Table 1. Antibiotic prophylaxis for early-VAP prevention in comatose patients: patients, interventions, and outcomes.

			Desirable effect		Undesirable effect		Desirable effect
Sample	Treatment	Control	Outcome	Outcome quality	Outcome	Outcome quality	Treatment arm % (95%-CI)	Controls % (95%-CI)	Difference % (95%-CI)	RR (95%-CI)
Ventilated patients, cerebral haemorrhage, GCS ≤8	Ampicillin-sulbactam 3-day course - 19 patients	Noplacebo -19 patients	Early VAP	Weak, surrogate of death/severe disability	Antibiotic- resistantbacteria selection not Adequately investigated	Robust, Clinicallyrelevant for thepatient/crucial forthe health-careorganization	21.1 (8.5 to 43.3)	57.9 (36.3 to 76.9)	–36.8	0.36 (0.14 to 0.94)
Ventilated patients, cerebral haemorrhage, GCS ≤12	2-dosecefuroxime - 50patients	Noplacebo - 50 patients	Early VAP	Weak, surrogate of death/severe disability	Antibiotic- resistantbacteria selection not investigated	Robust, Clinically relevant for the patient/crucial for the health-care organization	16 (8.3 to 28.5)	36 (24.1 to 49.9)	–20 (–35.8 to −2.8)	0.44 (0.21 to 0.93)