How can we deal with an unexpected preoperative prolongation of the activated partial thromboplastin time (APTT)? —a real world experience

Background: This article describes a tentative diagnostic algorithm for investigating patients with unexpected prolongation of activated partial thromboplastin time (APTT), detected during preoperative laboratory testing. Methods: The study population consisted of all patients undergoing preoperative laboratory testing (thus including APTT) during a 3-month period. Whenever prolonged APTT results were observed, abnormal data were investigated according to an algorithm combining familiar history of bleeding (yes/no), personal history of bleeding (yes/no) and signs/symptoms of bleeding with results of hemostasis testing, including mixing test, lupus anticoagulant (LAC) and clotting factor assays. Results: Overall, 1,433 patients underwent preoperative APTT testing throughout the 3-month study period, 76 (5.3%) of whom had prolonged values (i.e., >36.5 s). The remaining 1,357 (94.7%) patients underwent successful surgery, without further investigation. According to the algorithm, 55/76 (72.4%) of patients with prolonged APTT were excluded from supplementary laboratory testing. In the remaining 21/76 (27.6%) patients, a mixing test was scheduled. In 5/21 (23.8%) of these patients the mixing test was positive (i.e., no normalization), so that LAC testing was planned. In all these five patients LAC testing was positive. In the remaining 16/21 (76.2%) patients, mixing test was negative (i.e., normalization), so that clotting factors assays were scheduled. Seven patients were finally diagnosed with factor (F)XII deficiency, two patients with mild FXI deficiency (i.e., FXI activity of 47% and 51%), whilst 5 had clinically insignificant decreased values of one or more clotting factors and 2 patients had no identifiable explanation for APTT prolongation. The sensitivity and specificity of the algorithm for detecting clinically significant conditions (i.e., FXI deficiencies or LAC) were 1.00 and 0.99, respectively. Conclusions: The results of this study suggest that a preoperative algorithm including clinical and laboratory data may be a reasonable and sustainable approach for identifying patients at enhanced risk of perioperative bleeding or thrombosis.