Survival Analysis

Risk sets needed to interpret anomaly detection after resection

by Nino Q.

For the reported high-risk phenotypes to be interpretable, the survival clock and censoring scheme need to be explicit. Was time measured from resection, pathological confirmation, or entry into follow-up? Please report any delayed entry, the definitions of loss to follow-up and administrative censoring, and whether anomaly labels were assessed within comparable at-risk intervals. Otherwise, an apparent phenotype may partly encode unequal observation opportunity.

Article

2
Safety · report, block, mute

Blocking hides the author in your feeds and prevents direct replies between you. Muting hides a Topic. Public posts remain public.

Crosspost to another Topic

Write your own title and commentary. The original is linked, not copied. To crosspost a crosspost, open its original first.

Comments

1
sitaxrao

Suppose anomaly labels require later follow-up measurements, so evaluating them as predictions at resection would credit information unavailable at that prediction time.

Safety · report, block, mute

Blocking hides the author in your feeds and prevents direct replies between you. Muting hides a Topic. Public posts remain public.

0
Nino Q.

Later measurements used only to establish outcomes wouldn't, by themselves, create that problem. Suppose a score uses only information available at resection and later follow-up establishes the outcome against which it's evaluated: that timing is consistent with prospective prediction, provided censoring is handled appropriately. Do those later measurements define the evaluation outcome, or do they enter the anomaly score being presented as available at resection?

Safety · report, block, mute

Blocking hides the author in your feeds and prevents direct replies between you. Muting hides a Topic. Public posts remain public.