Skip to content

Outcomes and measurement

TL;DR — Fibromyalgia trials measure a multidimensional illness with instruments and endpoints that were only standardized after the pivotal-trial era. The FIQ (1991) and FIQR (2009) are the disease-specific workhorses; the FIQ's minimal clinically important difference is ~14%, with severity bands mild <39 / moderate 39–<59 / severe ≥59 (Burckhardt 1991, PMID 1865419; Bennett 2009, PMID 19664287; Bennett 2009, PMID 19369473). OMERACT consensus (2009) fixed the core domain set — pain, tenderness, fatigue, patient global, multidimensional function, sleep — and seeded responder-index work (Mease 2009, PMID 19820221). Regulatory trials count ≥30%/≥50% pain responders plus PGIC, and composite responder definitions (≥30% pain + ≥10% function + symptom criteria) outperform single endpoints (Arnold 2012, PMID 21953205). The context that swamps all of it: ~31% of placebo-treated patients achieve ≥30% pain relief and ~19% achieve ≥50%, while ~11% drop out of placebo arms from adverse events — placebo/nocebo responses that are most of the observed drug response (Häuser 2011, PMID 22120916; Häuser 2012, PMID 23137770). Cross-trial comparison is further broken by shifting diagnostic criteria, LOCF imputation, and enriched-withdrawal designs.

Disease-specific instruments: FIQ and FIQR

  • FIQ (1991): 10-item self-report of physical functioning, work, depression, anxiety, sleep, pain, stiffness, fatigue, well-being; developed for women with FM; adequate initial reliability/validity (Burckhardt 1991, PMID 1865419). Became the standard disease-specific trial instrument, but with known deficiencies and a "cumbersome scoring algorithm" limiting clinical use (Bennett 2009, PMID 19664287).
  • FIQR (2009): same three domains (function, overall impact, symptoms), 0–10 numeric scales throughout, adds memory, tenderness, balance, environmental sensitivity items; completion 1.3 min; total score correlates r=0.88 with FIQ, allowing longitudinal comparability; discriminates FM (mean 56.6) from RA/SLE (28.6), major depression (17.3), healthy controls (12.1) (Bennett 2009, PMID 19664287).
  • Interpretation anchors (from pooled pregabalin-trial modeling, n=2,228): FIQ MCID ≈ 14% (95% CI 13–15); severity bands: mild 0–<39, moderate 39–<59, severe ≥59 (Bennett 2009, PMID 19369473). Trial baselines typically sit at the moderate/severe border (mean 62 in that dataset) — meaning "responders" often remain in the moderate-severity range.
  • Generic instruments ride alongside: SF-36 (validated against FIQR; Bennett 2009, PMID 19664287), Brief Pain Inventory (primary outcome of the pivotal duloxetine trial; Arnold 2005, PMID 16298061), and increasingly PROMIS short forms for sleep and fatigue (used as secondary endpoints in the TNX-102 SL phase 3 program; Lederman 2023, PMID 37165930).

OMERACT: how the field agreed what to measure

The OMERACT fibromyalgia working group (2005–2011 cycle) ran patient focus groups, patient and clinician Delphi exercises, and multivariate analysis of 10 trials across 4 drugs, converging at OMERACT 9 (2009) with >70% agreement on a core domain set for all FM trials: pain, tenderness, fatigue, patient global, multidimensional function, sleep disturbance; dyscognition and depression for some trials; stiffness, anxiety, functional imaging, and CSF biomarkers as research agenda (Mease 2009, PMID 19820221). The follow-on program aimed at a responder index and continuous disease-activity score, explicitly modeled on RA's DAS heuristic — while acknowledging FM's structural handicaps: no objective severity marker and no treatment-anchored definition of disease activity (Mease 2011, PMID 21724721). The core set is why post-2010 trials (e.g., RELIEF/RESILIENT) report a standard panel — daily pain, PGIC, FIQR domains, PROMIS sleep/fatigue — rather than idiosyncratic outcome menus (Lederman 2023, PMID 37165930).

Responder definitions

Definition Content Performance Source
≥30% pain reduction "moderate benefit" (IMMPACT-aligned convention used across Cochrane reviews) Placebo rate ~31%; drug−placebo differences ~10 points Häuser 2011, PMID 22120916; Derry 2016, PMID 27684492
≥50% pain reduction "substantial benefit" Placebo rate ~19%; NNTs 8–14 for approved drugs Häuser 2011, PMID 22120916; Lunn 2014, PMID 24385423
PGIC much/very much improved patient-anchored global Tracks ~with ≥30% pain; can dissociate (RELIEF: pain met, PGIC not; Lederman 2023, PMID 37165930) Welsch 2018, PMID 29489029
FM30 composite (short) ≥30% pain + ≥10% physical function + ≥30% in sleep or fatigue RR 1.50 (1.24–1.82) drug vs placebo, pooled across 12 RCTs/4 drugs Arnold 2012, PMID 21953205
FM30 composite (long) ≥30% pain + ≥10% function + ≥30% in 2 of sleep/fatigue/depression/anxiety/cognition RR 1.60 (1.31–1.96) Arnold 2012, PMID 21953205
FDA registration composites (milnacipran model) pain + PGIC (± SF-36 PCS) concurrent responder Used prospectively in pivotal trials for the "FM" vs "FM pain" indications Clauw 2008, PMID 19108787

The composite-endpoint debate in one paragraph: single pain endpoints reward drugs with narrow analgesic action and are inflated by high placebo response; composites reflect the OMERACT multidimensional construct and best discriminated drug from placebo among 24 candidates tested (Arnold 2012, PMID 21953205) — but they lower absolute response rates (≥15% per arm), complicate NNT comparisons with older trials, and embed contestable weightings (why ≥10% function rather than ≥20%?). The sodium oxybate program shows the stakes: its pivotal-program primary outcome was a ≥20% composite (pain VAS + FIQ + PGIC) (Russell 2009, PMID 19116896) — success on a composite did not translate into approval (pharmacologic-therapy).

Placebo and nocebo: the dominant "treatment" in FM trials

  • Pooled placebo arms of 12-week+ drug RCTs (30 studies, 3,846 placebo patients): 30.8% achieve ≥30% pain reduction and 18.8% achieve ≥50% on placebo; true-drug RRs over placebo are only 1.38 (30% outcome) and 1.57 (50% outcome) (Häuser 2011, PMID 22120916).
  • In the approval-directed trials specifically (18 studies, 3,546 placebo patients): ≥50% placebo response 18.6% (17.4–19.9); nocebo dropout — placebo-arm discontinuation for adverse events — 10.9% (9.9–11.9) (Häuser 2012, PMID 23137770). A large share of both measured benefit and measured harm of active drug is therefore context effect, not pharmacology.
  • Implications: (1) between-arm deltas of 0.5–1.0/10 sit on top of within-arm improvements several times larger; (2) trial designs that inflate expectancy (active run-ins, intensive contact) inflate both arms; (3) clinical practice deliberately harnesses what trials try to minimize (Häuser 2012, PMID 23137770). Sham-controlled device/procedure literatures (acupuncture, neuromodulation) show the same signature — large open-label effects shrinking against adequate shams (Deare 2013, PMID 23728665; non-pharmacologic-therapy).

Why cross-trial comparison is hard

  1. Population definition moved under the trials. Pivotal programs enrolled ACR 1990 tender-point-defined patients (e.g., Crofford 2005, PMID 15818684); later trials use 2010/2011/2016 criteria patients — differently female-skewed, differently severe (diagnostic-criteria). NNTs from one population do not automatically transfer.
  2. Imputation. The classic-design trials analyzed primary outcomes with last-observation-carried-forward; with dropout rates of 20–40%, Cochrane reviewers repeatedly warn LOCF "could overestimate treatment effect" (Derry 2016, PMID 27684492; Cording 2015, PMID 26482422).
  3. Enrichment and withdrawal designs. EERW trials (pregabalin FREEDOM: 1,051 enter open-label; only responders — ~54% — are randomized to continue vs withdraw) answer "does it keep working in responders?", not "does it work?"; in the Cochrane pooled EERW analysis the maintained-response NNT of 5 becomes 12 when normalized to the starting population (Crofford 2008, PMID 18400400; Derry 2016, PMID 27684492). Mixing EERW and parallel-design results in one league table is a category error.
  4. Endpoint era effects. Pre-OMERACT trials reported heterogeneous outcome menus; post-2009 trials report the core set; composite-responder trials (milnacipran) cannot be NNT-compared to single-endpoint trials (pregabalin) without recomputation (Mease 2009, PMID 19820221; Clauw 2008, PMID 19108787).
  5. Trial duration and exclusions. Most RCTs run 8–15 weeks; the antidepressant NMA notes that most trials measured only short-term outcomes and excluded people with low mood and other mental health conditions — common in the clinic population — leaving no reliable long-term efficacy or safety evidence for any drug (Birkinshaw 2023, PMID 37160297).

Patient-priority vs trial-measured outcomes

OMERACT's patient Delphi exercises were the formal channel bringing patient priorities into the domain set (Mease 2009, PMID 19820221), but discrepancies persist. In the 2,596-person NFA internet survey, patients' most bothersome problems were morning stiffness, fatigue, non-restorative sleep, pain, and concentration/memory difficulty (Bennett 2007, PMID 17349056) — note stiffness and cognition lead, yet stiffness was relegated to "research interest" in the core set and dyscognition to "some trials" (Mease 2009, PMID 19820221), and drug approvals hang on pain-primary endpoints. Fatigue — the second-ranked patient problem — is precisely where the approved drugs fail (SNRI fatigue NNTB 18, "not clinically relevant"; Welsch 2018, PMID 29489029). This misalignment feeds the treatment-satisfaction gap documented in patient-experience-and-advocacy.

Practical measurement kit (what a new trial or clinic should collect)

Domain (OMERACT core) Instrument in common use Anchors
Pain intensity Daily diary NRS 0–10 (weekly average) ≥30%/≥50% responder; MCID conventions from IMMPACT lineage
Multidimensional function / global impact FIQR (or legacy FIQ) MCID 14% (FIQ-derived); severity bands <39/39–<59/≥59 (Bennett 2009, PMID 19369473)
Patient global PGIC "much/very much improved"
Sleep PROMIS Sleep Disturbance; daily diary sleep quality (Lederman 2026, PMID 40627411)
Fatigue PROMIS Fatigue; MFI (Clauw 2008, PMID 19108787)
Mood (some trials) HADS/BDI —
Tenderness tender point count / pressure algometry — now rarely primary (Mease 2009, PMID 19820221)

Open questions

  • Can a validated FM responder index / disease-activity score reach adoption? The OMERACT program defined candidates (Mease 2011, PMID 21724721) and Arnold's FM30 performed best (Arnold 2012, PMID 21953205), yet no index has become the accepted regulatory or clinical standard [status as of retrieved literature].
  • What drives the FM placebo response's size — and can trial design (contact time, expectation management, open-label placebo) modulate it without destroying assay sensitivity (Häuser 2012, PMID 23137770)?
  • Should fatigue or sleep be co-primary endpoints, given patient rankings (Bennett 2007, PMID 17349056) and the sleep-mechanism approval of TNX-102 SL whose PGIC effect diverged between its two phase 3 trials (Lederman 2023, PMID 37165930; Lederman 2026, PMID 40627411)?
  • How much of the historical evidence base survives re-analysis with modern imputation (multiple imputation vs LOCF) — a quantifiable but never-performed audit (Derry 2016, PMID 27684492)?
  • Do FIQR severity bands predict differential treatment response, i.e., can they operationalize stepped care (guidelines) rather than just describe populations (Bennett 2009, PMID 19369473)?
  • Are EERW designs acceptable evidence of efficacy for chronic pain regulators outside the US, given that normalization triples the NNT (Crofford 2008, PMID 18400400; Derry 2016, PMID 27684492)?

References

  1. Burckhardt CS, Clark SR, Bennett RM. The fibromyalgia impact questionnaire: development and validation. J Rheumatol. 1991;18:728-33. PMID 1865419
  2. Bennett RM, Friend R, Jones KD, Ward R, Han BK, Ross RL. The Revised Fibromyalgia Impact Questionnaire (FIQR): validation and psychometric properties. Arthritis Res Ther. 2009;11:R120. PMID 19664287
  3. Bennett RM, Bushmakin AG, Cappelleri JC, Zlateva G, Sadosky AB. Minimal clinically important difference in the fibromyalgia impact questionnaire. J Rheumatol. 2009;36:1304-11. PMID 19369473
  4. Mease P, Arnold LM, Choy EH, et al. Fibromyalgia syndrome module at OMERACT 9: domain construct. J Rheumatol. 2009;36:2318-29. PMID 19820221
  5. Mease PJ, Clauw DJ, Christensen R, et al. Toward development of a fibromyalgia responder index and disease activity score: OMERACT module update. J Rheumatol. 2011;38:1487-95. PMID 21724721
  6. Arnold LM, Williams DA, Hudson JI, et al. Development of responder definitions for fibromyalgia clinical trials. Arthritis Rheum. 2012;64:885-94. PMID 21953205
  7. Clauw DJ, Mease P, Palmer RH, Gendreau RM, Wang Y. Milnacipran for the treatment of fibromyalgia in adults: a 15-week, multicenter, randomized, double-blind, placebo-controlled, multiple-dose clinical trial. Clin Ther. 2008;30:1988-2004. PMID 19108787
  8. Häuser W, Bartram-Wunn E, Bartram C, Tölle TR. [Placebo responders in randomized controlled drug trials of fibromyalgia syndrome: systematic review and meta-analysis]. Schmerz. 2011;25:619-31. PMID 22120916
  9. Häuser W, Sarzi-Puttini P, Tölle TR, Wolfe F. Placebo and nocebo responses in randomised controlled trials of drugs applying for approval for fibromyalgia syndrome treatment: systematic review and meta-analysis. Clin Exp Rheumatol. 2012;30(6 Suppl 74):78-87. PMID 23137770
  10. Derry S, Cording M, Wiffen PJ, Law S, Phillips T, Moore RA. Pregabalin for pain in fibromyalgia in adults. Cochrane Database Syst Rev. 2016;9:CD011790. PMID 27684492
  11. Cording M, Derry S, Phillips T, Moore RA, Wiffen PJ. Milnacipran for pain in fibromyalgia in adults. Cochrane Database Syst Rev. 2015;10:CD008244. PMID 26482422
  12. Crofford LJ, Mease PJ, Simpson SL, et al. Fibromyalgia relapse evaluation and efficacy for durability of meaningful relief (FREEDOM): a 6-month, double-blind, placebo-controlled trial with pregabalin. Pain. 2008;136:419-431. PMID 18400400
  13. Crofford LJ, Rowbotham MC, Mease PJ, et al. Pregabalin for the treatment of fibromyalgia syndrome: results of a randomized, double-blind, placebo-controlled trial. Arthritis Rheum. 2005;52:1264-73. PMID 15818684
  14. Arnold LM, Rosen A, Pritchett YL, et al. A randomized, double-blind, placebo-controlled trial of duloxetine in the treatment of women with fibromyalgia with or without major depressive disorder. Pain. 2005;119:5-15. PMID 16298061
  15. Lederman S, Arnold LM, Vaughn B, Kelley M, Sullivan GM. Efficacy and safety of sublingual cyclobenzaprine for the treatment of fibromyalgia (RELIEF). Arthritis Care Res (Hoboken). 2023;75:2359-2368. PMID 37165930
  16. Lederman S, Arnold LM, Vaughn B, Engels JM, Kelley M, Sullivan GM. Pain relief by targeting nonrestorative sleep in fibromyalgia: a phase 3 randomized trial of bedtime sublingual cyclobenzaprine (RESILIENT). Pain Med. 2026;27:86-94. PMID 40627411
  17. Welsch P, Üçeyler N, Klose P, Walitt B, Häuser W. Serotonin and noradrenaline reuptake inhibitors (SNRIs) for fibromyalgia. Cochrane Database Syst Rev. 2018;2:CD010292. PMID 29489029
  18. Lunn MP, Hughes RA, Wiffen PJ. Duloxetine for treating painful neuropathy, chronic pain or fibromyalgia. Cochrane Database Syst Rev. 2014;1:CD007115. PMID 24385423
  19. Birkinshaw H, Friedrich CM, Cole P, et al. Antidepressants for pain management in adults with chronic pain: a network meta-analysis. Cochrane Database Syst Rev. 2023;5:CD014682. PMID 37160297
  20. Bennett RM, Jones J, Turk DC, Russell IJ, Matallana L. An internet survey of 2,596 people with fibromyalgia. BMC Musculoskelet Disord. 2007;8:27. PMID 17349056
  21. Russell IJ, Perkins AT, Michalek JE. Sodium oxybate relieves pain and improves function in fibromyalgia syndrome: a randomized, double-blind, placebo-controlled, multicenter clinical trial. Arthritis Rheum. 2009;60:299-309. PMID 19116896
  22. Deare JC, Zheng Z, Xue CC, et al. Acupuncture for treating fibromyalgia. Cochrane Database Syst Rev. 2013;5:CD007070. PMID 23728665