Symptom scores and clinical measures in benign prostatic hyperplasia

Figure 1  PRISMA flow diagram for study identification and selection.

Figure 2 Forest plots. A: Forest plot for the International Prostate Symptom Score (IPSS)-quality of life meta-analysis (k = 5, n = 705). Pooled random-effects r = 0.713 [95% confidence interval (CI): 0.557-0.820; I² = 3.26%]. Studies ordered by publication year; B: Forest plot for the prostate volume-IPSS meta-analysis (k = 12, n = 1446). Pooled random-effects r = 0.457 (95%CI: 0.257-0.619; I² = 38.37%). Studies ordered by publication year; C: Forest plot for the IPSS-post-void residual meta-analysis (k = 11, n = 2663). Pooled random-effects r = 0.363 (95%CI: 0.150-0.543; I² = 51.18%). Studies ordered by publication year.

Figure 3 Funnel plots. A: Funnel plot for the International Prostate Symptom Score (IPSS)-quality of life meta-analysis. No significant asymmetry detected (Egger’s P = 0.076; Kendall’s P = 0.233); B: Funnel plot for the prostate volume-IPSS meta-analysis. No evidence of publication bias (Egger’s P = 0.510; Kendall’s P = 0.630); C: Funnel plot for the IPSS-post-void residual meta-analysis. Borderline asymmetry noted (Egger’s P = 0.087; Kendall’s P = 0.218). Fazal et al[12] is a high-Fisher’s-Z outlier consistent with a small-study effect.