Reflex sympathetic dystrophy (RSD) is a complex regional pain syndrome characterized by variable dysfunctions of the musculoskeletal, skin, and vascular systems. The most common features are severe persistent spontaneous pain in a limb, sensory and vascular alterations, and major associated disability, which often causes profound psychological dysfunction (, 2).
RSD is characterized by spontaneous pain, swelling, dysaethesia, and allodynia usually of a distal site of the affected extremity (3, 4). Other features relate to the autonomic nervous system and include cyanosis, mottling, sweating (which is less common in children), and reduction in temperature. In some cases, particularly if there is delay in recognition or treatment, more permanent or serious features may develop; these include muscle atrophy, demineralization of bone, and contractures of soft tissue around the affected joint.
In the early stages of RSD, there are no specific laboratory tests available to help diagnose the condition, and the radiologic and bone scanning findings characteristic in adults are often not seen in children (5–9). The lack of diagnostic investigations, the relative lack of recognition of the condition, and the more variable clinical presentation present in children often result in a delay in diagnosis and treatment. This often is complicated by the diversity of medical opinions sought from the onset of symptoms. This may lead to a more severe and chronic presentation. Although a number of studies have reported RSD in childhood, a variety of approaches to its management have been suggested with variable or undetermined outcomes (10).
RSD may also occur in, or be associated with, a variety of other clinical conditions and may affect different areas of the body to a greater or lesser extent. There are also a variety of descriptive names that are used as synonyms for RSD, including algodystrophy, reflex neurovascular dystrophy, causalgia, Sudeck's atrophy (although this refers to patients with the classic macular osteopenia, not often seen in children), and complex regional pain syndrome type 1 (2). In this article we prefer the older but perhaps still more commonly used term, RSD (11–13).
Population studies in schoolchildren indicate that chronic or recurrent musculoskeletal pain is quite common, with back pain being the most frequent (20%), followed by limb pain (16%) and fibromyalgia (6%) (14–17). There is little work on the incidence of RSD, but ∼5–8% of new referrals to North American pediatric rheumatology centers have a diagnosis of idiopathic musculoskeletal pain syndromes, a small percentage of which are RSD (18, 19). Idiopathic musculoskeletal pain syndromes have been described in patients as young as 3 years of age, though RSD has not been reported at this age. The majority of reports involve children in late childhood or early adolescence, with a mean age of onset of 12–13 years and a higher frequency in girls than boys (∼4:1) (3, 4, 13, 20–22). There have been no reports of specific ethnic differences in the presentation of RSD. Currently, there is no single or specific biologic cause identified or pathogenetic pathway described for RSD, and it is clear that the clinical features may vary widely from children to adults. The etiology is likely to be multifactoral in some children. In many pediatric cases, however, there is no or only a very minor precipitating incident; psychological factors may have greater influence. This hypothesis is supported with the differences found between technetium bone scintigraphy in children compared with adults and by the fact that children respond more readily to physical therapies (4, 23). However, it appears that the triggering factor in many children is related to injury, illness, or psychological distress in any combination (3–5, 8, 9, 13–17, 20, 24–39), in contrast to adults in whom surgery or myocardial infarction often precedes the onset of RSD (31, 32, 40).
Many methods of managing this condition have been reported with varying success, including exercise therapy (36), transcutaneous electrical nerve stimulation, sympathetic nerve blocks (41), spinal cord stimulation (42), and such medications as gabapentin (43), calcitonin (44), in trathecal baclofen (45). The majority of these studies include small numbers and actual controlled trials are lacking.
In this retrospective study, we report on the clinical course and treatment response of 23 children with RSD referred to a tertiary pediatric rheumatology unit. We highlight the particular treatment approach utilized in the care of these children, which emphasizes exercise therapy, desensitization, and psychological counseling.
This study identified patients with RSD from a prospectively collected diagnostic registry. The medical and physiotherapy clinical documentation of the patients who had been referred to our unit at Great Ormond Street Hospital, London and who had a diagnosis of RSD were reviewed. This review was completed on 23 patients who were seen at the unit from 1993 to 2000. Of these 23 children, 4 had been followed for up to 6 months and 19 had been followed for >6 months. Diagnosis was made by a pediatric rheumatologist based on the clinical findings at the initial assessment, including documentation of a typical history of severe persistent spontaneous pain of an extremity; descriptions of burning, tingling, or paraesthesia; dysethesia, including temperature sensitivity or hypersensitivity to touch or movement (hyperesthesia) and allodynia (pain induced from light touch or movement). They also presented with clinical evidence on examination of a variable combination of autonomic signs, such as cyanosis, edema, mottling, cold, altered hair growth, and altered perspiration. Diagnosis was in concordance with Kozin et al (32) and with the modified International Association for the Study of Pain criteria (modified by the exclusion of the presence of the initiating noxious event, as this is often not present in the pediatric population) (2). Patients were considered to have definite RSD if they had chronic severe pain and tenderness in a limb region or extremity with some evidence of neurovascular of sympathetic disturbance including at least one of hyperesthesia, cyanosis or skin color changes, swelling or edema, and dystrophy.
After the initial consultation and diagnosis, the pediatric rheumatologist decided on the management course. Table 1 describes the criteria used for deciding the most appropriate course of management for each child.
Table 1. Criteria for management approaches in RSD*
Criteria for outpatient management
Criteria for inpatient management
RSD = reflex sympathetic dystrophy.
Good local therapy support
No or very little local therapy support
Family prepared to provide intensive home treatment
Family would find providing intensive home treatment difficult
Outpatient treatment consisted of a detailed explanation of RSD and the treatment ideology of using physical exercise to restore movement and function. This explanation included a description of the theory of altered sensation stressing that the pain was “useless pain” and that it did not signify a serious underlying pathology. The treatment philosophy was explained using the symbol of “pins and needles:” remaining motionless reduces the discomfort but does not get rid of the sensation, a small amount of movement causes a great deal of pain and discomfort, but the quickest way of alleviating the pain is to move normally and as much as possible.
Referral was then made to the local medical and therapy teams for outpatient treatment to begin. The child was shown the exercise regimen used routinely at this hospital as a guide for an initial home exercise program. The child was given advice about desensitization of the affected area by massage and touch. This exercise program focused upon exercising each muscle group individually before combining muscle work in functional activities. The importance of “little and often” and gradual progression of the treatment program was stressed. Patients were managed as outpatients only if the local team had sufficient time to provide the regular input and followup required.
When inpatient treatment was recommended, the children were admitted as soon as a bed was available, usually within a few weeks. On the admission day, a multidisciplinary approach was instituted comprising a medical physical reassessment and complete musculoskeletal assessment by a rheumatology physiotherapist, occupational therapist, and a specialist nurse. The assessment consisted of a full comprehensive history of the symptoms (the pain, the autonomic changes, and the functional consequences of the RSD), relevant family history, psychosocial functioning, and risk factors. The child was specifically examined for active and passive joint range of motion, muscle strength, and pain pattern related to such activities. Functional assessment included posture, gait, and summary of impact of the RSD on school functioning and general activities of daily living. From this assessment, a detailed plan of treatment was designed to ensure that all aspects of management were coordinated. The goals of inpatient therapy are shown in Table 2. The long-term goals for all children were to regain normal movement and full independent function, with complete pain relief being deemphasized.
Table 2. Aims of inpatient physiotherapy
Increased functional joint range of movement of affected limb
Increased function of affected limb
Improved muscle strength
Normal sensitivity to touch and movement
Normal balance, proprioception
Reduced or no pain
Normal independent function in all activities of daily living
The children, families, and therapists designed a detailed treatment timetable that included physiotherapy, occupational therapy, nursing treatment, psychology, and hospital school attendance.
Each child received 1–2 hours of hydrotherapy and 2 hours of land-based physiotherapy daily. Each session was in a group setting, but was supervised by at least 1 qualified physiotherapist and 1 physiotherapy assistant, both experienced in the management of RSD. These treatment sessions used various techniques to achieve the aims described in Table 2. The child was given a program of massage to complete with either the nursing staff or their family as an important part of the program of desensitisation of the affected limb. The intensity and length of massage was increased gradually.
The children also received occupational therapy during the admission period with a variety of techniques used to fulfill the treatment aims already mentioned. Any previously prescribed splinting was removed and was avoided because immobilization of the affected area was felt to exacerbate the symptoms and prolong recovery time.
In addition, nonsteroidal antiinflammatory drugs or acetaminophen were used occasionally for analgesia and psychological support but more frequently these analgesics, including narcotics, were stopped because the children reported that they received no benefit from them.
Counseling from the psychologist was recommended for the child and family to elucidate important precipitating or perpetuating factors, and to enhance their coping and pain-management strategies. Issues around sleep disturbance were also discussed in this setting.
An initial treatment period of 2 weeks rehabilitation was planned with weekend home leave allowed between the 2 weeks. This enabled the children and their families to practice the techniques that they had learned during the admission time. For the more severe cases, this inpatient admission was extended up to 4 weeks.
After discharge, the children were seen frequently until the clinical features had resolved completely or had improved significantly and were being appropriately managed by the family. The children were seen again after an additional 3 and 6 months. Any patients not seen in the 6 months prior to the case-notes review were contacted for a telephone interview.
The case-note review identified the patient demographics, the preexisting history, the precipitating events, and previous management provided (including medication, therapy, administration of nerve blocks, etc.). Clinical features on history or examination were documented, including region or limb affected, pain (hyperesthesia or allodynia), vasomotor changes (sweating or color changes), tissue swelling or edema, and dystrophic changes (hair loss, soft tissue atrophy, or bone atrophy).
The length of time to diagnosis was calculated from the onset of symptoms, and the length of time to recovery was also established. Full recovery was defined as a resolution of physical symptoms and complete return to full independent function, i.e., full attendance at school and full participation in age-appropriate activities. This recovery may or may not have included complete resolution of the pain, because often the experience of pain continued after full function was regained. Any child who had a relapse in their symptoms was also noted. Those who had a recurrence of RSD signs and symptoms, including reoccurrence of pain after it had resolved completely, within 3 months of apparent resolution were not considered to have made a full recovery. Those who had recurrence in another site or in the same site after 3 months were considered to have a new episode and to have had full recovery from the initial episode.
Statistical analysis was completed using the SPSS 10.1 package (SPSS, Chicago, IL). Descriptive statistics were used to describe frequency and means of the appropriate data, and parametric independent t-testing was carried out to compare the outcomes of the inpatient and outpatient management. This was possible because the Q-Q plots indicated normal data despite the small numbers. A significant result was accepted if P = 0.05.
RSD was diagnosed in 23 patients referred between January 1993 and December 1999 (Table 3). Eighteen were girls (78%) and 5 were boys (22%). The median age at onset was 11.9 years (range 8.6–17.5) with a median age at diagnosis of 12.1 years (8.7–18.1). Many of the patients had seen multiple medical and health practitioners prior to referral and diagnosis. Six patients also presented with 1 or more preexisting chronic conditions: 4 had juvenile idiopathic arthritis, 2 had recurrent transient synovitis of the hip, 1 had rhabdomyosarcoma, 1 had mild cerebral palsy, and 1 patient was diagnosed with asthma. Fourteen patients also reported 1 or more other chronic symptoms: 5 had recurrent headaches, 1 complained of chronic abdominal pain, 5 reported sleep disorders, 2 reported feelings of anxiety, and 2 complained of depression. Three patients also reported a history of conversion disorders and 1 reported anxiety-related hyperventilation. In addition, 16 of 23 (70%) had 1 or more close relative with a chronic or debilitating illness, of which 13 were of musculoskeletal origin. Five patients had a total of 7 first-degree relatives with a diagnosed psychiatric condition.
Table 3. Clinical characteristics of 23 children with RSD*
RSD = reflex sympathetic dystrophy.
Female, no. (%)
Upper limb, no. (%)
Lower limb, no. (%)
Left side, no. (%)
Right side, no. (%)
Both sides, no. (%)
Age at onset, median (range) years
Age at diagnosis, median (range) years
Duration onset to diagnosis, mean (median; range) weeks
All patients reported severe pain in the affected limb(s) as the presenting symptom. In addition to this, there were reports of restricted joint range of motion in 20 (91%) children, sensory symptoms in 14 (61%), swelling in 11 (50%), color change in 13 (56%), and abnormal sweating in 6 (26%). Nineteen patients (83%) had used crutches at some point during their illness and 14 (61%) had used a wheelchair.
Investigations prior to diagnosis were usually extensive. Eighteen patients (82%) had had plain radiographs (4 of which showed regional osteopenia). Eight patients (36%) had undergone a nucleotide bone scan, of which 3 demonstrated reduced uptake in the affected limb. Other investigations, including magnetic resonance imaging, electromyography, thermography, and nerve conduction studies, had been performed prior to referral and were not felt to be contributory to diagnosis or management.
Six patients (26%) were admitted to another hospital prior to referral to our unit, of which 4 (17%) were reported to have received inpatient physiotherapy. Ten patients (43%) had outpatient physiotherapy on average once weekly. In contrast, 19 patients (83%) had been treated with a period of limb immobilization using casts or splints.
Table 4 shows the results of examination findings in the affected limb in this study. Eighteen of 23 patients were admitted to our unit, and all received physiotherapy as described in addition to simple analgesics and nonsteroidal antiinflammatory drugs as required. All patients received intensive education and were taught self-management strategies, and all were given a tailored home-exercise program at discharge. The remaining 5 patients were treated as outpatients by their local services following initial treatment advice and ongoing support from the multidisciplinary team.
Table 4. Results of examination findings at assessment
Description of findings
Percentage of involvement
In 5 of the inpatients where adverse psychological factors were felt to be contributory, formal referral for clinical psychology assessment and appropriate intervention was initiated. The average followup time from diagnosis was 26 months in our group (range 5 months to 7 years), with 4 children being followed for 6 months or less. Current status was determined from the time of last clinic visit or by telephone interview.
Full recovery was defined as recovery with no recurrence of symptoms within the initial 3 months following resolution. Recurrences after this time were considered new episodes. A full recovery from the original episode was achieved in 78% (18/23) of patients.
We found that there was no significant reduction in the time reported to complete recovery between inpatient and outpatient management, with the inpatients' mean time to recovery being 6.2 months and outpatients' being 6.35 months (P = 0.94). However, there was a difference in the time to definite improvement in symptoms in patients receiving inpatient treatment compared with those receiving outpatient management. Patients who were admitted had a significantly reduced time interval between diagnosis and definite improvement, with a mean time for inpatients being 2.8 weeks and outpatients being 6.2 weeks from diagnosis (P = 0.026).
In our cohort, there was a trend for a higher proportion of patients in whom the diagnosis was made within 4 months of onset of RSD to make a full recovery earlier than those who had a delay in diagnosis; however the difference was not statistically significant (P = 0.96; Figure 1). On the contrary, when correlations were performed between time to diagnosis and time to improvement, there was a positive correlation (r = 0.587, P = 0.01).
Full outcome information for followup of >6 months was available for 19 of 23 patients, 11 of whom had remained completely well (58%), 5 had major functional improvement but were still symptomatic (26%, i.e., were still having some minor new pain episodes in the same [3 of 5] or another [2 of 5] limb), and 3 patients reported that they had had minimal improvement since diagnosis (15%). Therefore, a good to excellent outcome lasting >6 months was seen in 84% (16 of 19) of the patients that remained in the study for full followup. Of those who had recovered fully from the initial episode, the mean time to definite improvement was 4 weeks (range 2–10 weeks) and the mean time to full recovery was 25 weeks (range 4 weeks to 15 months).
RSD is a condition affecting both adults and children that results in significant morbidity. This condition is often challenging to manage, but can be treated very successfully with noninvasive but aggressive conservative management (46). This is supported by the evidence provided in this study as well as other studies reported in the literature (7, 47–50).
This study of 23 children with a definite diagnosis of RSD, who were managed over a 7-year period, provided good evidence for the successful provision of an intensive management program. Although this program requires inpatient admission at the present time, outpatient management would be possible with the correct ambulatory care facilities. Because we are a tertiary referral unit, it is likely that the sample represents a bias toward more severe cases and perhaps more associated psychosocial disturbance leading to urgency of referral. Although it is recognized that RSD is uncommon in children (51, 52) clinicians are increasingly recognizing and diagnosing this condition.
We document a higher incidence of RSD in girls compared with boys (∼3:1), as has been found in other studies (7, 10, 51). The majority of cases occurred from early adolescence (11–12 years) onward, and this is also consistent with other data. Most cases occurred predominantly in the leg compared with the arm (10). Although dominant handedness or leg use was not routinely recorded, there was no significant difference in which side of the body the episodes occurred, which also is supported by other studies (7).
Many potential causative or precipitating factors were identified in the history of these patients and could be identified as risk factors. The main psychosocial indicator was the high prevalence of relatives in the same family with chronic pain conditions. Sixteen children (70%) had 1 or more first-degree relatives with a chronic painful illness and of those, 13 had a musculoskeletal pain condition. Five patients had a first-degree relative with a diagnosed psychiatric condition. This high incidence of chronic illness within the family is likely to have influences on their understanding of pain and the mechanisms for managing pain, and thus a significant impact upon the pain experiences of the young person and their uptake of the “sick-role.” It was also evident that a proportion of this group had a preexisting knowledge of pain because 22% already had a preexisting painful condition themselves and an additional 14 (61%) reported other chronic symptoms such as headaches, sleep disturbance, and anxiety. This indicates that children with RSD may also suffer from more widespread pain syndromes and other possible psychosomatic symptoms, as has been reported previously. There have been several psychosocial indicators related to RSD, and those identified were related to role models of pain and school stress. Parental education and more chronic family problems, including the issue that children with RSD were more likely to be in a home with only 1 biologic parent, were also identified as significant factors (53–55).
About half of our cohort recalled an episode of trauma to the affected limb prior to the onset of the RSD, and this has also been reported in other studies (7, 10, 48, 56, 57). This trauma had often been investigated at length before a diagnosis was made and was felt to have contributed to a delay in diagnosis, because nonspecific or spurious laboratory results or diagnostic imaging findings led to further delay in management.
Many of the patients had been admitted to other hospitals, with management usually directed toward multiple investigations, with the diagnosis usually still uncertain, and no multidisciplinary approach evident. Only a very few children had received inpatient therapy (average 1 hour daily) before admission to our unit. Although 10 patients (53%) had received outpatient physiotherapy, this was generally a maximum of 1–2 times weekly and indicates that this level of input was inadequate to manage the condition.
Many of the children had been treated with a period of immobilization; all of these children reported an increase in symptoms (58). We think this indicates the importance of early specific diagnosis and intensive multidisciplinary team input, and demonstrates the importance of mobilization of the affected extremity, as has been described by others (7, 56).
Of the 19 children that we followed for >6 months, a few had a recurrence of symptoms following initial complete recovery (recovery lasting for 3 months). This was classed as a new episode of RSD, but the family, with a home management program and the support of the local therapy team, managed all these recurrences. This relatively high frequency of recurrence of RSD features after initial recovery has also been reported in other studies (7). However, if the relapse had occurred only once and recovery continued to completion, this was still classified as a complete recovery (47, 51). Five of the patients relapsed 2 or more times, with 1 reporting more than 4 relapses and 2 developing clinical features in another limb. This may indicate that once this condition has occurred, there may be a tendency for the limb to react in the same way to other minor traumas or for the predisposition to exist in other limbs.
Considerable evidence supports the approach of using a multidisciplinary team rehabilitation therapy program to improve the outcome of these children (59) because only 3 of 23 patients made minimal or no recovery. We would also advocate the “demedicalization” of care by encouraging another professional to lead the management of these children.
We found no difference in time from diagnosis to complete recovery for children managed either as inpatients or outpatients; however, the inpatients had worse symptoms at diagnosis, which would have had an impact on this time. There was a significant difference in the time to definite improvement between inpatient and outpatient management. Children who received inpatient therapy made a definite improvement in half the time (2.8 weeks compared with 6.2 weeks; P = 0.058). Given that, in general, there was a bias toward admission of more severe cases, this provides further evidence that inpatient, and perhaps more importantly, regular intensive therapy is an effective management for RSD. It was also important that early diagnosis seemed an important predictor of outcome with full recovery, because there was a significant correlation between time to diagnosis and time to recovery.
Subsequent to the retrospective study and associated audit, our management program was altered to address some of the issues that arose. From the data in this and other studies, it seems evident that there is either a considerably greater degree of psychological dysfunction present in many of the families than first appreciated. Thus the team psychologist is now included as part of the management from the onset, rather than initiating a referral when major problems become evident (4, 20, 47, 53, 54, 60). This has been further enhanced by the development of a dedicated preadmission clinic. After initial diagnosis, the young person is referred to this clinic where the physiotherapist, occupational therapist, and clinical psychologist fully assess him or her, and a management plan is developed with the young person and their family. Goals for the treatment program are discussed and decided with the young person, which is vital for maximum adherence to the program. The impact of this revised program on outcome is currently being investigated and will be compared with the cohort reported here.
RSD is an important diagnosis in childhood and can cause considerable pain and morbidity. Although this diagnosis is rare, it should not be overlooked because a delay in diagnosis can cause a delay in recovery.
We strongly think that there is only a limited role for regional or sympathetic nerve blocks or narcotic analgesia in the management of RSD in children. We further think that immobilization of the limb in this condition is inappropriate and should be considered contraindicated.
Appropriate management should include early diagnosis, avoidance of overinvestigation, and rapid instigation of a multidisciplinary team approach. This should encompass an aggressive active exercise program with the aim of regaining full movement and function, desensitization of the painful area, and psychological support and intervention for the young people and their families.
In summary, we feel that a fully multidisciplinary program will benefit the young people and their families (61) and that treatment should begin immediately, thus increasing the likelihood of full recovery. The program we describe requires dedicated staff, adequate time allocation, and regular team meetings to enhance communication and support.