= Primary graft patency
<Primary graft patency> measures how long a graft remains open without an intervention that terminates the primary-patency endpoint. It concerns the history since the original reconstruction, not merely whether a <bypass graft> is open on today’s image. Under the 1997 reporting standard, work confined to disease beyond the graft and its <anastomosis> can be an exception; an intervention on the graft to prevent its failure is not.<Recommended standards for reports dealing with lower extremity ischemia: Revised version by Robert B. Rutherford and others>[Printed p. 525, primary-patency definition and native-vessel exceptions]{ref}
For an explicit time-to-event model, define $T_p$ as the first qualifying loss of primary patency. Its survival function is
$$
S_p(t)=\Pr(T_p>t).
$$
Opening a narrowed but still patent graft by a qualifying preventive procedure can end <primary graft patency> even though blood never stopped flowing. Later rescue does not put that same graft back into the original primary-patency history. <Assisted primary graft patency> and <secondary graft patency> answer different questions about preserving or restoring function.
An original five-graft example demonstrates the denominator. All enter follow-up together. One has a qualifying intervention at month 2, another is last observed event-free at month 3, and a third occludes at month 4; the other two remain event-free through month 4. The <Kaplan–Meier estimator> gives
$$
\widehat S_p(4)=\left(1-\frac15\right)\left(1-\frac13\right)=\frac8{15}\approx0.533.
$$
The <graft patency risk set> falls after both the first endpoint and the later <right censoring>. The censored graft is neither declared a failure nor assumed patent indefinitely. Interpreting this estimate requires an appropriate independent-censoring assumption.
A report should identify the original procedure date, qualifying interventions, documented occlusions and observation limits. A status image alone cannot reconstruct every earlier event. <Primary graft patency> is therefore most informative alongside the intervention history and separately defined clinical outcomes, rather than as an unexplained percentage.
Back to article page
