Article Text

PDF

Consensus for dry needling for plantar heel pain (plantar fasciitis): a modified Delphi study
  1. Matthew P Cotchett,
  2. Karl B Landorf,
  3. Shannon E Munteanu,
  4. Anita M Raspovic
  1. Department of Podiatry and Musculoskeletal Research Centre, La Trobe University, Melbourne, Victoria, Australia
  1. Correspondence to Matthew P Cotchett, Department of Podiatry and Musculoskeletal Research Centre, La Trobe University, Kingsbury Drive, Bundoora, Melbourne, VIC 3086, Australia; m.cotchett{at}latrobe.edu.au

Abstract

Hypothesis Plantar heel pain (plantar fasciitis) is a common and disabling condition. A variety of treatment options are available to patients with plantar heel, however the evidence for these treatments is generally weak and the best way to manage plantar heel pain remains unclear. Trigger point dry needling is increasingly used as an adjunct therapy for musculoskeletal pain. In patients with plantar heel pain this technique is thought to improve muscle activation patterns, increase joint range of motion and alleviate pain. However, there have been no randomised controlled trials that have evaluated the effectiveness of dry needling for plantar heel pain.

Methods In order to develop a treatment protocol to evaluate the effectiveness of dry needling for plantar heel pain we conducted a three stage modified Delphi process using a web-based survey technique. Over a series of three iterations, 30 experts (participants) worldwide indicated their level of agreement on specific issues relating to the use of dry needling for plantar heel pain including their treatment rationale, needling details and treatment regimen. Consensus for a dry needling protocol for plantar heel pain was achieved when >60% of participants (IQR ≤ 1.0 category on 5-point Likert scale) agreed the protocol was adequate.

Results The response rate was 75% (n=30), 100% (n=30) and 93% (n=28) in the first, second and third rounds respectively. Round 1 helped generate a list of 10 items that were deemed important for developing a dry needling protocol for plantar heel pain. These 10 items were subsequently presented in Round 2. Of these, 5 of the 10 items met the criteria to be included in a dry needling protocol for plantar heel pain that was presented in the final round, Round 3. Items that did not meet the criteria were either removed or amended and then presented in Round 3. In the final round, 93% of participants (IQR range = 1) agreed the proposed dry needling protocol for plantar heel pain was adequate. The protocol can now be used in future research projects designed to evaluate the effectiveness of dry needling for plantar heel pain.

Statistics from Altmetric.com

Background

Despite the high prevalence of plantar heel pain1,,3 and the detrimental impact it has on quality of life,4 the optimal treatment for this disorder remains unclear.5 Consequently, an alternative adjunctive treatment such as trigger point dry needling is increasingly being used by health practitioners.6

Trigger point dry needling is an invasive procedure that involves stimulation of myofascial trigger points (MTrPs) using an acupuncture needle.6

The exact mechanism of action of dry needling is still largely unknown; however, dry needling has been shown to alter the biochemical environment surrounding a MTrP.7 In addition, the elicitation of a local twitch response using dry needling has been shown to reduce spontaneous electrical activity within the MTrP region of skeletal muscle in rabbits.8

While these findings are important for understanding how dry needling works, an equally important question is whether dry needling is effective in a clinical setting. A recent meta-analysis of randomised controlled trials found that dry needling and acupuncture of MTrPs was not significantly better than placebo.9 However, the meta-analysis only included four trials and each trial scored poorly when assessed for quality.9 At this stage, therefore, the evidence for the effectiveness of dry needling for musculoskeletal pain is also uncertain.

Clearly, there are few high-quality, randomised controlled trials that have evaluated the effectiveness of this intervention. With this in mind, we recently conducted a systematic review,10 but to widen the pool of information we included both randomised controlled trials and quasi-experimental trials. Including non-randomised trials in systematic reviews can be appropriate when there are a limited number of randomised trials available.11 12 Our review was also different in that we focused on the use of dry needling specifically for plantar heel. Our search identified two quasi-experimental trials that combined dry needling with acupuncture.13 14 Both trials found a statistically significant reduction in pain but the methodological quality, as measured by the Quality Index tool,15 was poor.

Therefore, further good-quality evaluations of dry needling are required to help inform practice. Importantly, to ensure these trials have external validity, a treatment protocol is required that has a broad consensus from experts practising in the area. Treatment protocols to evaluate the effectiveness of acupuncture for tennis elbow16 and depression17 have been developed by consensus, although at present no protocol for plantar heel pain has been documented. The aim of this study was to seek information from experts who use dry needling for plantar heel pain, in order to develop a consensus-driven treatment protocol to be used in future clinical trials.

Methods

Consensus method

Approval for this study was obtained from the La Trobe University, Faculty of Health Sciences human ethics committee (reference number FHEC09/200). A modified Delphi process was used to develop a consensus for use of dry needling for plantar heel pain. This process attempts to achieve a convergence of opinion among experts on a specific topic, over a series of rounds or iterations.18 The advantage of this method is that participants remain anonymous during all rounds of the survey and can express opinions without influence of dominant characters.18

Selection of experts

Little consensus exists for the definition of an expert and also the criteria used to select experts for consensus studies.19 We selected experts based on topics that have arisen in the literature, which have defined experts based on knowledge, and the ability to influence policy.19 20

Experts with knowledge of dry needling for plantar heel pain

Therapists involved in the instruction and facilitation of dry needling courses were considered leaders in the profession with substantial knowledge of the study topic. In addition, the Delphi panellists also considered authors of peer-reviewed articles to be leaders within the profession, although there is no guideline to determine how many peer-reviewed articles need to be published before an author is considered an expert.21 Nevertheless, some researchers have used this criterion previously to select experts for consensus studies.22 For our study, we considered the chief investigators of each of these publications to be experts and invited them to participate.

Experts with the ability to influence policy

Therapists who were members of a professional association and linked to an acupuncture and dry needling interest group were invited to participate. The Delphi panellists considered these therapists to have first-hand experience of dry needling with the potential to influence clinical practice guidelines within their profession.

Methods to reduce bias in the selection of experts

To minimise bias in selection of experts and increase the external validity of the study,19 we invited a heterogeneous sample consisting of therapists from multiple countries who practised varying treatment rationales. Invited participants were also largely unknown to the Delphi panellists.23

Expert identification

Participants were identified in July 2009 using two methods. First, a nomination process was used whereby well-respected individuals within the target population were chosen.24 For our study, a physical therapist with 17 years' clinical experience, and the author of nine peer-reviewed articles in the field of myofascial pain, nominated eight therapists who she considered leaders within the profession. Second, an internet search was performed by the chief investigator to locate dry needling courses taught world wide. The email address corresponding to the course and/or the course instructor(s) was identified and saved for future correspondence. In addition, an internet search of acupuncture and dry needling special interest groups within physiotherapy, osteopathy, myotherapy, podiatry and medical associations world wide, was conducted. Nominated experts and therapists linked to dry needling courses and special interest groups were invited by email to participate.

The Delphi panellists

The survey was constructed by two physical therapists, each with more than 8 years of dry needling experience and four university-based podiatrists (the investigators on this project) with an average of 17 years of practice-related experience, including the management of plantar heel pain.

Procedure

In order to develop a dry needling protocol, with specific details of needling techniques, we used a modified Delphi process that incorporated a structured questionnaire with the opportunity for open-ended responses in round 1. This is a common method employed when there is adequate information about the issue of interest.25 Similar to a traditional Delphi process, we developed a list of items in round 2 based on round 1 responses. However, instead of asking participants to rank order or prioritise items in rounds 2 and 3 (ie, to develop consensus), we rephrased items following round 2 based on highlighted themes. This approach is often undertaken in a modified Delphi process to move towards consensus.26

Figure 1 highlights the flow of information throughout the study. The initial invitation contained a URL to the Survey Monkey website (http://www.surveymonkey.com) that included the participant information sheet, consent form and round 1 of the survey. This website was used to deliver the questionnaire in all rounds of the study. Participants were asked to complete the survey electronically within 2 weeks of receiving the email. Up to two reminders were sent to those participants who did not complete the survey in the prespecified time (2 weeks). All participants provided electronic consent to participate.

Figure 1

Flow of information through the study.

In round 1, participants answered a series of questions and provided comments relating to their use of dry needling for plantar heel pain (see online supplementary file 1). Each question related to a section of the STandards for Reporting Interventions in Controlled Trials of Acupuncture (STRICTA) guidelines.27 After completed surveys had been received by the chief investigator, the quantitative data and open-ended responses were analysed by the Delphi panel (see ‘Data analysis’ below).

In round 2, participants were presented with the results of round 1. The Delphi panel generated a list of 10 items considered a priority for development of a dry needling protocol for plantar heel pain taking into account round 1 responses of the participants (see supplementary file 2). The 10 items selected were subjectively chosen by the panel; however, the items reflected the content of the STRICTA guidelines, which are heavily weighted toward reporting of ‘needling details’ and less to ‘needling rationale’, ‘treatment regimen’ and use of ‘co-interventions’. Therefore, we asked participants to answer eight statements relating to ‘needling details’ and two to treatment regimen. No items relating to use of co-interventions were incorporated because our planned trial was not designed to include alternative therapies.

For nine of the 10 items, participants were required to rate their level of agreement on a five-point Likert scale (from 1, indicating strongly disagree to 5, indicating strongly agree). For the 10th item, participants were asked to select the optimal needle retention time in minutes when dry needling for plantar heel pain.

The findings of rounds 1 and 2 were used to form the round 3 questionnaire (see supplementary file 3). In this final round (ie, round 3), a dry needling protocol for plantar heel pain was proposed. The protocol was presented in line with the STRICTA guidelines and included information relating to treatment rationale, needling details and treatment regimen. Participants were asked to indicate their level of agreement with the proposed protocol.

Data analysis

In round 1, the Delphi panel analysed quantitative and qualitative data (ie, from the open-ended responses). Both forms of data were exported to a Microsoft Excel spreadsheet and placed into a category linked to the STRICTA guidelines. The categories included:

  1. Treatment rationale;

  2. Muscles dry needled;

  3. Depth of needle insertion;

  4. Number of needles inserted;

  5. Needle retention time;

  6. Manual needle stimulation;

  7. Needle response sought;

  8. Treatment regimen.

For each category, the Delphi panel performed a thematic analysis28 to identify and list key themes from open-ended responses. In addition, data from the structured questions were analysed for the percentage agreement for each statement. Items with >60% support were considered to have an acceptable level of agreement. The use of percentage measures, to represent the collective responses of participants, is also common in the Delphi process,21 although there is no consensus on what percentage of participant responses constitutes an acceptable level of agreement.

The combination of themes originating from open-ended responses and analysis of quantitative data from the structured questions helped form the round 2 questionnaire, which was linked to the STRICTA guidelines. Items from round 1 that displayed >60% agreement were included in the questionnaire while items with <60% support were rephrased to reflect what the majority of respondents indicated in round 2. The questionnaire consisted of 10 items and was phrased using wording that was commonly used in open-ended responses made by participants.

The level of consensus for items 1–9 was determined by calculating the median, IQR and percentage agreement. Subsequently, each item was rated according to its level of consensus. Items with an IQR≤1.0 and >60% of votes within the ‘agree’ and ‘strongly agree’ categories were considered to have an adequate level of consensus while items with an IQR>1.0 were considered to have a low consensus.29,,31 Item 10, which asked participants to discuss needle retention time, was not analysed using an IQR because an answer to the question could not be measured on a Likert scale (ie, an ordinal scale). Instead, the Delphi panellists analysed the percentage agreement in five time intervals (see supplementary file 2).

At the end of round 2 consensus ratings and open-ended responses were exported to a Microsoft Excel spreadsheet and placed into a category linked to the STRICTA guidelines, similar to round 1. A thematic analysis was undertaken to highlight the majority theme. Items that displayed an adequate level of consensus (IQR≤1.0 and >60% of votes within the ‘agree’ and ‘strongly agree’ categories) were considered an important component of a dry needling treatment for plantar heel pain and were subsequently placed in a protocol presented to participants in round 3. The Delphi panel discussed items that displayed a low level of consensus and a decision was made to rephrase or remove each item so that a move towards consensus could be achieved. The rephrased item was then included in a protocol that was presented in round 3. The procedure to rephrase items between rounds is a process not uncommon to a Delphi process and has been used in a study that explored the use of specific statistical tests to measure consensus among a group of participants.26

The protocol presented in round 3, consisted of 10 items. Each item was based on sections outlined in the STRICTA guidelines. Consensus for the proposed dry needling protocol for plantar heel pain was achieved if >60% of participants (IQR≤1) agreed that the protocol was adequate for a clinical trial to evaluate the effectiveness of dry needling for plantar heel pain.

Results

Round 1

The response rate was 75% (n=30). Of the 75% who did respond, the chief investigator knew four participants personally and the other three members of the Delphi panel knew none. Of the 30 participants, the majority were from Australia 53% (n=16) while nine other countries were represented including the UK 10% (n=3), USA 10% (n=3), Spain 7% (n=2), Belgium 3% (n=1), Ireland 3% (n=1), Japan 3% (n=1), The Netherlands 3% (n=1), New Zealand 3% (n=1) and South Korea 3% (n=1). An invitation was also sent to therapists in Brazil, Canada, Denmark, Germany and Turkey but none responded.

The results of round 1 are presented in table 1. In regards to treatment rationale, 93% (n=28) of participants practised dry needling according to the MTrP model. However, 33% (n=10) of participants practised MTrP dry needling in combination with the radiculopathy model; 7% (n=2) with traditional Chinese acupuncture; 3% (n=1) with Western medical acupuncture; 3% (n=1) with the Baldry technique and 3% (n=1) with a ‘layering approach’. Two participants did not employ the MTrP model but rather applied a ‘neurophysiological’ approach.

Table 1

Round 1 findings*: a breakdown of participants' responses to specific details relating to dry needling for plantar heel pain (N=30)

When questioned about needling details there was substantial variability among participants for (1) muscles dry needled (a total of 35 different muscles (mean=7, SD=6) were ‘usually’ dry needled for plantar heel pain); (2) optimal needle retention time (range <1 min to 30+ min); (3) number of needle insertions per muscle (range 1–10) and (4) number of treatment sessions to manage a patient with plantar heel pain (range 3–8). In contrast, responses were more consistent for depth of needle insertion; use of manual needle stimulation; type of response elicited and frequency of treatment following the first consultation.

Based on the results of round 1, participants in round 2 were informed that the MTrP model would be adopted for use during subsequent development of the dry needling treatment protocol as 93% of participants used this rationale. In addition, as the range of muscles dry needled by participants was large (35 in total) it would not be possible to effectively dry needle 35 different muscles in a standard consultation of 20–30 min. As the aim of the study was to develop a treatment protocol that could be pragmatically applied in a clinical setting, the Delphi panellists presented two items in round 2 (items 1 and 2) to determine if participants would prefer a trial that was (1) standardised,32 with a fixed number of muscles to be dry needled; (2) semistandardised,33 which would include a set number of muscles and an additional prespecified list of muscles that could be assessed based on the participant's presentation or (3) pragmatic,34 which would involve dry needling any muscle that was clinically relevant. Participants were also given the opportunity to respond to these items in an open-ended manner.

Round 2

All 30 participants (100% response rate) completed the survey in round 2. Five of nine items (item numbers 1, 3, 4, 5 and 7) met the criteria to be included in the dry needling protocol that was to be proposed in round 3 (table 2).

Table 2

Round 2 findings: participants' responses to 10 items that were formulated, based on the results of round 1 (N=30)

Substantial variability was evident for the optimal needle retention time. Based on the findings in round 2, a dry needling protocol for plantar heel pain was presented in round 3 rather than a third structured survey. Items that did not meet the criteria for inclusion in the protocol proposed in round 2, were either removed or amended and then presented in the final round, round 3. Table 3 provides an explanation for amendment of items that displayed substantial variability in responses.

Table 3

Explanation for amending items that did not meet consensus criteria in round 2 (amended for round 3)

Round 3

Twenty-eight participants (93% response rate) completed the survey in round 3. One participant failed to respond and the Delphi panellists excluded another because the participant did not routinely use dry needling as a first-line management for plantar heel pain. The median level of agreement was 4 on a five-point Likert scale (IQR=1). Ninety-six per cent of respondents either strongly agreed (29%, n=8) or agreed (68%, n=19)with the proposed protocol, which was developed with a clinical trial to evaluate the effectiveness of dry needling for plantar heel pain in Mind. Only 4% (n=1) disagreed with the proposed protocol. Table 4 provides a detailed description of the dry needling treatment protocol based on the results of the previous two rounds. After completion of round 3, the Delphi panel decided to exclude the posterior tibial muscle as a structure to be dry needled owing to a recent publication that highlighted the hazardous nature of needling the this muscle without ultrasound guidance.35

Table 4

Dry needling protocol for plantar heel pain that was presented in the final round, round 3

Discussion

In preparation for future clinical trials to evaluate the effectiveness of dry needling for plantar heel pain, we conducted a modified Delphi process to develop and obtain consensus for a dry needling protocol. Experts in the use of dry needling for plantar heel pain, from varying allied health and medical backgrounds, participated in this project by indicating their level of agreement about specific dry needling topics. Questions put to the participants were in accordance with the STRICTA recommendations.27

After a series of three rounds, 93% of the experts surveyed agreed that the dry needling protocol proposed would be adequate for a clinical trial to evaluate the effectiveness of dry needling for plantar heel pain. As far as we know, this is the first study to have established a protocol for the use of dry needling for plantar heel pain. The protocol provides a detailed outline of a dry needling treatment, including treatment rationale; muscles to be assessed; type of acupuncture needle used; depth of needle insertion; needle response elicited; use of manual needle stimulation; needle retention time; frequency and total duration of treatment.

Although not a limitation, the final protocol established by consensus underwent one minor modification after round 3 without approval from the Delphi participants. The Delphi panellists removed the posterior tibial muscle as a structure that might be assessed and if appropriate, dry needled. This was in response to a recent study recommending that needle insertion into the tibialis posterior should only be undertaken using ultrasound guidance owing to close proximity of neurovascular bundles.35

The study also needs to be viewed in light of some limitations. First, the criteria used to select experts might not have adequately identified participants with sufficient clinical experience. The criteria set were based on themes that are commonly used to define experts.19 Although we believe that the criteria were sufficient to identify experts, we did not set a criterion for the minimum number of years of clinical and/or research experience required for inclusion in the study. While 63% of participants had practised dry needling for >5 years, 37% of participants had practised dry needling for less than this. It might be argued that practitioners with <5 years of dry needling experience might not be experts, but we believe it would be unreasonable to suggest that a set number of years are required before a therapist develops expertise. In addition, it was difficult to locate a sufficient number of therapists who had practised dry needling for a long period of time. As dry needling was only popularised in the late 1970s,36 construction of a list of therapists with extensive experience with dry needling will take time.

Second, we might have considered a criterion based on how frequently participants dry needled for plantar heel pain. While 73% (n=22) of participants indicated that they ‘usually’ or ‘always’ used dry needling for plantar heel pain, 23% (n=7) of participants were neutral and one participant did not use it as a first-line treatment for plantar heel pain. Despite these results, we do not believe this affected the outcome because participants might still demonstrate expertise when they decide to use dry needling for plantar heel pain. Nevertheless, future studies may choose to consider this subject when developing inclusion criteria for selection of experts.

Third, 47% (n=14) of participants were physiotherapists and the remainder comprised medical practitioners, podiatrists, myotherapists, osteopaths and researchers. A limitation of this selection might be that nearly half of participants were physiotherapists, although we believe this reflects the increased use of dry needling among physiotherapists compared with other professions.6 However, we recognise that there are no good-quality data to support this assertion.

Fourth, the survey results may not entirely reflect the views of experts world wide. Fifty-three per cent (n=16) of participants were from Australia while nine other countries represented the remaining 14 participants. It might have been more appropriate and valid had we invited an even number of experts from multiple countries to assist in developing consensus. For example, if we had a greater percentage of experts from Asia in our sample it might have disclosed increased use of the traditional Chinese medicine model for the treatment of plantar heel pain. Further, had experts from Canada responded to the invitation, the radiculopathy model might have been favoured as this model originated in Vancouver.

An additional limitation of our study might be the absence of a second structured survey in round 3. This process involves administering the same survey from round 2 in round 3, which allows participants to alter their response in view of group findings and enables stability of participants' responses to be assessed.26 However, because a high level of agreement had been achieved in round 2 for the majority of key variables, we chose to present a treatment protocol in round 3 that reflected the findings of the previous two rounds. The rationale for this approach was supported by the fact that only one person disagreed with the proposed protocol in round 3. If there had been substantial disparity (ie, IQR>1.0), it would have been shown by greater disagreement at this stage.

Summary points

  • We plan a trial of dry needling acupuncture for plantar fasciitis

  • We conducted a modified Delphi study to establish a consensus for treatment

  • Three rounds of consultation generated a protocol with 93% agreement

Conclusion

In preparation for future clinical trials, a modified Delphi process was used to develop and obtain consensus for a dry needling protocol for plantar heel pain. The vast majority of participants agreed that the proposed protocol, including treatment rationale, needling details, frequency and total duration of treatment was adequate for use in a clinical trial to evaluate the effectiveness of dry needling for plantar heel pain. The protocol can now be used in future research projects designed to evaluate the effectiveness of dry needling for plantar heel pain.

References

View Abstract

Footnotes

  • Competing interests None.

  • Ethics approval This study was conducted with the approval of the La Trobe University human ethics committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.