Massage Therapy for Phantom Limb Pain: Feasibility and Pilot Study

Massage therapy is a treatment self-reported by those with phantom limb pain (PLP) as moderately to extremely effective, but no research to date has been specifically designed to examine such efficacy or effectiveness in this regard. The investigators propose that massage therapy can affect PLP by applying treatment to the corresponding intact limb for amputees. Theoretical underpinnings of massage and mirror therapy have guided this study's development to address PLP through the development of a novel massage intervention. The proof-of-concept experimental study examines the feasibility and outcomes of a combined massage and mirror therapy treatment for individuals with lower limb phantom limb pain. Phantom pain severity, intensity, and interference were measured with visual analogue scales (collected twice weekly for the study's duration) and the Brief Pain Index (collected 4 total times across 12 weeks) using a single-subject, withdrawal design with three, 4-week study phases. Intervention consisted of twice weekly, 20-minute treatments for four weeks. Tailored treatments addressed individual phantom pain experience as if occurring on the corresponding, non-removed limb. Participants viewed treatment as though it were being applied to the amputated limb. Specific Aim: Determine the extent to which massage therapy applied in combination with mirror therapy to the intact leg of an amputee addresses PLP severity, intensity, and interference. Hypothesis: massage and mirror treatment session(s) focused on an intact limb will have acute and cumulative benefits (more pronounced and lasting) for PLP severity, intensity, and interference. All study activity to occur on Indiana University's IUPUI campus.

Potential participants self-identify themselves to study personnel by contacting the PI via office phone or IU email. Potential participants were informed of the research study through word-of-mouth and fliers at pain and prosthetic clinics and amputee support groups in the Indianapolis area which will invite interested individuals to contact the PI for more information and a telephone screening for eligibility. If an interested individual screens eligible and is interested after hearing more about the study, an appointment for an Informed Consent Visit will be made. All subjects who enroll in the study will receive the intervention. There is no randomization. Study Procedures Subjects were initially informed of what the study entails upon completion of the phone screening for eligibility and their indication of interest. After the study details were described, the potential subject was invited to schedule a time to come to campus for a formal informed consent and familiarization session. Time between the phone discussion and the scheduled appointment could range from hours to days. The actual environment for collecting informed consent was in the PI's lab were all study activity took place. The Informed Consent Document was reviewed in its entirety and potential subjects given the opportunity to ask any additional questions. After all questions were answered, the potential subject was invited to provide their informed consent by signing the IC document. After consent was obtained, the familiarization session and collection of baseline measures commenced. The proof-of-concept experimental study examines the feasibility and outcomes of a combined massage and mirror therapy treatment for individuals with lower limb phantom limb pain. Phantom pain severity, intensity, and interference, were measured with visual analogue scales (collected twice weekly for the study's duration) and the Brief Pain Index (collected 4 total times across 12 weeks) using a single-subject, withdrawal design with three, 4-week study phases; also referred to as an ABA withdrawal design. Study intervention consisted of twice weekly, 20-minute treatments for four weeks and is delivered during the second 4-week study phase. Tailored treatments addressed individual phantom pain experience as if occurring on the corresponding, non-removed limb. Participants viewed treatment as though it were being applied to the amputated limb. Study participation was approximately 3 months long for each subject and involved 11 on-site (IUPUI campus) study visits. All information and data gathered from subjects were self-report and in the form of pencil and paper surveys. Data for each subject was collected at baseline and twice weekly, 2-5 days apart, over a 12-week time period. For data collection during off-site collection points (first and final 4-week study phases), participants were prompted to complete the measures (provided in a binder at the start of the study) via a phone call, text message, or email reminder from the PI or research assistant. Measures completed off-site during the first 4 week study phase was collected by the PI or research assistant at the study visit appointment which also began the intervention study phase. During the second 4-week study phase, all data was collected on site just prior to and after the massage intervention. Measures completed off-site during the final 4-week study phase (withdrawal phase) was collected by the PI or research assistant at the final on-site, data collection study visit. Payments in the form of $20 gift cards were made to participants at six points totaling $120 if all aspects of the study were completed. Gift cards were distributed at the beginning and completion of each study phase and at the 4th and 8th intervention visit. All 8 massage treatments will be administered by a trained massage therapy provider in line with Indiana practice specifications. Subjects will not disrobe and will be asked to wear or bring shorts for treatments. Massage providers will be independent contractors hired and trained specifically for this purpose. Based on the PLP experienced by the subject as described in a pretreatment interview, the massage provider will develop and apply a tailored treatment to address the specific PLP experience as if the subject experienced the pain in the intact limb. Massage treatments will be administered in 20-25 minute timeframes on the intact limb that corresponds with the amputated limb. The massage provider will document the treatment in the form of SOAP (treatment) notes. During the application of each massage treatment, a digital camera is positioned to view, stream, and record the applied treatment on the intact limb. The only images captured by the camera will be the bare intact limb, the therapist hands, portions of the therapist's body (dressed in black), and minimal surrounding environment which will be draped in black cloth for visual neutralization. The resulting effect: Subjects will see on the monitor, a mirror image of their intact limb being massaged in the place of their amputated limb. Subjects will be asked to watch the monitor during the treatment. Considering the exploratory nature and small N of the study, descriptive statistics will primarily be used to consider and report results. Of particular interest is whether or not and the extent to which massage therapy that incorporates principles from mirror therapy affects PLP when applied to the intact limb. For the primary outcome variables (VAS 1-3), an average of each VAS measured in the first A phase will serve as baseline for each subject. Acute effects will be quantified by considering the change score (Δ) from baseline VAS measures to the VAS measures collected just before the 2nd massage treatment application (after one massage treatment). Cumulative effects will be quantified by considering the Δ from baseline VAS measures to the VAS measures collected at the first collection point in the second A phase of the study (after all massage treatments).

Repeated measures, single-subject withdrawal design with three, 4-week study phases; also referred to as an ABA withdrawal design. Study intervention consisted of twice weekly, 20-minute treatments for four weeks (8 total treatments) and is delivered during the second 4-week study phase. All subjects who enroll in the study will receive the intervention during the second 4-week study phase. There is no randomization.

Each study participant will progress through the three, 4-week study periods in the ABA withdrawal design in the same, designated order. The first and third 4-week study periods (or the A periods) have no intervention and only consist of twice weekly data collection. The second 4-week study period (or the B period) will include the twice weekly delivered massage therapy combined with components of mirror therapy intervention.

Individualized 20-minute, Swedish massage to address participant's unique phantom limb pain experience as if occurring on the corresponding, non-amputated limb. A specially designed apparatus is positioned such that a fixed, real-time image in place of the amputated limb is viewed by the participant during the massage session. The apparatus has a mirror, masking material to hide therapist and participant's lower body, electronic video display, electronic video camera, and equipment to electronically communicate the video signal to the video display. The mirror and video camera positioned to "see" a reflection of the existing limb being massage feds the image to a live video display situated in front of the participant to the correct orientation naturally represent the missing limb.

PLP severity is measured via visual analogue scales (VAS) of current, average, and worst PLP over the past 2-3 days. Each VAS is a 10cm line with the 10cm end-point indicating "worst pain imaginable" and the 0cm end-point indicating "no pain"

VAS scores were collected twice weekly (25 times total) throughout the three month study period. Participants were asked to consider the last 2-3 days for the "average" and "worst" VAS and in the moment for the "current" VAS

PLP intensity and interference were measured with the Brief Pain Inventory (BPI) which has been used in other studies examining PLP, consists of eleven, 11-point questions (0 indicates "no pain" while 10 indicates "worst pain imaginable"), and intensity and interference subscales. The BPI intensity subscale has four questions about the worst, least, and average pain in the past week and current ("right now") pain. The seven BPI interference subscale related questions address pain interference on activity related aspects such as walking ability, normal work, and sleep.

Participants complete the BPI four times, once at the beginning and/or end of each 4-week study period. The BPI's recall period is one week.

Health related quality of life is measured using the SF-36v2. The SF-36v2 is a widely-used and validated 36-item patient questionnaire to assess health-related quality of life. It includes eight health domains and yields two summary scales (physical health and mental health).

Participants complete the SF36v2 four times, once at the beginning and/or end of each 4-week study period. The SF36vs's recall period is the past four weeks.

Subjects opinions as to the benefits or effectiveness of the intervention were collected via an 11-point scale asking, "Overall, how beneficial was this treatment for your phantom limb pain? End points for the 11-point scale were: 0="Not beneficial at all" ; 10=Extremely beneficial". Participants were also asked to provide comment on whether they thought "additional treatment as provided in this study would improve [their] PLP".

Participants were asked about their perceived treatment benefit twice, at the end of the second and third 4-week study period.

Inclusion Criteria: ability to understand verbal or written English single limb/appendage, lower limb amputation persistent PLP for at least 1 year routinely has at least 3 occurrences of PLP per week PLP is greater than mild in pain severity according to the pain intensity component of the Brief Pain Inventory (3 or greater out of 10) Exclusion Criteria: presence of mild or greater traumatic brain injury known uncontrolled systemic diseases skin issues incompatible with massage application in the area for massage application (open or persistent wounds, fragile skin, etc) mental psychosis

All IPD that underlie results in a publication will be shared upon request. Sharing requests for IPD not included in current publications will be granted as long as secondary or further analyses are not compromised. All sharing requests should be made to the study PI.

IPD will be available for sharing 3 months after publication. Supporting Information items are available at and after publication submissions.