Meaningful change and patient-reported outcomes
A change in a patient-reported score that a statistical test can detect does not necessarily matter to the patient. Whether it matters is judged against the instrument's minimum clinically important difference, a threshold set by one or more of several approaches for a named instrument in a named condition.
What the minimum clinically important difference is
An 8-point difference from baseline to endpoint on a quality-of-life questionnaire scored from 0 (worst) to 80 (best) does not by itself say what it means to the patient or how it should inform decisions in disease management. Certain surrogate biologic variables are tightly associated with, or predictive of, hard clinical endpoints such as hospitalization, progression-free survival and overall longevity: lipid levels and blood pressure in cardiovascular diseases, CD4 T-cell counts and viral loads in HIV, liver enzymes in hepatic diseases and protein expression by tumour cells in cancer. The consequences to patients of changes in their scores on quality-of-life instruments, by contrast, are typically much less clear.
Jaeschke and colleagues introduced the concept of the minimum clinically important difference in 1989 and defined it as "the smallest difference in score in the domain of interest which patients perceive as beneficial and which would mandate, in the absence of troublesome side effects and excessive cost, a change in the patient's management." The definition comprises two constructs related to change: the minimum change discerned, and reported, by the patient, and the minimum variation on a patient-reported outcome that alters disease management. A difference, to be a difference, must make a difference, and the minimum clinically important difference is also called the critical effect size.
Setting the threshold
A minimum clinically important difference can be determined by four approaches, one informal and three more formal. The informal approach designates a change of at least 10% of the total score on an instrument, index or questionnaire as clinically meaningful.
The three more formal approaches are anchor-based, distribution-based and the Delphi method. The anchor-based approach compares changes on a questionnaire with some reference, the anchor, that patients consider meaningful. The anchor can be elicited prospectively by asking a patient or population of patients what improvement in quality of life, as measured by the instrument, they would require to agree to receive an intervention. The distribution-based approach uses some measure of population dispersion, for instance one standard error or one-half of the standard deviation. The Delphi method is an iterative process based on the average scores of ratings by experts.
If an intervention improves quality of life in the vast majority of individuals while leaving test scores unchanged in very few, the anchor method may need to be adjusted by using a higher threshold, a substantial rather than minimal clinically important difference, which separates patients who improve greatly from those who merely improve. If possible, the minimum clinically important difference should be computed by more than one approach.
The threshold in a report
A manuscript's Methods section should say what the minimum clinically important difference is, so that a reader can interpret results based on changes in assessment scores. If a responder definition or another qualitative modifier, such as an acceptable clinical response, is used in the Results, the Methods should say how it was defined quantitatively, whether imprecision boundaries were included, and whether it is biologically sound, anchored in some other way, such as in assessment norms or minimum clinically important differences, or arbitrary.
In one trial report of women treated for depression, the difference between the treatment and control groups' mean Beck Depression Inventory scores after the intervention was 3 points, and the critical effect size for the instrument was not given. The threshold turns out to be 5 points in the 2016 account of this case and appears to be about 5 points in the 2013 account. The mean scores were 5 in the treatment group and 8 in the control group, which differ by less than 5 points, so the difference is not clinically important by the 2016 account and probably not clinically important by the 2013 account. The trial's authors seemed to have based their interpretation solely on a significant P value, without considering the clinical implications of the results. Confusing statistical significance with clinical importance is the most common error of statistical interpretation in the literature.
Published thresholds by condition
Minimum clinically important differences on both the SF-36 (36-Item Short Form Health Survey) and the EQ-5D (EuroQol 5D) vary across disease states. The table below gives thresholds for representative chronic conditions with adverse effects on patients' quality of life and well-being, and some of its ranges encompass values computed by both anchor- and distribution-based methods. Apart from the six-minute walk test, most of its measures are patient-reported outcomes.
| Clinical category | Condition | Measure | Threshold (point or unit) |
|---|---|---|---|
| Autoimmune/allergic | Angioedema | AE-QoL (Angioedema Quality of Life questionnaire) | 6 |
| Autoimmune/allergic | Systemic lupus erythematosus | SLEDAI-2000 (Systemic Lupus Erythematosus Disease Activity Index 2000) | ≥1 decrease (improvement); ≥3 increase (worsening) |
| Cardiovascular | Coronary heart disease (cardiac rehabilitation) | 6MWT (six-minute walk test) | 25 m |
| Cardiovascular | Depression and anxiety in coronary heart disease | HADS-A (Hospital Anxiety and Depression Scale, Anxiety) | 3.80 |
| Cardiovascular | Depression and anxiety in coronary heart disease | HADS-D (Hospital Anxiety and Depression Scale, Depression) | 3.99 |
| Cardiovascular | Heart failure | 6MWT | 45 m |
| Cardiovascular | Heart failure | MLHFQ (Minnesota Living with Heart Failure questionnaire) | 5 to 7 (overall score) |
| Cardiovascular | Peripheral arterial disease (chronic critical limb ischaemia) | VascuQoL (Vascular Quality of Life questionnaire) | 0.36 to 0.48 (total score) |
| Acne | Acne-QoL (Acne-Specific Quality of Life questionnaire) | 0.50 to 10.3 mean change per item | |
| General (thresholds for severely impaired health-related quality of life) | Skindex-29 | Symptoms scale ≥52 points; Emotions scale ≥39; Functioning scale ≥37; ≥44 overall | |
| Hyperhidrosis | HidroQOL (Hyperhidrosis Quality of Life index) | 3 points (overall); 1.94 to 3.07 (generalized); 2.16 to 4.36 (axillary); 2.15 to 3.39 (palmoplantar) | |
| Psoriasis | DLQI (Dermatology Life Quality Index) | 3.2 | |
| Psoriasis | PASI (Psoriasis Area and Severity Index) | >75% | |
| Urticaria | DLQI | 2.24 to 3.10 | |
| Urticaria | CU-Q2oL (Chronic Urticaria Quality of Life Questionnaire) | 15 (Thai version) | |
| Endocrine | Cushing's disease | CushingQoL | 10.1 |
| Endocrine | Diabetes mellitus | HFS-II (Hypoglycaemia Fear Survey) Worry scale | 2.0 to 5.8 |
| Endocrine | Diabetes mellitus | DHP-18 (Diabetes Health Profile) | Psychological distress domain 6.99 to 10.59; Barriers to activity domain 6.48 to 9.89; Disinhibited eating domain 7.52 to 11.39 |
| Endocrine | Gout | GIS (Gout Impact Scale, 0–100) | 5 to 8 |
| Endocrine (thyroid disorders) | Graves orbitopathy | GO-QOL (Graves Ophthalmology Quality of Life) | ≥6 (minor treatment); ≥10 (orbital decompression or major treatment) |
| Endocrine (thyroid disorders) | Various | ThyPRO (Thyroid-Related Patient-Reported Outcome) | 0.2 = no change |
| Genitourinary | Benign prostatic hyperplasia | IPSS (International Prostate Symptom Score) | 3 points or >25% reduction on total IPSS; 3 = slight symptom improvement; 5 = moderate improvement |
| Genitourinary | Erectile dysfunction | Urinary flow rate | 2 mL/s |
| Genitourinary | Erectile dysfunction | IIEF (International Index of Erectile Function) | Overall 4; 2 for mild, 5 for moderate and 7 for severe erectile dysfunction |
| Genitourinary | Female pelvic floor disorders | UDI (Urinary Distress Inventory) | 6.4 to 22.4 |
| Genitourinary | Female pelvic floor disorders | UDI Stress | 4.6 to 16.5 |
| Genitourinary | Female pelvic floor disorders | UIQ (Urinary Impact Questionnaire) | 6.5 to 17.0 |
| Genitourinary | Stress urinary incontinence | I-QOL (Incontinence Quality of Life) | 2.5 to 6.3 |
| Gastrointestinal | Gastro-oesophageal reflux | NDI | 10 |
| Gastrointestinal | Inflammatory bowel disease | SIBDQ (Short Inflammatory Bowel Disease Questionnaire) | 16 |
| Gastrointestinal | Crohn's disease | CDAI | 50 |
| Gastrointestinal | Crohn's disease | SF-36 PCS | 1.6 to 7.0 |
| Gastrointestinal | Crohn's disease | SF-36 MCS | 2.3 to 8.7 |
| Gastrointestinal | Crohn's disease | EQ-5D VAS | 4.2 to 14.8 |
| Gastrointestinal | Ulcerative colitis | Mayo Score | 2.5 |
| Gastrointestinal | Irritable bowel syndrome | SIP (Sickness Impact Profile) | 2.8 |
| Gastrointestinal | Irritable bowel syndrome | IBS-QOL (Irritable Bowel Syndrome Quality of Life) | 10 to 14 |
| Gastrointestinal | Irritable bowel syndrome | IBS pain severity | 50% |
| Musculoskeletal | Ankylosing spondylitis | BASFI (Bath Ankylosing Spondylitis Function Index) | 7 mm or 17.5% |
| Musculoskeletal | Ankylosing spondylitis | BASDAI (Bath Ankylosing Spondylitis Disease Activity Index) | 10 mm or 22.5% |
| Musculoskeletal | Ankylosing spondylitis | BAS-G (Bath Ankylosing Spondylitis Patient Global score) | 15 mm or 27.5% |
| Musculoskeletal | Fibromyalgia | BPI (Brief Pain Inventory) Pain (average) | 2.1 (32.3% reduction from baseline) |
| Musculoskeletal | Fibromyalgia | BPI Severity | 2.2 (34.2% reduction from baseline) |
| Musculoskeletal | Pain and disability after anterior cervical discectomy | VAS (visual analogue scale) | Neck pain 2.6; arm pain 4.1 |
| Musculoskeletal | Pain and disability after anterior cervical discectomy | NDI | 17.3% |
| Musculoskeletal | Pain and disability after anterior cervical discectomy | SF-12 PCS | 8.1 |
| Musculoskeletal | Pain and disability after anterior cervical discectomy | SF-12 MCS | 4.7 |
| Musculoskeletal | Pain and disability after lumbar-spine surgery | ODI (Oswestry Disability Index) | 12.8 |
| Musculoskeletal | Pain and disability after lumbar-spine surgery | SF-36 PCS | 4.9 |
| Musculoskeletal | Duchenne muscular dystrophy | 6MWT | 20 to 30 m |
| Musculoskeletal | Osteoarthritis and other chronic musculoskeletal pain, including rheumatoid arthritis | PI-NRS (Pain Intensity Numerical Rating Scale) | 1 point or 15% |
| Musculoskeletal | Rheumatoid arthritis | CDAI | 12 for high, 6 for moderate and 1 to 2 for low disease activity |
| Musculoskeletal | Fatigue in rheumatoid arthritis | FSS (Fatigue Severity Scale) | 20.2 |
| Musculoskeletal | Fatigue in rheumatoid arthritis | Vitality scale of the SF-36 | 14.8 |
| Musculoskeletal | Fatigue in rheumatoid arthritis | MAF (Multidimensional Assessment of Fatigue) | 18.7 |
| Musculoskeletal | Fatigue in rheumatoid arthritis | MFI (Multidimensional Fatigue Inventory) | 16.6 |
| Musculoskeletal | Fatigue in rheumatoid arthritis | FACIT-F (Functional Assessment of Chronic Illness Therapy-Fatigue) | 15.9 |
| Musculoskeletal | Fatigue in rheumatoid arthritis | CFS (Chalder Fatigue Scale) | 9.9 |
| Musculoskeletal | Fatigue in rheumatoid arthritis | 10-point numerical rating scale | 19.7 for normalized scores (0–100) |
| Neurologic | Alzheimer's disease | MMSE (Mini-Mental State Examination) | 1.40 to 3.72 |
| Neurologic | Alzheimer's disease | BADLS (Bristol Activities of Daily Living Scale) | 3.5 |
| Neurologic | Alzheimer's disease | NPI (Neuropsychiatric Inventory) | 8 |
| Neurologic | Generalized anxiety disorder | HAM-A (Hamilton Anxiety Rating Scale) | 50% |
| Neurologic | Cerebrovascular disease (stroke) | ARAT (Action Research Arm Test) | 12 (nondominant hand and arm) and 17 (dominant hand and arm) |
| Neurologic | Upper-extremity motor function | WMFT (Wolf Motor Function Test) and MAL (Motor Activity Log) | 1.0 and 1.2 |
| Neurologic | Depression in acute stroke | HADS (Hospital Anxiety and Depression Scale) | >8 indicates depression |
| Neurologic | Multiple sclerosis | 6MWT | 53.35 to 55.06 m |
| Neurologic | Multiple sclerosis | BBT (Box and Block Test, dominant hand) | 3.48 to 5.23 blocks |
| Neurologic | Multiple sclerosis | GAS (Goal Attainment Scale) | 10 for mean GAS |
| Neurologic | Early Parkinson's disease | UPDRS (Unified Parkinson's Disease Rating Scale) | UPDRS II 1.8 to 2.0; UPDRS III 6.1 to 6.2; UPDRS II + III 8.0 to 8.1 |
| Neurologic | Advanced Parkinson's disease | UPDRS | UPDRS II 1.8 to 2.3; UPDRS III 5.2 to 6.5; UPDRS II + III 7.1 to 8.8 |
| Neurologic | Post-traumatic stress disorder | EQ-5D | 0.04 to 0.10 |
| Neurologic | Post-traumatic stress disorder | QWB-SA (Quality of Well-Being Scale-Self-Administered) | 0.02 to 0.05 |
| Neurologic | Schizophrenia | PANSS (Positive and Negative Syndrome Scale) | 20.2 |
| Neurologic | Schizophrenia | Lenert utility score | 0.15 |
| Neurologic | Schizophrenia | HCQLS (Heinrichs-Carpenter Quality of Life Scale) | 1.13 |
| Neurologic | Schizophrenia | QWB | 0.17 |
| Oncology | Oropharyngeal cancer | UW-QOL (University of Washington Quality of Life Questionnaire) | 13.07 (global quality of life) |
| Oncology | Oropharyngeal cancer | EORTC C30 (European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30) | 9.43 (global quality of life) |
| Oncology | Gynecologic cancers | FACT-G (Functional Assessment of Cancer Therapy-General) | Positive 4.37 (4.0%) to 5.48 (4.9%) |
| Oncology | Gynecologic cancers | PROMIS (Patient-Reported Outcomes Measurement Information System) | 3 to 5 (fatigue, anxiety and depression) |
| Oncology | Prostate cancer | PCI (Prostate Cancer Index) | Urinary function 8; bowel function 7; sexual function 8; urinary bother 9; bowel bother 8; sexual bother 11 |
| Oncology | Prostate cancer | SF-36 domains | Physical 7; role physical 14; bodily pain 9; general health 8; mental health 6; role emotional 12; energy/vitality 9; social function 9 |
| Respiratory | Asthma | AQLQ (Asthma Quality of Life Questionnaire) | Total score 0.52 |
| Respiratory | Cystic fibrosis | CFQ-R (Cystic Fibrosis Questionnaire-Revised) | 4 (stable disease) to 8.5 (exacerbation) |
| Respiratory | Chronic obstructive pulmonary disease | TDI (Transition Dyspnoea Index) | 1 |
| Respiratory | Chronic obstructive pulmonary disease | SGRQ (St. George's Respiratory Questionnaire) | 4 |
| Respiratory | Chronic obstructive pulmonary disease | CCQ (Clinical COPD Questionnaire) | 0.4 |
| Respiratory | Chronic heart failure and chronic lung disease | CRQ (Chronic Respiratory Disease Questionnaire) | Overall approximately 0.50; dyspnoea 0.43; fatigue 0.64; emotional function 0.49 |
| Respiratory | Obstructive sleep apnoea | SAQLI (Sleep Apnea Quality of Life Index) | 1.0 |
| Respiratory | Rhinoconjunctivitis | RQLQ (Rhinoconjunctivitis Quality of Life Questionnaire) | 0.5 |
| Respiratory | Rhinoconjunctivitis | Mini-RQLQ | 0.4 |
| Respiratory | Rhinoconjunctivitis | TNSS (Total Nasal Symptom Score) | 0.55 |
Some abbreviations in the table stand for more than one instrument: CDAI is either the Crohn's Disease Activity Index or the Clinical Disease Activity Index used in rheumatoid arthritis, and NDI is either the Neck Disability Index or the Nepean Dyspepsia Index.