Jonathan Ball

Statistics review 13: Receiver operating characteristic curves

This review introduces some commonly used methods for assessing the performance of a diagnostic test. The sensitivity, specificity and likelihood ratio of a test are discussed. The uses of the receiver operating characteristic curve and the area under the curve are explained.

Statistics review 12: Survival analysis

This review introduces methods of analyzing data arising from studies where the response variable is the length of time taken to reach a certain end-point, often death. The Kaplan–Meier methods, log rank test and Coxs proportional hazards model are described.

Clinical review: Goal-directed therapy-what is the evidence in surgical patients? The effect on different risk groups

Patients with limited cardiac reserve are less likely to survive and develop more complications following major surgery. By augmenting oxygen delivery index (DO2I) with a combination of intravenous fluids and inotropes (goal directed therapy (GDT)), postoperative mortality and morbidity of high-risk patients may be reduced. However, although most studies suggest that GDT may improve outcome in high-risk surgical patients, it is still not widely practiced. We set out to test the hypothesis that GDT results in greatest benefit in terms of mortality and morbidity in patients with the highest risk of mortality and have undertaken a systematic review of the current literature to see if this is correct. We performed a systematic search of Medline, Embase and CENTRAL databases for randomized controlled trials (RCTs) and reviews of GDT in surgical patients. To minimize heterogeneity we excluded studies involving cardiac, trauma, and paediatric surgery. Extremely high risk, high risk and intermediate risks of mortality were defined as >20%, 5 to 20% and <5% mortality rates in the control arms of the trials, respectively. Meta analyses were performed and Forest plots drawn using RevMan software. Data are presented as odd ratios (OR; 95% confidence intervals (CI), and P-values). A total of 32 RCTs including 2,808 patients were reviewed. All studies reported mortality. Five studies (including 300 patients) were excluded from assessment of complication rates as the number of patients with complications was not reported. The mortality benefit of GDT was confined to the extremely high-risk group (OR = 0.20, 95% CI 0.09 to 0.41; P < 0.0001). Complication rates were reduced in all subgroups (OR = 0.45, 95% CI 0.34 to 0.60; P < 0.00001). The morbidity benefit was greatest amongst patients in the extremely high-risk subgroup (OR = 0.27, 95% CI 0.15 to 0.51; P < 0.0001), followed by the intermediate risk subgroup (OR = 0.43, 95% CI 0.27 to 0.67; P = 0.0002), and the high-risk subgroup (OR 0.56, 95% CI 0.36 to 0.89; P = 0.01). Despite heterogeneity in trial quality and design, we found GDT to be beneficial in all high-risk patients undergoing major surgery. The mortality benefit of GDT was confined to the subgroup of patients at extremely high risk of death. The reduction of complication rates was seen across all subgroups of GDT patients.

Statistics review 9: One-way analysis of variance

This review introduces one-way analysis of variance, which is a method of testing differences between more than two groups or treatments. Multiple comparison procedures and orthogonal contrasts are described as methods for identifying specific differences between pairs of treatments.

Statistics review 10: Further nonparametric methods

This review introduces nonparametric methods for testing differences between more than two groups or treatments. Three of the more common tests are described in detail, together with multiple comparison procedures for identifying specific differences between pairs of groups.

Statistics review 8: Qualitative data – tests of association

This review introduces methods for investigating relationships between two qualitative (categorical) variables. The χ2 test of association is described, together with the modifications needed for small samples. The test for trend, in which at least one of the variables is ordinal, is also outlined. Risk measurement is discussed. The calculation of confidence intervals for proportions and differences between proportions are described. Situations in which samples are matched are considered.

Comfort and patient-centred care without excessive sedation: the eCASH concept

We propose an integrated and adaptable approach to improve patient care and clinical outcomes through analgesia and light sedation, initiated early during an episode of critical illness and as a priority of care. This strategy, which may be regarded as an evolution of the Pain, Agitation and Delirium guidelines, is conveyed in the mnemonic eCASH—early Comfort using Analgesia, minimal Sedatives and maximal Humane care. eCASH aims to establish optimal patient comfort with minimal sedation as the default presumption for intensive care unit (ICU) patients in the absence of recognised medical requirements for deeper sedation. Effective pain relief is the first priority for implementation of eCASH: we advocate flexible multimodal analgesia designed to minimise use of opioids. Sedation is secondary to pain relief and where possible should be based on agents that can be titrated to a prespecified target level that is subject to regular review and adjustment; routine use of benzodiazepines should be minimised. From the outset, the objective of sedation strategy is to eliminate the use of sedatives at the earliest medically justifiable opportunity. Effective analgesia and minimal sedation contribute to the larger aims of eCASH by facilitating promotion of sleep, early mobilization strategies and improved communication of patients with staff and relatives, all of which may be expected to assist rehabilitation and avoid isolation, confusion and possible long-term psychological complications of an ICU stay. eCASH represents a new paradigm for patient-centred care in the ICU. Some organizational challenges to the implementation of eCASH are identified.

Statistics review 2: Samples and populations

The previous review in this series introduced the notion of data description and outlined some of the more common summary measures used to describe a dataset. However, a dataset is typically only of interest for the information it provides regarding the population from which it was drawn. The present review focuses on estimation of population values from a sample.

Statistics review 11: assessing risk.

Relative risk and odds ratio have been introduced in earlier reviews (see Statistics reviews 3, 6 and 8). This review describes the calculation and interpretation of their confidence intervals. The different circumstances in which the use of either the relative risk or odds ratio is appropriate and their relative merits are discussed. A method of measuring the impact of exposure to a risk factor is introduced. Measures of the success of a treatment using data from clinical trials are also considered.

An under-recognized complication of treatment of acute severe asthma

A 39-year-old man presented to the emergency department (ED) in severe respiratory distress. He had a prior diagnosis of brittle asthma and had been admitted on several occasions but never previously ventilated. Therapy given in the first 3 hours of arrival included nebulized salbutamol (5 mg, x5), ipratropium bromide (0.5 mg), intravenous hydrocortisone (200 mg), and magnesium sulfate (2 g). His arterial blood gases continued to deteriorate. He was then given an intravenous bolus of salbutamol (250 microg) and heliox via facemask. His worsening status necessitated invasive ventilation. His hypercapnia and resultant respiratory acidosis improved rapidly, but there was a concurrent accumulation of lactic acid resulting in acidemia. This patient had lactic acidosis as a direct effect of administration of salbutamol. The development of hazardous salbutamol-induced toxicity in acute severe asthma is discussed.