Validity of the Patient-Reported Outcome Measurement Information System (PROMIS) Sexual Interest and Satisfaction Measures in Men Following Radical Prostatectomy

Abstract

PURPOSE

Patient-reported outcomes after radical prostatectomy have focused on erectile function. To date, no studies have validated the Patient-Reported Outcome Measurement Information System (PROMIS) Sexual Interest and Satisfaction single item measures in patients with prostate cancer, nor have studies examined how these measures relate to erectile function. In addition, data are lacking with regard to the clinical responsiveness of these measures to treatment. We sought to validate and examine the clinical utility of these measures in men after radical prostatectomy.

PATIENTS AND METHODS

We identified men who underwent a robotic radical prostatectomy from May 2014 to January 2016 in the Michigan Urological Surgery Improvement Collaborative. A single item from the PROMIS Global Satisfaction With Sex Life subdomain and a single item from the PROMIS Interest in Sexual Activity subdomain were administered. Erectile function was also assessed. Differences between baseline and 24-month T-scores for both PROMIS interest and satisfaction were examined. Multilevel models were constructed to examine change over time.

RESULTS

A total of 1,604 patients were included in the analysis. Convergent and discriminant validity of the PROMIS measures was supported. The mean PROMIS interest T-score decreased significantly from baseline to 3 months (P = .001) and significantly increased from 3 months to 24 months in this cohort, with 24-month scores exceeding baseline scores (P < .001). The mean PROMIS satisfaction T-score declined from baseline to 3 months and increased from 3 months to 24 months (P < .002).

CONCLUSION

PROMIS Global Satisfaction With Sex Life and Interest in Sexual Activity single-item measures are fundamental measures in prostate cancer survivorship. Patients are interested in sex despite functional losses and can salvage satisfaction, thereby giving insight into attainable patient-centered survivorship goals for sexual recovery after radical prostatectomy.

INTRODUCTION

Prostate cancer is the most prevalent noncutaneous cancer among men; it is estimated that by 2030, it will be the most commonly diagnosed malignancy in North America.¹ The treatment of prostate cancer is preference sensitive. Patients have a choice of more than one therapy (eg, radical prostatectomy, radiotherapy, active surveillance), and all therapies have similar oncologic outcomes but differ in their impact on health-related quality of life (HRQOL).^2-6 Radical prostatectomy has the most deleterious effect on erectile function, an important component of HRQOL.⁷ Patient-reported outcomes after radical prostatectomy have been widely studied⁸ and are a priority in survivorship care, the goal of which is to improve men’s lives after prostate cancer treatment. Sexual dysfunction after radical prostatectomy is directly related to treatment regret, feelings of loss, and distress in relationships.^9-11 The importance of sexuality in prostate cancer care is evidenced by the fact that some patients are willing to sacrifice survival to preserve potency.¹²

Various measures have been developed to assess symptom burden after prostate cancer treatment, including the University of California, Los Angeles Prostate Cancer Index (UCLA-PCI),¹³ Expanded Prostate Index Composite-26 (EPIC-26),^14,15 and the International Index of Erectile Function (IIEF).¹⁶ All include a sexual outcomes domain. The use of these measures has led to research focused on items such as erectile hardness and erections sufficient for intercourse, with little attention to other domains of sexuality, including satisfaction and interest as well as the interplay between sexual function, satisfaction, and interest.^17,18 Sexual satisfaction, interest, and function are different, distinct concepts and warrant distinct evaluation. Moreover, sexual satisfaction and interest may guide treatments during the survivorship period. A recently developed measure, the Patient-Reported Outcomes Measurement Information System (PROMIS) Interest in Sexual Activity and Satisfaction With Sex Life measures, attend to sexuality domains beyond erectile function.^19,20

The challenge of focusing only on erectile function is the reality that it generally trends toward decline after prostate cancer treatment⁵ and that, for the majority of men, it does not return to baseline levels. Wittmann et al²¹ suggested that survivorship care should support patients in improving satisfaction with sex life as a patient-centered attainable outcome after prostate cancer treatment, which has the potential to affect self-efficacy in prostate cancer survivors. Targeting such an outcome sets realistic goals for men, given that they live with altered sexual function after radical prostatectomy. Moreover, such a goal can potentially help men come to terms with feelings of loss and regret. To date, no studies have validated the PROMIS Interest in Sexual Activity and Global Satisfaction With Sex Life single-item measures in patients with prostate cancer, nor have studies examined how these measures relate to sexual function. In addition, data to support the responsiveness of these measures to treatment are currently lacking. The objective of this study was to validate the PROMIS Interest in Sexual Activity and Global Satisfaction With Sex Life single-item measures in patients with prostate cancer who underwent radical prostatectomy from diverse practices across Michigan and examine the longitudinal relationship of sexual interest, satisfaction, function, and urinary function.

PATIENTS AND METHODS

Study Setting and Design

The Michigan Urological Surgery Improvement Collaborative (MUSIC)²² is a physician-led collaborative that was established in 2011 with the objective of improving quality and decreasing costs of prostate cancer and urologic care, with more than 90% of urologists in Michigan participating in the program. A key priority of MUSIC is improving patient reported outcomes after radical prostatectomy. This objective led to development of the MUSIC Patient-Reported Outcomes (MUSIC-PRO), a statewide Web-based system for measuring outcomes and HRQOL after radical prostatectomy.²³ Each MUSIC practice has an exemption or approval for collaborative participation from a local institutional review board.

Study Participants

Patients in the current analysis included men who underwent radical prostatectomy from May 2014 to January 2016 and were enrolled in MUSIC-PRO.

Measures

PROMIS.

Single items from the PROMIS Interest in Sexual Activity and Global Satisfaction With Sex Life measures²⁴ were administered to each MUSIC-PRO participant at baseline and 3, 6, 12, and 24 months after radical prostatectomy. Single items from the Global Satisfaction With Sex Life and Interest in Sexual Activity measures are all calibrated so that scores can be calculated using item response theory provided that at least one item is completed.¹⁹ For this study, we examined a single item from the PROMIS Global Satisfaction With Sex Life subdomain (“When you have had sexual activity, how satisfying has it been?”) and a single item from the PROMIS Interest in Sexual Activity subdomain (“How interested have you been in sexual activity?”). Each item was rated on a 5-point Likert scale ranging from not at all to very much. Scores for each item are based on a T metric (mean score, 50; standard deviation, 10), with higher scores indicating higher satisfaction and interest in satisfaction and interest subdomains, respectively.

MUSIC-PRO.

The initial MUSIC-PRO survey²³ used the Memorial Sloan Kettering Cancer Center (MSKCC) questionnaire, the IIEF-6, PROMIS questions on interest and satisfaction, including single-item measures from the PROMIS Interest in Sexual Activity and Global Satisfaction With Sex Life subdomains. The surveys were administered at baseline (before radical prostatectomy) and 3, 6, 12, and 24 months after radical prostatectomy.

MSKCC urinary and bowel domains and IIEF-6.

The MSKCC questionnaire consists of self-reported functional recovery after radical prostatectomy, which includes a urinary domain, bowel domain, and overall quality-of-life items.²⁵ Each domain is scored separately, with higher scores indicating better function. The IIEF-6 evaluates erectile function. Participants who were not sexually active were prompted to bypass the IIEF-6 and are not included in the analyses. Responses for the urinary symptom items were dichotomized to represent no dysfunction (responses of “none,” “rarely or not at all,” or “no problem”) versus dysfunction (all other responses). Similarly, responses on the IIEF-6 erectile function items were dichotomized to represent no dysfunction (responses of “almost always/always,” “very good,” or “not difficult”) versus dysfunction (all other responses).

MUSIC Registry.

The MUSIC Registry includes detailed clinical and demographic information including patient age, race, prostate-specific antigen level, comorbidities, Gleason score, pathologic T stage, body mass index, nerve sparing (no nerve sparing, unilateral nerve sparing, or bilateral nerve sparing), surgeon, and MUSIC-PRO survey responses.

D’Amico risk groups.

D’Amico risk groups were created using a combination of clinical stage, Gleason scores, and prostate-specific antigen level to categorize patients into low-, intermediate-, or high-risk groups.^26,27 These scores are used to estimate prognosis and treatment outcomes.

Statistical Analyses

Convergent and discriminant validity.

Spearman correlation coefficients were used to determine convergent and discriminant validity. Evidence of convergent validity consisted of higher correlations between PROMIS T-scores for satisfaction and measures of a similar cognitive construct and between PROMIS T-scores for interest and items of a similar cognitive construct. Correlations less than 0.3 were considered poor, correlations of 0.3 to 0.6 were adequate, and correlations of 0.6 or greater were good to very good evidence of convergent validity. Evidence of discriminant validity consisted of lower correlations between the PROMIS T-scores and measures of a dissimilar cognitive construct (eg, bowel function or general quality of life).²⁸

Known groups validity.

Known groups validity was used to determine whether PROMIS scores and IIEF-6 Sexual Function scores differed with regard to nerve sparing (none, unilateral, or bilateral) at each time point. Kruskal-Wallace nonparametric tests were used to analyze overall group differences. Significant findings were further examined using Mann-Whitney U tests to identify where these differences lie. We expected participants who received bilateral nerve sparing to have better satisfaction with sex life (but not sexual interest) than individuals who received unilateral or no nerve sparing.

Pre-post treatment effects.

We examined differences between baseline and 3-month scores and between 3-month and 24-month scores for both PROMIS Interest in Sexual Activity and PROMIS Satisfaction With Sex Life using Wilcoxon signed rank tests. Changes in scores were calculated by subtracting the earlier time point score from the later time point score (baseline from 3 months and 3 months from 24 months).

Examining change over time.

Linear mixed effects models (with random effects for intercept and time) were used to examine change over time. Separate models were conducted for each PROMIS measure; all models controlled for baseline PROMIS score and age.

Relationship between urinary symptoms and sexual function with PROMIS measures.

A series of linear mixed effects models (with random effects for intercept and time) were used to examine the relationship of each of the PROMIS measures with urinary function and/or erectile function. Separate models were run for each PROMIS measure and for each symptom group for the five urinary and four erectile function items (no dysfunction v dysfunction). This resulted in 18 models (10 models for the urinary items [five for interest and five for satisfaction] and eight models for the erectile function items [four for interest and four for satisfaction]); all models controlled for age. All analyses for the current study were conducted using SAS version 9.4 (SAS Institute, Cary, NC).

RESULTS

Sample Description

A total of 1,604 participants were examined for the current analysis (Table 1). The average age of the sample was 63.2 years. The average body mass index was 29.1 kg/m², and the majority of men (80.6%) were married. The majority of participants (84.1%) were white, and 57.8% of the participants were intermediate risk by D’Amico scores. Bilateral nerve sparing was used for the majority of patients (80.4%).

TABLE 1.

Patients Demographic and Clinical Characteristics

Convergent and Discriminant Validity

Convergent and discriminant validity of the two PROMIS measures was generally supported (Table 2). Specifically, the PROMIS measures had moderate to large correlations with other measures of sexual satisfaction and interest (supporting convergent validity). In addition, there was little to no correlation with measures of bowel function and general quality of life (supporting discriminant validity).

TABLE 2.

Baseline Spearman Correlations Between PROMIS Measures and Comparator Measures

Known Groups Validity

In general, there were no group differences among the nerve-sparing groups in PROMIS Interest in Sexual Activity (Table 3). For PROMIS Satisfaction With Sex Life, there were no differences between groups at baseline or 24 months, but there were significant differences at 3 months (χ² = 9.89; P = .007), 6 months (χ² = 14.17; P < .001), and 12 months (χ² = 6.23; P = .044), because the bilateral group had higher T-scores than the group that received no sparing. With regard to sexual function, there were significant differences between nerve-sparing groups at all time points (P < .001).

TABLE 3.

Group Comparisons by Nerve-Sparing Group

Pre-Post Treatment Effects

Figure 1 depicts the average T-scores for PROMIS interest and PROMIS satisfaction across time points. For PROMIS interest, there was a significant decrease in mean T-scores from baseline to 3 months (rank sum [S] = −28,538; P = .001). In addition, there was a significant increase in PROMIS interest mean T-scores from 3 months to 24 months (S = 6,725.5; P < .001). Interest T-scores were higher at 24 months than at baseline (S = 5,096.5; P < .001).

FIG 1.

Average Patient-Reported Outcome Measurement Information System (PROMIS) Interest in Sexual Activity and Global Satisfaction With Sex Life T-scores from baseline to 24-month follow-up.

With regard to PROMIS satisfaction, there was a significant decrease in T-scores from baseline to 3 months (S = −53,649; P < .001). In addition, there was a significant increase in PROMIS satisfaction T-scores from 3 months to 24 months (S = 773.5; P = .002); T-scores did not return to baseline levels (S = −1,232; P = .003).

Examining Change Over Time

Both PROMIS Interest in Sexual Activity and Satisfaction With Sex Life single-item measures showed significant change over time. For PROMIS interest, T-scores increased by 1.32 points (95% CI, 1.04 to 1.59 points) at each follow-up visit and were reduced by 0.06 points (95% CI, 0.02 to 0.09 points) for every 1-year increase in age. Baseline PROMIS interest T-scores increased follow-up T-scores by 0.04 points (95% CI, 0.01 to 0.08 points) for each one-unit increase in baseline interest. Similarly, PROMIS satisfaction T-scores increased by 1.15 points (95% CI, 0.74 to 1.56 points) at each follow-up visit. Satisfaction T-scores were reduced by 0.07 points (95% CI, 0.2 to 0.13 points) for each 1-year increase in age. Baseline satisfaction did not have any effect on follow-up T-scores.

Relationship Between Urinary Symptoms and Sexual Function With PROMIS Measures

In general, PROMIS interest was not related to the presence of symptoms of urinary dysfunction (Table 4). However, there were group differences in interest for erectile function items at 3, 12, and 24 months after surgery. At all time points, the erectile dysfunction group reported less interest than the group with no erectile dysfunction. Considering that some individuals who are not sexually active but have good erectile function may be automatically categorized into the dysfunction group, we evaluated responses to “Confidence to get and keep an erection.” For PROMIS interest T-scores, the no problem group had an initial decline in PROMIS interest T-scores at 3 months followed by an increase at 6 and 12 months; 24-month PROMIS interest T-scores for the no problem group were higher than baseline T-scores. The group that reported problems getting and maintaining erections had a continuous increase in PROMIS interest T-scores from baseline to 24 months.

TABLE 4.

PROMIS Interest in Sexual Activity T-Scores for Urinary and Sexual Function Measures

For PROMIS satisfaction (Table 5), participants who used pads or had urinary leakage consistently reported less satisfaction. Those with any urinary symptoms reported less satisfaction than those with no symptoms at 12 months postoperatively. For erectile function symptoms, although there were generally no group differences in satisfaction at baseline, there were group differences at all postsurgical visits. Differences were as expected; those with erectile dysfunction consistently reported less satisfaction than those with no erectile dysfunction.

TABLE 5.

PROMIS Global Sexual Satisfaction T-Scores for Urinary and Sexual Function Measures

DISCUSSION

To our knowledge, this is the first study to demonstrate construct validity and examine the longitudinal trajectory of the PROMIS Interest in Sexual Activity and Global Satisfaction With Sex Life single-item measures in men with prostate cancer. The PROMIS measures had moderate to large correlations with general measures of interest and satisfaction, which support convergent validity. Little to no correlation between PROMIS measures and measures of bowel function and general quality of life was observed, supporting discriminant validity. Although erectile function is the measure that is more often reported in the literature, this study demonstrates that erectile function is not highly correlated with sexual interest and satisfaction, which supports the need for validated patient-reported measures that comprehensively assess key domains of sexual interest and satisfaction in both research and clinical settings. Despite functional losses, interest and satisfaction are dynamic during the recovery period and not completely dependent on erectile function. As expected, there was an initial decline in PROMIS satisfaction T-scores from baseline to 3 months; however, T-scores significantly increased from 3 months to 24 months. Unexpectedly, PROMIS interest T-scores increased from baseline to 24 months. Although there were no differences among nerve-sparing groups for PROMIS interest at any time point, the bilateral nerve-sparing group had higher satisfaction at 3, 6, and 12 months than the unilateral and no nerve-sparing groups. PROMIS Interest in Sexual Activity and Satisfaction With Sex Life measures can prove meaningful for patients and may enable enhanced coping strategies as patients face the reality of erectile function loss.

It has been demonstrated that, in discussing treatment options for men with localized prostate cancer, urologists rarely discuss patients’ interest in sex.²⁹ We demonstrate the importance of the evaluation of interest and satisfaction measures over time and show that, despite preconceived notions from prior studies indicating low interest in sex after radical prostatectomy,³⁰ this cohort of patients had high interest in sex after radical prostatectomy with a significant increase in interest from 3 to 24 months after prostatectomy. This may indicate that as men are cured of prostate cancer, they shift their focus from survival to quality of life.

For PROMIS Global Satisfaction With Sex Life, compared with men with poor baseline erectile function, men with no baseline erectile difficulties had higher satisfaction at 3, 6, 12, and 24 months, which is consistent with findings from a study that showed that having good sexual function before treatment was strongly related to sexual satisfaction.³¹ Individuals who underwent bilateral nerve sparing had higher IIEF-6 erectile function scores than those who underwent no nerve sparing, which was significant at all time points. This bilateral nerve-sparing group also had significantly higher satisfaction than the unilateral and no nerve-sparing groups at 3, 6, and 12 months. Our findings suggest that satisfaction with sex life is largely recoverable because there was a significant increase in PROMIS satisfaction T-scores from 3 months to 24 months. Although generalized instruments, such as the IIEF, are more often used in clinical practice^32,33 and are advantageous in detecting treatment effects, single items from the IIEF are not validated. The PROMIS Global Satisfaction With Sex Life single-item measure is validated and broad, has no limitation on how an individual defines sex life, and captures psychological aspects of sexuality. An evaluation of satisfaction in men 4.3 years after radiotherapy or radical prostatectomy found that factors related to sexual satisfaction included cognitive and behavioral elements,³¹ which begs the question of whether we are measuring the right outcome.²¹ This notion is supported by a study comparing men with and without erectile dysfunction in which relationship and psychological characteristics were significant predictors of sexual satisfaction.³⁴ This supports the concept that satisfaction with sex life can be both an obtainable and valuable survivorship goal independent of erectile function.

A limitation of this study is that the time points did not extend beyond 24 months. However, studies show that 24 months is the time point of optimal recovery of sexual function,³⁵ with rates of sexual activity in some cases peaking at 3 to 6 years after surgery.³⁰ It is possible that beyond 24 months, we might see changes in interest and satisfaction because there may be continued improvement in erectile function and adaptation of patients and their partners. In addition, we only evaluated these measures in patients after radical prostatectomy; they could potentially be useful in patients after other treatments for prostate cancer, including radiotherapy.

These limitations notwithstanding, this study has several strengths. Unlike other tools that measure symptom burden, PROMIS Interest in Sexual Activity and Satisfaction With Sex Life single-item measures are validated and allow for interpretation of a domain of sexuality from single-item responses because the measures use item response theory.¹⁹ To our knowledge, this is the first study to show that PROMIS interest and satisfaction single-item measures carry much promise and can potentially serve as additional measures of longitudinal patient-level indicators of sexual recovery. In addition, this study included a large number of patients from both community and academic practices.

This study has implications for patients, providers, and the health care system. PROMIS interest and satisfaction measures allow patients to target realistic goals in their recovery and sexual relationships. The goals are attainable and can leverage self-management as well as the involvement of partners, a key element in sexual recovery and prostate cancer survivorship.²¹ For patients to shift focus to attainable goals, however, providers must also shift focus, which may allow patients to realize that in treating prostate cancer, they do not have to have a trade-off between sex and survival.¹² For several years, research on sexual outcomes of survivors of prostate cancer has continuously identified decline in erectile function and erectile hardness after treatment; it is time for a paradigm shift for several reasons. Up to 50% of men demonstrate some degree of erectile dysfunction preoperatively.^36,37 The average patient in this cohort was 63 years old, and age is an independent risk factor for erectile dysfunction.³⁶ Compounding this, prostate cancer treatments result in continuous decline of erectile function, which lasts beyond 10 to 15 years.^5,38 Finally, patient-reported outcomes have been purported as a quality metric with a focus on efficacy.^39,40 Unlike erectile function, which may focus solely on the individual surgeon or individual patient characteristics, PROMIS satisfaction and interest measures have the potential to offer an evaluation of the health care system and the patient experience.

In conclusion, construct validity for the PROMIS measures of Interest in Sexual Activity and Global Satisfaction With Sex Life is supported in men with prostate cancer after radical prostatectomy. The pursuit of erectile function as the sole sexual outcome, which is not attainable for all patients taking into account disease extent, baseline erectile function, and nerve sparing, has the potential to be futile and discouraging to patients and their partners alike. When combined with tools to measure symptom burden, these PROMIS measures can give survivors of prostate cancer hope that, although they may have functional losses, they can still maintain sexual relationships and recover sexuality.

AUTHOR CONTRIBUTIONS

Conception and design: Nnenaya Q. Agochukwu, Rodney L. Dunn, James E. Montie, David C. Miller, James Peabody

Administrative support: Nnenaya Q. Agochukwu, Tae Kim, David C. Miller

Provision of study material or patients: Tae Kim

Collection and assembly of data: Nnenaya Q. Agochukwu, Tae Kim

Data analysis and interpretation: Nnenaya Q. Agochukwu, Daniela Wittmann, Nicholas R. Boileau, Rodney L. Dunn, James E. Montie, David C. Miller, James Peabody, Noelle E. Carlozzi

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Validity of the Patient-Reported Outcome Measurement Information System (PROMIS) Sexual Interest and Satisfaction Measures in Men Following Radical Prostatectomy

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/jco/site/ifc.

James E. Montie

Stock and Other Ownership Interests: Histosonics

Consulting or Advisory Role: Histosonics

David C. Miller

Research Funding: Blue Cross Blue Shield of Michigan

James Peabody

Stock and Other Ownership Interests: Intuitive Surgical (I)

Noelle E. Carlozzi

Consulting or Advisory Role: Teva

No other potential conflicts of interest were reported.