Adaptive designs in critical care trials: a simulation study.

BACKGROUND: Adaptive clinical trials are growing in popularity as they are more flexible, efficient and ethical than traditional fixed designs. However, notwithstanding their increased use in assessing treatments for COVID-19, their use in critical care trials remains limited. A better understanding of the relative benefits of various adaptive designs may increase their use and interpretation. METHODS: Using two large critical care trials (ADRENAL. CLINICALTRIALS: gov number, NCT01448109. Updated 12-12-2017; NICE-SUGAR. CLINICALTRIALS: gov number, NCT00220987. Updated 01-29-2009), we assessed the performance of three frequentist and two bayesian adaptive approaches. We retrospectively re-analysed the trials with one, two, four, and nine equally spaced interims. Using the original hypotheses, we conducted 10,000 simulations to derive error rates, probabilities of making an early correct and incorrect decision, expected sample size and treatment effect estimates under the null scenario (no treatment effect) and alternative scenario (a positive treatment effect). We used a logistic regression model with 90-day mortality as the outcome and the treatment arm as the covariate. The null hypothesis was tested using a two-sided significance level (α) at 0.05. RESULTS: Across all approaches, increasing the number of interims led to a decreased expected sample size. Under the null scenario, group sequential approaches provided good control of the type-I error rate; however, the type I error rate inflation was an issue for the Bayesian approaches. The Bayesian Predictive Probability and O'Brien-Fleming approaches showed the highest probability of correctly stopping the trials (around 95%). Under the alternative scenario, the Bayesian approaches showed the highest overall probability of correctly stopping the ADRENAL trial for efficacy (around 91%), whereas the Haybittle-Peto approach achieved the greatest power for the NICE-SUGAR trial. Treatment effect estimates became increasingly underestimated as the number of interims increased. CONCLUSIONS: This study confirms the right adaptive design can reach the same conclusion as a fixed design with a much-reduced sample size. The efficiency gain associated with an increased number of interims is highly relevant to late-phase critical care trials with large sample sizes and short follow-up times. Systematically exploring adaptive methods at the trial design stage will aid the choice of the most appropriate method.

Early Magnesium Treatment After Aneurysmal Subarachnoid Hemorrhage: Individual Patient Data Meta-Analysis.

BACKGROUND AND PURPOSE: Delayed cerebral ischemia (DCI) is an important cause of poor outcome after aneurysmal subarachnoid hemorrhage (SAH). Trials of magnesium treatment starting <4 days after symptom onset found no effect on poor outcome or DCI in SAH. Earlier installment of treatment might be more effective, but individual trials had not enough power for such a subanalysis. We performed an individual patient data meta-analysis to study whether magnesium is effective when given within different time frames within 24 hours after the SAH. METHODS: Patients were divided into categories according to the delay between symptom onset and start of the study medication: <6, 6 to 12, 12 to 24, and >24 hours. We calculated adjusted risk ratios with corresponding 95% confidence intervals for magnesium versus placebo treatment for poor outcome and DCI. RESULTS: We included 5 trials totaling 1981 patients; 83 patients started treatment<6 hours. For poor outcome, the adjusted risk ratios of magnesium treatment for start <6 hours were 1.44 (95% confidence interval, 0.83-2.51); for 6 to 12 hours 1.03 (0.65-1.63), for 12 to 24 hours 0.84 (0.65-1.09), and for >24 hours 1.06 (0.87-1.31), and for DCI, <6 hours 1.76 (0.68-4.58), for 6 to 12 hours 2.09 (0.99-4.39), for 12 to 24 hours 0.80 (0.56-1.16), and for >24 hours 1.08 (0.88-1.32). CONCLUSIONS: This meta-analysis suggests no beneficial effect of magnesium treatment on poor outcome or DCI when started early after SAH onset. Although the number of patients was small and a beneficial effect cannot be definitively excluded, we found no justification for a new trial with early magnesium treatment after SAH.

Resuscitation fluid use in Australian and New Zealand Intensive Care Units between 2007 and 2013.

INTRODUCTION: Recent evidence indicates that the choice of intravenous fluids may affect outcomes in critically ill patients. METHODS: We recorded the administration of resuscitation fluids in patients admitted to Australian and New Zealand adult intensive care units (ICUs) for a 24-h period at 6 time points between 2007 and 2013. Changes in patterns of fluid use over this period were determined using regression analyses. RESULTS: Of the 2825 patients admitted to the 61 ICUs on the 6 study days, 754 (26.7%) patients received fluid resuscitation. Of those receiving fluid resuscitation, the proportion of patients receiving crystalloid significantly increased from 28.9% (41/142) in 2007 to 50.5% (48/95) in 2013 (adjusted odds ratio (OR) 2.93; 95% confidence intervals (CI) 1.35-6.33; p = 0.006); of these, the proportion of patients receiving buffered salt solutions significantly increased from 4.9% (7/142) in 2007 to 31.6% (30/95) in 2013 (OR 7.00; 95% CI 2.14-22.92; p = 0.001). The use of colloids significantly decreased from 59.9% (85/142) in 2007 to 42.1% (40/95) in 2013 (adjusted OR 0.34; 95% CI 0.16-0.74; p = 0.007) due to a significant decrease in the proportion of patients receiving gelatin; 28.9% (41/142) to 2.1% (2/95) (OR 0.10; 95% CI 0.03-0.29; p ≤ 0.001). CONCLUSION: Fluid resuscitation practice in Australia and New Zealand adult ICUs has changed over the 6-year study period. Crystalloid use increased primarily due to an increase in the use of buffered salt solutions while overall the use of colloid has decreased.

International albumin use: 1995 to 2006.

Over the last ten years more reliable information regarding the risks and benefits of the use of albumin for fluid resuscitation has emerged. To determine what influence this has had on clinical practice, we sought to document albumin use (from mass of albumin supplied to hospitals) in 16 industrialised countries between 1995 and 2006. Data on national albumin and synthetic colloid use was sought from independent intensive care researchers and albumin issuers. The mass of albumin supplied per 10,000 persons on an annual basis by country and aggregated across the study countries was calculated. Volumes of synthetic colloid supplied per 10,000 persons were calculated. Data were obtained for 15 countries. Albumin use varied significantly between countries and throughout the observation period. Overall, aggregate albumin use decreased from a peak of 2.54 kg per 10,000 persons in 1995 to 1.40 kg per 10,000 persons in 1999; use has remained relatively constant since. Data on supply of synthetic colloids was available in only three countries and varied from 11.7 litres per 10,000 persons in Canada in 1995, to 231.8 litres per 10,000 persons in Denmark in 2004. Between 1995 and 1999 albumin use decreased and has been materially constant since; where data were available, use of synthetic colloids increased. Whether these practice changes have resulted in a net health gain or in harm requires further research.

Intensive insulin treatment.

Hyperglycaemia is common in acute illness and more severe hyperglycaemia is associated with worse outcomes in critically ill patients in general and after acute myocardial infarction, stroke, and trauma. Normalization of blood glucose by intensive insulin therapy has been shown to reduce morbidity and mortality in one study in surgical intensive care patients; a subsequent study in medical intensive care patients resulted in reduced morbidity but not a reduction in mortality. Multicentre studies and current meta-analyses in the critically ill have not demonstrated improved outcomes when normalization of blood glucose was targeted; furthermore all studies to date have detected an increased risk of hypoglycaemia in patients subjected to intensive insulin therapy. At present, universal treatment guidelines or recommendations to target strict normoglycaemia must be considered premature. Further data will be available after the completion of the NICE-SUGAR study which has recruited 6103 patients; the NICE SUGAR study will add significant power to future meta-analyses and may help define the role of intensive insulin therapy in critically ill patients.

Estimate of the number of patients eligible for treatment with drotrecogin alfa (activated) based on differing international indications: post-hoc analysis of an inception cohort study in Australia and New Zealand.

We aimed to estimate the potential number of patients eligible for treatment with drotrecogin alfa (activated) when applying different international criteria. The study was a post-hoc analysis of inception cohort study of 691 patients with severe sepsis during 5878 consecutive intensive care unit admission episodes in 23 closed multi-disciplinary ICUs of 21 hospitals (16 tertiary and 5 university-affiliated) in Australia and New Zealand. Outcomes assessed were presence of contraindications to treatment with drotrecogin alfa (activated), an admission APACHE II score of 25 or greater and dysfunction of two or more organs. During 5878 consecutive intensive care admission episodes, 691 patients had severe sepsis, 553 (80.0%, 95% CI 77.0-83.0%) had no relative or absolute contraindication, 64 (9.3%, 7.1-11.4%) had a relative contraindication and 74 (10.7%, 8.4-13.0%) had an absolute contraindication. Two hundred and six patients (3.5%, 3.0-4.0%) had an APACHE II score of 25 or greater, 452 (7.7%, 7.0-8.4%) had dysfunction of two or more organs, 469 (8.0%, 7.3-8.7%) had either dysfunction of two or more organs or an APACHE II score of 25 or greater Relatively few patients had an absolute contraindication to treatment with drotrecogin alfa (activated). Selection based on the APACHE II score results in fewer eligible patients than selection based on multiple organ dysfunction. Depending on the selection criteria used, for every hundred admissions to intensive care, between 3.5 and 8.0 of patients may be eligible for treatment with drotrecogin alfa (activated).

Life-threatening pulmonary hypertension and right ventricular failure complicating calcium and phosphate replacement in the intensive care unit.

A 43-year-old man developed septic shock and acute lung injury after surgery to drain an ischiorectal abscess. In the intensive care unit he initially improved but developed severe hypoxaemia, right ventricular failure and pulmonary hypertension 90 min after receiving intravenous calcium gluconate and potassium phosphate, best explained by the formation of a calcium-phosphate precipitant that resulted in aggregate anaphylaxis. His rapid deterioration and lack of response to conventional therapies necessitated support with extracorporeal membrane oxygenation that was life saving. This adverse event has altered local practice regarding calcium and phosphate replacement and has implications for all intensive care units.

Impact of operator expertise on collection of the APACHE II score and on the derived risk of death and standardized mortality ratio.

We assessed the impact of operator expertise on collection of the APACHE II score, the derived risk of death and standardized mortality ratio in 465 consecutive patients admitted to a multi-disciplinary tertiary hospital ICU. Research coordinators and junior clinical staff independently collected the APACHE II variables; experts (senior clinical staff) rescored 20% of the records. Agreement was moderate between junior clinical staff and research coordinators or senior clinical staff for most variables of the acute physiology score (weighted kappa<0.6); agreement between research coordinators and senior clinical staff data collectors was good (weighted kappa >0.75). The APACHE II score and its derived risk of death (ROD) were significantly lower using the junior clinical staff dataset compared to research coordinators and senior clinical staff (APACHE II score: 13.4+/-9.2 vs 16.8+/-8.5 vs 17.1+/-7 7, P<0.001; ROD: 14.7%+/-22.4% vs 21.6%+/-22.6% vs 20.8%+/-22.4%, P<0.01 respectively). The discriminative capacity was not altered by the lack of agreement (area under Receiver Operator Characteristic curve >0.8) but calibration of ROD from the junior clinical staff dataset was poor (Goodness-of-fit: P= 0.001). The standardized mortality ratio (SMR) was higher with the junior clinical staff dataset (SMR: 1.22, 95% CI: 0.96-1.52 vs 0.87, 95% CI: 0.70-1.06 vs 0.76, 95% CI: 0.40-1.3 calculated from junior clinical staff research coordinators and senior clinical staff datasets respectively). We conclude that the expertise of data collectors significantly influences the APACHE II score, the derived risk of death and the standardized mortality ratio. Given the importance of such scores, ICUs should be provided with sufficient resources to train and employ dedicated data collectors.

Improving patients' safety locally: changing clinician behaviour.

Safety initiatives in hospitals should focus on common health care interventions that when used appropriately can improve important health outcomes, and when used inappropriately or not at all, result in substantial harm. We suggest that errors of omission should be a safety priority. We focus on preventive health care interventions, and describe five steps that can improve patients' safety by changing clinician behaviour. The steps are to: do an environmental scan; understand current behaviour, target behaviour for change (why, what, when, where, and who); adopt effective strategies to change behaviour; and synergise.

Papaverine angioplasty to treat cerebral vasospasm following traumatic subarachnoid haemorrhage.

The management of vasospasm associated with traumatic subarachnoid haemorrhage presents many challenges. We present a 20-year-old male admitted after sustaining a closed head injury complicated by a Fisher grade III traumatic subarachnoid haemorrhage. Despite treatment with intravenous nimodipine he developed a delayed ischaemic neurological deficit due to cerebral arterial vasospasm. The vasospasm was successfully managed with serial papaverine angioplasty.

Prolonged thiopentone infusion for neurosurgical emergencies: usefulness of therapeutic drug monitoring.

Serial serum thiopentone concentrations were measured during and following completion of an intravenous infusion of thiopentone in 20 patients with neurosurgical emergencies. The concentration data from a further 55 patients who had had some such measurements were reviewed retrospectively. The patients received an infusion for longer than 24 hours at a rate adjusted to maintain EEG burst suppression. The data were interpreted in terms of thiopentone pharmacokinetics and used to produce statistical models relating to clinical outcomes. In these patients, the one-month mortality rate following commencement of thiopentone treatment was 20%; the mean durations of pupillary and motor unresponsiveness following cessation of an infusion were 22 and 91 hours, respectively. Predictors of a prolonged duration of motor unresponsiveness included a prolonged duration of pupillary unresponsiveness, a low thiopentone clearance and a high maximum serum concentration of thiopentone. From pooled logistic regression, median effective serum thiopentone concentrations (EC50) were found to be 50 mg x l(-1) for recovery of pupillary responsiveness and 12 mg x l(-1) for the recovery of motor responsiveness. Because prolonged high-dose thiopentone leads to prolonged residual serum concentrations, it is difficult to distinguish the residual pharmacological effects of thiopentone from the clinical condition. This study suggests that, based on EC50 values for responses, monitoring of post-infusion serum thiopentone concentrations may help determine whether a patient's clinical state is due to residual thiopentone pharmacological effects.

Severe traumatic brain injury.

In western countries, injuries remain the leading cause of death in young adults (Jennett B. Epidemiology of head injury. J Neurol Neurosurg Psychiatry 1996; 60: 362-369). Worldwide, injuries are estimated to account for 15% of the burden of death and disability, and are projected to account for 20% in 2020 (Ad Hoc Committee on Health Research Relating to Future Intervention Options. Investing in Health Research and Development (Document TDR/Gen/96.1). Geneva: World Health Organisation, 1996). In developing countries road traffic injuries in particular are increasing in incidence and injuries are projected to be the third leading cause of death and disability worldwide by 2020 (Ad Hoc Committee on Health Research Relating to Future Intervention Options. Investing in Health Research and Development (Document TDR/Gen/96.1). Geneva: World Health Organisation, 1996). Head injury accounts for up to half of all deaths from trauma (Kraus J. Epidemiology of head injury. In: Cooper PR, Ed. Head Injury, 3rd ed. Baltimore, MD: William Wilkins, 1993), and in addition to causing death often causes severe and long-lasting functional impairment in survivors.

Aggressive management of aneurysmal subarachnoid haemorrhage based on a papaverine angioplasty protocol.

The aim of this study was to analyse the results of intensive therapy unit management of aneurysmal subarachnoid haemorrhage incorporating angioplasty in the protocol. Two hundred consecutive patients were treated using a detailed protocol that included nimodipine, early aneurysm repair, and surveillance angiography. Angiography was performed on days 5 to 7 (or when the clinical state suggested the presence of vasospasm). If angiographic vasospasm was identified, irrespective of whether clinical vasospasm was present or absent, papaverine was selectively administered. In patients with vasospasm blood pressure was elevated to 160-180 mmHg and selective papaverine administration was repeated daily until vasospasm resolved. In cases requiring more frequent administration of papaverine, or in whom papaverine failed to adequately reverse spasm, balloon angioplasty was considered and for clinically refractory cases barbiturate coma was introduced. 43% of patients underwent papaverine administration and of these the average number of separate papaverine procedures was four (maximum 23). 26% of patients developed neurological deficits though to be due to vasospasm whilst 17% underwent papaverine angioplasty without clinical signs of vasospasm. Twelve patients (6%) were entered into barbiturate coma. There was a 5.5% mortality and no difference in outcome between patients who developed angiographic vasospasm and those who did not. For those developing clinical vasospasm, 71% were independent and 10% were dead at follow up compared with 84% reaching independent grades and 4% dead in those not developing clinical vasospasm. These differences failed to reach a significant difference. The average Intensive Therapy Unit stay for aneurysmal subarachnoid haemorrhage patients was 13.1 days with a mean cost to the hospital of $AUD 24,379. This protocol appears to be both a clinically and cost effective method of managing aneurysmal subarachnoid haemorrhage.