*I5LUE

R

1

R

Wrddiptftcopadi

*ER

*MBONE

R

1

2

AIC

M(ad

wcwpsmsIbv

fnffeil

bfcaebiwf

icat

SFP*

*DCL0IE

702 Correspondence JACC Vol. 55, No. 7, 2010February 16, 2010:700–3

International Centre for Circulatory Healthmperial College and St. Mary’s Hospital9-61 North Wharf Roadondon W2 1LAnited Kingdom-mail: claire.raphael@gmail.com

doi:10.1016/j.jacc.2009.09.044

EFERENCE

. Goldstein N, Bradley E, Zeidman J, Mehta D, Morrison RS. Barriersto conversations about deactivation of implantable defibrillators inseriously ill patients: results of a nationwide survey comparing cardiol-ogy specialists to primary care physicians. J Am Coll Cardiol 2009;54:371–3.

eply

e thank Dr. Raphael and colleagues for their interest in oureport (1). The investigators report their own findings, whichemonstrate that patients with implantable cardioverter-efibrillators have relatively poor understanding of their device ast relates to options for deactivation. Their findings confirm ourrevious qualitative work with patients as well (2). The investiga-ors point out that if physicians think patients understand optionsor deactivation, but in reality they do not, then this can be a reasonhat discussions about deactivation occur so rarely. We are inomplete agreement with these investigators and did mention inur original contribution that this apparent incongruity betweenhysician perception of patients’ knowledge and what patients’ctually understand poses a challenge to communication abouteactivation. We thank Dr. Raphael and colleagues for highlight-ng this important issue in their letter.

Nathan Goldstein, MDlizabeth Bradley, PhD. Sean Morrison, MD

Department of Geriatricsount Sinai School of Medicine

ox 1070ne Gustave Levy Placeew York, New York 10029-mail: nathan.goldstein@mssm.edu

doi:10.1016/j.jacc.2009.11.028

EFERENCES

. Goldstein NE, Bradley E, Zeidman J, Mehta D, Morrison RS. Barriersto conversations about deactivation of implantable defibrillators inseriously ill patients: results of a nationwide survey comparing cardiol-ogy specialists to primary care physicians. J Am Coll Cardiol 2009;54:371–3.

. Goldstein NE, Mehta D, Siddiqui S, et al. “That’s like an act of suicide”patients’ attitudes toward deactivation of implantable defibrillators.J Gen Intern Med 2008;23 Suppl 1:7–12.

Closer Look atncidental Findings onardiac Computed Tomography

acHaalany et al. (1) recently reported that incidental findingsIFs) on computed tomography performed to diagnose coronaryrtery disease (CAD) are common but do not predict noncardiaceath, and investigating them further “is not without cost or risk.”

Although we appreciate the detailed analysis of IFs and costs,e believe a significant flaw affects the study design. Drawing

onclusions on whether mortality differs between patients with andithout IFs becomes statistically unsound if some patients receiveotentially lifesaving (or at least life-prolonging) interventions,uch as lobectomy for lung carcinoma or chemotherapy forediastinal lymphoma. In other words, although Kaplan-Meier

urvival curves show no difference in survival between patients withFs and those without, any intervention that may prolong survivaleyond the reported follow-up time significantly impairs thealidity of the analysis.

Moreover, the investigators recognize that an 18-monthollow-up time may be inadequate to correctly evaluate indetermi-ate IFs, as some of them may become significant with a longerollow-up time. We concur, but we also add that an 18-monthollow-up time is probably inadequate even for the prognosticvaluation of the smaller number of IFs that were already signif-cant, as they include disorders with a natural course that may beonger than 18 months (2).

Further studies are certainly necessary to clarify whether anyenefit lies in further investigating indeterminate findings, butrom the point of view of evidence-based medicine, that questionannot be answered by comparing a group of patients with IFs andnother group without. A more appropriate study design wouldxclude patients who already have a clear management pathway setefore them (i.e., those without any IFs and those with anmmediately significant IF) and randomize the remaining patientsith indeterminate IFs to either further investigations or simple

ollow-up.As the number of procedures increases, invasive cardiologists

ncreasingly will be called to acquire sufficient preparation toonsider the global significance of imaging findings, and weppreciate the relevance of the work of MacHaalany et al. (1) inhat direction.

tefano Bartoletti, MSrancesco Perna, MDasquale Santangeli, MD

Michela Casella, MD, PhD

Institute of Cardiologyepartment of Cardiovascular Medicineatholic University of the Sacred Heartargo Gemelli 80168 Rometaly-mail: michelacasella@hotmail.com

doi:10.1016/j.jacc.2009.11.025

R

1

2

R

Ws

uibuoAtit

fpeTmwpao

niHcmucj

AJ*

*4OCE

PFAHsfI

R

1

703JACC Vol. 55, No. 7, 2010 CorrespondenceFebruary 16, 2010:700–3

EFERENCES

. MacHaalany J, Yam Y, Ruddy TD, et al. Potential clinical andeconomic consequences of noncardiac incidental findings on cardiaccomputed tomography. J Am Coll Cardiol 2009;54:1533–41.

. Henschke CI, Yankelevitz DF, Libby DM, Pasmantier MW, Smith JP,Miettinen OS, for the International Early Lung Cancer Action Pro-gram Investigators. Survival of patients with stage I lung cancer detectedon CT screening. N Engl J Med 2006;355:1763–71.

eply

e thank Drs. Bartoletti and colleagues for their interest in ourtudy (1).

If one could operate in a world without risk and that isnconstrained by costs, all incidental findings (IFs) could benvestigated. However, the investigation of IFs must be temperedy the reality that benefits may be offset by risks and costs. In thisnblinded observational study, mortality was chosen as the primaryutcome measure because it is least influenced by subjectivity.lthough we agree that when interventions prevent or delay death

he mortality in 2 groups will likely favor equivalence, it is alsomportant to recognize that survival benefits may also be attributedo lead-time bias.

We also agree that a longer follow-up duration may have betteracilitated appreciation of differences in outcomes between com-arison groups, and this limitation has been acknowledged. How-ver, none of the indeterminate IFs became clinically significant.he majority of the patients were followed up until a diagnosis wasade or until no further follow-up was recommended (i.e., the IFas deemed benign). Even if inferior outcomes were noted in IFatients, further studies would be needed to explain whether thessociation is causal or serendipitous, given the benign disposition

f IFs.

Most important is the readers’ observation that more studies areeeded and that a randomized controlled trial of patients with

ndeterminate IF would be ideal and should be encouraged.owever, one must accept that randomizing patients with IFs

ould be seen as unethical and may not be clinically feasible atany centers. Thus, it is important that results such as ours be

sed to encourage discussion and to cast doubt on our currentlinical practice, thus opening the opportunity for researchers toustify such randomized controlled trials.

run Abraham, MBBSimmy MacHaalany, MDBenjamin J. W. Chow, MD

University of Ottawa Heart Institute0 Ruskin Streetttawa, Ontario K1Y 4W7anada-mail: bchow@ottawaheart.ca

doi:10.1016/j.jacc.2009.11.024

lease note: Dr. Abraham was supported by the Heather and Whit Tucker Researchellowship in Cardiology. Dr. Chow is supported by the CIHR New Investigatorward MSH-83718, and receives research and fellowship training support from GEealthcare, research support from Pfizer Inc. and AstraZeneca, and educational

upport from TeraRecon Inc. Dr. Chow’s research is supported in part by the Imagingor Cardiovascular Therapeutics Project RE02-038 and the Canada Foundation fornnovation 11966.

EFERENCE

. MacHaalany J, Yam Y, Ruddy TD, et al. Potential clinical andeconomic consequences of noncardiac incidental findings on cardiac

computed tomography. J Am Coll Cardiol 2009;54:1533–41.