[Show abstract][Hide abstract] ABSTRACT: The PROSPECT trial reported no single echocardiographic measurement of dyssynchrony is recommended to improve patient selection for cardiac resynchronization therapy (CRT).
In 100 consecutive patients who received CRT, we analyzed 27 ECG and echocardiographic variables to predict a positive response to CRT defined as a left ventricular (LV) end systolic volume decrease of ≥ 15% after CRT.
Right ventricular (RV) pacing-induced left bundle branch block (LBBB), time difference between LV ejection measured by tissue Doppler and pulsed wave Doppler (T(TDI-PW)), and wall motion score index (WMSI) were significantly associated with positive CRT response by multivariate regression. We assigned 1 point for RV pacing-induced LBBB, 1 point for WMSI ≤ 1.59, and 2 points for T(TDI-PW) > 50 ms. Overall mean response score was 1.79 ±1.39. Cutoff point for response score to predict positive response to CRT was > 2 by receiver operating characteristic (ROC) analysis. Area under ROC curve was 0.97 (p = 0.0001). Cardiac resynchronization therapy responders in patients with response score > 2 and ≤ 2 were 36/38 (95%) and 7/62 (11%, p < 0.001), respectively. After age and gender adjustment, the response score was related to CRT response (OR = 45.4, p < 0.0001).
A response score generated from clinical, ECG and echocardiographic variables may be a useful predictor for CRT response. However, this needs to be validated.
Archives of medical science : AMS. 08/2011; 7(4):627-33.
[Show abstract][Hide abstract] ABSTRACT: We investigated in 136 consecutive patients with heart failure receiving cardiac resynchronization therapy (CRT) the effect of carvedilol versus metoprolol CR/XL versus no beta blocker on mortality. Of the 136 patients, 42 (31%) were on carvedilol, 80 (59%) were on metoprolol CR/XL, and 14 (10%) were not on a beta blocker. A decrease of left ventricular end-systolic volume ≥15% after CRT was defined as a positive response to CRT. Of the 136 patients, 62 (46%) responded to CRT. It was found that both carvedilol and metoprolol CR/XL were not related to CRT response on using Cox univariate regression analysis. Twenty-two of the 136 patients (16%) died during follow-up of 17 ± 10 months after initiating CRT. Mortality occurred in 14 of 80 patients (18%) on metoprolol CR/XL, in 3 of 42 patients (7%) on carvedilol, and in 5 of 14 patients (36%) not on beta blockers (P = 0.04). After adjustment for age, gender, and the variables with significant differences by Cox univariate regression, both carvedilol (hazard ratio = 0.14; P = 0.03; 95% confidence interval = 0.02-0.86) and metoprolol CR/XL (hazard ratio = 0.19; P = 0.02; 95% confidence interval = 0.04-0.80) were found to be related to mortality by Cox multivariate regression.
American journal of therapeutics 06/2011; · 1.29 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: We hypothesized a patient selection score (PSS) may improve patient selection for cardiac resynchronization therapy (CRT).
Of 136 patients who received CRT, group A included 100 study patients and group B 36 patients for validation. A positive response to CRT was a left ventricular (LV) end-systolic volume decrease of ≥15% and survival from heart failure at end of follow-up.
Of 100 group A patients, 37 (37%) were CRT responders during 14-month follow-up. A 7-point PSS was generated based on six variables. The cutoff point for PSS to predict a positive response to CRT was >4 by receiver operating characteristic (ROC) analysis. The area under the ROC curve (AUC) for PSS to predict CRT response was 0.94 (P = 0.0001). CRT responders in patients with a PSS > 4 and ≤4 were 33/40 (83%), and 4/60 (7%), respectively (P < 0.001). Multivariate Cox proportional regression analysis showed that PSS was related to CRT response (hazard ratio = 10.3, P < 0.0001). The CRT response rate in patients with a PSS > 4 in Group B was also significantly higher compared to a PSS ≤ 4 (88% vs. 16%, P < 0.001). The AUC for PSS to predict a CRT response in Group B was 0.91 (P = 0.0001).
Patients with a PSS >4 are the most likely to respond to CRT. Using this score system, a PSS score >4 can predict the probability of a CRT response up to 88% in patients with heart failure and a wide QRS duration.
[Show abstract][Hide abstract] ABSTRACT: The combination of pulsed wave (PW) and tissue Doppler imaging (TDI) has been proposed as a new method to assess left ventricular (LV) mechanical dyssynchrony (LVMD), but results have not been validated. We investigated the correlation of a combination of PW and TDI with a positive response to cardiac resynchronization therapy (CRT).
We studied 108 consecutive patients who received CRT. Patients with atrial fibrillation were excluded. The time difference (T(PW-TDI)) between onset of QRS to the end of LV ejection by PW (T(PW)) and onset of QRS to the end of the systolic wave in LV basal segments with greatest delay by TDI (T(TDI)) was measured before CRT and during short-term and long-term follow-up.
The T(PW-TDI) interval before CRT was 74 ±48 ms. Intra-observer variabilities for T(PW) and T(TDI) were 1.5 ±0.24% and 1 ±0.17%. Inter-observer variabilities for T(PW) and T(TDI) were 1 ±0.36% and 1 ±0.64%, respectively. T(PW-TDI) > 50 ms was defined as the cutoff value for diagnosis of LVMD by receiver operating curve (ROC) analysis. During follow-up of 15 ±11 months, the sensitivity and specificity of TP(PW-TDI) to predict a positive response to CRT were 98% and 82%, respectively. The area under the ROC curve was 0.92. There was a significant agreement between LVMD determined by T(PW-TDI) and the positive response to CRT (κ=0.80).
Left vertricular dyssynchrony detected by the method combining PW and TDI demonstrated a high reproducibility, sensitivity, specificity and agreement with a positive response to CRT.
Archives of medical science : AMS. 08/2010; 6(4):519-25.
[Show abstract][Hide abstract] ABSTRACT: Background. Due to underrepresentation of patients with chronic kidney disease (CKD) in large Implantable-Cardioverter Defibrillator (ICD) clinical trials, the impact of ICD remains uncertain in this population. Methods. Consecutive patients who received ICD at Creighton university medical center between years 2000-2004 were included in a retrospective cohort after excluding those on maintenance dialysis. Based on baseline Glomerular filtration rate (GFR), patients were classified as severe CKD: GFR < 30 mL/min; moderate CKD: GFR: 30-59 mL/min; and mild or no CKD: GFR ≥ 60 mL/min. The impact of GFR on appropriate shocks and survival was assessed using Kaplan-Meier method and Generalized Linear Models (GLM) with log-link function. Results. There were 509 patients with a mean follow-up of 3.0 + 1.3 years. Mortality risk was inversely proportional to the estimated GFR: 2 fold higher risk with GFR between 30-59 mL/min and 5 fold higher risk with GFR < 30 mL/min. One hundred and seventy-seven patients received appropriate shock(s); appropriate shock-free survival was lower in patients with severe CKD (GFR < 30) compared to mild or no CKD group (2.8 versus 4.2 yrs). Conclusion. Even moderate renal dysfunction increases all cause mortality in CKD patients with ICD. Severe but not moderate CKD is an independent predictor for time to first appropriate shock.
Cardiology research and practice. 01/2010; 2010:150285.
[Show abstract][Hide abstract] ABSTRACT: Lead perforation is a major complication of cardiac rhythm management devices (CRMD), occurring in about 1%. While most lead perforations occur early, numerous instances of delayed lead perforation (occurring >30 days after implantation) have been reported in the last few years. Only about 40 such cases have been published, with the majority occurring <1 year after implantation. Herein, we describe the case of an 84-year-old female who presented with recurrent syncope and was diagnosed to have delayed pacemaker lead perforation 4.8 years after implantation. Through this report, we intend to highlight the increasing use of CRMD in elderly patients, and the lifelong risk of complications with these devices. Presentation can be atypical and a high index of suspicion is necessary for diagnosis.
[Show abstract][Hide abstract] ABSTRACT: Background: We hypothesized that segmental wall motion abnormalities (WMAs) are related to cardiac resynchronization therapy (CRT) response. Methods: We studied 108 patients who received CRT, 69 with ischemic and 39 with nonischemic heart disease. A wall motion score index (WMSI) was analyzed using a 17-segment model and calculated by the total score/number of segments analyzed. A decrease of left ventricular end systolic volume ≥15% after CRT was defined as a positive response to CRT. Results: Of 108 patients, 1,054/1,836 segments (57%) had WMAs. The mean WMSI was 2.06 in patients with ischemic heart disease and 1.04 in patients with nonischemic heart disease (P < 0.0001). The area under the receiver operating characteristic curve for a WMSI predicting a positive response to CRT was 0.70 (P = 0.0001). The cutoff point was a WMSI ≤2 for prediction of a positive response to CRT. After adjustment for age, gender, and clinical features, the WMSI persistently related to CRT responders (P = 0.01). During 15-month follow-up, the percentage of CRT nonresponders in patients with a WMSI >2 was significantly higher (82%) compared to patients with a WMSI ≤2 (47%, P = 0.005) and nonischemic heart disease (36%, P < 0.001). In 59 patients with left ventricular mechanical dyssynchrony, the percentage of negative responders to CRT in patients with a WMSI >2, ≤2, and nonischemic heart disease were 53% (8 of 15), 16% (3 of 19) and 0% (0 of 25), respectively (P < 0.001). Conclusions: A large extent of WMAs and a WMSI >2 predicted a poorer CRT response. (ECHOCARDIOGRAPHY, Volume 26, November 2009)
[Show abstract][Hide abstract] ABSTRACT: We studied 99 consecutive patients with class III-IV systolic heart failure with a left ventricular ejection fraction (LVEF) < or =35% and a QRS duration <120 milliseconds. Patients with cardiac resynchronization therapy were excluded. Echocardiography was performed in all patients before and after optimal standard heart failure therapy. The septal-to-posterior wall motion delay (SPWMD) > or =130 milliseconds on echocardiogram was defined as left ventricular mechanical dyssynchrony (LVMD). Sixty-nine of 99 patients (70%) had ischemic heart disease. During follow-up of 15.2 +/- 9.8 months, LVEF improvement > or =15% was greater patients in nonischemic group (50%, 15/30) than in ischemic group (9%, 6/69; P < 0.001). After adjustment for age, gender, and clinical and echocardiographic characteristics, ischemic heart disease and grade of coronary disease were persistently related to LVEF improvement > or =15% (P = 0.03 and 0.02, respectively). Twenty of 99 patients (20%) had SPWMD > or =130 milliseconds (LVMD group), and 79 of 99 patients (80%) had SPWMD <130 milliseconds (non-LVMD group). LVEF increased in both groups (P = 0.005) during follow-up, but the percentage of patients with LVEF improvement > or =15% in LVMD was greater compared with patients without LVMD (40% versus 16%, respectively, P = 0.03). In conclusion, the improvement of LVEF in patients with systolic heart failure and narrow QRS was greater in patients with nonischemic heart disease and LVMD compared with patients with ischemic heart disease and absence of LVMD during medical therapy without cardiac resynchronization therapy.
American journal of therapeutics 02/2009; 17(1):e1-7. · 1.29 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Background: Amiodarone is an effective antiarrhythmic agent with a substantial risk of toxicity. Monitoring is recommended with the following minimum: 2 clinic visits per year including thyroid function tests (TFTs) and liver function tests (LFTs), annual eye exam (AEE) and chest x-ray (CXR).
PURPOSE: The objective of this study was to evaluate adherence to the minimum monitoring standards for amiodarone in two cohorts of patients (one treated between 1988-1992 and the other 2000-2004). Adherence to monitoring was correlated with adverse drug reactions and outcome of therapy.
METHODS: Consecutive patients initiated on amiodarone during 1988-1992 and 2000-2004 were followed prospectively. Only patients remaining on amiodarone for ≥ 12 mos were included. Patients with a minimum of 2 clinic visits having TFTs and LFTs, AEE and CXR were considered to be adherent to monitoring. Outcomes were categorized as (1) discontinued for lack of efficacy (DC-LOE); (2) discontinued due to SE (DC-SE); (3) continued with SE (C-SE); and (4) continued without SE (C-NoSE).
RESULTS: 577 patients initiated on amiodarone during 1988-1992 and 553 during 2000-2004 were included. Mean follow-up was 22 months. Outcomes for all 1130 patients include: DC-LOE 18%; DC-AE 38%; C-NoAE 22%; and C-AE 22%. The proportion of patients compliant with follow-up was significantly greater in 1988-1992 (67%) compared to 2000-2004 (55%; p < 0.05). Patients with the outcome of DC-AE were significantly more likely to be non-compliant with follow-up in both time periods (p < 0.05). The most common types of side effects in the DC-AE groups were pulmonary, hyperthyroidism, neurologic (tremor), and hepatic.
Conclusion: Patients compliant with follow-up are less likely to discontinue amiodarone due to side effects than non-compliant patients. It is hypothesized that compliant patients have side effects that can be managed clinically which reduces the severity of side effects and reduces the number requiring drug discontinuation.
2008 American College of Clinical Pharmacy Spring Practice and Research Forum; 04/2008
[Show abstract][Hide abstract] ABSTRACT: Delayed intracardiac lead perforation has been defined as migration and perforation after one month of implantation. It is a rare complication; pathophysiology and optimal management are currently unclear. Recognition of these cases becomes important with increasing use of these devices. We describe such a case of delayed lead perforation.
Pacing and Clinical Electrophysiology 02/2008; 31(1):10-2. · 1.75 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: This study investigated the circadian, daily, and seasonal distributions of ventricular arrhythmias in patients with new implantable cardioverter-defibrillator placement at Creighton University Medical Center from January 2000 to December 2004. The incidence and distribution of ventricular tachyarrhythmias as recorded by implantable cardioverter-defibrillators were analyzed with respect to season, month, day of the week, and average daily temperature. Data from 154 consecutive patients (mean age 67 +/- 14 years; 78% men, 71% with ischemic heart disease [IHD], mean left ventricular ejection fraction 34 +/- 15%) were analyzed. During a mean follow-up of 35 +/- 19 months, a total of 1,055 episodes of spontaneously terminated ventricular tachycardia (VT) and 612 episodes of VT or ventricular fibrillation with appropriate device therapy occurred. Distributions in the incidence of VT and VT or ventricular fibrillation receiving appropriate therapy were similar in patients with IHD and non-IHD. Spontaneously terminated VT and appropriately treated VT or ventricular fibrillation episodes occurred with the greatest incidence in the winter months and the lowest incidence in summer, spring, and fall. A linear regression between the number of episodes and the average daily temperature showed a greater likelihood of the 2 events occurring on cooler days, irrespective of the cause of cardiac disease. A weekly distribution was also observed, with the greatest proportion of episodes occurring on Fridays and the lowest on Saturdays and Sundays. A bimodal circadian distribution was present, with the greatest peak occurring from 8 a.m. to 1 p.m. and a smaller peak occurring from 5 p.m. to 10 p.m. In conclusion, the occurrence of ventricular tachyarrhythmias appears to follow circadian, daily, and seasonal distributions that are similar in patients with IHD and non-IHD. The incidence inversely correlates with average daily temperatures.
The American Journal of Cardiology 11/2007; 100(7):1134-8. · 3.21 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Warfarin anticoagulation significantly reduces the risk of thromboembolism in patients with atrial fibrillation (AF). However, there are many patients with AF who begin anticoagulation only after left atrial thrombus (LAT) is detected by transesophageal echocardiography (TEE). The impact of anticoagulation in these patients has not been clearly described. The purpose of this study was to investigate the incidence of cerebrovascular accident (CVA) among AF patients who began warfarin before LAT was detected by TEE compared to those who began warfarin only after TEE demonstrated LAT and those did not receive warfarin at any point.
Of the 90 consecutive AF patients with LAT (male 48, female 42, age 71.5 +/- 10.1 years), 49 began warfarin more than 3 weeks before TEE (Group I); 29 began warfarin after TEE (Group II); and 12 did not receive warfarin at all (Group III).
The incidence of CVA in Group I (14%, 7/49, prior CVA 5, new CVA after TEE 2) was significantly lower than Group II (45%, 13/29, prior CVA 10, new CVA after TEE 3, P = 0.006) and III (42%, 5/12, prior CVA 3, new CVA after TEE 2, P = 0.047). Patients with persistent LAT had significantly higher incidence (64% vs 23%, P = 0.024) of CVA and lower CVA free survival than those with resolved LAT.
The incidence of CVA among AF patients, who began warfarin before LAT detection, is significantly lower than those who began warfarin after LAT detection as well as those who did not receive warfarin at all.
[Show abstract][Hide abstract] ABSTRACT: Epidemiologic studies suggest that inhibition of renin-angiotensin system with angiotensin-converting enzyme inhibitors and angiotensin receptor blockers may prevent development of atrial fibrillation (AF).
The objective of the study was to assess if there is significant indication for using angiotensin-converting enzyme inhibitors and angiotensin receptor blockers in the prevention of new-onset AF and to identify the target patient population.
PubMed and Cochrane clinical trials database were searched from 1980 through March 2005 together with the review of citations. Nine randomized controlled human trials reporting the prevention of new-onset AF by inhibition of renin-angiotensin system were identified. Information about study design, follow-up, intervention, population, outcomes, and methodology quality was extracted.
The mean follow-up of the studies ranged from 6 months to 6.1 year. The pooled estimate using random effects model was 0.82 (95% CI 0.70-0.97) for prevention of new-onset AF and 0.61 (95% CI 0.46-0.83) for primary prevention of AF. The angiotensin-converting enzyme inhibitors (0.75, 95% CI 0.57-0.99) had greater protective effect than angiotensin receptor blockers (0.81, 95% CI 0.62-1.06). Patients with heart failure benefited the most (0.57, 95% CI 0.37-0.89). The test for heterogeneity between studies was significant. There was no consistent visual or statistical evidence of publication bias.
The use of angiotensin-converting enzyme inhibitors and angiotensin receptor blockers had an overall effect of 18% risk reduction in new-onset AF across the trials and 43% risk reduction in patients with heart failure.
American heart journal 09/2006; 152(2):217-22. · 4.65 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Patients who have pacemakers and sinus node dysfunction frequently have atrial fibrillation (AF). The need for continued pacemaker therapy after conversion to permanent AF remains uncertain. This study showed that, among 174 patients who received pacemaker implantation for sinus node dysfunction, 38% (n = 62) had the minimum intrinsic ventricular rate of >60 beats/min after conversion to AF. The pacemaker memory showed that 30 patients (18%) never used ventricular pacing during permanent AF. The study results suggest that patients who have a stable intrinsic ventricular rate during permanent AF by serial assessment may no longer need continued pacemaker therapy.
The American Journal of Cardiology 10/2005; 96(7):942-5. · 3.21 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Infection is a devastating complication of permanent pacemakers (PMs) implantable cardioverter defibrillators (ICDs). Many implanting physicians commonly use povidone-iodine solution to irrigate the device pocket before implanting the device. We sought to assess if such a measure would alter the rate of infection.
A total of 2,564 consecutive patients who received implantable PM or ICD devices between 1994 and 2002 were studied. Povidone-iodine was used for pocket irrigation in 53% and saline in 47%. A total of 18 (0.7%) patients developed pocket infections with 0.7% (10/1,359) in povidone-iodine (group I) and 0.6% (8/1,205) in saline (group II) pocket irrigation (p = ns). Groups I and II were studied for various clinical and demographic variables described in the results section.
There was no statistical difference in the baseline demographic and clinical characteristics between groups I and II, respectively. ICDs were most frequently infected than PMs (56% vs 44%). Most (83%) of the devices were dual chamber. Reopening of the pocket for either lead or generator replacement had a higher incidence of infection than new implants (61% vs 39%). There was no difference in the use of preimplantation antibiotic prophylaxis. Late (61%) and deep pocket infections (78%) were more common than early (39%) and superficial infections (22%). Blood cultures were positive in 67% and Staphylococcus aureus was the common most pathogen (50%). The mean duration of antibiotics use after the diagnosis of device infection was 35 +/- 23 days with 72% requiring device explantation. The device was reimplanted on the contralateral side in 56% cases. One patient in each group died due to device infection and related complications. No significant allergy to iodine was seen in either group.
Povidone-iodine irrigation of the subcutaneous pocket did not alter the rates of pocket infection after pacemaker/defibrillator implantation.
Pacing and Clinical Electrophysiology 09/2005; 28(8):789-94. · 1.75 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The AFFIRM investigators have recommended rate control as the preferred strategy for recurrent atrial fibrillation (AF), but the appropriate strategy for new-onset persistent AF is uncertain. Our study evaluated the AF recurrence rate and the impact of rhythm-control drugs (class 1A, 1C, and 3 antiarrhythmic drugs) on patients with new-onset persistent AF after successful direct-current (DC) cardioversion. Consecutive patients who underwent DC cardioversion of AF from January 1, 1996 to December 31, 1999 were screened for new-onset persistent AF, and 150 patients met the inclusion criteria. After the first DC cardioversion, 50 patients received rhythm-control drugs (rhythm-control group) and the other 100 did not (rate-control group). The 2 groups had similar clinical characteristics except for a lower ejection fraction (44 +/- 14% vs 49 +/- 14%, p <0.01) and a higher proportion of idiopathic dilated cardiomyopathy (20% vs 1%, p = 0.03) in the rhythm-control group versus the rate-control group. During the follow-up period there was a trend toward a lower rate of early AF recurrence at 24 hours after DC cardioversion in the rhythm-control group versus the rate-control group (6% vs 16%, p = 0.11), but there was a high recurrence rate of AF in both groups at 1 month (30% for the rhythm-control group vs 41% for the rate-control group, p = 0.25). At the end of the follow-up period, rhythm-control therapy was abandoned in 78% of the rhythm-control group patients after the failure of 1 to 3 rhythm-control drugs. In the rate-control group, rhythm-control therapy was attempted after AF recurrence in 62 patients but was later abandoned in 48 patients (77%) because of treatment failure. Therefore, the high incidence of treatment failure with rhythm-control therapy suggests that rate control with anticoagulation should be preferred in patients with new-onset persistent AF if AF recurs after DC cardioversion.
The American Journal of Cardiology 02/2004; 93(1):45-8. · 3.21 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: BACKGROUND: Atrial fibrillation (AF) is a recurrent problem that frequently requires repeat cardioversion. Transesophageal echocardiography (TEE) is indicated before cardioversion in patients who are underanticoagulated (warfarin therapy <3 weeks or international normalized ratio [INR] <2.0). It remains uncertain if TEE should be repeated in underanticoagulated patients who had no atrial thrombi detected by previous TEE. Methods and results From January 1996 to June 2001, 76 patients (43 men, 33 women; mean age, 68.8 +/- 10.4 years) who were underanticoagulated and had no atrial thrombi in previous TEE underwent repeat TEE before cardioversion of recurrent AF. The duration of recurrent AF at the time of the second TEE was 5.1 +/- 9.3 months (1 day to 4 years). The underlying diseases included coronary artery disease (n = 30), hypertension (n = 22), valvular heart diseases (n = 8), dilated cardiomyopathy (n = 4), hypertrophic cardiomyopathy (n = 2), and others (n = 10). Eight (10.5%) patients (2 men, 6 women; mean age, 68.6 +/- 6.6 years) were found to have intra-atrial thrombi on the second TEE. Of these 8 patients, 3 had coronary artery disease, 1 had hypertension, 2 had dilated cardiomyopathy, 1 had hypertrophic cardiomyopathy, and 1 had AF of unknown cause. The duration of recurrent AF in patients with and without thrombi was not significantly different (3.6 +/- 4.7 versus 5.3 +/- 9.7 months, P =.22). Of the 8 patients with intra-atrial thrombi on the second TEE, 5 had been taking warfarin for 3 to 4 weeks but had subtherapeutic INR and 3 were taking aspirin only. Compared with patients without intra-atrial thrombi, patients with intra-atrial thrombi had lower ejection fraction (32.5% +/- 18.1% versus 49.9% +/- 14.1%, P =.015), slower left atrial appendage empty velocity (0.22 +/- 0.08 versus 0.41 +/- 0.17 m/s, P <.01), and higher prevalence of spontaneous echo contrast (87.5%) than in patients without intra-atrial thrombi (19.1%, P <.05) but similar left atrial size (49.5 +/- 5.3 versus 47.3 +/- 7.1 mm, P =.15). Cardioversion was cancelled in all patients with atrial thrombi. CONCLUSIONS: In underanticoagulated patients, repeat TEE is necessary before cardioversion of recurrent AF even if the previous TEE showed no atrial thrombi.
American heart journal 10/2003; 146(4):741-5. · 4.65 Impact Factor