Design: Nationwide propensity-matched cohort, Korean insurance database
Population: 20,707 matched pairs of TKA patients with and without Parkinson disease
Primary outcome: 5-year cardiovascular, cerebrovascular and periprosthetic events
Effect: Cerebrovascular aHR 1.11 (1.07-1.16); no rise in fracture or infection

This South Korean national insurance study followed unilateral knee replacements from 2010 to 2018 for five years, excluding patients with known cardiovascular or cerebrovascular disease. It propensity-matched 20,707 patients with Parkinson disease diagnosed before surgery to patients without it.

Parkinson disease was associated with a slightly higher five-year risk of cerebrovascular events (aHR 1.11, 95% CI 1.07 to 1.16), but not of cardiovascular events, periprosthetic fracture or prosthetic joint infection. In those diagnosed within five years of surgery, cardiovascular (aHR 1.05) and cerebrovascular (aHR 1.14) risks were slightly higher; in those diagnosed more than five years before, neither was raised.

The authors caution against routine vascular screening simply because a patient has Parkinson disease. The more useful reassurance for surgeons is the absence of excess periprosthetic fracture and infection. The vascular differences are small and may reflect unmeasured frailty or autonomic disease.

Do not decline knee replacement on Parkinson disease alone; fracture and infection risks were not raised.
Ask about orthostatic dizziness and falls before surgery in patients with Parkinson disease.
Consider a medical review for those with labile blood pressure or autonomic symptoms.
Counsel about a small rise in stroke risk over five years, not a large one.

Why it matters

Surgeons often hesitate to replace knees in Parkinson disease, and this addresses the complications they fear most.

The statistics, in plain English

A hazard ratio of 1.11 means about an 11% relative increase in the rate of cerebrovascular events, which on a low baseline is a small absolute difference. Large databases detect small effects with narrow intervals; statistical significance here does not mean clinical importance.