0
Editorial |

Slow and Study: Support for a Randomized Trial of β-Blockade in Sepsis-Associated Atrial Fibrillation FREE TO VIEW

Matthew W. Semler, MD; Arthur P. Wheeler, MD, FCCP
Author and Funding Information

FINANCIAL/NONFINANCIAL DISCLOSURES: None declared.

CORRESPONDENCE TO: Matthew W. Semler, MD, Pulmonary and Critical Care Medicine, Vanderbilt University Medical Center, 1161 21st Ave S., T-2220 MCN, Nashville, TN 37232-2650


Copyright 2016, American College of Chest Physicians. All Rights Reserved.


Chest. 2016;149(1):9-10. doi:10.1016/j.chest.2015.08.020
Text Size: A A A
Published online

Among the 1 million patients hospitalized each year for sepsis in the United States, nearly one in four will experience atrial fibrillation (AF). Despite the use of β-blockers, calcium channel blockers, digoxin, and amiodarone for decades in the treatment of sepsis-associated AF, the effect of medication choice on outcomes has never been studied.

In this issue of CHEST (see page 74), Walkey et al examined this important question using an administrative database capturing 20% of all discharges from nonfederal US hospitals from 2010 to 2013. The authors identified patients receiving intravenous treatment for sepsis-associated AF via: (1) International Classification of Diseases, Ninth Revision, Clinical Modification codes for sepsis and AF and (2) pharmacy billing records for administration of antibiotics and intravenous calcium channel blockers, β-blockers, digoxin, or amiodarone. Among > 500,000 patients with a first hospitalization for sepsis, 20% had AF, 35% of whom were treated with intravenous therapy. Initial treatment was most commonly a calcium channel blocker (36%) followed by a β-blocker (28%), digoxin (20%), and amiodarone (16%). Choice of therapy was linked to patient characteristics (eg, higher rates of digoxin and amiodarone use in patients receiving vasopressors) but also inexplicably to geographic region and hospital characteristics unrelated to patient condition or prognosis. For example, there was a strong preference for β-blockers over calcium channel blockers in the northeast region of the country, whereas in hospitals in western areas, the preference was reversed by an equal magnitude. After propensity-matching, in-hospital mortality was lower with β-blockers than with calcium channel blockers (risk ratio, 0.92; 95% CI, 0.86-0.97), digoxin (risk ratio, 0.79; 95% CI, 0.75-0.85), or amiodarone (risk ratio, 0.64; 95% CI, 0.61-0.69).

The benefit associated with β-blockers persisted in the subgroup of patients receiving vasopressors and when potential for unmeasured confounding by indication was addressed (through the leveraging of variation in practice patterns among physicians, hospitals, and regions). Sensitivity analyses including patients who received > 1 therapy or excluding patients with hypertension as a potential indication for β-blocker therapy were supportive of the authors’ primary finding: Among patients receiving intravenous therapy for AF in the context of sepsis, β-blocker use was associated with lower hospital mortality.

This report shares limitations inherent to retrospective studies of administrative data. The “explicit diagnosis code” method used by the authors to identify patients with sepsis achieves near 100% specificity and positive predictive value at the cost of sensitivity as low as 10%. Thus, as many as 90% of patients with sepsis in the administrative database may have been excluded from the current analysis. Using the more common methods of Angus et al or Martin et al might have captured a broader sample including less severe sepsis cases, at some cost of a decrease in positive predictive value. Identifying AF via International Classification of Diseases, Ninth Revision, Clinical Modification code 427.31 is well validated in other populations, although the reassurance of internal validation in a subset of this population was unavailable. Use of “present on admission” data may have missed comorbidities that would have been identified by incorporating earlier claims data and may have misclassified some preexisting AF as new onset. Despite the lack of granularity in the sequence of events and specific indications for treatment, it is a reasonable fundamental assumption that when an inpatient identified as having sepsis and AF receives intravenous β-blockers, calcium channel blockers, digoxin, or amiodarone, it is for the treatment of AF rather than for another indication.

The central limitation that Walkey et al strive to address is confounding by indication bias. A clinician’s decision to prescribe a drug to a specific patient may involve factors linked to that patient’s outcome (eg, patients who are in shock are more likely to die and also more likely to receive amiodarone for their AF). The inclusion of all available covariates to perform 1:1 propensity-matching is a robust approach to limiting the influence of indication bias. However, residual confounding by unmeasured covariates may remain, particularly in the comparison of β-blockers vs digoxin and amiodarone, in which the differences in indication, patient characteristics before matching, and risk ratios after adjustment are most dramatic. According to the authors, the variations in choice of sepsis-associated AF treatment are also influenced by factors independent of the individual patient. The finding that a preference for β-blockers at the hospital level is associated with lower mortality after adjusting for patient characteristics is a strong argument against residual confounding and supports the conclusion that β-blockers reduce mortality in sepsis-associated AF.

The idea that β-blockers might improve outcomes for patients with sepsis-associated AF is biologically plausible. AF occurring during sepsis is mediated by high levels of endogenous and exogenous catecholamines in addition to acute inflammation, changes in atrial volume, and myocardial ischemia. Excessive sympathetic outflow causes myocardial damage, immunosuppression, and insulin resistance. β-blockers are efficacious for preventing and treating AF after cardiac surgery and are the recommended first-line AF treatment in other hyperadrenergic states such as hyperthyroidism. Similar to therapy for congestive heart failure, early attempts to augment the physiologic response to sepsis by increasing β-adrenergic tone worsened outcomes., Moreover, emerging data suggest that opposition of adrenergic outflow may protect patients who have septic shock. Although routine β-blockade in septic shock remains premature, results of the study by Walkey et al should sway clinicians away from calcium channel blockers and toward β-blockers for the patient with sepsis who requires intravenous AF treatment.

The study authors concluded that a randomized clinical trial comparing the effectiveness of AF treatments during sepsis is warranted. We agree but go further. The imperative to scrupulously test common interventions for common conditions, including those “grandfathered” into practice, applies not just to the management of sepsis-associated AF but also to the type and volume of intravenous fluid we administer, how inpatients are bathed, and every other “precedent-based” intervention we take for granted. The promise of comparative effectiveness research for a condition as common as sepsis-associated AF is that a difference in hospital mortality as small as 1% would mean thousands of lives saved every year.

References

Walkey A.J. .Hammill B.G. .Curtis L.H. .Benjamin E.J. . Long-term outcomes following development of new-onset atrial fibrillation during sepsis. Chest. 2014;146:1187-1195 [PubMed]journal. [CrossRef] [PubMed]
 
Walkey A.J. .Evans S.R. .Winter M.R. .Benjamin E.J. . Practice patterns and outcomes of treatments for atrial fibrillation during sepsis: a propensity-matched cohort study. Chest. 2016;149:74-83 [PubMed]journal
 
Iwashyna T.J. .Odden A. .Rohde J. .et al Identifying patients with severe sepsis using administrative claims: patient-level validation of the Angus implementation of the international consensus conference definition of severe sepsis. Med Care. 2014;52:e39-e43 [PubMed]journal. [CrossRef] [PubMed]
 
Angus D.C. .Linde-Zwirble W.T. .Lidicker J. .Clermont G. .Carcillo J. .Pinsky M.R. . Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit. Care Med. 2001;29:1303-1310 [PubMed]journal. [CrossRef] [PubMed]
 
Martin G.S. .Mannino D.M. .Eaton S. .Moss M. . The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med. 2003;348:1546-1554 [PubMed]journal. [CrossRef] [PubMed]
 
Connolly S.J. .Cybulsky I. .Lamy A. . Beta-Blocker Length Of Stay (BLOS) Studyet al Double-blind, placebo-controlled, randomized trial of prophylactic metoprolol for reduction of hospital length of stay after heart surgery: the Beta-Blocker Length Of Stay (BLOS) study. Am Heart J. 2003;145:226-232 [PubMed]journal. [PubMed]
 
January C.T. .Wann L.S. .Alpert J.S. . and the ACC/AHA Task Force Memberset al 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines and the Heart Rhythm Society. Circulation. 2014;130:2071-2104 [PubMed]journal. [CrossRef] [PubMed]
 
Hayes M.A. .Timmins A.C. .Yau E.H. .Palazzo M. .Hinds C.J. .Watson D. . Elevation of systemic oxygen delivery in the treatment of critically ill patients. N Engl J Med. 1994;330:1717-1722 [PubMed]journal. [CrossRef] [PubMed]
 
Gates S. .Perkins G.D. .Lamb S.E. .et al Beta-Agonist Lung injury TrIal-2 (BALTI-2): a multicentre, randomised, double-blind, placebo-controlled trial and economic evaluation of intravenous infusion of salbutamol versus placebo in patients with acute respiratory distress syndrome. Health Technol Assess Winch Engl. 2013;17:v-vi [PubMed]journal
 
Morelli A. .Ertmer C. .Westphal M. .et al Effect of heart rate control with esmolol on hemodynamic and clinical outcomes in patients with septic shock: a randomized clinical trial. JAMA. 2013;310:1683-1691 [PubMed]journal. [CrossRef] [PubMed]
 
Yunos N.M. .Bellomo R. .Hegarty C. .Story D. .Ho L. .Bailey M. . Association between a chloride-liberal vs chloride-restrictive intravenous fluid administration strategy and kidney injury in critically ill adults. JAMA. 2012;308:1566-1572 [PubMed]journal. [CrossRef] [PubMed]
 
Noto M.J. .Domenico H.J. .Byrne D.W. .et al Chlorhexidine bathing and health care-associated infections: a randomized clinical trial. JAMA. 2015;313:369-378 [PubMed]journal. [CrossRef] [PubMed]
 

Figures

Tables

References

Walkey A.J. .Hammill B.G. .Curtis L.H. .Benjamin E.J. . Long-term outcomes following development of new-onset atrial fibrillation during sepsis. Chest. 2014;146:1187-1195 [PubMed]journal. [CrossRef] [PubMed]
 
Walkey A.J. .Evans S.R. .Winter M.R. .Benjamin E.J. . Practice patterns and outcomes of treatments for atrial fibrillation during sepsis: a propensity-matched cohort study. Chest. 2016;149:74-83 [PubMed]journal
 
Iwashyna T.J. .Odden A. .Rohde J. .et al Identifying patients with severe sepsis using administrative claims: patient-level validation of the Angus implementation of the international consensus conference definition of severe sepsis. Med Care. 2014;52:e39-e43 [PubMed]journal. [CrossRef] [PubMed]
 
Angus D.C. .Linde-Zwirble W.T. .Lidicker J. .Clermont G. .Carcillo J. .Pinsky M.R. . Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit. Care Med. 2001;29:1303-1310 [PubMed]journal. [CrossRef] [PubMed]
 
Martin G.S. .Mannino D.M. .Eaton S. .Moss M. . The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med. 2003;348:1546-1554 [PubMed]journal. [CrossRef] [PubMed]
 
Connolly S.J. .Cybulsky I. .Lamy A. . Beta-Blocker Length Of Stay (BLOS) Studyet al Double-blind, placebo-controlled, randomized trial of prophylactic metoprolol for reduction of hospital length of stay after heart surgery: the Beta-Blocker Length Of Stay (BLOS) study. Am Heart J. 2003;145:226-232 [PubMed]journal. [PubMed]
 
January C.T. .Wann L.S. .Alpert J.S. . and the ACC/AHA Task Force Memberset al 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines and the Heart Rhythm Society. Circulation. 2014;130:2071-2104 [PubMed]journal. [CrossRef] [PubMed]
 
Hayes M.A. .Timmins A.C. .Yau E.H. .Palazzo M. .Hinds C.J. .Watson D. . Elevation of systemic oxygen delivery in the treatment of critically ill patients. N Engl J Med. 1994;330:1717-1722 [PubMed]journal. [CrossRef] [PubMed]
 
Gates S. .Perkins G.D. .Lamb S.E. .et al Beta-Agonist Lung injury TrIal-2 (BALTI-2): a multicentre, randomised, double-blind, placebo-controlled trial and economic evaluation of intravenous infusion of salbutamol versus placebo in patients with acute respiratory distress syndrome. Health Technol Assess Winch Engl. 2013;17:v-vi [PubMed]journal
 
Morelli A. .Ertmer C. .Westphal M. .et al Effect of heart rate control with esmolol on hemodynamic and clinical outcomes in patients with septic shock: a randomized clinical trial. JAMA. 2013;310:1683-1691 [PubMed]journal. [CrossRef] [PubMed]
 
Yunos N.M. .Bellomo R. .Hegarty C. .Story D. .Ho L. .Bailey M. . Association between a chloride-liberal vs chloride-restrictive intravenous fluid administration strategy and kidney injury in critically ill adults. JAMA. 2012;308:1566-1572 [PubMed]journal. [CrossRef] [PubMed]
 
Noto M.J. .Domenico H.J. .Byrne D.W. .et al Chlorhexidine bathing and health care-associated infections: a randomized clinical trial. JAMA. 2015;313:369-378 [PubMed]journal. [CrossRef] [PubMed]
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
PubMed Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543