A Review of Some of the History of Factorial Invariance and Differential Item Functioning.

The concept of factorial invariance has evolved since it originated in the 1930s as a criterion for the usefulness of the multiple factor model; it has become a form of analysis supporting the validity of inferences about group differences on underlying latent variables. The analysis of differential item functioning (DIF) arose in the literature of item response theory (IRT), where its original purpose was the detection and removal of test items that are differentially difficult for, or biased against, one subpopulation or another. The two traditions merge at the level of the underlying latent variable model, but their separate origins and different purposes have led them to differ in details of terminology and procedure. This review traces some aspects of the histories of the two traditions, ultimately drawing some conclusions about how analysts may draw on elements of both, and how the nature of the research question determines the procedures used. Whether statistical tests are grouped by parameter (as in studies of factorial invariance) or across parameters by variable (as in DIF analysis) depends on the context and is independent of the model, as are subtle aspects of the order of the tests. In any case in which DIF or partial invariance is a possibility, the invariant parameters, or anchor items in DIF analysis, are best selected in an interplay between the statistics and judgment about what is being measured.

Health-related quality of life profiles in adolescents and young adults with chronic conditions.

PURPOSE: To assess health-related quality of life (HRQOL) among adolescents and young adults (AYAs) with chronic conditions. METHODS: AYAs (N = 872) aged 14-20 years completed NIH's Patient-Reported Outcomes Measurement Information System® (PROMIS®) measures of physical function, pain interference, fatigue, social health, depression, anxiety, and anger. Latent profile analysis (LPA) was used to group AYAs into HRQOL profiles using PROMIS T-scores. The optimal number of profiles was determined by model fit statistics, likelihood ratio test, and entropy. Multinomial logistic regression models were used to examine how LPA's HRQOL profile membership was associated with patient demographic and chronic conditions. The model prediction accuracy on profile membership was evaluated using Huberty's I index with a threshold of 0.35 for good effect. RESULTS: A 4-profile LPA model was selected. A total of 161 (18.5%), 256 (29.4%), 364 (41.7%), and 91 (10.4%) AYAs were classified into Minimal, Mild, Moderate, and Severe HRQOL Impact profiles. AYAs in each profile had distinctive mean scores with over a half standard deviation (5-points in PROMIS T-scores) of difference between profiles across most HRQOL domains. AYAs who were female or had conditions such as mental health condition, hypertension, and self-reported chronic pain were more likely to be in the Severe HRQOL Impact profile. The Huberty's I index was 0.36. CONCLUSIONS: Approximately half of AYAs with a chronic condition experience moderate to severe HRQOL impact. The availability of risk prediction models for HRQOL impact will help to identify AYAs who are in greatest need of closer clinical care follow-up.

Applying latent profile analysis to identify adolescents and young adults with chronic conditions at risk for poor health-related quality of life.

The impact of chronic diseases on health-related quality of life (HRQOL) in adolescents and young adults (AYAs) is understudied. Latent profile analysis (LPA) can identify profiles of AYAs based on their HRQOL scores reflecting physical, mental, and social well-being. This paper will (1) demonstrate how to use LPA to identify profiles of AYAs based on their scores on multiple HRQOL indicators; (2) explore associations of demographic and clinical factors with LPA-identified HRQOL profiles of AYAs; and (3) provide guidance on the selection of adult or pediatric versions of Patient-Reported Outcomes Measurement Information System® (PROMIS®) in AYAs. A total of 872 AYAs with chronic conditions completed the adult and pediatric versions of PROMIS measures of anger, anxiety, depression, fatigue, pain interference, social health, and physical function. The optimal number of LPA profiles was determined by model fit statistics and clinical interpretability. Multinomial regression models examined clinical and demographic factors associated with profile membership. As a result of the LPA, AYAs were categorized into 3 profiles: Minimal, Moderate, and Severe HRQOL Impact profiles. Comparing LPA results using either the pediatric or adult PROMIS T-scores found approximately 71% of patients were placed in the same HRQOL profiles. AYAs who were female, had hypertension, mental health conditions, chronic pain, and those on medication were more likely to be placed in the Severe HRQOL Impact Profile. Our findings may facilitate clinicians to screen AYAs who may have low HRQOL due to diseases or treatments with the identified risk factors without implementing the HRQOL assessment.

Validation of the Multidimensional Assessment of Parenting: An application of item response theory.

Parenting is a critical mechanism contributing to child and adolescent development and outcomes. The Multidimensional Assessment of Parenting Scale (MAPS) is a new measure that aims to address gaps in the literature on existing self-report parenting measures. Research to date on the MAPS includes essential steps of scale development and validation; however, replicating scale dimensionality and examining differential item functioning (DIF) based on child age and a parent or child gender is a critical next step. The current study included 1,790 mothers and fathers of sons and daughters, spanning childhood to adolescence in the United States. Item response theory (IRT) confirmed initial factor-analytic work revealing positive and negative dimensions; however, the best-fitting multidimensional model included six nested dimensions from the original seven. A few notable items displayed DIF based on child age and parent gender; however, DIF based on child gender had minimal impact on the overall score. Future directions, clinical implications, and recommendations are discussed. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Determining a transitional scoring link between PROMIS® pediatric and adult physical health measures.

PURPOSE: Having independent versions of the PROMIS® scales (for Pediatric and Adults) is problematic as scores cannot be evaluated longitudinally as individuals move from childhood into adulthood. The primary aim of this research project is to use item response theory (IRT) to develop a transitional scoring link (or "crosswalk") between the PROMIS adult and pediatric physical health measures. SETTING: Sample 1 was collected at 6 rehabilitation hospitals in the U.S., and participants in Sample 2 were recruited from public health insurance programs and an online research panel. METHODS: PROMIS pediatric and adult physical function, mobility, upper extremity, fatigue, and pain measures were administered to a sample of 874 individuals aged 14-20 years old with special health needs and a sample of 641 individuals aged 14-25 years with a disability. IRT-based scores were linked using a linear approximation to calibrated projection. RESULTS: Estimated latent variable correlations ranged between 0.84 and 0.95 for the PROMIS pediatric and adult scores. Root Expected Mean Square Difference values were below the 0.08 threshold in all cases except when comparing genders on the Mobility (0.097) and Pain (0.10) scales in the special health care needs sample. Sum score conversion tables for the pediatric and adult PROMIS measures are presented. CONCLUSIONS: The linking coefficients can be used to calculate scale scores on PROMIS adult measures from pediatric measure scores and vice versa. This may lead to more accurate measurement in cross-sectional studies spanning multiple age groups or longitudinal studies that require comparable measurement across distinct developmental stages.

Assessing responsiveness over time of the PROMIS® pediatric symptom and function measures in cancer, nephrotic syndrome, and sickle cell disease.

PURPOSE: Previous studies provided evidence for the validity of the PROMIS Pediatric measures in cross-sectional studies. This study evaluated the ability of the PROMIS Pediatric measures to detect change over time in children and adolescents with cancer, nephrotic syndrome (NS), or sickle cell disease (SCD). METHODS: Participants (8-17 years) completed measures of fatigue, pain interference, anger, anxiety, depressive symptoms, mobility, upper extremity, and peer relationships at three or four time points (T1-T4). Between T1 and T2, children with cancer received chemotherapy and children with SCD experienced a pain exacerbation. Children with NS were first assessed during active disease (T2), with T3 and T4 conducted at disease remission. For the primary analysis of responsiveness, we expected better scores at T3 (recovery) compared to T2 (event) for all diseases. T1 and T4 are also expected to have better scores than T2. Linear mixed models were used and adjusted for time, gender, age, race/ethnicity, education, comorbid conditions, and disease. RESULTS: Enrolled were 96 children with cancer, 121 children with SCD, and 127 children with NS. Fatigue, pain interference, mobility, and upper extremity scores worsened from T1 (baseline) to T2 (event) (p < 0.01), and significantly improved from T2 to T3 and T4 (p < 0.01). Similarly, anxiety and depressive symptoms significantly improved from T2 to T3 and T4 (p < 0.01). CONCLUSIONS: This study provides evidence for the responsiveness of seven PROMIS Pediatric measures to clinical disease state in three chronic illnesses. The findings support use of PROMIS Pediatric measures in clinical research.

United States Population-Based Estimates of Patient-Reported Outcomes Measurement Information System Symptom and Functional Status Reference Values for Individuals With Cancer.

Purpose To estimate cancer population-based reference values in the United States for eight PROMIS (Patient-Reported Outcomes Measurement Information System) domains by age and stage of disease. Patients and Methods For the Measuring Your Health (MY-Health) study, persons newly diagnosed with cancer (prostate, colorectal, non-small-cell lung, non-Hodgkin lymphoma, breast, uterine, or cervical) from 2010 to 2012 (N = 5,284) were recruited through the National Cancer Institute's SEER Program. Participants were mailed surveys 6 to 13 months after diagnosis. Raking by race/ethnicity, age, and stage generated weighted average PROMIS scores for pain interference, fatigue, anxiety, depression, sleep disturbance, physical function, ability to participate in social roles, and cognitive function. PROMIS measures are standardized to a T-score metric, with a score of 50 representing the general US population mean. Clinically meaningful differences were defined as a 3-point difference in scores. Results Several reference values (means) for patients with cancer were worse than the general United States population norms of 50. These include pain interference (52.4), fatigue (52.2), and physical function (44.1). Reference values were highest (ie, showed greatest symptom burden) in lung cancer (pain interference, 55.5; fatigue, 57.3; depression, 51.4) and cervical cancer (anxiety, 53.2; sleep disturbance, 53.4). Reference values for patients age 65 to 84 years reported lower sleep disturbance, anxiety, and depression, and better cognitive function than younger patients. Cancer reference values were poorer among those with advanced disease compared with patients with limited or no evidence of disease, specifically physical function (41.1 v 46.6, respectively), fatigue (55.8 v 50.2, respectively), and pain interference (55.2 v 50.9, respectively). Conclusion In a large, population-based sample of patients with recently diagnosed cancer, we observed symptom severity and functional deficits by age, stage, and cancer type consistent with the expected impact of cancer diagnosis and treatment. These United States cancer reference values can help facilitate interpretation of the PROMIS domain scores in research studies or in clinical applications that measure and evaluate the symptom and functional burden patients with cancer experience after initial treatment.

Psychometric evaluation and design of patient-centered communication measures for cancer care settings.

OBJECTIVE: To evaluate the psychometric properties of questions that assess patient perceptions of patient-provider communication and design measures of patient-centered communication (PCC). METHODS: Participants (adults with colon or rectal cancer living in North Carolina) completed a survey at 2 to 3 months post-diagnosis. The survey included 87 questions in six PCC Functions: Exchanging Information, Fostering Health Relationships, Making Decisions, Responding to Emotions, Enabling Patient Self-Management, and Managing Uncertainty. For each Function we conducted factor analyses, item response theory modeling, and tests for differential item functioning, and assessed reliability and construct validity. RESULTS: Participants included 501 respondents; 46% had a high school education or less. Reliability within each Function ranged from 0.90 to 0.96. The PCC-Ca-36 (36-question survey; reliability=0.94) and PCC-Ca-6 (6-question survey; reliability=0.92) measures differentiated between individuals with poor and good health (i.e., known-groups validity) and were highly correlated with the HINTS communication scale (i.e., convergent validity). CONCLUSION: This study provides theory-grounded PCC measures found to be reliable and valid in colorectal cancer patients in North Carolina. Future work should evaluate measure validity over time and in other cancer populations. PRACTICE IMPLICATIONS: The PCC-Ca-36 and PCC-Ca-6 measures may be used for surveillance, intervention research, and quality improvement initiatives.

Reliability and validity of PROMIS measures administered by telephone interview in a longitudinal localized prostate cancer study.

PURPOSE: To evaluate the reliability and validity of six PROMIS measures (anxiety, depression, fatigue, pain interference, physical function, and sleep disturbance) telephone-administered to a diverse, population-based cohort of localized prostate cancer patients. METHODS: Newly diagnosed men were enrolled in the North Carolina Prostate Cancer Comparative Effectiveness and Survivorship Study. PROMIS measures were telephone-administered pre-treatment (baseline), and at 3-months and 12-months post-treatment initiation (N = 778). Reliability was evaluated using Cronbach's alpha. Dimensionality was examined with bifactor models and explained common variance (ECV). Ordinal logistic regression models were used to detect potential differential item functioning (DIF) for key demographic groups. Convergent and discriminant validity were assessed by correlations with the legacy instruments Memorial Anxiety Scale for Prostate Cancer and SF-12v2. Known-groups validity was examined by age, race/ethnicity, comorbidity, and treatment. RESULTS: Each PROMIS measure had high Cronbach's alpha values (0.86-0.96) and was sufficiently unidimensional. Floor effects were observed for anxiety, depression, and pain interference measures; ceiling effects were observed for physical function. No DIF was detected. Convergent validity was established with moderate to strong correlations between PROMIS and legacy measures (0.41-0.77) of similar constructs. Discriminant validity was demonstrated with weak correlations between measures of dissimilar domains (-0.20--0.31). PROMIS measures detected differences across age, race/ethnicity, and comorbidity groups; no differences were found by treatment. CONCLUSIONS: This study provides support for the reliability and construct validity of six PROMIS measures in prostate cancer, as well as the utility of telephone administration for assessing HRQoL in low literacy and hard-to-reach populations.

Development and Standardization of the Diagnostic Adaptive Behavior Scale: Application of Item Response Theory to the Assessment of Adaptive Behavior.

The Diagnostic Adaptive Behavior Scale (DABS) was developed using item response theory (IRT) methods and was constructed to provide the most precise and valid adaptive behavior information at or near the cutoff point of making a decision regarding a diagnosis of intellectual disability. The DABS initial item pool consisted of 260 items. Using IRT modeling and a nationally representative standardization sample, the item set was reduced to 75 items that provide the most precise adaptive behavior information at the cutoff area determining the presence or not of significant adaptive behavior deficits across conceptual, social, and practical skills. The standardization of the DABS is described and discussed.

Mode effects between computer self-administration and telephone interviewer-administration of the PROMIS(®) pediatric measures, self- and proxy report.

OBJECTIVE: To test equivalence of scores obtained with the PROMIS(®) pediatric Depressive Symptoms, Fatigue, and Mobility measures across two modes of administration: computer self-administration and telephone interviewer-administration. If mode effects are found, to estimate the magnitude and direction of the mode effects. METHODS: Respondents from an internet survey panel completed the child self-report and parent proxy-report versions of the PROMIS(®) pediatric Depressive Symptoms, Fatigue, and Mobility measures using both computer self-administration and telephone interviewer-administration in a crossed counterbalanced design. Pearson correlations and multivariate analysis of variance were used to examine the effects of mode of administration as well as order and form effects. RESULTS: Correlations between scores obtained with the two modes of administration were high. Scores were generally comparable across modes of administration, but there were some small significant effects involving mode of administration; significant differences in scores between the two modes ranged from 1.24 to 4.36 points. CONCLUSIONS: Scores for these pediatric PROMIS measures are generally comparable across modes of administration. Studies planning to use multiple modes (e.g., self-administration and interviewer-administration) should exercise good study design principles to minimize possible confounding effects from mixed modes.

Estimating minimally important difference (MID) in PROMIS pediatric measures using the scale-judgment method.

OBJECTIVE: To assess minimally important differences (MIDs) for several pediatric self-report item banks from the National Institutes of Health Patient-Reported Outcomes Measurement Information System(®) (PROMIS(®)). METHODS: We presented vignettes comprising sets of two completed PROMIS questionnaires and asked judges to declare whether the individual completing those questionnaires had an important change or not. We enrolled judges (including adolescents, parents, and clinicians) who responded to 24 vignettes (six for each domain of depression, pain interference, fatigue, and mobility). We used item response theory to model responses to the vignettes across different judges and estimated MID as the point at which 50 % of the judges would declare an important change. RESULTS: We enrolled 246 judges (78 adolescents, 85 parents, and 83 clinicians). The MID estimated with clinician data was about 2 points on the PROMIS T-score scale, and the MID estimated with adolescent and parent data was about 3 points on that same scale. CONCLUSIONS: The MIDs enhance the value of PROMIS pediatric measures in clinical research studies to identify meaningful changes in health status over time.

Modeling and Testing Differential Item Functioning in Unidimensional Binary Item Response Models with a Single Continuous Covariate: A Functional Data Analysis Approach.

Differential item functioning (DIF), referring to between-group variation in item characteristics above and beyond the group-level disparity in the latent variable of interest, has long been regarded as an important item-level diagnostic. The presence of DIF impairs the fit of the single-group item response model being used, and calls for either model modification or item deletion in practice, depending on the mode of analysis. Methods for testing DIF with continuous covariates, rather than categorical grouping variables, have been developed; however, they are restrictive in parametric forms, and thus are not sufficiently flexible to describe complex interaction among latent variables and covariates. In the current study, we formulate the probability of endorsing each test item as a general bivariate function of a unidimensional latent trait and a single covariate, which is then approximated by a two-dimensional smoothing spline. The accuracy and precision of the proposed procedure is evaluated via Monte Carlo simulations. If anchor items are available, we proposed an extended model that simultaneously estimates item characteristic functions (ICFs) for anchor items, ICFs conditional on the covariate for non-anchor items, and the latent variable density conditional on the covariate-all using regression splines. A permutation DIF test is developed, and its performance is compared to the conventional parametric approach in a simulation study. We also illustrate the proposed semiparametric DIF testing procedure with an empirical example.

Linkage between the PROMIS® pediatric and adult emotional distress measures.

PURPOSE: Research studies that measure health-related quality of life (HRQOL) in both children and adults and longitudinal studies that follow children into adulthood need measures that can be compared across these age groups. This study links the PROMIS pediatric and adult emotional distress measures using data from participants with diverse health conditions and disabilities. METHODS: Analyses were conducted and compared in two separate samples to confirm the stability of results. One sample (n = 874) included individuals aged 14-20 years with special health care needs and who require health services. The other sample (n = 641) included individuals aged 14-25 years who have a physical or cognitive disability. Participants completed both PROMIS pediatric and adult measures. Item response theory-based scores were linked using the linear approximation to calibrated projection. RESULTS: The estimated latent-variable correlation between pediatric and adult PROMIS measures ranged from 0.87 to 0.94. Regression coefficients ß 0 (intercept) and ß 1 (slope), and mean squared error are provided to transform scores from the pediatric to the adult measures, and vice versa. CONCLUSIONS: This study used a relatively new linking method, calibrated projection, to link PROMIS pediatric and adult measure scores, thus expanding the use of PROMIS measures to research that includes both populations.

PROMIS(®) pediatric self-report scales distinguish subgroups of children within and across six common pediatric chronic health conditions.

PURPOSE: To conduct a comparative analysis of eight pediatric self-report scales for ages 8-17 years from the National Institutes of Health (NIH) Patient Reported Outcomes Measurement Information System (PROMIS(®)) in six pediatric chronic health conditions, using indicators of disease severity. METHODS: Pediatric patients (N = 1454) with asthma, cancer, chronic kidney disease, obesity, rheumatic disease, and sickle cell disease completed items from the PROMIS pediatric mobility, upper extremity functioning, depressive symptoms, anxiety, anger, peer relationships, pain interference, and fatigue self-report scales. Comparisons within the six pediatric chronic health conditions were conducted by examining differences in groups based on the disease severity using markers of severity that were specific to characteristics of each disease. A comparison was also made across diseases between children who had been recently hospitalized and those who had not. RESULTS: In general, there were differences in self-reported health outcomes within each chronic health condition, with patients who had higher disease severity showing worse outcomes. Across health conditions, when children with recent hospitalizations were compared with those who had not been hospitalized in the past 6 months, we found significant differences in the expected directions for all PROMIS domains, except anger. CONCLUSIONS: PROMIS measures discriminate between different clinically meaningful subgroups within several chronic illnesses. Further research is needed to determine the responsiveness of the PROMIS pediatric scales to change over time.

Item-level informant discrepancies between children and their parents on the PROMIS(®) pediatric scales.

OBJECTIVE: The study objective was to describe the individual item-level discrepancies between children ages 8-17 years and their parents for the PROMIS(®) pediatric scales. Contextual effects on item-level informant discrepancies for the pediatric pain interference items were further analyzed conditional on whether the child, the parent, or anyone else in the household experienced chronic pain. METHODS: Parallel pediatric self-report and parent proxy-report items were completed by approximately 300 parent-child dyads depending on form assignment and individual nonresponse. Agreement between parent and child responses to individual items was measured using the polychoric correlation coefficient and weighted κ. The Chi-square test of symmetry was utilized for a comparison of the pattern of parent-child item discrepancies on the response scales, and the differences between the child and parent responses on the 1-5 item response scale are summarized . RESULTS: A continuum of higher item-level parent-child discrepancies was demonstrated starting with peer relationships, anger, anxiety, and depressive symptoms, followed by progressively lower parent-child discrepancies for energy, fatigue, asthma impact, pain interference, upper extremity, and mobility items. Parent-child discrepancies for pain interference items were lower in the context of chronic pain either in the child or in the parent. CONCLUSIONS: Parent-child item-level discrepancies were lower for more objective or visible items than for items measuring internal states or less observable items measuring latent variables such as peer relationships and fatigue. Future research should focus on the child and parent characteristics that influence domain-specific item-level discrepancies, and under what conditions item-level parent-child discrepancies predict child health outcomes.

Comparing score tests and other local dependence diagnostics for the graded response model.

Score tests for identifying locally dependent item pairs have been proposed for binary item response models. In this article, both the bifactor and the threshold shift score tests are generalized to the graded response model. For the bifactor test, the generalization is straightforward; it adds one secondary dimension associated only with one pair of items. For the threshold shift test, however, multiple generalizations are possible: in particular, conditional, uniform, and linear shift tests are discussed in this article. Simulation studies show that all of the score tests have accurate Type I error rates given large enough samples, although their small-sample behaviour is not as good as that of Pearson's Χ2 and M2 as proposed in other studies for the purpose of local dependence (LD) detection. All score tests have the highest power to detect the LD which is consistent with their parametric form, and in this case they are uniformly more powerful than Χ2 and M2 ; even wrongly specified score tests are more powerful than Χ2 and M2 in most conditions. An example using empirical data is provided for illustration.

Using item response theory to enrich and expand the PROMIS® pediatric self report banks.

BACKGROUND: The primary objective was to enhance the content coverage of some of the pediatric self-report item banks for ages 8-17 years from the National Institutes of Health (NIH) Patient Reported Outcomes Measurement Information System (PROMIS®), and extend the range of precise measurement to higher levels of physical functioning. METHODS: Data from 1,419 pediatric patients with cancer, chronic kidney disease, obesity, rehabilitation needs, rheumatic disease, and sickle cell disease were combined with item responses from the original standardization sample of 3,048 children to calibrate new items for the pediatric PROMIS Anger, Anxiety, Depressive Symptoms, Pain Interference, Fatigue, and physical functioning Upper Extremity and Mobility scales. Simultaneous or concurrent calibration using the graded item response theory model placed all of the items on the same scale. RESULTS: Twenty-two of 28 potential new items were added across the seven scales. A recommended short form was proposed for the Anger scale, and the recommended short forms for the Anxiety and Depressive Symptoms scales were revised. Unfortunately, we were not particularly successful at extending the range of measurement for the physical functioning banks. CONCLUSIONS: The present study expanded PROMIS pediatric item banks to add new content and to increase the range of measurement. Using item response theory, the banks were revised and expanded without changing the underlying scale of measurement. For Anger, Anxiety, and Depressive Symptoms, we successfully added new content that may render those banks more robust and flexible.

The diagnostic adaptive behavior scale: evaluating its diagnostic sensitivity and specificity.

The Diagnostic Adaptive Behavior Scale (DABS) was constructed with items across three domains--conceptual, social, and practical adaptive skills--and normed on a representative sample of American individuals from 4 to 21 years of age. The DABS was developed to focus its assessment around the decision point for determining the presence or absence of significant limitations of adaptive behavior for the diagnosis of Intellectual Disability (ID). The purpose of this study, which was composed of 125 individuals with and 933 without an ID-related diagnosis, was to determine the ability of the DABS to correctly identify the individuals with and without ID (i.e., sensitivity and specificity). The results indicate that the DABS sensitivity coefficients ranged from 81% to 98%, specificity coefficients ranged from 89% to 91%, and that the Area Under the Receiver Operating Characteristic Curve were excellent or good. These results indicate that the DABS has very good levels of diagnostic efficiency.