Summary

This pilot focuses on evaluating the Cambridge Knee Injury Tool (CamKIT), which is designed to improve the diagnosis and management of soft tissue knee injuries, demonstrating its potential to expedite care and reduce healthcare inefficiencies without missing clinically significant injuries.

Abstract

Soft tissue knee injuries (STKIs) remain a prominent issue in healthcare. While there has been an increased focus on injury prevention and rehabilitation, escalating rates of injury and preference for surgical intervention underscore the growing need to optimise the diagnostic and management pathway of STKIs.

The Cambridge Knee Injury Tool (CamKIT) is a novel clinical prediction tool developed as a 12-point scoring tool involving patient factors, external factors, and signs and symptoms. The tool was tested on a retrospective cohort involving 229 patients presenting to a major Regional Trauma Center with acute knee pain over 3 months. Data on the twelve variables, injury outcomes and details of the treatment pathway were extracted from the Electronic Medical Records (EMR). Conclusions were made about the model utility based on potential changes to time-to-event for consult, MRI, and surgery and comparisons of resource allocation in discharge, consult, and MRI rates.

In total, 70 clinically significant injury events were recorded, with 44% (31/70) occurring in isolation and 56% (39/70) being concomitant. ACL injuries were recorded in 54% (38/70) of events, with 90% (34/38) of all ACL injuries being concomitant. Collateral ligament injuries were prevalent with 33% (23/70) attributed to MCL injuries and 16% (11/70) attributed to LCL injuries. 100% of collateral ligament injuries were concomitant. Meniscal injuries were common, with 39% (27/70) being the medial meniscus and 21% (15/70) being the lateral meniscus. Of the 229 individuals who presented to A&E with acute knee pain, only 30% (70/229) had a confirmed injury and 17% (39/229) required surgery. Of the 116 individuals who were referred for specialist consultation, 59% (68/116) had a confirmed injury and 34% (39/116) required surgery. A total of 37% (84/229) of the cohort underwent MRI, with 76% (64/84) having a confirmed injury and 46% (39/84) requiring surgery. The CamKIT yielded a median score of 7.5 (IQR = 3) in the injured cohort, compared to a median score of 2 (IQR = 3) in the non-injured cohort, with a statistically significant difference (p < 0.0001). When constructed as a three-tier risk stratification tool, the CamKIT produces a sensitivity of 100%, a specificity of 94.3%, a positive predictive value of 89%, and a negative predictive value of 100%. Implementing the CamKIT model would expedite the imaging pathway for 70% of patients to initiate the surgery pathway faster for 85% of patients. The CamKIT model also reduced total specialist consultations by 34%. Crucially, these improvements were achieved without missing any clinically significant injuries.

The low proportion of clinically significant injuries among the cohort presenting to A&E and the subsequent inefficiencies in the referral pathways underscore the critical need for accurate and accessible diagnostic tools in acute settings. The novel analysis of the CamKIT advocates for developing a tool that empowers primary healthcare workers in challenging scenarios by instilling confidence and promoting accuracy in clinical decision-making. The CamKIT also promotes efficiency in the secondary healthcare setting, enabling more targeted and timely use of specialist resources and reducing patient wait times and healthcare costs.