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 Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ghali, W. A. Articles by Moskowitz, M. A. Search for Related Content PubMed PubMed Citation Articles by Ghali, W. A. Articles by Moskowitz, M. A.

Ann Thorac Surg 1999;67:441-445
© 1999 The Society of Thoracic Surgeons

---

Original Articles

Variation in hospital rates of intraaortic balloon pump use in coronary artery bypass operations

^a Health Care Research Unit, Evans Department of Medicine, Boston University School of Medicine, Boston, Massachusetts, USA

Accepted for publication July 6, 1998.

Address reprint requests to Dr Ghali, Faculty of Medicine, University of Calgary, 3330 Hospital Dr NW, Calgary, Alberta, Canada, T2N 4N1
e-mail: wghali{at}ucalgary.ca

	Abstract

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
Background. Little is known about regional patterns of intraaortic balloon pump (IABP) use in coronary artery bypass graft (CABG) operations. Our objectives were (1) to identify clinical variables associated with IABP use, and (2) to examine risk-adjusted rates of IABP use for 12 Massachusetts hospitals performing CABG operations.

Methods. We used hospital discharge data to identify 6944 CABG surgical cases. Logistic regression was used to identify clinical variables associated with IABP use, and the resulting multivariate model was then used to risk adjust hospital rates of IABP use.

Results. The IABP was used in 13.4% of the CABG surgical cases. The clinical variables independently associated with IABP use were cardiogenic shock, same admission angioplasty, prior CABG operation, cardiac arrest, congestive heart failure, recent myocardial infarction, and urgent admission status. Risk-adjusted rates of IABP use varied widely across hospitals from 7.8% to 20.8% (p < 0.0001).

Conclusions. Hospital rates of IABP use vary considerably in Massachusetts. This practice variation may be related to the persistent uncertainty regarding the precise clinical indications for the IABP in this patient population.

	Introduction

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
Coronary artery bypass graft (CABG) operations have been the focus of several studies evaluating outcomes of care. Many of these studies have demonstrated regional and between-hospital variability in postoperative outcomes that are not explained by adjusting for patients’ baseline clinical characteristics [1–4]. Such variation in outcomes across providers raises the possibility of differences in quality of care across regions and hospitals. These findings have prompted calls [5, 6] for more research focusing on process of care variables—rather than just outcomes—to identify factors that contribute to variation in outcomes across providers.

Intraaortic balloon pump (IABP) use in CABG operation is an important process variable that may vary across hospitals and affect patient outcomes. The IABP has established physiologic benefits in patients with compromised cardiac function [7], and underuse of this technology has the potential to produce poor patient outcomes. However, IABP use itself can lead to complications (eg, infection, hemorrhage, peripheral embolism, aortic rupture) in as many as 15% to 20% of cases [8, 9]. Overuse of the IABP therefore also has the potential to produce poor patient outcomes. Little is known about regional patterns of IABP use, and whether rates of use are consistent across providers—an important question considering the potential impact of this technology on patient outcomes.

Our objectives were (1) to identify clinical and sociodemographic variables associated with IABP use, and (2) to use a multivariate model to examine risk-adjusted rates of IABP use across Massachusetts hospitals performing CABG operations.

	Material and methods

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
Data source
We used a statewide database maintained by the Massachusetts Health Data Consortium to identify all CABG surgical cases in Massachusetts in the fiscal year 1994 (October 1, 1993, through September 30, 1994). The database contains one record for each hospital discharge in Massachusetts [10], and each inpatient record contains up to 15 International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnoses and 15 procedure codes [11]. The records also contain information on patient age, sex, race, payer (eg, Medicaid, Medicare, health maintenance organization [HMO]), length of stay, admission type (elective, urgent, emergent), and diagnosis related group (DRG) [12].

Derivation of study cohort
We identified CABG surgical cases by screening all hospital discharge abstracts for ICD-9-CM procedure codes representing CABG operations (36.10 through 36.19). This case selection criterion captured all patients who had a simple CABG procedure, as well as those who had a combined CABG plus valve procedure. Cases outside of DRGs 104 through 108 and 483 were excluded (n = 42) because the clinically dissimilar cases assigned to other DRGs usually involved hospitalizations with other noncardiac surgical procedures. Cases in DRG 483 (denoting tracheostomy) were reclassified to the DRG category (104 through 108) to which they would have been assigned had tracheostomy not occurred. All cases from a single Veterans Affairs hospital were also excluded because of important data inconsistencies. This case selection process yielded a study cohort of 6,944 CABG surgical cases performed in 12 hospitals. The volume of cases per hospital ranged from a low of 330 to a high of 946.

Definitions of study variables
The process variable of interest was IABP use, defined by the presence of ICD-9-CM procedure code 37.61. Because we could not distinguish between preoperative and postoperative IABP use in our database, the focus of this study is on overall hospital rates of IABP use. A number of other variables were assessed as potential correlates of IABP use. These were grouped into three categories: sociodemographic, comorbidity, and disease-specific variables.

The sociodemographic variables were age, sex, race (white versus other), and socioeconomic status (defined as "low" for cases in which care was free or the payer was Medicaid). Fifteen comorbidity variables (eg, diabetes, peripheral vascular disease, cerebrovascular disease, chronic lung disease) were defined using an ICD-9-CM coding algorithm derived by Deyo and colleagues [13]. We then used these comorbidity variables to assign a comorbidity score to each CABG surgical case. The score consisted of the 17-variable Charlson index [14], altered by removing two disease-specific variables (congestive heart failure and myocardial infarction), which were evaluated separately.

The disease-specific variables and corresponding ICD-9-CM code definitions were congestive heart failure (ICD-9-CM code 428), recent myocardial infarction (410.0x through 410.9x), prior CABG operation (v458.1), angioplasty on the same admission (procedure codes 36.01, 36.03, or 36.09), acquired ventricular septal defect (429.71), ruptured chordae tendinae (429.5 or 429.6), cardiogenic shock (785.51), and cardiac arrest (427.4, 427.5, 427.41, or 427.42). Combined CABG plus valve procedures were identified by patient assignments to DRGs 104 and 105. Urgent admission status (ie, urgent or emergent admission as opposed to elective) was determined from "admission type."

Analysis
We used ² and Fisher’s exact tests, when appropriate, to test potential associations between IABP use and each of the variables defined above. From the results of these bivariate analyses, we selected variables that appeared in at least 10 cases and showed even weak (p 0.15) associations with IABP use for presentation to a logistic regression model. In our multivariate analysis, we retained only variables significantly associated with IABP use (p 0.05).

We assessed the resulting multivariate model’s discrimination for predicting IABP use with the c statistic, which equals the area under the receiver operating characteristic curve. Model calibration was assessed using the Hosmer-Lemeshow test of goodness-of-fit across categories of predicted risk.

We then used the model to calculate risk-adjusted rates of IABP use for each of the 12 Massachusetts hospitals performing CABG operations as follows. First, we calculated the hospitals’ observed and expected rates of IABP use, and calculated each hospital’s ratio of observed-to-expected use. We then multiplied this ratio by the statewide rate of use to yield a risk-adjusted rate.

Because it is not clear whether higher or lower rates of IABP use are better (unlike mortality where lower is clearly better), we did not seek to identify outlier hospitals. Rather, we present the full range of IABP rates along with the extremal quotient (ratio of highest rate divided by lowest rate) and the coefficient of variation (ratio of the standard deviation of hospital rates divided by the average hospital rate) to quantify the degree of variation. We used a ² test with 11 degrees of freedom to assess the statistical significance of the differences in risk-adjusted IABP rates across hospitals. This was achieved by evaluating the increment in our logistic model’s ^-2 log likelihood ² statistic when dummy variables for the hospital were added to the model.

	Results

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
Table 1 presents the baseline characteristics of the 6,944 CABG surgical cases studied. Mean age was 66.0 years, and 62.8% were admitted urgently. The overall statewide rate of IABP use was 13.4%. The statewide in-hospital mortality rate was 3.9%.

View larger version (22K): [in this window] [in a new window]   Fig 1. Unadjusted and adjusted rates of intraaortic balloon pump (IABP) use in 12 Massachusetts hospitals performing coronary artery bypass graft operations.

	Comment

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
We have identified a number of clinical variables that correlate with IABP use in CABG operations, and despite adjusting for these variables have found wide variation in hospital rates of IABP use. This is an important finding because IABP use is a process variable with the potential to have both a positive and negative impact on patients’ clinical outcomes after CABG operations.

The multivariate correlates of IABP use (Table 3) all fall within our predefined category of disease-specific variables. This suggests that it is clinical parameters relating to cardiac disease that have the greatest influence on clinicians’ decisions to use the IABP. Interestingly, burden of comorbidity (measured by the Charlson comorbidity index [13, 14]) was not independently associated with IABP use.

Reassuringly, sociodemographic variables were not independently associated with IABP use. This contrasts with published findings of lower use of cardiac procedures in women (versus men) [15, 16], blacks (versus whites) [17], and Medicaid (versus insured) patients [18]. Our findings thus indicate that once a patient has been selected for CABG operation, the IABP appears to be used equally regardless of patients’ age, sex, race, and socioeconomic status.

Most published articles reviewing clinical indications for IABP use distinguish between preoperative and intraoperative or postoperative indications [19–22]. There is little controversy about the clinical indications for intraoperative and postoperative IABP use. These are (1) inability to wean patients from cardiopulmonary bypass within 30 minutes of completing the operation, (2) poor hemodynamic status (systolic blood pressure <80 mm Hg, left ventricular end-diastolic pressure >20 mm Hg) despite inotropic therapy and vasodilators, (3) persistent requirements for inotropic therapy, and (4) arrhythmias attributable to cardiac ischemia.

In contrast, there is considerable disagreement regarding preoperative indications. Some authors [19, 20] advocate use of the IABP for the following indications: (1) acute myocardial infarction with new mechanical defect (ventricular septal defect or acute mitral regurgitation), (2) cardiogenic shock, (3) left main coronary artery stenosis, (4) impaired left ventricular function, and (5) preoperative unstable angina. Others [21, 22] have questioned the need for IABP use in patients with the latter three indications. In arguing against widespread preoperative IABP use, they report that many patients with left main coronary artery stenosis, poor ventricular function, and unstable angina do well without the IABP. The question of patient selection for preoperative IABP use remains unresolved. A recent retrospective analysis of patients undergoing CABG operation with ejection fraction of <25% suggests benefit from IABP use, but the study was not randomized, and controls differed considerably from IABP-treated patients [23]. A randomized controlled trial is needed to definitively assess the efficacy of preoperative IABP use for selected patients undergoing CABG operation.

This controversy about indications for IABP use may be responsible for the apparent practice variation in Massachusetts. Indeed, others [24] have found the largest surgical rate variations when procedural indications are unclear (ie, clear indications result in less variation). Wennberg [25] has stated that "variations occur because the profession lacks consensus on the correct way to practice medicine," and that clinical research should target those interventions for which there is greatest variation.

Our data do not clarify which rate of IABP use is right. Although IABP-treated patients were more likely to die and to experience postoperative complications (data not shown), this probably reflects the fact that patients who received the IABP were initially sicker. We also cannot tell whether the differences in rates of IABP use arise from preoperative or postoperative practice variation. This latter question needs to be examined in databases that distinguish between preoperative and intraoperative or postoperative IABP insertion.

Another important limitation of our study is our use of ICD-9-CM administrative data. Concerns with such data include the possibilities of undercoding of certain data elements and variable coding practices across hospitals. Furthermore, administrative data lack important clinical variables, which may explain at least some of the observed practice variation. For example, a greater proportion of the variation in hospital rates of IABP use might have been explained had we been able to adjust hospital IABP rates for clinical variables such as left ventricular ejection fraction or end-diastolic pressures. Reassuringly, despite these concerns, we were nonetheless able to develop and apply a clinically plausible multivariate model (Table 3) that is strongly predictive. We encourage investigators with access to more detailed regional clinical databases to see whether our findings can be replicated after adjustment for additional clinical variables such as ejection fraction.

Despite these limitations, this study identifies clinical correlates of IABP use and finds significant variation in hospital rates of IABP use. Further work is now needed to clarify the true extent of interhospital variation after adjustment for more detailed clinical variables, as well as the specific timing of practice variation (preoperative versus intraoperative or postoperative). More importantly, randomized controlled trials are needed to evaluate the true efficacy of elective IABP use in selected CABG surgical patients. Until such studies are done, clinicians will continue to struggle with the difficult question of which rate is right.

	Acknowledgments

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
We acknowledge the support of Hude Quan, MD, PhD, (University of Calgary) in data management and analysis. This project was supported by the Massachusetts Health Data Consortium, and by a grant from the Walnut Medical Charitable Trust. Doctor Ghali is supported by a Population Health Investigator grant from the Alberta Heritage Foundation for Medical Research.

	References

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 

1. Hannan E.L., Kilburn H., Jr, O’Donnell J.F., Lukacik G., Shilds E.P. Adult open heart surgery in New York state: an analysis of risk factors and hospital mortality rates. JAMA 1990;264:2768-2774.[Abstract]
2. O’Connor G.T., Plume S.K., Olmstead E.M., et al. Northern New England Cardiovascular Disease Study Group. A regional prospective study of in-hospital mortality associated with coronary artery bypass grafting. JAMA 1991;266:803-809.[Abstract]
3. Hartz A.J., Kuhn E.M. Comparing hospitals that perform coronary artery bypass surgery: the effect of outcome measures and data sources. Am J Public Health 1994;84:1609-1614.[Abstract/Free Full Text]
4. Silber J.H., Rosenbaum P.R., Schwartz J.S., Ross R.N., Williams S.V. Evaluation of the complication rate as a measure of quality of care in coronary artery bypass graft surgery. JAMA 1995;274:317-323.[Abstract]
5. Huff E.D. Have quality improvement efforts decreased mortality rates following cardiac surgery?. JAMA 1997;277:1518-1519.[Medline]
6. Naylor C.D. Assessing processes and outcomes of medical care. Ann R Coll Physicians Surg Can 1997;30:157-161.
7. Kusiak V.M., Goldberg S. Percutaneous intra-aortic balloon counterpulsation. Cardiovasc Clin 1985;15:281-302.[Medline]
8. Miller J.S., Dodson T.F., Salam A.A., Smith R.B. Vascular complications following intra-aortic balloon pump insertion. Am Surg 1992;58:232-238.[Medline]
9. Pennington D.G., Swartz M., Codd J.E., Merjavy J.P., Kaiser G.C. Intra-aortic balloon pumping in cardiac surgical patients: a nine-year experience. Ann Thorac Surg 1983;36:125-131.[Medline]
10. Densen P.M., Fielding J.E., Getson J., Stone E. The collection of data on hospital patients—the Massachusetts Health Data Consortium approach. N Engl J Med 1980;302:171-173.[Medline]
11. International classification of diseases, ninth revision, clinical modification. Washington, DC: Public Health Services, US Department of Health and Human Services, 1988.
12. Diagnosis related groups, version 10.0: definitions manual. Murray, UT: 3M Health Information Systems, 1992.
13. Deyo R.A., Cherkin D.C., Ciol M.A. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992;45:613-619.[Medline]
14. Charlson M.E., Pompei P., Ales K.L., MacKenzie C.R. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373-383.[Medline]
15. Ayanian J.Z., Epstein A.M. Differences in the use of procedures between women and men hospitalized for coronary heart disease. N Engl J Med 1991;325:221-225.[Abstract]
16. Naylor C.D., Levinton C.M. Sex-related differences in coronary revascularization practices: the perspective from a Canadian queue management project. Can Med Assoc J 1993;149:965-973.[Abstract]
17. Escarce J.J., Epstein K.R., Colby D.C., Schwartz J.S. Racial differences in the elderly’s use of medical procedures and diagnostic tests. Am J Public Health 1993;83:948-954.[Abstract/Free Full Text]
18. Young G.J., Cohen B.B. Inequities in hospital care, the Massachusetts experience. Inquiry 1991;28:255-262.[Medline]
19. Gustensen J., Goldman B.S., Scully H.E., Huckell V.F., Adelman A.G. Evolving indications for preoperative intra-aortic balloon pump assistance. Ann Thorac Surg 1976;22:535-543.[Abstract]
20. Bolooki H., Williams W., Thurer R.J., et al. Clinical and hemodynamic criteria for use of the intra-aortic balloon pump in patients requiring cardiac surgery. J Thorac Cardiovasc Surg 1976;72:756-768.[Abstract]
21. Mundth E.D. Preoperative intraaortic balloon pump assistance. Ann Thorac Surg 1976;22:603-604.[Medline]
22. Kaplan J.A., Craver J.M., Jones E.L., Sumpter R. The role of the intra-aortic balloon in cardiac anesthesia and surgery. Am Heart J 1979;98:580-586.[Medline]
23. Dietl C.A., Berkheimer M.D., Woods E.L., Gilbert C.L., Pharr W.F., Benoit C.H. Efficacy and cost-effectiveness of preoperative IABP in patients with ejection fraction of 0.25 or less. Ann Thorac Surg 1996;62:401-409.[Abstract/Free Full Text]
24. Hall R.E., Cohen M.M. Variations in hysterectomy rates in Ontario: does the indication matter?. Can Med Assoc J 1994;151:1713-1719.[Abstract]
25. Wennberg J.E. The paradox of appropriate care. JAMA 1987;258:2568-2569.[Medline]

This article has been cited by other articles:

				R. J. F. Baskett, G. T. O'Connor, G. M. Hirsch, W. A. Ghali, K. Sabadosa, J. R. Morton, C. S. Ross, F. Hernandez, W. C. Nugent Jr, S. J. Lahey, et al. A multicenter comparison of intraaortic balloon pump utilization in isolated coronary artery bypass graft surgery Ann. Thorac. Surg., December 1, 2003; 76(6): 1988 - 1992. [Abstract] [Full Text] [PDF]

				J. T. Christenson Invited commentary Ann. Thorac. Surg., December 1, 2003; 76(6): 1992 - 1992. [Full Text] [PDF]

				M. Cohen, P. Urban, J. T. Christenson, D. L. Joseph, R. J. Freedman Jr, M. F. Miller, E.M. Ohman, R. C. Reddy, G. W. Stone, J. J. Ferguson III, et al. Intra-aortic balloon counterpulsation in US and non-US centres: results of the Benchmark(R) Registry Eur. Heart J., October 1, 2003; 24(19): 1763 - 1770. [Abstract] [Full Text] [PDF]

				J. T. Christenson, M. Cohen, J. J. Ferguson III, R. J. Freedman, M. F. Miller, E. M. Ohman, R. C. Reddy, G. W. Stone, and P. M. Urban Trends in intraaortic balloon counterpulsation complications and outcomes in cardiac surgery Ann. Thorac. Surg., October 1, 2002; 74(4): 1086 - 1090. [Abstract] [Full Text] [PDF]

				R. J.F. Baskett, W. A. Ghali, A. Maitland, and G. M. Hirsch The intraaortic balloon pump in cardiac surgery Ann. Thorac. Surg., October 1, 2002; 74(4): 1276 - 1287. [Abstract] [Full Text] [PDF]

				W. L. Holman, Q. Li, C. I. Kiefe, D. C. McGiffin, E. D. Peterson, R. M. Allman, V. G. Nielsen, and A. D. Pacifico Prophylactic value of preincision intra-aortic balloon pump: Analysis of a statewide experience J. Thorac. Cardiovasc. Surg., December 1, 2000; 120(6): 1112 - 1119. [Abstract] [Full Text] [PDF]

				J. T. Christenson Preoperative intraaortic balloon therapy Ann. Thorac. Surg., October 1, 1999; 68(4): 1438 - 1438. [Full Text] [PDF]

				W. A. Ghali Reply Ann. Thorac. Surg., October 1, 1999; 68(4): 1438 - 1439. [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 Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ghali, W. A. Articles by Moskowitz, M. A. Search for Related Content PubMed PubMed Citation Articles by Ghali, W. A. Articles by Moskowitz, M. A.