Skip to main content
. 2018 Jul 6;38(4):BSR20180214. doi: 10.1042/BSR20180214

Table 5. Assessment of potential sources of heterogeneity amongst studies by meta-regression.

Covariates Outcomes P-value Regression coefficient (95% CI)
Tumor type (UTUC compared with BC)
OS 0.82 0.92 (0.36–2.37)
CSS 0.15 0.52 (0.20–1.35)
RFS 0.02 0.27 (0.18–0.79)
30dC 0.03 0.48 (0.25–0.92)
90dM 0.02 0.36 (0.18–0.74)
Ethnicity (Asian compared with non-Asian)
OS 0.10 1.68 (0.90–3.14)
CSS 0.81 1.11 (0.39–3.16)
RFS 0.43 1.58 (0.37–6.80)
Analysis type (univariate compared with multivariate)
OS 0.74 0.88 (0.36–2.10)
CSS 0.54 0.77 (0.29–2.04)
RFS 0.05 0.39 (0.19–1.02)
30dC 0.16 0.62 (0.30–1.30)
90dM 0.66 0.66 (0.06–7.13)
Cut-off value (3.5 g/dl compared with others)
OS 0.08 2.50 (0.89–7.01)
CSS 0.72 0.86 (0.32–2.30)
RFS 0.78 1.18 (0.27–5.22)
30dC 0.32 0.64 (0.21–1.92)
90dM 0.42 0.52 (0.07–4.04)
Sample size (>600 compared with ≤600)
OS 0.46 1.38 (0.54–3.49)
RFS 0.40 1.63 (0.38–6.97)
30dC 0.35 1.64 (0.47–5.72)
90dM 0.82 1.20 (0.16–9.20)