Routine Unselected Access to Day Case Electrical Cardioversion of Persistent Atrial Fibrillation: Anticoagulant preparation is the key factor

SMJ 2003 49(1): 26-28

Jonathan Wilshire¹ , House Physician

Alison Smith¹ , Cardiac Research Nurse

Annegret Böhm¹ , Cardiology Senior House Officer

Robert J. MacFadyen^{1, 2} , Consultant Cardiologist

 ¹ Cardiac Unit (7^th Floor), Raigmore Hospital, Inverness, IV2 3UJ, UK

² Departments of Cardiology and University Department of Medicine,

City Hospital, Dudley Road, Birmingham B18 7QH, UK

Address for correspondence:

Dr. Robert J. MacFadyen , Department of Cardiology, , City Hospital NHS Trust, , Dudley Road, , Birmingham B18 7QU, United Kingdom

E-mail:  robert.macfadyen@swbh.nhs.uk

Web :   http://www.cityhospital.org.uk/teaching

 Disclaimers: NONE                  Grant in aid: NONE

Abstract

Objectives: To determine the routine efficacy of a day case cardioversion system operating in a remote geographical area with an unselected case referral base and a proportion of direct GP access

Design: Prospective analysis of all admitted cases in a 13 week period

Participants: 47 consecutive patients admitted for 49 planned episodes of electrical cardioversion.

Measurements: Pre procedural investigations and preparation, immediate and three-month outcome of rhythm following ECV

Results: We found a predominant use by the cardiac unit despite working within a general medical service unit suggesting low case selection from non-cardiac sources. We suspected and confirmed a high rate of point of care treatment cancellation. Poor management of anticoagulation was the dominant reason for canceling planned treatment. Procedural preparation in terms of anti arrhythmic drug therapy and investigations seemed well preserved.

Conclusion: New strategies for initiating and sustaining adequate outpatient warfarin therapy are needed to allow such systems to operate efficiently.

Key words: Atrial fibrillation, Routine electrical cardioversion, Case turnover, cancellations; poor warfarin titration; rural dependence; effectiveness and efficacy

Background

Atrial flutter/fibrillation is the most common arrhythmia in the elderly; it is often asymptomatic and has a wide range of aetiology¹. The main clinical risk is cardio embolism, which is decreased but not eliminated by whole body anticoagulation in conjunction with rate control. However symptomatic limitation is more likely with this strategy². Alternatively achieving a return to maintained sinus rhythm has its advocates³. There are benefits in exercise capacity but not in mortality from this approach. Electrical cardioversion following whole body anticoagulation is the simplest and safest means to achieve sinus rhythm and is widely applied⁴. However effective case selection and preparation is essential and there is a surfeit of reports that do not reflect real life clinical practice. The “routine” can be impaired by misconceptions over roles and responsibilities among clinicians, practical application and post procedural management which can lead to under utilisation, procedural delay, a reduced rate of successful treatment and inappropriate or repeated use of the procedure where there is little chance of success.

In the UK, where the majority of routine cardiac care is conducted by general physicians and GP’s⁵, the use of DC cardioversion systems can be affected by unfamiliarity with the strategy and/or technique⁶. Firstly in our institution we were concerned that the pattern of referral for DC cardioversion was not optimal in terms of case selection and preparation leading to poor success rates or high early recurrence of AF. Secondly our rate of cancellation on the day of cardioversion was felt to be too high. We therefore chose to prospectively audit these factors. The findings should be typical for a district hospital service albeit in this case one serving a geographically remote region where complete supervision by the hospital service is problematic.

Methodology

The population studied were all patients referred to the cardioversion clinic from any source with ECG confirmation of persistent atrial flutter/fibrillation during the over a 13-week period audit. No cardiology led pre-selection occurs in this system with a treatment session every Wednesday run by the cardiology junior staff from CCU or ward with help from staff grade or consultant cardiologist. All patients were lightly sedated for the procedure and anaesthetised with propofol while the attending anaesthetist controlled ventilation. All cases had routine antero-posterior pad placement for escalating synchronised DC counter shock.

Data was collected from the attending patients’ case notes on the afternoon of cardioversion using an audit pro-forma. The demography of the sample, their drug treatment and aetiological factors in success on cardioversion were studied and ECG outcomes defined at three months. The aetiology and duration of arrhythmia, previous attempts at cardioversion, and pre treatment investigations e.g. Echocardiography reports (left atrial size, mitral valve function and LV mass) were screened to assess quality of pre procedural care. Follow up clinical and ECG assessment was arranged at 3 months for all patients.

Results

Demography

The direct access system for cardioversion of AF provided only three referrals from general practitioners. Cardiology staff reviewed these requests but the patients were not seen in person prior to admission for treatment. 47 treatments were booked in 12 sessions during the 13-week period. One treatment session was lost due to the unavailability of anaesthetic cover. The number of patients attending on any day varied from none (cancelled session) to five with 12 of the potential 13 treatment sessions during the audit period used. From the total admissions group (mean age 63.2yr (SD, ±10.1) range 41-82) there were 33 males, (mean age 62.0 (SD ±11.1); mean BMI 28.7 (SD ±3.3)), and 14 females, mean age 66.1 (SD ±6.9); mean BMI 28.9 (SD ±10.6). Of the 47 treatments booked, 45 presented to hospital for day case treatment (two lost due to patient initiated cancellation) in the 13-week period of audit. Of the treatments eleven patients attended twice and one patient attended three times within the audit period. Of the repeat treatments, four were due to failed anaesthetic cover from one session, eight due to inadequate anticoagulation (one patient cancelled twice) only one repeated due to treatment failure.

Of the total procedures, 22 planned treatments (47%) were cancelled. Two patients attended already in sinus rhythm. Two patients failed to attend for personal reasons, one was cancelled due to failure to take amiodarone as intended, two had arterial BP measurements felt to be “too high” by the anaesthetist, one was cancelled due to hypokalaemia. The predominant reason for cancellation was where ten treatments (22%) were cancelled due to inadequate anticoagulation with INR values <1.5 on attendance.

Rhythm, Aetiology and Investigation and Pre treatment Strategies

Two male patients had cardioverted and presented in sinus rhythm. These individuals had a structurally abnormal hearts with both having confirmed echocardiographic LVH and MR in the context of chronic hypertension but only mild left atrial enlargement (>4cm, <5cm). Both patients had been pre-treated with a class III drug in the form of Sotalol (1) or Amiodarone (and beta blockade)(1). Five patients (3 female) had atrial flutter rather than atrial fibrillation and all four who attended achieved cardioversion despite two patients having no anti arrhythmic pre-treatment. The patient with atrial flutter who cancelled due to personal circumstances had alcohol as the most probable aetiology of his arrhythmia.

The medical pre treatment strategies of the cardioversion attempts are given in Table 1. Combination therapy was common but four patients (two patients with atrial flutter) had no pre-treatment anti arrhythmic drug therapy. Two patients were pre-treated with flecainide and three patients received digoxin monotherapy. Four patients received Atenolol and three patients had Sotalol monotherapy (one had two attempts) pre-treatment.

Amiodarone pre-treatment was the dominant medical therapy either as monotherapy (5) or in combination with beta blockade (6) and (3) /or digoxin (8). The concomitant use of calcium channel blocking drugs (CCBD), while not strictly agents that qualify as assisting in cardioversion, was common. Six cases were treated with CCBD monotherapy, three cases had a combination of amiodarone and CCBD, and

Initial and 3 month Cardioversion Efficacy

Of the twenty treatments, nine treated patients failed electrical cardioversion at maximal energy (55% success rate). Eight of those nine had echocardiography with normal hearts found in three studies. Three had one previous failed attempted DC cardioversion with two having two previous attempts. Two of these failures had no pre treatment drug therapy (one of whom had been instructed to take Sotalol but this had not been provided) but none had had previous cardioversion attempts. Five of the group were hypertensive, two IHD and one had alcohol abuse in addition to IHD. Four patients were pre treated with amiodarone, three in combination with a beta-blocker and one in combination with digoxin. One of the patients treated at two sessions failed on both occasions to restore SR.

Eleven cases had successful treatment at initial discharge. One of these had been cancelled on a previous occasion due to low INR within the study period. Two of these patients had no anti arrhythmic drug therapy including the patient cancelled from previous treatment. Of these eleven successfully cardioverted patients ten remained in SR at three month follow up. The one patient who quickly reverted to atrial fibrillation was an obese normotensive 49-year-old man with a structurally normal heart treated only with digoxin monotherapy.
Discussion

As in most UK hospitals, patients selected for attempted conversion to SR are admitted for day case cardioversion⁷. While there is little doubt that recent trial results show increasingly marginal benefits for strategies aimed at restoring sinus rhythm⁸, policies for elective DC cardioversion will remain relevant for sub groups of patients who remain significantly symptomatic. While the optimal strategy is dependent on the pre treatment risk of embolism and stroke, this is generally to administer cardioversion alone (i.e. without prior anti arrhythmic drug therapy) followed by cardioversion repeated with amiodarone pre treatment, should the initial procedure fail or if AF recurs^9,10. While performing day case electrical cardioversion is obviously more cost effective than hospitalisation, poor utilisation (delays and or deferred treatments) remains inappropriate and it may be complicated by various practical failures as we saw in our sample. While there is some interest in completing procedures under sedation without an anaesthetist in the UK¹¹ and this is widely used in the USA¹², most units still provide full anaesthetic cover to provide titration¹³ of any one of a range of suitable agents¹⁴.

System failures can be due to administrative or procedural problems as well as the details of individual medical case management. Cancellations or repeated admissions through ineffective preparation cause delay and will reduce the likelihood of successful restoration of sinus rhythm which is clearly time dependent^15,16. As we had suspected and even with a relatively small sample we found a large amount of deferred activity, only two of which were clinically acceptable, due to spontaneous reversion to sinus rhythm. The case could be made that even these were “lost” activity, as immediate pre admission rhythm was not defined. Patient deferred cancellation of treatment was remarkably infrequent.

While the small procedural risk related to embolism can be minimised with anticoagulation¹⁷, patients need also to understand the role of preparative and continued drug treatment particularly where a high proportion fail to sustain sinus rhythm under “real life” conditions7 . Issuing simple clear patient information leaflets (such as those issued in many hospitals including our own) can deal with this, provided that the referring clinicians are familiar with the routine. Providing strategies for acceptable anticoagulation was the key problem in our sample of patients and was the major reason for cancellation, amounting to nearly a 30% problem rate. This always involved under anticoagulation.

In our geographic setting is often left with the general practitioner to initiate and stabilise the INR. Clearly this is not a good arrangement, as they appear ill equipped to achieve a stable INR and tend to under treat patients. This is a common difficulty either in preparation for electrical cardioversion or in chronic management of AF and other conditions^18.19. While this function can be completed by the hospital service, in our institution it is not used routinely due to the distances involved in patients attending for venesection or finger prick testing. In future we propose to document predetermine control of anticoagulation and ECG rhythm three days prior to attendance to potentially refine therapy or reallocate the beds to alternative patients or different day case activity in the event of difficulty or spontaneous return to SR. Inevitably the over burdened general practitioner will still be required to fulfil this function given the huge geographic spread of patients in our region. It is hard to see how this could be improved upon unless there were specific deficiencies in understanding the principles of dose titration of warfarin or a lack of facilities to document a rhythm strip of ECG.

The main hospital issues in effective utilisation are in the selection of cases for treatment, and the preparation and maintenance of sinus rhythm^20. All factors associated with increasing the likelihood of successful and sustained conversion to SR should have been documented in each case (such as good control of HBP, left atrial size, mitral valve and thyroid function). Pre procedural therapies such as anti-arrhythmic drug treatment should be optimised. Our sample raised a number of points in this regard.

One positive feature was what appeared to be direct use of the system by local GP’s in conjunction with the cardiology service although realistically this was very limited. Alternatively we also saw a striking lack of use by any source other than hospital cardiology. These numbers imply poor utilisation elsewhere as the numbers of patients treated is far lower than that required to cover the expected (albeit potentially reducing8 ) demand for a unit serving 210,000 people. This theme of low service uptake of general cardiac services by non cardiology units is common in general hospitals²¹. While many aspects of case selection here were acceptable (we found no cases where attempted DC cardioversion was inappropriate e.g. multiple previous attempts untreated thyrotoxicosis etc), we found often extensive delays to treatment that were far too common.

The role of anti arrhythmic drug therapy in the sample was perhaps less problematic than anticipated. Some publications suggest that DC cardioversion of persistent AF, without pre treatment drug therapy is of little clinical value²². Unsupported DC cardioversion was a remarkably infrequent occurrence in our sample. This may represent a cardiology based pre-selection bias of cases referred for treatment. Similarly while a high proportion of patients had relevant pre-procedural investigational triage such as a very high utilisation of echocardiography, this need not have been reflecting targeted activity as it might also reflect a degree of indiscriminate testing. The prominent use of both beta blockade and/or Amiodarone is reassuring given their efficacy either as monotherapy or in combination to promote cardioversion and sustenance of sinus rhythm²³. Why we found such prominent use of Digoxin, as an adjuvant is unclear. Generally it is now accepted that this drug has little evidence base of either facilitating cardioversion or providing good ambulatory rate control as monotherapy²⁴. The successful medium term rate of sustaining sinus rhythm was predictably well preserved 7 presumably due to the high rate of adjuvant pharmacological treatment.

In summary this real life sample shows that the key factor creating quite a large problem with inefficient utilisation of service was out patient anticoagulant preparation. GP open access to cardioversion occurred but was at a very low level and much of the service limitation stemmed from poor community anticoagulation.

Back to February Contents