Predictors of pacemaker dependency in patients receiving devices following cardiac surgery

Introduction: Permanent pacemaker (PPM) implantation is a well-recognised short-term complication after cardiac surgery; however, in some cases, conduction system function may recover, leaving the patient with an unnecessary device and its associated morbidity. Aims: In order to aid in decision-making process for PPM implantation post cardiac surgery, we sought to determine predictors of pacemaker dependency. Methods: We identified patients who had undergone PPM implant within 30 days of cardiac surgery (bypass grafting and/or valvular surgery) at our institution between May 2016 and December 2019. Parameters including pre-operative ECG, surgical details and device follow-up data were collected. Pacemaker dependency (PPM-D) was defined as any of the following at the initial post discharge device check: no underlying rhythm at VVI 40, underlying junctional rhythm, RV pace percentage >80% or underlying rhythm rate of 40 bpm or less. Using a stepwise approach, we constructed a series of logistic regression models to predict PPM-D, minimising the Akaike Information Criterion (AIC) and maximising the AUROC/C-statistic to optimise discriminative performance. Results: We identified 143 patients who had undergone PPM implantation for whom device follow-up information was available. Only seventy eight (54.5%) patients were PPM-D as defined. After multivariate logistic regression analysis two factors predicted PPM-D. The presence of post-operative complete AV block (OR 11.41; CI, 2.55–73.82; p=0.004) and QRS duration >120 ms on the post-operative ECG (OR 6.88; CI, 1.72–34.73; p=0.010) (Figure 1). Interestingly, neither surgery type (valve, bypass, etc.) nor any pre-operative ECG parameters were significantly associated in this cohort. A model incorporating just these factors had high discriminative capability with AUROC=0.749. Conclusions: In post cardiac surgical patients, pacemaker dependency was only present in approximately half of patients at the post-discharge device check. Although this data requires external validation, assessing for complete AV block or a broad QRS may be helpful in deciding whether to implant PPMs in this patient group. [Image Omitted]