Approval from Ethical Committee of Faculty of Medicine Universitas Indonesia was acquired prior conducting the study. Subjects were given informed consent before enrolling the study and randomized into two groups; single dose (SD) and fractionated dose (FD). All subjects will receive spinal anesthesia with 2,5 ml of 10 mg bupivacaine hyperbaric 0,5% and fentanyl 25 mcg. After IV cannulas (18 - 20 G) were properly placed, premedication with intravenous (IV) ranitidine 50 mg and metoclopramide 10 mg were given. Additional IV access was added if MgSO4 must been continuously administered intravenously perioperatively. Standard monitor device were placed after patients had been transferred to operating room. Additional premedication such as intravenous midazolam 0.02 mg/BW or fentanyl 1 mcg/BW were given if the patients were anxious or in pain. Baseline haemodynamic measurement was done two minutes after premedication. Spinal anesthesia was performed while patients in sitting position at L3-4 or L4-5 level with median or paramedian approach, with total dose 2,5 ml of 10 mg bupivacaine hyperbaric 0,5% and fentanyl 25 mcg. After skin wheal of lidocaine at the intended spinous interspace were placed, spinal needle (SpinocanR, B Braun, 27 G) were inserted and advanced until subarachnoid space were reached. In group FD, 1,5 ml of total dose followed by 1 ml remaining dose after 90 s interval were given. In SD group, 2,5 ml total dose were given with 0.2 ml/second rate, and subjects were asked to remain sitted for 90 seconds. If there were more than three injection attempts, the patients were excluded from the study. Co-loading 5 - 10 ml/kg ringer lactate in 20 minutes were given for all subjects.
