MILITARY MEDICINE, 175, 7:529, 2010 Decreasing Outpatient Cardiac Catheterization Rates Associated With Cardiology Clinic Volume but Not With Increasing Cardiac Computed Tomography Utilization MAJ Eddie Hulten, MC USA * ; MAJ Salvatore Carbonaro, MC USA * ; LTC Barnett Gibbs, MC USA * ; CPT Michael Cheezum, MC USA ; Irwin Feuerstein, MD ; LTC Daniel Simpson, MC USA (Ret.) * ; COL Allen J. Taylor, MC USA (Ret.) * ; MAJ Todd C. Villines, MC USA * ABSTRACT Cardiac computed tomographic angiography (CTA) is an accurate noninvasive test for diagnosing coronary artery disease (CAD). To investigate whether increasing use of CTA is correlated with left heart catheterization (LHC) rates, we performed a retrospective review of existing outpatient and inpatient catheterization lab and CTA electronic medical records from July 1, 2004 to June 30, 2008. Comparing the previous 2 years (July 2004 June 2006) to the 2 years after addition of CTA (July 2006 June 2008), monthly LHC rates decreased 20 ± 6% ( p = 0.08) and percutaneous coronary intervention (PCI) rates decreased 47 ± 6% ( p < 0.001). Cardiology clinic volume declined 34%. CTA rates increased 64 ± 7% ( p < 0.001). Radionuclide myocardial perfusion scan (MPS) usage remained stable. Despite increased utilization over the past 2 years, CTA was not correlated with significantly reduced LHC rates. The decline of outpatient LHC rates at our institution over 4 years is mainly influenced by decreasing outpatient Cardiology clinic volume. INTRODUCTION Cardiac catheterization has remained for over 30 years the gold standard diagnostic modality for anatomic assessment of coronary artery disease. However, the last decade has witnessed the emergence of marked technological advances that allow for accurate noninvasive coronary anatomic description via cardiac computed tomography angiography (CTA). A recent comparison of cardiac CTA with diagnostic left heart catheterization (LHC) demonstrated noninferiority, with sensitivity 95% and specificity 83%. The positive predictive value of cardiac CTA was 64% and negative predictive value 99%. 1,2 Thus, cardiac CTA is an excellent modality to rule out coronary artery disease in select patients (use is limited in the morbidly obese and those with irregular heart rhythms). 3 A multidisciplinary update of the 2006 American College of Cardiology appropriateness criteria for CTA has listed its use to rule out coronary artery disease (CAD) among low and intermediate pretest probability chest pain patients as appropriate. 4 Because it offers detailed coronary visualization with good sensitivity to rule out significant disease without the * Cardiology Service, Department of Internal Medicine, Walter Reed Army Medical Center, 6900 Georgia Ave. NW, Washington, DC 20307. Department of Internal Medicine, Walter Reed Army Medical Center, 6900 Georgia Ave. NW, Washington, DC 20307. Department of Radiology, Walter Reed Army Medical Center, 6900 Georgia Ave. NW, Washington, DC 20307. Division of Cardiology, Washington Hospital Center, 110 Irving St. NW, Washington DC, 20010. The data herein were previously presented at the 2008 meeting of the U.S. Army Chapter of the American College of Physicians (ACP) and at the 2009 meeting of the Society of Coronary Angiography and Intervention (SCAI). The data have not been submitted elsewhere for publication. The opinions and assertions contained herein are the authors alone and do not represent the views of the Walter Reed National Military Medical Center, the U.S. Army, or the Department of Defense. risk of vascular complications associated with conventional angiography, increasing use of cardiac CTA may replace LHC in select populations. Walter Reed Army Medical Center (WRAMC), a military tertiary referral center, may offer an ideal population to test this hypothesis, given the high number of referrals for low-risk rule-out patients for whom any degree of risk of coronary disease may be unacceptable due to their high-risk occupation (e.g., patient with atypical chest pain symptoms who may be military pilots or imminently deploying to remote combat zones). Previous authors have reported significant associations of changing cardiac catheterization rates with changing rates of diagnostic CTA. In a recent presentation at the Society of Cardiac CT, Karlsberg and colleagues reported a 45% decline in left heart catheterization rates after the addition of cardiac CTA. Furthermore, their percentage of catheterization cases leading to percutaneous coronary intervention (PCI) was increased, suggesting that CTA was replacing low-risk catheterizations that were unlikely to need PCI. 5 Therefore, we hypothesized that addition of a 64-slice multidetector cardiac CTA at a military tertiary referral hospital may change our rates of outpatient (nonemergent) diagnostic left heart catheterizations. We would anticipate a decline of LHC procedures with increasing CTA use and a higher ratio of PCI to LHC (less unnecessary diagnostic LHC). METHODS We conducted a retrospective cohort study of an available electronic database of diagnostic cardiac catheterization and PCI patient records collected from July 1, 2004 through June 30, 2008 and cardiac CTA records available from July 1, 2006 through June 30, 2008. Inpatients and emergent cases (e.g., ST elevation myocardial infarction) were excluded so that MILITARY MEDICINE, Vol. 175, July 2010 529
elective LHC rates could be compared to elective outpatient CTA. As a control measure for potentially confounding variables, we controlled for workload volume by measuring outpatient clinic visit rates during the study time period. Unadjusted procedural rates and adjusted rates (by 1,000 outpatient encounters) are reported. Thallium/sestamibi dual isotope myocardial perfusion stress (MPS) tests may compete as a noninvasive risk stratification test with cardiac CTA; therefore rates of MPS were collected from July 1, 2004 through June 30, 2008. Demographic information (age and gender) was extracted electronically from the Composite Health Clinical Systems (CHCS) records. Statistics were conducted using SPSS v13.0 (SPSS Inc., Chicago, Illinois) and Stata v8.0.2 (StataCorp, College Station, Texas). Continuous variables were compared using Student s t -test. Rates of LHC, PCI, and CTA were correlated with Pearson s r. Since the outpatient referral volume was not constant over the 4-year study period, unadjusted rates and rates after adjusting for clinic visits were compared. Variables significantly correlated with rates of outpatient LHC were evaluated further using multivariate regression analysis. RESULTS Table I describes the changing demographics among Cardiology clinic, LHC, and CTA patients during the 4-year study period. The mean age for the entire LHC population from July 2004 through June 2008 was 64.2 years. There was a nonsignificant trend for LHC patients to be older after the addition of CTA (61.1 years versus 61.8 years, p = 0.057). The mean age for CTA patients was 49.8 years, significantly lower than the age for LHC patients, p < 0.0001. CTA patients were 59% male compared to 73% male for LHC, p < 0.001. Outpatient LHC and PCI rates declined over the 4-year period (Fig. 1A and Table II ). CTA rates increased. MPS rates remained stable. Outpatient clinic volume declined significantly (see Figure 2 ). After controlling procedural rates by clinic volume, there was no significant change in outpatient LHC or PCI, but there was a relative increase in MPS and CTA ( Fig. 1B ). Rates of outpatient LHC were significantly TABLE I. negatively correlated with MPS rates and positively correlated with inpatient LHC and clinic visit rates. LHC rates were not correlated with cardiac CTA ( Table III ). The ratio of outpatient PCI to LHC changed opposite the hypothesized direction (anticipated a ratio of fewer diagnostic LHC) during the study period, from 4.1 LHC per PCI before CTA versus 5.9 LHC per PCI after CTA, p = 0.04. Linear regression of unadjusted and adjusted LHC rates is depicted in Figure 3. The LHC rates (slope of the regression line) did not significantly differ before addition of CTA (May 2006) when compared with the rate after the addition of CTA. Table III portrays the significant predictors of outpatient LHC rates by univariate analysis. After multivariable regression analysis, only inpatient LHC rates ( b = 1.03, 95% CI 0.70 1.36) and clinic volume ( b = 0.008, 95% CI 0.006 0.10), p < 0.001) remained significant. DISCUSSION We expected that addition of cardiac CTA would have been strongly correlated with a reduction in LHC rates during the observed study period. However, although CTA utilization markedly increased (64% absolute increase) and LHC rates significantly declined (20% absolute decrease), the two rates were not correlated. The LHC rates began to decline even before addition of CTA and the rate of change in LHC (slope) did not appear to be influenced by CTA. In our system, CTA patients were younger and more likely to be women than LHC patients, supporting the concept that this imaging modality is used as a low-risk rule-out test in the military health care system, owing to its high negative predictive value (NPV). However, use of CTA is becoming more widespread and not all centers have the ability to perform diagnostic LHC. With its high NPV reported in a recent multicenter study, cardiac CTA will likely replace many diagnostic LHC. Nevertheless, we did not see a shift in the ratio of PCI to LHC that would suggest performance of less unnecessary diagnostic angiograms. We found the exact opposite, an increased ratio of LHC to PCI, although this is likely confounded by the publication of the clinical outcome utilization Demographics Entire 4 Years, July 2004 June 2008 July 2004 June 2006 (Pre-CTA) July 2006 June 2008 (Post-CTA) p value Clinic Patients ( n ) 151,145 84,681 66,464 Age 64.2 ± 17.6 65.0 ± 17.9 63.0 ± 17.3 <0.001 % Male 62 62 62 0.92 Outpatient LHC ( n ) 1,376 778 597 Age 61.1 ± 12.5 60.5 ± 12.8 61.8 ± 11.9 0.057 % Male 73 72 75 <0.001 CTA ( n ) 530 N/A 530 N/A Age 49.8 ± 13.0 49.8 ± 13.0 % Male 59 59 LHC, left heart catheterization; CTA, cardiac CT angiography. 530 MILITARY MEDICINE, Vol. 175, July 2010
FIGURE 1. (A, upper). Trends in unadjusted monthly procedural rates at Walter Reed Cardiology, Jul 2004 Jun 2008. (B, lower). Monthly procedural rates at Walter Reed Cardiology adjusted per 1,000 clinic patients. LHC, left heart catheterization; PCI, percutaneous coronary intervention; CTA, cardiac CT angiography. TABLE II. Procedure Before and After Addition of a 64-Slice MDCT, Changes in Rates of LHC, PCI, Stress Thallium, and CTA Pre- and Post-CTA, Unadjusted and Adjusted per 1,000 Clinic Patients Cases per Month, July 2004 June 2006 (Pre-CTA) Cases per Month, July 2006 June 2008 (Post-CTA) D% p value Outpatient LHC* 22.5 18.0 20 0.008 Adjusted LHC 6.9 7.7 11 0.216 Outpatient PCI* 6.8 3.6 47 <0.001 Adjusted PCI* 5.4 4.2 23 0.005 MPS 115.6 115.0 0 NS Adjusted MPS* 37.3 50.3 35 <0.001 CTA 2006 2008 N/A 21.2 64 N/A Adjusted CTA N/A 9.5 110 N/A LHC, left heart catheterization; PCI, percutaneous coronary intervention; CTA, cardiac CT angiography. * p < 0.05. revascularization and aggressive drug evaluation (COURAGE) trial during the study period, changes in physician staffing, and changes in physician acceptance of CTA as a diagnostic alternative.6 In our system, however, MPSs are more frequently requested for patients apparently at higher pretest probability of CAD, for whom LHC might otherwise have been performed. Of concern for accredited cardiology, cardiothoracic surgery, and internal medicine training programs, changes in referral patterns during the study period associated with a MILITARY MEDICINE, Vol. 175, July 2010 531
FIGURE 2. Thirty-four percent decline in Walter Reed Cardiology clinic visits ( y -axis) over time from July 2004 to June 2008. TABLE III. Univariate Predictors of Rate of LHC Predictors of Outpatient LHC Rates Pearson s r p value Inpatient LHC* 0.77 <0.01 Inpatient PCI 0.19 0.19 CTA 0.121 0.56 Cardiology Clinic Workload* 0.68 <0.01 Stress Thallium* 0.437 <0.01 Multivariate predictors of outpatient LHC rates were inpatient LHC ( b = 1.03, 95% CI 0.70 1.36) and clinic volume ( b = 0.008, 95% CI 0.006 0.10), p < 0.001. * p < 0.05. FIGURE 3. LHC rates pre- and postaddition of cardiac CTA, unadjusted and adjusted by clinic volume. The rate of decline in LHC (slope) was not affected by CTA but was associated with clinic volume. change in health maintenance policies (decreased access to primary care for patients over 65 years of age) directly impacted our Cardiology clinic referral volume, which was strongly negatively correlated with the change in LHC rates. Thus, declining clinic visits resulted in lower procedural volume. This drop in LHC and PCI volume is of concern given that WRAMC serves as an accredited cardiovascular disease fellowship training site for military cardiologists as well as a PCI referral center. The American College of Cardiology recommends such centers maintain 200 PCI/36 ST elevation myocardial infarction (STEMI) procedures annually (we did not include STEMI in this study of outpatient elective PCI). Our study is limited by the change in clinic volume during the study time period. Although we adjusted for clinic volume, a large change in Cardiology clinic volume likely resulted from changes in access to primary care for the over 65 years of age population in our health care system due to military deployment of primary care providers in support of a global war mission. The mean age for Cardiology clinic outpatients has decreased a significant 2.0 years ( Table I ), likely reflective of decreased access to care for older patients due to shifting demographics in support of WRAMC s wartime mission. Thus, the referral base for LHC and PCI was apparently diminished by this reduction in primary care access. Our study was also limited by changes in practice that likely occurred during the study period, which cannot be controlled for, especially as a result of the publication of the COURAGE trial in March 2007. 6 This study, which demonstrated no difference in PCI versus medical therapy for asymptomatic patients with CAD, likely may have contributed to decreased LHC and PCI volume. Finally, we could not control for changes in Cardiology staffing, which occurred during the study time period and are associated with the frequent staff turn-over at any military tertiary referral center due to military deployment, reassignments, and retirements from active duty military medical service. CONCLUSIONS In conclusion, LHC, PCI, and outpatient Cardiology clinic volume declined significantly at our hospital from 2004 to 2008. The mean age for Cardiology clinic outpatients has decreased a significant 2 years, likely reflective of decreased access to care for older patients due to shifting demographics in support of WRAMC s wartime mission. MPS rates remained stable. CTA use increased 64%, but this increase was not associated with changing LHC rates. LHC decline was associated strongly with declining clinic volume. In spite of decreasing LHC, the ratio of PCI to LHC decreased, likely due to changes in practice associated with publication of the COURAGE trial. Changes in primary care referral patterns and Cardiology clinic volume dramatically affect procedural volume. 532 MILITARY MEDICINE, Vol. 175, July 2010
ACKNOWLEDGMENTS The authors thank Ms. Linda Ridgle, RN BSN, Walter Reed Cardiac Catheterization lab head nurse, for her assistance with catheterization records and Ms. Sharrie Booth, Composite Health Care System (CHCS) data administrator, for her assistance with electronic medical record data extraction. REFERENCES 1. Budoff MJ, Dowe D, Jollis JG, et al : Diagnostic performance of 64- multidetector row coronary computed tomographic angiography for evaluation of coronary artery stenosis in individuals without known coronary artery disease: results from the prospective multicenter ACCURACY (Assessment by Coronary Computed Tomographic Angiography of Individuals Undergoing Invasive Coronary Angiography) trial. J Am Coll Cardiol 2008 ; 52 (21) : 1724 32. 2. Raff GL : Interpreting the evidence: how accurate is coronary computed tomography angiography? J Cardiovasc Comput Tomogr 2007 ; 1 (2) : 73 7. 3. Hendel RC, Patel MR, Kramer CM, et al : ACCF/ACR/SCCT/SCMR/ ASNC/NASCI/SCAI/SIR 2006 appropriateness criteria for cardiac computed tomography and cardiac magnetic resonance imaging: a report of the American College of Cardiology Foundation Quality Strategic Directions Committee Appropriateness Criteria Working Group, American College of Radiology, Society of Cardiovascular Computed Tomography, Society for Cardiovascular Magnetic Resonance, American Society of Nuclear Cardiology, North American Society for Cardiac Imaging, Society for Cardiovascular Angiography and Interventions, and Society of Interventional Radiology. J Am Coll Cardiol 2006 ; 48 (7) : 1475 97. 4. Carbonaro S, Villines TC, Hausleiter J, Devine PJ, Gerber TC, Taylor AJ : International, multidisciplinary update of the 2006 appropriateness criteria for cardiac computed tomography. J Cardiovasc Comput Tomogr 2009 ; 3 (4) : 224 32. 5. Karlsberg RP : CCTA integrated in the office setting dramatically reduces the need for invasive angiography with enhanced identification of critical disease not identified by conventional testing and clinical presentation. J Cardiovasc Comput Tomogr 2008 ; 2 (4S) : S11 2. 6. Boden WE, O Rourke RA, Teo KK, et al : Optimal medical therapy with or without PCI for stable coronary disease. N Engl J Med 2007 ; 356 (15) : 1503 16. MILITARY MEDICINE, Vol. 175, July 2010 533