Treatments Available for Phantom Limb Pain

ensation. To determine the efficacy of such treatments by critically reviewing the literature available.

Methods: A Review of relevant studies identified by systematic searching of the National Health Services Knowledge Network databases, libraries of University of Strathclyde, the Robert Gordon University and medical libraries in Woodend Hospital, Aberdeen and Perth Royal Infirmary, as well as the use of paper resources provided by staff in Mobility and Rehabilitation Services (MARS) in Aberdeen. Journals articles where evaluated using “What is a critical review?” information from the “What is…? Series” from Hayward Medical Communications that includes the quality criteria of the Jadad scale (see Table 2).

Results: Two randomly controlled trials, which were deemed of moderate quality on the Jadad scale (3 points out of 5), five deemed of low quality on the Jadad scale (1 point out of 5) and one systematic review, were deemed suitable for inclusion in the review as well as additional information gained from the textbooks; Phantom and Stump Pain by Siegfried and Zimmermann and Phantom Pain by Sherman. The randomly controlled trials included in the review included treatments in the form of Mirror Therapy, phantom exercises with prosthetic training, hypnosis, reflexology and the use of active low frequency transcutaneous electrical nerve stimulation (TENS). The systematic review focused on the efficacy of mirror therapy in treating phantom limb sensation and pain and the qualitative reviews focused on perception of phantom limbs and sensation and the mechanisms which are the possible causes of the phantom limb phenomena.

Conclusion: There are many available treatments for phantom limb pain and phantom limb sensation, however due to the lack of quality in the current literature it is difficult to be able to draw strong conclusions on which treatments are better than others. More research is required to compare the treatments available in order to draw definitive conclusions on the most reliable treatment protocols for patients who experience these phantoms. In addition, more information needs to be made available to patients pre and post-amputation on what Phantom Limb Related Phenomena (PLRP) is and to inform them that it is a common issue within the amputee population.


Phantom limb sensations and phantom limb pain are vivid feelings that an amputated limb or body part is still present in an individual who has undergone amputation surgery to this area. It is reported up to 98% of the limb loss population experience phantom sensations, (Ramachandran and Hirstein, 1998), and as many as 80% experience phantom pain, (Sherman & Sherman, 1983; Sherman et al, 1984 as cited in Sherman, 1997, page 9). These phantoms range from the sensation of itching or pins and needles, to painful burning or cramping sensations all which can affect an individual’s day to day quality of life.

Phantom pain and sensations are often difficult to manage as the causes are not precisely understood, nevertheless phantom pain and sensations should not be considered as being trivial issues, as they appear in startling reality to the patient experiencing them (Wall, 1981). The most frequently used treatment is the use of analgesics, tricyclic anti-depressants and anti-convulsant drugs (Nikolajsen and Jensen, 2001). However, there are several other non-medicinal treatments available as alternatives to or to be used in conjunction with these medications. These treatments include: Mirror Box Therapy, Zuckweiler’s Image Imprinting (ZIPS), virtual immersive therapy, phantom exercises, reflexology, hypnosis and Transcutaneous Electrical Neural Stimulation (TENS). Surgical intervention for the treatment of phantom limb pain has been attempted in many ways since the first reports of phantom pains in the 1500s by Ambroise Paré. Most of these treatments have had limited success and the pain often resurfaces in time (Nikolajsen and Jensen, 2001) and the majority of these surgical interventions have been discarded due to this lack of success.


The aim of this literature review is to determine the efficacy of some of the non-medicinal and non-surgical treatments that are available for person’s with limb absence who suffer from phantom limb sensation and phantom limb pain.


A systematic review of the literature was carried out to identify suitable journal articles published up to and including 2009. This was achieved by utilizing the electronic database search feature of ‘The Knowledge Network’: Scotland website, as well as manual systematic searches in the libraries of the University of Strathclyde, the Robert Gordon University and the medical libraries at Woodend Hospital, Aberdeen and the Perth Royal Infirmary. Paper resources were also provided by the departmental staff at the Mobility and Rehabilitation Service in Aberdeen.

An initial basic search of “phantom limb pain” resulted in 922 articles and the following additional search terms where used to determine the preliminary literature for the review: “phantom”, “limb”, “pain”, “treatments”, “NOT medicine”, “NOT surgery”. These search terms then narrowed down the literature to 41 articles to which the initial inclusion and exclusion criteria (see Table 1) was applied in order to ascertain the articles suitable for appraisal. Unpublished papers were not sought after, however, non-English articles where included within the 41 resultant papers, alongside secondary articles from the reference pages of some of these resultant papers.


Once these criteria had been met, there were a total of two textbooks from the University of Strathclyde Library, one textbook chapter available from online and eleven journal articles which were subsequently critically appraised with information from the “What is…? Series” (Hill and Spittlehouse, 2001) to establish the final journal count (see appendices 1 and 2). Out of the eleven original journals, eight were deemed suitable to include in the review due to the quality of the articles determined during the appraisal process. Out of the three systematic reviews appraised only one was found to be of suitable quality to be included in the review.

The randomly controlled trials had to register on the JADAD-scale (see Table 2) in order to be included in the review. The JADAD-scale is an instrument that has been developed by Alejandro Jadad in order to assess the quality of reports of randomly controlled trials (Jadad et al, 1996). Out of the eight randomly controlled trials, two scored a 3 on the JADAD-scale establishing them as being of moderate quality, five scored a 1 on the JADAD-scale establishing them as low quality and the remaining randomly controlled trial did not register on the JADAD-scale and was therefore discarded.

The first article to be included in this review is “Effectiveness of phantom exercises for phantom limb pain: a pilot study” by Ülger et al (2009). In which they carried out a pilot study on the efficacy of incorporating phantom exercises within prosthetic training versus routine prosthetic rehabilitation. A Jadad scoring of three for this trial resulted in this paper achieving moderate quality. The study involved twenty traumatic upper and lower limb amputees, between the ages of 30 and 45 years old, who experienced phantom limb pain which registered at least a seven on the visual analogue scale. Each subject was randomly assigned into one of two groups depending on their arrival time at the clinic. The first group consisted of ten patients who each received phantom exercises along with prosthetic training. The second group included the remaining ten patients who received routine prosthetic training and a general exercise programme. The trial was not blinded and there were no withdrawals from the trial. After four weeks, according to the visual analogue scale, there had been a statistically important result in the reduction of phantom limb pain between the phantom exercises group and the general exercise group (ρ=0.05). The subjects where followed up two months after discharge from hospital by telephone. Subjects reported at this time, that they still performed the exercises if they were required during experiences of phantom limb pain however the frequency of phantom limb pain had decreased overall. This paper concluded that the phantom exercises seemed to be effective in the reduction of phantom pain in patients with upper and lower limb loss, although it does suggest that further research is necessary in the field to confirm this result.

The second randomly controlled trial included in this review is “Transcutaneous Electrical Nerve Stimulation after Major Amputation” by Finsen et al (1998), which also achieved moderate quality on the Jadad scale. This study focused on 52 patients either with an ankle disarticulation, trans-tibial or knee disarticulation amputation due to ischemic changes as a result of diabetes or atherosclerosis and what the effect of Transcutaneous Electrical Nerve Stimulation (TENS) had on stump healing and on post-operative and late phantom limb pain. The study was randomised but not perfectly as there was an uneven spread of amputation levels throughout the groups, although this was taken into account when collating results. Study was also single blinded with the patients not knowing what treatment they received as sham TENS was used on two of the groups with electrodes being placed in the inactive channel of the stimulator and only one group receiving active TENS. The patients were split into 3 groups: Group 1 received sham TENS and chlorpromazine, Group 2 received sham TENS only, and Group 3 received low frequency active TENS. The active TENS stimulation utilized two electrodes positioned over the femoral and sciatic nerves to deliver low frequency pulses and the sham TENS groups had the electrodes positioned over the femoral nerve whilst being connected to inactive ports. Patients who were withdrawn from the treatment included thirteen whose stumps did not heal, two patients who died and one who changed departments. The results from this study showed that after six weeks and at nine weeks there were significantly more (ρ=0.05) healed stumps in the ten patients from Group C who received the active TENS stimulation compared with the 23 patients in Groups A and B who received sham TENS. Patients were also followed up at 16 weeks and one year post-operatively. Results were not statistically significant in relation to median analgesic doses between the three groups as there was a large variation in the consumption of analgesics within the groups themselves, although the median number of analgesics consumed was lower in Group C. This paper concluded that the use of TENS helped increase the rate of wound healing in the first instance. Stimulation helps reduce phantom pain in the middle term but has no long term effect on the difference between the groups with significant reduction of phantom pain.

The remaining randomly controlled trials only managed to score a one on the Jadad scale indicating that they are of low quality thus representing a lack of quality research in this field which will need to be addressed in future studies.

Chan et al (2007) conducted a randomised sham-control trial of 22 lower limb amputees who suffered from phantom limb pain in order to determine the efficacy of visual feedback versus illusions of movement in treating phantom limb pain and sensation with the use of mirror therapy. Mirror therapy is a treatment that has been developed in order to treat chronic pain disorders including the rehabilitation of amputees who suffer from phantom limb pain and sensations. It involves amputees attempting to perform movements with the residual limb whilst viewing a reflected image of their intact limb in the mirror. 18 patients completed the trial, however details of the four withdrawals was not given. There were six amputees assigned to each group; Group 1 undertook mirror therapy, Group 2 where the control group who performed mirror therapy but with a mirror covered in an opaque sheet and Group 3 where trained in mental visualisation, where they closed their eyes and imagined performing the movements with their phantom limbs. Prior to treatment commencing baseline scores for number and duration of pain episodes and intensity of pain episodes was recorded and the results were similar amongst each group (ρ=0.62). After four weeks of treatment, Group 1 resulted in 100% of patients reporting decreased pain intensity and episodes of pain, Group 2 resulted in 17% (1 patient) reporting a decrease in pain and 50% reporting an increase in pain and Group 3 had 33% of patients reporting a decrease in pain and 67% reporting worsening pain. There were four withdrawals from this trial, although none of these were described. This trial concluded that mirror therapy was an effective treatment for phantom limb pain however two patients did experience a brief period of grieving after viewing their reflected intact limb in the mirror The underlying mechanisms that account for the positive result of mirror therapy are not known although the authors suggest that they are due to either activation of mirror neurons in the contra-lateral hemisphere of the brain to the amputated side, or due to the visual input of what appears to be movement of the amputated limb that may reduce activity of the systems that perceive pain signals.

“Training with Virtual Visual Feedback to Alleviate Phantom Pain” by Mercier and Sirigu (2009), attempted to explore the factors that influence the success of Virtual Visual Feedback (VVF) based on the individual response of each patient. Virtual Visual Feedback has been developed off the back of Mirror Box Therapy in which instead of a reflected and inverted image of the patient’s limb being viewed in the mirror box, a computer generated image of the limb is viewed by the patient on a screen. This study used a convenience sample of eleven male participants who had either traumatic trans-humeral amputation or a brachial plexus avulsion and who experienced phantom limb pain. Participants completed the individual training sessions consisting of a virtual image of the paralysed or missing arm performing specific exercises which the phantom limb had to follow, twice a week over an eight week period. All of the subjects reported that the phantom exercises and movements were easier to perform with the aid of visual feedback. There was an average pain reduction of 38% at the end of treatment, with five participants reporting a decrease of more than 30%. At the end of the 4-week follow-up stage these results were maintained in four of these five patients. Three participants withdrew from treatment, two due to work commitments and one due to geographical distance from treatment centre. This study concludes that virtual visual feedback does alleviate phantom limb pain if applied to the right patients. It suggests that physiological and psychological measures should be in place to identify patients who are likely to be receptive to treatment of phantom limb pain with virtual visual feedback, as the success of this treatment depends on the susceptibility of the patient to the ideas of VVF.

Another randomly controlled trial that scored a one on the Jadad scale was “Reflexology treatment for patients with lower limb amputations and phantom limb pain – An exploratory pilot study” by Brown and Lido (2008). This pilot study involved twelve pre-selected patients with unilateral lower limb loss who were six months post-amputation and who suffered with phantom limb pain. The aim of this study was to evaluate the potential usage of reflexology as a non-invasive treatment for phantom limb pain and phantom limb sensation. The study was split into five, six week block’s throughout which the subjects had to record a pain diary in order for individual pain patterns to be recorded. Phase one was purely for recording the initial pain log, the second phase introduced weekly reflexology sessions on which a subject’s foot, on the residual side and hand, on their amputated side, was manipulated. Phase three was a period of rest and phase four was the second set of weekly reflexology, during which subjects where taught self-manipulation of their own hand (on the amputated side). The final phase involved the patients self-treating themselves on a weekly basis. Two patients withdrew from the study as one had an undiagnosed pedal infection and the other had family commitments. The results of the study showed that there had been an improvement in the perception of the existence of subjects’ phantom limb pain as well as the intensity with which it presented. There was an improvement in the duration of the pain and how the pains affected the quality of life of the subjects, and this improvement was able to be maintained by the self-treatment of patients after completion of the 30 week trial. The follow-up questionnaire showed that the ability to self-treat allowed the patients to feel that they had some measure of control over their phantom pains which again helped with the improvement in their quality of life.

Bamford (2006) carried out research to assess the efficacy of using hypnosis in a multifaceted approach to relieve phantom limb pain. The four therapeutic interventions utilised in this study were hypnotic analgesia, visualisation and movement of the imaginary limb to facilitate movement of the phantom limb, psychological therapy and self-hypnosis. Thirty-four patients, with both upper and lower limb amputations who suffered from phantom limb pain, were referred for hypnosis after gaining no relief from various other methods of pain relief. Three of these patients showed cognitive impairment and six failed to attend after the first session, the remaining 25 patients continued with the 6 weekly treatment sessions as well as agreeing to undertake thrice daily self-hypnosis sessions at home. There was a follow-up questionnaire sent to patients six months after completion of the trial. The journal concluded that because the patients could relieve their own pain that this had a motivational effect on them. The treatment reduced the number of post-amputation pain clinics, reduced the number of appointments with rehabilitation consultants with regards to pain and allowed patients to be able to reduce the amount of medications they were taking as well as taking away some of the side effects that these medications caused. The author is currently still monitoring how effective self-hypnosis is due to the nearly significant (ρ=0.07) difference between the baseline and self-hypnosis results, she is also monitoring a modified version of the self-hypnosis versus a personalised recorded CD of the patients hypnotic sessions. She also recommends further study with brachial plexus avulsion and complex regional pain syndrome patients.

Mortimer, Steedman, McMillan, Martin, and Ravey published: “Patient information on phantom limb pain: a focus group study of patient experiences, perceptions and opinions” in 2002. This study focused on patients with lower limb amputations who experienced phantom limb pain and aimed to explore their experiences and perceptions of the information they received before and after amputation on phantom limb pain and sensations. The initial convenience sample, who met the inclusion criteria, consisted of 62 patients who were identified from a prosthetic centre in central Scotland and another five who volunteered from the Prosthetic Users Self-Help group (PUSH). Withdrawals included 28 patients who declined to take part, five were found to be unsuitable, two were deceased and one had moved away, therefore a total of thirty-one subjects took part in the study. This study involved seven groups, one which contained the PUSH volunteers and six others from the prosthetic service. Each participant was provided with written information, a consent form and an initial questionnaire to gather demographic information before they took part and each group session was held in convenient locations for the participants. The group sessions where led by a moderator who audio recorded the sessions and in all but one session an observer was also present in order to make additional notes in relation to body posture and other interesting points. Six areas of interest; which were developed by Mortimer in a previous study, were used to direct the discussions. These areas were relating to phantom limb sensation and pain and the support and information provided to patients pre and post-amputation. After the group sessions where collated it was concluded that phantom sensations and pain are significant parts of an amputees experience and should be addressed appropriately by the relevant professionals. The findings agreed that there are common experiences within the amputee population with the types of sensations and pains they experience, in addition to those being experienced frequently with varying degrees of severity, for longer than six months post-amputation. The information that participants felt the professionals should be supplying them with was inadequate both pre and post-amputation. The study concluded that in order for the multi-disciplinary team to be able to supply patients with the correct information at different stages of the rehabilitation process, a standardised approach to developing an information pack should be developed. This pack should involve peer sources and should be utilised more efficiently throughout the rehabilitation process in order to provide up-to-date information to the patients at all times.

Only one systematic review was deemed of high enough quality to include in the review and this was the 2009 paper “Spiegeltherapie bei Phantomschmerzen, Eine systematische Übersichtsarbeit” which translates as “Mirror Therapy for Phantom Limb Pain, A Systematic Review” by Seidel, Kasprian, Sycha and Auff. The original study was a German study which was translated into English with the use of Google Translate. The review focused on patients who suffered with Complex Regional Pain Syndrome (CRPS) or patients with Phantom Limb Related Phenomena (PLRP) as a result of upper or lower limb amputations and the effect that mirror therapy had as a treatment for these conditions. The papers used in the study were located, using the relevant search terms, from the online Medline Database and the Cochrane Libraries and after the initial screening of the titles and abstracts, 1193 papers remained. These papers included foreign language papers and papers yet to be published. The reviewers then used the Jadad-scale to determine the quality of the papers and the results showed that the three remaining randomly controlled trials where of low quality, only scoring one or two on the Jadad-scale. The results were expressed with tables, images and in textual discussion and the reviewers concluded that better methodology needs to be developed in order to assess more effectively the treatments of phantom pain and CRPS and blinding needs to be utilised more effectively to increase the quality of the research. The three included studies also concluded that even with the small size of the studies all the results pointed towards mirror therapy being a successful, low cost, non-invasive treatment for complex regional pain syndrome and phantom limb related phenomena.


On appraisal of the included journal articles, it is clear that phantom limb pain and phantom limb sensations are prevalent within the amputee population and that there has been no agreement of the definitive causes. The most popular theory is that post-amputation, cortical reorganisation takes place and the amount of reorganisation correlates to the amount of phantom limb pain experienced (Mercier & Sirigu 2009, Nikolajsen & Jensen 2001, Mortimer et al 1998, Zuckweiler 2005 and Fraser et al 2001). Another potential cause of phantom pain and sensations was described by Ülger (2009) and is thought to occur due to the absence of normal stimulation and the over-discharges from damaged neurones being interpreted as pain signals by the changed structure of the neuromatrix. Ülger also proposes that somatosensory pain memories can revive after amputation and cause phantom limb pain, especially in traumatic or painful vascular amputations.

One common theme found within the journal articles was the lack of quality of the available literature, as most of the relevant journals for this review where found to score only a one on the Jadad-scale. Further research is required for more thorough trials in order to compare the many non-medicinal and non-surgical treatments available for the treatment of phantom limb related phenomena, with each other and also in conjunction with medications to find the optimum treatment path for each patient. Within this requirement for more research, higher quality research needs to be carried out for the current treatments in order to allow the professionals to be able to construct these reliable treatment paths, as they will have a better quality of literature on which to make their recommendations. However, in spite of this requirement for better research, several of the articles did point towards the success of their relevant treatments. With statistical results of ρ≤0.05 being regarded as a significant result, many of them can be used for the treatment of PLRP if in the right context and if the correct patients are identified. The benefit of several of these treatments is that they can often result in the patients being able to reduce the amount of medication they are taking as a result of the reduction in PLRP due to the use of non-invasive treatments. These treatments can also often result in the patient being able to self-treat when necessary, for example Bamford (2006) and Brown (2008) include teaching the participants how to self-treat within their papers, using hypnosis and reflexology respectively and this was shown to be able to give the participants the freedom and confidence that if they do experience PLRP they know how to reduce their symptoms by following the treatment guidelines they had been provided with.


Up until now there have been no definitive results on what the causes of phantom limb pain and phantom limb sensations actually are, although there are many theories being researched currently. There are many available treatments for phantom limb pain and phantom limb sensation, however due to the lack of quality in the current literature it is difficult to be able to draw strong conclusions on which treatments are better than others. More research is required to compare the treatments available in order to draw definitive conclusions on the most reliable treatment protocols for patients who experience these phantoms. In addition, more information needs to be made available to patients pre and post-amputation on what PLRP is and to inform them that it is a common issue within the amputee population. Professionals also need to be aware of the range of treatments available and this should be done within the development of better quality research and information.

Cite This Work

To export a reference to this article please select a referencing stye below:

Related Services

View all 


Leave a Reply