[The effects of long-term sedation on intestinal function].

Gastrointestinal integrity with intact function are of main importance in critically ill patients, and not only as a route of nutritional support. Drugs used for long-term sedation can lead to disordered gastrointestinal motility. In this study we compared the influence of different combinations of analgesics and sedatives on the intestinal function in mechanically ventilated, critically ill patients. METHODS. A total of 190 patients were evaluated retrospectively. All patients required controlled mechanical ventilation and deep sedation (Ramsay Score 5-6) for 7 days or more due to acute respiratory failure or elevated intracranial pressure. In none of these patients was enteric tube feeding contraindicated. Intact intestinal function was assumed when full enteric tube feeding was achieved on days 5 and 6 of the treatment period. Furthermore, other gastrointestinal motility disorders (e.g. constipation) had to be absent. In all patients the feeding tube was placed in the stomach by the nasogastric route. Corresponding to different combinations of analgesics and sedatives, the 190 patients were divided into 11 groups. The following combinations were used: group 1 (n = 20), fentanyl+flunitrazepam; group 2 (n = 20), fentanyl+midazolam; group 3 (n = 20), fentanyl+thiopentone; group 4 (n = 20) piritramide+midazolam; group 5 (n = 20), piritramide and continuous epidural administration of bupivacaine+midazolam; group 6 (n = 20), piritramide+gamma-aminobutyric acid (GABA); group 7 (n = 20), ketamine+midazolam; group 8 (n = 10), ketamin+methohexitone; group 9 (n = 20), ketamine+propofol; group 10 (n = 10), ketamine+midazolam and GABA; group 11 (n = 10), sufentail+midazolam and methohexitone. Patients in groups 3, 8, 9, 10, and 11 all had severe head injury and elevated intracranial pressure. Group 6 was made up exclusively of elderly patients (> 65 years) without head trauma. RESULTS. The patients in groups 1, 2, and 3 received fentanyl for analgesia and were completely fed by enteric tube in 30%, 35%, and 15% of cases, respectively. In group 3 deep sedation was necessary because of elevated intracranial pressure. In groups 4, 5, and 6, piritramide was administered for analgesia, and normal enteric tube feeding was achieved in 70%, 75%, and 90% of cases. The best results were seen in group 6, and these elderly patients needed smaller amounts of piritramide for analgesia. In groups 7, 8, 9, and 10, ketamine was given for analgesia, and complete enteric tube feeding was carried out in 75%, 30%, 45%, and 60% of these patients. The best results in the ketamine groups were found in combination with midazolam as the sedating drug; however, the patients in group 7 did not have elevated intracranial pressure, in contrast to the patients in groups 8, 9, and 10. The last group received the combination of sufentanil, midazolam and methohexitone to achieve a deep sedation. The rate of normal enteric tube feeding in these patients with severe head trauma was 30%. CONCLUSIONS. In patients with severe head trauma who need deep sedation to prevent dangerous high intracranial pressure, gastrointestinal motility disorders are very commonly found. The results obtained suggest that ketamine should be regarded as the analgesic drug of choice, combined with propofol rather than a high-dose barbiturate therapy. The combination of ketamine with midazolam and GABA is an unusual strategy for long-term sedation, which resulted from our own clinical studies directed at an effective and well-tolerated regime for this high-risk patient group. Obviously, high-dose barbiturates and short-acting opioids, especially when combined, make enteric tube feeding more difficult. Therefore, we recommend piritramide or ketamine for analgesia. The basic sedating drug is midazolam, in special cases combined with or replaced by propofol. The position of GABA in long-term sedation is not yet clear, but a lack of side effects on the gastrointestinal tract became evident in this stud