Intrathecal opioids are now used routinely in the UK for intra- and postoperative analgesia. The opioids of choice have altered over recent years and the dosage regimens used can vary between institutions. Concerns over safety have been reduced probably because much lower doses of opioids are now being used. This survey explored the practice of intrathecal opioid usage in the UK.

We sent a questionnaire survey to 270 anaesthetic departments and received 199 replies, a response rate of 73.7%.

Intrathecal opioids were used in 175 (88.4%) departments. Of these departments, 107 (61.1%) had local guidelines or protocols in place. Opioids such as diamorphine (used in 136 (78.2%) of departments) and fentanyl (129 (74.1%)) with a shorter duration of action are now more commonly used than morphine (37 (21.3%)) for intrathecal analgesia. In 96 (54.5%) departments, patients were nursed on regular surgical wards following administration of spinal opioids.

The use of low-dose lipophilic intrathecal opioids for postoperative analgesia is widespread in the UK. Patients are commonly nursed in low-dependency post-anaesthetic care areas. The low incidence of adverse events reported by the respondents along with the popularity of the technique suggests that low-dose spinal opioid administration is safe.

Tonsillectomy is frequently associated with postoperative pain of considerable duration, which is usually accompanied by the substantial consumption of both opioid and non-opioid analgesics. Despite the use of different surgical and anaesthetic techniques in the search for safe and effective post-tonsillectomy pain relief, this problem remains a clinical dilemma. The aim of the current study was to evaluate the potential effects of topically administered ketamine and morphine by an oral rinse into the tonsillar fossae.

In all, 60 children, 15 for each group, aged between 3 and 12 yr scheduled for tonsillectomy were randomly assigned to one of four groups. Study drugs were administered to both tonsillar fossae for 5 min. Group K received 0.4 mL (20 mg) ketamine in 10 mL artificial saliva, Group KM received 0.4 mL (20 mg) ketamine + 5 mL (20 mg) 4‰ morphine aqueous solution in 5 mL artificial saliva, Group M received 5 mL (20 mg) 4‰ morphine aqueous solution in 5 mL artificial saliva, Group C received only 10 mL artificial saliva. Postoperative pain, nausea, vomiting, sedation and bleeding were evaluated.

Pain scores were higher in the control group at arrival in the recovery ward (P < 0.05). Morphine and ketamine groups had longer effective analgesia time than the morphine + ketamine and control groups. The 24-h analgesic consumption was significantly higher in the control group.

Topical ketamine and morphine seems to be a safe and easy analgesic approach for decreasing adenotonsillectomy pain.

A randomized and prospective study was performed to compare anaesthetic characteristics and stress hormone responses of two anaesthetic techniques.

Forty-two patients undergoing day case excisional biopsy of breast mass were randomly assigned to receive propofol–remifentanil or sevoflurane–N2O. Anaesthesia was induced and maintained either with sevoflurane and 50% N2O in oxygen or with target-controlled remifentanil and propofol in 50% oxygen and air. Anaesthetic depth was monitored by the bispectral index.

The times for induction (2.9 vs. 1.7 min) and for laryngeal mask insertion (5.7 vs. 3.3 min) were longer in the sevoflurane–N2O group than in the propofol–remifentanil group. However, apnoea (57.1% vs. 9.5%) and bradycardia (23.8% vs. 0%) were more prevalent with propofol–remifentanil. In the sevoflurane–N2O group, the emergence times to a verbal response (10.6 vs. 3.7 min), to extubation (11.8 vs. 4.0 min) and to orientation (14.7 vs. 4.8 min) were longer than in the propofol–remifentanil group. There were significantly more nausea (38.1% vs. 4.8%) and vomiting (19.2% vs. 0%) in the sevoflurane–N2O group than in the propofol–remifentanil group. The time to discharge was similar although there was less postoperative pain in the sevoflurane–N2O group. There were no differences in the perioperative cortisol responses in the two groups.

Smoother induction of anaesthesia was seen with sevoflurane–N2O. Propofol–remifentanil showed a quicker emergence with less nausea/vomiting. There were similar perioperative cortisol responses in the two anaesthetic techniques.

To investigate whether preemptive administered lornoxicam changes perioperative platelet function during thoracic surgery.

A total of 20 patients scheduled for elective thoracic surgery were randomly assigned to receive either lornoxicam (16 mg, i.v.; n = 10) or placebo (n = 10) preoperatively. All patients underwent treatment of solitary lung metastasis and denied any antiplatelet medication within the past 2 weeks. Blood samples were drawn via an arterial catheter directly into silicone-coated Vacutainer tubes containing 0.5 mL of 0.129 M buffered sodium citrate 3.8% before, 15 min, 4 h and 8 h after the study medication was administered. Platelet aggregation curves were obtained by whole blood electrical impedance aggregometry (Chrono Log®).

Platelet aggregation was significantly reduced 15 min, 4 h and 8 h after lornoxicam administration compared to placebo (P < 0.05) for collagen, adenosine diphosphate and arachidonic acid as trigger substances. Adenosine diphosphate-induced platelet aggregation decreased by 85% 15 min after lornoxicam administration, and remained impaired for 8 h.

Platelet aggregation assays are impaired for at least 8 h after lornoxicam application. Therefore perioperative analgesia by use of lornoxicam should be carefully administered under consideration of subsequent platelet dysfunction.

Controlled hypotension is frequently used for obtaining better exposure during tympanoplasty. The aim of this study was to compare dexmedetomidine, a selective, short-acting, central α2-adrenergic agonist with remifentanil, an ultra-short-acting opioid with properties similar to other μ-specific agonists, regarding their effects in achieving controlled hypotension and improving surgical field exposure and surgeon’s satisfaction during tympanoplasty.

In this prospective, double-blind pilot study, 24 consecutive patients scheduled for elective tympanoplasty were randomly assigned to receive either dexmedetomidine 1 μg kg−1 over 10 min at anaesthesia induction followed by 0.4–0.8 μg kg−1 h−1 infusion during maintenance or remifentanil 1 μg kg−1 over 1 min at anaesthesia induction followed by 0.2–0.4 μg kg−1 min−1 infusion during maintenance. Mean arterial pressure and heart rate were recorded before induction, at incision, 30, 60, 90 and 120 min after incision and 10 min after stopping the infusion. Surgical field exposure condition and satisfaction scores were assessed by the surgeon, blinded to the study drugs.

Mean arterial pressure and heart rate were significantly lower in the remifentanil group compared with the dexmedetomidine group at all times (P = 0.03 and 0.036, respectively). Surgical field exposure condition (3 ± 0.01 vs. 2.3 ± 0.7; P = 0.039) and surgeons’ satisfaction (3 ± 0.01 vs. 2.25 ± 0.87; P = 0.039) scores were significant after remifentanil compared with dexmedetomidine.

Infusion of dexmedetomidine, at the doses used in this study, was less effective than remifentanil in achieving controlled hypotension, good surgical field exposure condition and surgeons’ satisfaction during tympanoplasty.

This study was designed to determine the protective effects of zinc on halothane-induced hepatotoxicity.

Forty-five male Sprague-Dawley rats were divided into three groups. The halothane group received normal drinking water and diet; the zinc-halothane group received 227 mg L−1 zinc sulphate in the drinking water and diet for 2 weeks; and the control group received normal diet and water. At the end of 2 weeks, rats were housed in an anaesthesia box and 1 MAC (minimum alveolar concentration) halothane was administered at 6 L min−1 in room air for 2 h. This was repeated 48 h later. After the rats were sacrificed, we measured alkaline phosphatase, aspartate aminotransferase, alanine aminotransferase, gamma-glutamyl transpeptidase, glutathione-S-transferase, serum electrolytes and bilirubin in samples. The degree of liver toxicity was assessed by light microscopic examination.

We demonstrated a reduction of alanine aminotransferase, aspartate amino transferase, glutathione-S-transferase levels and a reduction in liver damage in the zinc-halothane group.

The study concludes that zinc has the potential to alleviate the toxic effects of halothane in rat liver.

The electroencephalographic Narcotrend Index was evaluated as a measure of sedation in mechanically ventilated intensive care unit patients. Narcotrend Index and conventional electroencephalography parameter values were compared to the Richmond Agitation-Sedation Scale and a simplified three-level sedation scale.

In all, 100 mechanically ventilated patients, admitted to the cardiac surgical intensive care unit after open-heart surgery, were enrolled in this prospective observational study. The Narcotrend Index was recorded while patients were either sedated by propofol infusion or without sedative medication while being weaned off the ventilator. Clinical assessment of the patients’ level of sedation was performed by means of the Richmond Agitation-Sedation Scale by a single observer who was blinded to the Narcotrend Index.

With the six-level Richmond Agitation-Sedation Scale, the prediction probability (PK) for the Narcotrend Index (0.81) was better than for all other electroencephalography parameters (P < 0.01) except for relative power in the β band (PK 0.75). Using the three-level sedation scale instead, PK values for the Narcotrend Index (0.88) and all electroencephalography parameters improved (P < 0.01), and the Narcotrend Index was now superior to all electroencephalography parameters. Narcotrend Index values were distributed among the various sedation levels with significant overlap.

When used for assessment of propofol sedation in postsurgical cardiac intensive care unit patients, the Narcotrend Index can distinguish between very light and deep sedation. With respect to differentiation between light and moderate or moderate and deep sedation, the Narcotrend Index was not reliable in this study.

Limited and inconsistent data exist on simple, readily available predictors of long-term mortality of critically ill chronic obstructive pulmonary disease patients requiring invasive mechanical ventilation. We therefore examined the influence of arterial blood gas derangement and burden of comorbidities on 90-day and 1-yr mortality of chronic obstructive pulmonary disease patients treated with invasive mechanical ventilation.

We identified all chronic obstructive pulmonary disease patients (n = 230) treated with invasive mechanical ventilation between 1994 and 2004 at a Danish primary-level hospital. Data on arterial blood gas specimens and comorbidity were obtained from medical records and Hospital Discharge Registries. We used Cox’s regression analysis to estimate mortality ratios according to arterial blood gas values and level of comorbidity.

Ninety-day and 1-yr mortality among chronic obstructive pulmonary disease patients requiring invasive mechanical ventilation was 30.8% and 40.5%, respectively. All 90-day and 1-yr mortality ratios according to arterial blood gas values were close to one and one was included in all 95% CI. Among patients with a high level of comorbidity 90-day mortality ratio was 1.3 (95% CI: 0.6–2.7) when compared with patients without comorbidity. The corresponding 1-yr mortality ratio was 1.4 (95% CI: 0.7–2.9).

Chronic obstructive pulmonary disease patients treated with invasive mechanical ventilation have substantial long-term mortality. Neither the levels of arterial blood gas values measured immediately before invasive mechanical ventilation was initiated nor the burden of comorbidity were strong determinants of long-term mortality among these patients.

Lornoxicam is a novel non-steroidal anti-inflammatory drug with marked analgesic property. It has been shown to have analgesic potency similar to morphine and meperidine. Laryngoscopy and tracheal intubation provoke major haemodynamic responses. Minimizing perioperative adverse events in elderly patients is of utmost importance. The aim of this study was to demonstrate the effect of preoperative administration of Lornoxicam on haemodynamic changes during laryngoscopy and tracheal intubation in the elderly.

Fifty patients aged between 65 and 75 yr were randomly recruited to this randomized, double-blind, placebo-controlled study. They were divided into two groups to receive either Lornoxicam 8 mg or placebo half an hour before surgery. Systolic and diastolic blood pressure, mean arterial pressure, and heart rate were recorded before and after administration of the intravenous anaesthetic, also at 1, 3, 5 and 10 min after tracheal intubation.

In the control group, significant increases in the haemodynamic parameters were observed during the first 10 min following tracheal intubation (P < 0.05).

Preoperative administration of Lornoxicam attenuates the haemodynamic response to laryngoscopy and tracheal intubation in the elderly.

A significant proportion of preschool children experiences severe emergence agitation after anaesthesia. The symptoms of disorientation, restlessness, inconsolable crying and thrashing resemble an acute psychosis similar to an agitated central anticholinergic syndrome. The primary aim of this randomized controlled study was to assess the efficiency of the cholinesterase-inhibitor physostigmine in these children and to identify adverse effects.

We anaesthetized 211 children (1–5 yr) with sevoflurane after midazolam premedication for varying operative procedures. Multimodal intraoperative and prophylactic pain therapy combined alfentanil, piritramide, diclofenac and regional/local bupivacaine. A 5-step score assessed emergence agitation. Severely agitated children were treated immediately with physostigmine (30 μg kg−1) or placebo in a randomized, double-blind fashion. The primary variable was the agitation score after 5 min.

Severe delirium occurred in 19% of all children. Five minutes following injection, severe agitation was still present in 10 out of 20 patients treated with physostigmine and 16/20 with placebo. This difference did not reach statistical significance (P = 0.1). Rescue therapy with intravenous propofol was given after 15 min of severe agitation to four children following physostigmine and nine following placebo (non-significant). An increased rate of postoperative nausea and vomiting (45% vs. 15%, P < 0.05) was the only adverse effect observed.

Severe emergence agitation might be related to a central anticholinergic syndrome as diagnosed empirically with a successful treatment with physostigmine. However, the results of this study do not support its routine use. The substance may augment the therapeutic options if injected slowly and after suitable prophylaxis to avoid postoperative nausea and vomiting.

Four percent gelatine is an alkaline compound due to NH2 groups, whereas 6% hydroxyethyl starch 130/0.4 (HES130) has acidic features. We investigated whether these solutions lead to differences in acid–base balance in pigs during acidaemia and correction of pH.

Anaesthetized pigs were randomized to HES130 or gelatine infusion (n = 5 per group). Animals received acid infusion (0.4 M solution of lactic acid and HCl diluted in normal saline) and low tidal volume ventilation (6–7 mL kg−1, PaCO2 of 80–85 mmHg, pH 7.19–7.24). Measurements were made before and after induction of acidaemia, before and after correction of pH with haemofiltration (continuous venovenous haemofiltration) and tris-hydroxymethylaminomethane infusion. We measured parameters describing acid–base balance according to Stewart’s approach, ketone body formation, oxygen delivery, haemodynamics, diuresis and urinary pH.

Acid–base balance did not differ significantly between the groups. In HES130-treated pigs, the haemodilution-based drop of haemoglobin (1.4 ± 1.0 g dL−1, median ± SD) was paralleled by an increase in the cardiac output (0.5 ± 0.4 L min−1). Lacking increases in cardiac output, gelatine-treated pigs demonstrated a reduction in oxygen delivery (149.4 ± 106.0 mL min−1). Tris-hydroxymethylaminomethane volumes required for pH titration to desired values were significantly higher in the gelatine group (0.7 ± 0.1 mL kg−1 h−1 vs. HES130: 0.5 ± 0.2 mL kg−1 h−1).

The buffer capacity of gelatine did not lead to favourable differences in acid–base balance in comparison to HES130.

Postoperative shivering and pain are frequent problems in patients recovering from anaesthesia with particularly high incidences being observed after remifentanil–isoflurane-based general anaesthesia. The opioid tramadol is generally effective in preventing shivering and treating pain, but its effects are not characterized after remifentanil-based general anaesthesia. This randomized, placebo-controlled, double-blind study evaluated the effects of intraoperative intravenous tramadol on postoperative shivering and pain after remifentanil-based general anaesthesia.

After Ethics Committee approval, 60 patients scheduled for lumbar disc surgery were included. Surgery was performed under general anaesthesia (remifentanil, isoflurane). Patients were randomly assigned to receive 2 mg kg−1 tramadol in 30 mL 0.9% saline infused intravenously (n = 30) or 30 mL saline (n = 30) 45–30 min before skin closure. The following parameters were assessed every 10 min for 2 h: shivering, pain, postoperative nausea and vomiting, sedation, heart rate, non-invasive blood pressure and peripheral oxygen saturation. The primary outcome variable was the incidence of shivering during the first 2 postoperative hours. Secondary variables were: shivering intensity, pain, postoperative nausea and vomiting, sedation, heart rate, non-invasive blood pressure and peripheral oxygen saturation.

Shivering was less frequent in patients treated with tramadol (20% vs. 70%, P = 0.0009) and was of lower intensity (severe shivering: 10% vs. 46.7%, P = 0.003). Pain scores were similar between the groups and all other secondary outcome variables failed to reveal significant differences.

Compared with placebo, intraoperative intravenous administration of 2 mg kg−1 tramadol reduces the incidence and extent of postoperative shivering without alterations in pain perception after lumbar disc surgery under remifentanil–isoflurane-based general anaesthesia.

Hypoxia-reperfusion injury is an important determinant of secondary brain injury. In the acute phase of cerebral reperfusion, pro-inflammatory events enhance expression of cerebral endothelial (intercellular adhesion molecule-1 and P-selectin) adhesion molecules, which play an important role in brain hypoxia-reperfusion injury. Midazolam is the most commonly used sedative in patients with brain injury. The objective of this investigation was to examine the effect of midazolam on the expression of cerebral endothelial intercellular adhesion molecule-1 and P-selectin during hypoxia-reperfusion injury invitro.

The up-regulation of mouse cerebral endothelial cells intercellular adhesion molecule-1 and P-selectin was assessed following hypoxia-reoxygenation (hypoxia-reperfusion). Cells were pre-treated with three different concentrations of midazolam (0, 5 and 50 μg mL−1) prior to hypoxia. Flow cytometry was used to estimate adhesion molecule expression mean channel fluorescence. Data are presented as mean ± SD.

Mouse cerebral endothelial cell intercellular adhesion molecule-1 and P-selectin expression was greater after exposure to hypoxia-reoxygenation compared to normoxia (mean channel fluorescence) 241 ± 12 vs. 140 ± 7 and 120 ± 14 vs. 46 ± 7, respectively, P < 0.05. Intercellular adhesion molecule-1 and P-selectin expression was decreased by midazolam (5 μg mL−1) pre-incubation compared to control, mean channel fluorescence 184 ± 10 vs. 241 ± 12 and 51 ± 7 vs. 120 ± 14, respectively, P < 0.05. Midazolam at 50 μg mL−1 had the same effect as 5 μg mL−1.

Pre-treatment with midazolam diminishes increased expression of cerebral endothelial intercellular adhesion molecule-1 and P-selectin expression following hypoxia-reoxygenation.

With the increasing demand for one-lung ventilation in both thoracic surgery and other procedures, identifying the correct placement becomes increasingly important. Currently, endobronchial intubation is suspected based on a combination of auscultation and physiological findings. We investigated the ability of the visual display of airflow-induced vibrations to detect single-lung ventilation with a double-lumen endotracheal tube.

Double-lumen tubes were placed prior to surgery. Tracheal and endobronchial lumens were alternately clamped to produce unilateral lung ventilation of right and left lung. Vibration response imaging, which detects vibrations transmitted to the surface of the thorax, was performed during both right- and left-lung ventilation. Geographical area of vibration response image as well as amount and distribution of lung sounds were assessed.

During single-lung ventilation, the image and video obtained from the vibration response imaging identifies the ventilated lung with a larger and darker image on the ventilated side. During single-lung ventilation, 87.2 ± 5.7% of the measured vibrations was detected over the ventilated lung and 12.8 ± 5.7% over the non-ventilated lung (P < 0.0001). It was also noted that during single-lung ventilation, the vibration distribution in the non-ventilated lung had a majority of vibration detected by the medial sensors closest to the midline (P < 0.05) as opposed to the midclavicular sensors when the lung is ventilated.

During single-lung ventilation, vibration response imaging clearly showed increased vibration in the lung that is being ventilated. Distribution of residual vibration differed in the non-ventilated lung in a manner that suggests transmission of vibrations across the mediastinum from the ventilated lung. The lung image and video obtained from vibration response imaging may provide useful and immediate information to help one-lung ventilation assessment.

The aim of this study was to compare the postoperative analgesic efficacy of intraperitoneal tramadol with intravenous tramadol or normal saline in patients undergoing laparoscopic cholecystectomy.

Sixty-one patients undergoing laparoscopic cholecystectomy were randomized to one of three groups in a double-blind manner via coded syringes. All patients received an intravenous and an intraperitoneal injection after installation of the pneumoperitoneum and again before removal of the trocars. In the control group, all injections were with normal saline. In the intravenous tramadol group, patients received intravenous tramadol 100 mg and intraperitoneal saline. In the intraperitoneal tramadol group, patients received intravenous saline and intraperitoneal tramadol 100 mg. All patients had a standard anaesthetic. Postoperative analgesia was with morphine. Postoperatively, numeric pain scores for parietal and visceral pain, 1 h and 24 h morphine consumption, and adverse effects were recorded.

Parietal and visceral pain scores were lowest in the intravenous tramadol group during the first postoperative hour (P < 0.016 compared with control). The delay until the first analgesic administration was longest in the intravenous tramadol group (median 23 min, range 1–45), when compared with the intraperitoneal tramadol group (10, 1–120 min, P = 0.263) or with the control group (1, 1–30 min, P = 0.015). One-hour morphine consumption was significantly lower in the intravenous tramadol group (mean ± SD; 3.4 mg ± 2.5) and in the intraperitoneal tramadol group (4.4 ± 4.3 mg) compared with the control group (6 ± 2 mg) (P = 0.044). There was no difference between the three groups regarding pain scores, morphine consumption and incidence of shoulder pain or adverse effects at 24 h.

Intravenous tramadol provides superior postoperative analgesia in the early postoperative period after laparoscopic cholecystectomy compared with an equivalent dose of tramadol administered intraperitoneally and with normal saline in patients undergoing laparoscopic cholecystectomy.

The generation of shunt criteria by somatosensory evoked potentials is controversial. This study was designed to validate somatosensory evoked potentials’ tracings and their derived shunt criteria compared with clinical-neurological examination.

We prospectively studied the somatosensory evoked potentials’ tracing in 102 patients who underwent carotid endarterectomy under cervical plexus blockade, comparing different somatosensory evoked potentials’ derived thresholds of impaired cerebral blood flow with the neurological examination.

The 50% reduction of the primary cortical response to median nerve stimulation as threshold of critical cerebral blood flow showed a sensitivity of 87% and a specificity of 66%. In case of total collapse of primary cortical response, we observed a sensitivity of 85% and a specificity of 89%. Concerning the prolongation of central conduction time between the spinal response at N13 and the cortical response at N20/P25 as critical threshold of impaired cerebral blood flow, the sensitivity was 87% and specificity 40%. The need-to-shunt index with its threshold at >0.5 showed a sensitivity of 85% and a specificity of 88%.

Compared to other studies, we have shown a much lower sensitivity and specificity of somatosensory evoked potentials in detecting critical cerebral perfusion; thus its use remains controversial and has to be assessed in further investigations.

Massive post-partum haemorrhage continues to be one of the world’s leading causes of maternal morbidity and mortality. Any new treatment that potentially helps at risk parturients should be thoroughly investigated. Recombinant factor VIIa (rVIIa) is increasingly being used in the treatment of massive haemorrhage. We performed a case-matched analysis of its use since 2003 in the treatment of massive post-partum haemorrhage at our hospital.

Twenty-eight cases of massive post-partum haemorrhage were identified over a 3-yr period since 2003. In six of these cases, rVIIa was used as part of their management. Six case-matched controls were sought. The six women with the greatest requirement for packed red cell transfusion who also had a deranged prothrombin time were included. The groups were then compared for differences. The worst prothrombin time in each group was noted as was the best prothrombin time within 6 h, this was used as our measure of response to treatment.

There was no statistical difference in age, gestation, parity, transfusion requirements, mode of delivery or the severity of the coagulopathy between the two groups. In both groups the prothrombin time improved with management. There was no significant difference in either the magnitude of the improvement in the value of the prothrombin time or the absolute value of the best prothrombin time (P = 0.09). Five out of the six women in the rFVIIa group had normal or low prothrombin times within 6 h yet only one woman who did not receive rFVIIa had a normal prothrombin time within 6 h though this was not significant (P = 0.08).

This case-matched analysis supports the management of massive post-partum haemorrhage with appropriate resuscitation, surgical intervention and use of blood and blood products. This study does not support the routine use of rFVIIa in the management of massive obstetric haemorrhage. rFVIIa may have a role to play in this management but further studies and analyses will be required.

Although the PROWESS trial demonstrated a mortality benefit, subsequent studies in different patient populations have not reproduced the effect. As a result, concerns have been expressed about the clinical effectiveness of drotrecogin alfa (activated). Therefore the aim of this audit was to review the clinical impact of drotrecogin alfa (activated) when used outside clinical trials.

A retrospective review of ICU charts and medical records of patients who had received drotrecogin alfa (activated) in the five largest users of drotrecogin alfa (activated) in England. Patients characteristics details at ICU admission and vital status at hospital discharge were recorded. The severity of illness was assessed by the APACHE II score (using first 24 h admission data) and the number of organ dysfunctions. Adverse incidents were recorded and any sequence effect explored.

In all, 351 patients received drotrecogin alfa (activated) between December 2002 and November 2005. Of those, 201 (57.2%) were male, and 177 (50.4%) were admitted after recent surgery. The patients’ average age was 61.8 yr. The mean admission APACHE II score was 23.3 and the average number of dysfunctional organs on admission was 3.3. The hospital mortality was 46.7% (164 deaths). The expected number of deaths calculated by using the APACHE II risk of death was 173 (49.3%) and by number of sepsis induced organ failures 210 (59.7%). Overall, there were 33 (9.4%) adverse incidents.

Expected mortality derived from both the APACHE II score and organ dysfunctions suggests that drotrecogin alfa (activated) does reduce mortality. Serious adverse incidents occurred in 5.1% patients; however, the direct contributing effect of drotrecogin alfa (activated) cannot be established from this type of audit.