Clinical Screening of Acute Ruptures of the Anterior Cruciate Ligament of the Knee

A observational, Prospective, non-randomized, multi-centric study. The main objective of this study will be to determine the diagnostic performance (sensitivity and specificity) of a functional score for the detection of a rupture of the knee ACL in the immediate post-traumatic period.

It will follow the methodology for the evaluation of a new diagnostic test according to the STARD (Standards for Reporting of Diagnostic Accuracy) checklist.

Phases of the study

• Build score This phase will quantify the weight of each item and sub-items of the questionnaire already conceived from a statistical analysis adapted and thus to give a value to each item then a maximum value to the score.
• Validation of the score The score obtained will be validated in a different population versus MRI to determine its performance parameters. This score will be called the ACL-TS screening score (Tear Screening). An inter-evaluative reproducibility analysis of the score will be carried out in the investigative centers which have an emergency department
• Planned sample size For phase 1, a logistic regression will be performed. Empirically, there are 10 patients per variable for this questionnaire with 17 sub-items, 170 patients.

For Phase 2, the workforce was calculated on sensitivity because it is important to detect LCA ruptures. To obtain a sensitivity of 98% with a 2% CI, and taking into account a 10% rate of incomplete records, 210 patients should be included.

For the inter-evaluative ancillary reproducibility study, to obtain a Pearson coefficient of 0.8 (strong correlation if coefficient greater than 0.5) with a confidence interval of 0.67 to 0.88 and a risk of first 60 patients should be included.

Statistical analysis

• Analysis of normality of distributions
• Descriptive statistics: mean, standard deviation, minimum, maximum and median (quantitative variables); Frequency and percentage (qualitative variables).
• Comparisons of means: Student or Mann-Whitney according to the normality of the distributions
• Percentage comparisons: chi 2 or Fisher tests, idem
• The threshold of significance of the set of tests will be 0.05

Phase 1:

• A bi-varied study will then be carried out for each of the factors taken individually (search for an individual relationship, a rupture of the LCA confirmed by MRI vs the factors to be taken into account). These analyzes will provide an OR, an 95% CI of the OR, and a p-value measuring the severity of the association between the severe sprain and the factor taken into account.
• Follow a multivariate analysis (here logistic since the dependent variable will be "breaking ACL O / N"). To carry out this logistic regression, all factors for which p <0.25 will be included in the model. A descending step-by-step method will arrive at the final model which will contain only those factors whose p will be less than 5% and whose 95% CI of the OR will not contain 1. Possibly it will be possible to adjust The model on certain parameters and to look for possible interactions. The ORs and IC95% obtained will allow the gravity scores to be established.

Phase 2:

• Performance parameters of the score with respect to the MRI: ROC curves with determination of the optimal threshold value, calculation of the area under the curve (AUC), Se, Sp, the likelihood ratios positive and negative and the proportion of Subjects.
• Identify the factors that can modify the sensitivity of the score (age, sex, sport level, laxity, meniscal or cartilage lesions) by bi-varied analyzes between the group correctly classified by the MRI and the score And on the other hand the group presenting a discrepancy between the reference examination and the score. If several factors are identified, these analyzes may be supplemented by a multivariate logistic regression analysis
• Pearson correlation coefficient and intra-class correlation coefficient to evaluate all correlations between quantitative variables (score reproducibility, differential score / laxity)
• Model the diagnostic approach including the score by developing a decision tree

Inclusion procedure The study will target 380 patients seen in an emergency. After a knee sprain, patients usually go to hospital or a treatment center for medical consultation. For the study, patients will be recruited either in the emergency department if the investigating center has an emergency department or in the corresponding medical-surgical centers of the investigating center. Investigating surgeons at the centers should inform physicians of the initiation of the study and the modalities of implementation.

Usually, patients are received by an emergency physician, orthopedic surgeon on duty or a sports physician. A standard radiographic assessment will be performed to eliminate the presence of a bone fracture. If the patient meets the test selection criteria, he / she will be informed of the study and his / her right to oppose the collection of his / her medical data. The patient will ask all the questions necessary for his understanding before confirming his non-opposition. The questionnaire will then be completed by the physician with the patient. In the centers where the reproducibility study will be carried out, a second doctor will complete the questionnaire.

Once the consultation has been completed, the physician will issue prescriptions for the usual medical treatment, laximetry prescription (GNRB) to be performed between 3 to 6 weeks after the accident when the knee is indolent and the prescription of the MRI to be performed between 7 days and 1 month after the accident. The physician will inform the patient that they should consult with the investigator's orthopedic surgeon once the results of GNRB and MRI are available within 6 weeks of the accident.

During this delayed consultation, the orthopedic surgeon will realize in this patient the dynamic maneuvers of ligament testing. The study will end there for the patient. The surgeon will then follow his usual diagnostic procedure to decide on the future care of the patient.

Ethical aspects Agreement will be sought from the local Ethics committee. All patients will be fully informed and will be provided with an information form. The data will be entered into an internet page and will be analysed anonymously.