Minimally important change values of a measurement instrument depend more on baseline values than on the type of intervention

doi:10.1016/j.jclinepi.2014.07.008

Journal of Clinical Epidemiology

Volume 68, Issue 5, May 2015, Pages 518-524

https://doi.org/10.1016/j.jclinepi.2014.07.008 Get rights and content

Abstract

Objectives

Multi-item questionnaires are frequently used to measure outcomes in randomized controlled trials (RCTs) in patients with sciatica. Knowing the minimaly important change (MIC) values for these instruments will facilitate interpretation of change scores. MIC values have been shown to be dependent on baseline values. The question is whether they also depend on the type of intervention. To estimate the MIC of the Roland Morris Disability Questionnaire (modified 23 item version) (RMDQ) and of intensity of leg pain measured by a Visual Analogue Scale (VAS) in patients with sciatica and to assess to what extent MIC values depend on type of intervention and on baseline values.

Study Design and Setting

This is a secondary analysis of RCT data of the effects of early surgery vs. prolonged conservative treatment in patients with sciatica. Baseline and 8-week data were used to assess MIC of the RMDQ-23 and VAS leg pain. We used the receiver operator characteristic (ROC) method to assess the MIC. Global Perceived Recovery (rated 8 weeks after baseline) was used as anchor. Subgroups were created based on type of treatment and baseline severity.

Results

The MIC value of the RMDQ-23 for the total group of sciatica patients was 7.5. The values were 8.1 and 6.9 for surgery and conservative treatment, respectively. For high and low baseline values, the MICs were 9.0 and 4.9, respectively, irrespective of treatment received. The MIC values of the VAS leg pain were 34.4 for the total group. For surgery and conservative treatment, the MIC values were 38.5 and 30.4, respectively, whereas for groups with high and low baseline values, MIC values of 53.5 and 17.2 were found.

Conclusion

The MIC values of the RMDQ-23 and VAS leg pain were found to be highly dependent on their baseline values, although the type of intervention appeared to influence the MIC value only slightly.

Introduction

Sciatica is a common health problem causing severe pain and disability in individual patients and high costs to society [1]. Sciatica (lumbar radiculopathy) is characterized by radiating pain in the leg typically served by one nerve root in the lumbar or sacral spine. It is sometimes also associated with sensory and motor deficits. The most common cause of sciatica is a herniated disk with associated nerve root compression. The most important symptoms are pain in the leg and related disability. The diagnosis is mainly based on history taking and physical examination. Diagnostic imaging is indicated directly when severe underlying pathology in suspected (infections, malignancies), otherwise only after 6–8 weeks when severe symptoms fail to respond to conservative care. Surgery is indicated when computed tomography and/or magnetic resonance imaging show disk herniation, but only when the clinical findings and symptoms (eg, location of leg pain) correspond well with the imaging findings [2].

The natural history of acute sciatica is favorable, with resolution of leg pain within 8 weeks from the onset in most patients [2], [3]. From this perspective, optimal care starts with symptomatic and conservative treatment. Only in case of persistent severe complaints, more invasive treatment might be indicated.

To assess health status and outcome of interventions for sciatica patients, patient-reported outcomes assessing disability and pain are most relevant. The Roland Morris Disability Questionnaire (RMDQ) [4] is frequently used to measure functional status in back pain-related disability. The RMDQ was modified for sciatica patients by Patrick et al. [5], by deleting one item and exchanging four others with alternative statements resulting in a 23-item questionnaire with increased responsiveness to sciatica. In this study, we used this RMDQ-23 for measuring disability in sciatica patients.

As leg pain is often more pronounced than back pain in patients with sciatica, we also assessed pain intensity in the leg, using a visual analogue scale (VAS) [6].

In randomized controlled trials (RCTs), statistical significance of a difference in change of scores over time does not necessarily imply that this change is clinically relevant. To facilitate interpretation of outcomes, benchmarks in pain improvement have been suggested, for example by the IMMPACT group, for minimal important improvement (10–20% pain reduction), moderate improvement (≥30%), and substantial improvement (≥50% pain reduction). [7] A minimally important change (MIC), defined as the smallest change in score that patients perceive as important, is assumed to be a characteristic of a measurement instrument and therefore assumed not to depend on the type of intervention that patients receive. However, there is little evidence to support the assumption. Some studies have shown that the MIC is dependent on the initial baseline scores [8], [9] and it has been suggested that MIC might also depend on the invasiveness or inconvenience of the intervention [10]. We used data from a RCT [ISRCTN26872154] that compared early surgery with prolonged conservative treatment among patients with sciatica due to lumbar disc herniation [11]. In this population with patients receiving quite different treatments, we examined whether the assumption holds that MIC values are independent of type of intervention received. Additionally, we also examine the influence of baseline values.

Section snippets

Materials and methods

In the multicenter RCT, 283 patients who had severe sciatica for 6 to 12 weeks were randomly assigned to early surgery or to prolonged conservative treatment with surgery if needed. The details of this study can be found elsewhere [11], [12]. Briefly, patients recruited were aged 18 to 65 years, with a radiologically confirmed disc herniation and lumbosacral radicular syndrome diagnosed by the attending neurologist who had lasted for 6 to 12 weeks. Early surgery was scheduled within 2 weeks

Results

Some demographic characteristics and baseline values of 243 patients are presented in Table 1. Table 2 lists the mean change scores with their standard deviations on the RMDQ-23 and the VAS leg pain for each score on the general and leg pain–specific GPR question.

As shown in the last rows of Table 2, 139 of 243 patients (57.4%) were considered as “importantly improved” (GPR score 1–2) and 95 of 243 (38.8%) as “not importantly changed” (GPR score 3–5). The Spearman correlation of RMDQ-23 change

Discussion

For subgroups according to the treatment patients received, we found only slight differences in the MIC value for RMDQ-23 and VAS for leg pain, although they appeared to differ substantially for subgroups with high and low baselines for these measures.

In this analysis, we deviated from the intention-to-treat analysis [12] because we wanted to compare the MIC values for patients who indeed underwent surgery and who had actually received conservative treatment. Therefore, we excluded 17 patients

References (24)

S.L. Collins et al.
The visual analogue pain intensity scale: what is moderate pain in millimetres?
Pain
(1997)
R.H. Dworkin et al.
Interpreting the clinical importance of treatment outcomes in chronic pain clinical Trials: IMMPACT Recommendations
J Pain
(2008)
J.T. Farrar et al.
Clinical importance of changes in chronic pain intensity measured on an 11-point rating scale
Pain
(2001)
R.A. Deyo et al.
Assessing the responsiveness of functional scales to clinical change: an analogy to diagnostic test performance
J Chronic Dis
(1986)
C.-Y. Lai et al.
Exact confidence interval estimation for the Youden index and its corresponding optimal cut-point
Comput Stat Data Anal
(2012)
C. Demoulin et al.
What factors influence the measurement properties of the Roland-Morris disability questionnaire?
Eur J Pain
(2010)
R. Jaeschke et al.
Measurement of health status. Ascertaining the minimal clinically important difference
Control Clin Trials
(1989)
M.L. Ferreira et al.
The smallest worthwhile effect of nonsteroidal anti-inflammatory drugs and physiotherapy for chronic low back pain: a benefit-harm trade-off study
J Clin Epidemiol
(2013)
L.C. Lambeek et al.
The trend in total cost of back pain in the Netherlands in the period 2002 to 2007
Spine
(2011)
B.W. Koes et al.
Diagnosis and treatment of sciatica
BMJ
(2007)

P.C. Vroomen et al.

Conservative treatment of sciatica: a systematic review

J Spinal Disord

(2000)

M. Roland et al.

A study of the natural history of low-back pain. Part II: development of guidelines for trials of treatment in primary care

Spine (Phila Pa 1976.)

(1983)

Cited by (41)

How much is enough? Considering minimally important change in youth mental health outcomes
2022, The Lancet Psychiatry
To make decisions in mental health care, service users, clinicians, and administrators need to make sense of research findings. Unfortunately, study results are often presented as raw questionnaire scores at different time points and regression coefficients, which are difficult to interpret with regards to their clinical meaning. Other commonly reported treatment outcome indicators in clinical trials or meta-analyses do not convey whether a given change score would make a noticeable difference to service users. There is an urgent need to improve the interpretability and relevance of outcome indicators in youth mental health (aged 12–24 years), in which shared decision making and person-centred care are cornerstones of an ongoing global transformation of care. In this Personal View, we make a case for considering minimally important change (MIC) as a meaningful, accessible, and user-centred outcome indicator. We discuss what the MIC represents, how it is calculated, and how it can be implemented in dialogues between clinician and researcher, and between youth and clinician. We outline how use of the MIC could enhance reporting in clinical trials, meta-analyses, clinical practice guidelines, and measurement-based care. Finally, we identify current methodological challenges around estimating the MIC and areas for future research. Efforts to select outcome domains and valid measurement instruments that resonate with youth, families, and clinicians have increased in the past 5 years. In this context, now is the time to define demarcations of changes in outcome scores that are clinically relevant, and meaningful to youth and families. Through the use of MIC, youth-centred outcome measurement, analysis, and reporting would support youth-centred therapeutic decision making.
Minimal important change was on the lower spectrum of previous estimates and responsiveness was sufficient for core outcomes in chronic low back pain
2022, Journal of Clinical Epidemiology
The objective of this study was to estimate the minimal important change (MIC) and responsiveness of core patient reported outcome measures for chronic low back pain (LBP) and Modic changes.
In the Antibiotics in Modic changes (AIM) trial we measured disability (RMDQ, ODI), LBP intensity (NRS) and health-related quality of life (EQ5D) electronically at baseline, three- and 12-month follow-up. MICs were estimated using Receiver Operating Curve (ROC) curve and Predictive modeling analyses against the global perceived effect. Credibility of the estimates was assessed by a standardized set of criteria. Responsiveness was assessed by a construct and criterion approach according to COSMIN guidelines.
The MIC estimates of RMDQ, ODI and NRS scores varied between a 15–40% reduction, depending on including “slightly improved” in the definition of MIC or not. The MIC estimates for EQ5D were lower. The credibility of the estimates was moderate. For responsiveness, five out of six hypotheses were confirmed and AUC was >0.7 for all PROMs.
When evaluated in a clinical trial of patients with chronic LBP and Modic changes, MIC thresholds for all PROMs were on the lower spectrum of previous estimates, varying depending on the definition of MIC. Responsiveness was sufficient.
Minimal important changes and differences were estimated for Oxford hip and knee scores following primary and revision arthroplasty
2022, Journal of Clinical Epidemiology
To create estimates for clinically meaningful changes and differences in pain and joint function for the Oxford Hip and Knee Scores (OHS/OKS) in primary and revision joint replacement.
694,487 primary and revision joint replacement procedures were analyzed from the NHS PROMs dataset between 2012-2020. Minimal important changes (MIC) and differences (MID) were calculated using distribution and anchor-based methods (including receiver-operating characteristic (ROC) curve and predictive-modelling techniques).
For comparison of two or more groups (such as in a clinic trial), MID estimates were ∼5 points. For cohort studies investigating changes over time in a single group of patients, $M I C_{g r o u p}$ estimates were 12.4 points (primary hip replacement), 8.6 points (revision hip replacement), 10.5 points (primary knee replacement) and 9.4 points (revision knee replacement). For studies investigating changes over time at the individual patient level, $M I C_{a d j u s t e d}$ estimates were ∼8 points, ∼6 points, ∼7 points and ∼6 points respectively.
This study has calculated contemporary estimates of clinically important changes and differences for the OHS/OKS for primary and revision hip and knee replacement. These estimates can be used to inform sample size calculations and to interpret changes in joint function over time and differences between groups.
Thresholds for meaningful improvement in WOMAC scores need to be adjusted to patient characteristics after hip and knee replacement
2022, Journal of Orthopaedics
To calculate unstratified and patient-specific meaningful improvement (MI) and patient acceptable symptom states (PASS) for the WOMAC total score in patients after total hip (THR) or total knee replacement (TKR).
A retrospective observational cohort study. Anchor-based receiver operator characteristics curves were used to estimate MI and PASS thresholds.
Recovery paths were specific to individual characteristics of patients. An unstratified 12-months MI threshold of 28.1 (PASS: 13.3) and 17.8 (PASS: 15.8) for patients after THR and TKR, respectively, would unfairly detect critical recovery paths.
Thresholds for treatment success need to be as patient-specific as possible.
The minimal important change for the seven-item disability of the arm, shoulder, and hand (DASH 7) questionnaire – Assessing shoulder function in patients with subacromial pain
2021, JSES International
Citation Excerpt :
An AUC greater than 0.7 indicates an adequate ability to distinguish between the improved patients and those not improved,42 as this result is considered as evidence for good responsiveness of the DASH 7. As mentioned in the introduction, MIC values are reported to possibly differ depending on patient population,5,6,8 context,5 and also on the method used for calculation.1,5,8 Rysstad et al (2017)35 reported an MIC value at 4.4 points for the original 30-item DASH in patients with subacromial pain which can be considered in line with 6.5 for the DASH 7.
The DASH 7 is a recently published activity-related 7-item short form of the disability of the arm, shoulder, and hand (DASH) questionnaire developed to assess shoulder function in patients with subacromial pain. Before implementation in both intervention studies and in clinical practice, it is essential to evaluate its responsiveness. The objective of this study was to determine the minimal important change (MIC) in the DASH 7 questionnaire for patients with subacromial pain after a 3 months exercise intervention in a primary care context.
In this psychometric study the anchor-based MIC-distribution method was used to establish the MIC. The Patient Global Impression of Change (PGIC) was used as external criterion. Data from a clinical implementation study, aimed to implement a specific exercise strategy for patients with subacromial pain among physiotherapists in primary care, were used. Data from 70 patients were included in the analyses.
The correlation coefficient between Patient Global Impression of Change and the DASH 7 score change was 0.67 and the area under the curve was 0.94 (95% confidence interval: 0.88-1.0). The MIC_ROC for improvement was detected at a mean change in 6.5 points with the sensitivity at 0.98 (98%) and the specificity at 0.78 (78%), and the MIC_{95% limit} for improvement was detected at a mean change of 25.7 points. There were 77% of the patients who reached at least this MIC_ROC and 51% who reached at least the MIC_{95% limit} after 3 months of exercise intervention_.
The DASH 7 is responsive to change over time and can discriminate between patients considered to be improved and patients considered not improved. These MIC values for patients with subacromial pain in the primary care setting can be used in clinical practice and in intervention studies as an indication on the patients clinically important level of score change for improvement.
Baseline health status and setting impacted minimal clinically important differences in COPD: an exploratory study
2019, Journal of Clinical Epidemiology
Citation Excerpt :
This could possibly be a solution and may be applied in clinical practice to interpret individual change scores. It has been hypothesized that disease severity—measured in the present study by spirometry—could impact the MCID [16,17,20,23–39]. Previous—but also present—research demonstrated that worse health status was correlated with worse lung function [9–13], although this correlation was only weak to moderate [4–7].
Minimal clinically important differences (MCIDs) are used as fixed numbers in the interpretation of clinical trials. Little is known about its dynamics. This study aims to explore the impact of baseline score, study setting, and patient characteristics on health status MCIDs in chronic obstructive pulmonary disease (COPD).
Baseline and follow-up data on the COPD Assessment Test (CAT), Clinical COPD Questionnaire (CCQ), and St. George's Respiratory Questionnaire (SGRQ) were retrospectively analyzed from pulmonary rehabilitation (PR) and routine clinical practice (RCP). Anchor- and distribution-based MCID estimates were calculated and tested between settings, gender, age, Global initiative for Obstructive Lung Disease (GOLD) classification, comorbidities, and baseline health status.
In total, 658 patients were included with 2,299 change score measurements. MCID estimates for improvement and deterioration ranged for all subgroups 0.50–6.30 (CAT), 0.10–0.84 (CCQ), and 0.33–12.86 (SGRQ). Larger MCID estimates for improvement and smaller ones for deterioration were noted in patients with worse baseline health status, females, elderly, GOLD I/II patients, and patients with less comorbidities. Estimates from PR were larger.
Baseline health status and setting affected MCID estimates of COPD health status questionnaires. Patterns were observed for gender, age, spirometry classification, and comorbidity levels. These outcomes would advocate the need for tailored MCIDs.

View all citing articles on Scopus

: Conflict of interest: None.

View full text

Original ArticleMinimally important change values of a measurement instrument depend more on baseline values than on the type of intervention

Abstract

Objectives

Study Design and Setting

Results

Conclusion

Introduction

Section snippets

Materials and methods

Results

Discussion

Pain

J Pain

Pain

J Chronic Dis

Comput Stat Data Anal

Eur J Pain

Control Clin Trials

J Clin Epidemiol

The trend in total cost of back pain in the Netherlands in the period 2002 to 2007

Spine

Diagnosis and treatment of sciatica

BMJ

Conservative treatment of sciatica: a systematic review

J Spinal Disord

A study of the natural history of low-back pain. Part II: development of guidelines for trials of treatment in primary care

Spine (Phila Pa 1976.)

Original Article
Minimally important change values of a measurement instrument depend more on baseline values than on the type of intervention