How should cost data in pragmatic randomised trials be analysed?

By S. Thompson, J. Barber. British medical journal, 2000

818 citations recorded by the source corpus.

Pragmatic randomised trials are usually large scale multicentre studies in which interventions or medical policies are compared in a realistic setting.1 The intention is that conclusions from these trials, if accepted, can be adopted directly into medical practice.2 Economic evaluations carried out alongside these trials are increasingly common because it is often important to assess costs and cost effectiveness as well as clinical outcomes.3 Costs are usually derived from information about the quantity of healthcare resources used by each patient in the trial. The quantities of each resource used are multiplied by fixed unit cost values and are then summed over the separate types of resource to give a total cost per patient.4 Distribution of costs from a trial comparing endometrial resection with hysterectomy in women with menorrhagia. Costs are based on health resource use from randomisation to two years; they include preoperative, operative, hospital stay, complications, retreatment, and primary care components5 This information leads to a range of different costs across participants in the trial. As an example, the figure shows the distribution of costs in women with menorrhagia randomised to treatment with endometrial resection or abdominal hysterectomy.5 Such highly skewed distributions are typical of cost data; the long right hand tail reflects the fact that some patients incur high costs because of factors such as medical complications, reoperation, or extended hospital stay. ### Summary points Health economic evaluations are now commonly included in pragmatic clinical trials that inform policy decisions Despite the usual skewness in the distribution of costs, it is the arithmetic mean that is the most informative measure Measures other than the arithmetic mean do not provide information about the cost of treating all patients, which is needed as the basis for healthcare policy decisions Statistical analysis based on transforming cost data or …

Subfields: Economics and Econometrics, General Health Professions

View Paper: https://www.semanticscholar.org/paper/d7f4c47d94bcd5eb60f5e502e87f935e2e73f3d0

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Economics and Econometrics

In General Health Professions

Citations in the corpus

Mindfulness-based cognitive therapy to prevent relapse in recurrent depression. (2008, background): “Despite the skewed nature of cost data, this approach is recommended to enable inferences to be made about the arithmetic mean ( Thompson & Barber, 2000 ), a more meaningful summary statistic for costs than is the median or geometric mean.”
Early goal-directed therapy after major surgery reduces complications and duration of hospital stay. A randomised, controlled trial [ISRCTN38797445] (2005, methodology): “Confidence intervals were constructed for the difference in mean duration of stay between the two groups by bootstrapping within treatment groups [17].”
Trial of early, goal-directed resuscitation for septic shock. (2015, background): “Salford Royal Hospital Non-teaching January 2012 - July 2014 661 88000 53 27 ED (22)/CCU (19)/Ward (4) 17 (63.”
Essential Medical Statistics (2003, methodology): “An example of the use of bootstrapping is provided by Thompson and Barber (2000), who consider the analysis of data on costs of treatment in clinical trials.”
Review of Statistical Methods for Analysing Healthcare Resources and Costs (2010, methodology): “Two common approaches to directly evaluate uncertainty in costs, effects and/or cost-effectiveness in randomised trials are the Central Limit Theorem and the bootstrap approach (Barber and Thompson, 2000; O’Hagan and Stevens, 2003; Thompson and Barber, 2000).”
A Bootstrap Test of Stochastic Equality of Two Populations (2005, background): “One should test that particular hypothesis, which fits to the specific biological question of interest, and should not change it for purely statistical reasons (Thompson and Barber 2000; Zhou et al. 2001).”
Parametric versus non-parametric statistics in the analysis of randomized trials with non-normally distributed data (2005, neutral): “A good example comes from health economics [18]: we want to know the difference between the mean costs of two treatments because multiplying this difference by the number of patients we expect to treat gives us the expected financial impact of choosing one treatment over the other; the difference in”