Incidence of postoperative atrial fibrillation

Of the 15 Dutch cardiothoracic centres, 14 (93%) responded to the survey, but not all centres provided an answer to each question. Completion rates per question are outlined in Table S1 in the Electronic Supplementary Material. Thirteen centres reported a POAF incidence, ranging from 20 to 30% (median 27%). The reported incidences per centre are presented in Fig. 1.

Fig. 1

Reported incidence of postoperative atrial fibrillation (POAF) after cardiac surgery in 13 Dutch cardiothoracic centres

Preoperative prophylaxis

Nine of the 14 centres (64%) used local protocols for POAF prophylaxis and treatment, consisting of administration of prophylactic beta-blockers and an anticoagulant regimen once POAF occurred. Two centres specified the prescribed dose of beta-blockers (i.e. 25 mg twice daily or 80 mg once daily). One centre also described perioperative electrolyte regulation in its protocol. One centre followed European guidelines instead of a protocol.

Six centres (43%) prescribed pharmaceutical prophylaxis preoperatively. This included non-sotalol beta-blockers (4/6; 67%), calcium antagonists (1/6; 17%), sotalol (1/6; 17%) and other current medication (1/6; 17%). Three centres prescribed prophylaxis to all patients, 2 centres prescribed prophylaxis only to patients who were already on beta-blockers or calcium antagonists, and 1 centre prescribed an anti-arrhythmic drug to patients already being treated with the specific drug. Colchicine was not used at all. The reported prescription of preoperative medications for AF prophylaxis in 6 centres is outlined in Fig. 2.

Fig. 2

Reported prescription of preoperative prophylactic medication for atrial fibrillation in Dutch cardiothoracic centres

Non-pharmaceutical prevention

Of the respondents, 12 (86%) believed posterior left pericardiotomy can prevent POAF, but only 1 centre used this intervention. Two centres (14%) were positive about this technique. The major reason for non-use was a lack of positive evidence (11/14; 79%). Similarly, most centres (11/14; 79%) believed perioperative pericardial flushing potentially reduces POAF, but only 1 centre used it in a research setting and none of them in a clinical context. All 11 centres found the evidence to be insufficient, and 2 centres (14%) stated there were no perceived benefits. One centre was positive about its effectiveness. Eight centres (57%) were familiar with BoNT/A injections in the atrial fat pad, but none used it to prevent POAF due to a lack of scientific evidence.

Left atrial appendage closure

Most centres (10/14; 71%) routinely performed concomitant LAAC for patients with pre-existent AF. Of them, 30% was aware this entailed procedure may increase the risk of POAF. Of the remaining 4 centres not routinely performing LAAC, 3 were not aware of this risk and 1 centre was. Indications to perform LAAC varied. Eight of the 14 centres (57%) expected to alter current indications in response to the LAAOS III trial results [19]. Two centres (14%) performed LAAC exclusively in surgical ablation, but one of them seldomly performed LAAC for patients with chronic AF, 3 centres reserved LAAC for patients with pre-existent AF undergoing open-heart surgery, and 1 centre did not state an indication.

Postoperative prophylaxis and treatment

All 14 centres administered postoperative medication. Their first line of medication was standard AF prophylaxis, while the second and subsequent lines of medication were administered in the clinical phase. The majority (12/14; 86%) prescribed non-sotalol beta-blockers first, and 8 of them (57%) specified the use of metoprolol. Only 2 centres (14%) prescribed sotalol as standard prophylaxis.

Digoxin, amiodarone and calcium antagonists were only used therapeutically when patients presented with POAF. All centres had a second line of medication, which comprised sotalol (43%), digoxin (21%), amiodarone (14%), non-sotalol beta-blockers (14%) and calcium antagonists (7%). Eight centres had a third line of treatment, which mostly consisted of a calcium antagonist (50%). One centre administered either amiodarone or digoxin as third-line medication. Three centres had a fourth line of treatment. One centre prescribed either amiodarone or sotalol as its fourth line. Furthermore, 2 centres prescribed combination therapy of non-sotalol beta-blockers and digoxin as its third- or fourth-line treatment. The reported first and subsequent lines of medication for the prevention and treatment of POAF are outlined in Fig. 3.

Fig. 3

Reported prescription of postoperative medication for prevention and treatment of postoperative atrial fibrillation in Dutch cardiothoracic centres. Prophylaxis is first line of medication only

Overall, non-sotalol beta-blockers and sotalol took up 38 and 24%, respectively, of all medication prescribed postoperatively for POAF. Digoxin, calcium antagonists and amiodarone comprised 14, 13 and 10% of the prescriptions, respectively.

Anticoagulant usage

All centres prescribed anticoagulants whilst treating POAF, but the indication differed. Six centres (43%) based their indication area once POAF occurred on national guidelines, 2 centres (14%) used the CHA2DS2-VASc score and national guidelines, another 2 centres only used the CHA2DS2-VASc score, 3 centres (21%) followed local protocols, and 1 centre used both its local protocol and the CHA2DS2-VASc score.

The question about the annual number of patients discharged with vitamin K antagonists (VKAs), was answered by 13 centres (93%): 4 (31%) replied this number was unknown and they could not provide a description of the trend, 7 (54%) reported an annual incidence of 0–56% (median 21%), and 2 (15%) could not provide an exact number but either stated a decrease in VKA prescriptions or an increase in direct oral anticoagulant (DOAC) prescriptions. The centre that did not prescribe VKAs to patients with POAF administered DOACs.

With regard to prescription duration, 1 centre was not aware how long its patients took VKAs. The majority (12/14; 86%) included the duration of VKA prescription in the referral letter to the Dutch Thrombosis Service. One centre prescribed anticoagulants indefinitely. None of the centres arranged a follow-up for their patients pertaining to VKA usage.

Electrical cardioversion

Only 3 of the 14 centres (21%) did not perform standard ECV if chemical conversion failed, although 2 of them performed ECV on indication or in hemodynamically impaired patients. Twelve centres answered the question about the annual number of ECVs: 3 (25%) did not know the exact number and 9 (75%) provided an estimate, which ranged from < 10 to 200. Figure 4 shows the reported annual number of ECVs performed per centre.

Fig. 4

Reported number of electrical cardioversions performed yearly in Dutch cardiothoracic centres

Perceived complications

Thirteen respondents (93%) noticed longer hospitalisation for patients with POAF. Additionally, 6 centres (43%) replied these patients generally had other complications. Anaemia and pleural effusion each comprised 21% of the perceived complications, pneumonia made up 14%, and the incidence of sepsis, excessive pericardial fluid, hypoxia, overfilling, renal damage and neurological complications was 7% each. The perceived complications associated with POAF are outlined in Fig. 5.

Fig. 5

Perceived complications associated with postoperative atrial fibrillation in Dutch cardiothoracic centres

Five respondents (36%) did not notice POAF occurred more often after specific procedures, but 9 (64%) stated they did. Of those that did, 5 centres reported observing POAF mainly after mitral or any valvular surgery, 1 centre noticed POAF after coronary bypass, and the remaining 3 saw POAF more frequently after complex procedures, longer perfusion and clamping duration, and impaired left ventricle function.

Postoperative atrial fibrillation risk score

None of the centres used the POAF risk score [20]. However, when asked about their willingness to implement it, 7 (50%) were positive, 5 (36%) were negative, and 2 (14%) were indifferent. Reasons for unwillingness were insubstantial support, no effective prevention, and no added benefits either due to low POAF rates at their centre or because the centre administered prophylaxis to all patients. Of the 13 centres that further elaborated on their answer, 12 were willing to use a similar risk score if proven sufficiently effective. One centre was indifferent, as it considered all patients undergoing surgery as high-risk.

Recommendation on improvements

Half of the respondents (7/14) answered the question on what could be an improvement for Dutch centres regarding POAF management. They mainly wanted better national agreements or a consensus on preferred treatment for more uniform management. Others expressed wanting better designed guidelines for anticoagulant prescription and increased guidance on duration, especially after referral to the Thrombosis Service.