Background: The formation, prevalence, intensity, course, and predisposing factors of phantom limb pain were investigated to determine possible mechanisms of the origin of phantom limb pain in traumatic upper limb amputees.
Methods: Ninety-six upper limb amputees participated in the study. A questionnaire assessed the following question: side, date, extension, and cause of amputation; preamputation pain; and presence or absence of phantom pain, phantom and stump sensations or stump pain or both.
Results: The response rate was 84%. Sixty-five (81%) participants returned the questionnaire. In 64 (98.5%) participants a traumatic injury led to amputation; the amputation was necessary because of infection in one patient (1.5%). The median follow-up time (from amputation to evaluation) was 3.2 years (range, 0.9-3.8 years) The prevalence of phantom pain was 44.6%, phantom sensation 53.8%, stump pain 61.5%, and stump sensation 78.5%. After its first appearance, phantom pain had a decreasing course in 14 (48.2%) of 29 amputees, was stable in 11 (37.9%) amputees, and worsened in 2 (6.9%) of 29 amputees. Stump pain had a decreasing course in 19 (47.5%) of 40 amputees but was stable in 12 (30%) amputees. Phantom pain occurred immediately after amputation in 8 (28%) of 29 amputees between 1 month and 12 months in 3 (10%) amputees and after 12 or more months in 12 (41%) amputees.
Conclusion: Stump pain and stump sensation predominate traumatic amputees' somatosensory experience immediately after amputation; phantom pain and phantom sensations are often long-term consequences of amputation. Amputees experience phantom sensations and phantom pain within 1 month after amputation, a second peak occurs 12 months after amputation. Revised diagnostic criteria for phantom pain are proposed on the basis of these data.