After Siriraj Ethical Committee approval, a prospective randomized control trial was conducted. The patients who were scheduled for TAH and met the criteria, including ASA (American Society of Anesthesiologists) classification 1-2, age from 18-60 years old, body mass index (BMI) below 35 and who accepted combined spinal-general anesthesia as the choice of anesthesia for TAH and intrathecal morphine as post-operative pain control, were enrolled. The patients who had any previous history of Hydroxyzine allergy or had urticaria, pruritus, nausea vomiting or motion sickness as underlying diseases or complications from a previous procedure or anesthesia were excluded. The patients who had drugs that effected pruritus and nausea vomiting or who had epidural morphine for post-operative pain control were also excluded.

From previous studies, the incidence of post-operative pruritus are from 25% to 85%, (mean is 60%).^{1,2,3,4,9,12,13,14,15} If Hydroxyzine can decrease the incident of pruritus by 50%, the target incident is 30%. We calculated the number of patients per group by this formula, N = [2(zα + zb)²P(1-P)]/(p_c-p_t).2 We indicated that p_c = 0.6, p_t  = 0.3, P = (0.6 + 0.3)/2 = 0.45, zα = 1.65 and zb = 0.84. After that we added 10% to the result for compensation for the patients who quit this study, finally we had 40 patients per group.

Eighty patients were randomized by the program Research Randomizer from www.randomozer.org/form.htm to receive either oral hydroxyzine 75 mg (3 tabs of 25 mg Hydroxyzine) and oral midazolam 7.5 mg (1/2 tab of 15 mg midazolam) (ATR group) or placebo (3 tabs of starch that mimic Hydroxyzine and oral midazolam 7.5 mg (C group) from a coated envelope as premedication at least half an hour before their operations. The observers and evaluators never knew the groups of patients. This could prevent the bias from scoring data and evaluating the results.  
In the operative room the sedation score was evaluated. Ringer’s Lactate was administered, 15 ml/kg intravenous within 30 minutes, then 150 ml/hour continuously until the operation was completed. Non-invasive blood pressure (NBP), pulse oximetry and electrocardiogram (ECG) were monitored in all patients. Spinal block with 0.5% heavy bupivacaine 2 ml with 0.3 mg preservative free morphine was performed by a number 25 Whitacre’s spinal needle and the injection rate was 2 ml in 20 seconds. General anesthesia was performed by thiopenthal sodium 5 mg/kg as induction, intubated with atracurium 0.6 mg/kg and maintained with nitrous oxide in 50% oxygen and isoflurane as needed. NBP was recorded every minute for 15 minutes, then every 2 minutes for 30 minutes and then every 5 minutes until the completion of the operation. If systolic blood pressure was below 110 mmHg with heart rate below 90 beats per minute, 6 mg of ephedrine was intravenously administered to patients and if the heart rate was above 90 beats per minute, 4 μg of levophedR was intravenously administered to patients. A conventional reversal technique with 1.2 mg atropine and 2.5 mg prostigmine was performed in all patients. After extubation patients were sent to the recovery room where vital signs, pruritic score, nauseous score, and sedation score were recorded until patients were sent to the ward. At the ward, vital signs, pruritic score, nauseous score and sedation score were recorded until 48 hours post-operation. Intravenous fentanyl was used for pain as needed, oral chropheniramine syrup 2 tsp (4 mg/10 ml) every 4 hours was used for pruritus and intravenous ondanzetron (8 mg) was used for nausea vomiting in the post-operative period. The amounts of these supplemental drugs were recorded too.

Statistical analysis
The data were processed using SPSS for Windows (Statistical Package for the Social Science) version 13.0. Ordinary data such as age and BMI were analyzed by mean and standard deviation. The differences in pruritic score, nauseous score, sedation score and other complications were analyzed by Chi-Square. The P value of < 0.05 was considered statistically significant.