| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2004;77:488-495
© 2004 The Society of Thoracic Surgeons
a Cardiac Surgical Associates, P.A., St. Paul, Minnesota, USA
Accepted for publication July 21, 2003.
* Address reprint requests to Dr Northrup, Cardiac Surgical Associates, P.A., 2356 University Ave W, Suite 258, St. Paul, MN 55407, USA.
e-mail: north7{at}earthlink.net
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
METHODS: We retrospectively reviewed all 30,319 adult CAB and VO in our group from 1979 to 1999 according to specific procedures.
RESULTS: Coronary artery bypass volumes peaked in 1996 at 1,895 cases, declining 15.3% to 1,605 cases in 1999 with a decrease in risk profile and percent reoperations and an increase in mean age and percent octogenarians, prior percutaneous coronary interventions (PCI), left internal mammary artery (LIMA) graft usage, off-pump technology usage, and hospital mortality of reoperations. Right internal mammary grafts were employed infrequently and radial artery grafts transiently. Overall VO volumes continued to increase 24.0% since 1996, from 470 to 583 cases with a decreased risk profile, increased mean age, and percent octogenarians and prior PCI. The percentage of mechanical valve implants decreased, while the percentage of various tissue solutions for valve disease increased. Limited access incisions and port-access were employed transiently with CAB and VO.
CONCLUSIONS: Coronary artery bypass volumes are decreasing, with an increasing percentage of LIMA grafts and off-pump cases. Valve operation volumes are steadily increasing, with a decreasing percentage of mechanical valve implants, in favor of various tissue solutions.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
Although PCI technologies may be eroding current surgical volume, this trend may be offset by the competing effects of in-stent restenosis [1] and population demographics. A recent analysis of these demographics has documented a contemporary increase in seniors, and especially octogenarians, and has predicted a dramatic increase in the need for cardiovascular services for coronary artery and valve disease in the next three decades as the large "baby-boomer" population subgroup ages [4]. In light of these opposing forces, we undertook a retrospective review of all our adult CAB and valve operations (VO) for the past 21 years to analyze current volume and procedure trends.
Because we are a large group of community-based surgeons working in multiple hospitals and different cities with different cardiology groups, serving urban, suburban, rural, blue-collar, and white-collar populations, we speculated that the results of this analysis might be more or less representative of current cardiac surgical trends in the developed world. We specifically looked beneath the trends in technology adoption at the level of the individual surgeon and the hospital where the surgeon practiced to infer, if possible, surgical paradigms.
|
Material and methods |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
Statistical analysis
Data were summarized by frequencies and percentages for categorical factors and means and standard deviation for continuous factors. Various comparisons were performed using univariate analysis (
2, Fischer's exact test) of categorical data. Fischer's exact test was performed if the results of the 2 test resulted in an expected value of less than 5. Univariate analyses of normally distributed continuous variables were carried out with the Student's t test. Variables were analyzed and significance was determined if the p value was equal to, or less than, 0.05. P values are the result of univariate analysis, unless otherwise stated. P values more than 0.05 are designated NS for not significant.
Multivariate analysis for predictors of hospital mortality was performed with 1980s and 1990s data combined for CAB and VO data. The four possible binary risk factors analyzed were: time period of surgery (1980s vs 1990s; 1990 to 94 vs 1995 to 99); age (< 65 years vs 
65 years); first operation versus reoperation; and isolated versus combined. To determine which variable or combination of variables were significantly associated with hospital mortality, multiple logistic regression was performed using a stepwise process. For those variables remaining in the model, odds ratios (OR) and 95% confidence intervals (CI) were obtained.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
Most of the procedures were done in six different hospitals, located in four different cities. By the second half of the 1990s, we were mainly in five hospitals. Cardiology groups tended to work in only one hospital. Three hospitals had only one cardiology group, while two hospitals in the same city shared two competing cardiology groups. We were the only cardiac surgical group in three hospitals in two cities, while competing with two different cardiac surgical groups in the other two hospitals in the other two cities. Among the seven active surgeons during the entire decade of the 1990s, the mean annual case volume was 179.7 cases/yr, declining 5.4% from 184.7 to 174.7 cases/yr (p < 0.001) between the first and second halves of the decade.
Coronary artery bypass surgery
Coronary artery bypass volumes increased in all but three years from 103 cases in 1979 to a peak of 1,895 cases in 1996, declining 15.3% to 1,605 cases in 1999.
Patient demographics
Trends in clinical variables for CAB over various time intervals are shown in Table 1,
with the risk profile and surgical complexity of CAB decreasing despite increasing age. Trends in age-group distribution for CAB, 1979 to 1999, are shown in Table 2,
with a decrease in patients less than 65 years of age and an increase in seniors, especially octogenarians.
|
|
The use of arterial grafts appeared in the 1990s. Left internal mammary artery (LIMA) usage increased from 1.5% in 1992 to 87.5% in 1999 (p = 0). Between 1993 and 1999, the use of the right internal mammary artery (RIMA) was consistently below 1.0% with no trend change. Between 1997 and 1998, radial artery (RA) usage increased from 4.8% to 8.3% (p < 0.001), decreasing to 3.6% in 1999 (p < 0.001).
Off-pump CAB (OPCAB) increased from 0.1% in 1994 to 28.9% in 1999 (p < 0.001). In 1999, OPCAB as a percentage of all CAB, varied widely among surgeons: three surgeons did more than 75%, two surgeons did approximately one-third, four surgeons did less than 10%, and five surgeons did none. Similarly, OPCAB varied among the five primary hospitals from 9.4% to 67.7%, with only one hospital more than 25%. Port-access technology for CAB was employed briefly and rarely. Limited access incisions (any incision other than a full sternotomy) peaked at 4.0% of all CAB in 1996, decreasing to 1.4% in 1999 (p < 0.001).
Hospital mortality
Predictors and trends of hospital mortality for CAB during various time intervals are shown in Tables 3 and 4.
Overall CAB mortality decreased between the decades of the 1980s and 1990s, but did not decrease further during the 1990s.
|
|
Age greater than 65 years was a predictor of hospital mortality for CAB during all time periods, with the mortality rates increasing stepwise among the younger patients, young seniors (age 65 to 79), and octogenarians. For each age group except octogenarians, there was a decrease in hospital mortality rates between the decades of the 1980s and 1990s but not during the 1990s.
Overall valve surgery
Procedure volume
The 6,524 VO accounted for just more than 20% of all cardiac procedures, with 6,030 (92.4%) single VO. Among single VOs, there were 3,917 aortic VO and 2,055 mitral VO, with a small number of tricuspid and pulmonic VO. Valve operation volumes increased in 14 out of 21 years, since 1979. There was a continuing 24% increase in VO volumes, from 470 to 583 since 1996, when CAB volumes peaked. During the 1990s, VO as a percentage of all cardiac procedures varied among surgeons from 0.7% to 41.6% (mean 21.5%), with a wide variation: one surgeon, less than 10%; six surgeons, 10% to 20%; five surgeons, 20% to 30%; one surgeon, 30% to 40%; one surgeon, more than 40%. For 1979 to 1999, VO as a percentage of all procedures varied among six hospitals from 15.1% to 26.7%.
Patient demographics
Trends in clinical variables for VO are shown in Table 5,
with an increase in mean age over the entire time period and increase in combined operations between the first and second halves of the 1990s. Trends in age-group distribution for VO over the past two decades are shown in Table 2. The less than 65-year age group decreased between the decades of the 1980s and 1990s, with no further decrease during the 1990s. The young seniors increased between the decades but decreased during the last half of the 1990s, while the octogenarians steadily increased throughout the entire time period.
|
Port-access was employed in the second half of the 1990s, peaking at 7.3% of VO in 1998, and falling to 2.2% in 1999 (p < 0.001). Similarly, limited access incisions peaked at 12.5% of VO in 1998, falling to 5.3% in 1999 (p < 0.001). During the decade of the 1990s, mechanical valve usage fell by 25.3%, from 83.7% in 1990 to 62.5% in 1999 (p < 0.001). Conversely, the application of various "tissue solutions" (including valve repair, aortic homografts, and bioprostheses) rose by 82.0%, from 12.4% in 1990 to 27.5% in 1999 (p < 0.001).
Hospital mortality
Predictors of hospital mortality for VO during various time intervals are shown in Tables 3 and 4. Overall VO mortality decreased approximately by half between the decades of the 1980s and 1990s. Surgical complexity was a predictor of hospital mortality during both decades combined. Either reoperation or the addition of CAB to VO nearly doubled hospital mortality. Age greater than 65 years was a predictor of hospital mortality during both decades combined, with a mortality for senior VO more than double that of younger patients. For the entire study period, the mortality rates increased stepwise among the younger patients, young seniors, and octogenarians. For each age group, there was a decrease in mortality for all age groups between the decades of the 1980s and 1990s, and no trend changes for any age group during the 1990s, except for a slight upward trend among octogenarians.
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
The recent decline in CAB volumes is likely due to a combination of an increase in the application of PCI to patients with coronary artery disease and a possible decrease in the prevalence of coronary artery disease [10]. The increasing prevalence of prior PCI in our CAB patients supports the likely role of PCI in the declining CAB volumes. Current guidelines for PCI have expanded to include interventions without on-site cardiac surgical backup, acute MI, three-vessel coronary artery disease, and prior CAB [3]. With a steady upward trend during the last decade, nearly one-third of all our CAB patients had had a prior PCI by 1999. Contrary to our data, others [6] have reported a decline in the prevalence of prior PCI during the 1990s to only 15% by the end of the decade. Perhaps, as previously speculated, this reflects less aggressive cardiologists in their series compared to ours.
Parallel developments in "minimally invasive" CAB technology, including off-pump CAB (OPCAB), have taken place mainly in the last half of the 1990s. OPCAB was quickly adopted and refined by several surgeons in our group [11]. These early adopters or visionaries [12] of new technologies profoundly influenced the conduct of CAB in our group, resulting in an increase of up to 28.9% OPCABs by 1999.
Port-access technology was adopted by three of the same surgeons who did OPCAB, but it was only applied to a few CAB patients before being abandoned. The robot was pursued in the laboratory but not applied clinically. Limited access incisions made a brief appearance but faded quickly when full sternotomy became the standard approach to OPCAB. The larger incision was used primarily for gaining access to the entire left ventricle in order to accomplish complete revascularization. Therefore, for CAB, minimally invasive came to be defined by the elimination of cardiopulmonary bypass rather than by the length or location of the incision. Concerns about the adverse effects of cardiopulmonary bypass on neurocognitive function and growing concerns about cannulating and clamping diseased aortas have likely contributed to our apparent long-term adoption of OPCAB. However, technology adoption has been very individual, regional, and frequently transient, sometimes following an adoption/abandonment, or evanescent, pattern as shown in Table 6.
|
The absolute values and nearly threefold increase in hospital mortality rates of our reoperative CAB, compared to first operations, are similar to other reports [9]. The time-trend to higher mortality for reoperations may be due to a recent bias of more difficult surgical cases in the wake of prior PCI, with cardiologists referring only more difficult cases for surgery. The trend of an increased expected mortality between the first and second halves of the 1990s may lend some support to this contention. The stepwise increase in mortality with increasing age in our patients is similar to other reports [8].
Overall valve surgery
Although the Society for Thoracic Surgeons (STS) recently developed a risk stratification for cardiac valve replacement from the National Cardiac Surgery Database with similar risk factors as for CAB [14, 15], the software was not available at the time of our data analysis. Accordingly, the more sophisticated analysis of risk profile, which was available for our CAB population, was not performed on our VO patients.
Although others have either documented increasing numbers of older patients [7] or increasing mean age [14, 16], in their VO patients in whom valve disease is more prevalent, an actual increase in the volume of VO has not apparently been widely reported. Consistent with our data, others [14, 17] have reported the stable frequency of gender distribution over time. Contrary to our data, others [14, 17] have found their patients to be sicker by multiple criteria over time. As for surgical complexity, the report by Thourani and colleagues [17] was similar to our data. They found an increase in the frequency of concomitant CAB but no change in frequency of prior CAB, suggesting an unchanged frequency of first operations and reoperations over time. By contrast, the STS Database found an increase in the percentage of reoperations during the decade of their analysis, 1986 to 1995 [14].
In the report by Thourani and colleagues [17], the frequency of prior PTCA among VO patients was less than 2% every year during the period 1988 to 1997 with no trend change. As suggested earlier, without supporting data, this was perhaps due to a less aggressive interventional cardiology group involved with their patients compared to ours.
Minimally invasive approaches to valve surgery were also employed by three surgeons in our group, including port-access technology and limited access incisions [18]. Concerns about the cost/benefit ratio of additional technological components and time, added complexity and risks with the peripheral cannulation strategy of the port-access technique probably contributed to its transient appearance in our practice. Limited access incisions followed the same pattern of early adoption and rapid abandonment as for port-access. Doubts about limited access permitting the achievement of a "perfect" operation, particularly with complex mitral repair and aortic root procedures, may have contributed to the transient appearance of these technologies in our group, as shown in Table 6. Moreover, concerns about the long-term importance of the length and location of the incision in comparison with the specific valve operation performed (eg, mitral valve repair vs replacement) probably also contributed to the decision to emphasize the procedure rather than the incision. The additional time and difficulty of doing operations through small incisions was possibly a barrier to adoption for some surgeons. The increasing numbers of older patients, who seem to be generally less concerned about the cosmetic appearance of a surgical incision than younger patients, may also have been a factor. Thus, the adoption of technology followed the same pattern for valve surgery as for CAB surgery, as an individual and often transient affair, as shown in Table 6.
Our large series of St. Jude mechanical valve implants (St. Jude Medical, St. Paul, MN, USA) [19], since the first implant in 1977 [20], represents an obvious early "mechanical" paradigm in our group for the surgical treatment of valve disease. Moreover, adoption of this new bileaflet technology by the original early adopter [16] in our group, also represents an obvious paradigm shift away from earlier mechanical valve designs. The bileaflet design has remained an established technology in our group over the entire study period, as shown in Table 6. A similar paradigm of "mechanical solutions" for mitral valve disease clearly existed in Carpentier's Hopital Broussais in Paris until 1976, when the downward trend lines for two different mechanical valves crossed the upward trend lines for the tissue solutions of mitral repair and the bioprosthesis [21].
In the recent STS Cardiac Valve Replacement Database report [14], there was no change in the overall distribution of mechanical valve and bioprosthetic implants between the two time periods of 1986 to 1990 and 1991 to 1995. However, the overall mechanical valve implant rate in this report was lower than the overall 
80% mechanical valve implant rate in our series during the same time interval of the first half of the 1990s. This likely reflects a more established mechanical paradigm for the surgical treatment of valve disease in our practice. The major reasons for the steady erosion of mechanical valve implant rates during the 1990s appear to be the increasing frequency of mitral valve repair [22] and the accelerated growth of octogenarians during the 1990s [4], in whom tissue solutions for valve disease are especially attractive.
Others [16] have also observed a decrease in operative mortality for VO with time. However, we did not observe a decline in hospital mortality during the last decade as reported by Birkmeyer and colleagues [23]; perhaps because our mortality rates were already lower to begin with. Although these authors did not identify a cause for the improvement in hospital mortality, we speculate that in our series it is likely due to improved myocardial protection as a consequence of the uniform adoption of retrograde cardioplegia technology by our group, as shown in Table 6.
As with our data, others [14, 15, 24] have reported an increased mortality with surgical complexity, including both reoperations and the addition of CAB to the VO. The absolute values and essential doubling of hospital mortality rates with complex VO over simple VO in these series are similar to our data. The increase in mortality with age has also been reported consistently by others [7, 15, 24] and has been shown to be a stepwise increase for patients in their 60s, 70s, and 80s by Craver and colleagues [7], and for patients age more than 55 years overall, with a stepwise increase in groups 55 to 64 years, 65 to 74 years, 75 to 84 years, and more than 85 years, by Hannan and colleagues [24].
Population demographics
In this study, the most obvious trend over the past two decades is the increasing mean age for CAB and VO. In 1979, seniors only accounted for less than 20% of all CAB patients and just more than 40% of all valve patients, with no octogenarians. Currently, over half of all patients are greater than or equal to 65 years and approximately 10% are 80+ years. Our data show growth of the overall senior population in the 1980s but essentially no growth in the 1990s. However, the octogenarians have continued to increase through both decades. Although this is largely a function of population demographics [4], this progressive increase in surgical volume among older and older patients may also represent a subtle upward paradigm shift in our collective definition of "old" (and its associated perceived upper age limit for cardiac surgery), moving up gradually but steadily over the past two decades.
The age group profiles for our patients during the past two decades are similar to those already documented in the general United States population. Between 1980 and 1990, seniors increased their share of the population from 11.3% to 12.5%, with essentially no change during the 1990s. Although the "young seniors," age 65 to 74, showed almost no growth in the 1990s, the "senior-seniors," age 75 years and older, grew 26%. The overall flat senior growth rate in the 1990s is due to the young seniors born during the post-Depression 1930s, when the birth rate was low, while the effects of immigration and economic prosperity during the 1910s and 1920s have produced the current boom in the senior-seniors group [4]. The increasing mean age of our patients and the increasing volumes of valve operations in our practice are consistent with these data.
The importance of population demographics for the future is based on four premises: (1) The population will grow more than 25% in the next three decades. (2) Life expectancy will increase by approximately one year for each ensuing decade. (3) The pace of population aging will accelerate dramatically in the next three decades due to the baby-boomers, with 65 years age group nearly tripling its growth rate, increasing its share of the population from 12.6% in 2000 to 20% by 2030. (4) The prevalence of chronic heart conditions is a function of increasing age [4]. Assuming the accuracy of these assumptions, the next decade will reflect the concurrent outgoing tide of the seniors born in the 1920s, who will be in their 80s, and the incoming tide of the early baby-boomers born in the late 1940s, who will present in their mid 50s.
Study limitations
The primary limitation of this study is a patient database that may not represent a valid cross-section of the population we serve. Underrepresented subgroups include racial minorities, the indigent, the uninsured, and the unemployed. Most of our patients are middle class, employed or retired, and of Western European or Scandinavian extraction. They typically have health insurance and a primary care physician whom they see more or less regularly.
Another limitation is the potential mobility of our patient population. Although most patients are permanent residents, some migrate temporarily to warmer climates during the colder months. Moreover, patients may change insurance coverage (and, therefore, provider coverage) during the annual open enrollment of competing health plans. And, changing referral patterns among cardiologists and primary care physicians and the presence of competing surgical groups add to the possible patient mobility in or out of our practice.
An additional limitation of this study is the absence of a formal survey to define technology adoption or abandonment patterns among the various referring cardiologists or the members of our group. Accordingly, we can only speculate on the reasons behind the observations.
Conclusions
Population demographics have been responsible for increased patient volumes in some age groups up to the present time, primarily as a result of the rise in the senior population, and especially the octogenarians. Commensurate with the actual increase in the numbers of older patients, our changing paradigm of the definition of old, apparently rising gradually over the past two decades, seems to have allowed a progressively higher upper age limit for cardiac surgery. Population demographics also predict a future wave of aging baby-boomers and their burden of an increasing prevalence of heart disease over the next three decades.
Clearly, the biggest impact of new technology on CAB volume has been the PCI. Coronary artery bypass volumes in the next three decades will be proportional to the incoming baby-boomer wave, with their increased burden of coronary artery disease, but inversely proportional to the long-term success rates of the intracoronary stent and risk factor reductions. The new drug-eluting, coated stents appear to improve longer term post PCI patency rates [25], making it difficult to predict when and where the downward trend of CAB volumes will end. Unfortunately, at least to date, the minimally invasive OPCAB does not appear to have offset the impact of the even "less invasive" PCI.
The increase of patients needing valve operations is currently a reality, primarily because of the rapid rise in the octogenarian population. However, the baby-boomers are also beginning to appear, with their bicuspid aortic valve disease and degenerative and ischemic mitral valve disease. Whether this trend of increasing VO will continue, will, as with CAB volumes, probably largely depend on the success of emerging percutaneous technologies for valve disease and risk factor reduction.
Technology adoption has not been uniform in our group, as summarized in Table 6. However, a major paradigm shift for CAB seems to be the gradual elimination of arrested heart, cardiopulmonary bypass platforms in favor of beating-heart platforms. A paradigm shift for VO also seems to be under way in favor of various tissue solutions rather than mechanical solutions for valve disease.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. A. Groh, O. A. Binns, H. G. Burton III, G. L. Champsaur, S. W. Ely, and A. M. Johnson Epicardial Ultrasonic Ablation of Atrial Fibrillation During Concomitant Cardiac Surgery Is a Valid Option in Patients With Ischemic Heart Disease Circulation, September 30, 2008; 118(14_suppl_1): S78 - S82. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Gerber, C. S. Rihal, T. M. Sundt III, J. M. Killian, S. A. Weston, T. M. Therneau, and V. L. Roger Coronary Revascularization in the Community: A Population-Based Study, 1990 to 2004 J. Am. Coll. Cardiol., September 25, 2007; 50(13): 1223 - 1229. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Bardakci, F. H. Cheema, V. K. Topkara, N. C. Dang, T. P. Martens, M. L. Mercando, C. S. Forster, A. A. Benson, I. George, M. J. Russo, et al. Discharge to Home Rates Are Significantly Lower for Octogenarians Undergoing Coronary Artery Bypass Graft Surgery Ann. Thorac. Surg., February 1, 2007; 83(2): 483 - 489. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-Y. Dupuis Clinical Predictions and Decisions to Perform Cardiac Surgery on High-Risk Patients Seminars in Cardiothoracic and Vascular Anesthesia, June 1, 2005; 9(2): 179 - 186. [Abstract] [PDF]
|
|
|
|
|
|
C. Weissman Pulmonary Complications After Cardiac Surgery Seminars in Cardiothoracic and Vascular Anesthesia, September 1, 2004; 8(3): 185 - 211. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
