1.2.1

Constitution of the Steering Committee, formulation of the questions and designation of the pairs of literature reviewers

The SOFMER and the AFIGAP first designated a Executive Committee for the conference (Chairperson: BPV; Scientific Coordinator: MV; Secretary: TA; Scientific and Methodological Advisor: MR; see the list of authors).

The second step consisted of the constitution of a Steering Committee with representatives of the SOFMER and the AFIGAP ( Table 1 ).

The Steering Committee’s first meeting (at Cochin Hospital in Paris) enabled the formulation of 12 questions (some of which had sub-questions) judged to be relevant and which were inspired by the Agency for Healthcare Research and Quality’s (AHRQ) September 2001 report on the “Management of Chronic Neuropathic Pain following Traumatic Spinal Cord Injury” . The next step involved the constitution of a Scientific Committee ( Table 2 ), composed of designated SFETD and SFNC members and all the Steering Committee members. Lastly, the Scientific Committee designated a pair of literature reviewers for each question, whose medical and scientific specialties were chosen according to (i) the nature of the question and (ii) balanced representation of the sponsoring learned societies.

The following 12 questions were chosen:

• Q1:
  

  How can one evaluate chronic neuropathic pain in the SCI patient? Identification, classification and evaluation (excluding paraclinical exams).

  
  
• Q2:
  

  Which paraclinical investigations have the greatest diagnostic value in neuropathic pain in the SCI patient?

  
  
• Q3:
  

  What are the risk factors for the occurrence and chronicity of neuropathic pain in the SCI patient?

  
  
• Q4:
  

  What is the efficacy of systemic drug treatments?

  
  
• Q5:
  

  What is the efficacy of local and regional therapies?

  
  
• Q6:
  

  What is the efficacy of physical therapies?

  
  
• Q7:
  

  What is the efficacy of psychobehavioural management?

  
  
• Q8:
  

  What is the efficacy of social and environmental management?

  
  
• Q9:
  

  What is the efficacy of spinal stimulation?

  
  
• Q10:
  

  What is the efficacy of brain stimulation?

  
  
• Q11:
  

  What is the efficacy of surgical treatments other than central neurostimulation?

  
  
• 11a:
  

  DREZotomy

  
  
• 11b:
  

  other types of surgery (trauma site surgery, syrinx treatment, nerve root or trunk compression)

  
  
• Q12:
  

  What are the most effective organisational models?

1.2.2

Literature selection and analysis

An exhaustive literature search was performed by staff at the University of Nantes Library (Healthcare Section). Three databases were queried: Pubmed, Biological Abstracts and Pascal. The search was limited to publications in English and French. Lastly, the terms “human”, “clinical trial” and “adult” were used to focus more precisely on the objective. The initial search covered a period of ten years. The literature search was performed by combining the key words “spinal cord injury” and “pain” and “neuropathic” and “pain”. Logical combinations of more specific keywords were then necessary: “deep brain stimulation”, “transcranial magnetic stimulation”, “motor cortex stimulation”, “spinal cord stimulation”, “DREZotomy”, “spinal injections”, “post-traumatic syringomyelia”, “neuropathic pain”, “ulnar nerve compression syndrome”, “carpal tunnel syndrome”, “complex regional pain syndrome” and “spinal cord injury”.

As a result of this initial literature search, 934 abstracts were reviewed by BPV and 451 were subsequently selected (while excluding fundamental research reports). A second literature search based on the AHRQ report (which goes back several decades) yielded 591 references for analysis (286 of which were selected). In all, 737 articles were selected (BPV), sorted according to the question addressed (BPV, MV and TA) and then sent to the pairs of reviewers (either on a CD with listings and links or on paper).

The designated literature was independently examined by each reviewer. The data extracted from each article was graded for the level of proof according to the ANAES criteria .

1.2.3

Harmonisation of the levels of proof and guideline grading

Two one-day meetings were organized and moderated by the Scientific Committee in order to harmonize the levels of proof and grade the guidelines, with a view to producing a set of consensus guidelines for clinical practice. Lastly, questions for evaluating professional practice among French-speaking specialists were developed by each pair of reviewers and then validated by the Scientific Committee.

1.2.4

Presentation of the conclusions and guidelines

The conclusions and guidelines for each question were presented by one expert from each pair of reviewers during the SOFMER conference (Saint Malo, October 2007) on a day dedicated to expert consensus (October 3 ^rd ). During the meeting, the questions aimed at validating professional practice were put to the 116 physicians present, with electronic voting. Moreover, 50 practitioners replied via the Internet. Ninety-two percent of the respondees (total n = 166) were physical medicine and rehabilitation (PMR) specialists, with anaesthetists, pain specialists and neurosurgeons comprising the remainder. The entire discussion and all comments were noted by the session’s secretary.

1.2.5

Elaboration of the guidelines with respect to professional practice and levels of proof in the literature

The guidelines published here result from summary of this exhaustive literature review and the analysis of professional practice. Prior to publication, the guidelines were reviewed by a multidisciplinary review committee ( Table 3 ) composed of representative paramedical staff (an occupational therapist, a physiotherapist and a state registered nurse), a social worker, neurosurgeons, PMR specialists, a general practitioner and a patient representative appointed by the Association des Paralysés de France.

1.2.6

Conclusion and acknowledgements

This literature review notably underlined the lack of grade 1 studies in the field of non-pharmacological treatment of chronic neuropathic pain in the SCI patient. It is essential to perform rigorous, well-designed prospective studies in this area in the future.

After this lengthy work, we particularly wish to thank the reviewers & authors, the SOFMER (for its financial support for the AFIGAP) and Bénédicte Clément from the Centre Propara (for producing all the translations).