Estimation of minimal clinically important change of the Japanese version of EQ-5D in patients with chronic noncancer pain: a retrospective research using real-world data

Abstract

BackgroundQuality of life (QoL) is routinely assessed and evaluated in medical research. However, in Japan, there is a lack of solid cutoff criteria for evaluating QoL improvement in chronic noncancer pain management. The present study was conducted to identify the minimal clinically important change (MCIC) of the Japanese version of EuroQol-5D 3L(EQ-5D) utility score and numeric rating scale (NRS) with an emphasis on chronic noncancer pain.MethodsThe data source for this post hoc research was the post-marketing surveillance (PMS) data for a tramadol/acetaminophen combination tablet, which was previously conducted in real-world settings. The parameters extracted from the PMS data were sociodemographic characteristics, NRS, EQ-5D, and dichotomous physician’s global impression of treatment effectiveness (PGI). The optimal cutoff points of MCIC for EQ-5D utility and NRS scores were evaluated using receiver operating characteristics (ROC) analysis. An anchor-based approach using PGI was applied.ResultsData of 710 patients with chronic noncancer pain were extracted from the PMS database. The NRS score decreased by 2.7 (standard deviation, 2.3) points, whereas the EQ-5D score increased by 0.16 (0.20) points at 4 weeks from baseline. The changes from baseline in NRS and EQ-5D were significantly correlated (r = 0.53, p < 0.001). The estimated optimal cutoff points of MCIC for EQ-5D and NRS were 0.10 and −2.0 points, respectively. The area under the curve of ROC was > 0.80 in both analyses.ConclusionThese results demonstrated novel cutoff criteria for the Japanese version of EQ-5D, focusing on patients with chronic noncancer pain. The obtained criteria were fairly consistent and can be confidently utilized as an evaluation tool in medical research on chronic noncancer pain in Japan, with additional functionality and usability for QoL assessment in pain management practice.Trial registrationThe data source of this post hoc research was a PMS study with the identifier number UMIN000015901 at umin.ac.jp, UMIN clinical trial registry (UMIN-CTR).

Association between somatic symptom burden and health-related quality of life in people with chronic low back pain

“…The difference in EQ-5D between SSS-8 high vs. very high was 0.04 and was 0.22 for SSS-8 non to minimal vs. very high among all participants. Yoshizawa et al reported that the minimal clinically important change (MCIC) of EQ-5D in Japanese with chronic noncancer pain was 0.10 [ 30 ]. Soer et al reported a MCIC of EQ-5D as 0.03 in patients with chronic LBP in Netherland [ 31 ].…”

The effect of three exercise approaches on health-related quality of life, and factors associated with its improvement in chronic whiplash-associated disorders: analysis of a randomized controlled trial

“…Regarding EQ-5D we have not been able to find a suggested level of MCID in chronic neck pain following non-surgical interventions. In chronic pain including rheumatoid arthritis, post-herpetic neuralgia, lumbago, etc., the overall MCID of EQ-5D is suggested to be 0.1 following introduction of pain medication [ 47 ]. Since the MCID depends on the kind of disease/pain condition and of the intervention in question, this level may not be true in chronic WAD.…”

How is the minimal clinically important difference established in health-related quality of life instruments? Review of anchors and methods

“…-The ROC: created by plotting the sensitivity of the instrument (the true positive rate) against the specificity (the false positive rate). Some studies [43,51,54,60,76] identified the MCID as the upper corner of the curve, and other studies [45,65] identified the MCID as the point of the receiver operating characteristics curve in which sensitivity and specificity are maximized (Maximum (Sensibility+Specificity-1), Youden index). The area under the curve (AUC) was always calculated to measure the instrument responsiveness, suggesting AUC values upper than 0.…”

“…-Methodology: even for the same pathology, MCID values were variable. For example, for EQ-5D, MCID values, using anchor and/or distribution-based methods, varied from 0.01 to 0.39 for patients with rheumatology/musculoskeletal disorders [47][48][49][50]54] and from 0.08 to 0.15 for oncology patients [31,40,43]. For patients with psychology disorders, MCID ranged from 0.05 to 0.08 using anchor-based method, and 0.04 to 0.1 using distribution-based method [65].…”