Phantom pain, anxiety, depression, and their relation in consecutive patients with amputated limbs: case reports

BMJ 1998; 316 doi: (Published 21 March 1998) Cite this as: BMJ 1998;316:903
  1. K Fisher, consultant clinical psychologist,
  2. R S Hanspal, consultant in rehabilitation medicine
  1. Disablement Services Centre, Royal National Orthopaedic Hospital Trust, Brockley Hill, Stanmore, Middlesex HA7 4LP
  1. Correspondence to: Dr Fisher
  • Accepted 3 October 1997

Parkes suggested that emotional factors are influential in patients' experience of prolonged pain in a phantom limb after amputation and concluded that this may be prevented if patients are encouraged to express grief over their loss.1 However, Katz and Melzack found no significant difference in standardised tests of psychological dysfunction between patients who experienced phantom pain and those who did not. They concluded that the pain is more likely to vary with the experience of preamputation pain, even retaining many of its characteristics.2 A review of the literature on measures used to diagnose psychopathology found that many measures include items that confound emotional distress with the physical disorder and thus overestimate it.3 We investigated whether people who had had arms or legs amputated experienced emotional distress, and the relation between the distress and pain, using standardised screening techniques designed for patients with physical illness.

Patients, methods, and results

Calculations of sample size indicated that 21 patients per group would be needed to show a reliable difference at the 5% level of significance. The participants were 93 consecutive patients who had been referred to the prosthetic rehabilitation clinic and were aged 34-91 (mean 65) years; 54 were men. Time since amputation was 1-58 (9.7) years. Sixty patients had had a leg amputated for vascular illness, including diabetes, 10 of them losing both legs. Twenty four patients had lost a leg and nine an arm because of trauma. RSH obtained a clinical history including information about previous and concurrent medical and psychiatric problems. Phantom pain was assessed with the short form McGill pain questionnaire,4 the patients endorsing all words describing their phantom pain, if present. KF, who was blind to the pain report, then assessed them with the hospital anxiety and depression scale.5

Phantom pain (mostly mild) was reported by 29 patients. Fifty three of the remaining 64 patients reported non-painful sensations in the phantom limb. Mean scores on the anxiety and depression scale were 3.9 for anxiety and 2.9 for depression. Whereas 10 patients scored in the clinical range for anxiety, mainly about falling, only one patient scored in this range for depression. No patient gave a history of previous or concurrent psychiatric treatment.

The patients were divided according to whether they experienced pain, and their anxiety and depression scores and time from amputation were compared with non-parametric statistics. The table shows that the time from amputation, and anxiety and depression scores did not differ between the two groups. Time from amputation was not strongly significantly associated with distress, so anxiety and depression do not seem to vary consistently over time.

Table 1

Differences between patients with and without pain in phantom limb in time from amputation and scores on hospital anxiety and depression scale. Values are means (95% confidence intervals)

View this table:


The incidence of phantom pain in this study was 31%, in keeping with current reports.2 Only a few patients experienced emotional distress, anxiety being reported more often than depression. The prevalence of depression was low, suggesting that it is an uncommon reaction to amputation. In this elderly group of patients who had discomfort due to vascular illness, loss of the limb did not constitute a bereavement in the way that Parkes suggested.1

These results, in agreement with those of Katz and Melzack,2 show little support for the grief hypothesis, since it is difficult to sustain a concept of grief in the absence of depression on objective measures. In addition, we found no relation between the experience of pain and emotional distress, suggesting that phantom pain is not a function of emotional adjustment.


Dr M J Campbell and Dr Robert West commented on the statistics.

Contributors: RSH undertook the initial assessment of the patients and administered the McGill pain questionnaire. KF administered the hospital anxiety and depression scale, analysed the data, and wrote the paper. KF is the guarantor for the study.

Funding: None.

Conflict of interest: None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
View Abstract