Evaluation of clinical tests used in classification procedures in pregnancy-related pelvic joint pain

Pain in the pelvic joints and lower back, a major problem for pregnant wome n, has proved resistant to precise measurement and quantification. To devel op a classification system, the clinical tests used must be able to separat e pelvic from low back pain; they must also have a high inter-examiner reli ability, sensitivity and specificity, and preferably be easy to perform. Th e aim of this study was to describe a standardised way of performing tests for examining the pelvis, and to evaluate inter-examiner reliability, and e stablish the sensitivity and specificity of 15 clinical tests. It was desig ned as a longitudinal, prospective, epidemiological cohort study. First, 34 pregnant women were examined by blinded examiners to establish inter-exami ner reliability. Second, a cohort of 2269 consecutive pregnant women, each responded to a questionnaire and underwent a thorough and highly standardis ed physical examination (15 tests with 48 possible responses) of the pelvic joints and surrounding areas. The 535 women who reported daily pain from t he pelvic joints and had objective findings from the joints were divided, a ccording to symptoms, into four classification groups and one miscellaneous group. The results of the study showed inter-examiner agreement of the tes ts was high, calculated in percentage terms, at between 88 and 100%. Using the Kappa coefficient, most tests kept the high agreement: six tests had an inter-examiner agreement of between 0.81 and 1.00, three between 0.61 and 0.80, and two between 0.60 and 0.41. Five tests showed superior sensitivity . The specificity of the tests was between 0.98 and 1.00, except the value for pelvic topography, which was 0.79. These results show that it is possib le to standardise examination and interpretation of clinical tests of the p elvic joints, resulting in a high degree of sensitivity, specificity and in ter-examiner reliability.