Pilot intervention to enhance sexual rehabilitation for couples after treatment for localized prostate carcinoma


  • Andrea L. Canada Ph.D.,

    Corresponding author
    1. Department of Behavioral Science, The University of Texas M. D. Anderson Cancer Center, Houston, Texas
    • Department of Behavioral Science, Unit 1330, University of Texas M. D. Anderson Cancer Center, – P.O. Box 310439, Houston, TX 77230-1439===

    Search for more papers by this author
    • Fax: (713) 745-4286

  • Leah E. Neese M.A.,

    1. Department of Behavioral Science, The University of Texas M. D. Anderson Cancer Center, Houston, Texas
    Search for more papers by this author
  • Dawen Sui M.S.,

    1. Department of Behavioral Science, The University of Texas M. D. Anderson Cancer Center, Houston, Texas
    Search for more papers by this author
  • Leslie R. Schover Ph.D.

    1. Department of Behavioral Science, The University of Texas M. D. Anderson Cancer Center, Houston, Texas
    Search for more papers by this author

  • This study was presented in part at the Annual Meeting of the Society for Behavioral Medicine, Baltimore, Maryland, March 24–27, 2004.



The majority of prostate carcinoma survivors experience enduring sexual difficulties and associated distress in the years after definitive treatment. A counseling intervention aimed at improving levels of sexual satisfaction and increasing successful utilization of medical treatment for erectile dysfunction (ED) was developed and pilot-tested for both the survivor of prostate carcinoma and his partner.


All male participants were 3-month to 5-year survivors of localized prostate carcinoma who had been treated with radical prostatectomy or radiation therapy, and were married or in a committed relationship. Couples were randomized to attend four sessions of counseling together or to have the man attend alone. In both groups, partners completed behavioral homework. The sessions included education on prostate carcinoma and sexual function and options to treat ED as well as sexual communication and stimulation skills. Standardized questionnaires at baseline, posttreatment, and at 3-month and 6-month follow-up assessed sexual function, marital adjustment, psychologic distress, and utilization of treatments for ED.


Fifty-one of 84 couples randomized to treatment completed the intervention (61%). Attendance by the partner did not affect outcomes. Participants completing the intervention demonstrated improvment in male overall distress (P < 0.01), male global sexual function (P < 0.0001), and female global sexual function (P < 0.05) at 3-month follow-up, but regression toward baseline was noted at 6-month follow-up. However, utilization of ED treatments increased from 31% at the time of study entry to 49% at the 6-month follow-up (P = 0.003).


The results of this brief pilot counseling intervention demonstrated significant gains in sexual function and satisfaction and increased utilization of treatments for ED. However, modifications are needed in future randomized trials to reduce the rate of premature termination and to improve long-term maintenance of gains. Cancer 2005. © 2005 American Cancer Society.

Despite efforts to reduce the sexual morbidity of radical prostatectomy and definitive radiation therapy for localized prostate carcinoma, the great majority of men have erectile dysfunction (ED) at long-term follow-up. In a prospective cohort study of more than 33,000 nonphysician male health professionals, the 2109 who developed prostate carcinoma had rates of ED that were 10–15-fold higher than their same-aged peers, with rates of ED ranging from 65% in men ages 50–54 years to over 90% for men ages 85–90 years.1 The 5-year follow-up data from the Prostate Cancer Outcomes Study, which comprised a national sample of 1187 men treated either with surgery or external beam radiation therapy for localized disease, demonstrated that 79% of men had ED after radical prostatectomy and 64% after radiation therapy.2 In our own detailed cross-sectional survey of 1236 men treated for localized prostate carcinoma at 1 tertiary referral center, 85% of men reported ED at a mean of 4.3 years posttreatment, with no significant difference noted between those who had undergone surgery versus radiation therapy as their primary therapy.3 The prevalence of ED in the youngest quintile of the sample (ages 42–62 yrs), was 65%, increasing to 90% of men ages 76–88 years. Men who underwent bilateral nerve-sparing prostatectomy or brachytherapy had the best recovery of erectile function, with 18% and 19%, respectively, achieving reliable, firm erections without any use of a medical/surgical treatment for ED. These numbers increased to 33% and 26%, respectively, with the use of an ED treatment. Approximately 60% of the prostate cancer survivors with ED rated the problem as significantly distressing.2, 3

Survivors also indicate a desire for more information concerning the treatment of ED. In a sample of 204 men treated 1–2 years previously, information regarding the management of ED was rated seventh on a list of 135 unmet needs weighted by importance.4 In a Canadian survey of 200 men treated for prostate carcinoma within the past 3 years, 61% wanted more information concerning sexual rehabilitation than they had received.5

With the advent and popularization of the phosphodiesterase type 5 (PDE5) inhibiting drugs, more men have been consulting physicians about sexual dysfunction. A recent survey of 4422 men with ED from eight industrialized nations found that 58% had seen a physician, with 41% specifically requesting a PDE5-inhibitor.6 Another multinational survey of 1930 men with ED recruited from physician offices reported that 46% had sought help, including 56% of patients from the U.S.7 Similarly, after treatment for localized prostate carcinoma, 59% of men we surveyed had tried a treatment for ED, with 52% opting for PDE5-inhibitors.8

Unfortunately, prostate carcinoma survivors are less likely to be able to use PDE5-inhibitors successfully than men with other causes for their ED. In a 2.5-year follow-up of a large sample of men prescribed sildenafil, only 38% of 197 men who had undergone a radical prostatectomy were still using the medication compared with 54% of men with other causes for their ED (P < 0.001).9 Lack of efficacy was the most common reason for discontinuing the drug. In our own survey, only 39% of 549 prostate cancer survivors who tried sildenafil were still using it 4–5 years after surgery or radiation therapy.8 Even within the group of men still using sildenafil after prostate cancer treatment, many if not most did not achieve firm, reliable erections with the medication.8, 9 It is also unlikely that the two newer PDE5-inhibitors, tadalafil and vardenafil, will be more effective than sildenafil for survivors of prostate carcinoma because they impact the same pathway of neurotransmitters.10

Despite frequent disappointment with oral medication, men are reluctant to use more invasive treatments for ED, even though the efficacy of penile injection therapy, vacuum devices, or penile prosthetic surgery is far superior to that of sildenafil.8, 9 Less than 20% of men in our sample had tried any one of these therapies for ED. Not surprisingly, trying more than one type of treatment was a significant predictor of successful sexual rehabilitation outcome.3

ED is not the only sexual problem that occurs after treatment for prostate carcinoma. In the men we surveyed, 45% reported low desire for sex. Other surveys have found a prevalence of 50–55%.2, 11, 12 In our survey, 17% of prostate carcinoma survivors were unable to reach orgasm and 28% found their orgasms weak and disappointing.3 Men were just as distressed about desire and orgasm problems as about ED. In another sample of 239 men who had undergone radical prostatectomy, 37% complained of weak orgasms and 14% had significant pain at orgasm.13 Men also frequently lose urine at orgasm after prostatectomy, which may interfere with the resumption of sexual activity.11 Finally, radical prostatectomy results in penile shortening in up to 68% of men,14 which can be a source of low self-esteem. It is also important to note that two-thirds of men in committed relationships report that their female partner has at least one sexual problem, with loss of desire for sex the most common, at 42%.3

Improving sexual outcomes for prostate cancer survivors may also enhance their overall well-being. When men with ED of varied etiology are treated successfully, their perceived mental and physical quality of life improves.15, 16 Conversely, when men are unable to use PDE5-inhibitors successfully, their distress regarding ED may increase beyond its baseline level.16 In our previous survey, we identified several factors associated with successful sexual rehabilitation.8 Some, such as age and type of prostate carcinomatreatment, are not amenable to change. However, men who achieved a good sexual recovery had female partners who still enjoyed sex3 and had more positive attitudes on a new scale measuring cognitive beliefs regarding prostate carcinoma and sexuality.17 In particular, men who sought medical help for ED did not believe the problem to be a normal part of aging, but rather believed that staying sexually active was important to health and that a better sex life was also a priority for their partner. Satisfaction with and continued use of treatments for ED depended on willingness to try more invasive options until erections were normalized.17

In contrast to the plentiful literature on medical and surgical treatments for ED, to our knowledge only a few studies to date have evaluated psychosocial interventions for prostate carcinoma survivors,18, 19, 20 and none have focused primarily on sexuality. To determine what kind of treatment would be acceptable to survivors of prostate carcinoma, we conducted qualitative telephone interviews with 320 men from our survey who intended to seek help for ED in the next year, as well as with 164 of their sexual partners.21 Results suggested that both men and women preferred individual consultations with a health professional to a group format. Ninety-five percent of men would be willing to have a 1-on-1 consultation, whereas 74% would agree to have their partner present. In contrast, 93% of partners preferred a couple session compared with 84% who would be willing to see a health professional individually. Therefore, we designed our intervention as a randomized trial of two formats: four counseling sessions with the male alone versus the same content and duration but with both partners present. We did not use a usual care group in our study design because our survey data provided the benchmark of 85% of survivors having ED at 4.3-year follow-up after usual care.3 In addition, studies have documented that roughly 30% of survivors utilize a medical treatment for ED.6

To improve sexual rehabilitation after prostate carcinoma, treatment should also use techniques from sex therapy.22 It was important to include components within the intervention that would enhance the woman's sexual pleasure and desire, target negative beliefs that interfere with help-seeking for ED, and teach the partners skills in sexual communication and stimulation that would allow them to better integrate treatments for ED into their sex life.

Taking all of the above factors into consideration, we designed an intervention and hypothesized that the couple format would result in superior outcomes compared with the man alone condition, but that both formats would: 1) increase the successful utilization of medical treatments available for ED; and 2) improve sexual satisfaction for both the survivor of prostate carcinoma and his partner.



Men were eligible for the study if they: 1) had undergone treatment for localized prostate carcinoma (Stages A through C) with either surgery or radiation therapy between 3 and 60 months previously; 2) were married or living with a female partner for at least the past year who was also willing to participate; 3) were able to speak and read English; 4) resided within a reasonable distance of the University of Texas M. D. Anderson Cancer Center (UTMDACC), making repeated sessions possible; and 5) were willing to be randomized to 1 of 2 intervention conditions. Survivors were excluded from study participation if they were: 1) currently receiving hormonal therapy for prostate carcinoma (a treatment that directly reduces sexual desire); 2) able to achieve and maintain an erection sufficient for sexual intercourse without medical or mechanical assistance on ≥50% of attempts within the past 3 months; or 3) currently using a medical treatment for ED that was successful and satisfactory.


The UTMDACC Institutional Review Board approved this study. Men in the UTMDACC cancer registry who met eligibility criteria were mailed recruitment letters signed, when possible, by the primary treating surgeon or radiation oncologist or by another member of the faculty. Flyers were supplied to relevant UTMDACC patient care areas or the private practice offices of community urologists. Other recruitment strategies included physician encouragement during treatment follow-up and visits to the UTMDACC's Erectile Dysfunction (ED) Clinic and informational talks at community outreach programs such as cancer ministries at local African-American churches and Houston chapters of USToo and Man to Man.

Group Assignment

After both partners provided written informed consent, couples were assigned to attend four counseling sessions together or to have the man attend alone using a form of adaptive randomization termed minimization.23–26 Similar to stratification, minimization assigns participants to treatment conditions using chosen characteristics. However, minimization results in better group balance than stratification and does not suffer from the limitations of increasing the probability of group imbalance when several participant factors are used. Although statistical adjustment using covariates may improve estimates of treatment effect size, analyses are generally more credible when groups are comparable in demographic factors at baseline. Thus, minimization is preferable, particularly with smaller studies. The variables used to assign survivors to treatment conditions were age (60 yrs vs. >60 yrs), education (some college vs. college degree or more), ethnicity (white vs. other), and prostate carcinoma treatment category (surgery vs. radiation therapy).


Regardless of group assignment, couples received the intervention through either a Ph.D.-level clinical psychologist (A.L.C.) or a Master's-level counselor (L.E.N.). The treatment protocol was manualized and counselors received weekly supervision from a licensed psychologist (L.R.S.) to ensure standardization of the intervention. Table 1 presents an outline of the session content, which was identical in both couple and individual treatment formats. The first session, including a sexual history, was designed to take 60–90 minutes, with the remaining sessions lasting 45–60 minutes. Education was provided concerning the sexual impact of surgery or radiation therapy for prostate carcinoma, medical and surgical treatments for ED, ways to optimize the use of those treatments, and strategies to cope during sexual activity with urinary incontinence or postmenopausal vaginal atrophy. Skill training was provided to improve sexual communication, general communication of feelings, and open expression of affection. Cognitive-behavioral techniques targeted to each partner addressed negative beliefs regarding cancer and sexuality.

Table 1. Content of the Intervention Sessionsa
  • ED: erectile dysfunction.

  • a

    A therapist manual was created for the study and is available on request from the authors.

1• History of prostate carcinoma, couple's relationship, and sexuality• Each partner rates his/her willingness to try each possible treatment for ED
 • Reasons that sexual problems occur after prostate carcinoma treatment• Partners compare ratings and arrive at plan to seek help.
 • Medical treatments available for ED, with pros and cons of each 
 • Planning for partners to work together to choose the optimal treatment for ED 
2• Communication in relationships• Each partner suggests 2 steps to improve general communication in the relationship and 2 to improve sexual communication
 • How to talk about cancer-related emotions
 • How to talk about the sexual problem• First sensate focus exercise
 • Specific sexual communication skills 
 • Impact of the sexual problem on the relationship 
3• How to get started in resuming sex after cancer• Yes's and no's exercise in sexual assertiveness
 • Enjoying sex without a firm erection• Second sensate focus exercise
 • How to rekindle lost desire 
 • Feeling pleasure despite ED 
 • Using varied touch to achieve noncoital orgasms 
 • Coping with urinary incontinence 
 • Coping with menopausal dryness 
 • Planning for better sexual encounters 
4• Viewing cancer as a challenge for positive sexual change• Each partner defines 3 steps towards improving sex and his/her role in making them happen
 • Promoting feelings of passion, play and love during sex
 • Attitudes about sexual fantasy and variety• Share steps and agree on plan for the next month
 • Troubleshooting with unsatisfying medical treatments for ED 

Detailed informational handouts and behavioral exercise instructions were sent home with participants after each session. Partners were asked to complete identical behavioral homework assignments and questionnaire assessments in both the man alone and couple conditions. Examples of specific homework assignments include completing touching exercises designed to increase intimacy without intercourse; establishing a ‘plan of action’ to obtain, utilize, and troubleshoot medical treatments for ED; and keeping a ‘caring diary’ or record of behavioral goals intended to increase the expression of affection in the relationship. Homework assigned during the previous session was reviewed at the start of the next session. At the end of Session 4, participants set goals for the next month and scheduled a follow-up phone appointment to discuss their progress.

Assessment Measures

Both partners completed self-report questionnaires at the time of entry to the study, after the last session of the intervention, and at the 3-month and 6-month follow-up. The questionnaires assessed demographic and medical factors and included several standardized measures. They were returned by mail.

All men completed the International Index of Erectile Functioning (IIEF).27 With 15 multiple-choice items, the IIEF provides a total score as well as scores on five subscales (mean score ± standard deviation [SD] for volunteer sample without a history of sexual dysfunction): erectile function (25.8 ± 7.6), orgasmic function (8.8 ± 2.9), sexual desire (7.0 ± 1.8), intercourse satisfaction (10.6 ± 3.9), and overall sexual satisfaction (8.6 ± 1.7). One item measures the frequency of sexual activity. The IIEF has demonstrated good internal consistency on each subscale (ranging from 0.86–0.93), good 4-week test–retest reliability (0.64–0.84), and excellent ability to discriminate between men with sexual dysfunction and controls. (Total IIEF score averages 60.8 for volunteer sample vs. 35.2 for men with known sexual dysfunction in validation studies28 and 27.8 for respondents in our prostate survivor survey.3) Scores on the IIEF are not significantly correlated with measures of social desirability.

All women completed the Female Sexual Function Index, (FSFI)29 a parallel 19-item questionnaire, yielding an overall sexual function score in addition to subscale scores for sexual desire, arousal (both subjective and physiologic), lubrication, orgasm, satisfaction, and pain. Validation studies regarding sexually dysfunctional and matched control women ages 21–70 years have demonstrated excellent discriminant validity as well as internal consistency for the subscales and summary score (0.89–0.97). Test–retest reliabilities across 2–4 weeks also are quite good for all scores (0.79–0.88). Community volunteers without sexual dysfunction obtained a mean full scale score of 30.5 (SD of 5.3) of a maximum of 36 compared with a total score of 19.2 (SD of 6.6) for women with sexual dysfunction.30

To evaluate utilization of medical treatments for ED, each partner completed a questionnaire that was slightly modified from our previous survey.8 Items ask whether the man has obtained medical help for a sexual problem since prostate cancer treatment (at baseline), during the intervention period (at posttreatment), or within the past 3 months (at 3-month and 6-month follow-up assessments). Current medical treatments for ED, including oral medications, penile injections, penile suppositories, vacuum devices, and penile prosthesis surgery are listed, and for each treatment type the questionnaire asks whether the couple has ever tried it and, if so, whether they have used it within the past 4 weeks (to correspond to the period of time assessed by the IIEF). Participants then rate the impact of each utilized treatment on their sex life with four possible responses: “made our sex life worse,” “no effect on sex life,” “improved our sex life somewhat,” and “improved our sex life greatly.”

Participants were administered the Brief Symptom Inventory (BSI),31 which provides a global score of current psychological distress (global symptom inventory [GSI]), as well as subscale scores measuring depression, anxiety, hostility, tendency to somatize, and other dimensions of emotional adjustment. The BSI is frequently used to measure changes in distress over time in studies of medical patients. Raw scores on the BSI are transformed to T-scores with a mean of 50 (SD of 10).

Each partner also completed the seven-item abbreviated form of the Dyadic Adjustment Scale (A-DAS).32 The DAS is the most widely used measure of marital satisfaction, with proven validity, internal consistency, and test–retest reliability over a 2-week period.33 The reliability of the abbreviated scale is 0.96.32 Normative data from a sample of 196 community participants included a mean value of 25.6 with an SD of 4.8.34 Subsequent analyses revealed a cut-off score of less than 21 to be indicative of marital distress.35

To minimize subject burden, three scales were only filled out at baseline and at 6-month follow-up. Urinary and bowel symptoms after prostate carcinoma treatment can influence men's comfort in engaging in sexual intercourse. Therefore, male participants were asked to complete the urinary and bowel symptom scales from the UCLA Prostate Cancer Index (UCLA PCI).36, 37 Because menopausal symptoms in the female partners also may interfere with sexual pleasure, the women completed a seven-item menopausal symptom scale derived from a longer instrument in the Breast Cancer Prevention Trial (BCPT) Symptom Checklist.38 The items form three internally consistent subscale scores (urinary incontinence, vaginal irritation, and hot flashes) as well as yield a summary score. This abbreviated scale was used to assess outcome in a study of brief intervention for menopause symptoms with breast carcinoma survivors.39 To enable assessment of the correlation between sexual function, psychologic distress, and health-related quality of life, the SF-36 Short Form Health Survey (SF-36) was administered to study participants at the same two assessment points.40 In our analyses we used participants' scores on the two summary scales of the SF-36, the Physical (PCS) and Mental (MCS) Health components.41 These scores have a mean of 50 and a SD of 10 based on the large, community-dwelling normative sample used to validate the instrument.

Data Analyses

Descriptive analyses were conducted to characterize the study participants and their partners with respect to background information; rates of help-seeking; sexual frequency, function, and satisfaction; adherence in using ED treatments; psychologic distress; marital satisfaction; and health-related covariates as collected over the course of the study. Demographic differences between those participants who completed all four sessions of the intervention versus those who withdrew were analyzed using the Mann–Whitney–Wilcoxon test, Fisher exact test, or chi-square test, as appropriate.

To determine intervention effects across time (i.e., baseline, postcounseling, 3-month follow-up, and 6-month follow-up), a series of general linear mixed models with repeated measurements of time were conducted (SAS version 8.02; SAS Institute, Inc., Cary, NC). In these analyses, each outcome measure was regressed onto the treatment group, time of assessment, and the interaction of treatment group × time. No statistically significant interactions were observed between time and treatment group. Akaike Information Criterion (AIC) was selected to detect the best model fit by using different covariance structures (compound symmetric, unstructured, or autoregressive order one). The mixed-model F-tests were based on the Kenward–Roger adjusted degrees of freedom solution, a correction specifically designed for small sample settings. The Tukey–Kramer test was used to adjust for multiple comparisons in post-hoc significance testing of differences in outcome measures across time.



Eighty-four men and their wives/partners were initially recruited. All couples minimized to treatment attended counseling session 1, 76 (90%) finished the second session, 56 (67%) completed 3 sessions, and 51 (61%) attended all 4 counseling sessions.

Factors Associated with Withdrawing from Treatment

Reasons for withdrawal from the study among the 33 drop-out couples included the following: inability to focus on the specific sexual intervention because of high marital distress (n = 7 couples; 21%), scheduling conflicts due to work hours or extended business travel (n = 2 couples; 6%), discomfort with the explicit sexual topics of the program (n = 3 couples; 9%), serious illness (n = 2 couples, 6%), and unknown reasons (n = 19 couples; 58%, e.g., multiple ‘no shows,’ unreachable by phone or letter). Table 2 compares demographic information for participants who completed the study and those who withdrew. A general linear model analysis of variance showed that survivors who withdrew had a greater age discrepancy between partners than those who completed the intervention (P = 0.040) (i.e., men were on average 3.7 yrs older than the women in couples completing treatment vs. a 6.3-yr difference in those couples dropping out). Post hoc analysis revealed that the age discrepancy was only associated with dropping out in the couple treatment group (P = 0.005). In addition, a greater proportion of African-American couples (7 of 9 couples; 78%) withdrew compared with couples of other ethnicities (26 of 49 couples; 53%) (P = 0.033). Rates of withdrawal for African-American couples were similar in both treatment groups.

Table 2. Demographics of Couples Who Completed Treatment versus Dropouts
Variable Completed (n = 51)Dropped out (n = 33)P value
  • a

    P value was determined using the Mann–Whitney–Wilcoxon test.

  • b

    P value was determined using the Fisher exact test.

  • c

    P value was determined using the Chi-square test.

AgeMales65.5 ± 7.6 yrs62.4 ± 8.5 yrs0.097a
 Females61.8 ± 8.1 yrs56.1 ± 9.0 yrs0.013a
Education< High school6%3% 
 High school8%6% 
 Some college26%38%0.463b
Yrs in relationship 29.0 ± 15.625.0 ± 15.50.269a
Prostate carcinoma treatmentSurgery56.9%%75.0% 
 Radiation therapy31.3%25.0%0.083b
Mos since treatment 30.7 ± 21.024.7 ± 19.20.140a
Treatment groupMan only26200.387c

Baseline scores on psychosocial assessments, including marital satisfaction (A-DAS), overall psychologic adjustment (GSI score on BSI), and the MCS and PCS scores from the SF-36 did not differ significantly between those who completed treatment and those who dropped out, either for men or for women.

Differences in Outcome by Treatment Group for Couples Completing Treatment

Of the 51 couples completing the intervention, 26 survivors and their partners had been assigned to the man attending alone and 25 to the couple attendance conditions. Demographic differences between the two study conditions were examined (Table 3). Men seen alone were more likely to be Catholic (P = 0.010). However, religion was not a significant correlate of dropping out of treatment either overall or within the treatment group. Religion also was not found to be significantly correlated with any of the outcome measures (e.g., IIEF, GSI). After recoding education into two categories (with or without Bachelor's degree), there were no significant differences in education noted between men seen alone and men seen in the couple condition (P = 0.645). In addition, education was not significantly associated with study outcomes.

Table 3. Demographic Characteristics by Treatment Group for Couples Completing All Sessions
Variable Man alone (n = 26)Couple (n = 25)P value
  • a

    P value was determined using the Mann–Whitney–Wilcoxon test.

  • b

    P value was determined using the Fisher exact test.

AgeMales65.7 ± 7.2 yrs65.2 ± 8.2 yrs0.896a
 Females61.2 ± 8.1 yrs62.3 ± 8.1 yrs0.524a
Education< High school8%4% 
 High school0%16% 
 Some college35%16%0.005b
Yrs in relationship 31.2 ± 14.026.7 ± 17.20.411a
Cancer treatmentSurgery61%52% 
Mos since treatment 31.2 ± 20.230.3 ± 22.10.687a

Mann–Whitney–Wilcoxon tests were used to compare scores on psychologic outcome measures between treatment groups at the same assessment points. Neither men nor women differed significantly between treatment groups at baseline or at 6-month follow-up on marital happiness (A-DAS), overall psychologic distress (BSI), men's sexual function and satisfaction (IIEF), menopause symptoms, or bowel and bladder symptoms in men. For female sexual function and satisfaction (total FSFI score), the partners in the group in which men attended sessions alone had significantly better FSFI scores than partners in the couple group, both at baseline (19.68 ± 9.97 vs. 12.37 ± 10.01, respectively; P = 0.028) and at 6-month follow-up (23.71 ± 8.34 vs. 14.80 ± 10.21; P = 0.027), with no interaction of group × time.

Treatment Outcomes for the Combined Groups

Because outcome measures for the two treatment groups were not significantly different at baseline or follow-up or, in the case of the FSFI, no differential impact of treatment was noted on scores over time, all couples who completed treatment were combined in analyzing the efficacy of the intervention. The final regression models included assessment times (baseline, posttreatment, 3-month, and 6-month follow-ups), but did not adjust for health-related factors, including health-related quality of life (MCS and PCS), menopause symptoms for women, and bowel and urinary symptoms for men (Table 4). According to the Mann–Whitney–Wilcoxon tests, none of these variables changed significantly from baseline to 6-month follow-up assessment. Adding the variables to the models also did not appear to have a significant influence on the changes in the outcome measures across time.

Table 4. Mean Scores across Time for Health-Related Covariates
 Baseline (n = 51)6-mos follow-up (n = 39)P valuea
  • UCLA Prostate Cancer Index: University of California—Los Angeles Prostate Cancer Index; BCPT Menopause Symptom Checklist: Breast Cancer Prevention Trial Menopause Symptom Checklist.

  • a

    Significance was determined according to the Wilcoxon test comparing baseline with the 6-month follow-up.

SF-36 Health-Related Quality of Life   
 Mental Health Component–Males52.7 ± 8.452.4 ± 7.80.575
 Physical Health Component–Males48.8 ± 9.346.2 ± 8.90.134
 Mental Health Component–Females51.6 ± 8.553.8 ± 8.10.194
 Physical Health Component–Females49.1 ± 9.946.8 ± 10.20.345
UCLA Prostate Cancer Index   
 Total bowel function85.0 ± 21.880.5 ± 23.00.357
 Total urinary function76.4 ± 20.975.8 ± 22.80.997
BCPT Menopause Symptom Checklist   
 Total menopause0.55 ± 0.530.42 ± 0.520.144
 Hot flash0.64 ± 0.980.40 ± 0.810.203
 Urinary symptoms0.52 ± 0.740.50 ± 0.810.503
 Vaginal dryness0.50 ± 0.670.38 ± 0.560.574

The results of the regression analyses for men and women are summarized in Tables 5 and 6. The intervention did not improve marital adjustment, most likely because mean A-DAS scores reflected good marital satisfaction at baseline. The mean score (SD) of the standardization sample was 25.6 (4.8); for our sample it was 24.5 (5.6) for women and 25.3 (4.8) for men at baseline, indicating adequate marital adjustment. Changes across time were found to be statistically significant for men's scores (SD) on emotional distress (BSI-18 summary GSI score). Distress at baseline was equivalent to a T-score of 58, decreasing to 54 at the posttreatment assessment (i.e., a change of 0.4 SDs). Improvements also were noted across time in male sexual function/satisfaction (global IIEF score) and all IIEF subscales except for sexual desire. Women's scores changed significantly over time on sexual function/satisfaction (global FSFI score) and on each of the FSFI subscales as well. For the total FSFI, women's scores were below those of a sample of women with diagnosed female sexual arousal disorder. However, after treatment, women's scores were within 1 SD of the mean score obtained by the control group.

Table 5. Mean (± SD) Outcome Scores for Males across Time
Outcome measureBaseline (n = 51)Posttreatment (n = 44)3-mos follow-up (n = 38)6-mos follow-up (n = 39)P value for model across time
  • SD: standard deviation; GSI: global symptom inventory.

  • a

    Significance of P < 0.01 compared with baseline using the Tukey–Kramer post hoc test.

  • b

    Significance of P < 0.05 compared with baseline using the Tukey–Kramer post hoc test.

  • c

    Significance of P < 0.0001 compared with baseline using the Tukey–Kramer post hoc test.

  • d

    Significance of P < 0.001 compared with baseline using the Tukey–Kramer post hoc test.

Brief Symptom Inventory: GSI0.38 ± 0.290.29 ± 0.260.25 ± 0.20a0.29 ± 0.220.004
Abbreviated Dyadic Adjustment Scale25.3 ± 4.825.3 ± 4.725.7 ± 5.024.8 ± 4.90.640
International Index of Erectile Function:     
 Sexual desire5.4 ± 2.45.7 ± 2.0b5.6 ± 2.05.5 ± 1.70.171
 Erectile function7.6 ± 8.712.8 ± 9.5a15.3 ± 11.2c11.1 ± 10.7< 0.0001
 Orgasmic function3.7 ± 3.44.8 ± 3.14.8 ± 3.54.0 ± 3.20.025
 Intercourse satisfaction4.1 ± 4.66.7 ± 4.6d6.7 ± 5.1a4.8 ± 4.8< 0.0001
 Overall satisfaction4.3 ± 2.46.5 ± 2.3c6.4 ± 2.7c6.0 ± 2.7d< 0.0001
 Total24.8 ± 18.736.3 ± 17.3c38.9 ± 21.0c31.1 ± 20.1< 0.0001
Table 6. Mean (± SD) Outcome Scores for Females across Time
Outcome measureBaseline (n = 51)Posttreatment (n = 42)3-mos follow-up (n = 38)6-mos follow-up (n = 37)P value for model across time
  • SD: standard deviation; GSI: global symptom inventory.

  • a

    Significance of P < 0.01 compared with baseline using the Tukey–Kramer post hoc test.

  • b

    Significance of P < 0.0001 compared with baseline using the Tukey–Kramer post hoc test.

  • c

    Significance of P < 0.05 compared with baseline using the Tukey–Kramer post hoc test.

  • d

    Significance of P < 0.001 compared with baseline using the Tukey–Kramer post hoc test.

Brief Symptom Inventory: GSI0.38 ± 0.320.36 ± 0.390.39 ± 0.380.36 ± 0.400.789
Abbreviated Dyadic Adjustment Scale24.5 ± 5.624.5 ± 5.025.1 ± 5.224.0 ± 5.90.715
Female Sexual Function Index:     
 Desire3.0 ± 1.23.4 ± 1.2a3.2 ± 1.23.1 ± 1.30.011
 Arousal2.7 ± 2.24.3 ± 1.7b3.4 ± 2.23.2 ± 2.3< 0.0001
 Lubrication2.4 ± 2.44.2 ± 1.9b3.3 ± 2.22.7 ± 2.4< 0.0001
 Orgasm2.6 ± 2.54.3 ± 2.0b3.4 ± 2.42.9 ± 2.60.0001
 Pain2.1 ± 2.73.4 ± 2.7c3.0 ± 2.92.4 ± 2.90.030
 Satisfaction3.1 ± 1.64.6 ± 1.3b4.2 ± 1.7d3.8 ± 1.7< 0.0001
 Total score15.9 ± 10.625.1 ± 7.6b22.2 ± 9.5c19.0 ± 10.2< 0.0001

A caveat regarding these findings warrants discussion. As illustrated in Figures 1–3 (male GSI, IIEF, and FSFI) and demonstrated in the results of Tukey–Kramer post hoc tests (Tables 5 and 6), the shape of the outcomes curves was quadratic, reflecting highly significant improvements from baseline to the end of therapy for both men and women. For men, most gains remained significant at 3-month follow-up, but returned nearly to baseline at 6-month follow-up, except for overall sexual satisfaction. For women, this pattern of decline was more rapid, so that by 3-month follow-up the only outcome measure retaining significance was overall sexual satisfaction.

Figure 1.

Changes in Brief Symptom Inventory (BSI) Summary Score (GSI) across time.

Figure 2.

Changes in International Index of Erectile Functioning (IIEF) total score across time.

Figure 3.

Changes in Female Sexual Function Index (FSFI) total score across time.

Utilization of Treatments for ED

The percentage of men utilizing at least one medical treatment for ED was also evaluated at each assessment point. At baseline, 31% of men were using a medical treatment for ED. This rate improved to 52% at posttreatment and 55% at 3-month follow-up, and declined just slightly to 49% at 6-month follow-up assessments (P = 0.003).

Process Variables

Additional analyses revealed that completing homework assignments did not significantly affect outcomes on standardized questionnaires (IIEF, P = 0.138; FSFI, P = 0.372; male GSI, P = 0.434; and female GSI, P = 0.785). However, couples who completed homework assignments from all three counseling sessions (vs. fewer than two counseling sessions) were more likely to use a medical treatment for ED across the follow-up period (P = 0.028).

Participants in the man alone and couples conditions were evenly distributed between study counselors. Using analysis of covariance (ANCOVA), no significant differences were found in efficacy between counselors, comparing baseline scores with posttreatment scores on the major outcome measures (IIEF, P = 0.809; FSFI, P = 0.769; use of treatment for ED, P = 0.546; male GSI, P = 0.919; and female GSI, P = 0.349).


Our intervention produced positive changes in sexual function and satisfaction in both the men and women who completed it, as well as a significant increase in the prevalence of men's use of a medical or surgical treatment for ED. The presence of the female partner at the counseling sessions did not appear to significantly affect outcome. As might be expected with such a brief intervention, some of the gains diminished with time, even with a relatively short-term, 6-month follow-up assessment. We attempted to prevent relapse by having both partners set goals and make a plan to achieve them during the month after treatment, reinforced by a ‘booster’ telephone contact. Nevertheless, women's sexual function and satisfaction peaked at the conclusion of the four sessions, as did men's at 3-month follow-up, although adherence to medical treatments for ED did maintain significance at 6 months. By 6-month follow-up, 49% of couples were using such treatments, compared with only 30% of men in our original survey.8 The finding that couples who completed all requested homework were more likely to continue to use ED treatments successfully is evidence that this improvement is due to the intervention. In addition, despite the lack of significant improvements in IIEF scores at 6-month follow-up, 31% of men had an EF subscale score of 22 or better compared with only 21% in our original survey sample.

A similar pattern of early gains and relapse was observed in a study of efficacy of an eight-session telephone intervention designed to help men newly treated for prostate carcinoma to manage emotional and physical side effects.19 As in our study, men were randomized to receive the intervention alone or with the participation of a close family member. Gains were seen at 4-month assessment in problem-solving, cognitive reframing, and sexual satisfaction, but by 7 months the intervention participants were no more well-adjusted than a usual care control group. The participation of the family member also was found to make no significant contribution to efficacy.

It is not uncommon for a period of severe ED to disrupt a couple's emotional intimacy and reciprocal demonstrations of affection after treatment for localized prostate carcinoma.42 We believe that the focus in our intervention on exchanging sensual body touch, communicating more openly about sex, and increasing expression of affection, as well as advice on managing postmenopausal vaginal dryness, led to the significant, across-the-board changes in women's sexual function and satisfaction seen at the end of counseling. Apparently, however, our treatment was not powerful enough to create long-term changes in behavior or cognition. As men adopted treatments that improved their erectile function, couples may have reverted to previous patterns of lovemaking focused mainly on penile–vaginal penetration. As women then became disappointed in the return of perfunctory sex, their desire may have decreased, affecting men's sexual satisfaction in turn. A vivid personal description of this pattern is presented in the book Making Love Again, a couple's account of their sexual rehabilitation after the husband's prostate cancer.43 It is only when a period of sex therapy helps the husband accept that he can be a better lover by focusing less on erections and more on the leisurely sharing of pleasure that he reconciles himself to his dependence on injection therapy. It is also important to remember that the original sex therapy treatment programs evaluated as effective were delivered over 14–15 sessions, yet long-term follow-up studies revealed significant disease recurrence rates over time, particularly for male dysfunctions such as premature ejaculation and ED.44

Subsequent to this pilot study, we designed an Internet-based version of our treatment program, utilizing graphics, animations, videotaped personal experiences, and standardized instructions for behavioral homework. When a member of a couple completes a homework assignment, he or she completes a report form and e-mails it to the counselor, with feedback delivered also via e-mail. We are conducting a randomized trial comparing the computerized format with a streamlined, three-session version of the intervention presented in the current study. Prostate carcinoma survivors have a positive view of using the Internet to find medical information, particularly information regarding sexual rehabilitation.5, 21 In contrast, many men would not agree to consult a mental health professional for a sexual problem. The lack of counseling professionals experienced both in treating sexual dysfunction and in psycho-oncology, and poor insurance coverage for mental health services (particularly for sex therapy), are further barriers to disseminating a traditional, face-to-face counseling intervention.

We also hope that the Internet-based intervention will minimize the unacceptable drop-out rate of 39% and improve the generalizability of our intervention. As Table 2 illustrates, the couples completing the intervention tended to be highly educated and white. Our difficulty in recruiting and retaining African-American couples echoes our experience in our original survey, in which the return rate was only 28% for African-American men compared with 51% for white men.45

Despite their reluctance to complete questionnaires about sexuality or to try sexual counseling, African-American prostate carcinoma survivors are more willing to seek help in the form of a medical or surgical treatment option for ED.45 They also are reported to have higher standards for their own sexual function and greater distress about ED than white or Hispanic men.45, 46 In this pilot study, we did not utilize special strategies for minority recruiting or retention, such as asking for support from opinion leaders in the community or matching ethnicity of participant and counselor.20 In the Internet study, an African-American psychologist contacts interested African-American couples and is assigned as their counselor. Creating a culturally sensitive sexual intervention is a high research priority, because African-American men have a twofold risk of prostate carcinoma and an earlier mean age at diagnosis compared with white men.45

In essence, we have techniques that can improve the success of sexual rehabilitation after prostate carcinoma, but in order to be successful they require changes in traditional male beliefs regarding the importance of erections to women's sexual satisfaction,17 as well as changes in sexual communication and behavior from both partners. In contrast to our message, the media holds out the mirage of a simple pill that can restore sexual satisfaction.6, 7 It is not surprising that men and women prefer the magic pill, but if we can create more realistic expectations, perhaps they will be willing to try interventions that focus less on penile rigidity and more on relationship flexibility.