Advertisement

Psychometric Evaluation of the Korean Version of PROMIS Self-Efficacy for Managing Symptoms Item Bank: Item Response Theory

Open AccessPublished:August 30, 2022DOI:https://doi.org/10.1016/j.anr.2022.08.003

      Summary

      Purpose

      To evaluate the psychometric properties of the Patient-Reported Outcomes Measurement Information System (PROMIS) self-efficacy for managing symptoms of the version 1.0 item bank in Korea.

      Methods

      This study consisted of two phases: first, developing the Korean version of the item bank following the translation guidelines; and second, performing a cross-sectional study to evaluate its psychometric properties using the item response theory. This study enrolled 323 patients with type 2 diabetes mellitus between July and August 2020. Cronbach's α was used to assess the reliability of this item bank. Confirmatory factor analysis, using diagonally weighted least squares, was used to identify the assumptions of item response theory. Item parameter estimates including discrimination and thresholds were derived using the graded response model of the item response theory to reflect patient-reported outcomes as individualized responses.

      Results

      The Korean version of the item bank demonstrated good reliability (Cronbach's α = .98) and its discrimination ranged from 1.82 to 4.93. The thresholds resulted in the establishment of a category response curve for each item. However, no overlap was observed among the category curves. Moreover, the differential item functioning was not significant for age, gender, and income variables.

      Conclusion

      The graded response model and differential item functioning provided qualitative evidence that demonstrated acceptable psychometric properties of symptom management self-efficacy among patients. This item bank is expected to provide adequate assessments of self-efficacy of symptom management for patients with a chronic disease, which can contribute to nursing research and intervention.

      Keywords

      Abbreviations:

      CFA (Confirmatory Factor Analysis), CFI (Comparative Fit Index), COSMIN (Consensus-based Standards for the Selection of the Health Measurement Instruments), DIF (Differential Item Functioning), D-SMART (Diabetes Self-Management Assessment Report Tool), DWLS (Diagonally Weighted Least Squares), FACIT (Functional Assessment of Chronic Illness Therapy), GRM (Graded Response Model), HbA1c (hemoglobin A1c), IRT (Item Response Theory), PHO (PROMIS Health Organization), PRO (Patient-Reported Outcomes), PROMIS (Patient-Reported Outcomes Measurement Information System), RMSEA (Root Mean Square Error of Approximation), SDSCA (Summary of Diabetes Self-Care Activities Questionnaire), TLI (Tucker-Lewis Index)

      Introduction

      The Patient-Reported Outcomes Measurement Information System (PROMIS) was established in 2004 to develop improved patient-reported outcomes (PRO) [
      • Ader D.N.
      Developing the patient-reported outcomes measurement information system (PROMIS).
      ]. The multicenter collaborative PROMIS has produced more than 300 item banks within the physical, mental, and social domains. The PROMIS scales are advantageous for their high precision, which facilitates the assessment of a wide range of various aspects regarding patients’ contexts. The information compiled by the PROMIS Health Organization has been translated into multiple languages and used worldwide [
      • Jakob T.
      • Nagl M.
      • Gramm L.
      • Heyduck K.
      • Farin E.
      • Glattacker M.
      Psychometric properties of a German translation of the PROMIS® depression item bank.
      ]; thus, the study developed the PROMIS self-efficacy for managing symptoms item bank (version 1.0) in Korean, using the item response theory (IRT) and evaluated its psychometric properties.
      The PRO is pertinent for establishing a scientific framework for patient experience in healthcare research [
      • Willke R.J.
      • Burke L.B.
      • Erickson P.
      Measuring treatment impact: a review of patient-reported outcomes and other efficacy endpoints in approved product labels.
      ,
      US Department of Health and Human Services
      Food and drug administration, center for drug evaluation and research, center for biologics evaluation and research, center for devices and radiological health.
      ]. The United States National Institutes of Health recognized the need for PRO measurement tools to ensure validity and reliability in high-quality care [
      • Ader D.N.
      Developing the patient-reported outcomes measurement information system (PROMIS).
      ]. There has been a significant demand from patients for the expression and measurements of their “real” symptoms and experiences [
      • Solberg L.I.
      • Peterson K.A.
      • Fu H.
      • Eder M.
      • Jacobsen R.
      • Carlin C.S.
      Strategies and factors associated with top performance in primary care for diabetes: insights from a mixed methods study.
      ]. The PROMIS item banks from a physical category have been translated and validated more frequently than those in the psychosocial health categories [
      • Abma I.L.
      • Butje B.J.D.
      • Ten Klooster P.M.
      • van der Wees P.J.
      Measurement properties of the Dutch-Flemish patient-reported outcomes measurement information system (PROMIS) physical function item bank and instruments: a systematic review.
      ,
      • Kaat A.J.
      • Buckenmaier 3rd, C.T.
      • Cook K.F.
      • Rothrock N.E.
      • Schalet B.D.
      • Gershon R.C.
      • et al.
      The expansion and validation of a new upper extremity item bank for the Patient-Reported Outcomes Measurement Information System® (PROMIS).
      ].
      The evaluation of self-care abilities among patients with chronic diseases is important for the maintenance, monitoring, and management of their medical information. According to the self-care of chronic illness theory, the improved management of chronic diseases results in positive self-care outcomes [
      • Riegel B.
      • Jaarsma T.
      • Lee C.S.
      • Strömberg A.
      Integrating symptoms into the middle-range theory of self-care of chronic illness.
      ]. As shown by the health action process approach theory, initiating health-related behaviors, such as self-care, requires a pre-intentional motivation process. In a previous study, it was reported that self-efficacy had an effect on self-care in patients with chronic diseases such as heart failure, asthma, and hypertension [
      • Riegel B.
      • Dickson V.V.
      • Vellone E.
      The situation-specific theory of heart failure self-care: an update on thepProblem, person, and environmental factors influencing heart failure self-care.
      ,
      • Bodenheimer T.
      • Lorig K.
      • Holman H.
      • Grumbach K.
      Patient self-management of chronic disease in primary care.
      ,
      • Chalfont G.
      • Mateus C.
      • Varey S.
      • Milligan C.
      Self-efficacy of older people using technology to self-manage COPD, hypertension, heart failure, or dementia at home: an overview of systematic reviews.
      ]. Self-efficacy is relevant to this process, as it is the belief in one's own abilities to complete a task or achieve a goal [
      • Schwarzer R.
      • Renner B.
      Social-cognitive predictors of health behavior: action self-efficacy and coping self-efficacy.
      ,
      • Bandura A.
      Self-efficacy: the exercise of control.
      ]. Thus, self-efficacy for managing symptoms refers to a set of patients' beliefs about their ability to control their symptoms successfully.
      The PROMIS self-efficacy scales for managing chronic conditions fall within the mental health category [

      Northwestern University; HealthMeasures. List of adult measures. PROMIS® [Internet]. [updated 2021 April 28; cited 2022 May 4]. Available from: https://www.healthmeasures.net/explore-measurement-systems/promis/intro-to-promis/list-of-adult-measures.

      ]. Patients are impacted by various needs and symptoms depending on their respective contexts; hence, evaluating patients can help to provide effective individualized care [
      • Rossiter C.
      • Levett-Jones T.
      • Pich J.
      The impact of person-centred care on patient safety: an umbrella review of systematic reviews.
      ,
      • Yun D.
      • Choi J.
      Person-centered rehabilitation care and outcomes: a systematic literature review.
      ]. Standardized PRO measurements are necessary to evaluate the patients’ cultural backgrounds, which are done during psychometric evaluations. Standardizing PRO measurements is crucial because multiple understandings could arise from different cultural backgrounds, even in the same given sentence [
      • Zhang P.
      • Ouyang Z.
      • Fang S.
      • He J.
      • Fan L.
      • Luo X.
      • et al.
      Personality inventory for DSM-5 brief form(PID-5-BF) in Chinese students and patients: evaluating the five-factor model and a culturally informed six-factor model.
      ].
      According to the evidence, chronic diseases have consistent guidelines that include symptoms management and complication prevention. However, treatment goals and management processes vary among patients [
      • Kucharski D.
      • Lange E.
      • Ross A.B.
      • Svedlund S.
      • Feldthusen C.
      • Önnheim K.
      • et al.
      Moderate-to-high intensity exercise with person-centered guidance influences fatigue in older adults with rheumatoid arthritis.
      ,
      • Siebolds M.
      • Gaedeke O.
      • Schwedes U.
      Self-monitoring of blood glucose–psychological aspects relevant to changes in HbA1c in type 2 diabetic patients treated with diet or diet plus oral antidiabetic medication.
      ]. With an understanding of the patients' integrative context, nurses should be able to make sound clinical judgments [
      • Kautz D.D.
      • Kuiper R.
      • Pesut D.J.
      • Knight-Brown P.
      • Daneker D.
      Promoting clinical reasoning in undergraduate nursing students: application and evaluation of the Outcome Present State Test (OPT) model of clinical reasoning.
      ]. For instance, a previous scoping review study emphasized the increasing need for cross-cultural studies that analyze indicators of Diabetes Mellitus (DM) in the context of the patients’ life and experiences. Particularly, social factors can be considered in order to manage blood glucose levels [
      • Martin-Delgado J.
      • Guilabert M.
      • Mira-Solves J.
      Patient-reported experience and outcome measures in people living with diabetes: a scoping review of instruments.
      ]. Nurses' monitoring and intervention to manage chronic disease patients' symptoms make up a substantial axis of social factors. Thus, measuring self-efficacy for managing symptoms using the PROMIS self-efficacy scales is essential for patients with chronic disease.
      Self-efficacy of patients with chronic disease for self-care and symptom management is a significant topic that has continuously piqued interest in healthcare [
      • Ghizzardi G.
      • Arrigoni C.
      • Dellafiore F.
      • Vellone E.
      • Caruso R.
      Efficacy of motivational interviewing on enhancing self-care behaviors among patients with chronic heart failure: a systematic review and meta-analysis of randomized controlled trials.
      ,
      • Kalluri M.
      • Younus S.
      • Archibald N.
      • Richman-Eisenstat J.
      • Pooler C.
      Action plans in idiopathic pulmonary fibrosis: a qualitative study-’I do what I can do’.
      ,
      • Hansen C.W.
      • Esbensen B.A.
      • de Thurah A.
      • Christensen R.
      • de Wit M.
      • Cromhout P.F.
      Outcome measures in rheumatology applied in self-management interventions targeting people with inflammatory arthritis A systematic review of outcome domains and measurement instruments.
      ,
      • Hosseini A.
      • Jackson A.C.
      • Chegini N.
      • Dehghan M.F.
      • Mazloum D.
      • Haghani S.
      • et al.
      The effect of an educational app on hemodialysis patients' self-efficacy and self-care: a quasi-experimental longitudinal study.
      ], and instruments have been developed in response [
      • Lorig K.
      • Chastain R.L.
      • Ung E.
      • Shoor S.
      • Holman H.R.
      Development and evaluation of a scale to measure perceived self-efficacy in people with arthritis.
      ,
      • Nicholas M.K.
      The pain self-efficacy questionnaire: taking pain into account.
      ,
      • Lev E.L.
      • Owen S.V.
      A measure of self-care self-efficacy.
      ]. A systematic review of the self-efficacy instruments for patients with chronic diseases reported that most instruments had unclear purposes and measurement properties [
      • Frei A.
      • Svarin A.
      • Steurer-Stey C.
      • Puhan M.A.
      Self-efficacy instruments for patients with chronic diseases suffer from methodological limitations–a systematic review.
      ]. The widely used self-efficacy scale [
      • Sherer M.
      • Maddux J.E.
      The self-efficacy scale: construction and validation.
      ] is limited to general aspects of self-efficacy and not for assessment of patients' self-efficacy in managing symptoms. There is a need to assess patients’ psychological readiness for the management of complications or acute exacerbations through the incorporation of voluntary self-care strategies. The PROMIS self-efficacy for managing symptoms item bank assesses self-efficacy in a variety of domains, ranging from daily symptom management activities to strategies for coping with unexpected changes.
      The original PROMIS item banks were developed using the IRT model [
      • Rothrock N.E.
      • Amtmann D.
      • Cook K.F.
      Development and validation of an interpretive guide for PROMIS scores.
      ,
      • Reeve B.B.
      • Hays R.D.
      • Bjorner J.B.
      • Cook K.F.
      • Crane P.K.
      • Teresi J.A.
      • et al.
      Psychometric evaluation and calibration of health-related quality of life item banks: plans for the Patient-Reported Outcomes Measurement Information System (PROMIS).
      ]. The IRT analysis highlighted the functions of each item and encompassed the characteristics of items in the whole measurement [
      • Lee E.H.
      • Kang E.H.
      • Kang H.J.
      Evaluation of studies on the measurement properties of self-reported instruments.
      ]. With regard to measuring, IRT is concerned with the item of measurement, whereas classical test theory depends on the entire measurement [
      • Song Y.
      • Kim H.
      • Park S.Y.
      An item response theory analysis of the Korean version of the CRAFFT scale for alcohol use among adolescents in Korea.
      ]. Using IRT, it is possible to determine how each item contributes to a total measurement and how each item performs on the measurement [
      • Nguyen T.H.
      • Han H.R.
      • Kim M.T.
      • Chan K.S.
      An introduction to item response theory for patient-reported outcome measurement.
      ]. Each PROMIS item bank measures specific categories and domains and is considered a one-factor model [
      • Abma I.L.
      • Butje B.J.D.
      • Ten Klooster P.M.
      • van der Wees P.J.
      Measurement properties of the Dutch-Flemish patient-reported outcomes measurement information system (PROMIS) physical function item bank and instruments: a systematic review.
      ,
      • Rose M.
      • Bjorner J.B.
      • Gandek B.
      • Bruce B.
      • Fries J.F.
      • Ware Jr., J.E.
      The PROMIS Physical Function item bank was calibrated to a standardized metric and shown to improve measurement efficiency.
      ]. The PROMIS seeks to expand the understanding of patients’ experiences by using item banks in the global healthcare domain. Therefore, developing the PROMIS item bank in a different language through strict and systematic methods can help generate individualized PRO evidence.
      This study aimed to develop the PROMIS self-efficacy to manage symptoms using the version 1.0 item bank that has been translated and adapted culturally to Korean. Furthermore, to investigate psychometrics using the IRT model for patients with type 2 DM.

      Methods

      Design

      This is a methodological study designed to evaluate the validity and reliability of the Korean version of PROMIS self-efficacy for managing symptoms item bank (version 1.0) with original data from the survey.
      The current study comprises two main phases. First, the Korean version of PROMIS self-efficacy for managing symptoms item bank was developed. The details of the first step are described in the following section, “Translation including cross-cultural context.” Second, a cross-sectional study was conducted to evaluate the psychometric properties of the final version of the Korean item bank. After the survey, raw data was analyzed using the IRT model, in accordance with the reporting checklist for PROMIS [
      • Hanmer J.
      • Jensen R.E.
      • Rothrock N.
      A reporting checklist for HealthMeasures' patient-reported outcomes: ASCQ-Me, Neuro-QoL, NIH Toolbox, and PROMIS.
      ]. Furthermore, this study adhered to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [
      • von Elm E.
      • Altman D.G.
      • Egger M.
      • Pocock S.J.
      • Gøtzsche P.C.
      • Vandenbroucke J.P.
      The strengthening the reporting of observational studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies.
      ].

      Translation process and validation

      The license agreement to translate the PROMIS self-efficacy for managing symptoms item bank (version 1.0) into Korean was obtained from the PROMIS Health Organization (PHO). The PHO presented the guidelines for translation and development. Figure 1 presents an overview of the translation process. The translation process followed the PROMIS translation guidelines based on the Functional Assessment of Chronic Illness Therapy (FACIT) translation methodology [
      • Bonomi A.E.
      • Cella D.F.
      • Hahn E.A.
      • Bjordal K.
      • Sperner-Unterweger B.
      • Gangeri L.
      • et al.
      Multilingual translation of the Functional Assessment of Cancer Therapy (FACT) quality of life measurement system.
      ,
      • Eremenco S.L.
      • Cella D.
      • Arnold B.J.
      A comprehensive method for the translation and cross-cultural validation of health status questionnaires.
      ]. One of two independent Korean speakers translated the first version of the item bank, and the other reviewed the first version. A Korean-English bilingual translator back-translated the version, and three Korean speakers reviewed the back-translation. All the translators used during this process were healthcare providers. The research team finalized the translated item bank and reached a consensus with the PROMIS center. Thereafter, five Korean patients with type 2 DM were enrolled in the cognitive testing and linguistic validation process. The study research team reported the cognitive interview results to the PROMIS center. The final version of the Korean version of PROMIS self-efficacy for managing symptoms item bank was completed based on the cognitive interview report and discussion.
      Figure 1
      Figure 1Translation Process of the Korean Version of PROMIS Self-efficacy for Managing Symptoms Item Bank. Note. PROMIS=Patient-Reported Outcomes Measurement Information System, DM = Diabetes Mellitus.

      Sample/Participants

      A total of 354 patients with type 2 DM were recruited using convenience sampling from the DM center of a tertiary hospital in Seoul, South Korea. The study participants were adults aged 19 and above and had volunteered to participate in completing the questionnaires between July and August 2020. We chose type 2 DM patients as the study population. The first reason is to reduce participant heterogeneity and to control exogenous variables for psychometric evaluation. Second, DM is one of the most common chronic diseases that can be treated and managed with regular evaluations and treatments such as diet, physical activity, and medication [
      World Health Organization
      WHO website Health topics: Diatetes [Internet].
      ]. It thus becomes vital for patients with type 2 DM to manage symptoms through medication and treatment [
      • Yazdani F.
      • Abazari P.
      • Haghani F.
      • Iraj B.
      Restrictors of the effectiveness of diabetes self-management education: a qualitative content analysis.
      ].
      A total sample of 323 patients (91.2%) completed the survey, and thus were included in the study utilized. Confirmatory factor analysis (CFA) was used to test the assumptions of the IRT model. The minimum sample size for CFA was 200 cases [
      • Polit D.F.
      • Yang F.M.
      Measurement and the measurement of change.
      ], with a previous study reporting that the IRT model can be applied to at least 200 patients depending on the model complexity in healthcare research [
      • Forero C.G.
      • Maydeu-Olivares A.
      Estimation of IRT graded response models: limited versus full information methods.
      ]. Moreover, the sample size used in this study exceeded the minimum criterion for the IRT model.

      Data Collection

      Instrument

      The main instrument used was the PROMIS self-efficacy for managing symptoms version 1.0 item bank for adults, which includes 28 items. A five-point Likert scale was used to assess the responses, ranging from a scale of “1 = not at all confident” to “5 = very confident.” A higher score implies an increased sense of self-efficacy in managing symptoms. This item bank measures patients’ self-efficacy regarding the degree to which symptoms are controllable and the ability to prevent symptoms from worsening.
      In addition, two measurements were used to analyze the convergent validity, namely the Diabetes Self-Management Assessment Report Tool (D-SMART) and the Korean version of the Summary of Diabetes Self-Care Activities Questionnaire (SDSCA), after approval from the original authors. These two instruments have established good validity and reliability in previous studies.
      The original version of D-SMART was developed by the American Association of Diabetes Educators via Peyrot and colleagues to assess the self-management behavior of patients with DM [
      • Peyrot M.
      • Peeples M.
      • Tomky D.
      • Charron-Prochownik D.
      • Weaver T.
      Development of the American Association of Diabetes Educators' diabetes self-management assessment report tool.
      ]. The Korean version of D-SMART was used in previous studies [
      • Lee H.
      • Choi E.K.
      • Kim H.
      • Kim H.
      • Kim H.
      Factors affecting the self-management of adolescents with type 1 diabetes mellitus based on the information-motivation-behavioral skills model.
      ,
      • Lee S.
      • Kim H.
      Structural equation modeling on self-care behavior and quality of life in older adults with diabetes using citizen health promotion centers.
      ]. Among the D-SMART questions, 23 items were used to evaluate the self-management skills confidence [
      • Peyrot M.
      • Peeples M.
      • Tomky D.
      • Charron-Prochownik D.
      • Weaver T.
      Development of the American Association of Diabetes Educators' diabetes self-management assessment report tool.
      ]. The evaluation of skills confidence for DM self-management behavior in seven categories, including exercise/activity, nutrition, medication, and monitoring, is conducted using a 4-point Likert scale, with higher scores indicating greater skills confidence. In this study, Cronbach's alpha value of scale was 0.91.
      Toobert and colleagues revised SDSCA in 2000, which is used mainly in self-management activity studies for patients with DM and consists of 25 items, including six subscales: general and specific diet, exercise, blood sugar test, foot care, and smoking [
      • Toobert D.J.
      • Hampson S.E.
      • Glasgow R.E.
      The summary of diabetes self-care activities measure: results from 7 studies and a revised scale.
      ]. Chang and Song (2009) translated and modified the revised SDSCA in Korean and it has 17 items, excluding eight items that could not be scored [
      • Chang S.
      • Song M.-S.
      The validity and reliability of a Korean version of the summary of diabetes self-care activities questionnaire for older patients with type 2 diabetes.
      ], and five domains—diet, exercise, medication, blood sugar test, and foot care—were included [
      • Chang S.
      Structural equation moedling on health-related quality of life in older adutls with type 2 diabetes mellitus [dissertation].
      ]. This measurement asks participants to indicate on an 8-point scale (‘0 day’ to ‘7 days’), the number of days they engaged in self-care activities corresponding to each item during the previous week. Cronbach's alpha for the Korean version of this study was as follows: 0.58 for diet, 0.80 for diet, 0.36 for medication, 0.92 for blood sugar test, and 0.63 for foot care.

      Assumptions of the IRT

      The IRT model requires several robust assumptions, namely: unidimensionality, invariance, local independence, and monotonicity [
      • Polit D.F.
      • Yang F.M.
      Measurement and the measurement of change.
      ]. First, the CFA and coefficient omega (ωh) were used to analyze unidimensionality and invariance [
      • Mokkink L.B.
      • Prinsen C.
      • Patrick D.L.
      • Alonso J.
      • Bouter L.M.
      • De Vet H.
      • et al.
      COSMIN methodology for systematic reviews of patient-reported outcome measures (PROMs).
      ,
      • Choi H.
      • Kim C.
      • Ko H.
      • Park C.G.
      Translation and validation of the Korean version of PROMIS® pediatric and parent proxy measures for emotional distress.
      ]. The criteria of the CFA results of unidimensionality required the comparative fit index (CFI) or Tucker-Lewis Index (TLI) to exceed .95 or root mean square error of approximation (RMSEA) to be less than 0.06 [
      • Mokkink L.B.
      • Prinsen C.
      • Patrick D.L.
      • Alonso J.
      • Bouter L.M.
      • De Vet H.
      • et al.
      COSMIN methodology for systematic reviews of patient-reported outcome measures (PROMs).
      ]. In addition, the results of ωh were used to assess unidimensionality [
      • Choi H.
      • Kim C.
      • Ko H.
      • Park C.G.
      Translation and validation of the Korean version of PROMIS® pediatric and parent proxy measures for emotional distress.
      ]. The generally accepted criterion for ωh is .70 [
      • Reise S.P.
      • Scheines R.
      • Widaman K.F.
      • Haviland M.G.
      Multidimensionality and structural coefficient bias in structural equation modeling: a bifactor perspective.
      ]. Second, the chi-square (χ2) value assessed whether the model was fit for invariance. When the p-value of χ2 was not statistically significant, it was considered an appropriate model fit [
      • Mokkink L.B.
      • Prinsen C.
      • Patrick D.L.
      • Alonso J.
      • Bouter L.M.
      • De Vet H.
      • et al.
      COSMIN methodology for systematic reviews of patient-reported outcome measures (PROMs).
      ]. Confirming the assumption with χ2 is a theoretical concept, and every case does not meet the χ2 assumption. When χ2 was not satisfied, it could be assumed that each subgroup has a varied differential item functioning (DIF) [
      • Nguyen T.H.
      • Han H.R.
      • Kim M.T.
      • Chan K.S.
      An introduction to item response theory for patient-reported outcome measurement.
      ]; therefore, age, gender, and income were selected as the anchor variables to confirm the DIF in this study. Thirdly, using Yen's Q3, local independence was tested by residual correlations [
      • Yen W.M.
      Effects of local item dependence on the fit and equating performance of the three-parameter logistic model.
      ]. A study reported that local independence did not have a single critical value [
      • Christensen K.B.
      • Makransky G.
      • Horton M.
      Critical values for Yen's Q3: identification of local dependence in the Rasch model using residual correlations.
      ]. However, based on previous research and consensus-based standards for the selection of the health measurement instruments (COSMIN) manual for systematic reviews of PROMs, this study established criteria: <0.37 is suitable, and <0.7 is considered possible [
      • Mokkink L.B.
      • Prinsen C.
      • Patrick D.L.
      • Alonso J.
      • Bouter L.M.
      • De Vet H.
      • et al.
      COSMIN methodology for systematic reviews of patient-reported outcome measures (PROMs).
      ,
      • González-de Paz L.
      • Kostov B.
      • López-Pina J.A.
      • Solans-Julián P.
      • Navarro-Rubio M.D.
      • Sisó-Almirall A.
      A Rasch analysis of patients' opinions of primary health care professionals' ethical behaviour with respect to communication issues.
      ]. Finally, monotonicity was supported by an adequate graph of discrimination and thresholds [
      • Mokkink L.B.
      • Prinsen C.
      • Patrick D.L.
      • Alonso J.
      • Bouter L.M.
      • De Vet H.
      • et al.
      COSMIN methodology for systematic reviews of patient-reported outcome measures (PROMs).
      ,
      • Pinto M.
      • Pinto R.M.C.
      • Mendonça T.
      • Souza C.G.
      • da Silva C.H.M.
      Validation and calibration of the patient-reported outcomes measurement information system: pediatric PROMIS® Emotional Distress domain item banks, Portuguese version (Brazil/Portugal).
      ].

      Data analysis

      The data was analyzed using SPSS (version 25.0; IBM, Armonk, NY, USA) and the lavaan, psych, mirt, and lordif packages in R version 4.1.2. A descriptive statistical test was performed for the demographic and clinical variables. Univariate normality was confirmed before analysis to identify the selection bias of the study. Cronbach's α coefficients were used to confirm the reliability of the measurements. This study applied diagonally weighted least squares (DWLS) to determine the CFA results using the lavaan package in R, because the item bank was an ordinal variable, and the ceiling effect was identified [
      • Li C.-H.
      Confirmatory factor analysis with ordinal data: comparing robust maximum likelihood and diagonally weighted least squares.
      ,
      • Rosseel Y.
      Lavaan: an R package for structural equation modeling and more.
      ]. The ωh were estimated using the psych package, and the residual correlation was tested using mirt package in R.

      T-score

      Following the PROMIS scoring guide, the standardized T-score was used in this study [
      • Martin-Delgado J.
      • Guilabert M.
      • Mira-Solves J.
      Patient-reported experience and outcome measures in people living with diabetes: a scoping review of instruments.
      ]. T-score is a standard score of reference samples including United States (U.S.) general population [
      • Rothrock N.E.
      • Amtmann D.
      • Cook K.F.
      Development and validation of an interpretive guide for PROMIS scores.
      ]. The underlying T-score of the self-efficacy for managing symptoms item bank was calibrated to reach an average of 50, with a standard deviation of ±10 for the U.S. clinical sample. The PROMIS center provides the PROMIS T-score maps on the website for some short-form item banks. The T-score was obtained using the website of the Health Measures Scoring Service (powered by the Assessment Center℠) that provides underlying item parameters and scoring for the U.S.

      IRT model

      This study used the graded response model (GRM) of the IRT model because the item bank has ordered categories, such as the Likert scale [
      • Polit D.F.
      • Yang F.M.
      Measurement and the measurement of change.
      ]. For the GRM, discrimination and thresholds were estimated, and category response curves were derived. The IRT model was implemented to reflect the patients’ ability level for psychometric evaluation using the mirt package in R [
      • Polit D.F.
      • Yang F.M.
      Measurement and the measurement of change.
      ,
      • Andrich D.
      An extension of the Rasch model for ratings providing both location and dispersion parameters.
      ].

      Differential item functioning

      The DIF was analyzed to evaluate the validity of this item bank, which was constructed using a five-point Likert ordinal scale. Three group variables, including age, gender, and income, were used to analyze whether each question functions differently between groups. Among the group variables, the age group was divided into under 60 years [
      ] and above, with a male gender group as a reference. The income group was divided into less than four million South Korean won [
      Statistics Korea
      2020 Korean social indicators.
      ] and more.
      The lordif package used the ordinal logistic regression model for DIF estimating methods [
      • von Elm E.
      • Altman D.G.
      • Egger M.
      • Pocock S.J.
      • Gøtzsche P.C.
      • Vandenbroucke J.P.
      The strengthening the reporting of observational studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies.
      ]. The DIF analysis was conducted in two steps. First, the likelihood ratio χ2 test was carried out without using the anchor item. Second, the DIF item was extracted from 28 items. The DIF can be categorized as either a uniform DIF (if the effect is constant) or a non-uniform DIF (if the effect varies depending on the trait level) [
      • Güler N.
      • Penfield R.D.
      A comparison of the logistic regression and contingency table methods for simultaneous detection of uniform and nonuniform DIF.
      ,
      • Choi S.W.
      • Gibbons L.E.
      • Crane P.K.
      lordif: an R Package for detecting Differential Item Functioning using iterative hybrid ordinal logistic regression/item response theory and Monte Carlo simulations.
      ]. The χ2 difference test (df = 1) was conducted for each of the two types of DIF using logistic regression. The overall χ2 difference test (df = 2) for the total DIF was identified for the two inclusive types of DIF effect. A significance level of .01 was used as the criterion for each χ2 test. Thereafter, the DIF was evaluated using the items that were not extracted during the first step as anchor items. In this step, at least 2.0% of the items within McFadden's pseudo R2-change were extracted as a DIF [
      • Rimehaug S.A.
      • Kaat A.J.
      • Nordvik J.E.
      • Klokkerud M.
      • Robinson H.S.
      Psychometric properties of the PROMIS-57 questionnaire, Norwegian version.
      ].

      Results

      Demographics and clinical characteristics

      The average age of the patients was 62.16 ± 10.54 years, with a DM period of 14.23 ± 10.33 years in this study. Male patients made up 68.4% of the participants, and the majority of the participants were married (91.6%). Monthly income was reported as less than four million South Korean won by 52.3% of participants and as more than four million South Korean won by 47.7% of participants. The participants' average Body Mass Index (BMI) was 25.11 ± 3.57 kg/m2, ranging from 16.60 to 42.82 kg/m2. The average recent hemoglobin A1c (HbA1c), which measures the amount of glucose attached to hemoglobin, was 7.5 ± 1.5% according to patients’ electronic health records. The majority of participants (90.1%) managed their DM through oral administration, 33.4% via insulin injection, and 26.9% through a combination of medication and insulin. Most of the participants did not receive DM group education (77.4%) and managed their DM through administering oral medications (90.1%).

      Item analysis

      In total, 28 items were analyzed using mean and standard deviation (Table 1). Considering the criteria that the average value should be between 1.5 and 4.5 on a five-point Likert scale, all items were within the range [
      • Meir E.I.
      • Gati I.
      Guidelines for item selection in inventories yielding score profiles.
      ]. The patients in this study reported moderate self-efficacy for managing symptoms (T = 52.6, SD = 8.25). The T-Score differed merely by two points as compared to the T-score derived for the general U.S. population. Baseline self-efficacy for managing symptoms (T = 51.38, SD = 8.353) improved after two weeks of follow-up (T = 53.82, SD = 7.98).
      Table 1Item Analysis of the Korean Version of PROMIS Self-Efficacy for Managing Symptoms Item Bank (n = 323).
      ItemMean ± SDCronbach's α if deleted
      SEMSX0013.53 ± 1.16.98
      SEMSX0023.68 ± 1.07.98
      SEMSX0033.17 ± 1.26.98
      SEMSX0043.89 ± 1.05.98
      SEMSX0053.74 ± 1.11.98
      SEMSX0063.93 ± 1.05.98
      SEMSX0074.18 ± 0.88.98
      SEMSX0083.88 ± 1.01.98
      SEMSX0093.72 ± 1.14.98
      SEMSX0103.95 ± 1.01.98
      SEMSX0113.89 ± 1.02.98
      SEMSX0123.90 ± 1.05.98
      SEMSX0133.77 ± 1.09.98
      SEMSX0143.93 ± 1.01.98
      SEMSX0153.84 ± 1.08.98
      SEMSX0163.79 ± 1.05.98
      SEMSX0173.85 ± 1.03.98
      SEMSX0183.85 ± 1.07.98
      SEMSX0193.73 ± 1.08.98
      SEMSX0203.90 ± 1.03.98
      SEMSX0213.73 ± 1.14.98
      SEMSX0223.86 ± 1.02.98
      SEMSX0232.85 ± 1.03.98
      SEMSX0242.73 ± 1.08.98
      SEMSX0252.73 ± 1.08.98
      SEMSX0262.89 ± 1.03.98
      SEMSX0272.73 ± 1.14.98
      SEMSX0282.86 ± 1.02.98
      Total mean ± SD105.13 ± 23.38
      Minimum – Maximum35 – 140
      Coefficient α (Cronbach's α).98
      Coefficient omega (ωh).87
      Total T-score Mean ± SD52.6 ± 8.25
      Baseline T-score Mean ± SD51.38 ± 8.53
      Follow-up T-score Mean ± SD53.82 ± 7.98
      Note. SD = standard deviation.

      Reliability and convergent validity

      The Cronbach's α of this item bank was .98 (Table 1). All measures met the reliability criteria (>.70). In addition, if the items were deleted, lower levels of Cronbach's α would be observed as opposed to the total Cronbach's α (Table 1).
      This study tested convergent validity using the D-SMART and the revised SDSCA. The correlation coefficients of the item bank and D-SMART was r = .59 (p < .001). However, there was no statistical significance between the item bank and each domain of revised SDSCA: diet 0.11 (p = .054), exercise 0.11 (p = .059), medication −0.02 (p = .731), blood sugar test 0.05 (p = .420), and foot care 0.03 (p = .633).

      Assumptions and the expected scores curves for the IRT

      First, unidimensionality was the primary assumption for IRT [
      • Polit D.F.
      • Yang F.M.
      Measurement and the measurement of change.
      ,
      • Bond T.G.
      • Fox C.M.
      Applying the Rasch model: fundamental measurement in the human sciences.
      ]. For this study, the CFA results were verified by applying the DWLS. As PROMIS item banks were developed as a unidimensional model [
      • Sunderland M.
      • Batterham P.
      • Calear A.
      • Carragher N.
      Validity of the PROMIS depression and anxiety common metrics in an online sample of Australian adults.
      ,
      • Bjorner J.B.
      • Rose M.
      • Gandek B.
      • Stone A.A.
      • Junghaenel D.U.
      • Ware Jr., J.E.
      Difference in method of administration did not significantly impact item response: an IRT-based analysis from the Patient-Reported Outcomes Measurement Information System (PROMIS) initiative.
      ], we determined the unidimensionality by conducting CFA to test the convergent validity without exploratory factor analysis [
      • Lee E.-H.
      Psychometric properties of an instrument 2: structural validity, internal consistency, and cross-cultural validity/measurement invariance.
      ]. The estimation result of χ2 (df = 350) was 8809.65, and the model did not perfectly fit the data (p < .001). Both CFI and TLI of the Korean version of the item bank were 0.99 which met the criteria [
      • Mokkink L.B.
      • Prinsen C.
      • Patrick D.L.
      • Alonso J.
      • Bouter L.M.
      • De Vet H.
      • et al.
      COSMIN methodology for systematic reviews of patient-reported outcome measures (PROMs).
      ,
      • Hair J.F.
      • Anderson R.E.
      • Tatham R.L.
      • Black W.C.
      Multivariate data analysis with readings.
      ,
      • Hooper D.
      • Coughlan J.
      • Mullen M.R.
      Structural equation modelling: guidelines for determining model fit.
      ] but the RMSEA, 0.274, did not. As the COSMIN methodology for PRO measures [
      • Mokkink L.B.
      • Prinsen C.
      • Patrick D.L.
      • Alonso J.
      • Bouter L.M.
      • De Vet H.
      • et al.
      COSMIN methodology for systematic reviews of patient-reported outcome measures (PROMs).
      ] recommends that either CFI/TLI or RMSEA should be satisfied with the criteria for unidimensionality; thus, the unidimensionality of the Korean version of the item bank was identified. In addition, these results comply with the recommended value ωh for this item bank, which was .87 (Table 1). Second, the p-value of χ2 > .01 provides an appropriate criterion for the invariance model fit. The χ2 value of this item bank was 8809.649 (p < .001). Therefore, we determined the value of DIF to verify the invariance [
      • Nguyen T.H.
      • Han H.R.
      • Kim M.T.
      • Chan K.S.
      An introduction to item response theory for patient-reported outcome measurement.
      ]. Using age, gender, and income as anchor items, a DIF analysis was conducted, and the results confirmed that this model showed invariance. Third, the results of residual correlation among the items as a unidimensional model using Yen's Q3 were less than 0.37, except for the residual correlation between items 1 (SEMSX001) and 2 (SEMSX002), which ranged from −0.24 to 0.33 [
      • Mokkink L.B.
      • Prinsen C.
      • Patrick D.L.
      • Alonso J.
      • Bouter L.M.
      • De Vet H.
      • et al.
      COSMIN methodology for systematic reviews of patient-reported outcome measures (PROMs).
      ]. The residual correlation between items 1 and 2 was estimated to be 0.51, and a previous study confirmed local independence [
      • González-de Paz L.
      • Kostov B.
      • López-Pina J.A.
      • Solans-Julián P.
      • Navarro-Rubio M.D.
      • Sisó-Almirall A.
      A Rasch analysis of patients' opinions of primary health care professionals' ethical behaviour with respect to communication issues.
      ]. Lastly, all category response curves indicated an adequate monotonic relationship between the item thresholds and participants' self-efficacy ability. These inform the monotonicity of the basic assumptions for IRT. Figure 2 displays the example of the category response curve for item 22.
      Figure 2
      Figure 2Category Response Curve of the Korean Version of PROMIS Self-Efficacy for Managing Symptoms Item Bank Item 22 (SEMSX022). The graphs of P1 to P5 are Likert scale scores, one to five, of the item 22.

      Estimating graded response model

      The GRM evaluated item discrimination (a) and thresholds (b) based on participants’ response patterns regarding the IRT model (Table 2). Overall, the discrimination of this item bank was high, ranging from 1.82 to 4.93. The threshold values in the item bank were estimated in the order of low to high values according to the GRM (Table 2). For item 11, no patient selected the first category. Thus, the threshold of item 11 was analyzed using only four categories, from the initial two to five. When identifying the category response curve derived using the estimated item parameter, the category curve did not indicate complete overlap with another curve. The figures of the 28 items were interpreted to ensure that each item category had appropriate functions. An example of item 22 (SEMSX022) is shown in Figure 2.
      Table 2Estimated Item Parameters for the Korean Version of PROMIS Self-Efficacy for Managing Symptoms Item Bank Using the Graded Response Model.
      ItemDiscrimination a (SE)Threshold
      b1 (SE)b2 (SE)b3 (SE)b4 (SE)
      SEMSX0012.34 (.22)−2.43 (.24)−1.18 (.12)−0.08 (.09).73 (.10)
      SEMSX0021.99 (.19)−2.18 (.21)−0.84 (.11).27 (.09)1.15 (.13)
      SEMSX0032.52 (.23)−2.62 (.28)−1.10 (.12)−0.27 (.08).64 (.10)
      SEMSX0041.82 (.17)−1.98 (.20)−0.86 (.11).15 (.10)1.08 (.13)
      SEMSX0052.55 (.23)−1.84 (.17)−1.08 (.11)−0.19 (.08).76 (.10)
      SEMSX0062.42 (.23)−2.19 (.21)−1.38 (.13)−0.33 (.09).67 (.10)
      SEMSX0071.93 (.18)−1.73 (.17)−0.67 (.10).20 (.09)1.07 (.13)
      SEMSX0083.21 (.30)−2.44 (.23)−1.36 (.12)−0.51 (.08).30 (.08)
      SEMSX0092.68 (.25)−2.26 (.21)−1.22 (.12)−0.37 (.08).48 (.09)
      SEMSX0103.48 (.33)−2.37 (.22)−1.31 (.12)−0.52 (.08).26 (.08)
      SEMSX0113.50 (.35)N/A−1.62 (.14)−0.88 (.10).13 (.08)
      SEMSX0123.37 (.31)−2.50 (.25)−1.38 (.12)−0.55 (.08).37 (.08)
      SEMSX0132.52 (.24)−2.30 (.22)−1.10 (.11)−0.43 (.09).50 (.09)
      SEMSX0143.50 (.33)−2.23 (.21)−1.41 (.13)−0.56 (.08).30 (.08)
      SEMSX0154.20 (.40)−2.22 (.20)−1.30 (.11)−0.48 (.08).33 (.08)
      SEMSX0164.93 (.48)−2.10 (.18)−1.29 (.11)−0.49 (.08).32 (.07)
      SEMSX0174.75 (.45)−2.05 (.17)−1.23 (.11)−0.51 (.08).29 (.07)
      SEMSX0183.46 (.32)−2.21 (.21)−1.25 (.11)−0.43 (.08).36 (.08)
      SEMSX0193.50 (.32)−2.11 (.19)−1.14 (.11)−0.44 (.08).43 (.08)
      SEMSX0204.20 (.40)−2.44 (.25)−1.27 (.11)−0.55 (.08).28 (.07)
      SEMSX0213.33 (.31)−2.28 (.21)−1.18 (.11)−0.50 (.08).35 (.08)
      SEMSX0222.39 (.23)−2.50 (.25)−1.37 (.13)−0.47 (.09).54 (.09)
      SEMSX0232.30 (.22)−2.85 (.34)−1.44 (.14)−0.53 (.09).45 (.09)
      SEMSX0242.57 (.24)−2.25 (.22)−1.33 (.13)−0.62 (.09).43 (.09)
      SEMSX0252.90 (.27)−2.22 (.21)−1.21 (.12)−0.40 (.08).53 (.09)
      SEMSX0264.57 (.43)−1.99 (.17)−1.30 (11)−0.52 (.08).34 (.07)
      SEMSX0272.08 (.20)−2.32 (.23)−1.27 (.13)−0.50 (.09).53 (.10)
      SEMSX0283.70 (.34)−2.20 (.20)−1.28 (.12)−0.48 (.08).42 (.08)
      Range1.82 to 4.93−2.85 to −1.73−1.62 to −0.67−0.88 to .27.13 to 1.15
      Note. N/A = Not applicable, SE = standard error.

      Analyzing the DIF

      This study used three group variables for DIF analysis: age, gender, and income.
      First, as a result of conducting the likelihood ratio χ2 test using the age group variable, items 1 (SEMSX001), 21 (SEMSX021), and 27 (SEMSX027) had DIF. Items 1 and 21 represented the non-uniform DIF, and item 27 described the uniform DIF. All three DIF items (1, 21, and 27) had statistical significance in the total DIF effect (p < .001). Ordinal logistic regression was conducted again with other items, except these three, as the anchor item. Consequently, item 27 was identified with a McFaddens’ pseudo R2-change of over 2.0% or more; the R2-change value of uniform DIF was 3.4%, and the total DIF was 3.9% (p < .001). Figure 3 shows the test characteristic curves (TCC) of item 27. The effect of item 27 on the expected score of the entire item bank was interpreted to be minimal.
      Figure 3
      Figure 3Test Characteristic Curves (TCC) for Age Differential Item Functioning (DIF) in the Korean Version of PROMIS Self-Efficacy for Managing Symptoms Item Bank. The TCC total consequence of DIF of all items is left graph; the TCC for item 27(SEMSX027) with negligible DIF is right graph. Note. DIF = differential item functioning.
      Next, the χ2 test results according to the gender group variable were described. Item 27 had a uniform DIF (p < .008); however, the total DIF effect was not statistically significant (p = .211). After the ordinal logistic regression was re-conducted with the remaining items, excluding item 27, there was no item with McFadden's pseudo R2-change. Finally, the χ2 test was conducted with the income group variable, and no item indicated the DIF.

      Discussion

      This study developed the Korean version of PROMIS self-efficacy for managing symptoms item bank. The PROMIS item banks are globally used instruments to assess self-reported patient outcomes, which include integrative factors that identify patients as individualized people [
      • Nguyen T.H.
      • Han H.R.
      • Kim M.T.
      • Chan K.S.
      An introduction to item response theory for patient-reported outcome measurement.
      ]. Previous studies have translated psychometric evaluations into other languages using the IRT [
      • Chan S.W.W.
      • Chien C.W.
      • Wong A.Y.L.
      • Pang M.Y.C.
      Translation and psychometric validation of the traditional Chinese version of patient-reported outcomes measurement information system Pediatric-25 Profile version 2.0 (PROMIS-25) in Chinese Children with Cancer in Hong Kong.
      ,
      • Chen W.H.
      • Lenderking W.
      • Jin Y.
      • Wyrwich K.W.
      • Gelhorn H.
      • Revicki D.A.
      Is Rasch model analysis applicable in small sample size pilot studies for assessing item characteristics? An example using PROMIS pain behavior item bank data.
      ]. The IRT model underscores the functions of each item and outlines the item characteristics across the instrument [
      • Nguyen T.H.
      • Han H.R.
      • Kim M.T.
      • Chan K.S.
      An introduction to item response theory for patient-reported outcome measurement.
      ,
      • Roy C.
      • Bakan G.
      • Li Z.
      • Nguyen T.H.
      Coping measurement: creating short form of Coping and Adaptation Processing Scale using item response theory and patients dealing with chronic and acute health conditions.
      ]. Cleanthous and his colleagues verified that the IRT was suitable for PROMIS® measurement applications [
      • Cleanthous S.
      • Barbic S.P.
      • Smith S.
      • Regnault A.
      Psychometric performance of the PROMIS® depression item bank: a comparison of the 28- and 51-item versions using Rasch measurement theory.
      ]. The IRT model was advantageous for measuring human abilities, attitudes, and other attributes using actual survey data.
      Cronbach's α identified the reliability of this item bank as appropriate. This study used the D-SMART and the revised SDSCA to test convergent validity. The current item bank showed a significant correlation with D-SMART, which is evaluating the self-efficacy for self-management skills [
      • Lee E.-H.
      Psychometric properties of an instrument 2: structural validity, internal consistency, and cross-cultural validity/measurement invariance.
      ]. It indicates that the item bank was reliable and suited conceptually in terms of self-efficacy among participants of this study. On the contrary, none of the subdomains of the revised SDSCA, measuring self-care activities in the past week, showed statistical significance. A systematic review of measurements for self-care among DM patients reported that the revised SDSCA had low quality of comprehensiveness and comprehensibility [
      • Lee J.
      • Lee E.H.
      • Chae D.
      • Kim C.J.
      Patient-reported outcome measures for diabetes self-care: a systematic review of measurement properties.
      ]. This psychometric limitation of the revised SDSCA needs careful interpretation of the current result of convergent validity with the PROMIS item bank.
      This study partially fulfilled the four basic IRT assumptions. The study adopted the COSMIN guidelines even though there was no absolute standard for the criteria of IRT assumptions. The CFA was conducted to validate the unidimensionality of the original PROMIS scale. Since the item bank comprised ordinal data, the DWLS was selected for the estimation method in this study [
      • Nguyen T.H.
      • Han H.R.
      • Kim M.T.
      • Chan K.S.
      An introduction to item response theory for patient-reported outcome measurement.
      ]. A small sample of fewer than 200 participants may face an increased risk of an overestimated correlation using DWLS [
      • Li C.-H.
      Confirmatory factor analysis with ordinal data: comparing robust maximum likelihood and diagonally weighted least squares.
      ]. However, the number of participants in this study met this criterion (n = 326). The overall fit of this item bank fulfilled the requirements of the validity of CFA and supported the unidimensionality of CFI. The Root Mean Square Error of Approximation and Standard Root Mean Residual did not meet the inclusion criteria. These results implied the possibility that the Korean version of the PROMIS self-efficacy for managing symptoms item bank may possibly have a multiple factor structure. According to the original PROMIS item banks [
      • Rothrock N.E.
      • Amtmann D.
      • Cook K.F.
      Development and validation of an interpretive guide for PROMIS scores.
      ,
      • Reeve B.B.
      • Hays R.D.
      • Bjorner J.B.
      • Cook K.F.
      • Crane P.K.
      • Teresi J.A.
      • et al.
      Psychometric evaluation and calibration of health-related quality of life item banks: plans for the Patient-Reported Outcomes Measurement Information System (PROMIS).
      ] as well as previous studies based on a psychometric evaluation of the PROMIS item banks, analyses were performed using a single factor model [
      • Jakob T.
      • Nagl M.
      • Gramm L.
      • Heyduck K.
      • Farin E.
      • Glattacker M.
      Psychometric properties of a German translation of the PROMIS® depression item bank.
      ,
      • Choi H.
      • Kim C.
      • Ko H.
      • Park C.G.
      Translation and validation of the Korean version of PROMIS® pediatric and parent proxy measures for emotional distress.
      ,
      • Cleanthous S.
      • Barbic S.P.
      • Smith S.
      • Regnault A.
      Psychometric performance of the PROMIS® depression item bank: a comparison of the 28- and 51-item versions using Rasch measurement theory.
      ,
      • Gruber-Baldini A.L.
      • Velozo C.
      • Romero S.
      • Shulman L.M.
      Validation of the PROMIS® measures of self-efficacy for managing chronic conditions.
      ]. Since this study aimed to verify the results by applying PROMIS measurements to a Korean context, the IRT was performed without further modification of the items. Thus, further research is required to analyze the subcategories in the item bank across various settings and populations.
      Data for this study was collected from the diabetes center at a tertiary hospital in Korea. The participants displayed effective outcomes with regard to DM control. For example, the HbA1c was 7.5 ± 1.5%, performing lower than that reported in previous studies [
      • Lee E.-H.
      • Lee Y.W.
      • Lee K.-W.
      • Hong S.
      • Kim S.H.
      A New Objective Health Numeracy Test for patients with type 2 diabetes: development and evaluation of psychometric properties.
      ,
      • Choi S.
      • Kim S.H.
      Influences of patient activation on diabetes self-care activities and diabetes-specific distress.
      ]. In addition, over 90% of the participants controlled their glucose with oral medication. These results can result in the ceiling effect, indicating good control of their glucose levels. Ceiling effects negatively affect the CFA results [
      • Depaoli S.
      • Tiemensma J.
      • Felt J.M.
      Assessment of health surveys: fitting a multidimensional graded response model.
      ]. This study was analyzed using the DWLS in consideration of the ceiling effect. Statistical calibration serves as one method to solve this problem; however, the flooring or ceiling effects need to be considered when developing psychological evaluation tools such as self-efficacy instruments.
      As a result of the psychometric evaluation using the IRT model in this study, the Korean version of PROMIS self-efficacy for managing symptoms was a suitable instrument. The discrimination (a) range of the Korean version of this item bank was from 1.82 to 4.93. All the category response curves of the items were independent. The proper item showed discrimination that exceeded zero, indicating that the higher the values, the better the associated discrimination [
      • Polit D.F.
      • Yang F.M.
      Measurement and the measurement of change.
      ]. In a previous study that analyzed the PROMIS self-efficacy for managing daily activities item bank through the IRT model, discrimination was scored between 1.90 and 4.03 [
      • Hong I.
      • Velozo C.A.
      • Li C.Y.
      • Romero S.
      • Gruber-Baldini A.L.
      • Shulman L.M.
      Assessment of the psychometrics of a PROMIS item bank: self-efficacy for managing daily activities.
      ]. This is similar to the present study. The independent category response curves derived from the threshold (b) values indicated that the scale of the item (five-point Likert scale) had its own traits [
      • Polit D.F.
      • Yang F.M.
      Measurement and the measurement of change.
      ]. The results of category response curves suggest that each item of the Korean version of this item bank did not need to be tuned or revised.
      The major strength of this study was that it identified the global utility of the PROMIS item bank of self-efficacy for managing symptoms. The DIF results, comprising subgroups of age, gender, and income, suggested that specific general characteristics did not interfere with the total item bank. Psychological measurements generally target participants from various contexts. Each item should function similarly for the same ability of participants [
      • Wilson M.
      Constructing measures: an item response modeling approach.
      ]. In this study, item 27 was identified as the DIF in the age group variable. The test characteristic curve of item 27 (I can find the information I need to manage my symptoms), showed a negligible difference between the total and item 27 graphs. The confidence or ability to obtain health-related information was affected by the use and access level of digital devices [
      • Muscat D.M.
      • Cvejic E.
      • Bell K.
      • Smith J.
      • Morris G.M.
      • Jansen J.
      • et al.
      The impact of health literacy on psychosocial and behavioural outcomes among people at low risk of cardiovascular disease.
      ,
      • Scholz S.
      • Teetz L.
      Smart health via mHealth? Potentials of mobile health apps for improving prevention and adherence of breast cancer patients.
      ]. Although there was no significant difference observed from the graph, item 27 reflected the increased tendency of health literacy toward using smart devices to induce vulnerability among older adults [
      • Yu E.
      • Hagens S.
      Socioeconomic disparities in the demand for and use of virtual visits among senior adults during the COVID-19 pandemic: a cross-sectional study.
      ,
      • Jung S.O.
      • Son Y.H.
      • Choi E.
      E-health literacy in older adults: an evolutionary concept analysis.
      ]. This result suggests that nurses and nursing scientists should consider older adults’ self-efficacy for information-seeking behavior.

      Limitations

      The Korean version of PROMIS self-efficacy for managing symptoms can be used to enhance healthcare providers' understanding of patients with chronic diseases and to individualize care plans according to a person's self-efficacy. In addition, it has become possible to benchmark the self-efficacy of chronic diseases on a global level. However, this study has some limitations. First, the study was conducted in a single tertiary hospital and recruited patients with type 2 DM. Therefore, its application to patients with other chronic diseases may be limited. Since the item bank is intended for patients with chronic diseases in general, it is suggested that future studies expand to include other chronic diseases. Second, we evaluated convergent validity using self-care instruments that are frequently used in patients with DM, as the self-efficacy instrument for symptom management can rarely be found; thus, consideration needs to be given to the interpretation of the convergent validity results of this study. Finally, the data had a ceiling effect and the probability of multi-dimensionality. The statistical results indicated that the ceiling effect affected the outcome. In future studies, the inclusion of various patient groups or situations is required to evaluate the psychometric properties of self-efficacy while considering the ceiling effect and multi-dimensionality.

      Conclusions

      For the Korean version of PROMIS self-efficacy for managing symptoms item bank, the IRT model for psychometric testing was used. The results indicated decent reliability and validity of the measurement. Increasing self-efficacy for managing symptoms in patients with chronic diseases can play a significant role in improving the capability of maintaining their health. Thus, this instrument can facilitate healthcare providers’ evaluation of the degree of self-efficacy required to manage symptoms among patients as well as develop educational tools and interventions for their effective management.

      Funding

      This work was supported by funding [6-2018-0144] from Mo-Im Kim Nursing Research Institute and the Brain Korea 21 FOUR Project funded by National Research Foundation (NRF) of Korea, Yonsei University College of Nursing.

      Conflict of interest

      The authors declare no conflicts of interest.
      This manuscript is not under consideration by another journal and has not been published or presented elsewhere in part or in its entirety except Research Square of a preprint.

      Ethical approval

      This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted from the Institutional Review Board of Severance Hospital (4-2019-0257) prior to the translation and survey. The purpose and process of this study were explained by the researchers to the participants. The study participants were guaranteed confidentiality and voluntary participation and provided their written informed consent.

      Consent to participate

      Informed consent was obtained from all individual participants included in the study.

      References

        • Ader D.N.
        Developing the patient-reported outcomes measurement information system (PROMIS).
        Med Care. 2007; 45: S1-S2https://doi.org/10.1097/01.mlr.0000258615.42478.55
        • Jakob T.
        • Nagl M.
        • Gramm L.
        • Heyduck K.
        • Farin E.
        • Glattacker M.
        Psychometric properties of a German translation of the PROMIS® depression item bank.
        Eval Health Prof. 2017; 40: 106-120https://doi.org/10.1177/0163278715598600
        • Willke R.J.
        • Burke L.B.
        • Erickson P.
        Measuring treatment impact: a review of patient-reported outcomes and other efficacy endpoints in approved product labels.
        Control Clin Trials. 2004; 25: 535-552https://doi.org/10.1016/j.cct.2004.09.003
        • US Department of Health and Human Services
        Food and drug administration, center for drug evaluation and research, center for biologics evaluation and research, center for devices and radiological health.
        Guidance for Industry: Core Patient-Reported Outcomes in Cancer Clinical Trials. 2021; 2021: 1-8
        • Solberg L.I.
        • Peterson K.A.
        • Fu H.
        • Eder M.
        • Jacobsen R.
        • Carlin C.S.
        Strategies and factors associated with top performance in primary care for diabetes: insights from a mixed methods study.
        Ann Fam Med. 2021; 19: 110-116https://doi.org/10.1370/afm.2646
        • Abma I.L.
        • Butje B.J.D.
        • Ten Klooster P.M.
        • van der Wees P.J.
        Measurement properties of the Dutch-Flemish patient-reported outcomes measurement information system (PROMIS) physical function item bank and instruments: a systematic review.
        Health Qual Life Outcomes. 2021; 62: 19https://doi.org/10.1186/s12955-020-01647-y
        • Kaat A.J.
        • Buckenmaier 3rd, C.T.
        • Cook K.F.
        • Rothrock N.E.
        • Schalet B.D.
        • Gershon R.C.
        • et al.
        The expansion and validation of a new upper extremity item bank for the Patient-Reported Outcomes Measurement Information System® (PROMIS).
        J Patient Rep Outcomes. 2019; 69: 3https://doi.org/10.1186/s41687-019-0158-6
        • Riegel B.
        • Jaarsma T.
        • Lee C.S.
        • Strömberg A.
        Integrating symptoms into the middle-range theory of self-care of chronic illness.
        ANS Adv Nurs Sci. 2019; 42: 206-215https://doi.org/10.1097/ans.0000000000000237
        • Riegel B.
        • Dickson V.V.
        • Vellone E.
        The situation-specific theory of heart failure self-care: an update on thepProblem, person, and environmental factors influencing heart failure self-care.
        J Cardiovasc Nurs. 2022; (In press)https://doi.org/10.1097/jcn.0000000000000919
        • Bodenheimer T.
        • Lorig K.
        • Holman H.
        • Grumbach K.
        Patient self-management of chronic disease in primary care.
        JAMA. 2002; 288: 2469-2475https://doi.org/10.1001/jama.288.19.2469
        • Chalfont G.
        • Mateus C.
        • Varey S.
        • Milligan C.
        Self-efficacy of older people using technology to self-manage COPD, hypertension, heart failure, or dementia at home: an overview of systematic reviews.
        Gerontol. 2021; 61: e318-e334https://doi.org/10.1093/geront/gnaa045
        • Schwarzer R.
        • Renner B.
        Social-cognitive predictors of health behavior: action self-efficacy and coping self-efficacy.
        Health Psychol. 2000; 19: 487-495
        • Bandura A.
        Self-efficacy: the exercise of control.
        Worth Publishers, New York, NY1997
      1. Northwestern University; HealthMeasures. List of adult measures. PROMIS® [Internet]. [updated 2021 April 28; cited 2022 May 4]. Available from: https://www.healthmeasures.net/explore-measurement-systems/promis/intro-to-promis/list-of-adult-measures.

        • Rossiter C.
        • Levett-Jones T.
        • Pich J.
        The impact of person-centred care on patient safety: an umbrella review of systematic reviews.
        Int J Nurs Stud. 2020; 103658: 109https://doi.org/10.1016/j.ijnurstu.2020.103658
        • Yun D.
        • Choi J.
        Person-centered rehabilitation care and outcomes: a systematic literature review.
        Int J Nurs Stud. 2019; 93: 74-83https://doi.org/10.1016/j.ijnurstu.2019.02.012
        • Zhang P.
        • Ouyang Z.
        • Fang S.
        • He J.
        • Fan L.
        • Luo X.
        • et al.
        Personality inventory for DSM-5 brief form(PID-5-BF) in Chinese students and patients: evaluating the five-factor model and a culturally informed six-factor model.
        BMC Psychiatr. 2021; 107: 21https://doi.org/10.1186/s12888-021-03080-x
        • Kucharski D.
        • Lange E.
        • Ross A.B.
        • Svedlund S.
        • Feldthusen C.
        • Önnheim K.
        • et al.
        Moderate-to-high intensity exercise with person-centered guidance influences fatigue in older adults with rheumatoid arthritis.
        Rheumatol Int. 2019; 39: 1585-1594https://doi.org/10.1007/s00296-019-04384-8
        • Siebolds M.
        • Gaedeke O.
        • Schwedes U.
        Self-monitoring of blood glucose–psychological aspects relevant to changes in HbA1c in type 2 diabetic patients treated with diet or diet plus oral antidiabetic medication.
        Patient Educ Counsel. 2006; 62: 104-110https://doi.org/10.1016/j.pec.2005.06.013
        • Kautz D.D.
        • Kuiper R.
        • Pesut D.J.
        • Knight-Brown P.
        • Daneker D.
        Promoting clinical reasoning in undergraduate nursing students: application and evaluation of the Outcome Present State Test (OPT) model of clinical reasoning.
        Int J Nurs Educ Scholarsh. 2005; 2 (Article 1)https://doi.org/10.2202/1548-923x.1052
        • Martin-Delgado J.
        • Guilabert M.
        • Mira-Solves J.
        Patient-reported experience and outcome measures in people living with diabetes: a scoping review of instruments.
        Patient. 2021; 14: 759-773https://doi.org/10.1007/s40271-021-00526-y
        • Ghizzardi G.
        • Arrigoni C.
        • Dellafiore F.
        • Vellone E.
        • Caruso R.
        Efficacy of motivational interviewing on enhancing self-care behaviors among patients with chronic heart failure: a systematic review and meta-analysis of randomized controlled trials.
        Heart Fail Rev. 2021; 27: 1029-1041https://doi.org/10.1007/s10741-021-10110-z
        • Kalluri M.
        • Younus S.
        • Archibald N.
        • Richman-Eisenstat J.
        • Pooler C.
        Action plans in idiopathic pulmonary fibrosis: a qualitative study-’I do what I can do’.
        BMJ Support Palliat Care. 2021; (In press)https://doi.org/10.1136/bmjspcare-2020-002831
        • Hansen C.W.
        • Esbensen B.A.
        • de Thurah A.
        • Christensen R.
        • de Wit M.
        • Cromhout P.F.
        Outcome measures in rheumatology applied in self-management interventions targeting people with inflammatory arthritis A systematic review of outcome domains and measurement instruments.
        Semin Arthritis Rheum. 2022; 151995: 54https://doi.org/10.1016/j.semarthrit.2022.151995
        • Hosseini A.
        • Jackson A.C.
        • Chegini N.
        • Dehghan M.F.
        • Mazloum D.
        • Haghani S.
        • et al.
        The effect of an educational app on hemodialysis patients' self-efficacy and self-care: a quasi-experimental longitudinal study.
        Chronic Illn. 2022; (17423953211073365)https://doi.org/10.1177/17423953211073365
        • Lorig K.
        • Chastain R.L.
        • Ung E.
        • Shoor S.
        • Holman H.R.
        Development and evaluation of a scale to measure perceived self-efficacy in people with arthritis.
        Arthritis Rheum. 1989; 32: 37-44https://doi.org/10.1002/anr.1780320107
        • Nicholas M.K.
        The pain self-efficacy questionnaire: taking pain into account.
        Eur J Pain. 2007; 11: 153-163https://doi.org/10.1016/j.ejpain.2005.12.008
        • Lev E.L.
        • Owen S.V.
        A measure of self-care self-efficacy.
        Res Nurs Health. 1996; 19 (19:5<421::Aid-nur6>3.0.Co;2-s): 421-429https://doi.org/10.1002/(sici)1098-240x(199610
        • Frei A.
        • Svarin A.
        • Steurer-Stey C.
        • Puhan M.A.
        Self-efficacy instruments for patients with chronic diseases suffer from methodological limitations–a systematic review.
        Health Qual Life Outcomes. 2009; 86: 7https://doi.org/10.1186/1477-7525-7-86
        • Sherer M.
        • Maddux J.E.
        The self-efficacy scale: construction and validation.
        Psychol Rep. 1982; 51: 663-671
        • Rothrock N.E.
        • Amtmann D.
        • Cook K.F.
        Development and validation of an interpretive guide for PROMIS scores.
        J Patient Rep Outcomes. 2020; 16: 4https://doi.org/10.1186/s41687-020-0181-7
        • Reeve B.B.
        • Hays R.D.
        • Bjorner J.B.
        • Cook K.F.
        • Crane P.K.
        • Teresi J.A.
        • et al.
        Psychometric evaluation and calibration of health-related quality of life item banks: plans for the Patient-Reported Outcomes Measurement Information System (PROMIS).
        Med Care. 2007; 45: 472-476https://doi.org/10.1097/01.mlr.0000250483.85507.04
        • Lee E.H.
        • Kang E.H.
        • Kang H.J.
        Evaluation of studies on the measurement properties of self-reported instruments.
        Asian Nurs Res (Korean Soc Nurs Sci). 2020; 14: 267-276https://doi.org/10.1016/j.anr.2020.11.004
        • Song Y.
        • Kim H.
        • Park S.Y.
        An item response theory analysis of the Korean version of the CRAFFT scale for alcohol use among adolescents in Korea.
        Asian Nurs Res (Korean Soc Nurs Sci). 2019; 13: 249-256https://doi.org/10.1016/j.anr.2019.09.003
        • Nguyen T.H.
        • Han H.R.
        • Kim M.T.
        • Chan K.S.
        An introduction to item response theory for patient-reported outcome measurement.
        Patient. 2014; 7: 23-35https://doi.org/10.1007/s40271-013-0041-0
        • Rose M.
        • Bjorner J.B.
        • Gandek B.
        • Bruce B.
        • Fries J.F.
        • Ware Jr., J.E.
        The PROMIS Physical Function item bank was calibrated to a standardized metric and shown to improve measurement efficiency.
        J Clin Epidemiol. 2014; 67: 516-526https://doi.org/10.1016/j.jclinepi.2013.10.024
        • Hanmer J.
        • Jensen R.E.
        • Rothrock N.
        A reporting checklist for HealthMeasures' patient-reported outcomes: ASCQ-Me, Neuro-QoL, NIH Toolbox, and PROMIS.
        J Patient Rep Outcomes. 2020; 21: 4https://doi.org/10.1186/s41687-020-0176-4
        • von Elm E.
        • Altman D.G.
        • Egger M.
        • Pocock S.J.
        • Gøtzsche P.C.
        • Vandenbroucke J.P.
        The strengthening the reporting of observational studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies.
        Lancet. 2007; 370: 1453-1457https://doi.org/10.1016/s0140-6736(07)61602-x
        • Bonomi A.E.
        • Cella D.F.
        • Hahn E.A.
        • Bjordal K.
        • Sperner-Unterweger B.
        • Gangeri L.
        • et al.
        Multilingual translation of the Functional Assessment of Cancer Therapy (FACT) quality of life measurement system.
        Qual Life Res. 1996; 5: 309-320https://doi.org/10.1007/bf00433915
        • Eremenco S.L.
        • Cella D.
        • Arnold B.J.
        A comprehensive method for the translation and cross-cultural validation of health status questionnaires.
        Eval Health Prof. 2005; 28: 212-232https://doi.org/10.1177/0163278705275342
        • World Health Organization
        WHO website Health topics: Diatetes [Internet].
        2021 April 13 ([cited 2022 April 17]. Available from:)
        • Yazdani F.
        • Abazari P.
        • Haghani F.
        • Iraj B.
        Restrictors of the effectiveness of diabetes self-management education: a qualitative content analysis.
        J Educ Health Promot. 2021; 18: 10https://doi.org/10.4103/jehp.jehp_914_20
        • Polit D.F.
        • Yang F.M.
        Measurement and the measurement of change.
        Wolters Kluwer, Philadelphia, PA2016
        • Forero C.G.
        • Maydeu-Olivares A.
        Estimation of IRT graded response models: limited versus full information methods.
        Psychol Methods. 2009; 14: 275-299https://doi.org/10.1037/a0015825
        • Peyrot M.
        • Peeples M.
        • Tomky D.
        • Charron-Prochownik D.
        • Weaver T.
        Development of the American Association of Diabetes Educators' diabetes self-management assessment report tool.
        Diabetes Educ. 2007; 33: 818-826https://doi.org/10.1177/0145721707307614
        • Lee H.
        • Choi E.K.
        • Kim H.
        • Kim H.
        • Kim H.
        Factors affecting the self-management of adolescents with type 1 diabetes mellitus based on the information-motivation-behavioral skills model.
        Child Health Nurs Res. 2019; 25: 234-243https://doi.org/10.4094/chnr.2019.25.2.234
        • Lee S.
        • Kim H.
        Structural equation modeling on self-care behavior and quality of life in older adults with diabetes using citizen health promotion centers.
        J Korean Acad Nurs. 2017; 47: 514-525https://doi.org/10.4040/jkan.2017.47.4.514
        • Toobert D.J.
        • Hampson S.E.
        • Glasgow R.E.
        The summary of diabetes self-care activities measure: results from 7 studies and a revised scale.
        Diabetes Care. 2000; 23: 943-950https://doi.org/10.2337/diacare.23.7.943
        • Chang S.
        • Song M.-S.
        The validity and reliability of a Korean version of the summary of diabetes self-care activities questionnaire for older patients with type 2 diabetes.
        Korean J Adult Nurs. 2009; 21: 235-244
        • Chang S.
        Structural equation moedling on health-related quality of life in older adutls with type 2 diabetes mellitus [dissertation].
        Seoul National University, Seoul, Korea2010
        • Mokkink L.B.
        • Prinsen C.
        • Patrick D.L.
        • Alonso J.
        • Bouter L.M.
        • De Vet H.
        • et al.
        COSMIN methodology for systematic reviews of patient-reported outcome measures (PROMs).
        User Manual. 2018; : 1-78
        • Choi H.
        • Kim C.
        • Ko H.
        • Park C.G.
        Translation and validation of the Korean version of PROMIS® pediatric and parent proxy measures for emotional distress.
        J Patient Rep Outcomes. 2019; 36: 3https://doi.org/10.1186/s41687-019-0120-7
        • Reise S.P.
        • Scheines R.
        • Widaman K.F.
        • Haviland M.G.
        Multidimensionality and structural coefficient bias in structural equation modeling: a bifactor perspective.
        Educ Psychol Meas. 2013; 73: 5-26https://doi.org/10.1177/001316449831
        • Yen W.M.
        Effects of local item dependence on the fit and equating performance of the three-parameter logistic model.
        Appl Psychol Meas. 1984; 8: 125-145
        • Christensen K.B.
        • Makransky G.
        • Horton M.
        Critical values for Yen's Q3: identification of local dependence in the Rasch model using residual correlations.
        Appl Psychol Meas. 2017; 41: 178-194https://doi.org/10.1177/0146621616677520
        • González-de Paz L.
        • Kostov B.
        • López-Pina J.A.
        • Solans-Julián P.
        • Navarro-Rubio M.D.
        • Sisó-Almirall A.
        A Rasch analysis of patients' opinions of primary health care professionals' ethical behaviour with respect to communication issues.
        Fam Pract. 2015; 32: 237-243https://doi.org/10.1093/fampra/cmu073
        • Pinto M.
        • Pinto R.M.C.
        • Mendonça T.
        • Souza C.G.
        • da Silva C.H.M.
        Validation and calibration of the patient-reported outcomes measurement information system: pediatric PROMIS® Emotional Distress domain item banks, Portuguese version (Brazil/Portugal).
        Qual Life Res. 2020; 29: 1987-1997https://doi.org/10.1007/s11136-020-02447-z
        • Li C.-H.
        Confirmatory factor analysis with ordinal data: comparing robust maximum likelihood and diagonally weighted least squares.
        Behav Res Methods. 2016; 48: 936-949https://doi.org/10.3758/s13428-015-0619-7
        • Rosseel Y.
        Lavaan: an R package for structural equation modeling and more.
        J Stat Softw. 2012; 48: 1-36
        Version: Version 0.5–12 (BETA)
        • Andrich D.
        An extension of the Rasch model for ratings providing both location and dispersion parameters.
        Psychometrika. 1982; 47: 105-113https://doi.org/10.1007/BF02293856
      2. Korean Diabetes Association Diabetes fact sheet in Korea 2020. 2020
        • Statistics Korea
        2020 Korean social indicators.
        2021
        • Güler N.
        • Penfield R.D.
        A comparison of the logistic regression and contingency table methods for simultaneous detection of uniform and nonuniform DIF.
        J Educ Meas. 2009; 46: 314-329
        • Choi S.W.
        • Gibbons L.E.
        • Crane P.K.
        lordif: an R Package for detecting Differential Item Functioning using iterative hybrid ordinal logistic regression/item response theory and Monte Carlo simulations.
        J Stat Softw. 2011; 39: 1-30https://doi.org/10.18637/jss.v039.i08
        • Rimehaug S.A.
        • Kaat A.J.
        • Nordvik J.E.
        • Klokkerud M.
        • Robinson H.S.
        Psychometric properties of the PROMIS-57 questionnaire, Norwegian version.
        Qual Life Res. 2022; 31: 269-280https://doi.org/10.1007/s11136-021-02906-1
        • Meir E.I.
        • Gati I.
        Guidelines for item selection in inventories yielding score profiles.
        Educ Psychol Meas. 1981; 41: 1011-1016
        • Bond T.G.
        • Fox C.M.
        Applying the Rasch model: fundamental measurement in the human sciences.
        Mahwah NJ: Lawrence Erlbaum, 2001
        • Sunderland M.
        • Batterham P.
        • Calear A.
        • Carragher N.
        Validity of the PROMIS depression and anxiety common metrics in an online sample of Australian adults.
        Qual Life Res. 2018; 27: 2453-2458https://doi.org/10.1007/s11136-018-1905-5
        • Bjorner J.B.
        • Rose M.
        • Gandek B.
        • Stone A.A.
        • Junghaenel D.U.
        • Ware Jr., J.E.
        Difference in method of administration did not significantly impact item response: an IRT-based analysis from the Patient-Reported Outcomes Measurement Information System (PROMIS) initiative.
        Qual Life Res. 2014; 23: 217-227https://doi.org/10.1007/s11136-013-0451-4
        • Lee E.-H.
        Psychometric properties of an instrument 2: structural validity, internal consistency, and cross-cultural validity/measurement invariance.
        Korean J Women Health Nurs. 2021; 27: 69-74
        • Hair J.F.
        • Anderson R.E.
        • Tatham R.L.
        • Black W.C.
        Multivariate data analysis with readings.
        4 ed. 1995
        • Hooper D.
        • Coughlan J.
        • Mullen M.R.
        Structural equation modelling: guidelines for determining model fit.
        Electron J Bus Res Methods. 2008; 6: 53-60https://doi.org/10.21427/D7CF7R
        • Chan S.W.W.
        • Chien C.W.
        • Wong A.Y.L.
        • Pang M.Y.C.
        Translation and psychometric validation of the traditional Chinese version of patient-reported outcomes measurement information system Pediatric-25 Profile version 2.0 (PROMIS-25) in Chinese Children with Cancer in Hong Kong.
        Qual Life Res. 2021; 30 (in press): 1779-1791https://doi.org/10.1007/s11136-021-02759-8
        • Chen W.H.
        • Lenderking W.
        • Jin Y.
        • Wyrwich K.W.
        • Gelhorn H.
        • Revicki D.A.
        Is Rasch model analysis applicable in small sample size pilot studies for assessing item characteristics? An example using PROMIS pain behavior item bank data.
        Qualive of Life Research. 2014; 23: 485-493https://doi.org/10.1007/s11136-013-0487-5
        • Roy C.
        • Bakan G.
        • Li Z.
        • Nguyen T.H.
        Coping measurement: creating short form of Coping and Adaptation Processing Scale using item response theory and patients dealing with chronic and acute health conditions.
        Appl Nurs Res. 2016; 32: 73-79https://doi.org/10.1016/j.apnr.2016.06.002
        • Cleanthous S.
        • Barbic S.P.
        • Smith S.
        • Regnault A.
        Psychometric performance of the PROMIS® depression item bank: a comparison of the 28- and 51-item versions using Rasch measurement theory.
        J Patient Rep Outcomes. 2019; 47: 3https://doi.org/10.1186/s41687-019-0131-4
        • Lee J.
        • Lee E.H.
        • Chae D.
        • Kim C.J.
        Patient-reported outcome measures for diabetes self-care: a systematic review of measurement properties.
        Int J Nurs Stud. 2020; 103498: 105https://doi.org/10.1016/j.ijnurstu.2019.103498
        • Gruber-Baldini A.L.
        • Velozo C.
        • Romero S.
        • Shulman L.M.
        Validation of the PROMIS® measures of self-efficacy for managing chronic conditions.
        Qual Life Res. 2017; 26: 1915-1924https://doi.org/10.1007/s11136-017-1527-3
        • Lee E.-H.
        • Lee Y.W.
        • Lee K.-W.
        • Hong S.
        • Kim S.H.
        A New Objective Health Numeracy Test for patients with type 2 diabetes: development and evaluation of psychometric properties.
        Asian Nurs Res. 2020; 14: 66-72https://doi.org/10.1016/j.anr.2020.01.006
        • Choi S.
        • Kim S.H.
        Influences of patient activation on diabetes self-care activities and diabetes-specific distress.
        Korean J Adult Nurs. 2020; 32: 10-20https://doi.org/10.7475/kjan.2020.32.1.10
        • Depaoli S.
        • Tiemensma J.
        • Felt J.M.
        Assessment of health surveys: fitting a multidimensional graded response model.
        Psychol Health Med. 2018; 23: 13-31https://doi.org/10.1080/13548506.2018.1447136
        • Hong I.
        • Velozo C.A.
        • Li C.Y.
        • Romero S.
        • Gruber-Baldini A.L.
        • Shulman L.M.
        Assessment of the psychometrics of a PROMIS item bank: self-efficacy for managing daily activities.
        Qual Life Res. 2016; 25: 2221-2232https://doi.org/10.1007/s11136-016-1270-1
        • Wilson M.
        Constructing measures: an item response modeling approach.
        Routledge, London, UK2004
        • Muscat D.M.
        • Cvejic E.
        • Bell K.
        • Smith J.
        • Morris G.M.
        • Jansen J.
        • et al.
        The impact of health literacy on psychosocial and behavioural outcomes among people at low risk of cardiovascular disease.
        Prev Med. 2022; 156: 106980https://doi.org/10.1016/j.ypmed.2022.106980
        • Scholz S.
        • Teetz L.
        Smart health via mHealth? Potentials of mobile health apps for improving prevention and adherence of breast cancer patients.
        Digit Health. 2022; 20552076221074127: 8https://doi.org/10.1177/20552076221074127
        • Yu E.
        • Hagens S.
        Socioeconomic disparities in the demand for and use of virtual visits among senior adults during the COVID-19 pandemic: a cross-sectional study.
        JMIR Aging. 2022; 5: e35221https://doi.org/10.2196/35221
        • Jung S.O.
        • Son Y.H.
        • Choi E.
        E-health literacy in older adults: an evolutionary concept analysis.
        BMC Med Inform Decis Mak. 2022; 28: 22https://doi.org/10.1186/s12911-022-01761-5