The Canine Postamputation Pain (CAMPPAIN) initiative: a retrospective study and development of a diagnostic scale

Objective To develop a scale to diagnose and assess the severity of postamputation pain (PAP) in dogs. Study design Single-center retrospective study. Animals A total of 66 dogs that underwent thoracic or pelvic limb amputation and 139 dogs that underwent tibial plateau leveling osteotomy (TPLO) at a veterinary teaching hospital. Methods An online survey regarding postoperative behavioral changes was sent to owners. Categorical, multiple-choice responses were entered into a univariable logistic regression model and tested for association with amputation using the Wald test. If p < 0.2, variables were forwarded to a multivariable logistic regression model for manual build. Model simplicity and predictive ability were optimized using the area under the receiver operating curve (AUROC) characteristic, and model calibration was assessed using the Hosmer-Lemeshow test. The selected model was converted to an integer scale (0-10), the Canine Postamputation Pain (CAMPPAIN) scale. Univariable logistic regression related each dog's calculated score to the probability of PAP. Results Multivariable logistic regression identified four independent predictors of PAP (p < 0.05): 1) restlessness or difficulty sleeping, 2) episodes of panic or anxiety, 3) sudden vocalization, and 4) compulsive grooming of the residual limb. Score AUROC was 0.70 (95% confidence interval = 0.63-0.78) with good calibration (Hosmer-Lemeshow statistic p = 0.82). A score of 2 corresponded to a risk probability of 0.5. Taking a score >= 2 to indicate PAP, score specificity and sensitivity were 92.1% and 36.4%, respectively. When this score was used to diagnose PAP, prevalence was 36.4% (24/66) and 7.9% (11/139) in the amputation and TPLO groups, respectively. Conclusions and clinical relevance Postamputation pain is characterized by specific postoperative behaviors and appears to affect approximately one-third of canine amputees. The CAMPPAIN scale generated from these data could facilitate diagnosis, treatment and further study of PAP but requires external validation.