Correspondence
Volume of Coronary Angioplasty Procedures and Mortality Rates
N Engl J Med 1995; 332:1304-1305May 11, 1995
- Article
To the Editor:
The article by Jollis et al. (Dec. 15 issue)1 concerning the relation between the annual volume of percutaneous transluminal coronary angioplasty (PTCA) procedures and mortality is incomplete and misleading. Data on Medicare admissions for PTCA from October 1991 through June 1994 in New Hampshire suggest that Jollis et al. disregarded robust risk factors that explain most of the variation in outcome attributed to hospital volume. Consequently, the findings are systematically biased against hospitals that perform fewer procedures.
Key unexamined claims data include the type of admission, whether the surgery was elective or nonelective, and the risk associated with the primary diagnosis. Of 1104 admissions for PTCA in New Hampshire, 61 percent of those in the highest-volume group of hospitals were elective, whereas 95 percent of those in the lowest-volume group were nonelective (chi-square with 2 df = 196.5; P<0.001). Similarly, an analysis of empirically defined2 patient groups based on the risk of short-term mortality associated with principal diagnoses reveals marked differences among hospital groups. In New Hampshire, a disproportionately large number of patients with high-risk principal diagnoses, such as “other anterior-wall acute myocardial infarction” and “other inferior-wall acute myocardial infarction” (International Classification of Diseases, 9th Revision, Clinical Modification [ICD-9-CM] codes 410.11 and 410.41, respectively) were admitted to the lowest-volume hospitals (70 percent, vs. 13 percent for the highest-volume hospitals). For lower-risk diagnoses, such as “unspecified chest pain” and “coronary atherosclerosis” (ICD-9-CM codes 786.50 and 414.0, respectively), the highest-volume group of hospitals admitted a disproportionately large number of patients (71 percent, vs. 14 percent for the lowest-volume hospitals; chi-square with 8 df = 127.4; P<0.001).
These findings suggest a nonelective, high-risk profile of patient admissions at low-volume hospitals. Inclusion of these risk variables in the state mortality model effectively nullifies the predictive value of the volume of procedures and demonstrates the effect of unaccounted sources of systematic bias.
2 ReferencesEdwin D. Huff, Ph.D.
New Hampshire Foundation for Medical Care, Dover, NH 03820-28301
Jollis JG ,Peterson ED ,DeLong ER . The relation between the volume of coronary angioplasty procedures at hospitals treating Medicare beneficiaries and short-term mortality. N Engl J Med 1994;331:1625-1629
Full Text | Web of Science | Medline2
Health Care Financing Administration. Medicare hospital information report. Vol. 55. 1992 Technical suppl. Analytical categories. Section A, Appendix 1. Washington, D.C.: Government Printing Office, 1992.
To the Editor:
Jollis et al. examine the relation between the volume of PTCA procedures and short-term mortality among Medicare patients but do not provide any information about the volume of procedures per operator. Yet the authors support the guidelines established by the American College of Cardiology and the American Heart Association1 for the minimal volume of procedures per hospital and per operator. The authors of the guidelines, in turn, reached their conclusions by citing another article,2 which provided no data. Jollis et al. acknowledge that “the relation of mortality and volume per physician cannot be examined . . . [because there were] no physician identifiers,” yet they insist that their “findings are consistent with minimal annual volumes per hospital and per operator.” They do not examine the severity of illness, diagnoses, indications, angiographic findings, need for multiple procedures, or restenosis — all possible explanations of higher mortality rates. The authors also do not take into account whether acute myocardial infarction occurred before, during, or after angioplasty. Moreover, they cite an abstract3 to refute the possibility that lower-volume hospitals perform procedures in sicker patients. They do not explore the possibility that high-volume, aggressive angioplasty centers (or operators) perform more unnecessary procedures than low-volume, conservative, yet economically efficient, scientifically serious, and academically sound centers (or operators).
3 ReferencesSergio E. Schabelman, M.D.
Louisiana State University Medical Center, New Orleans, LA 701121
Ryan TJ ,Bauman WB ,Kennedy JW , et al. Guidelines for percutaneous transluminal coronary angioplasty. Circulation 1993;88:2987-3007
Web of Science | Medline2
Ryan TJ ,Klocke FJ ,Reynolds WA . Clinical competence in percutaneous transluminal coronary angioplasty: a statement for physicians from the ACP/ACC/AHA Task Force on Clinical Privileges in Cardiology. J Am Coll Cardiol 1990;15:1469-1474
CrossRef | Web of Science | Medline3
Krone RJ, Vetrovec GW, Noto TJ, Johnson LW, Registry Committee, Medical College of Virginia
Author/Editor Response
The authors reply:
To the Editor: Dr. Huff's inability to find a relation between the volume of angioplasty procedures and outcome that is similar to the relation we reported is not surprising, given the limited size of his data set (1104 patients treated at four hospitals). If one assumes the same difference in mortality as that found in our study for a study involving 1100 patients, with an alpha level of 0.1, he has more than a 90 percent chance of a type II error (missing a significant difference in mortality for patients treated at the lowest-volume hospitals). Our study involved 217,836 patients at 1194 hospitals, allowing us to detect the relation between volume and mortality despite the relatively infrequent occurrence of death.
Among the four hospitals in his data set, Dr. Huff also found marked variation in the prevalence of insurance-claims data involving the urgency of the admission, the location of “other” infarcts, and nonspecific diagnoses of coronary syndromes. With so few hospitals, differences in coding practices can result in marked variations in the prevalence of these elements that is independent of the severity of illness, and the reliability of such data should be adequately validated before they are used to account completely for differences in outcome. There are a number of alternative clinical descriptors in insurance-claims data that can be used to account (at least partially) for differences among patients. We believe that we made a conservative and reasonable selection of descriptors for our model, to the extent that claims data permit adjustments for the severity of illness.
In his account of the derivation of the guidelines for coronary angioplasty, Dr. Schabelman overlooked the study by Ritchie and colleagues cited by the guidelines document and our study demonstrating higher rates of bypass surgery for patients undergoing angioplasty at low-volume hospitals in California.1,2 We agree with Dr. Schabelman that more detailed clinical data on the severity of illness need to be collected and examined to substantiate the coronary-angioplasty guidelines of the American College of Cardiology and the American Heart Association. Some of the ongoing efforts to collect detailed data involve proprietary data bases such as the National Cardiovascular Network data base, state registries such as the New York Coronary Angioplasty Reporting System, and federal initiatives such as the Cooperative Cardiovascular Project supported by the Health Care Financing Administration. Until such detailed data are available, the American College of Cardiology–American Heart Association expert panel and the data from California,1,2 as well as our analyses, suggest that angioplasty programs should not be operated in hospitals with very low volumes of procedures (fewer than four per week).
2 ReferencesJames G. Jollis, M.D.
Eric D. Peterson, M.D., M.P.H.
Elizabeth R. DeLong, Ph.D.
Daniel B. Mark, M.D., M.P.H.
Duke University Medical Center, Durham, NC 277101
Ritchie JL ,Phillips KA ,Luft HS . Coronary angioplasty: statewide experience in California. Circulation 1993;88:2735-2743
Web of Science | Medline2
Ryan TJ ,Bauman WB ,Kennedy JW , et al. Guidelines for percutaneous transluminal coronary angioplasty. Circulation 1993;88:2987-3007
Web of Science | Medline
- Citing Articles (1)
Citing Articles
1
Thomas P Wharton, Nancy Sinclair McNamara, Frank A Fedele, Mark I Jacobs, Alan R Gladstone, Erik J Funk. (2000) Reply. Journal of the American College of Cardiology 36:1, 301-303
CrossRef
