Impact of viral multiplex real-time PCR on management of respiratory tract infection: a retrospective cohort study

Table 10 Change of antibiotic therapy after rtPCR, all patients with another indication for antibiotics were excluded

	Viral	Bacterial	Mixed	No pathogen	Clinically	Clinically	p-value ^c
	(n = 62)	(n = 37)	(n = 9)	(n = 102)	bacterial ^a (n = 106)	viral ^b (n = 28)	Viral vs. bacterial	Viral vs. mixed	Viral vs. no pathogen
No antibiotic treatment before and after rtPCR, n (%)	22 (35.5)	8 (21.6)	2 (22.2)	38 (37.3)	24 (22.6)	19 (67.9)	0.15	0.71	0.82
Antibiotic treatment stopped after rtPCR, n (%)	8 (12.9)	2 (5.4)	0 (0.0)	3 (2.9)	6 (5.7)	4 (14.3)	0.40	0.64	0.03
Antibiotic treatment started after rtPCR, n (%)	6 (9.7)	4 (10.8)	3 (33.3)	11 (10.8)	10 (9.4)	3 (10.7)	1.00	0.16	0.82
Antibiotic treatment before and after rtPCR, n (%)	26 (41.9)	23 (62.2)	4 (44.4)	50 (49.0)	66 (62.3)	2 (7.1)	0.05	1.00	0.38
Correct management, n (%)	30 (48.4)	27 (73.0)	7 (77.7)		76 (71.7)	23 (82.1)	0.02	0.19	n/a

n number; rtPCR, real-time polymerase chain reaction
^aCases without detection of bacteria were evaluated as clinically bacterial if they fulfilled one of the following criteria: unilobular or multilobular pulmonary infiltrate or CRP >100 mg/l or PCT >0.25 μg/l or antibiotic therapy before rtPCR and no bacterium detection and biomarkers indeterminate (CRP >100 mg/l and PCT ≤0.25 μg/l or CRP between 51 and 100 mg/l and PCT not available)
^bCases with a detected viral pathogen excluding those patients with a clinically bacterial co-infection (as described above)
^cMantel-Haenszel chi square test or Fisher exact test were used