Table 2.
Outcomes for each comparison.
| Outcome: Mortality (All-Cause) | |||||||
| Comparison | Included Studies | Number of Patients | OR, 95% CI | I² | Prediction Interval | E-Value | Comments |
| TMP/SMX versus FQs |
11 | 2407 | 1.46 (1.15–1.86) | 33% | 1.10–1.93 | For point estimate: 1.71; for CI: 1.35. | See forest plot (Figure 2) for subgroup analysis about different timing of mortality. All monotherapy studies. One pediatric study [27]. FQs: five studies about levofloxacin [24,26,31,35,44], one about ciprofloxacin [37], five mixed [25,28,32,33,39]. |
| TMP/SMX versus FQs-BSI |
4 | 234 | 2.61 0.75–9.02 |
67% | 0.01–503.12 | For point estimate: 2.61; for CI: 1. | Different timing of mortality: 30-day [24], in-hospital [28,33], 7-day [37]. One pediatric study [37]. FQs: one study about levofloxacin [24], one about ciprofloxacin [37], two mixed [28,33]. |
| TMP/SMX versus FQs not only monotherapy |
15 | 2806 | 1.58 (1.10–2.27) | 43% | 0.58–4.35 | For point estimate: 1.83; for CI: 1.28. | See forest plot for subgroup analysis about different timing of mortality (Figure 4). One pediatric study [37]. FQs: six studies about levofloxacin [24,26,31,35,36,44], two about ciprofloxacin [37,43], seven mixed [25,28,32,33,34,39,40]. |
| TMP/SMX versus FQs not only monotherapy-BSI |
7 | 469 | 2.45 (1.13–5.31) | 59% | 0.24–24.76 | For point estimate: 2.51; for CI: 1.32. | Different timing of mortality: 30-day [24,41], in-hospital [28,33], 60-day [34], 90-day [43], 7-day [37]. One pediatric study [37]. FQs: two studies about levofloxacin [24,41], two about ciprofloxacin [37,43], three mixed [28,33,34] |
| TMP/SMX versus TDs |
3 | 346 | 1.95 (0.79–4.82) | 0% | 0.01–685.99 | For point estimate: 2.14; for CI: 1. | All monotherapy studies. 30-day mortality. TDs: minocycline in two studies [27,39]. tigecycline in the other [23]. |
| TMP/SMX versus others |
5 | 791 | 1.33 (0.74–2.37) | 58% | 0.22–8.14 | For point estimate: 1.57; for CI: 1. | Different timing of mortality: 14-day [21], 30-day [32,39,41], in-hospital [40]. |
| FQs vs TDs | 3 | 174 | 0.80 (0.28–2.23) | 28% | 0.00–13,453.68 | For point estimate: 1.48; for CI: 1. | Different timing of mortality: 28-day [42], 30-day [39], 90-day [30]. TDs: minocycline as monotherapy in two studies [30,39], tigecycline in the other one mostly in combination for VAP [42]. |
| Monotherapy versus combination | 4 | 438 | 0.71 (0.41–1.22) | 0% | 0.16–3.08 | For point estimate: 1.66; for CI: 1. | See forest plot (Figure 5) for a subgroup analysis about different timing of mortality and population. |
| Outcome: Mortality—Adjusted Effect Size | |||||||
| Comparison | Included Studies | Number of Patients | OR, 95% CI | I² | Prediction Interval | E-Value | Comments |
| FQs versus TMP/SMX | 3 | 1912 | 0.73 (0.56–0.95) | 0% | 0.13–4.10 | For point estimate: 1.62; for CI: 1.19. | All monotherapy studies (Figure 3). |
| Outcome: Clinical Failure | |||||||
| Comparison | Included Studies | Number of Patients | OR, 95% CI | I² | Prediction Interval | E-Value | Comments |
| TMP/SMX versus FQs |
3 | 360 | 0.94 (0.53–1.67) | 0% | 0.02–39.64 | For point estimate: 1.21; for CI: 1. | All monotherapy studies. Different definitions of clinical failure. |
| TMP/SMX versus TDs |
3 | 346 | 0.78 (0.24–2.54) | 70% | 0.00–659,171.29 | For point estimate: 1.52; for CI: 1. | All monotherapy studies. Different definitions of clinical failure. TDs: minocycline in two studies [27,39], tigecycline in the other [23]. |
| TMP/SMX versus Others |
2 | 385 | 1.35 (0.77–2.35) | 0% | Incalculable | For point estimate: 1.6; for CI: 1. | TMP/SMX always in monotherapy, comparator group based prevalently (89%) on various combination regimens. Different definitions of clinical failure. |
| FQs vs TDs | 2 | 149 | 0.48 (0.15–1.54) | 64% | Incalculable | For point estimate: 2.24: for CI: 1. | TDs: minocycline as monotherapy in one study [39], tigecycline in the other one mostly in combination for VAP [42]. Different definitions of clinical failure. |
| Outcome: Safety-Adverse Events Onset | |||||||
| Comparison | Included Studies | Number of Patients | OR, 95% CI | I² | Prediction Interval | E-Value | Comments |
| TMP/SMX versus FQs |
4 | 461 | 1.89 (0.26–13.60) | 81% | 0.00–7492.40 | For point estimate: 2.09; for CI: 1. | All monotherapy studies. Definitions: “any adverse event” for 2 studies [24,35], drug discontinuation in another [33], acute kidney injury in the last one [39]. |
| Outcome: Length of Stay | |||||||
| Comparison | Included Studies | Number of Patients | MD, 95% CI | I² | Prediction Interval | E-Value | Comments |
| TMP/SMX versus FQs |
5 | 2064 | 2.90 (−4.19–9.99) | 84% | −14.25– 20.05 |
For point estimate: 1.56; for CI: 1. | All monotherapy studies except a minority of patients in Cho et al. [24] Infection-related LOS in Junco et al. [39] |
| TMP/SMX versus TDs (minocycline) |
2 | 301 | 16.33 (−252.49– 285.15) |
85% | Incalculable | For point estimate: 1.66; for CI: 1. | All monotherapy studies. Infection-related LOS in Junco et al. [39] |
Abbreviations: CI: confidence interval; FQs: fluoroquinolones; LOS: length of stay; MD: mean difference; OR: odds ratio; TDs: tetracycline derivatives; TMP/SMX: trimethoprim/sulfamethoxazole; VAP: ventilator-associated pneumonia.