[Premedication in retrobulbar anesthesia. A blood gas analysis comparison of sublingual flunitrazepam and intravenous midazolam].

Benzodiazepines for sedation may decrease the PaO2, the arterial O2 saturation (SaO2), and the CO2 response more in the elderly than in the young. The purpose of this study was to assess changes in blood gases due to i.v. midazolam or sublingual flunitrazepam given as premedication in elderly patients for unilateral cataract surgery. METHODS. Fifty patients over 65 years of age with treated arterial hypertension and other co-existing diseases (ASA III-IV) were randomly assigned to have: (1) i.v. midazolam titrated until they became drowsy (17 patients; 2.85 +/- 0.84 mg [mean +/- SD]); (2) sublingual flunitrazepam (16 patients; 0.005 mg/kg); or (3) no sedation (17 patients; controls). On entering the operating theatre, the radial artery was cannulated and the first blood gas analysis was obtained. The premedication was then given. At 5, 10, 20, and 30 min after premedication, before and 10 min after retrobulbar block, before operation, 5 and 15 min after the beginning of the operation, 10 and 20 min after administration of 500 mg acetazolamide i.v. during the operation, and 10 and 20 min after the operation additional arterial blood samples were analysed (a total of 15 measuring points). Pulse oximetry, invasive blood pressure, and ECG were continuously monitored. All patients received oxygen 3 l/min during the operation by nasal cannula. Differences between the three groups were analysed by Student's t-test or U-test and a P value < 0.05 was considered significant. RESULTS. The patient demography, including duration of anaesthesia and operation, was similar in the three groups (Table 1). No significant differences were seen in heart rate, mean arterial pressure, PaO2, pulse-oximetric oxygen saturation (SpO2), base excess, or serum bicarbonate levels. The PaCO2 increased in patients after midazolam (P < 0.01) and flunitrazepam (P < 0.05) until the beginning of the operation compared with the control group (Fig. 3); 20 min after the operation there was still a significant difference between the midazolam group and the controls. SaO2 was significantly (P < 0.05) lower in the midazolam group 10 and 20 min after administration of premedication compared with the control group, but was within physiological limits (Fig. 5). Despite titration, 2 patients had severe respiratory insufficiency 3 min after midazolam: the SpO2 decreased below 85% and the paO2 below 55 mmHg. The paCO2 was higher (P < 0.05) in the midazolam group 10 min after acetazolamide compared with the controls. CONCLUSIONS. The results of the study show the potential hazards of i.v. midazolam in the elderly. If sedation is required for cataract surgery under local anaesthesia, we recommend sublingual flunitrazepam or the use of benzodiazepines with lower hypnogenic effects in the elderly. A thorough preoperative discussion of anaesthesia and the operation might be an adequate substitute for any premedication in high-risk patients; the best blood gas analysis results were obtained in the control group.