In our validation study and in many other studies, the prediction rule has shown very good prediction as measured by the area under the ROC curve. Therefore, it will always show decreased diagnostic performance in new data sets. Do you have any theories as to why it performed less well in certain populations? A prediction rule is an multivariate model that is idealized to the initial derivation data set. Some subsequent prospective studies of the Kocher criteria have found much lower sensitivities than your original studies. However, the overall prediction performance is similar with CRP or ESR. Therefore, CRP (>2.0 mg/dL) could replace ESR in the prediction rule. CRP and ESR covary as they are measuring similar entities. Other studies, and our experience as well, has shown that CRP is an important predictor. As more data has come in, do you think some criteria should be weighed more heavily than others? Our original prediction rule utilized ESR as CRP was not available at that time. Your criteria currently weighs all predictors equally. If there is an effusion and there is a risk of septic arthritis, we usually aspirate under ultrasound guidance, although this may require sedation. If in doubt, aspirate the hip to make the definitive diagnosis! In our emergency department setting at Boston Children's Hospital, we routinely use ultrasound. If it’s a case of septic arthritis of the hip where missing the diagnosis of septic arthritis can have devestating consequences, I would recommend on erring on the side of caution. So I would still rely on your clinical judgement. Rather, they are designed as a tool in aiding diagnosis. What recommendations do you have for health care providers once they have applied the Kocher Criteria? Are there any adjustments or updates you would make to the score given recent changes in medicine, like CRP, ultrasound, or imaging? Prediction rules are not meant to replace clinical judgment. Also it was not developed in infants or adults, so it may not apply to these populations. It has been used for other joints such as the knee, however it was not developed for the knee and may not be valid. What pearls, pitfalls and/or tips do you have for users of the Kocher Criteria? Are there cases when it has been applied, interpreted, or used inappropriately? The prediction rule was developed for the hip in children. The goal of the prediction rule was to make the diagnosis in a more accurate, reliable, and timely manner. However, the differentiation is essential since septic arthritis is a surgical emergency with the potential for a poor outcome such as septic necrosis of the hip, whereas transient synovitis is treated with observation and has a generally benign outcome. This could often be a vexing diagnosis as the patients often presented similarly with atraumatic hip pain, limp, and fever. It was inspired from my clinical experience as a resident and fellow trying to differentiate between septic arthritis and transient synovitis of the hip in children in the emergency department. Why did you develop the Kocher Criteria? Was there a clinical experience that inspired you to create this tool for clinicians? The prediction rule was actually my thesis project for my MPH in Clinical Epidemiology from the Harvard School of Public Health.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |