Total hip or knee arthroplasty is associated with significant blood loss. Techniques such as the use of antifibrinolytics or desmopressin, or normovolaemic haemodilution have been used to reduce the need for allogeneic blood transfusion. Tranexamic acid has been used to reduce blood loss and transfusion requirement for total hip and knee arthroplasty, with variable results. This meta-analysis aims to evaluate whether intravenous tranexamic acid, when compared with placebo, reduces blood loss and transfusion requirement in total hip and knee joint replacement surgery and whether it might increase the risk of thromboembolic complications. The literature search was based on MEDLINE, EMBASE, Cochrane Controlled Trials Register, and information from the pharmaceutical company that produces tranexamic acid (Pharmacia-Upjohn). We identified 15 clinical trials and 12 were considered suitable for detailed data extraction. Tranexamic acid reduces the proportion of patients requiring allogeneic blood transfusion (OR 0.16, 95% CI: 0.09-0.26), total amount of blood loss (WMD 460 ml, 95% CI: 274-626 ml), and the total number of units of allogeneic blood transfused (WMD 0.85 unit, 95% CI: 0.36-1.33). Tranexamic acid does not increase the risk of thromboemobolic complications such as deep vein thrombosis, pulmonary embolism, thrombotic cerebral vascular accident, or myocardial infarction (OR 0.98, 95% CI: 0.45-2.12). Intravenous tranexamic acid appears effective and safe in reducing allogeneic blood transfusion and blood loss in total hip and knee arthroplasty.

The ability to accurately adjust for the severity of illness in outcome studies of critically ill patients is essential. Previous studies have showed that Sequential Organ Failure Assessment (SOFA) score and Acute Physiology and Chronic Health Evaluation (APACHE) II score can predict hospital mortality of critically ill patients. The effects of combining these two scores to predict hospital mortality of critically ill patients has not been evaluated. This cohort study evaluated the performance of combining the APACHE II score with SOFA score in predicting hospital mortality of critically ill patients. A total of 1,311 consecutive adult patients admitted to a tertiary 22-bed multidisciplinary intensive care unit (ICU) in Western Australia were considered. The APACHE II, Admission SOFA, Delta SOFA and maximum SOFA score were all related to hospital survival in the univariate analyses. Combining Max SOFA (area under receiver operating characteristic curve 0.875 vs. 0.858, P=0.014; Nagelkerke R 2 : 0.411 vs. 0.371; Brier Score: 0.086 vs. 0.090) or Delta SOFA score (area under receiver operating characteristic curve 0.874 vs. 0.858, P=0.003; Nagelkerke R 2 : 0.412 vs. 0.371; Brier Score: 0.086 vs. 0.090) with the APACHE II score improved the discrimination and overall performance of the predictions when compared with using the APACHE II score alone, especially in the emergency ICU admissions. Combining Max SOFA or Delta SOFA score with the APACHE II score may improve the accuracy of risk adjustment in outcome studies of critically ill patients.

Air embolism is a rare complication of gastrointestinal endoscopy. We present a 56-year-old male who developed both venous and systemic arterial air embolism during an endoscopic retrograde cholangiopancreatogram. Despite early treatment based on clinical diagnosis and confirmation by transthoracic echocardiography, the patient died as a result of severe cerebral ischaemia. Risk factors associated with air embolism in gastrointestinal endoscopic procedures include situations where the mucosa is not intact or where high pressures are generated in the lumen of the gastrointestinal tract. Clinical diagnosis of air embolism during endoscopy is difficult and urgent echocardiography is the investigation of choice. Treatment is largely supportive but hyperbaric oxygen therapy should be considered in any severe cases, especially if neurological injury is present.

Intrathecal midazolam binds with gamma aminobutyric acid-A receptors in the spinal cord leading to an analgesic effect. Clinical studies suggested that intrathecal midazolam may also reduce nausea and vomiting when used as an adjunct to other spinal medications. However, the potential neurotoxic effect of intrathecal midazolam remains a concern. This meta-analysis aims to evaluate the effectiveness and side-effects of intrathecal midazolam in the perioperative and peripartum settings. Thirteen randomised controlled studies from MEDLINE (from 1966 to July 1 2007), EMBASE and Cochrane Controlled Trials Register databases, involving a total of 672 patients, were considered. Volunteer, animal and chronic pain studies were excluded. Adding intrathecal midazolam to other spinal medications reduced the incidence of nausea and vomiting (odds ratio 0.50, 95% confidence interval [CI] 0.27 to 0.90, P=0.02; I 2 =4%) and delayed the time to request for rescue analgesia (weighted-mean-difference=98.7 min, 95% CI: 76.1 to 121.4, P <0.00001; I 2 =98.5%). Intrathecal midazolam did not affect the duration of motor blockade (weighted-mean-difference =25.1 min, 95% CI-7.6 to 57.8, P=0.13, I 2 =94.8%). The incidence of neurological symptoms after intrathecal midazolam was uncommon (1.8%) and did not differ from placebo (odds ratio 1.20, 95% CI 0.22 to 6.68, P=0.84). Based on the limited data available, intrathecal midazolam appears to improve perioperative analgesia and reduce nausea and vomiting during caesarean delivery. A multicentre registry or large randomised controlled study with a prolonged follow-up period would be useful to confirm the clinical safety of intrathecal midazolam.