Article Text

Download PDFPDF

Complex regional pain syndrome in children: incidence and clinical characteristics
  1. Hashem Abu-Arafeh1,2,
  2. Ishaq Abu-Arafeh2
  1. 1University of Edinburgh, Edinburgh, UK
  2. 2Department of Paediatrics, Forth Valley Royal Hospital, Larbert, Stirlingshire, UK
  1. Correspondence to Dr Ishaq Abu-Arafeh, Department of Paediatrics, Forth Valley Royal Hospital, Larbert, Stirlingshire FK5 4WR, UK; iabu-arafeh{at}


Objective To study the clinical and epidemiological characteristics of complex regional pain syndrome (CRPS) in children.

Patients and methods All children and adolescents under 16 years of age with a new diagnosis of CRPS who were reported to the Scottish Paediatric Surveillance Unit were included. Patients’ recruitment ran between 1 November 2011 and 31 October 2015. Information was collected on patients’ demography, clinical features, investigations, management and impact of disease on child and family. The diagnosis of CRPS was made on fulfilling the clinical criteria of the International Association for the Study of Pain.

Results 26 cases of CRPS were reported over 4 years, giving a minimum estimated incidence of 1.16/100 000 (95% CI 0.87 to 1.44/100 000) children 5–15 years of age. Nineteen patients were female (73%) and mean age at diagnosis was 11.9 (range 5.5–15.4 years). The median interval between onset of symptoms and diagnosis was 2 months (range 1–12). The majority of children have single site involvement, with legs been more often affected than arms and the right side is more often affected than the left. There was a clear trauma at onset of the illness in 19 children and possible nerve injury in one. All investigations were normal and several treatment modalities were used with variable success. The disease had significant impacts on the patients’ education and family lives.

Conclusions The estimated incidence of CRPS is 1.2/100 000 children 5–15 years old. The diagnosis of CRPS is often delayed. CRPS has a significant impact on children and their families.

  • Complex Regional Pain Syndrome
  • Reflex Sympathetic Dystrophy
  • Adolescents
  • Incidence
  • Causalgia

Statistics from

What is already known on this topic

  • Clinic-based case reports and case series are the only source of the clinical features of complex regional pain syndrome (CRPS).

  • Population-based studies suggest that <5% of patients with CRPS are children under the age of 20 years.

  • The criteria used for the diagnosis of CRPS are based on studies in adult patients.

What this study adds

  • Incidence of CRPS in children (5–15 years of age) is at least 1.2/100 000.

  • Adolescent girls are most likely to be affected.

  • The disease has a significant negative impact on the child‘s education, sports and family life.


Complex regional pain syndrome (CRPS) is a chronic debilitating disorder. It is characterised by continuous and severe pain affecting one part of the body, in most cases a limb following injury. The pain is disproportionate to the severity of the injury and cannot be explained on the extent of tissue damage alone. The pain is associated with autonomic features such as oedema, colour changes, temperature changes and allodynia. It may also be associated with muscle spasm, wasting or dystonia. CRPS causes considerable distress, pain and disability with significant negative impact on the quality of life of affected children. The cause of CRPS is not known and its pathogenesis is poorly understood. However, there is seemingly a complex interplay between genetic susceptibility,1 an excessive inflammatory response to injury2 and consequent central nervous system changes.3

The diagnosis of CRPS is made on fulfilling clinical criteria designed for adult patients and there are no specific criteria for children. There are no biochemical tests, biological markers or radiological findings to confirm the diagnosis. The clinical criteria of the International Association for the Study of Pain (IASP) were introduced in 1995 and offer an inclusive approach to cases of CRPS (box 1).4 The Budapest criteria, published, in 2007 were designed to be restrictive for inclusion of cases into research studies, but less strict for use in clinical practice (box 2).5

Box 1

IASP criteria for the diagnosis of CRPS4

  1. Develops after an initiating noxious event (type 1) or after a nerve injury (type II).

  2. Spontaneous pain or allodynia/hyperalgesia that is not limited to the territory of a single peripheral nerve and is disproportionate to the inciting event.

  3. There is or has been evidence of oedema, skin blood flow abnormality or abnormal sudomotor activity in the region of the pain since the inciting event.

  4. This diagnosis is excluded by the existence of conditions that would otherwise account for the degree of pain and dysfunction.

CRPS, complex regional pain syndrome; IASP, International Association for the Study of Pain.

Box 2

Budapest criteria for the diagnosis of complex regional pain syndrome (CRPS)5

The diagnosis of CRPS is made on fulfilling criteria A–D:

  1. The patient has continuing pain which is disproportionate to any inciting event

  2. The patient has at least one sign in two or more of the categories (all categories in research or trials)

  3. The patient reports at least one symptom in three or more of the categories

  4. No other diagnosis can better explain the signs and symptoms


  1. Sensory: Allodynia (to light touch and/or hyperaesthesia does temperature sensation and/or also qualify as a deep somatic pressure and/or symptom joint movement) and/or hyperalgesia (to pinprick)

  2. Vasomotor: temperature asymmetry and/or if you notice skin colour changes and/or skin temperature colour asymmetry: must be >1°C

  3. Sudomotor/oedema: oedema and/or sweating changes and/or sweating asymmetry

  4. Motor/trophic: decreased range of motion and/or motor dysfunction (weakness, tremor, dystonia) and/or trophic changes (hair/nail/skin)

The prevalence and incidence of CRPS in children are not known and data on epidemiology of CRPS in adults are limited. The incidence of CRPS in the general population in the USA (Olmsted County) is estimated at around 5.5/100 000 person-years and a subset analysis of incidence in children and adolescents under 20 years is estimated at 1.58/100 000 person-years.6

An epidemiological study of CRPS, on cases known to general practitioners over a 10-year period in the Netherlands, showed a higher incidence.7 Subset analysis shows an incidence of 5.2/100 000 person-years in children under 20 years of age.7

CRPS type I, also known as reflex sympathetic dystrophy, follows a minor injury to the affected limb and CRPS type II, known as causalgia, follows a nerve injury. In a small number of patients no definite history of injury can be identified. In patients with a lower number of signs or symptoms, or had documented CRPS signs/symptoms in the past, the disease is described as CRPS—not otherwise specified.8

Children with CRPS may present and receive treatment from different specialities including general paediatrics, orthopaedics, psychiatry, pain management clinics, paediatric neurology and paediatric rheumatology. Therefore, it is not uncommon for patients to be given the correct diagnosis of CRPS after several months and on occasions after several years from first presentation. The delay in diagnosis may adversely impact on the chronic course of the condition, including the child's function and emotion as shown in a systematic review of CRPS recovery and outcome.9

The clinical features in children are derived from case reports and case series from patients attending specialist paediatric clinics.10–16 It is not possible to be certain if these groups of patients make an accurate representation of the whole population of patients with CRPS.

The majority of patients are female in a ratio of 4:1 and the mean age at diagnosis is between 12 and 14 years. CRPS is extremely rare in preschool children, although patients as young as 2.5 and 3 years of age were reported.17 ,18 A history of injury is almost always demonstrated in affected children and the lower limbs are more often affected than the upper limbs, while the upper extremity was affected more frequently than the lower extremity in adults. Abnormal movements and dystonia have also been frequently reported in adults and children with CRPS and may complicate the clinical picture.19 ,20

This study aims to explore the incidence and the spectrum of clinical features of CRPS, and its impact on children, adolescents and their families.

Patients and methods

This is a prospective observational study of all children with CRPS in Scotland. The clinical criteria for the diagnosis of CRPS as defined by the IASP (box 1) were initially used for the diagnosis of CRPS when the study protocol was made in 2010, but patients were also assessed against the Budapest criteria (box 2).

Patients were identified by treating clinicians and were reported through the Scottish Paediatric Surveillance Unit (ScotPSU) in response to monthly email ( Around 600 clinicians are on the ScotPSU email list including almost all consultant and specialist paediatricians and paediatric subspecialists including rheumatology and neurology. The email list also includes paediatric accident and emergency physicians, paediatric orthopaedic surgeons, child psychiatrists and clinical psychologist working in Scotland. Clinicians who report new cases of CRPS are asked to complete a detailed structured questionnaire (see online supplementary appendix 1) in order to confirm the diagnosis and collect further clinical and demographic data anonymously. If clinicians do not respond after the first request, they will be sent up to two reminders over the following 3 months.

All patients under the age of 16 years and presenting between 1 November 2011 and 31 October 2015 were eligible for inclusion.

East of Scotland Research Ethics Committee approved this study as a service evaluation and approved by the Forth Valley research and development department.


Thirty cases were reported over a period of 4 years. In three cases, the reporting clinicians did not complete the questionnaire despite three reminders and in one the diagnosis could not be confirmed. In 26 patients, the diagnosis of CRPS was confirmed.

Epidemiology and demography of CRPS

Nineteen patients (73%) were female. Patients were between 5.5 and 15.4 years of age at onset of the disease with a mean age of 11.9 years (SD: 2.58, 95% CI 10.9 to 12.9). The childhood population of Scotland between 5 and 15 years of age, according to 2011 census, is 561 232. Therefore, the incidence of CRPS in children and adolescents between 5 and 15 years of age is estimated at 1.16/100 000 (95% CI 0.87 to 1.44/100 000) children per year.

Clinical features of CRPS in children and adolescents

The intervals between onset of symptoms and confirmed diagnosis varied between 1 and 12 months with a median of 2 months.

In 20 children (74%) the onset of CRPS followed a clear traumatic injury (CRPS type I), including one patient who also had haemophilia and suffered a minor trauma and right ankle haemarthrosis. In one child, CRPS followed a pulling injury of the wrist and forearm causing stretch/axonal injury of the ulnar nerve (CRPS type II), manifested with numbness over the little and the ring fingers (fourth and fifth digits). There was no definite history of injury in five patients.

All children had single site involvement at presentation; lower limb in 19 (73%) patients, upper limb in seven (27%). The right side was affected in 14 children and the left in 12. Ten children had an involvement of a second site. The secondary site was on the same side of the primary site in five and on the other side in five. The associated clinical features and characteristics are summarised in table 1.

Table 1

Clinical features in 26 patients with CRPS

Management of CRPS

Children were managed by several specialities and were reported to ScotPSU by general paediatricians (18 patients; 69%), paediatric rheumatologists (three patients), child psychiatrist/psychologist (2), paediatric neurologist (1), paediatric pain specialist/anaesthetist (1) and an orthopaedic surgeon (1).

Investigations included plain X-ray in 20, Doppler ultrasound in seven, CT in three, MRI scan in 13, erythrocyte sedimentation rate in 16, full blood count in 20, creatine kinase in six and liver and kidney function tests in all.

Common medical treatment included simple analgesia in all patients, codeine and opiate in 11 patients and pain-modulating agents in 16 (gabapentin, pregabalin, carbamazepine and amitriptyline) (table 2). Pain specialists (anaesthetists) managed five patients; two of whom received treatment with a nerve block and epidural anaesthesia.

Table 2

Treatment options in cases of CRPS

Impact of CRPS on children's life and education

Twenty-two children reported missing school days because of the pain and in one child the number of missed school days was described as significant. Twenty-three children stopped sports completely and information on one child was not available. Families of 19 children had to cancel social activities because of the child's pain and nine families had to cancel planned holidays. At least one parent of nine children had to take time off work to stay with the child due to CRPS. Stress in the family was reported by parents of seven children manifesting with parental distress in four, anxiety in one, excessive worry in one and changing work (work from home) by one father.


The choice for using the ScotPSU for this study was based on the well-defined population of Scotland and the excellent network of paediatricians within the Scottish Paediatric Society, which runs the ScotPSU since 2009. The ScotPSU is shown to have a reliable system with a good response rate and several ScotPSU studies have been completed and some were published.21 ,22 Patients with CRPS were reported to this study by clinicians from different specialties confirming the ScotPSU's wide reach.

The IASP's clinical criteria for the diagnosis of CRPS were chosen at the early stage of study design in 2010, when the Budapest criteria were still emerging and before becoming widely used. However, as the study progressed it was reasonable to assess the cases against the Budapest criteria as well. It is also worth noting that although both sets of criteria were designed for use in adult patients, there is no evidence to suggest that these criteria are not equally applicable in adolescents who constitute the majority of paediatric patients. However, investigations of children with possible CRPS may be different than in adults and treatment options available for children are limited when compared with adult patients.

Clinical guidelines are available on the assessment and management of CRPS in adult patients,8 ,23 ,24 but they are not for use in children. The lack of paediatric guidelines may explain the different referral pathways that may take patients to different medical specialities and may explain the reason for the occasional long delay in making the correct diagnosis in some patients of up to 12 months; a median of 2 months. Other reports have also shown a similar delay in the diagnosis of children with CRPS.11

The incidence of CRPS is around 1.2/100 000 children and adolescents between 5 and 15 years of age. This is close to the incidence in the USA (1.58/100 000) and about a quarter of the incidence in the Netherlands (5.2/100 000).6 ,7 The incidence of CRPS in Scottish children and adolescents is probably an underestimate of the true incidence, due to the nature of the surveillance study that may not capture all patients and there is an inevitable under-reporting of cases. Moreover, it is conceivable that some patients over the age of 13 years may have presented to adult services and have not been reported.

Clinical features of CRPS in Scottish children are in keeping with the reported clinical features in the literature. Our prospectively collected data confirm the uniformity of the clinical features and the severity of the disease as seen in specialist clinics. The diagnosis of CRPS is made on the basis of the typical clinical features and therefore, raising the awareness of the disease would be essential for an early diagnosis and commencement of appropriate treatment. Investigations are important to exclude other conditions such as venous thrombosis, infections, inflammatory disorders, connective tissue diseases and neoplasia.

Musculoskeletal traumatic triggers were documented in 74% of patients and stretch nerve injury in one patient. It is not impossible to assume that at least some of the other children may have had a minor or a trivial inciting trauma. Lack of traumatic trigger, therefore, should not deter from making the diagnosis of CRPS in the presence of typical clinical presentation. In one patient with haemophilia, CRPS followed haemarthrosis possibly due to minor trauma, but spontaneous intra-articular or intramuscular bleeding has been reported to trigger CRPS.25

A secondary site was not uncommon in our patients (10 patients; 38%). This is similar to the reported frequency of second or third site in adults and children. In five patients the extension of pain crossed to the other side.

Dystonia was common in our patients affecting 10 children (38%), in keeping with the reported incidence in the literature. In a study of 32 children with CRPS attending a paediatric service in the UK over 10 years, 10 patients (30%) also had dystonia.19 It is estimated that 25% of adult patients with CRPS also suffer from movement disorders mostly (20%) dystonia.26

The impact of CRPS on children's life and education was significant and is not unexpected. The severity of the pain and the loss of limb function have adverse impact on the children's ability to participate in sports and attend school. The negative impact also was felt by parents who themselves had to cancel social events and family holidays. Some parents had to give up work to look after their children, which may lead to economic loss to the family, adding to the distress caused by the disease. Children with CRPS were shown to report greater functional disability and more somatic symptoms than children with headache or back pain.27 However, they have normal scores for depression and anxiety.28 In regards to assessment of quality of life, a study of 73 children with CRPS showed that quality of life 4–0 score, at a mean of 37 months from disease onset, was significantly lower compared with controls.29

It was not possible to comment on the prognosis of childhood CRPS in this study. Long-term follow-up studies are needed.

Study limitations

Several limitations are inherent with this type of surveillance study. By its nature, our study depends on the voluntary reporting of cases by clinicians working in Scotland. Therefore, it is likely that some patients are missed or unreported and it is also possible that some patients may have presented to adult services. As such, it is reasonable to view the incidence of CRPS in this study as a credible minimum of the true incidence.

The description of the clinical features is likely to be accurate, but it will be impossible to know if mild cases or atypical presentations were missed or gone unreported especially if the cases were not assessed by clinicians with expertise in the condition.

Follow-up was limited for a period of 3–6 months and therefore, outcome data are not available.

Summary and conclusions

  • The estimated incidence of CRPS in children between 5 and 15 years of age is 1.2/100 000 (95% CI 0.87 to 1.44).

  • Girls are more often affected than boys at a ratio of 3:1.

  • Mean age of onset is 11.9 years and the disease is very rare under the age of 8 years.

  • Legs are more affected than arms and the right side is more affected than the left.

  • A second site may be involved in 40% of patients including the other side of the body.

  • The disease has a significant negative impact on the child’s education and sports and the family social life.


The authors are grateful to all clinicians who reported cases of complex regional pain syndrome and for their cooperation in providing clinical data.



  • Contributors IA-A: produced the concept, protocol, applied for ethics, supervised collection of data and took part in writing and reviewing the manuscript. HA-A: collected data, made analysis of data, helped in writing and reviewing the manuscript.

  • Funding The project received funding from NHS Forth Valley Research Endowment Fund.

  • Competing interests None declared.

  • Ethics approval East of Scotland Research Ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.