Ann Thorac Surg 1996;61:1879-1880
© 1996 The Society of Thoracic Surgeons
Correspondence
Outcome Versus Volume in Coronary-Bypass Operations
Victor Parsonnet, MD,
Alan D. Bernstein, EngScD
Division of Surgical Research Newark Beth Israel Medical Center 201 Lyons Ave Newark, Nj 07112
To the Editor:
The authors of several recent reports in The Annals on the relationship between surgical volume and operative mortality have concluded in one way or another that the relationship is weak at best, and nonexistent for high surgical volume [1–3].
We are concerned that readers may have been led to believe that there is no relationship between outcome and volume even when the volume is very small. From the reports by Hannan and associates [4] and the Pennsylvania group [5], and from our own ongoing study in New Jersey, we know that many surgeons perform fewer than 50 operations a year, and that low-volume surgeons tend to have relatively high mortality rates. Because that is so, it may not be appropriate for us to say that there is no relationship between outcome and volume no matter how small the surgical volume.
We surgeons all know that there is some relationship, having learned from our personal experience that more work honed our skills. Even the data shown in the scatter plots in Clark and colleagues' [2] report show that there is a wide spread of mortality among low-volume hospitals, whereas large-volume groups had uniformly low mortality rates. Several pieces of evidence from recent experience may illustrate the relationship between volume and outcome. In each case, we examined risk-adjusted mortality by plotting the ratio of observed to expected mortality for coronary-bypass procedures alone as a function of surgical volume. Figure 1
shows this relationship for 60 surgeons in New Jersey who performed at least 10 coronary-bypass operations during 1994. There is an obvious knee in the hyperbolic regression model at about 50 to 75 cases a year. Almost no surgeon with a volume of more than 100 cases per year had an observed/expected ratio greater than 1.0.
View larger version (32K):
[in this window]
[in a new window]
|
Fig 1. . Results for aortocoronary-bypass (ACB) procedures at New Jersey hospitals during 1994 by 60 surgeons who performed more than 10 such operations. The observed/expected (O/E) ratio (ordinate) was derived from ACB procedures alone; the volume (abcissa) includes all cases in which ACB was performed. A hyperbolic regression model is superimposed. The Pearson correlation for the model is 0.461, and the relationship is significant at the p < 0.001 level (two-tailed significance). Risk-adjusted  2 analysis shows that the surgeons who performed 75 procedures per year or more achieved significantly better results than those who performed fewer than 75 procedures (O/E ratios = 0.46 and 0.73 respectively; two-tailed p < 0.02).
|
|
Figure 2 shows data that we copied from Figure 3 of Clark and colleagues [2]; each point represents a single surgical center. Again, one can see that the poorest results occurred among the groups with the lowest volume. Hospitals with more than 500 cases per year rarely have observed/expected ratios greater than 1.0, whereas many hospitals with volumes less than 250 a year have poor observed/expected ratios.
View larger version (33K):
[in this window]
[in a new window]
|
Fig 2. . Scatter plot of observed/expected (O/E) ratio versus annualized group caseload as shown in Figure 3 of Clark and colleagues [2]. A logarithmic regression model is superimposed. The Pearson correlation of the O/E ratio and the natural logarithm of the annualized group caseload is -0.896, and the relationship is significant at the p < 0.001 level (two-tailed significance). (Adapted by permission of The Society of Thoracic Surgeons [Ann Thorac Surg 1996;61:21--6].)
|
|
View larger version (33K):
[in this window]
[in a new window]
|
Fig 3. . Results of aortocoronary-bypass (ACB) procedures performed by 143 Pennsylvania surgeons during 1993 [5]. A linear regression model is superimposed. The Pearson correlation is -0.160 and the relationship is significant at the level of p < 0.06 (two-tailed). Risk-adjusted 2 analysis shows that the surgeons who performed 80 procedures per year or more achieved significantly better results than those who performed fewer than 80 procedures (observed/expected [O/E] ratios = 0.94 and 1.27 respectively; two-tailed p < 0.001).
|
|
Figure 3 was derived from data tabulated in the recently distributed Pennsylvania Health Care Report [5]. The results of 143 surgeons were poorest for low-volume surgeons, with the cutoff point at roughly 80 cases per year. Figure 4 shows the results of 39 Pennsylvania hospitals as reported in the same document.
View larger version (29K):
[in this window]
[in a new window]
|
Fig 4. . Results of aortocoronary-bypass (ACB) procedures performed at 39 Pennsylvania hospitals during 1993 [5]. A logarithmic regression model is superimposed. The Pearson correlation is -0.311 and the relationship is significant at the level of p = 0.051 (two-tailed). Risk-adjusted 2 analysis shows that the centers whose volume was 275 procedures per year or more achieved significantly better results than those whose volume was fewer than 275 procedures (observed/expected [O/E] ratios = 0.94 and 1.41 respectively; two-tailed p = 0.001).
|
|
Consequently, from these analyses, it appears that there is indeed an inverse relationship between volume and quality.
References
- Shroyer ALW, Dauber I, Jones RH, et al. Provider perceptions in using outcome data to improve clinical practice. Ann Thorac Surg 1994;58:1877–80.[Abstract/Free Full Text]
- Clark RE, Ad Hoc Committee on Cardiac Surgery Credentialing of The Society of Thoracic Surgeons. Outcome as a function of annual coronary artery bypass graft volume. Ann Thorac Surg 1996;61:21–6.[Abstract/Free Full Text]
- Crawford FA, Anderson RP, Clark RE, et al. Volume requirements for cardiac surgery credentialing: a critical examination. Ann Thorac Surg 1996;61:12–6.[Abstract/Free Full Text]
- Hannan EL, Kilburn H Jr, O'Donnell JF, et al. Adult open heart surgery in New York State: an analysis of risk factors and hospital mortality rates. JAMA 1990;264:2768–74.[Abstract/Free Full Text]
- Coronary artery bypass graft surgery: technical report volume IV, 1993. Philadelphia: The Pennsylvania Health Care Cost Containment Council. 1993:52, 65–74.
