Skip to main content
. 2021 Apr 30;71(706):e372–e380. doi: 10.3399/bjgp20X714197

Table 2.

Study counts of the 10 possible predictors for ED prevalence and P-value for the uni-predictor (or smallest P-value for the uni-predictor dummy variables) in the univariate MA regression analysis

Predictor Study count, n Coefficient estimate 95% CI P-value
Stage 40 0.6210 −0.5253 to 1.7673 0.2883
Continent 40 1.2655 −0.3051 to 2.8362 0.1143
Mean age 40 0.0503 0.0243 to 0.0762 0.0002a
Age range 40 0.0057 −0.0210 to 0.0325 0.6739
ED assessment method 40 1.5236 −0.1845 to 3.2316 0.0804
Cancer site 40 −1.8135 −2.5841 to −1.0429 <0.0001b
Age standard deviation 17 −0.0078 −0.0733 to 0.0576 0.8144
Proportion of patients who underwent radiotherapy 15 1.0584 −1.0225 to 3.1393 0.3188
Proportion of patients with diabetes 12 −3.6361 −19.1360 to 11.8639 0.6457
Proportion of patients who underwent chemotherapy 12 0.1348 −1.9925 to 2.2621 0.9012
a

Significance code controlling for type I error rate <0.05 and >0.01.

b

Significance code controlling for type I error rate <0.0001. CI = confidence interval. ED = erectile dysfunction. MA = meta-analysis.