Table 4.
Delayed bleeding risk according to the two different sensitivity analyses performed
| Study | MTBI | Sensitivity |
|---|---|---|
| ANTONI_2019 | NA | 1.85% (0.23–6.53%) |
| BATTLE_2017 | NA | 2.27% (0.06–12.02%) |
| CHENOWETH_2018 | NA | 7.11% (3.94–11.64%) |
| ERNSTBRUNNER_2016 | 1.05% (0.29–2.66%) | 1.05% (0.29–2.66%) |
| GALLIAZZO_2019 | 1.53% (0.19–5.41%) | 1.53% (0.19–5.41%) |
| GANETSKY_2017 | NA | 3.61% (2.30–5.37%) |
| HILL_2018 | NA | 3.29% (1.33–6.65%) |
| HUANG_2019 | NA | 2.52% (0.52–7.19%) |
| MANN_2018 | NA | 0.00% (0.00–3.18%) |
| NISHIJIMA_2012 | 0.00% (0.00–1.53%) | 1.23% (0.26–3.57%) |
| PECK_2011 | NA | 0.00% (0.00–3.52%) |
| SCANTLING_2017 | 1.21% (0.15–4.31%) | 1.2% (0.15–4.28%) |
| STANITSAS_2016 | NA | 0.00% (0.00–8.81%) |
| SWAP_2016 | NA | 2.31% (0.85–4.95%) |
| TAUBER_2009 | 4.00% (1.10–9.93%) | 4.00% (1.10–9.93%) |
| TAYLOR_2012 | NA | 0.00% (0.00–4.25%) |
| No. studies | 5 | 16 |
| DB mean estimated risk | 1.04% (0.15–2.49%) | 1.70% (0.93–2.67%) |
| p value for heterogeneity | p = 0.0334 | p = 0.0012 |
| I2 statistic | 62% | 60% |
The results are presented as the mean estimates of delayed bleeding risk and 95% confidence intervals (in parentheses). GCS = Glasgow Coma Scale; MTBI = mild traumatic brain injury: includes only studies enrolling patients with TBI and a GCS ≥ 13; Sensitivity = includes all the studies considering unexplained deaths and patients lost at follow-up as events (i.e. delayed bleeding); No. studies = number of the studies included in the analysis; NA = not assessed in the corresponding original study