How PATHFx works

By providing basic clinical and physiologic variables, clinicians are able derive a personalized, objective, and unbiased “survival trajectory” for each patient.

Shared decision-making. PATHFx v3.0 offers control by limiting the amount of uncertainty. The tool provides objective estimates of survival, while allowing clinicians with expertise to provide a weighted estimate. Clinicians, families and patients may then use this information to determine the best course of action. For instance, those with short life expectancies may benefit most from palliative radiotherapy or less invasive surgery including intramedullary nailing or other fixation techniques[1, 5, 10]. In contrast, patients with longer survival estimates are generally thought to require more durable reconstructive options that can be expected to last many years, but increase both the perioperative risk and the duration of rehabilitation[2, 4, 8, 9, 11].

What’s different in PATHFx v3.0?

In version 2.0, we re-imagined how the PATHFx tool could integrate with providers, electronic health records, registries and industry. We transitioned away from expensive, oudated, proprietary software to open source architecture in an effort to stimulate collaboration, ease compatibility and streamline the development of new models. In doing so, we re-created the PATHFx models from scratch, using machine-learning code, developed in-house, specifically for this purpose. Version 3.0 represents the first of many modernization updates to ensure PATHFx remains responsive to emerging clinical challenges.  The new models incorporate additional patients who were treated between 2015 and 2018, while maintaining nearly two decades of clinical experience used to create the original PATHFx models. We updated the machine learning algorithms to keep pace with the ever-evolving industry standard, and fine-tuned the imputation strategies to better account for missing data in this unique patient population.