Using the random forest method to detect a response shift in the quality of life of multiple sclerosis patients: a cohort study

Predicting the conversion from clinically isolated syndrome to multiple sclerosis: An explainable machine learning approach

INTRODUCTION

Predicting the conversion of clinically isolated syndrome (CIS) to clinically definite multiple sclerosis (CDMS) is critical to personalizing treatment planning and benefits for patients. The aim of this study is to develop an explainable machine learning (ML) model for predicting this conversion based on demographic, clinical, and imaging data.

METHOD

The ML model, Extreme Gradient Boosting (XGBoost), was employed on the public dataset of 273 Mexican mestizo CIS patients with 10-year follow-up. The data was divided into a training set for cross-validation and feature selection, and a holdout test set for final testing. Feature importance was determined using the SHapley Additive Explanations library (SHAP). Then, two experiments were conducted to optimize the model's performance by selectively adding variables and selecting the most contributive variables for the final model.

RESULTS

Nine variables including age, gender, schooling, motor symptoms, infratentorial and periventricular lesion at imaging, oligoclonal band in cerebrospinal fluid, lesion and symptoms types were significant. The model achieved an accuracy of 83.6 %, AUC of 91.8 %, sensitivity of 83.9 %, and specificity of 83.4 % in cross-validation. In the final testing, the model achieved an accuracy of 78.3 %, AUC of 85.8 %, sensitivity of 75 %, and specificity of 81.1 %. Finally, a web-based demo of the model was created for testing purposes.

CONCLUSION

The model, focusing on feature selection and interpretability, effectively stratifies risk for treatment decisions and disability prevention in MS patients. It provides a numerical risk estimate for CDMS conversion, enhancing transparency in clinical decision-making and aiding in patient care.

If it's information, it's not "bias": a scoping review and proposed nomenclature for future response-shift research

BACKGROUND

The growth in response-shift methods has enabled a stronger empirical foundation to investigate response-shift phenomena in quality-of-life (QOL) research; but many of these methods utilize certain language in framing the research question(s) and interpreting results that treats response-shift effects as "bias," "noise," "nuisance," or otherwise warranting removal from the results rather than as information that matters. The present project will describe the various ways in which researchers have framed the questions for investigating response-shift issues and interpreted the findings, and will develop a nomenclature for such that highlights the important information about resilience reflected by response-shift findings.

METHODS

A scoping review was done of the QOL and response-shift literature (n = 1100 articles) from 1963 to 2020. After culling only empirical response-shift articles, raters characterized how investigators framed and interpreted study research questions (n = 164 articles).

RESULTS

Of 10 methods used, papers using four of them utilized terms like "bias" and aimed to remove response-shift effects to reveal "true change." Yet, the investigators' reflections on their own conclusions suggested that they do not truly believe that response shift is error to be removed. A structured nomenclature is proposed for discussing response-shift results in a range of research contexts and response-shift detection methods.

CONCLUSIONS

It is time for a concerted and focused effort to change the nomenclature of those methods that demonstrated this misinterpretation. Only by framing and interpreting response shift as information, not bias, can we improve our understanding and methods to help to distill outcomes with and without response-shift effects.

Norm-based comparison of the quality-of-life impact of ravulizumab and eculizumab in paroxysmal nocturnal hemoglobinuria

Aims

Paroxysmal nocturnal hemoglobinuria (PNH) is a rare and life-threatening intravascular hematologic disorder with significant morbidity and premature mortality. Clinical trials (NCT02946463 and NCT03056040) comparing ravulizumab with eculizumab for PNH have supported the non-inferiority of the former and similar safety and tolerability. This secondary analysis compared PNH trial participants after 26 weeks on either treatment (n = 438) to a general-population sample (GenPop) (n = 15,386) and investigated response-shift effects.

Methods

Multivariate analysis of covariance (MANCOVA) investigated function and symptom scores on the European Organisation for Research and Treatment of Cancer (EORTC) QLQ-C30 of people with PNH as compared to GenPop, after covariate adjustment. Risk-factor groups were created based on clinical indicators known to be associated with worse PNH outcomes, and separate MANCOVAs were computed for lower- and higher-risk-factor groups. Differential item functioning (DIF) analyses examined whether item response varied systematically (1) by treatment, (2) compared to GenPop, and (3) over time, the latter two suggesting and reflecting response-shift effects, respectively. DIF analyses examined 24 items from scales with at least two items. Recalibration response shift was operationalized as uniform DIF over time, reflecting the idea that, for a given group, the difficulty of endorsing an item changes over time, after adjusting for the total subscale score. Reprioritization response shift was operationalized as non-uniform DIF over time, i.e., the relative difficulty of endorsing an item over time changes across the total domain score.

Results

Across PNH risk-factor levels, people who had been on either treatment for 26 weeks reported better-than-expected functioning and lower symptom burden compared to GenPop. Ravulizumab generally showed larger effect sizes. Results were similar for lower and higher PNH risk factors, with slightly stronger effects in the former. DIF analyses revealed no treatment DIF, but did uncover group DIF (9 items with uniform DIF, and 11 with non-uniform) and DIF over time (7 items with uniform DIF, and 3 with non-uniform).

Conclusions

This study revealed that people with PNH on ravulizumab or eculizumab for 26 weeks reported QOL levels better than those of the general population. Significant effects of DIF by group and DIF over time support recalibration and reprioritization response-shift effects. These findings suggest that the treatments enabled adaptive changes.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13023-021-02016-8.

Quality of life of transplanted children and their parents: a cross-sectional study

BACKGROUND

Transplantation is a saving therapeutic that has heavy consequences. The quality of life (QoL) of transplanted children and their parents has been little studied and should help physicians better manage these patients. The objectives of the study were to assess: (1) the QoL of transplanted children and parents and compare it with that of children with other chronic conditions associated with long-term consequences, and (2) potential variables modulating the QoL.

METHODS

This cross-sectional study was performed in a multidisciplinary paediatric unit (Timone Hospital, Marseille, France). Children were less than 18 years old; had a liver, kidney or heart transplant; and had a time since transplantation of 1-10 years. Socio-demographics and clinical data were recorded from medical forms. The QoL was assessed using the VSP-A (Vécu et Santé Perçue de l'Adolescent et de l'Enfant) and the WhoQoL self-reported questionnaires.

RESULTS

Forty-five families were included (response rate: 76%). The transplanted organs were the liver for 20 children, the kidney for 15 children, and the heart for 10 children. The QoL of transplanted children reported by their parents was better than that of children with inborn errors of metabolism and similar to that of childhood leukaemia survivors. The QoL of parents of transplanted children was better than that of parents of children with inborn errors of metabolism and did not differ from French norms. The QoL did not differ according to the nature of the transplanted organ, sex or the main sociodemographic data. The main modulators decreasing QoL were residual treatment level, medications switch and the presence of another regular treatment.

CONCLUSION

Transplanted children and their families reported a fairly preserved QoL compared to children with other chronic health conditions. Special attention should be given to QoL modulators related to therapeutic management (medication switches, regular treatments) that might be amenable to improve the QoL. Trial registration Ethics committee of Aix-Marseille University, France (reference number: 2014-08-04-03, 24/4/2015; https://www.univ-amu.fr/fr/public/comite-dethique ).

Creating idiometric short-form measures of cognitive appraisal: balancing theory and pragmatics

BACKGROUND

The Rapkin and Schwartz appraisal theory and measure provided a path toward documenting response-shift effects and describing individual differences in ways of thinking about quality of life (QOL) that distinguished people in different circumstances. Recent work developed and validated the QOL Appraisal Profile_{version 2} (QOLAP_v2), an 85-item measure that taps response-shift-detection domains of Frame of Reference, Standards of Comparison, Sampling of Experience, and Combinatory Algorithm. Recent theoretical work proposed that appraisal measurement constitutes a new class of measurement (idiometric), distinct from psychometric and clinimetric. To validate an idiometric measure, one would document that its items reflect different circumstances and population characteristics, and explain variance in QOL. The present work sought to develop idiometric short-forms of the QOLAP_v2 item bank by examining which items were most informative, retaining the appraisal-domain structure.

METHODS

This secondary analysis (n = 1481) included chronically-ill patients and their caregivers from a longitudinal web-based survey (mean follow-up 16.6 months). Data included the QOLAP_v2, the Center for Disease Control Healthy Days Core Module, the PROMIS-10 Global Health, and demographic/medical variables. Appraisal items were measured at baseline (relevant to understanding cognitive appraisal processes); and with change scores (sensitive to response-shift effects). Multivariate analysis of covariance examined what demographic and health-status change variables were reflected by each of 85 appraisal items (in five sets), as dependent variables, and other demographic/medical variables. Multiple linear regression examined how appraisal items explained variance in global physical- and mental-health change, after covariate adjustment. A tally summarized item performance across all five sets of cross-sectional and longitudinal analyses.

RESULTS

The vast majority (i.e., 80%) of the QOLAP_v2 items performed well across the analyses presented. Using a relatively strict criterion of explaining meaningful variance across 60% of analyses, one would retain 68 items. A more lenient criterion (40%) would retain 71.

CONCLUSIONS

The present study provides heuristics to support investigators' creating 'discretionary' QOLAP_v2 short-forms to fit their study aim and amplifying individual differences in the cognitive processes underlying QOL. This approach enables adapting the measure to the study population, as per the expectation that respondent populations differ in the predominant cognitive processes used.

Prediction of Multiple sclerosis disease using machine learning classifiers: a comparative study

Introduction

Hamedan Province is one of Iran’s high-risk regions for Multiple Sclerosis (MS). Early diagnosis of MS based on an accurate system can control the disease. The aim of this study was to compare the performance of four machine learning techniques with traditional methods for predicting MS patients.

Methods

The study used information regarding 200 patients through a case-control study conducted in Hamadan, Western Iran, from 2013 to 2015. The performance of six classifiers was used to compare their performance in terms of sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (LR+), negative likelihood ratio (LR-) and total accuracy.

Results

Random Forest (RF) model illustrated better performance among other models in both scenarios. It had greater specificity (0.67), PPV (0.68) and total accuracy (0.68). The most influential diagnostic factors for MS were age, birth season and gender.

Conclusions

Our findings showed that despite all the six methods performed almost similarly, the RF model performed slightly better in terms of different criteria in prediction accuracy. Accordingly, this approach is an effective classifier for predicting MS in the early stage and control the disease.

Exploiting Machine Learning Algorithms and Methods for the Prediction of Agitated Delirium After Cardiac Surgery: Models Development and Validation Study

Background

Delirium is a temporary mental disorder that occasionally affects patients undergoing surgery, especially cardiac surgery. It is strongly associated with major adverse events, which in turn leads to increased cost and poor outcomes (eg, need for nursing home due to cognitive impairment, stroke, and death). The ability to foresee patients at risk of delirium will guide the timely initiation of multimodal preventive interventions, which will aid in reducing the burden and negative consequences associated with delirium. Several studies have focused on the prediction of delirium. However, the number of studies in cardiac surgical patients that have used machine learning methods is very limited.

Objective

This study aimed to explore the application of several machine learning predictive models that can pre-emptively predict delirium in patients undergoing cardiac surgery and compare their performance.

Methods

We investigated a number of machine learning methods to develop models that can predict delirium after cardiac surgery. A clinical dataset comprising over 5000 actual patients who underwent cardiac surgery in a single center was used to develop the models using logistic regression, artificial neural networks (ANN), support vector machines (SVM), Bayesian belief networks (BBN), naïve Bayesian, random forest, and decision trees.

Results

Only 507 out of 5584 patients (11.4%) developed delirium. We addressed the underlying class imbalance, using random undersampling, in the training dataset. The final prediction performance was validated on a separate test dataset. Owing to the target class imbalance, several measures were used to evaluate algorithm’s performance for the delirium class on the test dataset. Out of the selected algorithms, the SVM algorithm had the best F1 score for positive cases, kappa, and positive predictive value (40.2%, 29.3%, and 29.7%, respectively) with a P=.01, .03, .02, respectively. The ANN had the best receiver-operator area-under the curve (78.2%; P=.03). The BBN had the best precision-recall area-under the curve for detecting positive cases (30.4%; P=.03).

Conclusions

Although delirium is inherently complex, preventive measures to mitigate its negative effect can be applied proactively if patients at risk are prospectively identified. Our results highlight 2 important points: (1) addressing class imbalance on the training dataset will augment machine learning model’s performance in identifying patients likely to develop postoperative delirium, and (2) as the prediction of postoperative delirium is difficult because it is multifactorial and has complex pathophysiology, applying machine learning methods (complex or simple) may improve the prediction by revealing hidden patterns, which will lead to cost reduction by prevention of complications and will optimize patients’ outcomes.

Machine learning models reveal neurocognitive impairment type and prevalence are associated with distinct variables in HIV/AIDS

Neurocognitive impairment (NCI) among HIV-infected patients is heterogeneous in its reported presentations and frequencies. To determine the prevalence of NCI and its associated subtypes as well as predictive variables, we investigated patients with HIV/AIDS receiving universal health care. Recruited adult HIV-infected subjects underwent a neuropsychological (NP) test battery with established normative (sex-, age-, and education-matched) values together with assessment of their demographic and clinical variables. Three patient groups were identified including neurocognitively normal (NN, n = 246), HIV-associated neurocognitive disorders (HAND, n = 78), and neurocognitively impaired-other disorders (NCI-OD, n = 46). Univariate, multiple logistic regression and machine learning analyses were applied. Univariate analyses showed variables differed significantly between groups including birth continent, quality of life, substance use, and PHQ-9. Multiple logistic regression models revealed groups again differed significantly for substance use, PHQ-9 score, VACS index, and head injury. Random forest (RF) models disclosed that classification algorithms distinguished HAND from NN and NCI-OD from NN with area under the curve (AUC) values of 0.87 and 0.77, respectively. Relative importance plots derived from the RF model exhibited distinct variable rankings that were predictive of NCI status for both NN versus HAND and NN versus NCI-OD comparisons. Thus, NCI was frequently detected (33.5%) although HAND prevalence (21%) was lower than in several earlier reports underscoring the potential contribution of other factors to NCI. Machine learning models uncovered variables related to individual NCI types that were not identified by univariate or multiple logistic regression analyses, highlighting the value of other approaches to understanding NCI in HIV/AIDS.

Patient-reported outcomes in spine surgery: past, current, and future directions

The purpose of this article is to review the current state of outcome measurement in spine surgery, with an emphasis on patient-reported outcome measures (PROMs). The commonly used generic and disease-specific outcome measures used in spinal surgery and research will be discussed. The authors will introduce the concepts of response shift and appraisal processes, which may affect the face validity of PROMs, as well as their interpretation over time. It is not uncommon for there to be a discrepancy between the observed and expected outcome, which is not wholly explainable by objective measures. Current work on understanding how appraisal affects outcome measurement will be discussed, and future directions will be suggested to facilitate the continued evolution of PROMs.There has been an evolution in the way clinicians measure outcomes following spinal surgery. In moving from purely physical, objective measures to a growing emphasis on the patient's perspective, spine surgery outcomes are better able to integrate the impact at multiple levels of relevant change. Appraisal concepts and methods are gaining traction as ways to understand the cognitive processes underlying PROMs over time. Measurement of appraisal is a valuable adjunct to the current spine outcome tools.

Advancing quality-of-life research by deepening our understanding of response shift: a unifying theory of appraisal

Studies of response-shift phenomena in quality-of-life (QOL) research have grown steadily in the more than two decades of research. As a field, we have been calling a lot of different approaches "response shift" over the years, but the only approach that fully embodies the foundational (Sprangers and Schwartz, Soc Sci Med 48(11):1507-1515, 1999) theoretical model is appraisal. According to the definition proposed in this model, response shift is about individual cognitive change. This paper presents the argument that all response-shift research models need to be grounded in an understanding of cognitive appraisal. We present a reasoned argument for why appraisal is fundamental to QOL response-shift research. We describe current measurement and analytic methods for working with appraisal, and how these methods can be integrated into the current response-shift statistical 'tool box.' We propose future research directions on theory, methods, and cross-calibration of group- and individual-level methods. There are currently three tools available in multiple languages for assessing QOL appraisal processes. They have been tested and used to assess response shift in empirical studies of ~ 7000 people with chronic illness. The study findings illustrate how appraisal theory and methods can facilitate methodological investigations of and to enhance other response-shift detection methods. Future research directions are proposed to enrich QOL theory, response-shift methods, and interpretation of QOL change over time. Appraisal theory and methods are the closest approximation to a response-shift 'gold standard.' They provide the foundation for understanding response shift and point to a unified theory of QOL.

Perceptions on the value of bodily functions in multiple sclerosis

BACKGROUND

In neurological diseases presenting with a plethora of symptoms, the value of bodily functions for a given patient might be a guide for clinical management. Multiple sclerosis (MS) is paradigmatic in this respect, and little is known about the value of different bodily functions of patients and their physicians' perceptions.

METHODS

In a multicenter study, 171 patients with relapsing-remitting multiple sclerosis (RRMS), 61% with a clinically active disease within the last 2 years were followed over up to 3 years and yearly patients and their study physician rated on the perceived value of 13 bodily functions via a priority list. Differences between patients and physicians as well as modulating disease demographic factors were analyzed.

RESULTS

Patients with RRMS rated visual function followed by thinking and memory and walking highest while physicians stressed mobility, followed by thinking and memory and alertness most. Ratings were independent from disease duration or disability. Strongest value judgment differences were seen in swallowing regarded more relevant by patients and hand function regarded more relevant by physicians. In general, patients' and physicians' ratings through time were quite stable. Collapsing physical items into a physical functioning scale and mental items in a mental function scale, both dimensions were regarded equally important by patients while physicians underscored physical functioning (P = .016).

CONCLUSION

There are differences between patients and physicians in value statements of bodily functions in MS. In particular, visual functioning is under-recognized by physicians.

Scoping review of response shift methods: current reporting practices and recommendations

BACKGROUND

Response shift (RS) has been defined as a change in the meaning of an individual's self-evaluation of his/her health status and quality of life. Several statistical model- and design-based methods have been developed to test for RS in longitudinal data. We reviewed the uptake of these methods in patient-reported outcomes (PRO) literature.

METHODS

CINHAHL, EMBASE, Medline, ProQuest, PsycINFO, and Web of Science were searched to identify English-language articles about RS published until 2016. Data on year and country of publication, PRO measure adopted, RS detection method, type of RS detected, and testing of underlying model assumptions were extracted from the included articles.

RESULTS

Of the 1032 articles identified, 101 (9.8%) articles were included in the study. While 54.5 of the articles reported on the Then-test, 30.7% of the articles reported on Oort's or Schmitt's structural equation modeling (SEM) procedure. Newer RS detection methods, such as relative importance analysis and random forest regression, have been used less frequently. Less than 25% reported on testing the assumptions underlying the adopted RS detection method(s).

CONCLUSIONS

Despite rapid methodological advancements in RS research, this review highlights the need for further research about RS detection methods for complex longitudinal data and standardized reporting guidelines.

Using classification and regression tree modelling to investigate response shift patterns in dentine hypersensitivity

Background

Dentine hypersensitivity (DH) affects people’s quality of life (QoL). However changes in the internal meaning of QoL, known as Response shift (RS) may undermine longitudinal assessment of QoL. This study aimed to describe patterns of RS in people with DH using Classification and Regression Trees (CRT) and to explore the convergent validity of CRT with the then-test and ideals approaches.

Methods

Data from an 8-week clinical trial of mouthwashes for dentine hypersensitivity (n = 75) using the Dentine Hypersensitivity Experience Questionnaire (DHEQ) as the outcome measure, were analysed. CRT was used to examine 8-week changes in DHEQ total score as a dependent variable with clinical status for DH and each DHEQ subscale score (restrictions, coping, social, emotional and identity) as independent variables. Recalibration was inferred when the clinical change was not consistent with the DHEQ change score using a minimally important difference for DHEQ of 22 points. Reprioritization was inferred by changes in the relative importance of each subscale to the model over time.

Results

Overall, 50.7% of participants experienced a clinical improvement in their DH after treatment and 22.7% experienced an important improvement in their quality of life. Thirty-six per cent shifted their internal standards downward and 14.7% upwards, suggesting recalibration. Reprioritization occurred over time among the social and emotional impacts of DH.

Conclusions

CRT was a useful method to reveal both, the types and nature of RS in people with a mild health condition and demonstrated convergent validity with design based approaches to detect RS.

Appraisal assessment in patient-reported outcome research: methods for uncovering the personal context and meaning of quality of life

BACKGROUND

Recent work on patient-reported outcomes (PROs) focuses on precise, brief measures, which generally convey little about what an individual's rating actually means. Individual differences in appraisal are important and relevant to PRO research. This paper highlights the advantages of integrating appraisal assessment into clinical research.

METHODS

The most comprehensive method for assessing appraisal, the quality of life (QOL) Appraisal Profile, includes open-ended and multiple choice questions to assess four appraisal parameters: frame of reference, sampling of experience, standards of comparison, and combinatory algorithm. We illustrate with empirical findings four classes of investigation that would benefit from appraisal assessment: methodological, interpretation of change, the backstory of resilience, and clinical applications.

RESULTS

A methodological investigation of HIV/AIDS patients revealed a range of cognitive schemas induced by the then-test response shift detection method, only 15% of which reflected the presumed process invoked. In this same study and in a study of people with multiple sclerosis (MS), interpretation of change in positive versus negative mental-health response shifts was characterized by different appraisal processes. In studying resilience in MS patients, patients with more reserve-building activities were more likely to use appraisals that emphasized the positive and more controllable aspects of their illness experience, as compared to lower-reserve patients. In underserved cancer patients, the QOL Appraisal Profile was used as a clinical interview to articulate current concerns and for personalized treatment decision-making to reduce burden and promote adherence.

CONCLUSIONS

Integrating appraisal assessment can provide a more textured, person-centered understanding of person-factors not captured by standard PROs.

Exploring the Response Shift Effect on the Quality of Life of Patients with Schizophrenia

Background. Interpretation of quality of life (QoL) scores over time can be difficult because of possible changes in internal standards, values, and conceptualization of QoL by individuals. This effect is called a response shift (RS). The purpose of this study was to examine whether an RS effect occurred over a 24-mo period in patients who were suffering from schizophrenia. Methods. The random forest method was applied to detect any RS reprioritization in a multicenter cohort study. QoL was recorded using a generic questionnaire (SF36) at baseline (T0), 12 mo (T12), and 24 mo (T24). Patients were categorized into 3 groups based on psychotic symptoms and relapse (stable, improved, and worsened groups) from their clinical profiles. The random forest method was performed to predict the General Health score of the SF36 from the other QoL domain scores of the SF36. We estimated the average variable importance of the QoL domain for each of the 3 groups. Results. A total of 124 (53.2%) patients were defined as stable, 59 (25.3%) as improved, and 50 (21.5%) as worsened. Among the stable group, the Social Functioning domain became more important over time. Of those classified as improved, the Mental Health domain became more important over time, while the Vitality domain became less important. Among those in the group who worsened, the Mental Health domain became less important while the Vitality and Bodily Pain domains became more important. Conclusions. Our study identified differential RS reprioritization among patients with different clinical profiles. Further work is needed to determine whether RS should be interpreted as a measurement bias or as an effect integrated in a true change.

Identifying reprioritization response shift in a stroke caregiver population: a comparison of missing data methods

PURPOSE

Response shift (RS) is an important phenomenon that influences the assessment of longitudinal changes in health-related quality of life (HRQOL) studies. Given that RS effects are often small, missing data due to attrition or item non-response can contribute to failure to detect RS effects. Since missing data are often encountered in longitudinal HRQOL data, effective strategies to deal with missing data are important to consider. This study aims to compare different imputation methods on the detection of reprioritization RS in the HRQOL of caregivers of stroke survivors.

METHODS

Data were from a Canadian multi-center longitudinal study of caregivers of stroke survivors over a one-year period. The Stroke Impact Scale physical function score at baseline, with a cutoff of 75, was used to measure patient stroke severity for the reprioritization RS analysis. Mean imputation, likelihood-based expectation-maximization imputation, and multiple imputation methods were compared in test procedures based on changes in relative importance weights to detect RS in SF-36 domains over a 6-month period. Monte Carlo simulation methods were used to compare the statistical powers of relative importance test procedures for detecting RS in incomplete longitudinal data under different missing data mechanisms and imputation methods.

RESULTS

Of the 409 caregivers, 15.9 and 31.3 % of them had missing data at baseline and 6 months, respectively. There were no statistically significant changes in relative importance weights on any of the domains when complete-case analysis was adopted. But statistical significant changes were detected on physical functioning and/or vitality domains when mean imputation or EM imputation was adopted. There were also statistically significant changes in relative importance weights for physical functioning, mental health, and vitality domains when multiple imputation method was adopted. Our simulations revealed that relative importance test procedures were least powerful under complete-case analysis method and most powerful when a mean imputation or multiple imputation method was adopted for missing data, regardless of the missing data mechanism and proportion of missing data.

CONCLUSIONS

Test procedures based on relative importance measures are sensitive to the type and amount of missing data and imputation method. Relative importance test procedures based on mean imputation and multiple imputation are recommended for detecting RS in incomplete data.

Method variation in the impact of missing data on response shift detection

PURPOSE

Missing data due to attrition or item non-response can result in biased estimates and loss of power in longitudinal quality-of-life (QOL) research. The impact of missing data on response shift (RS) detection is relatively unknown. This overview article synthesizes the findings of three methods tested in this special section regarding the impact of missing data patterns on RS detection in incomplete longitudinal data.

METHODS

The RS detection methods investigated include: (1) Relative importance analysis to detect reprioritization RS in stroke caregivers; (2) Oort's structural equation modeling (SEM) to detect recalibration, reprioritization, and reconceptualization RS in cancer patients; and (3) Rasch-based item-response theory-based (IRT) models as compared to SEM models to detect recalibration and reprioritization RS in hospitalized chronic disease patients. Each method dealt with missing data differently, either with imputation (1), attrition-based multi-group analysis (2), or probabilistic analysis that is robust to missingness due to the specific objectivity property (3).

RESULTS

Relative importance analyses were sensitive to the type and amount of missing data and imputation method, with multiple imputation showing the largest RS effects. The attrition-based multi-group SEM revealed differential effects of both the changes in health-related QOL and the occurrence of response shift by attrition stratum, and enabled a more complete interpretation of findings. The IRT RS algorithm found evidence of small recalibration and reprioritization effects in General Health, whereas SEM mostly evidenced small recalibration effects. These differences may be due to differences between the two methods in handling of missing data.

CONCLUSIONS

Missing data imputation techniques result in different conclusions about the presence of reprioritization RS using the relative importance method, while the attrition-based SEM approach highlighted different recalibration and reprioritization RS effects by attrition group. The IRT analyses detected more recalibration and reprioritization RS effects than SEM, presumably due to IRT's robustness to missing data. Future research should apply simulation techniques in order to make conclusive statements about the impacts of missing data according to the type and amount of RS.

Changes in quality of life after epilepsy surgery: the role of reprioritization response shift

OBJECTIVES

Health-related quality of life (HRQOL) is an important indicator of well-being in patients with epilepsy. When assessing changes in HRQOL over time, some of the changes observed may be due to the patients' change in the internal standards, value, or meaning that they attribute to the HRQOL domain being measured, rather than actual change, a phenomenon known as response shift. Response shifts are increasingly recognized as an important explanation for the seemingly paradoxical or counterintuitive HRQOL results often observed in chronic conditions. We investigated the presence of changing values (reprioritization response shift) in data from a surgical randomized controlled trial of patients with temporal lobe epilepsy (TLE).

METHODS

Eighty patients with TLE, who were randomized to surgical and medical treatment, provided data on the epilepsy-specific 31-item Quality of life in Epilepsy (QOLIE-31) questionnaire at baseline and 1 year after randomization. Reprioritization response shift among the seven QOLIE-31 domains was assessed using changes in the relative importance weights derived from logistic regression and discriminant analysis.

RESULTS

The relative importance analysis showed a statistically significant increase over time in the importance of social function, but a significant decrease in the relative importance of seizure worry. There were no significant changes in the relative importance of the remaining five domains in distinguishing between surgical and medical group over time.

SIGNIFICANCE

Patients that receive surgical treatment are more likely to experience a decrease in the valuation of seizure worry and an increase in the value of social function as compared to patients who are medically treated. Changes in expectation about seizure freedom and social function may confound the assessment of longitudinal change on these outcomes, and highlight the importance of assessing response shift and the limitations of assessing HRQOL at a single point in time or without a control group. A PowerPoint slide summarizing this article is available for download in the Supporting Information section here.

Measuring quality of life in patients with schizophrenia:an overview

Quality of life (QoL) is of great importance to patients with schizophrenia and their families. Although the use of QoL measures may contribute to better adherence to therapeutic interventions, more satisfaction with care, improved health outcomes and reduction of health costs, QoL assessment remains underutilized in clinical practice. In this review, the authors propose a reflection on the limitations and lack of impact of QoL measures in clinical care. Our argument is based on three challenges regarding conceptual aspects and metrics, use and limits and the usefulness of measuring QoL. For each challenge, the authors have suggested pragmatic proposals and new research directions to promote the use of QoL measures in the future. These avenues of research involve a shared responsibility between QoL researchers, the medical community and decision makers. Close collaboration between all parties is necessary to promote the use of QoL measures in schizophrenia.

Shifting responses in quality of life: people living with dialysis

PURPOSE

People assess their quality of life (QoL) using internal standards, values, and priorities. With health changes, QoL responses shift to reflect current realities. This qualitative study investigated the life experience and QoL assessments of people living with dialysis (PWDs).

METHODS

Thirty-one interviews with 20 PWDs over the age of 70 (mean time on dialysis 34 months) and 11 family members investigated experience with disease and dialysis, related life challenges, medical decision-making, and future planning. Interviews were recorded, transcribed, and analyzed for themes and hypotheses using qualitative methods and description.

RESULTS

When beginning dialysis, PWDs' reported that life on dialysis was worth living and recalibrated their QoL assessments incorporating this standard into their daily lives. Three themes emerged as the disease progressed and dialysis became more difficult: PWDs first reported thriving on dialysis and then surviving with the support of dialysis, and when QoL was reported as poor, PWDs were reconsidering whether dialysis was still worth the life it was providing. Each of these steps involved changes in health circumstances, and PWDs explained their QoL assessments at each step using differing values and priorities. These steps in reconceptualization and reprioritization demonstrated PWDs' Response Shift and eventually led to another (recalibrated) standard: Life with dialysis was no longer worth living.

CONCLUSIONS

Quality-of-life assessment is an on-going process for older dialysis patients. Clinicians should be alert for changes in subjective QoL statements as dialysis and underlying diseases progress. Response Shift explains these changes in assessment.