Table 2.

Definition of the primary composite endpoint of treatment failure at 6 months since randomization

Component	Definition
All-cause mortality	All deaths observed among study participants within 6 months since randomization.
IVROBA	Sudden serious clinical deterioration and cranial imaging suggesting intraventricular rupture of brain abscess after time of randomization.
Unplanned neurosurgery	Unplanned (re-)aspiration or excision of brain abscess after randomization. This does not include, e.g., planned insertion of ventriculo-peritoneal drain or corrective surgery of dural leak.
Relapse	a. Cranial imaging showing an increase in brain abscess volume (height × depth × width/2) by 20% compared with maximum volume at time of randomization or formation of additional brain abscesses after randomization. b. Clinical deterioration attributable to treatment failure of brain abscess. This does not include clinical deterioration caused by unrelated infections or other medical conditionsa. c. The local investigator assesses that clinical cure cannot be obtained without IV antibiotics after 4 weeks or longer of oral antibioticsb. This does not include remnant cystic formation at location of brain abscess or residual enhancement on cranial imaging of an otherwise asymptomatic patient [33–36].
Recurrence	New brain abscess formation after end of antibiotic treatment

^a Patients randomized to early switch to oral antibiotic therapy will, if needed, be allowed a short course (≤ 7 days) of IV antibiotic therapy for unrelated infections during oral treatment

^bPatients in both treatment arms may sometimes require prolonged treatment and complete radiological resolution of brain abscess often lags behind clinical cure. As long as the clinical condition improves, this is not considered treatment failure and patients can continue the assigned treatment strategy until clinical cure