Objectives
Thrombocytopenia following Perceval aortic valve replacement has been described previously with variable outcome. Studies have lacked a robust analysis of platelet fluctuation and factors affecting it. We aimed to statistically describe the trend in thrombocyte variability as compared with conventional aortic valve replacement, and to assess predictors as well as impact on associated outcomes.

Methods
One hundred consecutive patients with first-time Perceval were retrospectively compared to 219 patients after Perimount Magna Ease valve replacement. The primary outcome was the serial thrombocyte count on day 0–6. Generalized estimating equations were used to analyse the data using fixed-effect models: for the effect of the post-operative day on platelet count, and random-effect models estimating both time-variant (platelets) and time in-variant variables (valve type, age, LV function, pre-op platelet level).

Results
Perceval patients were older (72 ± 1 vs 68 ± 1 years, p < 0.01) with higher NYHA status (3(2–3) vs 2(1–2), p < 0.001). Mean platelet count in the sutureless group was lowest on day 2 (91.9 ± 31.6 vs 121.7 ± 53.8 × 103 µl−1), and lower on day 4 (97.9 ± 44) and 6 (110.6 ± 61) compared to the conventional group (157.2 ± 60 and 181.7 ± 79) but did not result in a higher number of transfusions, bleeding or longer hospital stay (p > 0.05). Reduced platelet count was a strong predictor of red cell transfusion in the conventional (p = 0.016), but not in the sutureless group (p = 0.457). Age (Coef -1.025, 95%CI-1.649—-0.401, p < 0.001) and CPB-time (Coef 0.186, 95%CI-0.371—-0.001, p = 0.048) were predictors for lower platelet levels.

Conclusion
Considering the older patient profile treated with Perceval, postoperative thrombocytopenia does not impact on outcome in terms of transfusions, complications or hospital stay.