Update in Hospital Medicine FREE

The articles chosen for this update cover topics important for general internists with active inpatient practices. These original research papers published in 2002 gave new insights into common inpatient diagnoses and frequent challenges for inpatient physicians. The articles were identified through MEDLINE searches and review journals, such as ACP Journal Club and Journal Watch. Local and national experts were polled to help prioritize the list and highlight the work with the highest perceived impact. These papers question several conventional practices in inpatient medicine or point the direction into which the field may be heading.

Konstantinides S, Geibel A, Heusel G, et al.  Heparin plus alteplase compared with heparin alone in patients with submassive pulmonary embolism. N Engl J Med. 2002;347:1143-50. [PMID: 12374874]

Investigators asked whether the outcome of patients with submassive pulmonary embolism (PE) could be improved with thrombolytic therapy. The authors randomly assigned patients with PE diagnosed by ventilation–perfusion scanning, computed tomography (CT), or pulmonary angiography. Patients also had right ventricular dysfunction or pulmonary hypertension noted on echocardiography or electrocardiographic changes consistent with right ventricular strain or failure. Excluded patients were older than 80 years of age, were hemodynamically unstable, or had a contraindication to thrombolytic therapy. The primary outcome was a composite end point representing the failure of thrombolysis (for example, need for vasopressors, additional thrombolytics, and embolectomy), as well as such clinical outcomes as cardiopulmonary resuscitation, intubation, or death. The authors also studied secondary outcomes such as bleeding complications and recurrent PE.

A total of 256 patients from several European centers were randomly assigned, 138 were treated with unfractionated intravenous heparin alone, and 118 were treated with heparin and alteplase administered as a 10-mg intravenous bolus followed by 90 mg over 2 hours. Intravenous anticoagulation was continued in both groups for 3 days, and then patients began warfarin therapy. Outcomes were tabulated for the subsequent 30 days. The 2 groups were well matched, and an intention-to-treat analysis was done.

The patients who received thrombolytics had a higher probability of surviving for 30 days without an event (83% vs. 76%; P= 0.005). The relative risk for the primary end point in the heparin–placebo group compared with the heparin–alteplase group was 2.63 (95% CI, 1.32 to 5.26; P= 0.006), yielding a number needed to treat for benefit to prevent 1 composite event of 7. These findings seemed driven by a markedly lower need for rethrombolysis in the thrombolytic group (7.6% vs. 23.2%; P= 0.001).

Incidence of bleeding complications in this study was low. There were no differences in complication rates between the 2 groups, with 6 bleeding events overall and only 1 fatal bleeding event (in the heparin–placebo group). Incidence of bleeding in this study was far lower than that seen in other thrombolytic studies (1).

This paper suggests that thrombolysis has a role in management of acute submassive PE. Although its findings appeared to be driven by an end point (need for thrombolysis) that may not represent common clinical practice, the thrombolytic protocol of this study was at least as safe as standard heparin therapy and has potential for substantial clinical benefit. To apply the results of this paper in practice, patients with pulmonary emboli diagnosed by ventilation–perfusion scanning, CT, or pulmonary angiography who have no contraindications to thrombolysis should be risk-stratified with a baseline electrocardiogram to detect right ventricular strain or infarction. If no abnormalities appear on the electrocardiogram, echocardiography should be done to assess for signs of right ventricular dysfunction and pulmonary hypertension. As yet it is unclear whether dilated right ventricles or dilated pulmonary arteries seen on helical CT can replace the echocardiogram in this algorithm.

Salpeter SR, Ormiston TM, Salpeter EE.  Cardioselective β-blockers in patients with reactive airway disease: a meta-analysis. Ann Intern Med. 2002;137:715-25. [PMID: 12416945]

β-Blockers are grossly underused—as many as 35% of patients with myocardial infarction do not receive them (2 - 6). Part of this barrier may be fear of side effects, especially bronchospasm, even though this concern is based largely on older studies of nonselective β-blockers.

The authors of this paper aimed to quantify the risk for pulmonary compromise when a β₁-selective agent is used in patients with mild to moderate lung disease and to determine whether the cardiac drug blocks the effect of the patients' inhaled β-agonist. They performed a systematic search of MEDLINE using standard meta-analytic methods and identified randomized trials of β₁-selective β-blockers in patients who had mild to moderate chronic obstructive pulmonary disease (COPD) and asthma. Data were abstracted and pooled from 80 randomized trials that reported such outcomes as FEV₁ and use of medications, objective signs and symptoms, and clear descriptions of β-blocker dosing regimens.

Nineteen studies on single-dose treatment and 10 studies on continued treatment were included. The pooled results showed that the first dose of β-blocker had no adverse effects on symptoms and a negligible decrease in FEV₁(change, −7.46% [CI, −5.59% to −9.32%]). After continued use of the β-blocker for a few days or weeks, change in FEV₁ was reduced −0.42% [CI, −3.74% to 2.91%]) and there was no significant change in symptoms or inhaler use. Moreover, the single-dose and continued therapy β-blockers were associated with an increased response to inhaled bronchodilators. It is important to note that the follow-up periods of the studies included in this paper were short, so the safety of longer-term use requires further study.

To apply these data clinically, clinicians need to identify which patients have mild to moderate COPD or asthma; this task may be based on fairly subjective criteria that can be difficult to apply. Patients with mild disease may include those with cold- or exercise-induced asthma or single-allergen symptoms whose symptoms are stable or who are generally asymptomatic. Patients who have chronic lung disease and good functional status while taking a stable regimen of inhalers may also be eligible for β-blocker therapy.

Despite this potential difficulty, the key conclusion of this meta-analysis is that clinicians need to recalibrate their thinking with regard to risk for pulmonary compromise caused by cardioselective β-blockers. Patients with mild, well-controlled lung disease who need β-blockade for reduction of cardiovascular risk—whether the patient has had myocardial infarction or is a high-risk presurgical candidate—should no longer be considered ineligible for appropriate therapy. Hospitalization represents a clear opportunity to try a short course of these medications in a controlled environment, with the potential for continuing therapy on a long-term basis in patients who need it.

de Gans J, van de Beek D, European Dexamethasone in Adulthood Bacterial Meningitis Study Investigators.  Dexamethasone in adults with bacterial meningitis. N Engl J Med. 2002;347:1549-56. [PMID: 12432041]

In animal models, subarachnoid inflammation has been shown to cause the complications of meningitis. Studies have further shown that steroids reduce subarachnoid inflammation and reduce end-organ damage. Earlier trials of steroids in humans with meningitis were flawed because they focused only on children or were observational, retrospective trials with conclusions that were difficult to generalize to the broad population. In this randomized, double-blind, placebo-controlled trial, the authors sought to determine whether steroids improve the outcomes of bacterial meningitis.

Patients with the clinical diagnosis of meningitis were enrolled. On lumbar puncture, each patient had cloudy cerebrospinal fluid, Gram stain showing bacteria, or a leukocyte count greater than 1.100 × 10⁹ cells/L.

The primary outcome was the Glasgow Neurologic Outcome Scale measured 8 weeks after admission. On the basis of this score, patients were divided into groups of favorable or unfavorable outcomes. Patients with more than a minor disability were categorized as having an unfavorable outcome. Secondary outcomes included focal neurologic deficits, hearing loss, and paresis. Side effects of treatment, including hyperglycemia and bleeding, were measured. A total of 301 patients were enrolled at several sites in Europe; 144 were assigned to placebo, and 157 received dexamethasone. Dexamethasone, 10 mg intravenously, was given before antibiotics and then continued for 4 days along with empirical intravenous amoxicillin therapy.

The outcomes of patients with meningitis at 8 weeks are shown in Table 1. Patients who received dexamethasone were less likely to experience a negative neurologic outcome (15% vs. 25%; relative risk, 0.59 [CI, 0.37 to 0.94]; P= 0.03). In the subgroup of patients with Streptococcus pneumoniae meningitis, 26% in the dexamethasone group had an unfavorable outcome compared with 52% of the placebo recipients. Overall, fewer patients in the dexamethasone group died (7% vs. 15%; relative risk, 0.48 [CI, 0.24 to 0.96]; P= 0.04). The number needed to treat for benefit to prevent 1 unfavorable outcome was 10; however, in the streptococcus pneumonia group, the number needed to treat for benefit was only 4.

It should be noted that while 97% of the isolates were sensitive to amoxicillin, resistance patterns in the United States would probably necessitate broader antibacterial coverage. The study was not powered to detect a change in mortality, and this finding should be interpreted with caution. Side effects were rare and evenly split between the treatment and control groups, suggesting that the addition of dexamethasone in bacterial meningitis is associated with low morbidity and has potential for great benefit.

Turpie AG, Bauer KA, Eriksson BI, Lassen MR.  Fondaparinux vs enoxaparin for the prevention of venous thromboembolism in major orthopedic surgery: a meta-analysis of 4 randomized double-blind studies. Arch Intern Med. 2002;162:1833-40. [PMID: 12196081]

Francis CW, Davidson BL, Berkowitz SD, et al.  Ximelagatran versus warfarin for the prevention of venous thromboembolism after total knee arthroplasty. A randomized, double-blind trial. Ann Intern Med. 2002;137:648-55. [PMID: 12379064]

Two recent studies point out new developments in the field of anticoagulation and deep venous thrombosis (DVT) prophylaxis. These papers focus on 2 new anticoagulants: fondaparinux and ximelagatran. Fondaparinux is a synthetic pentasaccharide that binds antithrombin III specifically, leading to specific 10A inhibition. This medicine has several advantages over heparin: It has no antiplatelet activity, no activity against other factors, and no known drug interactions. It is administered subcutaneously on a daily basis, has 100% bioavailability, and is cleared by the kidneys. By contrast, ximelagatran, an engineered molecule, is a direct thrombin inhibitor. It also has no antiplatelet activity, no activity against other factors, and no known drug interactions. Ximelagatran is dosed orally twice daily and is 100% bioavailable. Both studies described here considered the advantages of these agents over currently approved DVT prophylaxis strategies in patients undergoing orthopedic surgery.

The fondaparinux study is a meta-analysis of 4 similar phase III, randomized, controlled trials performed by the same authors. These trials studied 7344 patients undergoing total hip, total knee, and hip open-reduction internal fixation. The primary outcome was venographic DVT or symptomatic PE after 11 days of therapy. The ximelagatran paper describes a multicenter, randomized, controlled trial of 680 patients undergoing total knee replacement. Researchers also used appearance of DVT or PE on venography as the primary outcome; venography was done between 7 and 12 days.

Fondaparinux therapy was initiated at least 6 hours after surgery and compared with enoxaparin, which was begun between 12 and 24 hours after surgery. In the ximelagatran study, therapy with the new medicine was started the morning after surgery and compared with a warfarin regimen, which is standard for patients having knee replacement surgery.

The meta-analysis of fondaparinux showed that venous thromboembolism was significantly less common in patients receiving fondaparinux than in those receiving enoxaparin (6.8% vs. 13.7%). The absolute risk reduction of 6.8% corresponded to an odds reduction of 55.2% (CI, 45.8% to 63.1%; P< 0.001). While bleeding was more common in the fondaparinux group, it did not lead to a significant increase in repeated surgery or mortality. The subgroup analysis showed that the bleeding risk was highest when the first dose was given within 9 hours after surgery. The ximelagatran trial described a similar absolute risk reduction, but it was not statistically significant. Venous thromboembolism occurred in 19.2% of patients receiving ximelagatran and 25.7% of those receiving warfarin (P= 0.070). There was no difference in major bleeding, which was seen in 1.7% of the ximelagatran group and in 0.9% of the warfarin group.

The first paper suggests that fondaparinux is more efficacious than enoxaparin for DVT prophylaxis in general orthopedics patients, potentially at the expense of higher bleeding complications. Further comparative studies are needed in which fondaparinux is administered later and the enoxaparin is administered sooner after surgery. It is important to note that fondaparinux cannot be used in patients with renal failure and should not be given to patients who received epidural or neuraxial anesthesia during or after surgery. As a synthetic relative of enoxaparin, the true safety of fondaparinux in terms of heparin-induced thrombocytopenia is still unknown.

The second article indicates that ximelagatran is at least as efficacious as warfarin in patients having total knee replacement surgery. But because most orthopedists have shifted away from warfarin to enoxaparin, future studies will need to compare ximelagatran with enoxaparin.

Both agents have substantial advantages with regard to convenience of dosing. Their cost-effectiveness and role in prolonged prophylaxis are still unknown. Currently, fondaparinux is not mentioned in guidelines for DVT prophylaxis in patients having orthopedic surgery. Ximelagatran is not yet approved by the Food and Drug Administration, but it will be soon. Therefore, until more studies emerge and the relative advantages of these new agents are better understood, enoxaparin and warfarin should remain the standard of care.

Auerbach AD, Wachter RM, Katz P, et al.  Implementation of a voluntary hospitalist service at a community teaching hospital: improved clinical efficiency and patient outcomes. Ann Intern Med. 2002;137:859-65. [PMID: 12458985]

Meltzer D, Manning WG, Morrison J, et al.  Effects of physician experience on costs and outcomes on an academic general medicine service: results of a trial of hospitalists. Ann Intern Med. 2002;137:866-74. [PMID: 12458986]

Two articles published in late 2002 compared outcomes of hospitalized patients treated by hospitalists with outcomes of patients treated by other physicians. Both studies sought to determine the effect of hospitalist care on patient outcomes and whether these effects change over time as hospitalists practice longer or whether hospitalist practice patterns revert back to earlier styles. The first study was an observational trial of 5308 patients admitted to a community teaching hospital in San Francisco over a 2-year period who were cared for by 5 hospitalists or several community-based physicians. Patients were followed in both the intensive care unit (ICU) and medical wards. The companion study examined 6511 patients at a university hospital in Chicago over a 2-year period. Patients were randomly assigned to 1 of 2 hospitalists or to a conventional ward service. The hospital used a closed ICU model, so critically ill patients were not studied.

In the San Francisco study, there were no differences in length of stay or cost in the first year, but in the second year, length of stay was 0.61 day shorter (P= 0.02) and costs were $822 lower (P= 0.002). In both years, patients cared for by hospitalists had a lower mortality rate at 30- and 60-day follow-up (adjusted relative hazard, 0.71 [CI, 0.54 to 0.93]). In the Chicago study, average adjusted length of stay for patients on the hospitalist service was 0.29 day shorter in the first year and 0.49 day shorter in the second year (CI, −0.79 to 0.15 day; P= 0.01). A significant reduction in average cost was also seen in the second year ($782 [CI, −$1313 to −$187]; P= 0.01). The authors also observed a difference in mortality, but the difference was statistically significant only in the second year—4.2% for patients being cared for by hospitalists and 6.0% for those being cared for by nonhospitalists (1.8% [CI, −3.6 to −0.1%]; P= 0.04). The California study is limited by substantial baseline differences in the patient groups, making the findings subject to residual confounding. Results of the Chicago study are limited because their findings are derived from the practice of only 2 hospitalists. Both studies were performed at teaching hospitals, and the results may not be generalizable to nonteaching community-based hospitals. Despite these weaknesses, the findings of these 2 papers are similar and underscore that experience with inpatient medicine is key to improving clinical efficiency and potentially reducing mortality in hospitalized patients.

Aviles RJ, Askari AT, Lindahl B, et al.  Troponin T levels in patients with acute coronary syndromes, with or without renal dysfunction. N Engl J Med. 2002;346:2047-52. [PMID: 12087140]

Patients with acute coronary syndromes have a poorer prognosis when troponin T levels are elevated, a finding independent of elevations in levels of creatine kinase–MB (6). As troponin levels may increase even in asymptomatic patients with renal dysfunction, the prognostic role of troponins in patients with acute coronary syndromes who have renal disease remains unclear (8). In this study, Aviles and colleagues used data from 7033 patients enrolled in the Global Use of Strategies To Open Occluded Coronary Arteries IV (GUSTO IV) trial to determine the prognostic value of baseline troponin T level in relation to renal function among patients with non–ST-segment elevation acute coronary syndromes. Acute coronary symdrome was defined as rest angina for at least 5 minutes with either ST-segment depression of at least 0.5 mm or elevated troponin T or I level. Included patients were not planned to have early percutaneous intervention. Patients were randomly assigned to abciximab, a glycoprotein IIb/IIIa inhibitor, or placebo, and the primary end point was a composite of death or myocardial infarction at 30 days.

Troponin T level was measured with a third-generation assay. For the primary analysis, troponin level greater than 0.1 ng/mL was considered abnormal. A secondary analysis used a troponin level cutoff of 0.03 ng/mL. To evaluate the association of troponin T level and 30-day adverse outcomes, patients were divided into quartiles of renal function on the basis of calculated creatinine clearance.

The median creatinine clearance of the patients studied was 76 mL/min, and troponin T levels were greater than 0.1 ng/mL in 52% of patients. In multivariate analysis, baseline elevation of troponin T levels predicted death or myocardial infarction in all quartiles of creatinine clearance (Table 2). This finding persisted when a cutoff level for troponin T of 0.03 ng/mL or 0.1 ng/mL was used. Renal dysfunction itself was associated with an increased risk for adverse outcomes. Compared with patients with more preserved renal function (creatinine clearance > 58 mL/min), patients with creatinine clearance less than 58 mL/min had poorer short-term prognosis, even without elevated troponin levels.

It is important to note that few patients in the study had severe renal dysfunction (only 11 patients had creatinine clearance < 10 mL/min). However, this makes the study results applicable to a large population with less severe renal dysfunction that is commonly encountered by practicing internists. In summary, elevated troponin T level is an accurate predictor of a poor 30-day outcome in patients with a non–ST-segment elevation acute coronary syndrome with and without renal dysfunction. This prognostic information should aid in identifying patients who will benefit from early aggressive interventions.

Wyse DG, Waldo AL, DiMarco JP, et al.  A comparison of rate control and rhythm control in patients with atrial fibrillation. N Engl J Med. 2002;347:1825-33. [PMID: 12466506]

Atrial fibrillation is a common reason for admission to acute care facilities and consumes substantial resources in terms of costs and length of stay (9). Traditionally, the goal of treatment has been to restore sinus rhythm and institute anticoagulation while atrial fibrillation persists. Whether this approach is superior to control of ventricular rate alone, which avoids the risks of antiarrhythmic therapy but may increase the likelihood of embolism, is unknown. In this randomized, multicenter trial, the authors compared the effects of a strategy of rate control with one of rhythm control on death and several secondary adverse outcomes in patients with atrial fibrillation. Eligible patients were older than 65 years of age or were believed by the investigators to be at high risk for stroke or death due to atrial fibrillation. All patients had atrial fibrillation that was likely to recur and require long-term therapy, with no contraindication to anticoagulation. In both groups, the drugs used were chosen by the treating physician but were administered according to the study protocol. All patients were to receive anticoagulation, which could be discontinued in the rhythm-control group after sinus rhythm had been maintained for at least 4 but preferably 12 weeks. Analysis was done by intention-to-treat.

A total of 4060 patients were enrolled. Mean age was 69.7 years, and 39% of patients were female. Mean left ventricular ejection fraction was 0.55, with left atrial enlargement was found on echocardiography in 65% of patients. Approximately one third of patients were enrolled with a first episode of atrial fibrillation. Rate control was achieved with digitalis, calcium-channel blockers, and β-blockers. The most commonly used antiarrhythmic agents were amiodarone and sotalol. At 5-year follow-up, sinus rhythm was present in 35% of the rate-control group and 63% of the rhythm-control group. The actuarial rate of crossover at 5 years was 15% in the rate-control group, primarily because of uncontrolled symptoms of atrial fibrillation or heart failure, and 38% in the rhythm-control group, most often from inability to maintain sinus rhythm or drug intolerance.

At 5 years, there was a trend toward increased mortality in the rhythm-control group (hazard ratio, 1.15 [99% CI, 0.99 to 1.34]; P= 0.08). The rate of the composite end point of death, disabling stroke, anoxic encephalopathy, major bleeding, or cardiac arrest was similar in the 2 groups. Small but statistically significant increases in hospitalization, proarrhythmia, and adverse events leading to discontinuation of drug therapy occurred in the rhythm-control group. Prespecified subgroup analysis showed significant mortality reductions with a rate-control strategy in patients older than 65 years of age and in those without congestive heart failure (CHF). Stroke occurred at an annual rate of approximately 1% in both groups. At the time of stroke, only 43% of patients were in atrial fibrillation, 44% had discontinued anticoagulation, and 28% of those receiving warfarin had a subtherapeutic international normalized ratio.

This study, as well as a smaller study published in the same issue of The New England Journal of Medicine, provides evidence that solely managing heart rate in high-risk patients with atrial fibrillation neither increases mortality nor diminishes quality of life (10). Both studies highlight the importance of effective anticoagulation and suggest that atrial fibrillation is often paroxysmal and asymptomatic. While it may still be reasonable to attempt restoration of sinus rhythm, especially in younger patients or those at lower risk for complications from atrial fibrillation, an initial management strategy of rate control is now supported by the literature.

Maisel AS, Krishnaswamy P, Nowak RM, et al.  Rapid measurement of B-type natriuretic peptide in the emergency diagnosis of heart failure. N Engl J Med. 2002;347:161-7. [PMID: 12124404]

Differentiating CHF from other causes of dyspnea, especially primary pulmonary disorders, is often difficult. B-type natriuretic peptide (BNP) is a cardiac neurohormone secreted by the ventricles in response to increased volume or pressure. Elevated levels of BNP have been shown to correlate with left ventricular dysfunction, New York Heart Association class, and prognosis (11 - 14).

In this multicenter trial, Maisel and colleagues aimed to determine the utility of a rapid assay for BNP level in distinguishing dyspnea due to CHF from other causes of acute dyspnea. In this multicenter trial, BNP levels were measured in 1586 adults presenting to the emergency department with dyspnea as the primary symptom. Patients with acute coronary syndromes, renal failure, or a clear alternative diagnosis were excluded. The actual diagnosis was made retrospectively by 2 cardiologists, blinded to BNP level, on the basis of the Framingham and National Health and Nutrition Examination Survey (NHANES) congestive heart failure scores and review of all pertinent medical records, including results of cardiac testing. Diagnosis was classified as dyspnea due to heart failure, dyspnea not due to cardiac causes in patients with a history of left ventricular dysfunction, and dyspnea not due to heart failure.

The final diagnosis was dyspnea due to CHF in 47% of patients, dyspnea due to noncardiac causes in 5% of patients with known left ventricular dysfunction, and dyspnea not due to CHF in 49%. While median BNP levels were significantly different between each group, there was a large degree of overlap (median BNP levels [±SD], 675 ± 450 pg/mL for dyspnea due to CHF, 346 ± 390 pg/mL for dyspnea not due to CHF in patients with known left ventricular dysfunction, and 110 ± 225 pg/mL for dyspnea due to noncardiac causes). The BNP levels were more accurate than clinical predictors, such as physical examination and Framingham and NHANES congestive heart failure scores, for diagnosing heart failure. A cutoff value of 100 mg/mL provided the highest accuracy, with a sensitivity and specificity of 90% and 76%. Given positive and negative likelihood ratios of 3.75 and 0.13 at this cutoff, it is clear that clinical judgment must still play an important role in the diagnosis of CHF, as interpretation of the BNP result remains dependent on the physician's estimate of pretest probability. For example, with a pretest probability of 50% for CHF, a BNP level greater than 100 pg/mL would increase the likelihood of CHF to 79%, while a level less than 100 pg/mL would reduce the likelihood that dyspnea was due to CHF to 12%. At very low values, BNP level is useful for excluding the diagnosis of acute CHF as the cause of dyspnea, with a negative predictive value for BNP level less than 50 pg/mL of 96%. The authors did not provide specificity data for BNP cutoff levels greater than 150 pg/mL, but it is reasonable to infer from these data that markedly elevated BNP levels do indicate heart failure.

B-type natriuretic peptide is a simple and rapid test that may become a useful aid in situations where the diagnosis of CHF is unclear. However, results from this study do not support the use of BNP alone as a screening test for CHF but rather as an adjunct to clinical decision making. It is important to keep in mind that BNP levels are likely to be elevated in patients with left ventricular dysfunction, even if the cause of dyspnea is not CHF. In these patients, elevated BNP level should not exclude consideration of a primary or coincident pulmonary cause of dyspnea.

Aversano T, Aversano LT, Passamani E, et al.  Thrombolytic therapy vs primary percutaneous coronary intervention for myocardial infarction in patients presenting to hospitals without on-site cardiac surgery: a randomized controlled trial. JAMA. 2002;287:1943-51. [PMID: 11960536]

Studies have consistently shown that patients with acute myocardial infarction have fewer cardiovascular events when they are managed with primary percutaneous coronary intervention (PCI) than with thrombolytic therapy (15). However, these trials have all been conducted in centers with on-site cardiac surgery. Percutaneous coronary intervention is not available to a large number of patients with acute myocardial infarction—almost two thirds of patients with myocardial infarction present to hospitals that do not have this capability (16).

The Atlantic Cardiovascular Patient Outcomes Research Team (C-PORT) trial was designed to determine whether the benefits of PCI over thrombolytics can be extended to patients with myocardial infarction cared for at hospitals without on-site cardiac surgery. Patients older than 18 years of age presenting with acute ST-segment elevation myocardial infarction were randomly assigned to either primary PCI or thrombolytic therapy with accelerated tissue plasminogen activator. Exclusion criteria were lenient, allowing inclusion of patients with comorbid conditions and those who had had coronary interventions. Patients were excluded if they had renal insufficiency and were receiving metformin or were allergic to aspirin or contrast media.

Eleven community hospitals in Massachusetts and Maryland without onsite cardiac surgery or PCI programs participated. At each center a formal PCI program was developed over a period of approximately 3 months that included staff training and institution of quality control and error management measures. Percutaneous coronary intervention operators were required to perform at least 50 interventions yearly. The goal time from hospital arrival to balloon inflation was 90 minutes, and for initiation of thrombolytics the goal time was 30 minutes. The primary outcome was a composite of death, recurrent myocardial infarction, and stroke at 6 months. Secondary outcomes were incidence of blood transfusion, nonprotocol catheterization and PCI, coronary artery bypass graft (CABG) surgery, and length of hospitalization. Analysis was done by intention-to-treat.

The study was stopped after the second interim analysis because of limited funding. This resulted in enrollment of only 451 patients instead of the planned 2550. Each center enrolled between 7 and 138 patients. There were no significant differences between groups at baseline. In the thrombolytic group, 9.7% of patients did not receive the assigned therapy, and 5.8% of patients assigned to PCI did not undergo angiography. Sixty-three percent of PCI patients had stents placed, and 76% were treated with a glycoprotein IIb/IIIa inhibitor. Median door-to-balloon time was 101 minutes, and median door-to-thrombolytic time was 46 minutes.

The composite end point was significantly reduced in the PCI group at 6 weeks (10.7% vs. 17.7%; P= 0.03) and at 6 months (12.4% vs. 19.9%; P= 0.03), primarily because of a statistically significant decrease in recurrent myocardial infarction at 6 months. Thus, for every 13 patients receiving PCI instead of thrombolytics, 1 cardiovascular event was prevented. There were more red blood cell transfusions in the PCI group (12.9% vs. 7.1%; P= 0.04), but at 6 months fewer nonprotocol PCIs (23.6% vs. 49.6%; P< 0.001) and fewer CABG surgeries (13.3% vs 19.5%; P< 0.001) had been done in the PCI group. Length of stay was shorter in the PCI group (4.5 vs. 6 days; P= 0.02). It is important to note that no patient receiving primary PCI required emergency CABG surgery.

This study suggests that centers without on-site cardiac surgery can effectively administer primary PCI. The participating centers in the Atlantic C-PORT trial did have the advantage of a controlled set-up, training program, and in-place quality assurance protocols, which small hospitals may not have the resources to replicate. Because the study was small and did not meet its enrollment goal, the data are not adequate to shift health policies. However, the findings certainly add to the evidence that primary PCI is the preferred strategy for management of acute myocardial infarction.