 68(2): 367 - 373. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Richard L. Prager Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Longo, K. M. Articles by Prager, R. L. Search for Related Content PubMed PubMed Citation Articles by Longo, K. M. Articles by Prager, R. L.