Measurement properties of the sagittal craniocervical posture photogrammetry

Commonly in clinical settings, the patient's posture is visually evaluated by the clinician using anatomical landmark references. However, this measurement is subjective and not quantifiable. Photogrammetry to assess posture was thought to be a possible good clinical alternative to the other methods because it is noninvasive, quantifiable, and less expensive. However, more tests were needed to determine its validity. This study tested the reliability and the validity of five angles measuring craniocervical posture using photogrammetry. Radiographs and photographs of the craniocervical posture of 39 healthy-female subjects were taken in a standardized sagittal standing position. Markers were placed on the back of the subject's neck and ear. A second photograph and radiograph was taken 1 week later using 21 of the 39 subjects to test reliability. The angles were analyzed using AlcimageRTM software. Intraclass-correlation coefficient and standard error of measurement was used to test the reliability. Concurrent validity was tested using Pearson correlation and regression analysis. Discriminant analysis was used to test the discriminant validity. Sensitivity/specificity and predicted values were also calculated. The results showed that photogrammetry ICC values were good to excellent when assessed by 2 raters (ICC=0.89-0.99). The posture of the subjects was reproducible when tested using radiographs (ICC=0.89-0.98). One rater was reliable in reattaching the markers (ICC=0.71-0.91) and precise in locating the reference spinous processes (87.8%). Craniovertebral angle (CVA) appeared to be valid in measuring the position of the head in relation to the cervical spine (r=0.84) and to be able discriminate subjects with aligned posture, slight forward head posture (SFHP), and forward head posture (FHP) assessed by 1 rater (84.6% correctly classified). Cervical inclination angle (CIA) appears to be valid in discriminating subjects with aligned and FHP (86% and 88% respectively) but moderate to predict the cervical spine inclination. The cervical lordosis angles were not able to discriminate postures and predict the cervical lordosis. CVA and CIA were able to detect postural differences through the sensitivity/specificity and predicted values analysis. This study supports the validation of CVA and CIA to assess craniocervical posture which may improve the ability of the clinician to detect and quantify craniocervical postural alterations.