This meta-analytic study aims to estimate the likely improvements of erectile dysfunction (ED) measured by the International Index of Erectile Function (IIEF) at the highest fixed dosages of the three available PDE-5-inhibitors: sildenafil, tadalafil, and vardenafil. MEDLINE and the Cochrane Library were searched electronically for efficacy trials of PDE-5-inhibitors for treating ED. In addition drug manufacturers were contacted to provide unpublished or unrecorded congress proceedings. Randomized, double-blind, placebo-controlled, parallel-group, maximum fixed-dose, broad-spectrum efficacy trials using IIEF were included in the analysis. Data were independently extracted by two reviewers. The results were pooled using weighted mean differences. A formal indirect comparison (including Bonferroni-correction) was conducted to estimate the differences between agents. A total of 14 trials were included in the meta-analysis (three with 100 mg sildenafil, eight with 20–25 mg tadalafil, and three with 20 mg vardenafil). All trials were of good methodological quality. Overall heterogeneity was moderate: I2=33.2%, χ2=19.47, P=0.11. The funnel plot suggested moderate likelihood of publication bias. Pooled results of IIEF-improvement were for sildenafil 9.65 (95% CI: 8.50, 10.79) points, tadalafil 8.52 (7.61, 9.42) points, and vardenafil 7.50 (6.50, 8.50) points, respectively. Sildenafil proved to be significantly more effective than vardenafil (d=2.15, P=0.006), other pairwise comparisons showed no difference in efficacy. All PDE-5-inhibitors are highly effective in the treatment of ED. At maximum dosage they improve erectile function 7–10 points on the IIEF compared to placebo-treatment. There is evidence that sildenafil might be more efficacious than vardenafil, although this is to be interpreted with caution. To prove higher efficacy truly independent comparative trials are needed.
According to high-quality international community-based epidemiological studies, erectile dysfunction (ED), the difficulty in developing or maintaining an erection suitable for satisfactory intercourse, affects around 16% of the men all over the world.1, 2 However, some estimates range up to 27%.3, 4 When using ad hoc definitions of ED reported prevalence rates exceed 40% (for summary see Kubin et al.5). The prevalence of ED increases with age as well as with comorbid conditions including cardiovascular diseases, diabetes, neurological diseases, and depression.1, 2, 5
There are several treatments for ED, such as hormone compounds, ointments, vacuum-constriction devices, intracavernosal injections of vasoactive agents, transurethral delivery of alprostadil, implantation of penile prostheses, venous or arterial surgery, and psychotherapeutic methods. Nevertheless, with the approval of sildenafil, a phosphodiesterase-5 (PDE-5) inhibitor, in 1998 an oral therapy for ED became available in the US that soon found worldwide acceptance. With the appearance of two other PDE-5 inhibitors, tadalafil and vardenafil in 2003, a competitive situation emerged, which was described by Wyllie as ‘the phosphodiesterase inhibitor war’.6 In his opinion, in 2003 it was high time to see, what the new agents (tadalafil and vardenafil) could and would add to treatment of ED.6 One aspect of this question, the topic of efficacy is the focus of the present meta-analysis.
Using evidence-based medicine (EBM) terms to date efficacy of sildenafil is empirically supported on the level I.A., that is, based on systematic review(s).7, 8, 9 The efficacy of tadalafil10 and vardenafil11, 12, 13, 14, 15 is supported by evidence level I.B., that is, based on high-quality randomized controlled trials (RCT). In these studies all three agents display both statistically as well as clinically significant improvement in erectile functioning.
In efficacy trials of treatments for ED numerous outcome measures exist. Besides physiological assessments, such as penile rigidity (RigiScan16), physician- (e.g. clinician's global impression (CGI)) and patient-rated instruments are available. General efficacy assessment questions (often abbreviated as GAQ, GEQ or GEAQ), for example, ‘Has the treatment improved your erections?’, are often preferred and presented as primary outcome to researchers, physicians and patients. Patient diaries (event logs), such as the sexual encounter profile (SEP), are also frequently used to assess ‘proportion of successful intercourse attempts’ or ‘proportion of patients with at least one successful intercourse attempt’ as outcomes. Patient questionnaires (e.g. International Index of Erectile Function (IIEF)17) are available and can be analyzed either as items or as scales. The latter is used in almost all newer studies of PDE-5-inhibitors. Most studies use more than one outcome for efficacy assessment.
The objective of the present meta-analysis is to evaluate the efficacy of the available PDE-5 inhibitors for ED and to assess possible difference between these agents using the Erectile Function Domain Score of the IIEF.17
Materials and methods
In July 2004, the MEDLINE and the Cochrane Central Register of Controlled Trials (CENTRAL) databases were searched for efficacy trials of sildenafil, tadalafil, and vardenafil. In the electronic search the substance and trade names were combined with the truncated terms ‘effic*’ and ‘effec*’. This highly sensitive strategy should enable a more extensive retrieval of relevant studies, than more restricted searches (e.g. by spectrum of disease or outcomes). Additionally, the website of the US Food and Drug Administration/Center for Drug Evaluation and Research was searched for approval studies since in many cases some approval studies are not available in a published format.
The national branches of drug manufacturers of the three examined agents were contacted to provide additional information, such as unpublished congress proceedings. Reference lists of all relevant publications were searched for possible additional papers (ancestry approach).
Inclusion of trials
Titles and abstracts of all identified publications were screened independently by two reviewers (LK and AH). Trials were included, when they fulfilled all of the following criteria: (1) randomized and controlled by placebo (RCT); (2) parallel-group design; (3) double-blinding; (4) administering the maximum fixed-dose regimen of the agent (sildenafil 100 mg, tadalafil 20–25 mg, vardenafil 20 mg); (5) conduction in broad-spectrum sample with ED; (6) administering the IIEF;17 (7) reporting relevant outcome, at least mean and sample size. When decision about inclusion was impossible on the basis of the abstract, the full text was acquired. Any disagreement between reviewers was resolved by involving a third reviewer (MB) to decide.
Data extraction and assessment of methodological quality
Basic data were extracted by LK using a structured form that included the following information: authors; year of publication; language; country; agent; daily dose; duration of treatment; sample size; age; ethnicity; duration of ED; cause of ED; and baseline severity of ED in patient population.
Methodological quality was assessed by LK by means of the Jadad Score.18 Due to the inclusion criteria (randomization and double-blinding) all studies scored at least two points on Jadad's scale. Further three points could be obtained by describing the (appropriate) randomization, (appropriate) double-blinding method and the study withdrawals. If study report was brief (e.g. only abstract available), methodological quality was not assessed.
Outcome reports were extracted independently by two reviewers (LK and AH). Any disagreement was resolved by discussion.
In case of insufficiently reported data, standard deviations were substituted. The estimation was based either on figures presenting confidence intervals, standard errors or standard deviations or on other studies with the same agent, which reported s.d. for treatment and placebo group. Means and sample sizes were not substituted, insufficient reporting led to study exclusion.
The outcome of the present meta-analysis was the Erectile Function (EF) domain score of IIEF.17 We chose this parameter because of several reasons. First, it is collected from the patient, providing direct data from the affected person. Second, it is a psychometrically tested instrument that validly and reliably measures several aspects of erectile functioning. Third, its continuous character allows a differentiated assessment of the ED (in contrast to categorizing patients in ‘responder’ and ‘nonresponder’), as well as the application of more sophisticated statistical methods. And fourth, it is a widespread assessment of disease severity in efficacy trials concerning ED. The EF domain score can range from 1 to 30.
For meta-analysis, weighted mean differences (WMD) between agent and placebo group in the EF domain score after treatment were calculated. As the stringent inclusion criteria probably led to reduced clinical heterogeneity, we also assumed a low statistical heterogeneity in primary study findings. We, therefore, used a fixed-effect model for pooling the data. Statistical heterogeneity was assessed by χ2- and I2-statistics. I2 indicates the percentage of the heterogeneity in effect estimates that is due to clinical or methodological diversity rather than chance.19
At the time of performing the present meta-analysis there was no information available from independent high-quality direct-comparative trials. Efficacy of sildenafil, tadalafil, and vardenafil was therefore indirectly compared using formal (statistical) methods suggested by several experts.20, 21, 22 The chosen procedure bases on the assumption, that agents are comparable through their relative effect vs a common comparator. In the present case it is placebo. In the analysis Bonferroni-correction of level of significance for multiple testing was carried out.
To test for possible publication bias, visual examination of funnel plot (scatter plot of effect size and the reciprocal standard error) was performed.
During the review process it was noted, that the highest fixed-dose regimen in case of tadalafil was not sufficiently defined. In some trials 25 mg tadalafil were administered, although the highest approved regimen in the US is 20 mg. Therefore, a post hoc decision was made to perform a sensitivity analysis with and without these trials.
Data analysis was performed with Review Manager 4.2, the Statistical Package for Social Sciences (SPSS) 11.5 and Microsoft Excel 8.0.
Identification of trials
Literature search retrieved 75 double-blind parallel group RCTs that evaluated efficacy of sildenafil, tadalafil, or vardenafil (see Figure 1). For sildenafil, five relevant studies were identified. Three of them were included in the meta-analysis23, 24, 25 and two were excluded due to insufficiently reported data.26, 27 A significant number of sildenafil studies were conducted before the development of the IIEF and were consequently excluded. Several studies could not be included due to the usage of flexible-dose regimen.
For tadalafil, eight studies were included in the analysis. Three of them were published in medical journals28, 29, 30 and five were abstracted from the documentation of the US National Food and Drug Administration.31 Three of these approval studies were already directly pooled in an analysis and published also elsewhere.10
For vardenafil, three trials were included in the meta-analysis. Two of them were published in medical journals.11, 12 One of them is an approval study that was also published a second time elsewhere.11, 32, 33 The third study was extracted from the documentation of the US National Food and Drug Administration.33
Altogether 14 studies with a total of 4836 patients were included in the meta-analysis (Figure 1).
Characteristics and methodological quality of the included trials is presented in Table 1.
The main results of the meta-analysis are shown in Figure 2. Pooled findings for sildenafil resulted in an effect of 9.65 (8.50, 10.79) points improvement. Heterogeneity between trial results was minimal (χ2=0.03, P=0.99, I2=0%).
Findings of eight tadalafil studies could be pooled into an effect of 8.52 (7.61, 9.42) points improvement. Minor but not substantial heterogeneity was present in the data (χ2=10.63, P=0.16, I2=34.1%).
Vardenafil showed an effect of 7.50 (6.50, 8.50) points improvement in a homogenous study sample (χ2=1.17, P=0.56, I2=0%).
Combining all studies lead to an estimated effect size of 8.46 (7.88, 9.04) points improvement. The whole data set was slightly heterogeneous (χ2=19.47, P=0.11, I2=33.2%).
In pairwise comparisons no significant difference was found between sildenafil and tadalafil (difference(d)=1.13 points, P=0.132) as well as between tadalafil and vardenafil (d=1.02 points, P=0.140). Sildenafil proved to be significantly more efficacious than vardenafil (d=2.15 points, P=0.006).
Visual examination of the funnel plot (Figure 3) suggested moderate likelihood of publication bias both for individual agents and for all studies, although results rest on a limited number of studies.
Sensitivity analysis with the exclusion of studies, in which a tadalafil dose of 25 mg was administered lead to a minor reduction of heterogeneity in the tadalafil study sample (I2=29.2%). Although there was some tendency favoring sildenafil over tadalafil it did not prove a significant effect on pairwise comparisons (sildenafil vs tadalafil: d=1.50, P=0.054; tadalafil vs vardenafil: d=0.65, P=0.367).
The present meta-analysis assesses the efficacy of the three available PDE-5 inhibitors (sildenafil, tadalafil, and vardenafil) for the treatment of ED. Its main limitation results from the fact, that it is not an overall systematic review, but merely a comparative meta-analysis of the best available evidence using the highest possible meta-analytic methodological standards. Thus it cannot provide an overall evaluation of all therapeutic properties of the three oral agents for ED. We aimed merely at one single efficacy outcome, the erectile function domain score of the IIEF.17 Although it is valid and reliable, in terms of ‘objective’ clinical epidemiology the best available outcome measure. But it cannot be excluded that other – especially generically created – outcome measures might have led to different results. In addition, we chose not to evaluate adverse effects and discontinuation rates, although these undesirable outcomes are important for the individual patient decision to favor one treatment over another. Other relevant aspects, such as onset latency, half-life or patient preference, were also not considered.
Nevertheless, some essential statements can be made based on this meta-analysis. First, there is no doubt, that all PDE-5 inhibitors are highly efficacious. An 8.5 point improvement in comparison to the control group is large enough, to successfully treat severe ED. Second, there seem to be possible efficacy differences between agents, mainly the statistically significant superiority of sildenafil over vardenafil. Although it might be too speculative to consider them as clinically relevant for an individual patient choice they might become more important in larger perspectives, for example, in cost-benefit analyses.
Since the comparison was carried out indirectly an important question with respect to the internal validity of the present meta-analysis is the comparability of the studies that evaluate the different agents. The differences that were found might be caused by other factors than the agent itself, such as baseline ED severity, etiology, age, ethnicity, or treatment history.34 To finally evaluate the likely differences between agents truly independent comparative trials are necessary.
This possible shortcoming with respect to internal validity, however, has some advantage considering external validity, that is, generalizibility of the results. The use of unselected, community-based patient samples with various etiologies, age, co-morbid conditions, ethnicity etc. allows to establish an efficacy prove for a broad spectrum of patients suffering from ED. This finding is also supported by a very low heterogeneity in the primary study results. With this respect, the present findings might support evidence not only for efficacy, but to a certain degree also for effectiveness in everyday practice, although highest dosage of these drugs may not be always be used. Effectiveness of the agents is also confirmed in open-label naturalistic studies.35, 36, 37
After establishing evidence for efficacy and pointing out towards possible efficacy differences, further information is needed in questions that mainly affect patient's compliance and preferences. This might include the effects on quality of life (e.g. effects on relationship issues), likely predictors of (non)compliance (e.g. myths about dangerous side effects) and true patient preferences (e.g. realistic desired frequency of intercourse). Indeed, the most essential step to overcome ‘the phosphodiesterase inhibitor war’6 would be to establish a shared-decision health care model that is oriented towards the patient's needs rather than marketing interests of third parties and a medicalization of the couple's sexuality.
Rosen RC, Fisher WA, Eardley I, Niederberger C, Nadel A, Sand M . The multinational Men's Attitudes to Life Events and Sexuality (MALES) study: I. Prevalence of erectile dysfunction and related health concerns in the general population. Curr Med Res Opin 2004; 20: 607–617.
Nicolosi A, Glasser DB, Moreira ED, Villa M, Erectile Dysfunction Epidemiology Cross National Study Group. Prevalence of erectile dysfunction and associated factors among men without concomitant diseases: a population study. Int J Impot Res 2003; 15: 253–257.
Heruti R, Shochat T, Tekes-Manova D, Ashkenazi I, Justo D . Prevalence of erectile dysfunction among young adults: results of a large-scale survey. J Sex Med 2004; 1: 284–291.
Auld RB, Brock G . Sexuality and erectile dysfunction: results of a national survey. J Sex Reprod Med 2002; 2: 50–54.
Kubin M, Wagner G, Fugl-Meyer AR . Epidemiology of erectile dysfunction. Int J Impot Res 2003; 15: 63–71.
Wyllie MG . The phosphodiesterase inhibitor ‘war’. Brit J Urol Int 2003; 91: 573.
Burls A, Gold L, Clark W . Systematic review of randomised controlled trials of sildenafil (Viagra) in the treatment of male erectile dysfunction. Br J Gen Pract 2001; 51: 1004–1012.
Fink HA, Mac Donald R, Rutks IR, Nelson DB, Wilt TJ . Sildenafil for male erectile dysfunction: a systematic review and meta-analysis. Arch Intern Med 2002; 162: 1349–1360.
Moore RA, Edwards JE, McQuay HJ . Sildenafil (Viagra) for male erectile dysfunction: a meta-analysis of clinical trial reports. BMC Urol 2002; 2: 6.
Brock GB, McMahon CG, Chen KK, Costigan T, Shen W, Watkins V et al. Efficacy and safety of tadalafil for the treatment of erectile dysfunction: results of integrated analyses. J Urol 2002; 168: 1332–1336.
Hellstrom WJG, Gittelman M, Karlin G, Segerson T, Thibonnier M, Taylor T et al. Vardenafil for treatment of men with erectile dysfunction: efficacy and safety in a randomized, double-blind, placebo-controlled trial. J Androl 2002; 23: 763–771.
Porst H, Rosen R, Padma-Nathan H, Goldstein I, Giuliano F, Ulbrich E et al. The efficacy and tolerability of vardenafil, a new, oral, selective phosphodiesterase type 5 inhibitor, in patients with erectile dysfunction: the first at-home clinical trial. Int J Impot Res 2001; 13: 192–199.
Brock G, Nehra A, Lipshultz LI, Karlin GS, Gleave M, Seger M et al. Safety and efficacy of vardenafil for the treatment of men with erectile dysfunction after radical retropubic prostatectomy. J Urol 2003; 170: 1278–1283.
Goldstein I, Young JM, Fischer J, Bangerter K, Segerson T, Taylor T et al. Vardenafil, a new phosphodiesterase type 5 inhibitor, in the treatment of erectile dysfunction in men with diabetes: a multicenter double-blind placebo-controlled fixed-dose study. Diabetes Care 2003; 26: 777–783.
Hatzichristou D, Montorsi F, Buvat J, Laferriere N, Bandel TJ, Porst H et al. The efficacy and safety of flexible-dose vardenafil (levitra) in a broad population of European men. Eur Urol 2004; 45: 634–641.
Burris AS, Banks SM, Sherins RJ . Quantitative assessment of nocturnal penile tumescence and rigidity in normal men using a home monitor. J Androl 1989; 10: 492–497.
Rosen RC, Riley A, Wagner G . The International Index of Erectile Function (IIEF): a multidimensional scale for assessment of erectile dysfunction. Urology 1997; 49: 822–830.
Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996; 17: 1–12.
Higgins JP, Thompson SG, Deeks JJ, Altman DG . Measuring inconsistency in meta-analyses. BMJ 2003; 327: 557–560.
Bucher HC, Guyatt GH, Griffith LE, Walter SD . The results of direct and indirect treatment comparisons in meta-analysis of randomized controlled trials. J Clin Epidemiol 1997; 50: 683–691.
Deeks JJ, Altman DG, Bradburn MJ . Statistical methods for examining heterogeneity and combining results from several studies in meta-analysis. In: Egger M, Smith GD, Altman DG, (eds.), Systematic Reviews in Health Care: Meta-Analysis in Context. London: BMJ Publishing Group, 2003 pp. 248–282.
Song F, Altman DG, Glenny AM, Deeks JJ . Validity of indirect comparison for estimating efficacy of competing interventions: empirical evidence from published meta-analyses. BMJ 2003; 326: 472.
Montorsi F, McDermott TED, Morgan R, Olsson A, Schultz A, Kirkeby HJ et al. Efficacy and safety of fixed-dose oral sildenafil in the treatment of erectile dysfunction of various etiologies. Urology 1999; 53: 1011–1018.
Padma-Nathan H, Stecher VJ, Sweeney M, Orazem J, Tseng L-J, deRiesthal H . Minimal time to successful intercourse after sildenafil citrate: results of a randomized, double-blind, placebo-controlled trial. Urology 2003; 62: 400–403.
Shirai M . Randomized, placebo-controlled, double-blind study of oral sildenafil in Japanese men with erectile dysfunction. Nishinihon J Urol 2000; 62: 373–382.
Goldstein I, Lue TF, Padma-Nathan H, Rosen RC, Steers WD, Wicker PA . Oral sildenafil in the treatment of erectile dysfunction. New Engl J Med 1998; 338: 1397–1404.
Lue TF . A study of sildenafil (viagra), a new oral agent for the treatment of male erectile dysfunction. J Urol 1997; 157: 181.
Brock G, Iglesias J, Toulouse K, Ferguson KM, Pullman WE, Anglin G . Efficacy and safety of IC351 treatment for ED. Eur Urol 2001; 39: 106.
Faria GE, Torres LO, Koff WJ, Claro JA, Cairoli CD, Fregonesi A et al. Efficacy and safety of tadalafil in Brazilian men with ED. Int J Impot Res 2003; 15: 10.
Padma-Nathan H, McMurray JG, Pullman WE, Whitaker JS, Saoud JB, Ferguson KM et al. On-demand IC351 (Cialis) enhances erectile function in patients with erectile dysfunction. Int J Impot Res 2001; 13: 2–9.
U.S. Food and Drug Administration/Center for Drug Evaluation and Research. NDA-21368 [review]. http://www.fda.gov/cder/approval/index.htm.2004.
Hellstrom WJ, Gittelman M, Karlin G, Segerson T, Thibonnier M, Taylor T et al. Sustained efficacy and tolerability of vardenafil, a highly potent selective phosphodiesterase type 5 inhibitor, in men with erectile dysfunction: results of a randomized, double-blind, 26-week placebo-controlled pivotal trial. Urology 2003; 61: 8–14.
U.S. Food and Drug Administration/Center for Drug Evaluation and Research. NDA-21400 [review]. http://www.fda.gov/cder/approval/index.htm.2003.
Berner M, Kriston L, Harms A . Covariates of treatment effect in trial with oral sildenafil in the treatment of erectile dysfunction – a meta-analytic regression analysis. J Sex Med 2005; 2: 56.
Montorsi F, Verheyden B, Meuleman E, Junemann KP, Moncada I, Valiquette L et al. Long-term safety and tolerability of tadalafil in the treatment of erectile dysfunction. Eur Urol 2004; 45: 339–345.
Potempa AJ, Ulbrich E, Bernard I, Beneke M, Vardenafil Study Group. Efficacy of vardenafil in men with erectile dysfunction: a flexible-dose community practice study. Eur Urol 2004; 46: 73–79.
Benchekroun A, Faik M, Benjelloun S, Bennani S, El Mrini M, Smires A . A baseline-controlled, open-label, flexible dose-escalation study to assess the safety and efficacy of sildenafil citrate (Viagra) in patients with erectile dysfunction. Int J Impot Res 2003; 15: 19–24.
Conflict of interest
There was no specific funding for this project. MB currently holds research grants from the German Ministry for Education and Research, Boehringer Ingelheim Inc., Pfizer Inc. and Willmar Schwabe Inc. He did receive tuition fees from Glaxo Smith Kline, Hormosan Kwizda, Lilly, Pfizer and Willmar Schwabe. He also received travel expenses from the European Sexual Dysfunction Alliance affiliate (ISG e.V.) that receives sponsoring from the pharmaceutical industry. LK did receive travel expenses reimbursement from Pfizer and Wilmar Schwabe. AH holds a paid part time position with the European Sexual Dysfunction Alliance affiliate (ISG e.V.).
About this article
Cite this article
Berner, M., Kriston, L. & Harms, A. Efficacy of PDE-5-inhibitors for erectile dysfunction. A comparative meta-analysis of fixed-dose regimen randomized controlled trials administering the International Index of Erectile Function in broad-spectrum populations. Int J Impot Res 18, 229–235 (2006). https://doi.org/10.1038/sj.ijir.3901395
- phosphodiesterase inhibitors
- treatment outcome
- randomized controlled trials
Direct comparison of tadalafil with sildenafil for the treatment of erectile dysfunction: a systematic review and meta-analysis
International Urology and Nephrology (2017)
International Journal of Impotence Research (2008)
The application of quantitative methods for identifying and exploring the presence of bias in systematic reviews: PDE-5 inhibitors for erectile dysfunction
International Journal of Impotence Research (2008)
International Journal of Impotence Research (2007)