46_Section_L _Exhibit_L-10_Sample_Quality_Improvement_Study.docx

Chest Pain – Final Report

Focused Study: Chest Pain

Exhibit L-10 Sample Quality Improvement Study

TQMC HT9402-15-R-0005, Section L, Exhibit L-10 Page 1 of 35

TRICARE FOCUSED STUDY CHEST PAIN

INDICATIONS FOR STUDY

An epidemiologic review of 27 months of Military Health System Data Repository (MDR) inpatient claims1 disclosed that three related ICD-9 codes (786.50, chest pain, unspecified, 786.51, precordial pain, and 786.59 chest pain, other discomfort/pressure/tightness in chest) together were the 6th most common primary diagnosis at discharge and accounted for $24,230,848 in paid claims. A total of 6,393 patients were admitted and discharged with a final primary diagnosis of one of these three non-specific diagnoses. On the basis of MDR inpatient claims, only 351 (22%) underwent cardiac catheterization, 12 underwent cardiovascular treadmill testing, 28 were subject to other cardiovascular stress test and 27 underwent cardiac diagnostic ultrasound. None of these patients were discharged with a more specific primary cardiac diagnosis (Table 1) nor were they discharged with a primary diagnosis suggesting pulmonary (e.g., pneumonia, bronchitis, pulmonary embolism), gastrointestinal (e.g., reflux esophagitis, gastric or duodenal ulcer), aortic (e.g., dissection, aneurysm) or musculoskeletal disease.

TABLE 1: Cardiac Diagnoses

· 411.81 - Coronary Occlusion without infarction

· 411.89 - Acute Coronary Insufficiency (or) subendocardial ischemia

· 424.0 - Mitral Valve Disorders

· 424.1 - Aortic Valve Disorders

· 424.2 - Tricuspid Valve Disorders

· 424.3 - Pulmonary Valve Disorders

In a review of 10,689 patients presenting to the emergency departments of 10 hospitals, Pope, et al.2 found that 2.1 % of patients with acute myocardial infarction (AMI) and 2.3% of patients with unstable angina were "mistakenly discharged" from the emergency department (ED) and Chest Pain – Final Report

TQMC HT9402-15-R-0005, Section L, Exhibit L-10 Page 9 of 35 had 1.9 and 1.7 times, respectively, the mortality compared to those patients in the study admitted with similar conditions. Lee and Goldman3 noted that patients with AMI mistakenly discharged from the ED have a short-term mortality of approximately 25%, twice that of patients with AMI who are admitted, and Farkouh et al.4 found that only 3.6% of patients presenting to an ED with chest pain were found to have myocardial infarction with ST-segment elevation.

In a multicenter controlled clinical trial, Selker et al. found that improved evaluation, using an acute cardiac ischemia time-insensitive predictive instrument (a mathematical model) was “associated with substantial reductions in admissions to the CCU, telemetry unit, and hospital, particularly in settings in which overuse is greatest, without adversely affecting care."5

MATERIALS

Charts were selected for review if the final primary diagnosis at discharge is any of the following:

METHODS

In a population-based cohort study four patients presenting to the ED with chest pain were classified on the basis of history, physical findings and electrocardiogram (ECG) findings into high, intermediate and low risk groups. Thirty-day mortality was 6.9%, 1.7% and 0% and non- fatal myocardial infarctions were noted in 2.0%, 2.4% and 0% in the high, intermediate and low risk groups respectively. Similarly, Lee et al. concluded that "improved diagnostic accuracy among low-risk patients may decrease costs of medical care, but the identification of these patients requires the interpretation of several pieces of data from the history and physical examination and usually requires an ECG as well."6

To assess patients presenting with chest pain a thorough patient history should be documented. We abstracted from the patient record the presence, absence or failure to record 17 key variables (Table 2) from the history of present illness.

TABLE 2: History of Present Illness

Onset of CP (min prior to presentation) Duration of CP (min) Last Meal (min prior to presentation) Pain Resolved (time since) Was Pain postprandial History of Palpitations History of Episode(s) of Similar Pain Pain related to Position Pain Relived by change in position

Pain Relieved by antacids Pain Relieved by Nitrates History of Trauma History of Ischemic CAD History of CABG History of Valve Surgery History of Diabetes History of Peripheral Vascular Disease

Several authors have emphasized the critical importance of the character of the chest pain. A final diagnosis of AMI or unstable angina was made in 54% of patients with pain described as "pressure," in 44% described as "burning/indigestion," in 32% characterized as "aching" and in 22% reported as "sharp or stabbing."6 The National Heart Attack Alert Program Coordinating Committee recommended7 that patients with chest pain described as pressure, tightness, heaviness, indigestion or heartburn, with pain radiating to the neck, jaw shoulders back or arms, or with nausea and/or vomiting require immediate evaluation. We abstracted from the patient chart the recorded characteristics of the chest pain including the character (Table 3) and the presence or absence as well as the location of radiation (Table 4) of the pain.

TABLE 3: Character of Chest Pain

Sharp Dull Pressure Squeezing Tightness Heaviness Burning Unspecified

TABLE 4: Radiation of Chest Pain

Neck Jaw Shoulders Arms Abdomen Teeth Back None Unspecified

The presence or absence of associated symptoms is also important as noted by the National Heart Attack Alert Program Coordinating Committee7 who suggested that patients with shortness of breath or with weakness, dizziness, lightheadedness or loss of consciousness be assessed immediately. The patient record, specifically the history of present illness (HPI) section, was reviewed and any associated symptoms (Table 5) were noted.

TABLE 5: HPI: Associated Symptoms

Nausea / Vomiting Diaphoresis Shortness of Breath Lightheadedness Dizziness None Unspecified

To further characterize the chest pain, and to qualify any associated activities, data abstraction included the presence or absence of any precipitating factors (Table 6) and whether or not the pain is relieved by rest.

TABLE 6: Precipitating Factors

Walking Walking Uphill/Fast Sexual Activity Emotion Other Activity

The presence or absence of cardiac risk factors has been shown to correlate with the long term probability of a cardiovascular event. Although several authors have suggested that such risk factors are not helpful in the identification of Acute Coronary Syndrome (ACS) the TRICARE population may be unique by virtue of a relatively young patient age. In a recent study, Han et al.8 found that the relationship of cardiac risk factors, specifically diabetes, hypertension, smoking, hypercholesterolemia, and family history of coronary artery disease, to the probability of ACS varied with patient age. These authors concluded that "the relationship between cardiac risk factor burden and acute coronary syndromes was significantly modified by age" and stated that "for patients younger than 40 years, we observed that as the number of total cardiac risk factors increased, the odds of acute coronary syndromes dramatically increased" whereas "for patients older than 65 years, cardiac risk factor burden was less useful for the prediction of acute coronary syndromes." To investigate the absence of each of these risk factors we abstracted information (or lack thereof) regarding these risk factors (Table 7) from the records identified for review. Each of these factors relating to patient and family history must be considered in context and none can be relied upon individually to establish the presence of absence of ACS.

TABLE 7: Risk Factors

Smoking Hypertension Diabetes Hyperlipidemia Family History of CAD

Although it is widely believed that chest pain relieved by nitroglycerin is likely to be cardiac in origin, several authors have suggested that this relationship is far from clear. Henrikson et al.9 studied patients admitted with a diagnosis of "rule out myocardial infarction" or "chest pain." These researchers defined nitroglycerin-responsive chest pain as "a 50% or greater reduction in the intensity of chest pain within approximately 5 minutes of administering 0.4 mg of sublingual or 0.4 mg of spray nitroglycerin," acute coronary artery disease (CAD) as "appropriate symptoms with at least 1 of the following during the index hospitalization or during the follow-up period: any measurement of elevated serum troponin T level (≥0.1 µg/L [normal level <0.1 µg/L in the study laboratory]), coronary angiography demonstrating a 70% or greater stenosis, a positive exercise test result with or without imaging, or the diagnosis of active CAD without testing (defined as the primary diagnosis for admission being active CAD as noted by the clinical attending physician, with concurrence by a study cardiologist who was blinded to nitroglycerin response, on the basis of standard clinical criteria)" and no active CAD as "no troponin T level elevation during the index hospitalization or the follow-up period and at least 1 of the following: coronary angiography without flow-limiting stenoses or a negative exercise test result with or without imaging." In patients with active CAD, 35% obtained pain relief with nitroglycerin. In patients with no active CAD, nitroglycerin provided pain relief in 41%. Similar findings were reported by Shry, et al.10 who reported that although 88% of patients with cardiac chest pain responded favorably to nitroglycerin, a similar response rate of 92% was found in patients with non-cardiac chest pain.

The relationship of chest pain to preceding events can also be misleading. Postprandial chest pain is often ascribed to gastroesophogeal disease, but the possibility of postprandial cardiac pain cannot be discounted. Berlinerblau and Shani11 found that postprandial angina was observed more often in males, and was associated with left main and three vessel coronary artery disease and with a high incidence of angina at rest.

The physical examination is often normal in patients with ACS but is an essential component of the initial evaluation. The finding of significant hypertension may suggest the possibility of an acute aortic dissection. The presence of a right-left arm-arm gradient or decreased femoral pulses would also be consistent with an aortic dissection, but are more commonly found with chronic obstructive peripheral vascular disease. Cardiac auscultation might reveal a cardiac murmur and such entities such as mitral valve prolapse may be associated with chest pain that can mimic that of CAD. Auscultation of the chest may also detect rales, rhonchi or egophony suggestive of pulmonary parenchymal disease. A pleural friction rub may also suggest pulmonary disease but can also be present with pulmonary embolism and other conditions involving the pleura. A pericardial friction rub would strongly suggest the presence of pericarditis. Reproduction of the patient's pain with palpation of the chest wall suggests, but does not establish, the diagnosis of musculoskeletal pain. The physical findings in the patient record were carefully reviewed and such information (Table 8) was abstracted and analyzed.

Table 8: Physical Examination

Blood Pressure (BAR, BAL) Femoral Pulses Cardiac Auscultation Chest auscultation Findings on palpation

The chest roentgenogram is also normal in most patients with ACS but, like the physical examination, may suggest an alternative etiology for the patient's chest pain. Parenchymal consolidation may suggest an infectious process, as may a pleural effusion. Pulmonary embolism may be suggested by the presence of pulmonary oligemia or a wedge-shaped defect suggestive of pulmonary infarction. Aortic dissection can result in a myriad of abnormalities including an abnormal aortic shadow, mediastinal widening, or an enlarged cardiac silhouette. Accordingly, the final radiology reports were reviewed and information (Table 9) extracted for review and analysis.

Table 9: Chest Roentgenogram

Parenchymal Abnormal Aortic Abnormality Mediastinal Widening Cardiac Silhouette Abnormal Other (Specify)

An essential component of the initial evaluation of a patient with chest pain in whom ACS is considered is the electrocardiogram (ECG). The ACC/AHA Guidelines for the Management of Patients With ST-Elevation Myocardial Infarction12 advises (Class I Recommendation, Level of Evidence: C) that a "12-lead ECG should be performed and shown to an experienced emergency physician within 10 minutes of ED arrival for all patients with chest discomfort (or anginal equivalent) or other symptoms suggestive of STEMI [ST elevated Myocardial Infarction]." These guidelines also recommend (Class I Recommendation, Level of Evidence: C) that "if the initial ECG is not diagnostic of STEMI but the patient remains symptomatic, and there is a high clinical suspicion for STEMI, serial ECGs at 5- to 10-minute intervals or continuous 12-lead ST- segment monitoring should be performed to detect the potential development of ST elevation".

Findings on admission ECG have significant diagnostic and prognostic value. Although ST- segment elevation is frequently considered synonymous with AMI, ST-segment depression can be just as ominous. In a review of 12,142 patients with symptoms of cardiac ischemia and an ECG within 12 hours of admission confirming myocardial ischemia, Savonitto et al.13 found both ST-elevation (≥.05mV in at least 2 contiguous leads) and ST-depression (≥.05mV) to predict outcomes at 30 days and at 6 months. Combined ST-segment elevation and depression was present in only 15% of patients but had the highest 6 month death or re-infarction rate (15.7%). ST-segment depression was found in 35% of patients and was associated with a 15.4% 6-month death or re-infarction rate. Interestingly, ST-segment elevation was less common, being present in 28% of patients, and was associated with a 12.3% 6-month combined death or re-infarction rate. Isolated T-wave inversion was present in 22% of patients and was associated with an 8.1% incidence of death or re-infarction at 6 months.

In a review of the data from the Thrombosis in Myocardial Infarction (TIMI) trial Antman et al.14 using multivariate analysis found ST deviation (ST-segment elevation or persistent ST depression of ≥.05mV) to be one of seven TIMI risk score predictor variables for the primary endpoint of all-cause mortality, new or recurrent MI, or severe recurrent ischemia requiring urgent revascularization. The importance of ECG findings in patients with ACS without persistent ST-segment elevation was emphasized by Boersma, et al.15 who reported that ST- segment depression (>0.5mm) on multivariate analysis was highly predictive (p<0.0001) at 30 days of both death (χ2 =20) and death or re-myocardial infarction (χ2 =14). The ACC/AHA Guideline Update for the Management of Patients With Unstable Angina and Non–ST-Segment Elevation Myocardial Infarction—200216 also reported that ST-segment depression was among the most important baseline features associated with death.

In addition to these widely reported associations, other findings on admission ECG, such as left bundle branch block (LBBB), pathologic Q-waves, PR depression and tachycardia have been associated with a higher risk of ACS. The final interpretation of the admission ECG, and any repeat ECGs (Tables 10 and 11) were reviewed and the findings abstracted for clinical review and analysis.

TABLE 10: Initial and Repeat ECG

Timing - Within 15 min ST Elevation ≥ 1 mm ST Elevation ≥ 2mm LBBB - New ST Depression T wave Inversion Pathologic Q Waves PR Depression (diffuse) Tachycardia

TABLE 11: Continuous ST-Segment Monitoring

ST Elevation ≥ 1 mm ST Depression

In addition to an initial ECG and, if necessary repeat ECG(s) or ST-monitoring, the importance of biomarkers cannot be overemphasized. The ACC/AHA 2007 Guidelines for the Management of Patients With Unstable Angina/Non-ST-Elevation Myocardial Infarction: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines17 recommends that "Cardiac biomarkers should be measured in all patients who present with chest discomfort consistent with ACS" (Class I Recommendation, Level of Evidence: B). Although the American College of Emergency Physicians Clinical Policies Subcommittee (Writing Committee) on Non–ST-Segment Elevation Acute Coronary Syndromes18 recommends (Level B) that a single negative CK-MB mass, Troponin I or Troponin T measured 8-12 hours after symptom onset can be used to exclude non-ST-segment elevation AMI, a more recent recommendation from the ACC/AHA17 states that "A cardiac specific troponin is the preferred marker, and if available it should be measured in all patients who present with chest discomfort consistent with ACS" (Class I Recommendation, Level of Evidence B). Similarly, the National Academy of Clinical Biochemistry Laboratory Medicine Practice Guidelines19 agree that "Biomarkers of myocardial necrosis should be measured in all patients who present with symptoms consistent with ACS" (Class I Recommendation, Level of Evidence: C) and stipulate that "cardiac troponin is the preferred marker for the diagnosis of MI" (Class I Recommendation, Level of Evidence: A) while acknowledging that "MB (CK-MB) by mass assay is an acceptable alternative when cardiac troponin is not available" (Class I Recommendations, Level of Evidence: A). At present, it is reasonable to expect Troponin I or Troponin T assays to be available at acute care facilities, especially those with inpatient populations. The ACC/AHA Guidelines17 state that "patients with negative cardiac biomarkers within 6 h of the onset of symptoms consistent with ACS should have biomarkers re-measured in the time frame of 8 to 12 h after symptom onset" and that "The exact timing of serum marker measurement should take into account the uncertainties often present with the exact timing of onset of pain and the sensitivity, precision, and institutional norms of the assay being utilized as well as the release kinetics of the marker being measured" (both: Class I Recommendation, Level of Evidence B). Although timing of the Troponin assay may be important, the time of symptom onset may be difficult to discern, especially on retrospective review, and the timing of sample collection may be imprecise. Accordingly, we reviewed the record to determine the sampling time of Troponin levels and determined if the sample was obtained more than 3 hours after the onset of symptoms (Table 12).

TABLE 12: Cardiac Biomarkers -Troponin (I or T)

Initial Assay (normal or abnormal) Timing > 3 hours after symptom onset

Repeat Assay (normal or abnormal) Timing > 3 hours after symptom onset

In patients with normal ECGs and normal cardiac biomarkers an exercise tolerance test (ETT) becomes the basis for further management. The ACC/AHA Guidelines for the Management of Patients With Unstable Angina and Non–ST-Segment Elevation Myocardial Infarction20 and the subsequent ACC/AHA 2007 Guidelines for the Management of Patients With Unstable Angina/Non-ST-Elevation Myocardial Infarction21 advise that "In patients with suspected ACS in whom ischemic heart disease is present or suspected, if the follow-up 12-lead ECG and cardiac biomarkers measurements are normal, a stress test (exercise or pharmacological) to provoke ischemia should be performed in the ED, in a chest pain unit, or on an outpatient basis in a timely fashion (within 72 h) as an alternative to inpatient admission. Low-risk patients with a negative diagnostic test can be managed as outpatients" (Class I Recommendation, Level of Evidence: C).

There is general agreement that if the patient has no symptoms on ETT, yet fails to achieve at least 85% of Maximum Predicted Heart Rate (MPHR) and fails to achieve at least 3 METS, the test is non-diagnostic. The ETT is considered positive if the ETT reproduces the patient's symptoms, if ≥1 mV ST-depression/elevation is documented, if systolic blood pressure decreases more than 10 mmHg or if there is deemed to be abnormal heart rate recovery. A retrospective review of the final interpretation of the ETT abstracted this information (Table 13) for review and analysis.

TQMC HT9402-15-R-0005, Section L, Exhibit L-10 Page 10 of 35

TABLE 13: Exercise Stress Test

≥ 1 mV ST-depression/elevation HR < 85% MPHR Symptom Limited Decrease in BP >10 mmHg <3 METs achieved Abnormal HR Recovery Myocardial imaging to assess the risk of ACS inpatients includes both Myocardial Perfusion Imaging (MPI) and echocardiography. In patients without documented myocardial ischemia or necrosis, imaging can be performed using either exercise or pharmacological stress. The ACC/AHA/ASNC Guidelines for the Clinical Use of Cardiac Radionuclide Imaging22 advise that "Radionuclide imaging is most appropriate in patients with possible ACS. After initial triage on the basis of symptoms, ECG, and history, rest single-photon emission CT (SPECT) imaging appears to be useful for identifying patients at high risk (those with perfusion defects), who should be admitted, and patients with low risk (those with normal scans), who in general may be discharged home with a low risk for subsequent ischemic events." For patients with "chest pain with non-diagnostic ECG and negative serum markers and enzymes or normal resting scan" the ACC/AHA/ASNC guidelines suggest same-day rest/stress perfusion imaging (Class I Recommendation, Level of Evidence B). If the serum markers and enzymes as well as the ECG is non-diagnostic, a rest MPI is suggested (Class I Recommendation, Level of Evidence A).

Echocardiography can yield similar information and avoids exposing the patient to radiation.

The ACC/AHA Guidelines for the Clinical Application of Echocardiography23 advises that echocardiography is indicated in the diagnosis of underlying cardiac disease in patients with chest pain and clinical evidence of valvular, pericardial, or primary myocardial disease (Class I Recommendation) and in the evaluation of chest pain in patients with suspected acute myocardial ischemia, when baseline ECG is non-diagnostic and when study can be obtained during pain or soon after its abatement (Class I Recommendation). In an update of these guidelines in 200324, with regard to the section relating to the Recommendations for Echocardiography in the Diagnosis of Acute Myocardial Ischemic Syndromes, resulted only in the movement of a recommendation from Class IIa to Class I and a minor wording change.

Accordingly, we reviewed the final reports and abstract information relating to these diagnostic studies (Table 14) for further review and analysis.

Table 14: Additional Studies

Resting Myocardial Perfusion Imaging Stress Myocardial Perfusion Imaging Resting Echocardiography Stress Echocardiography

TQMC HT9402-15-R-0005, Section L, Exhibit L-10 Page 19 of 35

More recent studies have advocated the use of coronary artery imaging for the evaluation of patients suspected to have ACS. These newer technologies may identify those patients with chest pain who may be safely discharged home from the ED, Observational Status or brief inpatient hospitalization. In a study of Coronary Artery Calcium (CAC) scoring in the ED involving 1,031 patients, Nabi et al.25 found that after a mean follow-up of 3.3 months, only two cardiac events occurred in 625 patients with a calcium score of 0. Although the positive predictive value (PPV) of CAC scoring was poor, a negative predictive value (NPV) of nearly 100% suggests that patients with a CAC score of 0 could be safely discharged. Another imaging modality, Computed Tomography Coronary Angiography (CTCA) may have applicability in low-risk patients such as those presenting to the ED with chest pain not particularly suggestive of ACS and those with low (0 to 2) TIMI risk scores. Among 368 low-risk patients, only 31 of which had ACS, Hoffman et al.26 reported that the absence of any plaque on CRCA had a NPV of 100% (183/183) for ACS and a NPV of 100% (183/183) for myocardial infarction. This, and similar studies, suggest that CTCA may be used in this low risk population to identify patients who may be safely discharged. The absence of significant (>50% luminal) stenosis had a similar NPV of 98% (293/300) for ACS and 99% (297/300) for myocardial infarction. To assess the use of these newer coronary artery imaging techniques, reviewed and abstracted findings from the final reports (TABLE 15) for analysis.

TABLE 15: Coronary Artery Imaging

Coronary Artery Calcium (CAC) scoring Computed Tomography Coronary Angiography

In the United States, chest pain is the second most common reason for presentation to the ED (after stomach pain, cramps and spasms) for patients over 14 years of age, and the most common reason in patients aged 65 years or over.27 Although ACS is a major concern in these patients and warrants diagnostic consideration, the vast majority of these patients will not have acute cardiac conditions. Many of these patients will have other cardiovascular, pulmonary or gastrointestinal conditions as the cause of their acute chest pain. Diagnostic studies to detect other acute, and in some cases life-threatening, conditions may be warranted. Of all non-cardiac causes of chest pain, aortic dissection, pneumothorax and pulmonary embolism are the most common life- threatening emergencies seen in the ED.28

Perhaps the most concerning non-cardiac condition that may present as sudden onset of severe chest pain is acute aortic dissection. This may start in the ascending or descending aorta and progress a variable length distally, potentially impairing renal, mesenteric and/or peripheral flow. Diagnosis is based on the history, physical findings and selected diagnostic studies. Pain often radiates to the back and there may be a blood pressure gradient between the upper extremities and/or a decrease in femoral pulses. Chest roentgenogram may show an abnormal aortic silhouette and/or mediastinal widening. The diagnosis can be made on Computed Tomography

(CT) of the chest and aorta. As an integral part of the retrospective chart review, we will abstract data from the final CT report if applicable.

A tension pneumothorax may be the result of trauma, or may occur spontaneously. Often presenting with acute respiratory distress the diagnosis is strongly suspected if auscultation of the chest documents reduced or absent breath sounds unilaterally. Pneumothorax can be seen on upright PA and lateral chest roentgenogram and can be clearly demonstrated on CT.

The incidence of pulmonary embolism (PE) in the U.S. has been widely reported as approximately 1 per 1,000 persons per year with variation dependent upon the sensitivity of the diagnostic modality used. However, the true incidence of this life-threatening disease is almost certainly underestimated. The chest pain associated with PE is often acute and associated with dyspnea as well as tachypnea and possibly tachycardia. If the parietal pleura is involved, there may be a pleuritic component to the pain. Auscultation of the chest may detect a pleural friction rub, but this finding is not diagnostic. A chest roentgenogram may show pulmonary oligemia or a wedge-shaped defect but cannot be relied upon for definitive diagnosis. A D-Dimer test will identify the breakdown products of the inciting thrombus and although the absence of D-Dimers has a high NPV, the presence of these breakdown products is non-specific. The vast majority of pulmonary emboli originate in the lower extremities and a venous duplex scan can often identify such lesions. A Ventilation/Perfusion (V/Q) scan of the chest will identify any ventilation perfusion mismatches which, in concert with the history and physical examination, can accurately predict the likelihood of a PE. The diagnosis can be established on pulmonary angiography or on CT scan using a PE protocol. The retrospective chart review abstracted these addition data points (Table 16) for review and analysis.

TABLE 16: Pulmonary Embolism Evaluation

D-Dimer Determination Chest Roentgenogram Ventilation/Perfusion scan Duplex Venous Ultrasound Pulmonary Angiography CT using PE Protocol

Although gastrointestinal causes of chest pain are usually not as immediately life-threatening as aortic dissection, PE or pneumothorax, there are a number of serious, and potentially fatal, gastrointestinal conditions that can mimic cardiac pain. Perhaps the most common is gastroesophogeal reflux disease (GERD) with esophagitis. Although the characteristic symptoms of GERD are often postprandial, exacerbated by recumbency and sometimes relieved with antacids, endoscopy is required for definitive diagnosis (and possible biopsy). A course of a proton pump inhibitor is often prescribed for long-term management in addition to antacids.

Acute pancreatitis, and to a lesser extent biliary disease, can present as epigastric and/or substernal pain which can be confused with cardiac discomfort. Although a variety of enzyme and imaging tests can be utilized to investigate these possibilities, abdominal CT scan is often relied upon for further evaluation. The charts were reviewed and data (Table 17) including information in the final radiology and endoscopy reports (if applicable) regarding any gastrointestinal evaluation was abstracted for clinical review and analysis.

TABLE 17: Gastrointestinal Evaluation

GI - Endoscopy Abdominal CT Scan Other - Specify Rx for Proton Pump Inhibitor

Following the initial evaluation, some patients may be considered to be at high risk for ACS and may undergo coronary angiography and possible thrombolytic therapy and/or coronary angioplasty with or without stent placement. Information regarding any such invasive procedures was abstracted from final reports, correlated with other patient-specific data and subject to further review and analysis.

If it is determined, after initial evaluation, that the patient may be safely discharged home, further evaluation may be scheduled on an outpatient basis. The hospital chart was carefully reviewed and any documentation for recommended or scheduled follow-up evaluation (Table 18) was recorded.

TABLE 18: Recommended/Scheduled Follow-up

Cardiology Outpatient Pulmonology Outpatient Gastroenterology Outpatient Other Outpatient Re-admission < 14 days Outpatient Diagnostics

Chart selection is based on the final primary diagnosis on discharge as recorded in the MDR inpatient claims database. The possibility must be considered that the ICD-9 code assigned as the final primary diagnosis may not accurately reflect the primary diagnosis recorded in the physician discharge note or in the dictated discharge summary. Therefore, the text of the final primary diagnosis as recorded in the dictated discharge summary was abstracted and recorded.

KEPRO reviewers, experienced in discharge coding, will assign the ICD-9 code that most accurately reflects that diagnosis at the highest level of specificity. This information will be compared and contrasted to the data from MDR claims.

RESULTS

Selection Criteria

Initially, 3,098 charts qualified and were requested for the study by the following inclusion criteria:

· In network

· Correct ACV group

· Date range: 6/2010 through 5/2012

· ICD-9 diagnosis at discharge of:

· 786.50 – chest pain, unspecified

· 786.51 – precordial pain

· 786.59 – chest pain, other discomfort/pressure/tightness in chest

· Provider information available

· Paid claims

· Correct institution type (no psych hospitals or psych wards, keep “General Medical and Surgical,” “Chronic Disease,” “Extended Care Facility” and “Sole Community”)

· Adults

Of the charts requested, 2,758 had QBuilder Questionnaires completed and the tables and figures correspond to these charts.

Table 19: Patient Characteristics - Results

Count (% of 2,758)

Patient was admitted from, n (%):

ICD-9 diagnosis at discharge, n (%)

786.59 – chest pain, other discomfort /pressure/tightness in chest 1,975 (71.61)

TRICARE Region, n (%)

Table 20: History of Present Illness - Results

Figure 1: History of Present IllnessHistory of Present Illness # of response 'Yes'

# of response 'Yes'

Table 21: Character of Chest Pain - Results

Figure 2: Character of Chest PainCharacter of Chest Pain 1,000

# of response

Table 22: Radiation of Chest Pain – Results

If yes, where:

Figure 3: Radiation of Chest Pain – LocationLocation of Radiation of Chest Pain -data available for 1,408 charts

Arms Shoulders Back Neck Jaw Unspecified Abdomen

Table 23: History of Present Illness, Associated Symptoms - Results

If yes, specify:

Figure 4: Associated Symptoms of Chest PainAssociated Symptoms of Chest Pain -data available for 1,765 charts

TQMC HT9402-15-R-0005, Section L, Exhibit L-10 Page 20 of 35

Table 24: Precipitating Factors - Results

If yes, specify:

Figure 5 - Precipitating FactorsPrecipitating Factors -data available for 640 charts Other Activity Walking Emotion Walking Uphill / Sexual Activity

TQMC HT9402-15-R-0005, Section L, Exhibit L-10 Page 29 of 35

Table 25: Risk Factors - Results

Figure 6: Risk FactorsRisk Factors

Hypertension? Hyperlipidemia? Family history of

CAD?

Diabetes?

Smoking?

Table 26: Physical Examination – Results

Table 27: Chest Roentgenogram – Results

Table 28: Initial ECG – Results

Table 29: Repeat ECG - Results

Table 30: Continuous ST-Segment Monitoring - Results

Table 31: Cardiac Biomarkers - Results

Table 32: Exercise Stress Test - Results

Outcomes/findings:

Table 33: Additional Studies - Results

Table 34: Coronary Artery Imaging - Results

Table 35: Pulmonary Embolism Evaluation - Results

Table 36: Gastrointestinal Evaluation - Results

Table 37: Recommended/Scheduled Follow-up - Results

DISCUSSION

A total of 2,758 charts were reviewed. The final diagnosis was 786.50 chest pain, unspecified (n= 713, 26%), 786.51 precordial pain (n=70, 2%), or 786.59 chest pain, other discomfort/pressure/tightness in chest (n=1,975, 72%). None of these patients had a more specific final diagnosis such as 410 acute myocardial infarction, 411.XX other acute and subacute ischemic heart disease, or 413 angina pectoris. The majority of patients were admitted via the ER and the average LOS was 1.5 days. The average onset of chest pain was 75 hours prior to presentation and the average duration was 35 hours. Only 1,936 (70%) still had chest pain at the time of presentation. The history of present illness was often poorly documented in the record. A history of similar pain was present in 36%, absent in 30% and not recorded in 32%. Pain relief with nitrates was described in 34%, denied in 19% and not recorded in 43%. The majority (n= 2,161, 78%) had pain described as pressure, tightness, heaviness, burning, radiating to the neck, jaw, shoulders, back or arms, or pain associated with nausea and/or vomiting. The National Heart Attack Alert Program Coordinating Committee recommended29 that such require immediate evaluation. The committee also recommended that patients with shortness of breath or with weakness, dizziness, or lightheadedness be assessed immediately, and 49% (n=1,340) had one or more of these associated symptoms. A precipitating factor was reported in 23%, denied in 40% and not recorded in 34%. Risk factors were common with 88% having a history of smoking, hypertension, diabetes, hyperlipidemia or a family history of CAD. Physical examination confirmed hypertension (Sys BP > 140mmHg or Dias BP > 90mmHg) in 38% but blood pressure was usually (95%) measured in only one arm. Palpation of the chest wall reproduced the pain in only 5% of patients and murmurs, rales or rhonchi were rarely noted. Chest roentgenograms were normal in the vast majority of patients.

The ACC/AHA Guidelines for the Management of Patients With ST-Elevation Myocardial Infarction30 advises (Class I Recommendation, Level of Evidence: C) that a "12-lead ECG should be performed and shown to an experienced emergency physician within 10 minutes of ED arrival for all patients with chest discomfort (or anginal equivalent) or other symptoms suggestive of STEMI [ST elevated Myocardial Infarction]." An initial ECG was performed in 98%, but was obtained within 15 minutes of arrival in only 74% of patients. Of the 2,241 patients with a normal initial ECG, 85% had a repeat ECG. Findings associated with Acute Coronary Syndrome (ACS), including ST elevation ≥ 1 mm, new LBBB, ST depression, pathological Q waves, diffuse PR depression, tachycardia and/or T-wave inversion were described in 18.7% (n=517) on the initial ECG and in an additional 1.5% (n= 28) on repeat ECG. Continuous ST segment monitoring was documented in 76% (n=2,086) and showed ST depression in 73% (n=2,012) of the 2,758 patients. Cardiac biomarkers (Troponin I or T) were obtained in 97% (n=2,672), the majority (93%, n=2,473) of which were sampled at least 3 hours after the onset of symptoms. Initial troponin levels were reported as normal in 93% (n=2,564) and abnormal in 4% (n=108). A repeat troponin assay was performed in 91% (n=2,497), 95% (n=2,382) of which were obtained at least three hours after symptom onset. Repeat troponin assays were reported as normal in 88% (n=2,425) and abnormal in 3% (n=72) of the entire group. In summary, a total of 605 patients (22%) had an abnormal ECG or abnormal cardiac biomarkers.

The ACC/AHA31 advises that "In patients with suspected ACS in whom ischemic heart disease is present or suspected, if the follow-up 12-lead ECG and cardiac biomarkers measurements are normal, a stress test (exercise or pharmacological) to provoke ischemia should be performed.” Interestingly, of the 605 patients with an abnormal ECG or cardiac biomarkers, 114 had an exercise stress test (ETT), while only 492 (23%) of the 2,152 patients who did not have either an abnormal ECG or abnormal biomarkers had an ETT as recommended by the ACC/AHA guidelines. Of the total of 606 patients who underwent an ETT, 75 (12%) were non-diagnostic (HR< MPHR and <3 METS achieved) and 288 (48%) were abnormal. Of the total group of 2,758 patients, 836 (30%) had an abnormal ECG, an abnormal Troponin or an abnormal ETT.

A total of 905 patients underwent myocardial perfusion imaging (MPI). Of the 836 patients who had an abnormal ECG, an abnormal Troponin or an abnormal ETT, 311 (37%) underwent MPI. Of the remaining 1,922 patients who did not have an abnormal ECG, an abnormal Troponin or an abnormal ETT, 594 (31%) underwent resting or stress MPI. Of these 594 patients, 85 (14%) had a positive resting or exercise MPI. Combining these results, a total of 921 (33%) patients had an abnormal ECG, an abnormal Troponin, an abnormal ETT or a positive MPI.

Echocardiograms (echo) were performed on 1,134 patients, 1,046 were reported as having a resting Echo and 832 as having a stress echo. Resting echo studies were abnormal in 120 (11%) and 137 (16%) of stress echo studies were abnormal. Cardiac CT-Calcium scans were rarely done (n=14). Chest and/or Aortic CT were reported as normal in 321 patients and abnormal in 86 patients. A total of 209 (8%) of patients underwent Coronary CT Angiography and 31 of these (15%) were abnormal. Coronary Angiograms were performed in 685 patients (25%) and 88 (13%) of these were abnormal. PTCA was reportedly performed in 203 (7%) of patients during the initial admission.

A significant number of patients underwent diagnostic studies to evaluate the possibility of Venous Thromboembolism (VTE) and/or Pulmonary Embolism (PE). A total of 1,048 (38%) of patients had D-dimer studies performed, 901 (86%) were normal and 147 (14%) were abnormal. A total of 62 patients underwent a Ventilation Perfusion (V/Q) scan with three (4.8%) being interpreted as abnormal. Interestingly, 16 (of 901) patients with a normal D-dimer assay underwent a V/Q scan only one (6.7%) of which was abnormal. Pulmonary angiography was done in 64 patients and eight (12.5%) were abnormal. Of the patients with normal D-dimer studies, pulmonary angiography was performed in 20 and three of these (15%) were abnormal. The most commonly employed imaging study for PE was a CT using a PE protocol with a total of 391 studies being performed, 26 of which (6.6%) were abnormal. In the group of 901 patients with a normal D-dimer study, 145 underwent a CT with PE protocol, seven of which (5.1%) were abnormal. No patient with a negative VQ scan underwent pulmonary angiography but nine

TQMC HT9402-15-R-0005, Section L, Exhibit L-10 Page 30 of 35 of 59 (15%) of those patients with a negative VQ scan did undergo a CT using PE protocol all of which were negative.

An abdominal CT scan was performed on 100 patients (3.6%) and 28 (28%) of these were abnormal. Only 71 patients (2.6%) underwent gastrointestinal endoscopy, but 40 of these (56%) showed abnormal findings. Interestingly, 286 patients were given a new prescription for a proton pump inhibitor although none had a final discharge diagnosis of a gastrointestinal disorder.

The majority of patients had recommendations for further evaluation and/or follow-up post- discharge. Outpatient diagnostic studies were ordered in 1,725 patients (63%) and 1,043 patients (38%) were scheduled for a cardiology outpatient appointment. An additional 684 (25%) were given other outpatient appointments including 122 (4.4%) gastroenterology appointments and 47 (1.7%) pulmonary appointments.

These findings raise several issues that may be worthy of further consideration. The average LOS for the entire cohort was 1.5 ± 1.2 days. It is unknown how many patients could have been evaluated in the ER and discharged or admitted on an “observation basis.” This may be…