California Cardiac Surgery and Intervention Project
A comprehensive report on California hospitals performing heart surgery and percutaneous coronary intervention
not here
not here

Procedure Summaries

  1. Coronary Artery Bypass Graft (CABG) and Percutaneous Coronary Intervention (PCI)
  2. Aortic Valve Surgery
  3. Mitral Valve Surgery
  4. Statewide Valve Data: Comparison with New York
  5. Aneurysms, Heart Transplants, and Other Procedures


Coronary Artery Bypass Graft (CABG) and Percutaneous Coronary Intervention (PCI) Procedures in California 1997-2009

Revascularization Procedures Decrease, Outcomes Improve

During first decade of the 21st century, there has been a shift away from surgery towards interventional procedures. In 2003, two major changes occurred: the introduction of drug-eluting stents, and mandated reporting for CABG surgery. As a result, outcomes have improved, and the number of procedures performed has decreased:

Statewide Volume

Coronary artery revascularization procedures (all CABG and PCI) peaked at nearly 80,000 in 2003, then decreased 25% to slightly more than 60,000 in 2009. CABG represented nearly 40% of all revascularization procedures in 1997 but only 20% in 2009. PCI procedures are reported in two groups, those performed for acute MI diagnoses (ICD-9-CM codes 410.xx)—PCI with acute coronary syndromes (PCI-ACS) and those performed for all other diagnoses (PCI-noACS).


The statewide rate of coronary revascularization procedures in California is much lower than the rate reported in other states, and similar to the rate reported in the United Kingdom. The figure and table below show the comparative rates for the most recently reported year, 2007.


State/CountryPopulationPCIPCICABGCABG
2007MillionsTotalNo./100KTotalNo./100K
California36.553,28114614,38639
New York19.351,69526811,44559
Massachusetts6.414,0632203,39653
United Kingdom61.277,37312621,24835

These differences are unexplained. There is a slight difference in age range between California as compared to New York and Massachusetts; the percentage of persons older than 65 years is 11.2% vs. 13.5%. Other demographic factors, such as race, gender and socioeconomic status have not been studied

In-Hospital Mortality

Mortality during the index hospitalization (in-hospital mortality) for PCI and CABG is relatively low, and many hospitals had no deaths, negating risk adjustment. We therefore have used in-hospital mortality (at index hospitalization or readmission) at 90 days (IHM-90) to better discriminate among providers. IHM-90 for PCI-ACS was 4.6% in 2009, and IHM-90 for PCI-noACS was 1.2%. Minimal change was noted over the decade. IHM-90 for CABG, however, improved markedly, decreasing from 4.3% prior to 2000 to 2.5% in 2009, despite decreasing volume and increasing predicted risk in the patient pool.


One Year Reintervention and Adverse Event Rates

Reintervention for another procedure (any PCI or cardiac surgery) within one year decreased markedly for PCI with or without ACS, but increased slightly for CABG. To account for elective return for planned PCI procedure on a non-target vessel, reintervention rates were also calculated for "unplanned" (non-elective) readmission for another procedure:


The one year adverse event rate including Mortality, Acute MI, Reintervention or Stroke (MARS) also decreased for PCI but remained stable for CABG. One year MARS (including only unplanned reinterventions) in 2008 was 17.6% for PCI+ACS, 13.3% for PCI noACS, and 9.8% for CABG.


Individual Hospital Data

Approximately 120 California hospitals offer open heart surgery and coronary stenting procedures, and an additional 17 sites offer PCI facilities only. The majority of hospitals perform a relatively low annual volume of revascularization procedures. Those with higher volumes (>500 PCI w/o ACS, >250 PCI w/ACS, >200 CABG) tend to have more consistent outcomes (most had O/E Ratio less than 1.0). There is wide variation in outcomes among lower volume hospitals.

The scatter plots (below) show the variation in volume and outcomes for PCI and CABG. The trendline for PCI with ACS suggests no relationship of volume to outcome, but the scatter plots for PCI w/o ACS and CABG show a trend towards better outcomes with higher volume. However, many of the low volume sites have no mortality, so there is little statistical relationship between volume and in-hospital mortality in any of the procedure groups.


In order to gain some insight into the aggressiveness of individual hospitals in PCI performance, the percent of patients undergoing multivessel procedures was calculated for 2008-2009, showing a variation from less than 10% to more than 40%:


The ratio of PCI to CABG procedures was also calculated. This showed a ratio of 1:1 to more than 14:1 for all California hospitals performing both procedures:


The percentage of PCI procedures performed for ACS diagnoses was also variable, with some sites performing less than 20%, and others more than 60%, PCI w/ACS:


It is unclear how these variations affect outcomes, but further studies are indicated.

Individual hospital data, showing comparative outcomes for all PCI and CABG procedures and trends over the last 13 years, is available to members. A list of hospitals and sample reports are provided on the login page


Aortic Valve Surgery

Isolated aortic valve replacement procedures have increased 26% in California over the last decade, most likely due to the age-related increasing incidence of degenerative aortic stenosis. In-hospital mortality rate improved slightly, from 4.3% to 3.5%. However, one year adverse outcome rate (MARS) has remained at about 10%.

Aortic valve procedures (including those associated with CABG) are more common than mitral valve procedures; overall, 4737 aortic valve and 2633 mitral valve procedures were performed in 2009. As a result, individual hospital volumes were greater for aortic valve surgery. The figure below shows individual hospital mortality rates in the two years covering 2008 and 2009 for all aortic valve procedures, expressed as observed/expected (O/E) ratio for all California hospitals performing 25 or more procedures, in relation to the number of procedures performed:


A larger number of hospitals perform more than 20 aortic procedures per year, but wide variation in outcome is evident among the lower volume sites as noted in mitral valve procedures.

The increasing incidence of frail, older patients with symptomatic aortic stenosis has stimulated the development of percutaneous techniques for correction of aortic stenosis. Several aortic valve stents have been designed and two models are now in clinical use. The numbers remain small but will be increasing.


Mitral Valve Surgery

Aging of the population has increased the number of patients with degenerative heart valve diseases such as mitral insufficiency and aortic stenosis. Surgeons have become increasingly proficient in mitral valve repair procedures during the last decade. As a result, the number of mitral valve repair procedures performed in California more than doubled over the last decade, and remarkably, the in-hospital mortality rate has been cut in half. In 2009, 1,043 mitral valve repair procedures were performed with only 15 in-hospital deaths (1.4%). In addition, one year followup found only 2.4% of patients requiring reintervention, indicating the durability of surgical techniques.

Valve Repair Procedures: The overall impact of increased use of mitral valve repair was studied by comparing outcomes for all patients undergoing mitral valve surgery during 1997-1999 to those undergoing procedures during 2006-2008. Mitral valve repair procedures increased from 32% (2736 of 8619) to 52% (4361 of 8209) of all isolated mitral and mitral combined with CABG cases. In-hospital mortality decreased from 9.1% to 6.2% and mortality at one year decreased from 12.4% to 9.6%. Stroke incidence at 90 and 365 days decreased from 1.9% and 2.9% to 1.6 and 2.4% for all MV surgery. This was entirely due to decreased incidence of stroke in MV repair, since stroke incidence actually increased in MV Replacement. Among all MV patients followed for a minimum of one year, there were 247 less deaths and 40 fewer strokes during 2006-2008 (n=8589) as compared to 1997-1999 (n=8619).

Individual Hospital Data: Wide variation in valve procedure volume and outcomes was noted among hospitals. Many hospitals perform only a few valve procedures. The figure below shows individual hospital mortality rates in the two years covering 2008 and 2009 for all mitral valve procedures, expressed as observed/expected (O/E) ratio for all California hospitals performing 25 or more procedures, in relation to the number of procedures performed:


Most hospitals perform fewer than 20 mitral valve replacements per year, and the risk-adjusted mortality rates vary widely. Consideration of the total volume of valve procedures performed by a hospital, together with early and late (one year) outcomes would provide a more valid comparison. For example, since the in-hospital mortality rate for mitral valve repair is low, one might conclude that it would make no difference where the procedure was performed. However, the risk of adverse post hospital outcomes was approximately 10% for all mitral repair procedures (including those associated with CABG) in 2008. Using the MARS one year outcome would improve the differentiation among providers.

Increasing use of mitral valve repair procedures has resulted in a net improvement in mortality and post hospital event rates for patients requiring surgery for mitral valve disease in California, which compares favorably with national data. However, individual hospital case volumes remain low. As market pressure shifts toward minimally invasive valve surgery, it will become increasingly important to track outcomes and maintain evidence based guidelines for these new surgical approaches, and the minimum volume to maintain success will become more important.


Statewide Valve Data: Comparison with New York

During the 13-year reporting period, heart valve procedure outcomes improved in all categories. The tables show IHM and one-year MARS outcomes for the six major valve procedures, comparing 1997 to 2009:

1997199720092009
TotalMortalityTotalMortality
ProcedureN%N%
Aortic Valve 2,1504.3 2,9643.5
Mitral Valve 1,2638.1 9245.4
Mitral Valve Repair 4235.9 1,0431.4
AV + CABG 1,988 8.6 1,8035.2
MV + CABG 67413.6 3038.9
MV Repair + CABG 43211.8 4886.1
Total 6,9307.77,5254.2


1997199720082008
TotalMARS 1yrTotalMARS 1yr
ProcedureN%N%
Aortic Valve 2,1509.9 2,8829.3
Mitral Valve 1,26314.6 93612.4
Mitral Valve Repair 42315.1 9278.4
AV + CABG 1,988 17.3 1,88113.3
MV + CABG 67424.2 32622.1
MV Repair + CABG 43222.5 52014.4
Total 6,93015.47,47211.4

A 45% improvement is noted in IHM over the time period. The incidence of adverse outcomes at one year also improved by 26%, however only isolated aortic valve replacement and mitral valve repair had an MARS of less than 10% at one year.

California vs New York

In New York, heart valve surgery outcomes are included in the public report on cardiac surgery. Standard STS reporting categories are included. The most recent report includes aggregate data for 2006-2008. The table shows data comparing the same time periods in California to New York:

California 2006-2008New York 2006-2008
TotalMortalityTotalMortality
ProcedureN%N%
Aortic Valve 8,2943.6 6,2143.1
Mitral Valve 2,9186.6 2,0055.3
Mitral Valve Repair 2,6962.4 2,3821.6
AV + CABG 5,770 6.2 4,8575.4
MV + CABG 1,09111.8 9539.8
MV Repair + CABG 1,6068.0 1,6846.3
Multiple Valve* 2,01610.0 3,3509.7
Total 24,3915.721,4455.2
*Tricuspid valve procedures are included in single valve categories in CA data

Procedure volume is higher in New York (34 procedures/year per 100,000 population) compared to California (22/100,000). IHM mortality is slightly lower in all procedure groups in New York. However, 2009 California data, noted in the first table above, is remarkably similar to the aggregate 2006-2008 New York data.


Aneurysms, Heart Transplants, and Other Procedures

Cardiac surgery procedures that were not coded within the specific CABG or Valve categories are reported with observed outcomes only (not risk adjusted). These procedure groups are listed in the following table, comparing annual volume and in-hospital mortality (IHM) for 1997 and 2009:

1997199720092009
TotalIHMTotalIHM
ProcedureN%N%
Heart Resection 35211.1 6636.2
Heart Repair 4825.0 15720.4
Aortic Aneurysm 52118.2 9857.5
Heart Transplant 202 6.9 1807.8
Multiple Valve 58513.9 6739.5
Other Valve 12716.5 2418.3
Other CABG 60611.7 2454.9
Total 2,44113.63,1448.2

These procedure categories include patients with various operations, mostly complicated cases requiring multiple operative interventions. The number of patients in each group increased over the interval, with the exception of transplants, which changed little over the 13-year period, and “other CABG” which decreased to a similar degree as isolated CABG procedures. IHM also improved in all categories except transplants.

Aortic Aneurysm procedures nearly doubled, while IHM mortality decreased from 18.2% to 7.5%. This is thought to be related to increased number of dissecting aortic aneurysms being discovered, probably related to improved awareness and diagnosis, coupled with advances in surgical techniques.

The complicated nature of these procedure groups is reflected in the one year outcomes. The table below shows the incidence of any adverse outcome (MARS, or Mortality, Acute MI, Reintervention and Stroke) at one year for each group, comparing 1997 to 2008:

1997199720082008
TotalMARS 1yrTotalMARS 1yr
ProcedureN%N%
Heart Resection 35237.0 68120.0
Heart Repair 4837.5 13828.3
Aortic Aneurysm 52130.9 101218.6
Heart Transplant 202 12.4 2129.9
Multiple Valve 58523.9 70318.6
Other Valve 12722.8 23316.3
Other CABG 60620.8 29819.8
Total 2,44125.83,27718.7

While the number of adverse outcomes has decreased, there continues to be a nearly 20% incidence of adverse events in most of these complicated patient categories.


© 2011 California Society of Thoracic Surgeons