What the as-needed program asked
Section 14.1 of the DailyMed Cialis label (setid bcd8f8ab-81a2-4891-83db-24a0b0e25895) is the source for the numbers below. In the seven primary studies, men took tadalafil as needed at doses from 2.5 mg to 20 mg, no more than once a day. They chose the gap between the tablet and the attempt. Food and alcohol were not restricted.
The erectile-function domain is a 4-week recall score. Higher is better, and the ceiling is 30. SEP question 2 asks whether insertion was possible. SEP question 3 asks whether the erection lasted long enough for successful intercourse. The label reports the percent of successful attempts for each man, then the group.
Two US trials enrolled 402 men, mean age 59 (range 27 to 87). More than 90% had had erectile dysfunction for at least a year. Study A was mostly academic centers. Study B was mostly community urology practices. Diabetes, hypertension, and other cardiovascular disease were in the mix. The treatment effect of 20 mg did not fade across the study.
Outside the United States, five primary studies enrolled 1,112 patients, mean age 59. There, 5 mg, 10 mg, and 20 mg each showed improvement versus placebo on the three primary measures in the tables where those doses were tested. One of those studies ran 6 months. This page does not paste every foreign cell into a new grid. The pattern is dose-ordered improvement, with placebo already moving a little.
The US 20 mg numbers people misquote
| US as-needed trial | Placebo, EF change | Cialis 20 mg, EF change | SEP3 at the end |
|---|---|---|---|
| Study A | -0.2 (n=49) | +6.9 (n=146) | 25% placebo, 50% on 20 mg |
| Study B | +0.3 (n=48) | +9.3 (n=159) | 23% placebo, 64% on 20 mg |
Study A, 20 mg (146 men) versus placebo (49 men): erectile-function endpoint 19.5 versus 13.5. Change from baseline 6.9 versus -0.2. Successful intercourse (SEP3) ended at 50% versus 25%, a change of 34 points versus 5. Insertion (SEP2) ended at 62% versus 39%.
Study B, 20 mg (159) versus placebo (48): erectile-function endpoint 22.5 versus 13.6. Change 9.3 versus 0.3. SEP3 ended at 64% versus 23% (change 44 points versus 4). SEP2 ended at 77% versus 43%.
Those are study-population means under a protocol. They are not the chance that a particular Friday works. Quoting 64% without the 23% on placebo turns a comparison into a promise. Both bars belong in the sentence.
A separate diabetes trial of as-needed 10 mg and 20 mg (216 men in that study, on top of 235 men with diabetes inside the general trials) still beat placebo. The erectile-function score rose 6.4 on 10 mg and 7.3 on 20 mg, versus 0.1 on placebo. After bilateral nerve-sparing prostatectomy (303 men), 20 mg lifted the score 5.3 points versus 1.1 on placebo, and SEP3 ended at 41% versus 19%. The drug still separated from placebo. The gap was smaller than in the broad US studies. Men who arrive with a cancer surgery or longstanding diabetes and expect the Study B percentage are reading the wrong row.
The 24-hour and 36-hour studies used 10 mg and 20 mg
Two studies timed intercourse to the clock. They are why the label says as-needed Cialis improved erectile function compared with placebo for up to 36 hours. They are not a duration claim for a 2.5 mg daily tablet, and they are not a coupon for a second dose inside that window.
First study: 348 men, placebo or 20 mg. At 24 hours (more precisely 22 to 26 hours), at least one successful intercourse was reported by 61% on 20 mg (84 of 138) and 37% on placebo (53 of 144). At 36 hours (33 to 39 hours), 64% on 20 mg (88 of 137) and 37% on placebo (49 of 133).
Second study: 483 men spread across placebo, 10 mg, and 20 mg, assigned to attempt intercourse at either 24 or 36 hours. At 36 hours the mean per-patient success was 33% on placebo, 56% on 10 mg, and 62% on 20 mg. At 24 hours those figures were 42%, 56%, and 67%. So 10 mg did clear placebo at those late windows. It did not match the early stopwatch story.
That early stopwatch enrolled 223 men. By 30 minutes, using a definition of at least one successful erection in four attempts, the rates were 35% placebo, 38% on 10 mg, and 52% on 20 mg. If a review says Cialis 'kicks in at half an hour,' it is leaning on the 20 mg arm and ignoring how close 10 mg sat to placebo.
Once-daily trials and the prostate score
Section 14.2 is a different program. Once-daily Cialis for erectile dysfunction was evaluated in two trials of 12 weeks and one of 24 weeks, 853 patients in all. Doses in the program ranged from 2.5 mg to 10 mg. Sex was not timed to the tablet. Food and alcohol were unrestricted.
In the 24-week US study, the erectile-function score rose 1.2 on placebo (94 men), 6.1 on 2.5 mg (96 men), and 7.0 on 5 mg (97 men). SEP3 change was 10, 31, and 35 points. End-of-study SEP3 was 31%, 50%, and 57%. In the 12-week study outside the United States, 5 mg (109 men) lifted the erectile-function score 9.7 points versus 0.9 on placebo (54 men).
The diabetes once-daily trial is the humbling row. Scores rose 4.8 on 2.5 mg and 4.5 on 5 mg, versus 1.3 on placebo, among about 100 men per arm. Still separated from placebo. Not the same lift as the general 24-week study. A review that quotes only the largest gap is choosing a friendlier room.
Prostate-symptom trials used the International Prostate Symptom Score, 0 to 35, higher meaning worse. In two 12-week studies of 5 mg once daily, the score fell 4.8 versus 2.2 on placebo in Study J (205 men each) and 5.6 versus 3.6 in Study K. Placebo improved urinary scores too. Maximum flow rate rose in both groups and the difference was not significant (about 1.6 mL/sec on 5 mg versus 1.2 and 1.1 on placebo). Symptom relief on the questionnaire is what the label claims. A stronger stream on the machine is not.
What a star rating skips
In eight primary as-needed studies, 3.1% of men on 10 mg or 20 mg stopped for adverse events, versus 1.4% on placebo. Headache in that as-needed safety table was 5% on placebo, 11% on 5 mg, 11% on 10 mg, and 15% on 20 mg. Dyspepsia was 1%, 4%, 8%, and 10%. Back pain was 3%, 3%, 5%, and 6%. The 5 mg column in that table is 151 men. It is an as-needed safety cell, not proof about a breakfast habit.
In three once-daily placebo-controlled trials of 12 or 24 weeks, 4.1% of men on tadalafil stopped for adverse events, versus 2.8% on placebo. In the 12-week safety table, headache on 5 mg was 6%, versus 5% on placebo and 3% on 2.5 mg. The bothersome effects are real and usually the familiar ones: headache, indigestion, back pain, muscle ache, congestion, flushing. Back pain and muscle ache in pharmacology trials generally arrived 12 to 24 hours after a dose and typically settled within 48 hours.
Across the development program tadalafil was given to over 9,000 men. Once-daily exposure included 1,434 men for at least 6 months, 905 for at least a year, and 115 for at least 2 years. As-needed exposure included more than 1,300 men for at least 6 months and 1,000 for at least a year. Those are duration counts, not a lifetime warranty.
Color-vision changes were rare, under 0.1% of patients across studies. The trials did not enroll men with a heart attack in the prior 90 days, unstable angina, NYHA Class 2 or greater heart failure in the last 6 months, uncontrolled arrhythmias, blood pressure under 90/50 mm Hg, uncontrolled hypertension, or a stroke in the last 6 months. A high intercourse percentage from Study B does not answer those excluded rooms. Nitrates were a contraindication, not a subgroup.
Consumer ratings mix expectation, counterfeit risk, and the wrong clock. They are not 22 randomized trials. For how the two schedules should be chosen before anyone orders, use the daily versus as-needed card. For how the product got from a 2003 approval to a prostate use, use the history card.
Questions the tables can answer
How many tadalafil trials are in the as-needed section? The label says efficacy and safety were evaluated in 22 trials of up to 24 weeks, with over 4,000 patients. Seven primary 12-week studies anchor the general-population story.
What did effectiveness mean? Higher erectile-function scores, more successful attempts at penetration, and more attempts where the erection lasted long enough for intercourse, each versus placebo.
Is the 36-hour figure a 5 mg result? No. The timed studies used 10 mg and 20 mg. At 36 hours in the first study, 20 mg beat placebo (64% versus 37% with at least one success). The once-daily program is a separate section.
Did placebo do nothing? No. Placebo arms moved, sometimes by several points on the diary. The drug effect is the gap that remains.
Does the prostate indication use the same score? No. Those trials used the International Prostate Symptom Score. Flow rate was not significantly better than placebo.
Where are the dosing rules? On the compound page and the schedule card. This URL stops at what was measured.