Revisiting the Concept of Minimal Detectable Change for Patient-Reported Outcome Measures

Translation, Cross-Cultural Adaptation, and Validation of the Persian Version of the Harris Hip Score

Background

The Harris hip score (HHS), a self-administered questionnaire, is widely used to evaluate hip pathology affecting health-related quality of life and physical function. This study's purpose was HHS translation to Persian (HHS-Pr) and validation in patients with different hip pathologies.

Methods

Translation and cultural adaptation followed existing guidelines. Hip pathology patients (n = 151) completed the HHS, 12-Item Health Survey, and the Western Ontario and McMaster Universities Arthritis Index (WOMAC). Criterion validity was determined from comparisons between the HHS measures and the different corresponding WOMAC domains. Internal consistency used Cronbach's alpha (α), content validity the "content validity index," and floor/ceiling effect the end-range 15%. Test-retest reliability used the intraclass correlation coefficient (subsample n = 30) at 3-7 days that compared baseline with a repeated measure. Measurement precision and change sensitivity used longitudinal assessment (subgroup n = 30) from the standard error of the measurement and minimal detectable change.

Results

Cross-cultural adaptation required minor wording changes. The mean HHS-Pr was 57.77 ± 19.69. Criterion validity was significant with the WOMAC (r = -0.76) and 12-Item Health Survey Physical Component Summary (r = 0.47). Internal consistency was high before (α = 0.75) and after standardization (α = 0.86). Content validity was satisfactory (content validity index = 0.88). No floor/ceiling effects were found. Test-retest reliability (intraclass correlation coefficient = 0.85) was excellent, as was standard error of the measurement (raw score = 5.8) and minimal detectable change (raw score = 11.4).

Conclusions

The HHS-Pr demonstrated adequate validity, reliability, and sensitivity to change. These psychometric properties sufficiently measure functional status in patients with hip pathologies in a Persian-speaking population.

Cervical Multifidus Stiffness Assessment in Individuals with and without Unilateral Chronic Neck Pain: An Inter-Examiner Reliability Study

This study aimed to evaluate the inter-examiner reliability of shear wave elastography (SWE) for measuring cervical multifidus (CM) muscle stiffness in asymptomatic controls and patients with chronic neck pain. A longitudinal observational study was conducted to assess the diagnostic accuracy of a procedure. SWE images, following a detailed procedure previously tested, were acquired by two examiners (one novice and one experienced) to calculate the shear wave speed (SWS) and Young's modulus. The painful side was examined for the experimental cases while the side examined in the control group was selected randomly. Data analyses calculated the intra-class correlation coefficients (ICCs), absolute errors between examiners, standard errors of measurement, and minimal detectable changes. A total of 125 participants were analyzed (n = 54 controls and n = 71 cases). The Young's modulus and SWS measurements obtained by both examiners were comparable within the asymptomatic group (both, p > 0.05) and the chronic neck pain group (both, p > 0.05). Nonetheless, a notable distinction was observed in the absolute error between examiners for shear wave speed measurements among patients with neck pain, where a significant difference was registered (p = 0.045), pointing to a sensitivity in measurement consistency affected by the presence of chronic neck pain. ICCs demonstrated moderate-to-good reliability across both groups, with ICC values for asymptomatic individuals reported as >0.8. Among the chronic neck pain patients, ICC values were slightly lower (>0.780). The study revealed moderate-to-good consistency, highlighting the practicality and generalizability of SWE.

Is the difference real, is the difference relevant: the minimal detectable and clinically important changes in the Montreal Cognitive Assessment

Background and aims

The Montreal Cognitive Assessment (MoCA) is a widely used instrument for assessing cognitive function in stroke survivors. To interpret changes in MoCA scores accurately, it is crucial to consider the minimal detectable change (MDC) and minimal clinically important difference (MCID). The aim was to establish the MDC and MCID of the MoCA within 6 months after stroke.

Methods

This cohort study analysed data from the EFFECTS trial. The MoCA was administered at baseline and at 6-month follow-up. The MDC was calculated as the upper limit of the 95 % confidence interval of the standard error of the MoCA mean. The MCID was determined using anchor-based and distribution methods. The visual analogue recovery scale of the Stroke Impact Scale (SIS [primary anchor]) and Euro Quality of Life-5 Dimensions index (EQ-5D [confirmatory anchor]) were used as anchors. The distribution-based method, the Cohen benchmark effect size was chosen.

Results

In total, 1131 (mean age [SD], 71 [10.6] years) participants were included. The mean (SD) MoCA scores at admission and 6-month follow-up were 22 (5.2) and 25 (4.2), respectively. The MDC of the MoCA was 5.1 points. The anchor method yielded the MCIDs 2 and 1.6 points for SIS and EQ-5D, respectively. Using the distribution method, the MCID for the MoCA was 1 point.

Conclusions

Even a small change in MoCA scores can be important for stroke survivors; however, larger differences are required to ensure that any difference in MoCA values is a true change and is not related to the inherent variation in the test. Due to small sample sizes, the results of the anchor analysis need to be interpreted with caution.

Construct validity, responsiveness, minimal detectable change, and minimal clinically important difference of the stroke self-efficacy questionnaire in individuals receiving stroke rehabilitation

PURPOSE

To assess the clinimetric properties of the Stroke Self-Efficacy Questionnaire (SSEQ) and estimate the minimal detectable change (MDC) and minimal clinically important difference (MCID) from the database of our randomized controlled trials (RCT) of individuals receiving stroke rehabilitation.

METHODS

We retrieved the pre- and post-intervention scores of the SSEQ and Stroke Impact Scale (SIS) from 80 stroke survivors. The analysis of clinimetric properties was performed using: (1) confirmatory factor analysis and item response theory modeling (IRT) for construct validity; (2) standardized response mean and Glass's delta for responsiveness; (3) MDC based on the standard deviation (SD) or standard error of measurement (SEM) of the SSEQ change scores; (4) MCID determined by the external anchor-SIS; (5) conditional MDC (cMDC) derived from the IRT analysis.

RESULTS

There was a bi-factorial construct with excellent model-data fit and marked responsiveness. The MDC determined by the SD and SEM were 1.5 and 3.0, respectively, and the MCIDs were 3.3 and 3.7.

CONCLUSIONS

This study confirmed that SSEQ is a valid and reliable assessment tool for patients receiving stroke rehabilitation. We also provided practical threshold values, especially demonstrating the benefit of using individualized cMDC, to help clinicians better interpret the change in the SSEQ scores.

Cross-Cultural Adaptation and Validation of the Arabic Version of the Prolapse Quality of Life Questionnaire in the United Arab Emirates

BACKGROUND

Given the extensive translation of the Prolapse Quality of Life Questionnaire (P-QoL) into many languages, it is imperative to develop an Arabic version to facilitate the study of pelvic organ health within the Arabian culture.

OBJECTIVE

The aim of this study was to investigate, cross-culturally adapt, and validate the Arabic version of the P-QoL.

STUDY DESIGN

This study involved cross-cultural adaptation and psychometric testing.

METHODS

A total of 90 participants were included in the study. This cross-sectional study was carried out in two phases; during phase I, the P-QoL was translated and adapted from English into Arabic. The Arabic version was psychometrically validated during phase II using the test-retest reliability and internal consistency with Cronbach's alpha coefficient, convergent construct (CC) validity among the four study tools using Spearman's coefficient (r), and discriminative validity using Mann-Whitney test to find the differences between the means of the two samples.

RESULTS

A satisfactory level of semantic, conceptual, idiomatic, and content comparability was reached in the cross-cultural adaptation of the Arabic version of the P-QoL. The internal consistency was high in terms of psychometric validation, with a Cronbach's alpha coefficient of 0.971 for the P-QoL. The test-retest results showed high reliability, with the interclass correlation coefficient (ICC) of the P-QoL determined as 0.987. The convergent construct validity was highly acceptable (moderately strong), reflecting a positive correlation between the Arabic version of the P-QoL and the Australian Pelvic Floor Dysfunction Questionnaire (APFD) (r = 0.68; p < 0.001). Similarly, a significant convergent validity of the Arabic version of the P-QoL and the Visual Analogue Scale (VAS) (r = 0.47; p < 0.001) was observed, as well as a correlation between the APFD and the VAS (r = 0.46; p < 0.001). However, there was no significant correlation between the 12-Item Short-Form Survey (SF-12), the P-QoL, the APFD, and the VAS.

CONCLUSION

Based on the significant correlation found between the Arabic APFD and the VAS, the results reveal good reliability, internal consistency, and construct validity. It is recommended that Arabic-speaking females with pelvic organ prolapse use the Arabic version of the P-QoL. More research is needed to assess the responsiveness of the P-QoL.

Responsiveness and meaningful thresholds of PROMIS pain interference, fatigue, and physical function forms in adults with idiopathic inflammatory myopathies: Report from the OMERACT Myositis Working Group

BACKGROUND

A series of qualitative studies conducted by the OMERACT Myositis Working Group identified pain interference, fatigue, and physical function as highly important life impact domains for adults with idiopathic inflammatory myositis (IIM). In this study, our goal was to assess the responsiveness and minimal important difference of PROMIS pain interference (6a), fatigue (7a), and physical function (8b).

METHODS

Adults with IIM from USA, Netherlands, Korea, Sweden, and Australia with two "clinical" visits were enrolled in this prospective study. Anchor questions on a Likert scale were collected at baseline, and manual muscle testing (MMT), physician and patient reported global disease activity, and PROMIS instruments were collected at both visits. Responsiveness was assessed with i) ANOVA, ii) paired t-test, effect size and standardized response mean, and iii) Pearson correlation. Minimal important difference (MID), minimal important change (MIC) and minimal detectable change (MDC) values were calculated.

RESULTS

114 patients with IIM (median age 60, 60 % female) completed both visits. Changes in PROMIS instruments were significantly different among anchor categories. Patients who reported improvement had a significant improvement in their PROMIS scores with at least medium effect size, while patients who reported worsening and stability did not show a significant change with weak effect size. PROMIS instruments had weak to moderate correlations with MMT, patient and physician global disease activity. MID was approximately 2-3 points for Pain Interference and 3-4 points for Fatigue and Physical Function forms based on the method used. MIC was approximately 4-5 for improvement of all the instruments, while MDC was 1.7-2 points for Pain Interference and Physical Function and 3.2-3.9 for Fatigue.

CONCLUSION

This study provides evidence towards the responsiveness of the PROMIS instruments in a large international prospective cohort of adults with IIM supporting their use as PROMs in adult myositis.

Reliability of Spectral Features of Resting-State Brain Activity: A Magnetoencephalography Study

Background Cognition is a vital sign and its deterioration is a major concern in clinical medicine. It is usually evaluated using neuropsychological assessments, which have innate limitations such as the practice effect. To compensate for these assessments, the oscillatory power of resting-state brain activity has recently become available. The power is obtained noninvasively using magnetoencephalography and is summarized by spectral parameters such as the median frequency (MF), individual alpha frequency (IAF), spectral edge frequency 95 (SEF95), and Shannon's spectral entropy (SSE). As these parameters are less sensitive to practice effects, they are suitable for longitudinal studies. However, their reliability remains unestablished, hindering their proactive use in clinical practice. Therefore, we aimed to quantify the within-participant reliability of these parameters using repeated measurements of healthy participants to facilitate their clinical use and to evaluate the observed changes/differences in these parameters reported in previous studies. Methodology Resting-state brain activity with eyes closed was recorded using magnetoencephalography for five minutes from 15 healthy individuals (29.3 ± 4.6 years old: ranging from 23 to 28 years old). The following four spectral parameters were calculated: MF, IAF, SEF95, and SSE. To quantify reliability, the minimal detectable change (MDC) and intraclass correlation coefficient (ICC) were computed for each parameter. In addition, we used MDCs to evaluate the changes and differences in the spectral parameters reported in previous longitudinal and cross-sectional studies. Results The MDC at 95% confidence interval (MDC95) of MF, IAF, SEF95, and SSE were 0.61 Hz, 0.44 Hz, 2.91 Hz, and 0.028, respectively. The ICCs of these parameters were 0.96, 0.92, 0.94, and 0.83, respectively. The MDC95 of these parameters was smaller than the mean difference in the parameters between cognitively healthy individuals and patients with dementia, as reported in previous studies. Conclusions The spectral parameter changes/differences observed in prior studies were not attributed to measurement errors but rather reflected genuine effects. Furthermore, all spectral parameters exhibited high ICCs (>0.8), underscoring their robust within-participant reliability. Our results support the clinical use of these parameters, especially in the longitudinal monitoring and evaluation of the outcomes of interventions.

Increasing Activity After Stroke: A Randomized Controlled Trial of High-Intensity Walking and Step Activity Intervention

BACKGROUND

Physical inactivity in people with chronic stroke profoundly affects daily function and increases recurrent stroke risk and mortality, making physical activity improvements an important target of intervention. We compared the effects of a high-intensity walking intervention (FAST), a step activity monitoring behavioral intervention (SAM), or a combined intervention (FAST+SAM) on physical activity (ie, steps/day). We hypothesized the combined intervention would yield the greatest increase in steps/day.

METHODS

This assessor-blinded multisite randomized controlled trial was conducted at 4 university/hospital-based laboratories. Participants were 21 to 85 years old, walking without physical assistance following a single, unilateral noncerebellar stroke of ≥6 months duration, and randomly assigned to FAST, SAM, or FAST+SAM for 12 weeks (2-3 sessions/week). FAST training consisted of walking-related activities at 70% to 80% heart rate reserve, while SAM received daily feedback and goal setting of walking activity (steps/day). Assessors and study statistician were masked to group assignment. The a priori-determined primary outcome and end point was a comparison of the change in steps/day between the 3 intervention groups from pre- to post-intervention. Adverse events were tracked after randomization. All randomized participants were included in the intent-to-treat analysis.

RESULTS

Participants were enrolled from July 18, 2016, to November 16, 2021. Of 2385 participants initially screened, 250 participants were randomized (mean [SE] age, 63 [0.80] years; 116 females/134 males), with 89 assigned to FAST, 81 to SAM, and 80 to FAST+SAM. Steps/day significantly increased in both the SAM (mean [SE], 1542 [267; 95% CI, 1014-2069] P<0.001) and FAST+SAM group (1307 [280; 95% CI, 752-1861] P<0.001) but not in the FAST group (406 [238; 95% CI, -63 to 876] P=0.09). There were no deaths or serious study-related adverse events.

CONCLUSIONS

Only individuals with chronic stroke who completed a step activity monitoring behavioral intervention with skilled coaching and goal progression demonstrated improvements in physical activity (steps/day).

REGISTRATION

URL: https://www.clinicaltrials.gov; Unique identifier: NCT02835313.

Conditional Minimal Detectable Change for the Cochlear Implant Quality of Life-35 Profile Associated With Improved Functional Abilities 12 Months After Cochlear Implantation

Importance

It is essential to measure an individual patient's baseline and follow-up abilities to demonstrate changes in clinical outcomes over time. Inherent in this strategy is interpreting whether the measured change is clinically significant and beyond measurement error. Conditional minimal detectable change (cMDC) values are widely used in many disciplines but have rarely been established for outcome measures in otolaryngology or hearing research, and never in cochlear implantation.

Objective

To determine cMDC values for the Cochlear Implant Quality of Life-35 (CIQOL-35) Profile instrument to enhance our understanding of the initial and ongoing changes in functional abilities from cochlear implants (CIs).

Design, Setting, and Participants

Item response theory analyses of responses from a multi-institutional cohort of 705 CI users at a tertiary CI center were used to derive standard error (SE) values for each possible CIQOL-35 domain score. Using an iterative approach, these SE values were used to calculate cMDC values for every possible pre-CI and post-CI domain score combination. We then compared pre-CI to 12-month post-CI CIQOL-35 domains scores in an independent cohort of 65 adult CI users to determine whether the measured change exceeded error to be clinically significant. The analysis took place on December 14, 2022.

Interventions

The CIQOL-35 Profile instrument and cochlear implantation.

Results

The cMDC values were smaller for the communication domain, and global measure and cMDC values for all domains were larger at the extremes of the measurement scale. Overall, 60 CI users (92.3%) demonstrated improvement beyond cMDC at 12 months post-CI for at least 1 CIQOL-35 domain, and no patients' scores declined beyond cMDC for any domain. The percentage of CI users demonstrating improvement beyond cMDC varied by domain, with communication (53 [81.5%]) showing the largest number of CI users improving, followed by global (42 [64.6%]) and entertainment (40 [60.9%]). In general, CI users who demonstrated improvement in CIQOL-35 domains had greater improvement in speech recognition scores than patients who did not, but the strength and significance of these associations greatly varied by domain and speech material.

Conclusions and Relevance

This multistep cohort study found that cMDC values for the CIQOL-35 Profile provided personalized thresholds for detecting real changes in patient self-reported functional abilities over time across multiple domains, which may inform clinical decision-making. Moreover, these longitudinal results reveal the domains with more or less improvement, which may aid in patient counseling.

Test-Retest Reliability and Random Measurement Error of the Multifactorial Memory Questionnaire in Older Adults With Subjective Memory Complaints

To examine the psychometric properties of the Multifactorial Memory Questionnaire (MMQ) in older adults with subjective memory complaints. The three MMQ subscale (Satisfaction, Ability, and Strategy) was administered twice, with a 3-month interval. The test-retest reliability was examined using intraclass correlation coefficients (ICCs). The random measurement error was examined by calculating the standard error of measurement (SEM) and minimal detectable change (MDC95). The test-retest reliabilities of the three MMQ subscales were generally acceptable. The SEM of the three MMQ subscales was higher than the acceptable criterion of 10%. Despite the influence of random measurement error, the change scores of the three MMQ subscales may represent true changes if they are larger than the MDC95 of 13.2 (Satisfaction), 18.4 (Ability), and 16.9 (Strategy). The MMQ appears to be a reliable measure for use in research settings, but may not yet be suitable for clinical use.