Skip to content

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

One-paragraph summary

Cochrane review of eight randomized, double-blind trials of pregabalin in adult fibromyalgia: five classic-design parallel trials (3,283 participants; 150–600 mg/day; 8–13 weeks) and two enriched-enrolment randomized-withdrawal (EERW) trials (1,492 titrated, 687 randomized; 13–26 weeks), plus one nightly-vs-twice-daily study (n=177). At 300–600 mg/day, substantial benefit (≥50% pain reduction) occurred in 22–24% of patients versus ~14% on placebo (450 mg: RR 1.8, 95% CI 1.4–2.1), and moderate benefit (≥30%) in 39–43% versus ~28% — high-quality evidence that about 1 patient in 10 gains substantial, drug-attributable benefit, with NNTs of 7–14 across outcomes. Harms were common and quantified: NNH 3.7 for dizziness, 7.4 for somnolence, 18 for weight gain, 19 for peripheral edema; adverse-event withdrawals ran ~10% above placebo while lack-of-efficacy withdrawals ran ~6% below.

Key findings

  • ≥50% pain relief: placebo ~14%; pregabalin 300–600 mg 22–24% (high-quality evidence); 450 mg RR 1.8 (1.4–2.1).
  • ≥30% pain relief: placebo ~28%; pregabalin 39–43%; 450 mg RR 1.5 (1.3–1.7); NNTs 7–14 across these and PGIC outcomes.
  • EERW trials: maintained therapeutic response 40% (pregabalin) vs 20% (placebo), NNT 5 — but normalized to the population entering titration, NNT ≈ 12, and only ~10% of starters achieve durable maintained response.
  • 70–90% of participants in all arms experienced adverse events; serious AEs did not differ (very low quality).
  • Explicit methods warning: LOCF imputation in the classic trials "could overestimate treatment effect."

Limitations

  • No active comparators anywhere in the program — efficacy is only ever relative to placebo.
  • 8–26 week horizons; nothing on long-term effectiveness, tolerance, or discontinuation effects.
  • Trial populations (ACR 1990-era, largely white women, severe baseline pain ~7/10) limit generalizability to criteria-set-2016 clinic populations.

Why it matters

The definitive quantification of what an FDA-approved FM drug actually delivers: a ~10% absolute increase in substantial responders, purchased with near-universal adverse-event exposure. Its side-by-side presentation of classic and EERW designs — and the tripling of the NNT when EERW results are normalized — is the clearest teaching example of why FM trial designs cannot be naively compared, which is exactly how it is used in outcomes-and-measurement.md. Together with the parallel SNRI reviews it underlies the "minority benefit, majority don't" framing now standard in guidelines.

Cited by wiki pages

  • pharmacologic-therapy
  • outcomes-and-measurement