CORRInsights®: Can A Multivariate Model for Survival Estimation in Skeletal Metastases (PATHFx) Be Externally Validated Using Japanese Patients?

    loading  Checking for direct PDF access through Ovid


Prediction models play an important role in informing patients about the future course of their disease and in aiding medical decision-making. For example, in a primary care setting, the Framingham cardiovascular risk score is used to identify patients who meet indications to receive cholesterol-lowering and antihypertensive drugs [6, 7]. In a surgical-oncology setting, patients undergoing resection for a soft-tissue sarcoma who have a low risk of local recurrence could be spared the short- and long-term complications of adjuvant radiotherapy, whereas those considered at high risk are commonly offered this treatment [10].
In their study, Ogura and colleagues performed external validation in a Japanese population of the PATHFx model ( that predicts 1, 3, 6 and 12-month postsurgical survival of patients with skeletal metastases. The PATHFx model was developed and validated for Western countries patients [2-4, 9], and the validation in an Asian population therefore represents a step forward in the evaluation of the applicability of model worldwide.
In this Japanese population, the PATHFx model showed a fair-to-good discrimination based on receiver operating characteristic curves, with the c index (the area under the curve, [AUC]) ranging between 0.75 and 0.83, comparable to previous development and validation studies (AUCs between 0.76 and 0.85). Interestingly, this occurred despite substantial differences in the characteristics of patients as compared to the original development sample, including a markedly better survival from 3 months and beyond.
    loading  Loading Related Articles