On “Effectiveness of trigger point dry needling…” Cotchett MP, Munteanu SE, Landorf KB. Phys Ther. 2014;94:1083–1094.

It was with interest and some concern that we read the report by Cotchett et al¹ published in the August 2014 issue of PTJ. Although the authors reported statistically significant differences in first-step pain and foot pain in favor of trigger point dry needling over sham dry needling, it appears that the actual palpatory methods used by Cotchett et al¹ to identify the location of the target trigger points and, therefore, the entry point, angulation, and depth of needle insertion have not yet been found to possess accurate diagnostic validity or acceptable intra- or inter-examiner reliability for muscles in the foot or lower leg. Therefore, the results of the study by Cotchett et al,¹ including the reported frequency counts of myofascial trigger points in specific foot intrinsic and lower leg muscles, should be questioned, or at least viewed cautiously. There are several original trials, literature reviews, and meta-analyses that support our contention on this issue.

In a recent systematic review, Tough et al concluded, “There is a lack of robust empirical evidence validating the clinical diagnostic criteria [for trigger point identification or diagnosis] proposed by both Travell & Simons (1999) and Fischer (1997).”² In another systematic review on the reliability of physical examination for the diagnosis of myofascial trigger points, Lucas et al concluded, “There is no accepted reference standard for the diagnosis of trigger points, and data on the reliability of physical examination for trigger points is conflicting.”³ In addition, a predictable pattern of pain referral and the local twitch response are each no longer considered sufficient or necessary for diagnosing trigger points.^2–4 Yet, regardless of the existing evidence, Cotchett et al¹ still decided to use “a characteristic pattern of referred pain” and “a local twitch response” as 2 of the 4 criteria for the identification and diagnosis of myofascial trigger points in the foot and lower leg.

We would like to encourage Cotchett et al¹ to provide evidence from an original, peer-reviewed experimental study, not a textbook written by Travell and Simons,^5,6 that supports the diagnostic accuracy and reliability for the use of a “characteristic pattern of referred pain” or the presence of “a local twitch response” in making the diagnosis of a myofascial trigger point within the quadratus plantae, flexor digitorum brevis, or other intrinsic muscles of the foot. To our knowledge, such a study does not exist, and, according to Myburgh et al,⁴ the gluteus medius (kappa range=.29 to .49) and the quadratus lumborum (kappa range=.36 to .50) are the only muscles in the human body with distinct and predictable pain referral patterns that have acceptable reproducibility on manual examination. Furthermore, in a systematic review on the reliability of physical examination for the diagnosis of myofascial trigger points, Lucas et al³ found an unacceptable interrater reliability range (kappa range=−.05 to .57) for the local twitch response. Therefore, the internal validity of the Cotchett et al¹ study appears to be jeopardized.

After reviewing 9 studies on reliability, Lucas et al concluded, “At present, there is no data on the reliability of pinpointing the exact location of active trigger points…. The existing data on reliability pertain only to agreeing if a muscle has the signs of a trigger point and not the exact location of the taut band or the nodule within the taut band.”³ Lew et al⁷ reported inter-examiner agreement of only 21%, and Sciotti et al⁸ reported error rates of 3.3 to 6.6 cm among examiners attempting to identify the specific location of trigger points in the upper trapezius muscle. In another recent literature review, Myburgh et al⁴ found poor inter-examiner reliability of manual palpation of trigger points in various muscle groups. Only “tenderness” of the upper trapezius muscle, not the actual location of the trigger point, was found to be moderately reliable. More recently, Myburgh et al⁹ similarly reported “good agreement between experienced practitioners” for the “presence or absence” of a clinically relevant trigger point in the upper trapezius muscle; however, like the many other studies,^10–14 Myburgh et al⁹ failed to investigate the inter-rater reliability for determining the specific location of the trigger point within the target muscle.

According to Lucas et al, “It is not yet evident that examiners can agree on the precise location of an active TrP [trigger point]; hence, they cannot be relied upon to accurately insert the needle into the nodule of the taut band.”³ If clinicians are not able to reliably identify the exact location of trigger points,^3,4,9 we are interested to know how Cotchett et al¹ were able to accurately locate and verify needle placement in myofascial trigger points in muscles such as the quadratus plantae or the flexor digitorum brevis. Given that high-quality evidence suggests that manual examination for identifying the specific location of the myofascial trigger point is not a valid^2,4,9 or reliable^3,4,7,8 process, we encourage the authors to disclose how they determined the frequency counts for the presence of myofascial trigger points within specific muscles of the foot and lower leg (eg, 132 myofascial trigger points for the quadratus plantae muscle and 92 myofascial trigger points for the flexor digitorum brevis muscle in the real dry needling group). Furthermore, in the absence of guided imagery technology, we are curious how Cotchett et al¹ were able to differentiate needle placement for the quadratus plantae muscle (within the second layer of the plantar foot) and the flexor digitorum brevis muscle (within the first layer of the plantar foot).

Although the authors cited a previous “consensus” study¹⁵ that Cotchett and colleagues authored in an attempt to justify leaving the needles in situ for a 5-minute duration after repeated pistoning (“withdrawn partially and advanced repeatedly”), it should be noted that 93% (28 of 30) of the participants on the Delphi panel practiced dry needling according to the Travell and Simons^5,6 myofascial trigger point model, with just 7% (2 of 30) of panel participants having previous training or practice patterns in Western medical or traditional Chinese acupuncture.¹⁵ Furthermore, none of the 3 observational studies^16–18 cited by Cotchett and colleagues^15,19 actually left needles in situ for just 5-minute durations. Aside from one cohort study that treated plantar heel pain with a single lidocaine injection in which the needle was immediately removed upon dispensing the injectate,¹⁶ one case series used manual acupuncture and left the needles in place for 15 minutes,¹⁸ and the other case series used electroacu-puncture and left the needles in place for 20 minutes.¹⁷ Additionally, a recent randomized controlled trial of patients with chronic plantar fasciitis reported a 69% reduction in foot pain and an 80% success rate following 5 weeks of electroacupuncture where the needles were left in place for a duration of 30 minutes.²⁰

In short, no previous case series,^17,18 cohort study,¹⁶ or randomized trial²⁰ has left the needles in place for just 5 minutes when treating patients with chronic plantar fasciitis; thus, we were surprised that Cotchett et al¹ chose to limit the dosage to just 5 minutes in their recent trial on the use of dry needling for plantar heel pain. The authors' justification should be based on more than a “consensus”¹⁵ opinion, comprising 28 of 30 panel members who were trained in dry needling according to the narrowly focused²¹ muscular “trigger point only” Travell and Simons^5,6 model—that would seem a little biased.

Notably, there is robust evidence from a recent meta-analysis²² that placement of needles in distal points of the upper and lower extremities activates specific brain areas such as the primary and secondary somatosensory cortex, insula, anterior cingulate cortex, and thalamus that are responsible for the sensory discrimination aspects of pain, commonly referred to as “deqi” (ie, soreness, aching, dull pain). According to functional magnetic resonance imaging and positron emission tomography studies,^23–29 dry needling also has been shown to activate areas of the brain and brain stem associated with descending pain inhibition such as the periaqueductal gray area, raphe nuclei, and locus coeruleus^30,31 while deactivating the limbic and paralimbic areas that modulate the cognitive and affective dimensions of pain.²² In the main, needles inserted into the body for the “central effects” on pain have typically been left in situ for between 8 and 25 minutes.^23–29 In this light, we find it odd that Cotchett et al¹ chose to confine the duration of their dry needling treatment to just 5 minutes.

Although the etiology of plantar heel pain remains unknown, it seems to be well accepted in the existing literature that the proximal attachment of the plantar aponeurosis, the medial tubercle of the calcaneus, is most often reported by patients with plantar fasciitis as the origin of symptoms and the site of greatest discomfort.^32–34 Moreover, although the plantar pain may initially present as diffuse, it usually localizes to the area of the medial calcaneal tuberosity with time.³³ Considering that 3 of the 4 previous studies on dry needling in patients with chronic plantar heel pain specifically targeted the insertion of the plantar fascia at or near the medial tubercle of the calcaneus,^16,17,20 we find it baffling why Cotchett et al¹ chose to omit needling of this key anatomic feature, instead focusing solely on muscular “trigger points” using manual palpation techniques that have yet to be validated or found to be reliable in their ability to locate such. Had Cotchett et al¹ targeted the insertion of the plantar fascia at or near the medial tubercle of the calcaneus^16,17,20 and left the needles in place for 15 minutes,¹⁸ 20 minutes,^17,35 or 30 minutes²⁰ as previous studies have done in patients with chronic plantar heel pain, perhaps the between-group difference in first-step pain (ie, the primary outcome measure at the primary end point of 6 weeks) of just −14.4 mm (visual analog scale, 0–100) would have shown a larger effect size and exceeded the minimum clinically important difference for that measure.^36–38

Thank you for the opportunity to comment on this most interesting paper. We appreciate the opportunity for a dialogue with the authors and believe that further discussion and debate can ultimately enhance both clinical practice and clinical research.

• © 2014 American Physical Therapy Association