Ejection fraction and QRS width as predictors of event rates in patients with implantable cardioverter defibrillators.

Objectives: The Multicenter Automatic Defibrillator Implantation Trial II in 2002 recommended implantable cardioverter defibrillators (ICDs) prophylactically for all patients with a prior myocardial infarction and ejection fraction (EF) of 30% or less. In June of 2003, the Centers for Medicare and Medicaid Services approved reimbursement for ICD placement in patients with an EF of 30% or less who have a QRS interval greater than 120 ms. The purpose of this study was therefore to determine the value of QRS duration in predicting the occurrence of ventricular arrhythmias within the first year after ICD implantation. These ICDs were placed over the past 5 years for various indications.

Methods: EF cut points of 30% or less and 31% or greater and QRS duration of 120 ms or less and 121 ms or greater were used to assess the risk of events.

Results: There was a significant increase in events in subjects with EF of 30% or less, compared with patients with EF of 31% or greater (P < 0.05), and there was a trend toward increased likelihood of arrhythmias in patients with widened QRS width.

Conclusions: This study confirms the conclusion of the Multicenter Automatic Defibrillator Implantation Trial II and implies that the Centers for Medicare and Medicaid Services criteria for reimbursement may not be scientifically valid.

Key Words: ejection fraction, implantable cardioverter defibrillators, QRS width

**********

Cardiovascular disease is the most common cause of death in the United States. (1-3) Sudden cardiac death (SCD) accounts for approximately 63% of all cardiovascular death. (1-3) More than 460,000 sudden cardiac deaths occur annually in the United States, (4) of which heart failure accounts for approximately 280,000. (5) Ventricular tachyarrhythmias, including ventricular tachycardia (VT) and ventricular fibrillation (VF), are responsible for 75 to 80% of cases. (6-8) The other events are due to bradycardia, asystole, or electromechanical dissociation. (9,10) Of the 1,100,000 myocardial infarctions that occur in 1 year in the United States, 5 to 6% have heart failure with an ejection fraction (EF) of less than 30%. (11) Patients with reduced left ventricular function secondary to ischemia have been shown to be at increased risk for sudden cardiac death from ventricular arrhythmias. Implantable cardioverter defibrillators (ICDs) have been shown to improve survival when compared with antiarrhythmic drugs in patients with inducible and nonsuppressible tachyarrhythmias, based on the Multicenter Automatic Defibrillator Implantation Trial (MADIT) I from 1996. (12) The MADIT II trial in 2002 recommended ICDs prophylactically for all patients with a prior myocardial infarction and decreased EF. (13) They reported a mortality rate of 14.2% in the defibrillator group compared with 19.8% in the conventional group. The MADIT II recommendation for ICD placement was a primary intervention without any electrophysiology studies. As a result of this recommendation, a great deal of emphasis has been placed on the use of ICDs for patients with decreased EF secondary to prior myocardial infarction. In July of 2002, the US Food and Drug Administration approved the ICD therapy for patients with EF of less than 30% who had an acute myocardial infarction 30 days or more prior, based on the MADIT II trial. (14) In June of 2003, the Centers for Medicare and Medicaid Services (CMS) approved reimbursement for ICD placement in patients with an EF of 30% or less and a QRS interval greater than 120 ms. (15)

Although the decision regarding EF was backed by the evidence from the MADIT II trial, there was no evidence for the QRS duration except for a subanalysis from the MADIT II trial. There was concern among the electrophysiology community that the criteria for QRS duration may not be scientifically sound. (16) Therefore, the aim of this study was to determine the association of QRS duration and ventricular arrhythmias in the first year after ICD implantation in patients who had ICDs placed over the past 5 years for various indications. We also aimed to determine the value of QRS duration in predicting ventricular arrhythmias in the first year as well as the correlation between EF and QRS duration.

Materials and Methods

Study patients

All patients who had ICDs placed over the past 5 years for various indications were included in the study. Patients with incomplete data and those who had ICDs placed for a period of less than 1 year and did not have any events were excluded from the analysis. In this study, event was defined as any ventricular arrhythmia recorded on the ICD. Ventricular arrhythmias were categorized into sustained VT, VF, and nonsustained ventricular tachycardia (NSVT). VT was defined as three or more consecutive, ventricular ectopic beats at a rate of more than 120 beats per minute. VT was considered sustained if it lasted for 30 seconds or more or resulted in shock delivery from the ICD. VF was defined as a fibrillatory state of the heart without coordinated contractions of the ventricle.

Variables for the study included the patient's age, sex, and race. EF and the method by which it was measured were recorded. Preimplantation PR and QRS durations and the rhythm were recorded for each eligible subject. Their date of ICD placement, time to first event, and events per year were also recorded. Patients were divided into two groups, based on having events in the first year (group 1) and not having any events in the first year (group 2). The indications for the study included ischemic cardiomyopathy (CM) with VT (51%), CM/NSVT with inducible VT (26.3%), and other indications (16.7%). Other indications included aborted SCD, long QT syndrome with family history of SCD, hypertrophic obstructive cardiomyopathy with family history of SCD, arrhythmogenic right ventricular dysplasia, recurrent syncope, bradycardia, and refractory paroxysmal atrial flutter.

Study protocol

All the patients have been followed regularly after ICD placement. Ninety-two percent of patients were stable, and 6% were lost to follow-up because of change of residence or transfer to a different electrophysiologist associated with the nursing home. Two percent of patients died, whereas 0.5% received cardiac transplantation. The cause of death has not been recorded. Preimplantation rhythm and PR and QRS intervals were recorded with the use of conventional 12-lead electrocardiography. The proportion of patients with sinus rhythm and atrial fibrillation was 83% and 5%, respectively. Only 6% and 1% of the patients had paced and junctional rhythm, respectively. A QRS width of 120 ms or more was used to define wide QRS width. EF was recorded by using catheterization in 63.6%, echocardiography in 32.8%, and nuclear imaging in 3.5%. It was obtained by computer analysis during echocardiography and catheterization. An EF of 30% or less was regarded as decreased EF. We reviewed the charts and recorded all the events they had recorded on their ICDs. The first event recorded was NSVT in 36.4%, sustained VT in 33.3%, and VF in 8.6%. Twenty-two percent did not have any events. Thirteen percent of patients were noted to have accompanying paroxysmal atrial fibrillation. The mean duration to first event was 6.4 months and the average number of events per year was 19.2.

Statistical analysis

Statistical analysis was performed with the use of SPSS Version 11.5. (17) An event versus no event was used as the dependent variable in the univariate analysis to determine clinical differences between subjects in group 1 and group 2. [chi square] testing was used to determine group differences for dichotomous variables, including race and sex, whereas the independent t test was used for continuous variables, including QRS duration, PR interval, age, and EF. Multivariate analysis, using logistic regression, was used to assess the odds of events adjusting for possible confounding variables including age, sex, and race. In the analysis of the relation between QRS and events, EF was considered a confounding variable and hence was statistically adjusted and vice versa. A P value of less than 0.05 was used to determine statistical significance.

Results

Of a total of 230 eligible patients, 32 were excluded. The basic characteristics of the eligible subjects in the two groups are shown in the Table 1. The population included 76% whites and 22% blacks, as reported by the patients. Race was unknown in 2% of the subjects. Seventy-nine percent of eligible subjects were males. Except for EF, there was no statistically significant difference between the two groups with respect to age, race, sex, or PR interval (Table 1).

Table 2 shows the distribution of patients in the two groups with QRS width of 120 ms or less and 121 ms or more and with EF of 30% or less and those with 31% or more. There is a statistically higher number of events in patients with EF of 30% or less compared with subjects with EF of 31% or more (P < 0.01), whereas the relation between patients with QRS of 121 ms or more and those with QRS of 120 ms or less did not have any statistical significance.

In the univariate analysis (Table 3), the two groups differed from each other only with regard to EF. Patients with EF of 30% or less had increased odds of events (P = 0.01). In the multivariate analysis (Table 4), after adjusting for other variables, the two groups continued to be different with regard to EF, and patients with a QRS width 121 ms or more showed a trend toward increased odds of events but did not attain statistical significance.

We used Pearson correlation analysis to determine the degree of linear relation between EF and QRS duration. Increase in QRS width was associated with a progressive decrease in EF (P < 0.01) (Figure). The association remained statistically significant after adjusting for age, race, and sex.

Finally, we excluded all subjects (n = 33) who had indications other than ischemia for placement of an ICD and studied the remaining 165 patients who represented the MADIT II population. A similar degree of association between dependent variables and independent variables was observed. For EF of 30% or less, the odds ratio was 0.35 (95% confidence interval, 0.17 to 0.71) and the odds ratio for QRS of 121 ms or more was 1.41 (95% confidence interval, 0.69 to 2.88).

[FIGURE OMITTED]

We finally combined widened QRS duration and decreased ejection fraction and saw that the addition of widened QRS duration did not improve the predictability of having arrhythmias.

Discussion

Mirowski et al (18) first used ICDs in humans in 1980. This electronic device is used to continuously monitor the heart and identifies malignant ventricular tachyarrhythmias. It then delivers electrical countershock or antitachycardia pacing to restore normal rhythm. All ICDs currently available will also render antibradycardia pacing. Implantable defibrillators were initially used as secondary prevention for patients with aborted cardiac arrest or documented life-threatening VT in the 1980s and early 1990s. In 1996, Moss et al (12) reported the results of the MADIT I trial and recommended ICDs in all patients with nonsustained VT and inducible nonsuppressible VT at electrophysiologic study. The authors reported a 16% total mortality rate in the defibrillator group compared with 39% in the conventional therapy group and a 54% reduction in all-cause mortality rate. (12) The Multicenter Unsustained Tachycardia Trial confirmed MADIT findings. (19) Bigger (20) in 1997 reported no improved survival with prophylactic placement of an ICD at the time of coronary bypass surgery. The AVID trial in 1997 concluded that among survivors of VF and VT, the ICD is superior to antiarrhythmic drugs for increased overall survival. (21) Connolly et al (22) and Kuck et al (23) in 1999 reported similar findings. All the above studies supported prior recommendations for secondary prevention of SCD.

In 2002, the MADIT II trial showed improved survival in patients with ischemic CM with an EF of 30% or less. The age of the prior myocardial infarction was greater than 30 days. ICD placement was performed without prior electrophysiology testing. This was a primary indication for ICD usage. (13) In July 2002, the US Food and Drug Administration approved the MADIT II indication for ICD therapy. (14) Recent CMS criteria for reimbursement of ICD placement include QRS duration of 121 ms or more in addition to EF of 30% or less to prevent ventricular arrhythmias and SCD. (15) The QRS duration was deduced by a subanalysis from the MADIT II trial.

In our retrospective study, QRS duration of 121 ms or more was used to designate wide QRS duration, similar to the CMS criteria. We found a trend toward increased likelihood of having an arrhythmia in the subjects who had QRS duration of 121 ms or more. An EF of 30% or less was used as the cutoff for decreased EF. There was a highly significant difference between EF between the two groups. The results were similar in the analysis using all the study patients (n = 198) and only the patients with history of ischemia and cardiomyopathy (n = 165). The second study population (n = 165) was the population studied in the MADIT II trial and also applied for the CMS guidelines. Interestingly, there was also a significant increase in the QRS duration with every onepoint decrease in EF. This relation raises the important question if we are looking at the same parameter in two different variables.

The purpose of our study was to assess if the CMS decision regarding using ICDs only in patients with QRS duration of 121 ms or more had any scientific basis to it. Our study confirmed previous studies, which concluded that patients with low EF are at a high risk for having inducible arrhythmias and have a higher chance of arrhythmias when compared with people with normal cardiac function. Whether analyzing all patients or just the population studied in other studies, there was a highly significant increase in ventricular arrhythmias in patients with low EF. There was no statistically significant increase in arrhythmias in patients with widened QRS duration. We also selected patients with combined widened QRS duration and decreased EF to see if the CMS criteria added to the MADIT II recommendations, and we found that adding widened QRS duration to the low EF does not improve the predictability of having an increased number of arrhythmias.

Finally, our findings confirm the findings of MADIT II, suggesting that the criteria of widened QRS duration necessary for implantation of an ICD, cited by the CMS, is thus far not proven. QRS duration of 121 ms or more was associated with a trend toward increased likelihood of having an arrhythmia, and EF of 30% or less was associated with significantly increased odds of having arrhythmias. Combining widened QRS duration to decreased EF does not improve the predictability.

Some limitations for this study are noteworthy. Our study had a small sample size and was a retrospective study. Whether a bigger study that is prospective and randomized, controlled, and double-blinded with appropriate sample size will produce any different results needs to be determined.

Only the dead have seen the end of war.
--Plato

Table 1. Basic descriptive statistics of the study population

                     Significant arrhythmias in first 2 years
                        Yes                    No
Variables            (n = 133)              (n = 65)                 P

Demographics
Age (yr)              65.83 [+ or -] 12.44   65.82 [+ or -] 11.83  0.992
Race
  White (%)           77.1                   77.8                  0.916
  Black (%)           22.9                   22.2
Sex
  Male (%)            78.2                   81.5                  0.586
  Female (%)          21.8                   18.5
Determinants
  Ejection fraction   29.73 [+ or -] 12.74   35.71 [+ or -] 14.43  0.005
  QRS interval       115.93 [+ or -] 29.44  110.18 [+ or -] 29.99  0.202
    (ms)
  PR interval (ms)   170.88 [+ or -] 40.93  166.26 [+ or -] 34.58  0.447

Study population includes patients who had implantable cardioverter
defibrillators implanted for all reasons (n = 198). Numbers signify
mean [+ or -] SD. P < 0.05 considered statistically significant.

Table 2. Distribution of QRS and ejection fraction by prevalence of
arrhythmias in the first year

                                             Arrhythmias in
                                             first 2 years
                                              Yes        No
Variables                                   (n = 133)  (n = 65)    P

QRS duration  [less than or equal to] 120   63%        37%         0.107
                                    > 120   74.3%      25.7%
Ejection       [less than or equal to] 30   75.2%      24.8%       0.009
  fraction                            >30   57.1%      42.9%

Table 3. Univariate analysis between ventricular arrhythmias and other
variables (a)

Variable       Odds ratio     95% CI        P

EF             2.28           1.22-4.26     0.01
QRS width      1.70           0.89-3.24     0.11
Age (yr)       1.00           0.98-1.03     0.99
Race           1.04           0.51-2.14     0.92
Sex            1.23           0.58-2.61     0.59

(a) CI, confidence interval; EF, ejection fraction.

Table 4. Multivariate analysis between arrhythmias and EF and QRS
width (a)

Variable              Odds ratio   95% CI        P

Adjusted for EF
  EF                  0.40         0.21-0.77     0.01
  Age (yr)            1.00         0.97-1.02     0.81
  Race                0.88         0.39-1.98     0.75
Adjusted for QRS
  QRS width           1.65         0.86-3.19     0.13
  Age (yr)            1.00         0.97-1.02     0.93
  Race                0.98         0.46-2.10     0.96
  Sex                 1.32         0.59-2.97     0.50

(a) CI, Confidence interval; EF, ejection fraction.

Acknowledgments

The authors would like to acknowledge Dr. Ike Okosun for his assistance with statistics and Dr. John Hudson for review of the manuscript.

Accepted October 4, 2004.

References

1. Myerburg RJ, Kessler KM, Castellanos A. Sudden cardiac death: epidemiology, transient risk, and intervention assessment. Ann Intern Med 1993;119:1187-1197.

2. Myerburg RJ, Interian A Jr, Mitrani RM, et al. Frequency of sudden cardiac death and profiles of risk. Am J Cardiol 1997;80:10F-19F.

3. Huikuri HV. Castellanos A. Myerburg RJ. Sudden death due to cardiac arrhythmias. N Engl J Med 2001;345:1473-1482.

4. Centers for Disease Control. State specific mortality from sudden cardiac death-United States, 1999. MMWR 2002;51:123-126.

5. American Heart Association. 2002 Heart and Stroke Statistical Update. Dallas, Tex, American Heart Association, 2001.

6. Zipes DP, Wellens HJJ. Sudden cardiac death. Circulation 1998;98:2334-2351.

7. Wit AL, Janse MJ. Experimental models of ventricular tachycardia and fibrillation caused by ischemia and infarction. Circulation 1992;85(Suppl 1):1-32-1-42.

8. Mehta D, Curwin J, Gomes JA, et al. Sudden death in coronary artery disease: acute ischemia versus myocardial substrate. Circulation 1997;96:3215-3223.

9. Epstein AE, Carlson MD, Fogoros RN, et al. Classification of death in antiarrhythmia trials. J Am Coll Cardiol 1996;27:433-442.

10. Pratt CM, Greenway PS, Schoenfeld MH, et al. Exploration of the precision of classifying sudden cardiac death: implications for the interpretation of clinical trials. Circulation 1996;93:519-524.

11. Huikuri HV, Castellanos A, Myerburg RJ. Sudden death due to cardiac arrhythmias. N Engl J Med 2002;345:1473-1482.

12. Moss AJ, Hall WJ, Cannom DS, et al. Improved survival with an implanted defibrillator in patients with coronary disease at high risk for ventricular arrhythmia. N Engl J Med 1996;335:1933-1940.

13. Moss AJ, Zareba W, Hall WJ, et al. Prophylactic implantation of a defibrillator in patients with myocardial infarction and reduced ejection fraction. N Engl J Med 2002;346:877-883.

14. Letter from Daniel G. Schultz, FDA, to Guidant Corporation. July 18, 2002. Available at http:www.fda.gov/cder/approval/index.htm. Accessed May 2004.

15. Decision memo from Phurrough S. Farrell J, Chin J to Administrative file CAG: 00157N. Implantable Cardioverter Defibrillators. Available at http://www.cms.hhs.gov/media/press/release.asp?counter=807. Accessed March 29, 2005.

16. NASPE/Heart Rhythm Society concerned with CMS ruling limiting ICDs for select patients. Washington Advocacy, June 09, 2003. Available at http://www.hrsonline.org/swpressfiles/press97689173.asp. Accessed March 29, 2005.

17. Norusis MJ. SPSS 11.5 for windows released 11.5.0 SPSS Inc. Chicago, 2002.

18. Mirowski M, Reid PR, Mower MM, et al. Termination of malignant ventricular arrhythmias with an implanted automatic defibrillator in human beings. N Engl J Med 1980;303:322-324.

19. Buxton AE. Lee KL, Fisher JD, et al. A randomized study of the prevention of sudden death in patients with coronary artery disease. N Engl J Med 1999;341:1882-1890.

20. Bigger JT. Prophylactic use of implanted cardiac defibrillators in patients at high risk for ventricular arrhythmias after coronary artery bypass graft surgery. N Engl J Med 1997;337:1569-1575.

21. AVID Investigators. A comparison of antiarrhythmic drug therapy with implantable defibrillators in patients resuscitated from near-fatal ventricular arrhythmias. N Engl J Med 1997;337:1576-1583.

22. Connolly SJ, Gent M, Roberts RS, et al. A randomized trial of the implantable cardioverter-defibrillator against amiodarone. Circulation 2000;101:1297-1302.

23. Kuck KH, Cappato R, Siebels J, et al. Randomized comparison of antiarrhythmic drug therapy with implantable defibrillators in patients resuscitated from cardiac arrest. Circulation 2000;102:748-754.

RELATED ARTICLE: Key Points

* Widened QRS duration does not predict the likelihood of ventricular arrhythmias.

* Decreased ejection fraction predicts increases in ventricular arrhythmias.

* Combining widened QRS duration and decreased ejection fraction does not improve the predictability of ventricular arrhythmias.

* Increasing QRS width is associated with a progressive decrease in ejection fraction.

Mahi Lakshmi Ashwath, MD, and Felix O. Sogade, MD, FACC 

From Mercer University School of Medicine, Macon, GA.

The authors have no financial interest in any procedure or device mentioned in this article.

Reprint requests to Dr. Mahi Lakshmi Ashwath, 11477 Mayfield Road, Apartment 318, Cleveland, OH 44106. Email: [e-mail]