State, Not Scar · source record · paper

Population study of tender point counts and pain as evidence of fibromyalgia

P. Croft; J. Schollum; A. Silman · 1994

AI-verified citationSRC-CROFT-1994

identifier / locator: doi:10.1136/bmj.309.6956.696
url: https://pmc.ncbi.nlm.nih.gov/articles/PMC2540791/
verification note: text retrieved and shown to the drafter on 2026-09-20 (run 2026-09-20-draft-vasocomputation-113013); passages checked by the verify verb
reliability notes: Peer-reviewed primary (BMJ 309:696–699, 17 Sep 1994); only the structured abstract was retrieved at PMC2540791 (full text is a scanned PDF not shown).
Two-stage cross-sectional design in two general practices in north-west England; a stratified random sample answered a pain questionnaire, and 250 responders were selected for tender-point examination on the basis of their reported pain complaints, of whom 177 participated. The examined subsample is therefore pain-selected, not an unselected community sample.
Outcome is the 18-site tender-point count (0–18) of the 1990 ACR fibromyalgia criteria; tissue damage is not assessed or mentioned in the abstract.
Arthritis and Rheumatism Council Epidemiology Research Unit, University of Manchester; independent of the trigger-point lineages in the ledger.

Evidence drawn from this source

Supports· moderateVASO-E062

what the source states: The source is a two-stage cross-sectional study set in "Two general practices in north west England", drawing a stratified random sample of adults from age–sex registers, in which 177 selected responders underwent tender-point examination with the outcome "Tender point count (0 to 18) grouped into four categories", the highest (11 or more) corresponding to the ACR fibromyalgia criteria.
what we infer: This is the primary the founding essay gestured at: the fibromyalgia-era 18-site exam administered outside the clinic. Note that the examined subsample was chosen on reported pain complaints, so it is a community-based but pain-selected sample.
limitations: Abstract only retrieved; full text not shown.
Examined subsample (177 of 250 selected) was chosen on the basis of reported pain complaints.

Context· weakVASO-E082

what the source states: Two people with tender point counts of 11 or more were in the group reporting no pain: "Two people with counts of 11 or more were in the group reporting no pain."
what we infer: Examiner-elicited tender points occur in people with no pain complaint at all, which is the population the case's opening observation describes. Weak because it is two individuals, the sample was pain-selected, and the 18-site protocol tests fixed anatomical sites rather than counting palpable knots.
limitations: Two individuals; pain-selected subsample.
Fixed-site tenderness testing is not a census of palpable knots.
Direction set to "context" (from "supports") by the second reader (claude-sonnet-5, 2026-09-21, run 2026-09-21-answer-vasocomputation-012456, verify-v7) at intake: The quote is accurate and correctly located in the abstract's Results section, and the sourceStatement fairly paraphrases it. But the observation—two pain-free individuals in a pain-selected subsample having ≥11 tender points—does not establish or support that tender points 'commonly' occur across the general population; it merely notes an anomaly within a small, non-representative subgroup, so it bears on the claim only as context, not as support. The record appears to be a legitimate part of a split evidence record, stating one discrete finding from the original compound without importing unrelated content.