Outcomes. In total, all 30 trials reported some information on the absence and/or occurrence of either total or serious adverse events

In total, all 30 trials reported some information on the absence and/or occurrence of either total or serious adverse events. In four trials, the incidence of any adverse events was not reported.^{121, 217, 221, 224, 232} Authors of 14 trials failed to report the absence or occurrence of serious adverse events.^{118, 121, 163, 216, 218, 219, 221, 225–227, 229, 230, 232, 237} The number of patients who withdrew as a result of adverse events was reported in all but two trials.^{221, 232}

Efficacy. The efficacy outcomes measured in the 30 included trials varied to some degree. The most commonly measured and reported outcomes across the trials were the mean or median endpoint score/mean score change on IIEF domains and/or individual questions, ^{103, 163, 214, 216, 217, 219, 220, 222–224, 226–229, 233, 234, 236–240, 242, 243} the mean endpoint/ mean change in the per-patient proportion of “yes” answers to the Sexual Encounter Profile questions 2 and 3 (i.e., SEP-2/3, “ Were you able to insert your penis into your partner's vagina? ” and “ Did your erection last long enough for you to have successful intercourse? ”), ^{103, 214–217, 219, 220, 222–230, 234, 236–240, 243} and the proportion of patients who answered “yes” to the Global Assessment Question 1 or 2 (i.e. GAQ-Q1: “ Has the treatment you have been taking improved your erections? ” and GAQ-Q2: “ If yes, has the treatment improved your ability to engage in sexual activity? ”).^{216, 217, 220, 222–224, 226–229, 234, 236–240} Of the 16 trials that reported outcomes based on GAQs, nine trials^{216, 217, 224, 226–229, 237 238} evaluated the proportion of patients who answered “yes” to GAQ-Q1 only, whereas the remaining seven trials^{220, 222, 223, 234, 236, 239, 240,} evaluated this parameter for both GAQ-Q1 and GAQ-Q2.

Eight trials additionally evaluated the efficacy (i.e., the mean per-patient percentage of successful intercourse attempts based on “yes” responses to SEP Q3) for different time-periods after dosing of tadalafil.^{216, 217, 219, 220, 224, 225, 227, 230}

The authors of one trial, ²²¹ derived logistic regression models based on the patient data obtained from a randomized placebo-controlled trial. These dose-response models assessed the relationship between the dose of tadalafil (2, 5, 10, or 25 mg) and the probability of getting an outcome (SEP questions 2 and 3 and/or on IIEF questions 3 and 4). The models included such covariates as baseline severity of ED and IIEF-EF domain score.

In five trials, ^{103, 163, 217, 222, 223, 241, 243} treatment satisfaction was measured using the Erectile Dysfunction Inventory of Treatment Satisfaction (EDITS) questionnaire.

The patients' treatment preference (e.g. percentage of patients who preferred the use of tadalafil to that of sildenafil) was evaluated and reported in six trials, all of which were of a crossover design in which the patients received both tadalafil and sildenafil^{103, 118, 121, 163, 241} or alternating dosing regimens of tadalafil (i.e., on demand versus fixed).^{214, 228, 244–247}

Endothelial function using percentage change (compared with baseline) in the mean flow mediated dilation (FMD) of brachial and cavernous arteries was measured in two trials.^{218, 232}