Cricoid pressure application to awake volunteers: discomfort cannot be used to indicate appropriate force.

To the Editor: The lowest platelet counts at which one can safely administer neuraxial anesthesia for labour and delivery is controversial. Published studies are few and sample sizes small.1–3 We report our institutional experience with thrombocytopenic parturients, probably the largest group to date. Medical records from 1997 to 2002 of parturients with platelet counts < 100,000·μL–1 during the peripartum period were reviewed for methods of anesthesia/analgesia for delivery, peripartum, and hospital course and incidence of neurological complications; 177 patients were identified. Of these, 170 (96%) received regional anesthesia. Ninety percent of identified patients had platelet counts > 70,000·μL–1; all received regional anesthesia for either vaginal or Cesarean delivery, as did all parturients with platelet counts 70,000·μL–1 to 60,000·μL–1 requesting regional anesthesia. In paturients with platelet counts between 50,000 to 60,000·μL–1, six received regional anesthesia and one was denied. Spinal, instead of epidural, was more often chosen in this group than in those with counts > 60,000·μL–1 (4/6 vs 29/160 P < 0.005, Chi square test), probably because five of seven parturients in this group presented for Cesarean delivery for worsening pre-eclampsia without being in labour and because of a perception among anesthesiologists of the lower risk of epidural hematoma associated with spinal (1:220,000) vs epidural anesthesia (1:150,000).4 Parturients with counts < 50,000·μL–1 received regional anesthesia only after platelet transfusion. In 82%, the platelet count was over 60,000·μL–1 at catheter removal in our study. In some women, platelet count decreased after epidural placement and the catheter was left in place for up to 36 hr until the platelet count showed a steady rise. Five patients had counts < 50,000·μL–1 (two of them, 36 and 38,000·μL–1, respectively) when catheters were removed. The causes of thrombocytopenia were gestational thrombocytopenia, pre-eclampsia, and idiopathic thrombocytopenia. There were no other associated coagulation abnormalities. No neurological complications were documented in any patients reviewed. Our retrospective study by no means proves that it is safe to administer regional anesthesia to parturients with platelet counts < 100,000·μL–1. Using the standard formulae for zero numerators,5 we estimate, with 95% confidence, that the incidence of epidural hematoma in our population is not greater than 1.8%. If our data are combined with other published reports, the expected incidence becomes even lower. Assessment of the incidence of a rare event is difficult in clinical practice. Therefore, in the absence of large, unfeasible studies, it is prudent to evaluate the risk-benefit ratio on a case-by-case basis before administering regional anesthesia to parturients.