Criteria used for the diagnosis of myofascial trigger points in clinical trials on physical therapy: updated systematic review

Key points

1. 01Spot tenderness was reported in 96.9% of the 129 studies that specified diagnostic criteria, making it the most common single criterion.
2. 02Referred pain was used in 73.6% of those studies, and local twitch response in 48.8%.
3. 03The most frequently used combination was spot tenderness plus referred pain plus local twitch response, applied in 28 studies (22%).
4. 0428.3% of included studies failed to report any diagnostic method, and 6.6% used expert-based definitions without specification.
5. 05Only 9 of 198 studies (4.5%) clearly distinguished between active and latent myofascial trigger points.

How it was conducted

Design

Updated systematic review (PRISMA-P guidelines), registered in PROSPERO (CRD42018087420)

Databases

MEDLINE and Physiotherapy Evidence Database (PEDro)

Publication period

January 1, 2007 to April 10, 2019

Included study types

Randomized controlled trials, nonrandomized trials, quasi-experiments, pre-post, cross-over, and parallel study designs

Inclusion criteria

Physical therapy clinical trials involving patients with musculoskeletal pain and at least one active or latent myofascial trigger point in any body area

Quality assessment

PEDro scale (0-10); median score of included studies was 5/10, with 65% scoring 6 or higher

What they found

• Of 478 records identified after deduplication, 198 (41%) met inclusion criteria.
• 129 of 198 studies (65.1%) specified diagnostic criteria for myofascial trigger points in the main text; 56 studies (28.3%) reported no diagnostic method; 13 studies (6.6%) used expert-based definitions without specification.
• Among the 129 studies reporting criteria, spot tenderness was used in 125 studies (96.9%), referred pain in 95 (73.6%), local twitch response in 63 (48.8%), pain recognition in 59 (45.7%), limited range of motion in 29 (22.5%), and jump sign in 10 (7.8%).
• 23 distinct combinations of the 6 most common criteria were identified.
• The most common combination was spot tenderness plus referred pain plus local twitch response, used in 28 studies (22%); this trio appeared in 56 studies (43%) when all overlapping combinations were counted.
• The second most common combination was spot tenderness plus referred pain alone, used in 17 studies (13%); this pair appeared in 93 studies (72%) in total.
• 12 studies used only a single criterion to diagnose myofascial trigger points: 9 used only spot tenderness, 2 only limited range of motion, and 1 only referred pain.
• Only 9 studies (4.5%) made a clear distinction between active and latent myofascial trigger points.
• PEDro scores ranged from 1 to 9 with a median of 5/10; 65% of studies scored 6 to 9.
• Cervicothoracic region was the target in 139 studies (70%); cranio-mandibular in 24 (12%); lower extremity in 15 (8%); pelvic girdle in 14 (7%); upper extremity in 11 (6%).

Limitations

• Only physiotherapy clinical trials were included, limiting generalizability to invasive or pharmacological treatment contexts.
• Approximately 35% of included studies either did not report any diagnostic procedure or used vague expert-based definitions, so findings may not extend to those studies.
• The review assessed frequency of criteria use, not the validity or reliability of the criteria themselves.
• Only English-language publications were included, which may introduce language bias.

Why it matters

For patients

Patients being diagnosed with myofascial trigger points should know there is no universally agreed diagnostic standard, meaning diagnosis can vary considerably between clinicians and studies.

For clinicians

Clinicians should document each specific diagnostic criterion used when reporting or publishing myofascial trigger point research, and clearly distinguish between active and latent trigger points to improve reproducibility.

For readers

This review shows that diagnostic inconsistency is a major barrier to comparing myofascial trigger point research, and that consensus on a standardized criterion cluster is still needed.