Table 3. Cox regression analysis for prognostic factors according to 90-day mortality in 357 patients with methicillin-sensitive S. aureus bacteraemia (SAB) with at least one deep infection focus.
Univariate analysis | Cox regression analysis 1 | |||||
---|---|---|---|---|---|---|
Survived N = 291 (82) | Died N = 66 (18) | OR (95% CI) | p- value | OR (95% CI) | p- value | |
Male sex | 181 (62) | 37 (56) | 0.78(0.45–1.33) | NS | — | — |
Age >60 years | 142 (49) | 49 (74) | 3.02(1.66–5.49) | <0.001 | 2.00(1.13–3.54) | 0.017 |
Healthy-nonfatal disease A | 233 (80) | 28 (42) | 0.18(0.10–0.32) | <0.001 | 0.23(0.14–0.39) | <0.01 |
Ultimately-rapidly fatal disease A | 58 (20) | 38 (58) | 5.45(3.09–9.61) | <0.001 | — | — |
Healthcare associated SAB | 153 (53) | 47 (71) | 2.23(1.25–3.99) | 0.006 | — | — |
Intensive care unit B | 42 (14) | 18 (27) | 2.22(1.18–4.19) | 0.013 | — | — |
Severe sepsis B | 14 (5) | 9 (14) | 3.12(1.29–7.57) | 0.008 | 3.64(1.76–7.53) | 0.01 |
Endocarditis C | 44 (15) | 18 (27) | 2.11(1.12–3.95) | 0.019 | 2.68(1.53–4.71) | 0.01 |
Fluoroquinolone therapy D | 156 (54) | 30 (45) | 0.72(0.42–1.23) | NS | — | — |
Aminoglucoside therapy | 45 (15) | 16 (24) | 1.75(0.92–3.34) | NS | — | — |
Lack of rifampicin therapy | 71 (24) | 25 (38) | 1.89(1.07–3.32) | 0.026 | — | — |
Rifampicin therapy of any duration E | 220 (76) | 41 (62) | 0.53(0.30–0.93) | 0.026 | — | — |
Rifampicin therapy ≥14 days, early onset F | 167 (57) | 18 (27) | 0.28(0.15–0.50) | <0.001 | 0.33(0.19–0.57) | <0.01 |
Rifampicin therapy ≥14 days, late onset G | 22 (8) | 4 (6) | 0.79(0.26–2.37) | NS | — | — |
Bedside IDS consultation H | 271 (93) | 56 (84) | 0.41(0.18–0.93) | 0.029 | — | — |
Telephone IDS consultation H | 14 (5) | 9 (14) | 3.12(1.29–7.57) | 0.008 | 2.11(1.01–4.42) | 0.04 |
Patients with alcoholism, acute or chronic liver diseases, lack of deep infection foci, MRSA bacteraemia (n = 6) or a fatal outcome within 3 days have been excluded. Values are expressed as N (%) and odds ratio (OR) for fatal outcome within 90 days. NS = non-significant.
1 Cox regression analysis results are shown only for parameters with significant prognostic impact (p<0.05).
A Classification according to McCabe and Jackson [32].
B Severity of illness at blood culture collection time point
C Endocarditis diagnosed within 90 days follow-up.
D Fluoroquinolone: levo-, moxi- or ciprofloxacin.
E Including all patients with rifampicin therapy of any length.
F Early onset i.e. onset within 7 days of positive blood culture.
G Late onset i.e. onset 7 days after positive blood culture.
H Infectious diseases specialist (IDS) consultation