Background
The aim of this study was to evaluate the clinical application of the American College of Cardiology Foundation and American Society of Echocardiography appropriateness criteria for stress echocardiography (SE) in a single-center university hospital.
Methods
Indications were determined for consecutive studies by two reviewers and categorized as appropriate, uncertain, or inappropriate.
Results
Of 477 studies for which primary indications could be determined, 188 specifically related to university transplantation programs were excluded. Of the remaining 289 studies, 88% were addressed in the appropriateness criteria for SE. Of these, 71% were appropriate, 9% were uncertain, and 20% were inappropriate. Inappropriate studies were more likely to be ordered on younger patients and women and were less likely to be ordered by cardiologists. Abnormal results on SE were more frequent among appropriate than inappropriate studies.
Conclusions
The appropriateness criteria for SE encompass and effectively characterize the majority of studies ordered in a single-center university hospital and appear to reasonably stratify the likelihood of abnormal results on SE. However, revisions will be required to fully capture and stratify appropriate clinical practice of SE.
The dramatic increase in the utilization of cardiovascular diagnostic imaging over the past decade is undeniable. Diagnostic imaging services reimbursed under Medicare’s physician fee schedule grew more rapidly than any other type of physician service from 1999 to 2003. This increased utilization has resulted in increased scrutiny of the appropriate use of cardiac imaging services. In response, the American College of Cardiology Foundation (ACCF) in conjunction with imaging subspecialty societies has published appropriateness criteria (AC) for selected patient indications for a variety of imaging modalities. The goal of these documents is to guide physicians and reimbursement agencies in determining a rational approach to the use of diagnostic imaging in the delivery of high-quality care.
Recently, the ACCF and American Society of Echocardiography (ASE) AC for stress echocardiography (SE) were published. As with other AC documents, these criteria use a standardized methodology, combining available evidence with expert opinion to identify common indications for SE and to determine their level of appropriateness. Similar AC for nuclear stress testing were first published in 2005 and then revised in 2009, after a number of investigations assessing their implementation in routine clinical practice identified gaps and improvements that were necessary. The ACCF and ASE are currently working to develop a revision of the AC for echocardiographic procedures. Although a number of studies have assessed the clinical application of the AC in transthoracic and transesophageal echocardiography, limited experience with the application of the AC for SE has been published.
Thus, the purposes of this study were to investigate the feasibility of clinical application of the published AC for SE and to describe their application to current clinical practice at a single-center university hospital.
Methods
Consecutive patients referred for exercise or pharmacologic SE at the University of Chicago Medical Center between May 2008 and December 2008 were eligible for inclusion. The protocol was approved by the institutional review board, and all patients provided informed consent. For each study, patient demographic information, referring physician specialty, outpatient or inpatient status, the primary indication for the study, and the results of the SE were recorded. For this analysis, SE performed for hemodynamic evaluation and for viability assessment were excluded.
Indication Determination
For each study, written requisitions and hospital and practice records were reviewed. A primary indication for each study was determined independently by two physician investigators blinded to the results of the SE. Investigators were asked to select one of two options; (1) one of the indications listed in the AC for SE or (2) not addressed (NA; the primary indication of the study was NA in the AC for SE). A primary consensus indication for each study was then determined for all further analysis. For studies in which the two investigators were in agreement, the indication and category selected served as the final primary consensus indication. For those studies in which there was disagreement between the two investigators, a third investigator independently reviewed the data and chose one of the two initial selections as a final primary consensus indication.
SE
All stress echocardiographic studies were performed according to the preferred recommendations of the ASE using a full platform echocardiographic instrument (iE33; Philips Medical Systems, Andover, MA). All studies were interpreted by one of seven expert echocardiographers, and the findings used for analysis in this study represent those reported on the final clinical echocardiographic report. Exercise testing consisted of symptom-limited exercise treadmill testing using a Bruce protocol, with baseline and peak stress imaging. All standard exercise testing parameters were recorded. Pharmacologic testing was performed with dobutamine using a standard rest, low-dose (5–10 μg/kg/min), high-dose (20–50 μg/kg/min), and recovery imaging protocol, with atropine supplementation (up to 1 mg) if necessary to achieve 85% of the maximum predicted heart rate response. Echocardiographic contrast was used if two or more contiguous endocardial segments were not visible on baseline images.
Abnormal results on SE were defined as any echocardiographic finding suggesting inducible myocardial ischemia, including any new stress-induced wall motion abnormality, stress-induced worsening of an existing wall motion abnormality, and/or stress-induced left ventricular dysfunction or chamber enlargement.
Statistical Analysis
Comparisons between study and patient characteristics, findings on SE, and levels of appropriateness were performed with χ 2 tests for categorical data and Student’s t tests for continuous data using a two-tailed p value < .05 for statistical significance. Interobserver variability in the determination of indication, category, and level of appropriateness is expressed as percentage agreement between the two independent reviewers. For indication and category determination, interobserver comparisons represent the frequency with which the reviewers agreed on the indication number or category (i.e., indication number or NA). For determination of level of appropriateness, interobserver comparisons represent the frequency with which reviewers selected indications with matching levels of appropriateness (appropriate vs inappropriate vs uncertain), though not necessarily an identical indication number. Studies for which both reviewers chose NA, and thus no appropriateness level was available for either reviewer selection, were excluded from comparisons of appropriateness level.
Results
Overall, 517 studies were enrolled. Of these, 8% ( n = 40) were excluded because of insufficient evidence to determine the primary indication. Additionally, in an effort to maximize the applicability of our results to the broad practice of SE, studies that were specifically related to stress testing protocols as part of solid organ transplantation programs at our institution (36% [ n = 188]) were excluded, leaving a final study group of 289 studies.
Of this study group, 93% ( n = 268) were outpatient studies, and 7% ( n = 21) were inpatient studies. The mean patient age was 59 ± 18 years, and 49% ( n = 141) of studies were performed on women ( Table 1 ). Fifty-one percent of the patients ( n = 145) underwent dobutamine SE, and 49% ( n = 144) underwent exercise SE. Cardiology was the most common referring specialty (45% [ n = 130]), followed by internal medicine or primary care (27% [ n = 78]), and anesthesiology from a preoperative evaluation clinic at our institution (25% [ n = 71]) ( Table 2 ).
| All studies ( n = 289) | Appropriate indication ( n = 180) | Inappropriate indication ( n = 50) | Uncertain indications ( n = 23) | NA studies ( n = 36) | |
|---|---|---|---|---|---|
| Age (years) | 59 ± 18 | 63.3 ± 12 | 50 ± 18 † | 66.4 ± 12 ‡ | 55 ± 16 |
| Women | 51% (147) | 46% (82) | 64% (32) ∗ | 43% (10) | 65% (23) |
| Outpatients | 93% (268) | 91% (163) | 98% (49) | 100% (23) | 93% (33) |
∗ P < .05 vs appropriate studies.
† P < .01 vs appropriate studies.
| Ordering physician specialty | All studies ( n = 289) | Appropriate indication ( n = 180) | Inappropriate indication ( n = 50) | Uncertain indications ( n = 23) | NA studies ( n = 36) |
|---|---|---|---|---|---|
| Cardiology | 45% (130) | 49% (89) | 36% (18) ∗ | 57% (13) † | 29% (10) |
| Internal medicine/primary care | 27% (78) | 28% (50) | 18% (9) ∗ | 30% (7) † | 34% (12) † |
| Anesthesiology (preoperative clinic) | 25% (71) | 18% (32) | 46% (23) ∗ | 9% (2) † | 37% (14) |
| Other | 3% (10) | 5% (9) | 0% | 4% (1) | 0% |
∗ P < .01 vs appropriate studies.
Of the final study group, 88% ( n = 253) were ordered for indications outlined in the AC document, while 12% ( n = 36) were ordered for indications NA by the document. Of the 253 studies for which the AC document could be applied, 71% ( n = 180) studies were appropriate, 9% ( n = 23) were uncertain, and 20% ( n = 50) (Figure 1) were inappropriate studies. The frequency of studies in each AC table and the frequency of studies ordered for the most common indications as outlined the AC document are summarized in Tables 3 and 4 , respectively. The most common indication for SE was indication 3, “Evaluation of chest pain syndrome or anginal equivalent in patients with intermediate pre-test probability of CAD [coronary artery disease] who have an interpretable ECG [electrocardiogram] who are able to exercise,” accounting for 34% of all studies.
| Table number | % (number of studies) ( n = 253) | Appropriate ( n = 180) | Inappropriate ( n = 50) | Uncertain ( n = 23) |
|---|---|---|---|---|
| 1 | 56.5% (143) | 85% | 15% | 0 |
| 2 | 9.9% (25) | 0 | 24% | 76% |
| 3 | 1.6% (4) | 75% | 0 | 25% |
| 4 | 0.8% (2) | 50% | 0 | 50% |
| 5 | 28.5% (72) | 72% | 0 | 28% |
| 6 | 0.4% (1) | 100% | 0 | 0 |
| 7 | 1.2% (3) | 67% | 33% | 0 |
| 8 | 0 | 0 | 0 | 0 |
| 9 | 1.2% (3) | 0 | 33% | 67% |
| Indication as listed in the AC for SE | % of all studies for which indication was covered in the AC for SE ( n = 253) |
|---|---|
| Indication 3: “Evaluation of chest pain syndrome or anginal equivalent in patients with intermediate pre-test probability of CAD who have an interpretable ECG and who are able to exercise” (appropriate) | 34% |
| Indication 30: “Pre-operative evaluation for an intermediate-risk noncardiac surgery in patients with poor exercise tolerance (less than or equal to 4 METs), and intermediate clinical risk predictors” (appropriate) | 22% |
| Indication 4: “Evaluation of chest pain syndrome or anginal equivalent in patients with an intermediate pre-test probability of CAD who have an uninterpretable ECG or unable to exercise” (appropriate) | 13% |
| Indication 1: “Evaluation of chest pain syndrome or anginal equivalent in patients with Low pre-test probability of CAD who have an interpretable ECG and able to exercise” (inappropriate) | 9% |
| Indication 13: “Detection of CAD and risk assessment in asymptomatic (without chest pain syndrome or anginal equivalent) general patient populations with high CHD risk (Framingham)” (uncertain) | 8% |
| Indication 29: “Pre-operative evaluation for an intermediate-risk non-cardiac surgery in patients with poor exercise tolerance (less than or equal to 4 METs), and minor or no clinical risk predictors” (inappropriate) | 5% |
Inappropriate studies were more likely to be ordered on women (23% of studies ordered on women were inappropriate vs 16% of studies ordered on men, P = .03) and younger patients (mean age for appropriate studies, 50 ± 18 years vs 63 ± 12 years for appropriate studies; P < .001). Cardiologists were less likely to order inappropriate studies compared with noncardiologists (15% of inappropriate studies vs 24% of inappropriate studies, P < .05). Inappropriate studies were most frequently ordered by anesthesiologists from an anesthesia preoperative clinic at our institution (46% of all inappropriate studies; Table 2 ). Among studies ordered for preoperative assessment, cardiologists had 4% inappropriate studies (one of 25), internal medicine or primary care physicians had 0% inappropriate studies (zero of six), and anesthesiologists had 51% inappropriate studies (19 of 37) ( P < .001, anesthesiologist-ordered vs non-anesthesiologist-ordered preoperative studies).
Indication 1 (“Evaluation of chest pain syndrome or anginal equivalent in patients with low pre-test probability of CAD who have an interpretable ECG and able to exercise”) was the most common inappropriate indication, accounting for 44% of all inappropriate indications. The frequencies of the five most commonly ordered inappropriate studies (accounting for >95% of all inappropriate studies) are summarized in Table 5 .
