Erectile dysfunction (ED) is a prevalent condition associated with complex psychological and interpersonal issues. The Psychological and Interpersonal Relationship Scales (PAIRS) is a valid and reliable measure of Sexual Self-Confidence, Spontaneity, and Time Concerns associated with ED and its treatment. This paper describes the development and validation of a 15-item, short form of PAIRS (SF-PAIRS). Domain scores and psychometric properties of SF-PAIRS were assessed using data from the validation studies of the full-length (23-item) PAIRS. Items in the two briefer scales (Spontaneity and Time Concerns) were identified using internal consistency analyses and crossvalidated during two multinational clinical trials and a community sample. The scales in SF-PAIRS showed good internal consistency, convergent and discriminant validity, were stable over a 2-week period, and showed sensitivity to treatment. SF-PAIRS is a valid, brief measure of psychological and interpersonal outcomes associated with ED.
Erectile dysfunction (ED) is a disorder that is associated with decreased psychological well-being and quality of life.1, 2, 3, 4 ED is a prevalent condition, although estimates differ across study designs and regions.5, 6, 7 A recent, large multinational study of men aged 20–75 years showed an overall prevalence rate of 16%, with prevalence rates ranging from 10 to 22% among men in various countries.8 A widely cited US study found that 52.0% of men aged 40–70 years experienced some degree of ED,9 and even higher prevalence rates of ED have been reported among men with comorbid conditions related to ED such as hypertension and diabetes.8, 10
The Psychological and Interpersonal Relationship Scales (PAIRS) is a new self-report measure that was developed and validated in conjunction with clinical trials of tadalafil (Lilly ICOS LLC, Indianapolis, IN and Bothell, WA, USA), an inhibitor of phosphodiesterase type 5 (PDE5) approved globally for the treatment of ED.11 The full-length PAIRS consists of 23 items that measure three conceptual domains associated with ED and current treatments for ED: Sexual Self-Confidence, Spontaneity and Time Concerns associated with sexual encounters.
PAIRS domains and items were developed from several sources, including patient interviews and focus groups, market research, clinical experts in the field of ED, and from the scientific literature on ED. Factor analyses were conducted to establish and confirm the final set of items. The psychometric characteristics of PAIRS were evaluated in four questionnaire studies associated with tadalafil trials.12 The scales demonstrated good internal consistency across samples, adequate test–retest reliability and convergent and divergent validity, and showed differentiation in a community sample of men with and without ED. An analysis of PAIRS in the community sample suggested that PAIRS domains were independent predictors of both sexual satisfaction and relationship satisfaction.13 In three post-treatment questionnaire studies that evaluated PAIRS scores associated with two PDE5 inhibitor therapies with different duration of action, tadalafil and sildenafil citrate (Pfizer Inc., New York, NY, USA), we found that tadalafil was associated with significantly higher Sexual Self-Confidence and Spontaneity scores, and lower Time Concerns scores.14 These findings supported the sensitivity of PAIRS to differentiate between PDE5 inhibitor treatments and provided further evidence of the measure's validity.15
For many studies it may not be possible to administer an instrument the length of PAIRS. This paper describes the development and psychometric characteristics of a 15-item, short form of PAIRS (SF-PAIRS) that can be used when a briefer scale is needed. SF-PAIRS consists of the full Time Concerns domain, and abbreviated Sexual Self-Confidence and Spontaneity domains. As the validation information on the Time Concerns domain has been published elsewhere,12, 14 this paper will focus primarily on the psychometric characteristics of the abbreviated domains.
Participants and procedures
The domain scores and psychometric properties of SF-PAIRS were assessed from the validation studies of the full-length (23-item) PAIRS. These data were evaluated in several questionnaire-based studies; four initial validation studies (Table 1; see Swindle et al.12 for a complete description of study samples, procedures and results), and three additional studies (Studies 5–7) to assess the sensitivity of PAIRS to PDE5 inhibitor treatments.14
In Studies 1–3 and 5–7 men with ED were recruited from clinical trials of tadalafil to complete voluntary, post-treatment, self-report questionnaires containing PAIRS. The trials and questionnaire studies were conducted in accordance with the ethical principles from the latest version of the Declaration of Helsinki and in conformity with the applicable laws and regulations of each country. The ethical review boards at the participating sites approved protocols for the clinical trials and questionnaire studies, and all participants provided written informed consent for all studies. Patients recruited from the clinical trials were men 18 years and older who reported at least a 3-month history of ED prior to study entry. ED was defined as the consistent inability to achieve and/or maintain an erection sufficient to permit satisfactory sexual intercourse and was diagnosed based on physicians’ clinical assessment.
Study 4 consisted of men with and without ED aged 40–70 years who were recruited from a standing, nationally representative panel of volunteers who participate in Internet-based surveys. Respondents were anonymous to the study authors. The study design and confidentiality safeguards conformed to US Federal Trade Commission privacy protection standards. We also obtained 2-week test–retest information on PAIRS items in a subset of these men (n=254). There were no differences on any measures or demographic characteristics between men who completed PAIRS once and those who completed PAIRS at both time points.
Items for SF-PAIRS were selected using internal consistency reliability analyses based on the items comprising the full-length PAIRS scales. In Study 2 (N=201), the full-length scales were examined for item reduction using Cronbach's alpha, with the goal of reducing the number of items and maximizing internal consistency. Each item was evaluated for its impact on alpha if removed. Items were removed sequentially in order of the least reduction in overall alpha. Items were removed until a substantial reduction in alpha would occur if any more items were removed. The Time Concerns domain in SF-PAIRS consists of the full 8-item scale in the full-length PAIRS in order to maintain adequate psychometric characteristics of the scale. The Sexual Self-Confidence domain was reduced from six to three items in SF-PAIRS, and the Spontaneity domain was reduced from nine to four items. The internal consistencies of the two new SF-PAIRS scales were then cross-validated in Studies 3 and 4, demonstrating no essential loss in internal consistency when applied to new samples.
Assessment of treatment sensitivity
The sensitivity of PAIRS and SF-PAIRS was evaluated using two methods. In Study 3, we examined whether domains were responsive to self-reported improvement in sexual function after 12 weeks of tadalafil treatment or placebo. Participants were classified as treatment responders or nonresponders based on International Index of Erectile Function (IIEF) erectile function (EF) domain scores.16 Possible scores range from 1 to 30; participants with scores of 22 or higher (indicating mild or no ED)17 at the end of treatment were classified as treatment responders (n=71, 63.4%), and those with scores of 21 or lower were classified as treatment nonresponders (n=38, 33.9%). Sensitivity was assessed by comparing post-treatment scores on PAIRS domains in these two responder groups using analysis of covariance (ANCOVA), controlling for pretreatment scores.
Sensitivity was also assessed by evaluating scores associated with the use of two PDE5 inhibitor therapies with different pharmacokinetic characteristics, tadalafil and sildenafil.14 As a result of the different characteristics of these medications, we hypothesized that PAIRS domains would demonstrate treatment sensitivity if tadalafil was associated with higher Sexual Self-Confidence and Spontaneity scores, and lower Time Concerns scores compared to sildenafil.
We examined sensitivity using this method in three voluntary, post-treatment questionnaire studies (see Rosen et al.14 for complete description of study samples, procedures and limitations). The first study (Study 5) consisted of a subsample of men from Study 2; PAIRS was assessed after 6 months of open-label tadalafil (20 mg) treatment and after 2 months of real-world sildenafil use (25–100 mg). In the other two studies (Studies 6 and 7), PAIRS was assessed at the end of treatment. In Study 6, men with a 6–24-week history of sildenafil use (25–100 mg) received 3 weeks open-label sildenafil treatment (dosage at study entry), and then switched to 9 weeks tadalafil (20 mg) treatment (for complete details on the clinical trial, see Stroberg et al.18). In Study 7, a double blind, crossover study, men were randomized to 12 weeks of either tadalafil (20 mg) or sildenafil (50 mg) treatment, followed by 12 weeks of the alternate treatment. Patients on sildenafil treatment had the option to increase their dose to 100 mg (limited to 35% of patients in each treatment period to reflect dose usage in clinical practice). PAIRS scores for the second treatment period were evaluated (for complete details on the clinical trial, see Von Keitz et al.19).
Items included in SF-PAIRS are presented in Table 2. In early validation studies, items in the Spontaneity and Time Concerns domains were scored on 1–4 Likert-type scale (Strongly Disagree to Strongly Agree), and Sexual Self-Confidence items were scored on a 1–5 Likert-type scale (Neutral response option included). In the PAIRS and SF-PAIRS measures currently being implemented in clinical trials (data not presented here) all items are scored using the 1–4 scale to maintain consistency across domains. Mean scores for PAIRS domains were derived from individual items. For each participant, if at least one-half of the items that comprised each domain were valid (i.e., nonmissing), then mean scores were computed from the completed (nonmissing) items. If more than one-half of the items in a domain were missing, then the domain score was considered missing for that participant. Higher scores indicated greater Sexual Self-Confidence, Spontaneity and Time Concerns.
Mean scores and standard deviations for the Sexual Self-Confidence and Spontaneity domains of SF-PAIRS and PAIRS are presented in Table 3. Mean scores for the 3-item Sexual Self-Confidence domain were slightly higher compared to the 6-item version of this scale in two of the studies (Study 2; Study 3, Time 1), although scores in the other studies were very similar. Mean scores for the Spontaneity domain were virtually identical for the 4-item and 9-item versions.
Across Studies 2–4, correlation coefficients between the full 6-item and short 3-item versions of the Sexual Self-Confidence domain ranged from 0.94 to 0.98, and correlation coefficients for the full 9-item and short 4-item versions of the Spontaneity domain ranged from 0.95 to 0.97, suggesting a very high degree of commonality between the short and long versions of these domains.
We estimated the internal consistency reliability of the Sexual Self-Confidence and Spontaneity domains of SF-PAIRS by examining Cronbach alphas across studies. Internal consistency reliability for the shortened domains was good to excellent in Studies 2–4 (Table 4). For the 3-item Sexual Self-Confidence domain, Cronbach alphas ranged from 0.84 to 0.96, and were only slightly lower than the full 6 items in most studies. For the 4-item Spontaneity domain, Cronbach alphas ranged from 0.79 to 0.85, which were somewhat lower than alphas for the 9-item version of this domain across studies (0.067 lower, on average), as would be expected with a reduction in items. Alphas were consistent with those of the Time Concerns domain, 0.73–0.81, which is represented in full in SF-PAIRS. Internal consistency reliability for Studies 5–7 was also good to excellent, and consistent with that from Studies 2–4 (alphas not shown).
We assessed the 2-week test–retest reliability of the SF-PAIRS domains using Pearson correlation coefficients in the Study 4 participants who completed PAIRS items at two time periods (n=254). Item-level test–retest Pearson correlation coefficients ranged from 0.67 to 0.76 for individual items in the 3-item Sexual Self-Confidence scale (average item test–retest correlation=0.71); correlations ranged from 0.39 to 0.52 for items in the 4-item Spontaneity scale (average item test–retest correlation=0.46). The test–retest correlation coefficient for the 3-item Sexual Self-Confidence scale was good, r=0.76, and similar to that of the 6-item scale (r=0.77). Test–retest correlation for the 4-item Spontaneity scale was low, r=0.59, and slightly lower than that of the 9-item Spontaneity scale (r=0.66).
Convergent and discriminant validity
We assessed the validity of the PAIRS-SF in Study 2 by examining the extent to which scales were correlated with measures of similar (convergent validity) and unrelated (divergent validity) constructs (Table 5). Validity correlations for the Sexual Self-Confidence and Spontaneity domains showed the same patterns for the full-length and short-form PAIRS, with the strength of correlations slightly lower for SF-PAIRS domains. As expected for the theoretical framework of PAIRS,12 measures of sexual function, sexual satisfaction and psychological aspects of sexuality were highly correlated with Sexual Self-Confidence, whereas they were moderately with the Spontaneity domain. PAIRS domains, which are sexuality-specific, were not correlated with general self-esteem, showing divergent validity. The relationship measures were more highly correlated with the interpersonal domain of Spontaneity than with Sexual Self-Confidence.
Although PAIRS/SF-PAIRS is not meant to be used as a diagnostic instrument for ED, known groups discriminant validity was explored in SF-PAIRS domains in men with and without ED (Study 4). Men (N=801) were defined as having ED (n=212) if they reported physician-diagnosed ED in the past 12 months (n=72), had an IIEF EF domain score of 16 or below (indicating moderate to severe ED in the past 4 weeks, n=141),17 or had reported ever using sildenafil (n=101). The remaining men were considered to have normal erectile function (n=589). Similar to the full-length versions, Sexual Self-Confidence (t799=16.18; P<0.001) and Spontaneity (t799=7.60; P<0.001) scores were significantly higher in men without ED compared to men with ED.13
In Study 3, men classified as treatment responders reported significantly higher mean scores on the 3-item Sexual Self-Confidence scale (mean score responders=4.09, s.d.=1.06; mean score nonresponders=2.54, s.d.=1.23; F1, 82=31.55; P<0.001) and on the 4-item Spontaneity scale (mean score responders=3.40, s.d.=0.48; mean score nonresponders=2.95, s.d.=0.67; F1, 82=12.94; P=0.001) compared to nonresponders. These findings are similar to those for the full-length domains and suggest that the short-form domains are sensitive to men's self-reported response to PDE5 inhibitor treatment.
In Studies 5–7, we compared scores for tadalafil and sildenafil on the two reduced-item SF-PAIRS domains and found evidence to support the sensitivity of the short scales to PDE5 treatments similar to that found for the full-length PAIRS. In general, compared to sildenafil, tadalafil was associated with higher Sexual Self-Confidence and Spontaneity scores (Table 6). The exception was that in the third comparative study (Study 7), mean differences for the 3-item Sexual Self-Confidence domain did not quite reach statistical significance (P=0.057).
Translations and linguistic validation
PAIRS items have been translated into numerous languages and linguistically validated by the MAPI Research Institute (Lyon, France). In general, the linguistic validation process included conceptual analysis of the items, forward and backward translation, expert review, cognitive debriefing with target respondents and international harmonization of translations. This process ensured that the translations were conceptually equivalent to the original items written in US English, and that the translated items were easily understood by the intended users. The psychometric characteristics (validity and reliability) of the MAPI-translated versions of PAIRS have yet to be established; however, some of the original validation studies were multinational and the translations (forward translations with verification) used in those studies showed consistent psychometric characteristics.
This paper described the development and psychometric properties of a 15-item, briefer version of the Psychological and Interpersonal Relationship Scales (PAIRS). SF-PAIRS is comprised of the same three domains as PAIRS; the Sexual Self-Confidence and Spontaneity domains have a reduced number of items, whereas the Time Concerns domain has the same number of items as the full-length PAIRS. As the psychometric properties of the full-length PAIRS (including the Time Concerns domain) are published elsewhere,12, 13, 14 the focus of this paper was on the two reduced-item domains.
Overall, there were very high correlations between the long and short versions of the Sexual Self-Confidence and Spontaneity domains of PAIRS across studies. In addition, the psychometric characteristics of the two reduced-item domains were similar to those of the full-length versions, with a couple of exceptions. The reduced-item domains showed good internal consistency reliability; the Cronbach alphas for the 4-item Spontaneity domain were lower than they were for the 9-item scale across studies, as would be expected with fewer items. Similarly, the 2-week test–retest reliability correlations were slightly lower for the 4-item Spontaneity domain. A potential limitation of the spontaneity scale, especially the shorter version, is the relatively low test–retest correlation. However, because sexual experiences may fluctuate over time, these correlations are still reasonable for the constructs measured by PAIRS.
Another difference found by reducing the number of items concerned the sensitivity of the 3-item Sexual Self-Confidence scale in Study 7. The 6-item scale showed a significant mean difference on scores for sildenafil compared to scores for tadalafil at P<0.05, whereas the difference in mean scores using the 3-item scale became nonsignificant, P=0.057. Across Studies 5–7, there were smaller effect sizes for the mean differences between treatments using the 3-item Sexual-Self Confidence scale compared to mean differences found for the 6-item version.14 This is likely due to the higher mean scores observed for the sildenafil assessment and larger standard deviations found with the 3-item scale. Thus, although the 3-item Sexual Self-Confidence scale demonstrates sensitivity, it appears to be slightly less sensitive than the full 6-item scale. Researchers may want to consider using the full 6-item set when the research goal is to demonstrate treatment differences and administration time permits the use of additional items.
One limitation is that all of the data reported in this paper were collected in the context of the full-item set; we do not yet have data on the SF-PAIRS items administered alone. Although this method of creating reduced-item measures from items administered in their entirety has been used elsewhere,20, 21 there exists the possibility of a context effect of SF-PAIRS items being embedded with the additional 8 items. The consequence of a context effect is that there could be an artificially high correlation between the short and long versions of the scales. Therefore, the psychometric characteristics of SF-PAIRS items administered alone will need to be evaluated further and compared to studies using the full-length PAIRS.
In summary, SF-PAIRS offers a briefer assessment of psychological and interpersonal domains associated with ED and its treatment. SF-PAIRS generally maintains the psychometric integrity of the full-length PAIRS; differences found between versions are not unexpected given the reduction in items. SF-PAIRS offers brevity for clinical use and in research applications when there is not enough time for the full 23-item PAIRS.
NIH Consensus Conference. NIH consensus development panel on impotence. JAMA 1993; 270: 83–90.
Althof SE . Quality of life and erectile dysfunction. Urology 2002; 59: 803–810.
Litwin MS, Nied RJ, Dhanani N . Health-related quality of life in men with erectile dysfunction. J Gen Intern Med 1998; 13: 159–166.
Shabsigh R, Klein LT, Seidman S, Kaplan SA, Lehrhoff BJ, Ritter JS . Increased incidence of depressive symptoms in men with erectile dysfunction. Urology 1998; 52: 848–852.
Lewis RW, Fugl-Meyer KS, Bosch R, Fugl-Meyer AR, Laumann EO, Lizza E et al. Epidemiology/risk factors of sexual dysfunction. J Sex Med 2004; 1: 35–39.
Laumann EO, Nicolosi A, Glasser DB, Paik A, Gingell C, Moreira E et al. Sexual problems among women and men aged 40–80 y: prevalence and correlates identified in the Global Study of Sexual Attitudes and Behaviors. Int J Impot Res 2005; 17: 39–57.
Nicolosi A, Laumann EO, Glasser DB, Moreira A, Paik A, Gingell C et al. Sexual behavior and sexual dysfunctions after age 40: the Global Study of Sexual Attitudes and Behaviors. Urology 2004; 64: 991–997.
Rosen RC, Fisher WA, Eardley I, Niederberger C, Nadel A, Sand M . The Multinational Men's Attitudes to Life Events and Sexuality (MALES) study: prevalence of erectile dysfunction and related health concerns in the general population. Curr Med Res Opin 2004; 20: 607–617.
Feldman HA, Goldstein I, Hatzichristou DG, Krane RJ, McKinlay JB . Impotence and its medical and psychosocial correlates: results of the Massachusetts Male Aging Study. J Urol 1994; 151: 54–61.
Guiliano FA, Leriche A, Jaudinot EO, Solesse de Gendre A . Prevalence of erectile dysfunction among 7689 patients with diabetes or hypertension, or both. Urology 2004; 64: 1196–1201.
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.
Swindle RW, Cameron AE, Lockhart DC, Rosen RC . The Psychological and Interpersonal Relationship Scales: assessing psychological and relationship outcomes associated with erectile dysfunction and its treatment. Arch Sex Behav 2004; 33: 19–30.
Cameron A, Rosen RC, Swindle RW . Sexual and relationship characteristics among an internet-based sample of US men with and without erectile dysfunction. J Sex Marital Ther 2005; 31: 229–242.
Rosen RC, Broderick GA, Shabsigh R, Swindle RW, Lockhart D, Cameron A . Sensitivity of the Psychological and Interpersonal Relationship Scales to Oral Therapies for Erectile Dysfunction. J Sex Med 2005; 2: 461–468.
Hays RD, Hadorn D . Responsiveness to change: an aspect of validity, not a separate dimension. Qual Life Res 1992; 2: 73–76.
Rosen RC, Riley A, Wagner G, Osterloh IH, Kirkpatrick J, Mishra A . The International Index of Erectile Function (IIEF): a multidimensional scale for assessment of erectile dysfunction. Urology 1997; 49: 822–830.
Cappelleri JC, Rosen RC, Smith MD, Mishra A, Osterloh IH . Diagnostic evaluation of the erectile function domain of the International Index of Erectile Function. Urology 1999; 54: 346–351.
Stroberg P, Murphy A, Costigan T . Switching patients with erectile dysfunction from sildenafil citrate to tadalafil: results of a European multicenter, open-label study of patient preference. Clin Ther 2003; 25: 2724–2737.
von Keitz A, Rajfer J, Segal S, Murphy A, Denne J, Costigan T et al. A multicenter, randomized, double-blind, crossover study to evaluate patient preference between tadalafil and sildenafil. Eur Urol 2004; 45: 499–509.
Rosen RC, Cappelleri JC, Smith MD, Lipsky J, Pena BM . Development and evaluation of an abridged, 5-item version of the International Index of Erectile Function (IIEF-5) as a diagnostic tool for erectile dysfunction. Int J Impot Res 1999; 11: 319–326.
Ware JE, Kosinski M, Keller SD . A 12-item short form health survey: construction of scales and preliminary tests of reliability and validity. Med Care 1996; 34: 220–233.
Snell Jr WE . The multidimensional sexual self-concept questionnaire. In: Davis CM, Yarber WL, Bauserman R, Schreer G, Davis SL (eds.), Handbook of Sexuality-Related Measures. Thousand Oaks: Sage, 1998, pp. 521–524.
Fichten CS, Spector I, Amsel R, Creti L, Brender W, Libman E . Sexual self-efficacy scale-erectile functioning. In: Davis CM, Yarber WL, Bauserman R, Schreer G, Davis SL (eds.), Handbook of Sexuality-Related Measures. Thousand Oaks: Sage, 1998, pp. 534–537.
Veroff J, Kulka RA, Douvan E . Mental Health in America: Patterns of Help-Seeking from 1957 to 1976. New York: Basic Books, 1981 352pp.
Hendrick S . A generic measure of relationship satisfaction. J Marriage Fam 1988; 50: 93–98.
Sprecher S, Regan PC . Passionate and companionate love in courting and young married couples. Sociol Inq 1998; 68: 163–185.
Cohen J . Statistical Power Analysis for the Behavioral Sciences 2nd edn. Hillsdale, NJ: Lawrence Erlbaum, 1988 567pp.
About this article
Cite this article
Swindle, R., Cameron, A. & Rosen, R. A 15-item short form of the Psychological and Interpersonal Relationship Scales. Int J Impot Res 18, 82–88 (2006). https://doi.org/10.1038/sj.ijir.3901381
- erectile dysfunction
- psychosocial assessment
- sexual self-confidence
- treatment outcomes
Improving sexual health in men with prostate cancer: randomised controlled trial of exercise and psychosexual therapies
BMC Cancer (2014)
Relationship between erectile function and sexual self-confidence: a path analytic model in men being treated with tadalafil
International Journal of Impotence Research (2014)
Establishing the content validity of the confidence in performing sexual intercourse and difficulty in performing sexual intercourse questionnaires
International Journal of Impotence Research (2013)
Sex Roles (2009)
ProsCan for Couples: Randomised controlled trial of a couples-based sexuality intervention for men with localised prostate cancer who receive radical prostatectomy
BMC Cancer (2008)