Statistical Analysis of Noncommensurate Multiple Outcomes

Reported Outcomes in Interdisciplinary Pain Treatment: An Overview of Systematic Reviews and Meta-Analyses of Randomised Controlled Trials

Background

There is considerable diversity of outcome selections and methodologies for handling the multiple outcomes across all systematic reviews (SRs) of Interdisciplinary Pain Treatment (IPT) due to the complexity. This diversity presents difficulties for healthcare decision makers. Better recommendations about how to select outcomes in SRs (with or without meta-analysis) are needed to explicitly demonstrate the effectiveness of IPT.

Objective

This overview systematically collates the reported outcomes and measurements of IPT across published SRs and identifies the methodological characteristics. Additionally, we provide some suggestions on framing the selection of outcomes and on conducting SRs of IPT.

Methods

Three electronic databases (PubMed, Cochrane Library, and Epistemonikos) and the PROSPERO registry for ongoing SR were supplemented with hand-searching ending on 30 September 2021.

Results

We included 18 SRs with data on 49007 people from 356 primary randomised controlled trials (RCTs); eight were followed by meta-analysis and ten used narrative syntheses of data. For all the SRs, pain was the most common reported outcome (72%), followed by disability/functional status (61%) and working status (61%). Psychological well-being and quality of life were also reported in half of the included SR (50%). The core outcome domains according to VAPAIN, IMMPACT, and PROMIS were seldom met. The methodological quality varied from critically low to moderate according to AMSTAR2. The AMSTAR2 rating was negatively correlated to the number of outcome domains in PROMIS, and VAPAIN was positively correlated with IMMPACT and PROMIS, indicating the intercorrelations between the reported outcomes.

Conclusion

This systematic overview showed wide-ranging disparity in reported outcomes and applied outcome domains in SRs evaluating IPT interventions for chronic pain conditions. The intercorrelations between the reported outcomes should be appropriately handled in future research. Some approaches are discussed as well.

Bayesian analysis of longitudinal quality of life measures with informative missing data using a selection model

Health-related quality of life consists of multi-dimensional measurements of physical and mental health domains. Health-related quality of life is often followed up to evaluate efficacy of treatments in clinical studies. During the follow-up period, a missing data problem inevitably arises. When missing data occur for reasons related to poor health-related quality of life, a complete-case only analysis can lead to invalid inferences. We propose a Bayesian approach to analyze longitudinal moderate to high-dimensional multivariate outcome data in the presence of non-ignorable missing data. To account for non-ignorable missing data, we employ a selection model for the joint likelihood factorization where we apply Bayesian spike and slab variable selection in the missing data mechanism to detect informative factors among multiple outcomes. We model the relationship between multiple outcomes and covariates using linear mixed effects models where multiple outcome correlations are captured by a hierarchical structure. We conduct simulation studies to evaluate the performance of the proposed method compared with the conventional last observation carried forward approach. We use a motivating example that originates from a longitudinal study of quality of life in gastric cancer patients who underwent distal gastrectomy. In this application, we demonstrate that our proposed method can offer efficiency gain in the marginal associations and provide the associations between outcomes and the absence of patients' information.

Moderate and Stable Pain Reductions as a Result of Interdisciplinary Pain Rehabilitation-A Cohort Study from the Swedish Quality Registry for Pain Rehabilitation (SQRP)

Few studies have investigated the real-life outcomes of interdisciplinary multimodal pain rehabilitation programs (IMMRP) for chronic pain. This study has four aims: investigate effect sizes (ES); analyse correlation patterns of outcome changes; define a multivariate outcome measure; and investigate whether the clinical self-reported presentation pre-IMMRP predicts the multivariate outcome. To this end, this study analysed chronic pain patients in specialist care included in the Swedish Quality Registry for Pain Rehabilitation for 22 outcomes (pain, psychological distress, participation, and health) on three occasions: pre-IMMRP, post-IMMRP, and 12-month follow-up. Moderate stable ES were demonstrated for pain intensity, interference in daily life, vitality, and health; most other outcomes showed small ES. Using a Multivariate Improvement Score (MIS), we identified three clusters. Cluster 1 had marked positive MIS and was associated with the overall worst situation pre-IMMRP. However, the pre-IMMRP situation could only predict 8% of the variation in MIS. Specialist care IMPRPs showed moderate ES for pain, interference, vitality, and health. Outcomes were best for patients with the worst clinical presentation pre-IMMRP. It was not possible to predict who would clinically benefit most from IMMRP.

A guide to missing data for the pediatric nephrologist

Missing data is an important and common source of bias in clinical research. Readers should be alert to and consider the impact of missing data when reading studies. Beyond preventing missing data in the first place, through good study design and conduct, there are different strategies available to handle data containing missing observations. Complete case analysis is often biased unless data are missing completely at random. Better methods of handling missing data include multiple imputation and models using likelihood-based estimation. With advancing computing power and modern statistical software, these methods are within the reach of clinician-researchers under guidance of a biostatistician. As clinicians reading papers, we need to continue to update our understanding of statistical methods, so that we understand the limitations of these techniques and can critically interpret literature.

The use of echocardiographic indices in defining and assessing right ventricular systolic function in critical care research

PURPOSE

Many echocardiographic indices (or methods) for assessing right ventricular (RV) function are available, but each has its strengths and limitations. In some cases, there might be discordance between the indices. We conducted a systematic review to audit the echocardiographic RV assessments in critical care research to see if a consistent pattern existed. We specifically looked into the kind and number of RV indices used, and how RV dysfunction was defined in each study.

METHODS

Studies conducted in critical care settings and reported echocardiographic RV function indices from 1997 to 2017 were searched systematically from three databases. Non-adult studies, case reports, reviews and secondary studies were excluded. These studies' characteristics and RV indices reported were summarized.

RESULTS

Out of 495 non-duplicated publications found, 81 studies were included in our systematic review. There has been an increasing trend of studying RV function by echocardiography since 2001, and most were conducted in ICU. Thirty-one studies use a single index, mostly TAPSE, to define RV dysfunction; 33 used composite indices and the combinations varied between studies. Seventeen studies did not define RV dysfunction. For those using composite indices, many did not explain their choices.

CONCLUSIONS

TAPSE seemed to be the most popular index in the last 2-3 years. Many studies used combinations of indices but, apart from cor pulmonale, we could not find a consistent pattern of RV assessment and definition of RV dysfunction amongst these studies.

Estimating scaled treatment effects with multiple outcomes

In classical study designs, the aim is often to learn about the effects of a treatment or intervention on a single outcome; in many modern studies, however, data on multiple outcomes are collected and it is of interest to explore effects on multiple outcomes simultaneously. Such designs can be particularly useful in patient-centered research, where different outcomes might be more or less important to different patients. In this paper, we propose scaled effect measures (via potential outcomes) that translate effects on multiple outcomes to a common scale, using mean-variance and median-interquartile range based standardizations. We present efficient, nonparametric, doubly robust methods for estimating these scaled effects (and weighted average summary measures), and for testing the null hypothesis that treatment affects all outcomes equally. We also discuss methods for exploring how treatment effects depend on covariates (i.e., effect modification). In addition to describing efficiency theory for our estimands and the asymptotic behavior of our estimators, we illustrate the methods in a simulation study and a data analysis. Importantly, and in contrast to much of the literature concerning effects on multiple outcomes, our methods are nonparametric and can be used not only in randomized trials to yield increased efficiency, but also in observational studies with high-dimensional covariates to reduce confounding bias.

Weak outcome predictors of multimodal rehabilitation at one-year follow-up in patients with chronic pain-a practice based evidence study from two SQRP centres

BACKGROUND

For patients with chronic pain, the heterogeneity of clinical presentations makes it difficult to identify patients who would benefit from multimodal rehabilitation programs (MMRP). Yet, there is limited knowledge regarding the predictors of MMRP's outcomes. This study identifies predictors of outcome of MMRPs at a 12-month follow-up (FU-12) based on data from the Swedish Quality Registry for Pain Rehabilitation (SQRP).

METHODS

Patients with chronic pain from two clinical departments in Sweden completed the SQRP questionnaires-background, pain characteristics, psychological symptoms, function, activity/participation, health and quality of life-on three occasions: 1) during their first visit; 2) immediately after the completion of their MMRP; and 3) 12 months after completing the MMRP (n = 227). During the FU-12, the patients also retrospectively reported their global impressions of any changes in their perception of pain and their ability to handle their life situation in general.

RESULTS

Significant improvements were found for pain, psychological symptoms, activity/participation, health, and quality of life aspects with low/medium strong effects. A general pattern was observed from the analyses of the changes from baseline to FU-12; the largest improvements in outcomes were significantly associated with poor situations according to their respective baseline scores. Although significant regressors of the investigated outcomes were found, the significant predictors were weak and explained a minor part of the variation in outcomes (15-25%). At the FU-12, 53.6% of the patients reported that their pain had decreased and 80.1% reported that their life situation in general had improved. These improvements were associated with high education, low pain intensity, high health level, and work importance (only pain perception). The explained variations were low (9-11%).

CONCLUSIONS

Representing patients in real-world clinical settings, this study confirmed systematic reviews that outcomes of MMRP are associated with broad positive effects. A mix of background and baseline variables influenced the outcomes investigated, but the explained variations in outcomes were low. There is still a need to develop standardized and relatively simple outcomes that can be used to evaluate MMRP in trials, in clinical evaluations at group level, and for individual patients.

Comparative analysis of outcome measures used in examining neurodevelopmental effects of early childhood anesthesia exposure

INTRODUCTION

Immature animals exposed to anesthesia display apoptotic neurodegeneration and neurobehavioral deficits. The safety of anesthetic agents in children has been evaluated using a variety of neurodevelopmental outcome measures with varied results.

METHODS

The authors used data from the Western Australian Pregnancy Cohort (Raine) Study to examine the association between exposure to anesthesia in children younger than 3 yr of age and three types of outcomes at age of 10 yr: neuropsychological testing, International Classification of Diseases, 9th Revision, Clinical Modification-coded clinical disorders, and academic achievement. The authors' primary analysis was restricted to children with data for all outcomes and covariates from the total cohort of 2,868 children born from 1989 to 1992. The authors used a modified multivariable Poisson regression model to determine the adjusted association of anesthesia exposure with outcomes.

RESULTS

Of 781 children studied, 112 had anesthesia exposure. The incidence of deficit ranged from 5.1 to 7.8% in neuropsychological tests, 14.6 to 29.5% in International Classification of Diseases, 9th Revision, Clinical Modification-coded outcomes, and 4.2 to 11.8% in academic achievement tests. Compared with unexposed peers, exposed children had an increased risk of deficit in neuropsychological language assessments (Clinical Evaluation of Language Fundamentals Total Score: adjusted risk ratio, 2.47; 95% CI, 1.41 to 4.33, Clinical Evaluation of Language Fundamentals Receptive Language Score: adjusted risk ratio, 2.23; 95% CI, 1.19 to 4.18, and Clinical Evaluation of Language Fundamentals Expressive Language Score: adjusted risk ratio, 2.00; 95% CI, 1.08 to 3.68) and International Classification of Diseases, 9th Revision, Clinical Modification-coded language and cognitive disorders (adjusted risk ratio, 1.57; 95% CI, 1.18 to 2.10), but not academic achievement scores.

CONCLUSIONS

When assessing cognition in children with early exposure to anesthesia, the results may depend on the outcome measure used. Neuropsychological and International Classification of Diseases, 9th Revision, Clinical Modification-coded clinical outcomes showed an increased risk of deficit in exposed children compared with that in unexposed children, whereas academic achievement scores did not. This may explain some of the variation in the literature and underscores the importance of the outcome measures when interpreting studies of cognitive function.