Patients with external fixators are at risk of pin site infection. A more objective assessment of possible pin site infection is warranted, particularly for future home-based monitoring of pin sites. The aim was to determine if thermography can detect signs of inflammation around pin sites by 1) Establishing a maximum temperature cut-off value 2) Investigating the correlation between local temperature and visual signs of inflammation 3) Adjust for anatomical location and ambient room temperature. This was a cross-sectional international multi-center study following STROBE guidelines. All patients with external ring-fixators scheduled for a visit in the out-patient clinic were eligible. Visual signs of inflammation were categorized using the Modified Gordon classification System (MGS, simplified sMGS). Thermographic imaging was done with an infrared camera (FLIR T540) and the maximum temperature within the ROI (MaxTp) was the primary outcome measure. Sample size and reliability were estimated. Cohen-Kappa, ROC-curve/AUC and Poisson regression were used for statistical analysis.Introduction
Method
There is no consensus on how to evaluate and grade pin site infection. A precise, objective and reliable pin site infectious score is warranted. The literature was reviewed for pin site infection classification systems, The Modified Gordon Score (MGS) grade 0-6 was used. The aim was to test the reliability of The Modified Gordon Infection Score. The observed agreement and inter-rater reliability were investigated between nurse and doctors. MGS was performed in the outpatient clinic at Aalborg University Hospital, Denmark on 1472 pin sites in 119 patients by one nurse and one of three orthopaedic surgeons blinded to each other's judgement. The data was stored in a Red Cap Database for further statistical analysis. The observed agreement between the nurse and the 3 orthopaedic surgeons was evaluated with a one-way random-effect model with interclass correlation with absolute agreement. Furthermore the observed agreement for each of the 3 surgeons with the nurse was calculated. The distribution of MGS infection grade in the 1472 pin sites was: Grade 0; n=1372, Grade 1; n=32, Grade 2; n=39, Grade 3; n=24, Grade 4; n=5, Grade 5; n=0, Grade 6; n=0. The observed agreement between the nurse and the surgeons was calculated as 98%. The ICC estimated between nurse and the surgeons was 0,8943 (ICC >0,85 = reliable). The grading was done by three different doctors with an agreement with the nurse as follows. Rater1 (n=416) =99,5 %, Rater2 (n=1440) =97,4%, Rater3 (n=1440) =96,6%. A limitation to this study is that the dataset represents mostly clean pin sites with MGS 0. Only 100 pin sites had signs of superficial infection MGS 1-4 none above 4. We found that the MGS infection score is highly reliable for low grade infections but we cannot conclude on reliability in severe infections.
Digital infra-red thermography may have the capability of identifying local inflammations. Nevertheless, the role of thermography in diagnosing pin site infection has not been explored yet and the reliability and validity of this method for pin site surveillance is in question. The purpose of this study was to explore the capability and intra-rater reliability of thermography in detecting pin site infection. This explorative proof of concept study follows GRRAS -guidelines for reporting reliability and agreement studies. After clinical assessment of pin sites by one examiner using Modified Gordon Pin Infection Classification (Grade 0 – 6), thermographic images of the pin sites were captured with a FLIR C3 camera and analyzed by the FLIR tools software package. The maximum skin temperature around the pin site and the maximum temperature for the whole thermographic picture was measured. Intra-rater agreement was established and test-retests were performed with different camera angles.Introduction and Objective
Materials and Methods
