Search

BackgroundAppropriate scale validity and internal consistency reliability have recently been documented for the new thyroid-specific quality of life (QoL) patient-reported outcome (PRO) measure for benign thyroid disorders, the ThyPRO. However, before clinical use, clinical validity and test–retest reliability should be evaluated.AimTo investigate clinical (‘known-groups’) validity and test–retest reliability of the Danish version of the ThyPRO.MethodsFor each of the 13 ThyPRO scales, we defined groups expected to have high versus low scores (‘known-groups’). The clinical validity (known-groups validity) was evaluated by whether the ThyPRO scales could detect expected differences in a cross-sectional study of 907 thyroid patients. Test–retest reliability was evaluated by intra-class correlations of two responses to the ThyPRO 2 weeks apart in a subsample of 87 stable patients.ResultsOn all 13 ThyPRO scales, we found substantial and significant differences between the groups expected to have high versus low scores. Test–retest reliability was above 0.70 (range 0.77–0.89) for all scales, which is usually considered necessary for comparisons among patient groups, but below 0.90, which is the usual threshold for use in individual patients.ConclusionWe found support for the clinical validity of the new thyroid-specific QoL questionnaire, ThyPRO, and evidence of good test–retest reliability. The questionnaire is now ready for use in clinical studies of patients with thyroid diseases.

Udgivelsesdato: 2009-May OBJECTIVE: To establish a reliable and valid scale structure of a patient-reported outcome measuring thyroid-specific quality of life. METHODS: The 98-item ThyPRO questionnaire was administered to patients with benign thyroid diseases at two university hospitals. Multi-trait scaling was performed, evaluating lack of convergent validity (item-own scale polyserial correlation 0.70, >0.80 with polychoric correlations) for all 13 scales. CONCLUSION: A reliable scale structure displaying complete convergent and almost complete discriminant validity was established in general analyses and in distinct clinical subgroups of patients with benign thyroid diseases.

In the past, a German Computerized Adaptive Test, based on Item Response Theory (IRT), was developed for purposes of assessing the construct depression [Computer-adaptive test for depression (D-CAT)]. This study aims at testing the feasibility and validity of the real computer-adaptive application. The D-CAT, supplied by a bank of 64 items, was administered on personal digital assistants (PDAs) to 423 consecutive patients suffering from psychosomatic and other medical conditions (78 with depression). Items were adaptively administered until a predetermined reliability (r ≥ 0.90) was attained. For validation purposes, the Hospital Anxiety and Depression Scale (HADS), the Centre for Epidemiological Studies Depression (CES-D) scale, and the Beck Depression Inventory (BDI) were administered. Another sample of 114 patients was evaluated using standardized diagnostic interviews [Composite International Diagnostic Interview (CIDI)]. The D-CAT was quickly completed (mean 74 seconds), well accepted by the patients and reliable after an average administration of only six items. In 95% of the cases, 10 items or less were needed for a reliable score estimate. Correlations between the D-CAT and the HADS, CES-D, and BDI ranged between r = 0.68 and r = 0.77. The D-CAT distinguished between diagnostic groups as well as established questionnaires do. The D-CAT proved an efficient, well accepted and reliable tool. Discriminative power was comparable to other depression measures, whereby the CAT is shorter and more precise. Item usage raises questions of balancing the item selection for content in the future. Copyright © 2009 John Wiley & Sons, Ltd.

Background: The aim of this study was to evaluate the Computerized Adaptive Test to measure anxiety (A-CAT), a patient-reported outcome questionnaire that uses computerized adaptive testing to measure anxiety. Methods: The A-CAT builds on an item bank of 50 items that has been built using conventional item analyses and item response theory analyses. The A-CAT was administered on Personal Digital Assistants to n=357 patients diagnosed and treated at the department of Psychosomatic Medicine and Psychotherapy, Charite Berlin, Germany. For validation purposes, two subgroups of patients (n=110 and 125) answered the A-CAT along with established anxiety and depression questionnaires. Results: The A-CAT was fast to complete (on average in 2 min, 38 s) and a precise item response theory based CAT score (reliability>.9) could be estimated after 4–41 items. On average, the CAT displayed 6 items (SD=4.2). Convergent validity of the A-CAT was supported by correlations to existing tools (Hospital Anxiety and Depression Scale-A, Beck Anxiety Inventory, Berliner Stimmungs-Fragebogen A/D, and State Trait Anxiety Inventory: r=.56–.66); discriminant validity between diagnostic groups was higher for the A-CAT than for other anxiety measures. Conclusions: The German A-CAT is an efficient, reliable, and valid tool for assessing anxiety in patients suffering from anxiety disorders and other conditions with significant potential for initial assessment and long-term treatment monitoring. Future research directions are to explore content balancing of the item selection algorithm of the CAT, to norm the tool to a healthy sample, and to develop practical cutoff scores. Depression and Anxiety, 2008. © 2008 Wiley-Liss, Inc.

Udgivelsesdato: 2008-Sep OBJECTIVE: To improve a newly developed patient-reported outcome measure for thyroid patients using cognitive interviewing. METHODS: Thirty-one interviews using immediate retrospective and expansive probing were conducted among patients with non-toxic goiter (n = 4), nodular toxic goiter (n = 5) Graves' disease (n = 6), thyroid eye-disease (n = 6), and primary hypothyroidism (n = 10). The questionnaire was revised successively. Six iterative rounds of interviews were conducted. Identified problems were categorized according to Tourangeau's four-stage model. RESULTS: Problems were identified 126 times in 43 of the 99 tested items, four of the 15 introductions, and four of the five response categories. Fifty-four problems involved comprehension, one retrieval, 23 judgment, 28 response, and 20 were not applicable to the four-stage model. Among all problems identified, 18 concerned attribution, i.e. whether or not to report only issues considered of thyroid causality. Within each round of interviews, the number of problems declined from an initial average of six per interview to two, mainly due to a reduction in the number of problems associated with comprehension. The least amount of reduction was within the set of problems involving attribution. CONCLUSION: The cognitive interview methodology was effective in identifying and reducing problems within the questionnaire responding process. Patients tended to selectively report problems they considered to be caused by their thyroid disease even when specifically instructed not to.

Patient reported outcomes (PRO) are considered central outcome measures for both clinical trials and observational studies in rheumatology. More sophisticated statistical models, including item response theory (IRT) and computerized adaptive testing (CAT), will enable critical evaluation and reconstruction of currently utilized PRO instruments to improve measurement precision while reducing item burden on the individual patient.We developed a domain hierarchy encompassing the latent trait of physical function/disability from the more general to most specific. Items collected from 165 English-language instruments were evaluated by a structured process including trained raters, modified Delphi expert consensus, and then patient evaluation. Each item in the refined data bank will undergo extensive analysis using IRT to evaluate response functions and measurement precision. CAT will allow for real-time questionnaires of potentially smaller numbers of questions tailored directly to each individual's level of physical function.Physical function/disability domain comprises 4 subdomains: upper extremity, trunk, lower extremity, and complex activities. Expert and patient review led to consensus favoring use of present-tense "capability" questions using a 4- or 5-item Likert response construct over past-tense "performance"items. Floor and ceiling effects, attribution of disability, and standardization of response categories were also addressed.By applying statistical techniques of IRT through use of CAT, existing PRO instruments may be improved to reduce questionnaire burden on the individual patients while increasing measurement precision that may ultimately lead to reduced sample size requirements for costly clinical trials.

Patient reported outcomes (PROs) have become standard study endpoints. However, little attention has been given to using item improvement to advance PRO performance which could improve precision, clarity, patient relevance, and information content of "physical function/disability" items and thus the performance of resulting instruments.The present study included 1860 physical function/disability items from 165 instruments. Item formulations were assessed by frequency of use, modified Delphi consensus, respondent judgement of clarity and importance, and item response theory (IRT). Data from 1100 rheumatoid arthritis, osteoarthritis, and normal ageing subjects, using qualitative item review, focus groups, cognitive interviews, and patient survey were used to achieve a unique item pool that was clear, reliable, sensitive to change, readily translatable, devoid of floor and ceiling limitations, contained unidimensional subdomains, and had maximal information content.A "present tense" time frame was used most frequently, better understood, more readily translated, and more directly estimated the latent trait of disability. Items in the "past tense" had 80-90% false negatives (p0.001). The best items were brief, clear, and contained a single construct. Responses with four to five options were preferred by both experts and respondents. The term physical function may be preferable to the term disability because of fewer floor effects. IRT analyses of "disability" suggest four independent subdomains (mobility, dexterity, axial, and compound) with factor loadings of 0.81-0.99.Major improvement in performance of items and instruments is possible, and may have the effect of substantially reducing sample size requirements for clinical trials.

In June 2004, the National Cancer Institute and the Drug Information Association co-sponsored the conference, "Improving the Measurement of Health Outcomes through the Applications of Item Response Theory (IRT) Modeling: Exploration of Item Banks and Computer-Adaptive Assessment." A component of the conference was presentation of a psychometric and content analysis of a secondary dataset.A thorough psychometric and content analysis was conducted of two primary domains within a cancer health-related quality of life (HRQOL) dataset.HRQOL scales were evaluated using factor analysis for categorical data, IRT modeling, and differential item functioning analyses. In addition, computerized adaptive administration of HRQOL item banks was simulated, and various IRT models were applied and compared.The original data were collected as part of the NCI-funded Quality of Life Evaluation in Oncology (Q-Score) Project. A total of 1,714 patients with cancer or HIV/AIDS were recruited from 5 clinical sites.Items from 4 HRQOL instruments were evaluated: Cancer Rehabilitation Evaluation System-Short Form, European Organization for Research and Treatment of Cancer Quality of Life Questionnaire, Functional Assessment of Cancer Therapy and Medical Outcomes Study Short-Form Health Survey.Four lessons learned from the project are discussed: the importance of good developmental item banks, the ambiguity of model fit results, the limits of our knowledge regarding the practical implications of model misfit, and the importance in the measurement of HRQOL of construct definition. With respect to these lessons, areas for future research are suggested. The feasibility of developing item banks for broad definitions of health is discussed.

Item banks and Computerized Adaptive Testing (CAT) have the potential to greatly improve the assessment of health outcomes. This review describes the unique features of item banks and CAT and discusses how to develop item banks. In CAT, a computer selects the items from an item bank that are most relevant for and informative about the particular respondent; thus optimizing test relevance and precision. Item response theory (IRT) provides the foundation for selecting the items that are most informative for the particular respondent and for scoring responses on a common metric. The development of an item bank is a multi-stage process that requires a clear definition of the construct to be measured, good items, a careful psychometric analysis of the items, and a clear specification of the final CAT. The psychometric analysis needs to evaluate the assumptions of the IRT model such as unidimensionality and local independence; that the items function the same way in different subgroups of the population; and that there is an adequate fit between the data and the chosen item response models. Also, interpretation guidelines need to be established to help the clinical application of the assessment. Although medical research can draw upon expertise from educational testing in the development of item banks and CAT, the medical field also encounters unique opportunities and challenges.

To compare the measurement properties of the Modified Health Assessment Questionnaire [MHAQ], the SF-36((R)) Health Survey 10 item Physical Functioning scale [PF10], and scores from an item response theory (IRT) based scale combining the two measures.Rheumatoid arthritis (RA) patients (n = 339) enrolled in a multi-center, randomized, double-blind, placebo-controlled trial completed the MHAQ and the SF-36 pre- and post-treatment. Psychometric analyses used confirmatory factor analysis and IRT models. Analyses of variance were used to assess sensitivity to changes in disease severity (defined by the American College of Rheumatism (ACR)) using change scores in MHAQ, PF10, and IRT scales. Analyses of covariance were used to assess treatment responsiveness.For the entire score range, the 95% confidence interval around individual patient scores was smaller for the combined (total) IRT based scale than for other measures. The MHAQ and PF10 were about 70% and 50% as efficient as the total IRT score of physical functioning in discriminating among ACR groups, respectively. The MHAQ and PF10 were also less efficient than the total IRT score in discriminating among treatment groups.Combining scales from the two short forms yields a more powerful tool with greater sensitivity to treatment response.

The importance of patient-reported outcomes such as health-related quality of life (HRQL) in clinical research is increasingly acknowledged. In order to yield valid results, the measurement properties of HRQL questionnaires must be thoroughly investigated. One aspect of such a validation process is the demonstration of content validity, i.e. that the questionnaire covers all relevant aspects. We review studies reporting on consequences of thyroid disorders and present the frequency of identified aspects, both overall HRQL issues and classical thyroid symptoms, in order to evaluate which issues are relevant for patients with thyroid diseases. Furthermore, existing questionnaires for thyroid patients are reviewed. A systematic search was performed in the Medline, Cinahl and Psycinfo databases and the reference lists of the relevant articles were hand-searched. Seventy-five relevant studies were identified. According to these studies, patients with untreated thyroid disease suffer from a wide range of symptoms and have major impairment in most areas of HRQL. Furthermore, the studies indicate that impairments in HRQL are also frequent in the long term. Six HRQL questionnaires for thyroid patients were identified. Generally, data supporting the validity of these questionnaires were sparse. According to the available literature, the quality of life of thyroid patients is substantially impaired over a wide range of aspects of HRQL in the untreated phase and continues to be so in many patients also in the long term. Studies systematically exploring the relative importance of these various aspects to thyroid patients are lacking, as is a comprehensive, validated thyroid-specific HRQL questionnaire.

Aim: To present the theoretical framework, design, methods, and baseline findings of the first Danish study on determinants and consequences of burnout, and the impact of workplace interventions in human service work organizations. Method: A 5-year prospective intervention study comprising 2,391 employees from different organizations in the human service sector: social security offices, psychiatric prison, institutions for severely disabled, hospitals, and homecare services. Data were collected at baseline and at two follow-ups. The authors developed a new burnout tool (the Copenhagen Burnout Inventory) covering work-related, client-related, and personal burnout. The study includes potential determinants of burnout (e.g. the psychosocial work environment, social relations outside work, lifestyle factors, and personality aspects) and consequences of burnout (e.g. poor health, low job satisfaction, turnover, and absenteeism). Here, the focus is on the description of the study population at baseline, including associations of work burnout with psychosocial work environment scales and absence. Results: Response rate at baseline was 80.1%. Midwives and homecare workers had high levels on both work- and client-related burnout. Prison officers had the highest level on client-related burnout. Supervisors and office assistants had low levels on both scales. Work burnout showed the highest correlations with job satisfaction (r=-0.51), quantitative demands (r=0.48), role-conflicts (r=0.44), and emotional demands (r=0.42). Sickness absence was 13.9 vs 6.0 days among participants in the highest and lowest work burnout quartile, respectively. Conclusion: The findings indicate that study design and methods are adequate for the upcoming prospective analyses of aetiology and consequences of burnout and of the impact of workplace interventions.

This study aimed at developing a shortened version of the EORTC QLQ-C30, one of the most widely used health-related quality of life questionnaires in oncology, for palliative care research. The study included interviews with 41 patients and 66 health care professionals in palliative care to determine the appropriateness, relevance and importance of the various domains of the QLQ-C30. Item response theory methods were used to shorten scales. Patients and health care professionals rated pain, physical function, emotional function, fatigue, global health status/quality of life, nausea/vomiting, appetite, dyspnoea, constipation, and sleep as most important. Therefore, these scales/items were retained in the questionnaire. Four scales were shortened without reducing measurement precision. Important dimensions not covered by the questionnaire were identified. The resulting 15-item EORTC QLQ-C15-PAL is a 'core questionnaire' for palliative care. Depending on the research questions, it may be supplemented by additional items, modules or questionnaires.

OBJECTIVE: Self-report questionnaires are widely used to measure health-related quality of life (HRQOL). Ideally, such questionnaires should be adapted to the individual patient and at the same time scores should be directly comparable across patients. This may be achieved using computerized adaptive testing (CAT). Usually, CAT is carried out for a single domain at a time. However, many HRQOL domains are highly correlated. Multidimensional CAT may utilize these correlations to improve measurement efficiency. We investigated the possible advantages and difficulties of multidimensional CAT. STUDY DESIGN AND SETTING: We evaluated multidimensional CAT of three scales from the EORTC QLQ-C30: the physical functioning, emotional functioning, and fatigue scales. Analyses utilised a database with 2958 European cancer patients. RESULTS: It was possible to obtain scores for the three domains with five to seven items administered using multidimensional CAT that were very close to the scores obtained using all 12 items and with no or little loss of measurement precision. CONCLUSION: The findings suggest that multidimensional CAT may significantly improve measurement precision and efficiency and encourage further research into multidimensional CAT. Particularly, the estimation of the model underlying the multidimensional CAT and the conceptual aspects need further investigations

Depression is one of the most prevalent mental health problems and measuring depressive symptoms becomes increasingly important in science as well as medical practice. Computer Adaptive Tests (CAT) based on the Item Response Theory (IRT) promise to enhance measurement precision and reduce respondent's burden. Our aim was to develop a CAT application to measure depressive symptoms. Three thousand two hundred seventy psychosomatic patients answered an overall of 11 mental health questionnaires at the University Clinic in Berlin. Three independent reviewers rated 144 items out of these questionnaires as indicative of depressive symptoms. All items underwent six empirical steps to analyze unidimensionality, local independence and item discrimination. Finally 64 items could be used to calculate item parameters applying a Generalized Partial Credit Model (GPCM). CAT scores were estimated using an 'expected a posteriori' algorithm (EAP). Two simulation experiments showed that for theta values within the range of 2SD around the mean (98% of the cases), the latent trait can be estimated out of approximately six items with a predefined standard error of [Symbol: see text] 0.32 (reliability rho [Symbol: see text] 0.90). The CAT-scores correlated high with scores of all depression items (r = 0.95), with the Beck Depression Inventory (r = 0.79) and with a CES-D 8 item short form (r = 0.76). We conclude that the Depression-CAT measures depressive symptoms with high precision and low respondent burden.

Background and Objectives: Most health-related quality-of-life questionnaires include multi-item scales. Scale scores are usually estimated as simple sums of the item scores. However, scoring procedures utilizing more information from the items might improve measurement abilities, and thereby reduce the needed sample sizes. We investigated whether item response theory (IRT)-based scoring improved the measurement abilities of the EORTC QLQ-C30 physical functioning, emotional functioning, and fatigue scales. Methods: Using a database of 13,0 10 subjects we estimated the relative validities of IRT scoring compared to sum scoring of the scales. Results: The mean relative validities were 1.04 (physical), 1.03 (emotional), and 0.97 (fatigue). None of these were significantly larger than 1. Thus, no gain in measurement abilities using IRT scoring was found for these scales. Possible explanations include that the items in the scales are not constructed for IRT scoring and that the scales are relatively short. Conclusion: IRT scoring of the three longest EORTC QLQ-C30 scales did not improve measurement abilities compared to the traditional sum scoring of the scales. (c) 2005 Elsevier Inc. All rights reserved

Background and objective This study examined the psychometric properties and equivalence of the six-item Headache Impact Test (HIT-6) across 11 languages in 14 countries. Methods A multicenter, international cross-sectional study conducted in a primary care setting. Data obtained from 1,171 adults from 14 countries who consulted their primary care physician for headache completed the HIT-6 questionnaire and a headache survey were included in this analysis. Item-level statistics (e.g., range of response choices used by participants), item–scale statistics (e.g., item–total correlations), scale level statistics (e.g., internal consistency reliability), and tests of differential item functioning were conducted to examine the psychometric properties of all HIT-6 translations and their comparability across translations. Results Across languages, missing data were low, item–scale correlations were high, reliability was adequate, and item-level statistics were generally comparable. We found only minor differential item functioning, suggesting that the HIT-6 translations are equivalent to the U.S. English form. Conclusions Psychometric analyses indicate that most HIT-6 translations (Canadian English, French, Greek, Hungarian, UK English, Hebrew, Portuguese, German, Spanish, and Dutch) are comparable to U.S. English. Improvements may be needed in the Finnish and Slovakian translations and the appropriateness of using the HIT-6 in South Africa should be explored further.

Measurement of headache impact is important in clinical trials, case detection, and the clinical monitoring of patients. Computerized adaptive testing (CAT) of headache impact has potential advantages over traditional fixed-length tests in terms of precision, relevance, real-time quality control and flexibility.To develop an item pool that can be used for a computerized adaptive test of headache impact.We analyzed responses to four well-known tests of headache impact from a population-based sample of recent headache sufferers (n = 1016). We used confirmatory factor analysis for categorical data and analyses based on item response theory (IRT).In factor analyses, we found very high correlations between the factors hypothesized by the original test constructers, both within and between the original questionnaires. These results suggest that a single score of headache impact is sufficient. We established a pool of 47 items which fitted the generalized partial credit IRT model. By simulating a computerized adaptive health test we showed that an adaptive test of only five items had a very high concordance with the score based on all items and that different worst-case item selection scenarios did not lead to bias.We have established a headache impact item pool that can be used in CAT of headache impact.

Item response theory (IRT) is a powerful framework for analyzing multiitem scales and is central to the implementation of computerized adaptive testing.To explain the use of IRT to examine measurement properties and to apply IRT to a questionnaire for measuring migraine impact--the Migraine Specific Questionnaire (MSQ).Data from three clinical studies that employed the MSQ-version 1 were analyzed by confirmatory factor analysis for categorical data and by IRT modeling.Confirmatory factor analyses showed very high correlations between the factors hypothesized by the original test constructions. Further, high item loadings on one common factor suggest that migraine impact may be adequately assessed by only one score. IRT analyses of the MSQ were feasible and provided several suggestions as to how to improve the items and in particular the response choices. Out of 15 items, 13 showed adequate fit to the IRT model. In general, IRT scores were strongly associated with the scores proposed by the original test developers and with the total item sum score. Analysis of response consistency showed that more than 90% of the patients answered consistently according to a unidimensional IRT model. For the remaining patients, scores on the dimension of emotional function were less strongly related to the overall IRT scores that mainly reflected role limitations. Such response patterns can be detected easily using response consistency indices. Analysis of test precision across score levels revealed that the MSQ was most precise at one standard deviation worse than the mean impact level for migraine patients that are not in treatment. Thus, gains in test precision can be achieved by developing items aimed at less severe levels of migraine impact.IRT proved useful for analyzing the MSQ. The approach warrants further testing in a more comprehensive item pool for headache impact that would enable computerized adaptive testing.

Item response theory (IRT) scoring of health status questionnaires offers many advantages. However, to ensure 'backwards comparability' and to facilitate interpretations of results, we need the ability to express the IRT score in the metrics of the traditional scales.To develop procedures to calibrate IRT-based scores on the Headache Impact Test (HIT) into the metrics of the traditional headache scales. To assess the degree to which the calibrated HIT scores agree with the observed traditional scores and lead to the same conclusions in group comparisons.We used telephone interview data (n = 1016) and Internet data (n = 1103) from general population surveys of recent headache sufferers. Analyses were conducted in four steps: (1) develop IRT models for all items, (2) for each IRT score level, calculate the expected score on each of the traditional scales (calibration), (3) adjust this calibrated score for measurement error in the IRT score, (4) for each of the traditional scales, assess agreement between calibrated HIT scores and observed scores using intraclass correlation (ICC) and evaluate the agreement of mean scores and the relative validity (RV) in discriminating among groups differing in migraine diagnosis, headache severity, and change in impact over time.For the traditional categorical questionnaire items (the Migraine Specific Questionnaire (MSQ) and the Headache Disability Inventory (HDI)) the calibrated HIT agreed with the observed traditional scores: ICC's were between 0.80 and 0.94. In RV analyses the maximum mean difference between the observed and expected scores was 1.7 points on a 0-100 scale for comparisons at one point in time. Analyses of change over time and analyses calibrating scores from the fixed-form HIT-6 to the metric of other questionnaires were also satisfactory although less precise. Analysis of non-standard questionnaire items (e.g. On how many days in the past 3 months did you have a headache, from the HIMQ and the MIDAS) required special IRT models. Agreement was less good: ICC's were between 0.56 and 0.61 and the maximum mean differences were 2.9 (on a 0-270 scale) and 3.8 (on a 0-450 scale) in RV analyses at one point in time. The ability of the calibrated scale scores to discriminate between groups was at least as good as the ability of the observed sum scales and often remarkably better.The theoretical advantage of IRT models in scale calibration is supported by our results. This approach to achieving comparability of new and widely-used scales and accelerating the accumulation of interpretation guidelines based on previous work warrant testing for measures of other generic and disease-specific concepts.

To evaluate the feasibility of computerized adaptive testing (CAT) and the reliability and validity of CAT-based estimates of headache impact scores in comparison with 'static' surveys.Responses to the 54-item Headache Impact Test (HIT) were re-analyzed for recent headache sufferers (n = 1016) who completed telephone interviews during the National Survey of Headache Impact (NSHI). Item response theory (IRT) calibrations and the computerized dynamic health assessment (DYNHA) software were used to simulate CAT assessments by selecting the most informative items for each person and estimating impact scores according to pre-set precision standards (CAT-HIT). Results were compared with IRT estimates based on all items (total-HIT), computerized 6-item dynamic estimates (CAT-HIT-6), and a developmental version of a 'static' 6-item form (HIT-6-D). Analyses focused on: respondent burden (survey length and administration time), score distributions ('ceiling' and 'floor' effects), reliability and standard errors, and clinical validity (diagnosis, level of severity). A random sample (n = 245) was re-assessed to test responsiveness. A second study (n = 1103) compared actual CAT surveys and an improved 'static' HIT-6 among current headache sufferers sampled on the Internet. Respondents completed measures from the first study and the generic SF-8 Health Survey; some (n = 540) were re-tested on the Internet after 2 weeks.In the first study, simulated CAT-HIT and total-HIT scores were highly correlated (r = 0.92) without 'ceiling' or 'floor' effects and with a substantial reduction (90.8%) in respondent burden. Six of the 54 items accounted for the great majority of item administrations (3603/5028, 77.6%). CAT-HIT reliability estimates were very high (0.975-0.992) in the range where 95% of respondents scored, and relative validity (RV) coefficients were high for diagnosis (RV = 0.87) and severity (RV = 0.89); patient-level classifications were accurate 91.3% for a diagnosis of migraine. For all three criteria of change, CAT-HIT scores were more responsive than all other measures. In the second study, estimates of respondent burden, item usage, reliability and clinical validity were replicated. The test-retest reliability of CAT-HIT was 0.79 and alternate forms coefficients ranged from 0.85 to 0.91. All correlations with the generic SF-8 were negative.CAT-based administrations of headache impact items achieved very large reductions in respondent burden without compromising validity for purposes of patient screening or monitoring changes in headache impact over time. IRT models and CAT-based dynamic health assessments warrant testing among patients with other conditions.

Evaluate a patient-reported outcomes questionnaire that uses computerized adaptive testing (CAT) to measure the impact of osteoarthritis (OA) on functioning and well-being.OA patients completed 37 questions about the impact of OA on physical, social and role functioning, emotional well-being, and vitality. Questionnaire responses were calibrated and scored using item response theory, and two scores were estimated: a Total-OA score based on patients' responses to all 37 questions, and a simulated CAT-OA score where the computer selected and scored the five most informative questions for each patient. Agreement between Total-OA and CAT-OA scores was assessed using correlations. Discriminant validity of Total-OA and CAT-OA scores was assessed with analysis of variance. Criterion measures included OA pain and severity, patient global assessment, and missed work days.Simulated CAT-OA and Total-OA scores correlated highly (r = 0.96). Both Total-OA and simulated CAT-OA scores discriminated significantly between patients differing on the criterion measures. F-statistics across criterion measures ranged from 39.0 (P.001) to 225.1 (P.001) for the Total-OA score, and from 40.5 (P.001) to 221.5 (P.001) for the simulated CAT-OA score.CAT methods produce valid and precise estimates of the impact of OA on functioning and well-being with significant reduction in response burden.

In cross-national comparisons based on questionnaires, accurate translations are necessary to obtain valid results. Differential item functioning (DIF) analysis can be used to test whether translations of items in multi-item scales are equivalent to the original. In data from 10,815 respondents representing 10 European languages we tested for DIF in the nine translations of the EORTC QLQ-C30 emotional function scale when compared to the original English version. We tested for DIF using two different methods in parallel, a contingency table method and logistic regression. The DIF results obtained with the two methods were similar. We found indications of DIF in seven of the nine translations. At least two of the DIF findings seem to reflect linguistic problems in the translation. 'Imperfect' translations can affect conclusions drawn from cross-national comparisons. Given that translations can never be identical to the original we discuss how findings of DIF can be interpreted and discuss the difference between linguistic DIF and DIF caused by confounding, cross-cultural differences, or DIF in other items in the scale. We conclude that testing for DIF is a useful way to validate questionnaire translations.

Analyses of cognitive aspects of survey methodology (CASM) and psychometric analysis are two methods that are able to complement each other. We use concrete examples to illustrate how psychometric analyses can test hypotheses from CASM. The psychometrics framework recognizes that survey responses are affected by other factors than the concept being assessed, for example by cognitive factors and processes. Such factors are subsumed under the concept of measurement error. Possible sources of measurement error can be tested, e.g. by randomized experiments. A standard way to reduce measurement error is to ask several questions about the same concept and combine the answers into a multi-item scale that is more precise than the individual items. Techniques like structural equation models use the item correlations to assess the magnitude of measurement error and to test the assumptions behind the multi-item scale, e.g. the effect of common response choices and item time frames. A central problem in modern psychometrics is how to model the mapping of the continuous latent variable onto the item response choice categories. This is achieved by threshold models (e.g. item response models and structural equation models for categorical data). These models can, for example, analyze the impact of mode of administration, test whether the items function in the same way for all people (measurement invariance/differential item functioning) and examine the consistency of responses from any single person. Such analyses provide new possibilities for combining psychometrics and cognitive methods.

Objective—To measure the levels of fatigue in the general population, and to examine how disease and sociodemographic factors influence fatigue. Design—Cross sectional questionnaire study in the Danish general population. Subjects—A random, age stratified sample of 1608 people aged 20‐77 with an equal gender distribution (response rate 67%). Main outcome measures—Five fatigue scales from the questionnaire Multidimensional Fatigue Inventory: General Fatigue, Physical Fatigue, Reduced Activity, Reduced Motivation and Mental Fatigue. Results—Fatigue scores were skewed towards absence of fatigue. The General Fatigue and Physical Fatigue scales showed the highest fatigue levels while the Reduced Motivation scale showed lowest levels.Gender diVerences in fatigue scores were small, but the variability among women was higher—that is, more women had high scores. A multiple linear regression analysis showed that respondents of low social status and respondents with a depression had high fatigue scores on all scales, independent of other factors. Chronic somatic disease had an independent direct eVect on Mental Fatigue, but for the rest of the scales, the eVect of somatic disease depended on age, gender and/or whether the person was living alone. For example, General and Physical Fatigue decreased with age among healthy people, whereas scores on these scales increased with age among those with a somatic disease. Conclusions—Physical and mental diseases play essential parts for the level of fatigue and as modulators of the associations between sociodemographic factors and fatigue. These interactions should be taken into account in future research on fatigue and sociodemographic factors and when data from clinical studies are compared with normative data from the general population. (J Epidemiol Community Health 2000;54:827‐833)

We used general population data (n = 4084) to examine data completeness, response consistency, tests of scaling assumptions, and reliability of the Danish SF-36 Health Survey. We compared traditional multitrait scaling analyses to analyses using polychoric correlations and Spearman correlations. The frequency of missing values was low, except for elderly people and people with lower levels of education. Response consistency was high and compared well with results for the U.S. SF-36. For respondents with computable scales in all eight domains, scaling assumptions (item internal consistency, item discriminant validity, equal item-own scale correlations, and equal variances) were satisfactory in the total sample and in all subgroups. The SF-36 could discriminate between levels of health in all subgroups, but there were skewness, kurtosis, and ceiling effects in many subgroups (elderly people and people with chronic diseases excepted). Concerning correlation methods, we found interesting differences indicating advantages of using methods that do not assume a normal distribution of answers as an addition to traditional methods.

Background: The SF-36 physical functioning scale (PF-10) and the Health Assessment Questionnaire disability index (HAQ-DI) are the most frequently used instruments for measuring self-reported physical function in rheumatoid arthritis (RA). The objective of this study was to develop a crosswalk between scores on the PF-10 and HAQ-DI in RA. Methods: Item response theory (IRT) methods were used to co-calibrate both scales using data from 1791 RA patients. The appropriateness of a Rasch-based crosswalk was evaluated by comparing it with crosswalks based on a two-parameter and a multi-dimensional IRT model. The accuracy of the final crosswalk was cross-validated using baseline (n = 532) and 6-month follow-up (n = 276) data from an independent cohort of early RA patients. Results: The PF-10 and HAQ-DI adequately fit a unidimensional Rasch model. Both scales measured a wide range of functioning, although the HAQ-DI tended to better target lower levels of functioning. The Rasch-based crosswalk performed similarly to crosswalks based on the two-parameter and multidimensional IRT models. Agreement between predicted and observed scale scores in the cross-validation sample was acceptable for group-level comparisons. The longitudinal validity in discriminating between disease response states was similar between observed and predicted scores. Conclusion: The crosswalk developed in this study allows for converting scores from one scale to the other and can be used for group-level analyses in patients with RA.

Objectives This study aimed to develop and evaluate a first computerized adaptive test (CAT) for the measurement of stress perception (Stress-CAT), in terms of the two dimensions: exposure to stress and stress reaction . Study Design and Setting Item response theory modeling was performed using a two-parameter model (Generalized Partial Credit Model). The evaluation of the Stress-CAT comprised a simulation study and real clinical application. A total of 1,092 psychosomatic patients (N1) were studied. Two hundred simulees (N2) were generated for a simulated response data set. Then the Stress-CAT was given to n =116 inpatients, (N3) together with established stress questionnaires as validity criteria. Results The final banks included n =38 stress exposure items and n =31 stress reaction items. In the first simulation study, CAT scores could be estimated with a high measurement precision (SE 0.90) using 7.0±2.3 (M±SD) stress reaction items and 11.6±1.7 stress exposure items. The second simulation study reanalyzed real patients data (N1) and showed an average use of items of 5.6±2.1 for the dimension stress reaction and 10.0±4.9 for the dimension stress exposure . Convergent validity showed significantly high correlations. Conclusions The Stress-CAT is short and precise, potentially lowering the response burden of patients in clinical decision making.

PAR10 IMPROVING THE SENSITIVITY OF PHYSICAL FUNCTION MEASURES IN RHEUMATOID ARTHRITIS: USE OF ITEM RESPONSE THEORY IN PATIENTS TREATED WITH ABATACEPT (CTLA4IG) Martin M, Emery P, Kosinski M,Ware J, Li T,Williams R, Maclean R, Bjorner J Quality Metric, Inc, Lincoln, RI, USA; University of Leeds, Leeds, United Kingdom; Bristol-Myers Squibb, Princeton, NJ, USA; Bristol-Myers Squibb, Brussels, Belgium OBJECTIVES: The Health Assessment Questionnaire (HAQ) and the Modified HAQ (MHAQ) are examples of common short-forms of physical function used to measure improvement in the treatment of rheumatoid arthritis (RA) which have been associated with ceiling problems. These problems, inherent to short-form surveys, pose risks of failing to detect a treatment response in clinical trials. Item Response Theory (IRT) methods were used to examine the properties of two physical function measures and construct a combined measure to better detect changes in disease activity and treatment response. METHODS: Data were from a 12-month, double-blind, multi-center study of 339 RA patients on a background of methotrexate randomized to Abatacept at 2mg/kg, at 10mg/kg, or placebo. MHAQ and SF-36 (with its Physical Functioning scale, PF10) were administered at pretreatment and 3, 6, and 12 months post-treatment. IRT methods were used to examine the surveys’ measurement properties and compute new IRT-based physical function scores. Analyses of variance were used to assess sensitivity to changes in disease severity and treatment response. Relative validity coefficients were used to compare the measures. RESULTS: A Rasch IRT model fit the data. IRT-based scores successfully lowered the floor and raised the ceiling of the physical function measured. IRT-based scores were 30% more efficient than MHAQ and 50% more efficient than PF10 in discriminating among ACR groups. In discriminating among treatment groups, IRT-based scores were 25% more efficient than MHAQ and 12% more efficient than PF10 at 6-months; and 16% and 17% more efficient at 12-months based on observed effect sizes. CONCLUSIONS: Using IRT methodology to estimate a combined score for physical functioning lead to greater range of the construct measured. The improved measure, with greater measurement precision and sensitivity to treatment response, further confirmed the beneficial effect of Abatacept on physical function in the treatment of RA.

OBJECTIVES: The aim of this study was to test the unidimensionality of 65 items from nine widely used pain questionnaires in preparation for computerized adaptive pain assessment. METHODS: Random samples of adults were contacted on the internet (N = 782) and telephone (N = 750). 65 items were selected mainly from Brief Pain Inventory, MOS Pain Measures, Oswestry Low Back Pain Disability Questionnaire, NottingHam Health Profile, McGill Pain Questionnaire, Health Disability Index, Migraine-Specific Questionnaire, Aberdeen Back Pain Scale, and Headache Impact Test. Mplus software was used to perform confirmatory and exploratory factor analyses of polychoric correlations among the items. Model was examined through eigenvalue analysis, interfactor correlations, and root mean square residuals. RESULTS: Confirmatory factor loadings ranged from 0.53 to 0.92 with 90% of them higher than 0.7. Three factors were rotated to an oblique solution. Examination of exploratory factor analysis revealed two subdimensions of pain based upon item content: pain severity and pain impact. The third dimension observed in the three-factor solution, interpreted as current pain, was defined by questions without recall period. Eigenvalue analyses showed that the first factor explained more than 65 percent of the total variance in all pain item scores. Three factors explained 74% of the variance. The interfactor correlations were high, ranged from 0.64 to 0.67. Root mean square residuals were low, 0.07 for one-factor solution and 0.05, 0.03 for two- and three-factor solution respectively. CONCLUSION: Most items from widely used questionnaires measure one single underlying construct of pain sufficiently to meet the assumption of Item Response Theory. Items selected will be calibrated on a standard metric to estimate scores at patient's specific pain level. Further investigation on the application of IRT methods and computer technology in pain assessment is warranted.

We have the potential to substantially advance the field of health status assessment by constructing and calibrating questionnaires based on item response theory (IRT) and administering them using computerized adaptive methods. This opportunity could hardly come at a better time. The health care industry needs more practical tools to monitor population health on a large scale as well as more precise tools to identify those who need and are most likely to benefit from treatment. Furthermore, we need to standardize the concepts and metrics of health sufficiently to permit comparisons of results across assessment tools and across diverse applications. To achieve the full potential of IRT models, many methodological issues must be addressed, including those raised by Hays et al1 and McHorney and Cohen.2 In responding to these authors, we have chosen to emphasize the practical implications of IRT models, particularly their usefulness for purposes of computerized adaptive health assessments. We briefly summarize ongoing efforts to calibrate and dynamically administer widely used measures of headache impact, note similarities and differences in our analytic approaches, and comment on the implications of IRT methodology.

Objective Previous research has indicated that health behaviours tend to cluster in social networks, but few have studied the cluster effect in workgroups. We examined the effect of workgroups on current state and change in three indicators of health behaviours (smoking, body mass index (BMI) and physical activity). Further, we examined whether health behaviours of the respondents at group level predicted lifestyle changes. Methods In a prospective cohort (n = 4730), employees from 250 workgroups in the Danish eldercare sector answered questionnaires at baseline (2005) and follow-up (2006). Multilevel regression models were used to examine the effect of workgroups. Results Workgroups accounted for 6.49% of the variation in smoking status, 6.56% of amount smoked and 2.62% of the variation in current BMI. We found no significant workgroup clustering in physical activity or lifestyle changes. Furthermore, changes in smoking status (cessation) and weight gain were seen in workgroups with high percentage of smokers and high levels of BMI. Conclusion We found modest evidence for clustering of some health behaviours within workgroups, which could be due to social learning or selection into and out of workgroups. Future health promotion programmes at worksites should recognize the potential clustering of lifestyle behaviours within workgroups.

Background The use of global health items permits an efficient way of gathering general perceptions of health. These items provide useful summary information about health and are predictive of health care utilization and subsequent mortality. Methods Analyses of 10 self-reported global health items obtained from an internet survey as part of the Patient-Reported Outcome Measurement Information System (PROMIS) project. We derived summary scores from the global health items. We estimated the associations of the summary scores with the EQ-5D index score and the PROMIS physical function, pain, fatigue, emotional distress, and social health domain scores. Results Exploratory and confirmatory factor analyses supported a two-factor model. Global physical health (GPH; 4 items on overall physical health, physical function, pain, and fatigue) and global mental health (GMH; 4 items on quality of life, mental health, satisfaction with social activities, and emotional problems) scales were created. The scales had internal consistency reliability coefficients of 0.81 and 0.86, respectively. GPH correlated more strongly with the EQ-5D than did GMH (r = 0.76 vs. 0.59). GPH correlated most strongly with pain impact (r = −0.75) whereas GMH correlated most strongly with depressive symptoms (r = −0.71). Conclusions Two dimensions representing physical and mental health underlie the global health items in PROMIS. These global health scales can be used to efficiently summarize physical and mental health in patient-reported outcome studies.

Background Lifestyle variables may serve as important intermediate factors between psychosocial work environment and health outcomes. Previous studies, focussing on work stress models have shown mixed and weak results in relation to weight change. This study aims to investigate psychosocial factors outside the classical work stress models as potential predictors of change in body mass index (BMI) in a population of health care workers. Methods A cohort study, with three years follow-up, was conducted among Danish health care workers (3982 women and 152 men). Logistic regression analyses examined change in BMI (more than +/− 2 kg/m2) as predicted by baseline psychosocial work factors (work pace, workload, quality of leadership, influence at work, meaning of work, predictability, commitment, role clarity, and role conflicts) and five covariates (age, cohabitation, physical work demands, type of work position and seniority). Results Among women, high role conflicts predicted weight gain, while high role clarity predicted both weight gain and weight loss. Living alone also predicted weight gain among women, while older age decreased the odds of weight gain. High leadership quality predicted weight loss among men. Associations were generally weak, with the exception of quality of leadership, age, and cohabitation. Conclusion This study of a single occupational group suggested a few new risk factors for weight change outside the traditional work stress models.

Background Physical function (PF) is a core patient-reported outcome domain in clinical trials in rheumatic diseases. Frequently used PF measures have ceiling effects, leading to large sample size requirements and low sensitivity to change. In most of these instruments, the response category that indicates the highest PF level is the statement that one is able to perform a given physical activity without any limitations or difficulty. This study investigates whether using an item format with an extended response scale, allowing respondents to state that the performance of an activity is easy or very easy, increases the range of precise measurement of self-reported PF. Methods Three five-item PF short forms were constructed from the Patient-Reported Outcomes Measurement Information System (PROMIS®) wave 1 data. All forms included the same physical activities but varied in item stem and response scale: format A (“Are you able to …”; “without any difficulty”/“unable to do”); format B (“Does your health now limit you …”; “not at all”/“cannot do”); format C (“How difficult is it for you to …”; “very easy”/“impossible”). Each short-form item was answered by 2217–2835 subjects. We evaluated unidimensionality and estimated a graded response model for the 15 short-form items and remaining 119 items of the PROMIS PF bank to compare item and test information for the short forms along the PF continuum. We then used simulated data for five groups with different PF levels to illustrate differences in scoring precision between the short forms using different item formats. Results Sufficient unidimensionality of all short-form items and the original PF item bank was supported. Compared to formats A and B, format C increased the range of reliable measurement by about 0.5 standard deviations on the positive side of the PF continuum of the sample, provided more item information, and was more useful in distinguishing known groups with above-average functioning. Conclusions Using an item format with an extended response scale is an efficient option to increase the measurement range of self-reported physical function without changing the content of the measure or affecting the latent construct of the instrument. Electronic supplementary material The online version of this article (doi:10.1186/s13075-017-1273-5) contains supplementary material, which is available to authorized users.