Desenvolvimento de método palpatório da coluna cervical e sua validação baseada em imagem radiológica

Abstract

Health care professionals who use palpation in rehabilitation share the need of an accurate and reliable clinical examination; they use manual procedures and interventions that are based on the complete understanding of the in vivo anatomy and on palpation and skills for location of anatomical structures. In particular, most of the cervical spine palpatory methods used have not been systematically studied and therefore, their accuracy and reliability remains without evidence. The goal of this study is to develop a palpation method for locating the spinous process of the seventh cervical vertebra (C7SP) and its validation based on radiological imaging. Methods: This study has cross-sectional observational design with non-probability sampling and was divided into three phases. Preliminary phase: radiologic imaging evaluation obtained from radiographs, CT scans, MRIs and consulting the literature for anatomical clues in order to develop a new palpation method of bone structures of the cervical spine; evaluation of anthropometric proportions to adjust the method to population diversity. Phase I: Comparative analysis between the best existing method (method flexion-Extension, FEM) and the new method (Thorax–rib static method - TRSM), using chest X-ray for evaluation of diagnostic performance. Phase II: Development and validation of a multiple linear regression model using the TRSM and personal information. Results: The accuracy of C7SP identification was 18% and 33% (p = 0.013) using the FEM and TRSM, respectively. Cumulative accuracy considering both caudal and cranial directions (C7SP±1SP) increased to 58% and 81% (p = 0.001) using the FEM and TRSM, respectively. Age had a significant effect on the accuracy of the FEM (p = 0.027), but not on the TRSM (p = 0.939). Sex, body weight, height and BMI had no significant effect on the accuracy of both methods. The intrarater reliability of the radiologist for the location of the markers was excellent (Κappa = 0.962, 95% CI = [0.888; 1.000]). The linear multivariate prediction model was designed, the distance vertex- spinous process C7 (cm) generated the equation 𝐷𝑉−𝐶7 = 0.986 × 𝐷𝑉−𝑀 + 0.018 × 𝑚𝑎𝑠𝑠 + 0.014 × 𝑎𝑔𝑒 − 1.008 (adjusted R2 = 0.797, SE of bias = 0.025 cm). The accuracy of identification of SPC7 in the development phase of the multiple linear regression model using the TRSM was 40% and the accuracy of PEC7 identification using the TRSM associated with the multiple linear regression model was 66% (p = 0.001). Discussion: The best accuracy of TRSM for locating the C7SP can be explained by the fact that it is a static method of palpation and therefore may be less prone to error due to cervical mobility limitation than the FEM method. The palpation was the most influential variable for the accurate location of C7SP. Age and body mass were also independent predictors to locate accurately the C7SP. Using objective adjustments based on personal information such as age and body mass could increase the accuracy of palpation. Conclusion: The location of C7SP is more accurate using the TRSM than the method of flexion-extension. The accuracy of the C7SP location is further enhanced with the use of multivariate prediction model in conjunction with palpation procedure of TRSM than the use of isolated palpation method. Keywords: palpation, reliability, rehabilitation.