The analysis of very small samples of repeated measurements II: A modified Box correction

Skene, Simon; Kenward, Michael G.

doi:10.1002/sim.4072

Cited by 17 publications

(43 citation statements)

References 24 publications

(37 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…the actual difference with age on this limited sample) and required sample size to avoid Type II error. In addition, according to the small sample inference of the linear mixed model [36], our result needs to be verified on larger sample size. Despite limitation about the statistical inference, the present result about MA opens new insight in the field of pain evaluation for osteoarthritic dogs.…”

Section: Discussionmentioning

confidence: 99%

Clinical validity of outcome pain measures in naturally occurring canine osteoarthritis

Rialland

Bichot²,

Moreau

et al. 2012

BMC Vet Res

View full text Add to dashboard Cite

BackgroundThe conceptual validity of kinetic gait analysis and disability outcome assessment methods has guided their use in the assessment of pain caused by osteoarthritis (OA). No consensus on the best clinical methods for pain evaluation in canine OA exists, particularly, when evaluating treatments where a smaller treatment effect is anticipated than with pharmacological pain killers. This study thus aimed at determining the technical validity of some clinical endpoints on OA pain in dogs using the green-lipped mussel (GLM)-enriched diet.Twenty-three adult dogs with clinical OA completed the prospective controlled study. All the dogs were fed a balanced diet over a 30-day control period followed by a GLM-enriched diet over a 60-day period. The kinetic gait analysis parameter (PVFBW, peak vertical force adjusted for body weight change), electrodermal activity (EDA), and a standardized multifactorial pain questionnaire (MFQ) were performed on day (D) 0 (inclusion), D30 (start) and D90 (end). The owners completed a client-specific outcome measures (CSOM) instrument twice a week. Motor activity (MA) was continuously recorded in seven dogs using telemetered accelerometric counts. We hypothesized that these methods would produce convergent results related to diet changes. A Type I error of 0.05 was adjusted to correct for the multiplicity of the primary clinical endpoints.ResultsNeither the EDA nor the MFQ were found reliable or could be validated. Changes in the PVFBW (Padj = 0.0004), the CSOM (Padj = 0.006) and the MA intensity (Padj = 0.02) from D0 to D90 suggested an effect of diet(s). Only the PVFBW clearly increased after the GLM-diet (Padj = 0.003). The CSOM exhibited a negative relationship with the PVFBW (P = 0.02) and MA duration (P = 0.02).ConclusionsThe PVFBW exhibited the best technical validity for the characterization of the beneficial effect of a GLM-enriched diet. The CSOM and MA appeared less responsive following a GLM-diet, but these measures appeared complementary to gait analysis. Apparently, the CSOM provides the capacity to rely on pain OA assessment influenced by both lameness quantification (PVFBW) and physical functioning (MA).

show abstract

Section: Discussionmentioning

confidence: 99%

Clinical validity of outcome pain measures in naturally occurring canine osteoarthritis

Rialland

Bichot²,

Moreau

et al. 2012

BMC Vet Res

View full text Add to dashboard Cite

show abstract

“…The purpose of this paper and the following companion paper [1] is to examine methods of analysis which are appropriate for very small samples of repeated measurements. These issues are highlighted in Brammer [2], who suggests using the general linear mixed model as an appropriate modelling framework for ascending dose design studies, with reference to two experiments involving isolated tissues and organs from guinea pigs and rats.…”

Section: Introductionmentioning

confidence: 99%

The analysis of very small samples of repeated measurements I: An adjusted sandwich estimator

Skene

Kenward

2010

Statistics in Medicine

Self Cite

View full text Add to dashboard Cite

The statistical analysis of repeated measures or longitudinal data always requires the accommodation of the covariance structure of the repeated measurements at some stage in the analysis. The general linear mixed model is often used for such analyses, and allows for the specification of both a mean model and a covariance structure. Often the covariance structure itself is not of direct interest, but only a means to producing valid inferences about the response. Existing methods of analysis are often inadequate where the sample size is small. More precisely, statistical measures of goodness of fit are not necessarily the right measure of the appropriateness of a covariance structure and inferences based on conventional Wald type procedures do not approximate sufficiently well their nominal properties when data are unbalanced or incomplete. This is shown to be the case when adopting the Kenward-Roger adjustment where the sample size is very small. A generalization of an approach to Wald tests using a bias adjusted empirical sandwich estimator for the covariance matrix of the fixed effects from generalized estimating equations is developed for Gaussian repeated measurements. This is shown to attain the correct test size but has very low power.

show abstract

“…Hence, the MLM gives higher weight to the means from the smaller clusters, which in the control group happened to have relatively low mean costs, thus reducing the overall costs in the control group. Although improved robust estimators have been proposed for setting with few clusters (Pan and Wall, 2002;Skene and Kenward, 2010), it is unclear how they would perform with unequal cluster sizes. Figure 2.…”

Section: Discussionmentioning

confidence: 99%

Methods for Covariate Adjustment in Cost‐effectiveness Analysis That Use Cluster Randomised Trials

et al. 2012

View full text Add to dashboard Cite

Statistical methods have been developed for cost-effectiveness analyses of cluster randomised trials (CRTs) where baseline covariates are balanced. However, CRTs may show systematic differences in individual and cluster-level covariates between the treatment groups. This paper presents three methods to adjust for imbalances in observed covariates: seemingly unrelated regression with a robust standard error, a 'two-stage' bootstrap approach combined with seemingly unrelated regression and multilevel models. We consider the methods in a cost-effectiveness analysis of a CRT with covariate imbalance, unequal cluster sizes and a prognostic relationship that varied by treatment group. The cost-effectiveness results differed according to the approach for covariate adjustment. A simulation study then assessed the relative performance of methods for addressing systematic imbalance in baseline covariates. The simulations extended the case study and considered scenarios with different levels of confounding, cluster size variation and few clusters. Performance was reported as bias, root mean squared error and CI coverage of the incremental net benefit. Even with low levels of confounding, unadjusted methods were biased, but all adjusted methods were unbiased. Multilevel models performed well across all settings, and unlike the other methods, reported CI coverage close to nominal levels even with few clusters of unequal sizes.

show abstract

The analysis of very small samples of repeated measurements II: A modified Box correction

Cited by 17 publications

References 24 publications

Clinical validity of outcome pain measures in naturally occurring canine osteoarthritis

Clinical validity of outcome pain measures in naturally occurring canine osteoarthritis

The analysis of very small samples of repeated measurements I: An adjusted sandwich estimator

Methods for Covariate Adjustment in Cost‐effectiveness Analysis That Use Cluster Randomised Trials

Contact Info

Product

Resources

About