Work Prognosis of Complex Regional Pain Syndrome Type I

To the Editor: Complex regional pain syndrome type I (CRPS I) is a persistent functional disorder of the control of microcirculation of one or several joints by the autonomic nervous system, usually secondary to trauma.1 This unpredictable disease2,3 is responsible for disabling sequelae with an impact on the patient's ability to return to work,4 as about 30% of the patients are unable to return to work. No published case series has assessed the work prognosis of patients with advanced CRPS I.4 We aimed to assess the medical and work outcome of patients managed in one of the occupational health centers in the Paris region following CRPS I and to analyze the factors associated with a favorable or unfavorable prognosis. This was a multicenter observational study conducted in occupational health centers of the Paris region (Raymond Poincaré, Cochin-Hôtel-Dieu and Fernand Widal public hospitals, and the Centre Hospitalier Intercommunal de Créteil). All patients, for whom the presenting complaint (principal or comorbidity) was CRPS I, were included. Clinical criteria used were described by Veldman et al in 1993,5 considering that no consensus had been reached concerning those described by Bruehl et al and Galer et al.6–8 Data were extracted from their medical charts for visits held between 2001 and 2006. Missing data were collected by means of a patient telephone questionnaire (with a maximum of three reminder calls to the patient) or by calling their respective doctors (attending physician or occupational health physician). Patients unable to answer the telephone questionnaire (invalid telephone number, poor comprehension, impaired cognitive function), patients not in contact with their doctors, and patients refusing to give their informed consent to participate were excluded. Several types of variables were investigated in the medical charts: sociodemographic variables (gender; age [in three classes]; level of education), medical variables (site, upper limb versus lower limb; clinical signs, treatments received and comorbidities [dyslipidemia, hyperuricemia, diabetes, smoking habits]) and work variables (size of current company, a nonsedentary job [job involving walking]). Two endpoints were considered: return to work at the time of the study and the interval between the accident and return to work. The association between the variables studied and return to work was studied by univariate (Fisher exact test) and multivariate analyses (logistic model). In view of the power of the study, only univariate analyses were performed on the time to return to work based on significant variables identified by the logistic model (log-rank test). In 6 years, 55 subjects were selected (six patients without CRPS I were not included), and no patients were excluded. A total of 22 men and 33 women with a mean age of 48 years (range: 24–64 years) participated in this study. Return to work was observed in 68% of the patients of this population. The upper limb was affected significantly more frequently, especially the wrist (78%). Edema was the most specific positive clinical sign (Table). The treatments most often prescribed were analgesics and physiotherapy (66% of the patients who took this treatment returned to work).TABLE: Factors Associated With Return to Work (Yes/No)On univariate analysis, criteria predictive of return to work were young age, a sedentary job, upper limb involvement, high level of education probably allowing a change of job more adapted to the patient's disability or a more sedentary job, and early introduction of analgesic treatment associated with physiotherapy (Table). Lower limb involvement was significantly negatively correlated with return to work after adjustment for age, gender, sector of activity, and level of education. The average time to return to work of 14.8 months (±11.5, 9 missing data) was also significantly longer for patients with lower limb involvement. The results of the descriptive analysis of this population in terms of return to work are consistent with the rare data of the literature on this subject, despite the small number of subjects. The occupational health centers only see patients presenting diagnostic difficulties of work-related diseases and problems concerning maintenance of employment and they therefore tend to overestimate the medical and social severity of the lesions observed. However, perfectly comparable data concerning return to work were observed with, in particular, 68% of the patients returning to work as in the studies by Thevenon et al (73%)9 and Grunert et al (70%).10 In the study by Paraskevas et al,11 the recovery rate was even 100%. The time to return to work also appeared to be consistent with the results reported by Dauty et al,12 with an average of 10.5 months.13 The definition of return to work (under what conditions—permanent or temporary?) can also be questioned, as for all studies on this subject. The site most frequently affected in this study was the wrist, as in the study by Geertzen et al.14 The comorbidities studied appeared to have a negative impact on return to work, although these results are difficult to interpret in view of the small sample size of workers presenting these comorbidities. Thevenon et al also identified diabetes as a pejorative factor for return to work.9 One of the limitations of this study is that other relevant characteristics, such as the effect of psychological factors, were not studied due to the retrospective design, based on review of medical charts.15 Lower limb lesions were associated with a lower rate of return to work than upper limb lesions with a longer duration of time off work. This result has only rarely been reported,4 but could be a useful risk factor if confirmed in other populations, as lower limb involvement limits the subject's functional capacity for a large number of nonsedentary jobs (such as access to work), while upper limb involvement only affects manual work. Complex regional pain syndrome type I appears to be responsible for disabling sequelae, but 68% of the patients in this population returned to work, especially when the lower limb was not involved. These results justify studies in other types of populations. Sophie Dumas, MD Bertrand Pichon, MD Anne-Claire Dapolito, MD AP-HP, Occupational Health Unit Poincaré University Hospital Garches, France Lynda Bensefa-Colas, MD AP-HP, Occupational Health Unit Cochin University Hospital Paris, France Pascal Andujar, MD, PhD CHIC-Université Paris Est-Inserm U955 Occupational Health and Pneumology Unit, Créteil, France Antoine Villa, MD, PhD AP-HP, Occupational Health Unit Fernand Widal University Hospital Paris, France Alexis Descatha, MD, PhD AP-HP, Occupational Health Unit Poincaré University Hospital Garches, France UVSQ, UMRS 1018, Inserm Centre for Research in Epidemiology and Population Health, Population-Based Epidemiological Cohorts Research Platform Villejuif, France