Advertisement

Willingness to Pay for Prostate Cancer Treatment among Patients and Their Family Members at 1 Year After Diagnosis

Open ArchivePublished:June 11, 2012DOI:https://doi.org/10.1016/j.jval.2012.03.003

      Abstract

      Objectives

      To explore an alternative approach to quantifying the burden of side effects at 1 year after treatment for prostate cancer among both patients and their partners.

      Methods

      We analyzed data from 75 couples in the Family and Cancer Therapy Selection study. Paired patients and family members were independently asked about their willingness to pay (WTP) for a hypothetical new treatment that cures prostate cancer without side effects if they could reconsider their treatment decision by indicating the maximum amount they would be willing to pay given 11 separate “bids” ranging from $0 to $1500 per month. Descriptive and regression analyses were conducted for patients and family members controlling for sociodemographic characteristics and health status; Spearman correlations were also examined.

      Results

      Among 75 couples analyzed, the income-adjusted mean WTP estimates per month were $400.8 (standard error [SE] $54.3) for patients and $650.2 (SE 72.2) for family members. The WTP between patients and family members was correlated (Pearson ρ 0.30; P = 0.01). After adjusting for covariates, the adjusted mean WTP per month was $588.1 (SE 65.77) for patients and $819.4 (SE 74.33) for family members. Wanting to avoid side effects at baseline predicted higher WTP for patients (P = 0.010). Experiencing sexual side effects was predictive of higher WTP for family members (P = 0.047).

      Conclusions

      Fairly high WTP amounts for a hypothetical treatment without side effects suggests that patients and their partners are experiencing important burdens 1 year after treatment. The higher amounts partners are willing to pay and the correlation with sexual side effects suggest that they are perceptive of significant treatment burdens.

      Keywords

      Introduction

      Prostate cancer is the most commonly diagnosed cancer, the second leading cause of cancer-related deaths among men in the United States [
      U.S. Cancer Statistics Working Group
      United States Cancer Statistics: 1999–2005 Incidence and Mortality Web-Based Report.
      ], and is an important public health problem in terms of economic and personal burden [
      • Ahmedin J.
      • Thun M.J.
      • Ries L.A.G.
      • et al.
      Annual report to the nation on the status of cancer, 1975–2005, featuring trends in lung cancer, tobacco use, and tobacco control.
      ,
      • Li C.
      • Ekwueme D.U.
      Years of potential life lost caused by prostate cancer deaths in the United States—projection from 2004 through 2050.
      ,
      • Brown M.L.
      • Riley G.F.
      • Schussler N.
      • et al.
      Estimating health care costs related to cancer treatment from SEER-Medicare data.
      ,
      • Roehrborn C.G.
      • Albertsen P.
      • Stokes M.E.
      • et al.
      First-year costs of treating prostate cancer: estimates from SEER-Medicare data.
      ]. Men diagnosed with clinically localized prostate cancer have a number of treatment options available to them [
      • Wilt T.J.
      • Shamliyan T.
      • Taylor B.
      • et al.
      Comparative-Effectiveness of Therapies for Clinically Localized Prostate Cancer Comparative Effectiveness Review No. 13.
      ]. All treatment options result in adverse effects (primarily urinary, bowel, and sexual), although the severity and frequency may vary between treatments. Urinary dysfunction, especially incontinence, appears to be more common with radical prostatectomy and bowel dysfunction with external beam radiation therapy. Sexual dysfunction is common following all treatments. These side effects result in compromised quality of life, causing burden for both patients and their family members, but quantifying the harm associated with sides effects of cancer therapies perceived to be life-saving is challenging because of cognitive dissonance and adaptation [
      • Sanda M.G.
      • Dunn R.L.
      • Mcihalski J.
      • et al.
      Quality of life and satisfaction with outcome among prostate-cancer survivors.
      ].
      One way to measure the personal benefit of a medical therapy is to assess an individual's maximum willingness to pay (WTP) for the treatment. WTP studies have been widely conducted to evaluate patients' and family members' preferences for treatment for cancer and other chronic conditions [
      • Lang H.C.
      Willingness to pay for lung cancer treatment.
      ,
      • Jonas D.E.
      • Russell L.B.
      • Chou J.
      • et al.
      Willingness-to-pay to avoid the time spent and discomfort associated with screening colonoscopy.
      ,
      • Milligan M.A.
      • Bohara A.K.
      • Pagán J.A.
      Assessing willingness to pay for cancer prevention.
      ,
      • Brown D.S.
      • Johnson F.R.
      • Poulos C.
      • et al.
      Mothers' preferences and willingness to pay for vaccinating daughters against human papillomavirus.
      ,
      • Anderson G.
      • Black C.
      • Dunn E.
      • et al.
      Willingness to pay to shorten waiting time for cataract surgery.
      ]. In settings where individuals cannot directly purchase the benefit under consideration (e.g., clean air, symptom-free quality of life), assessment of WTP through surveys and interviews has been argued to be a theoretically valid alternative [
      • Malin J.L.
      Wrestling with the high price of cancer care: should we control costs by individuals' ability to pay or society's willingness to pay?.
      ,
      • Ramsey S.D.
      • Sullivan S.D.
      • Psaty B.M.
      • et al.
      Willingness to pay for antihypertensive care: evidence from a staff-model HMO.
      ].
      Few studies have estimated patients' WTP for prostate cancer treatment [
      • Byrne M.M.
      • Ashton C.M.
      • Brody B.
      • et al.
      Willingness to pay values for prostate cancer and related treatment effects.
      ,
      • Diefenbach M.A.
      • Schnoll R.A.
      • Miller S.M.
      • et al.
      Genetic testing for prostate cancer: willingness and predictors of interest.
      ,
      • Sennfält K.
      • Carlsson P.
      • Sandblom G.
      • et al.
      The estimated economic value of the welfare loss due to prostate cancer pain in a defined population.
      ]. These limited WTP studies have been performed among patients with little attention paid to the preferences of family members. Family members are potentially significantly affected by the side effects of treatment their loved ones experience through many mechanisms [
      • Schumm K.
      • Skea Z.
      • McKee L.
      • et al.
      ‘They’re doing surgery on two people': a meta-ethography of the influences on couples' treatment decision making for prostate cancer.
      ,
      • Jones R.A.
      • Steeves R.
      • Williams I.
      Family and friend interactions among African-American men deciding whether or not to have a prostate cancer screening.
      ,
      • Zeliadt S.B.
      • Penson D.F.
      • Moinpour C.M.
      • et al.
      Provider and partner interactions in the treatment decision-making process for newly diagnosed localized prostate cancer.
      ]. Thus, including partners in quantifying the burden of side effects associated with treatment may be important to fully assessing the total impact of prostate cancer therapies.
      The purpose of this study was to 1) estimate WTP as a method of quantifying the burden associated with side effects 1 year after prostate cancer treatment among patients and their partners; 2) explore variations in WTP by population characteristics and reported declines in urinary irritation, urinary incontinence, and sexual function; and 3) examine the concordance of WTP estimates of patients with estimates of their family members. We hypothesized that patients (and family members) who experience fewer side effects would report lower WTP and patients experiencing more side effects would report higher WTP amounts.

      Methods

      Study design

      We conducted a retrospective study using data from the Family And Cancer Therapy Selection study, a 3-wave self-administered survey (baseline, 6-month follow-up, and 12-month follow-up) among patients with newly diagnosed prostate cancer in multiple clinics. Recruitment procedures and patient eligibility were described elsewhere [
      • Zeliadt S.B.
      • Moinpour C.M.
      • Blough D.K.
      • et al.
      Preliminary treatment considerations among men with newly diagnosed prostate cancer.
      ]. Briefly, newly diagnosed patients were approached in urology practice sites in California, South Carolina, and Texas. Interested patients signed consent forms and received a take-home survey to return by mail. Patients identified a family member to participate in a separate baseline survey. Mailed follow-up surveys to patients and family members were administered at 6 and 12 months to ask about treatment outcomes and satisfaction with care. WTP questions were asked at 12-month follow-up only. Study materials were approved by the institutional review board at each accrual site and the coordinating center. Participants received $25 after completing the baseline survey.
      A total of 423 patients were approached for participation (Fig. 1). Of these, 240 met eligibility criteria and 198 (83%) returned the survey prior to initiating treatment. Overall, 240 patients and 193 partners completed the baseline survey. Of these, 131 patients (89.1% of 147 who completed the 12-month survey) and 84 partners (87.9% of 99 who completed the 12-month survey) responded to the WTP questions at 12-month follow-up. Seventy-five couples responded to the WTP questions at 12-month follow-up and were included in the analysis.
      Figure thumbnail gr1
      Fig. 1Consort flow diagram for analyzed sample from the Family and Cancer Therapy Selection (n = 75 couples).

      Measures

      WTP

      On the basis of published survey methods for assessing WTP [
      • Lang H.C.
      Willingness to pay for lung cancer treatment.
      ,
      • Ramsey S.D.
      • Sullivan S.D.
      • Psaty B.M.
      • et al.
      Willingness to pay for antihypertensive care: evidence from a staff-model HMO.
      ], we used a payment-scale WTP question to determine the maximum amount a patient (or family member) was willing to pay for a hypothetical new treatment that could cure prostate cancer without side effects. Prior to our study, we developed the WTP questions and conducted focus groups of prostate cancer survivors to pilot evaluate the items. Changes were made on the basis of the feedback [
      • Zeliadt S.B.
      • Moinpour C.M.
      • Blough D.K.
      • et al.
      Preliminary treatment considerations among men with newly diagnosed prostate cancer.
      ].
      The specific question for patients and their family members was as follows:Imagine you/your loved one had just been diagnosed with prostate cancer and had to make your/his treatment decision again. However, a new drug-based treatment for prostate cancer is now available that has absolutely no side effects, such as fatigue, pain, incontinence, impotence, or changes in your/your loved one's ability to do the things you/he likes to do. This new treatment has the same chance of cure as the treatment you/he received. Assume for the moment that insurance would not cover the full costs of this treatment so you and your loved one would have to pay for it. You/your loved one would need to keep taking this therapy indefinitely. You/your loved one would not need any other treatment such as surgery or radiation. For each monthly cost listed below, please indicate how willing you would be to pay that amount of money out of your own pocket with the resources you and your loved one currently have for this new cancer treatment.The WTP question contained 11 separate “bids”: $25, $50, $75, $100, $150, $200, $250, $500, $750, $1000, and $1500. Respondents were asked to respond by checking one of four categories (No, definitely not; No, probably not; Yes, probably; Yes, definitely) for each bid. The question was designed to simulate a bidding game where respondents would be less likely to accept the bid as the dollar values rose in relation to their monthly income [
      • Ramsey S.D.
      • Sullivan S.D.
      • Psaty B.M.
      • et al.
      Willingness to pay for antihypertensive care: evidence from a staff-model HMO.
      ]. Providing “definitely” and “probably” options has been previously demonstrated to reduce pressure on respondents, encourage the bidding game scenario, and yield more accurate estimates than would a dichotomous choice [
      • Blumenschein K.
      • Johannesson M.
      • Yokoyama K.K.
      • Freeman P.R.
      Hypothetical versus real willingness to pay in the health care sector: results from a field experiment.
      ]. If patient/family member answered “Yes, probably” or “Yes, definitely” for a certain amount, then we defined them as willing to pay for that amount, and we chose the maximum amount as the final value for their WTP for a new treatment without any side effect.

      Population characteristics

      Population characteristics included sociodemographic and health status assessment. Sociodemographic characteristics were collected for both patients and their family members at baseline, which included age (<60 year, 60–69 year, 70+ year), race (white and nonwhite), education (some college or less vs. college graduate or higher), annual household income (<$40,000, $40,000–$74,999, $75,000+), and whether side effects were important in the initial treatment decision making (yes/no).
      Patients were asked “how important are these potential side effects (including problems with urinary and bowel function, sexual function and intimacy, pain, tiredness or fatigue) to your decision?” If the patient answered “very important” to any of those potential side effects, then he was identified as considering avoidance of side effects in his treatment decision. Family members were asked, “Besides curing the cancer, which aspects of treatment are important to you?” If they chose any of these (urinary and bowel function, sexual function and intimacy, pain, tiredness or fatigue), they were identified as considering avoidance of side effects in the treatment decision.
      In addition, information on insurance status (Medicare, non-Medicare private insurance, military insurance, other/unknown), first time diagnosed of cancer (yes/no), and received surgery (yes/no) was collected for patients.
      Health status assessment (medical characteristics) included disease classification (moderate/high risk of progression vs. low risk) [
      • D'Amico A.V.
      • Whittington R.
      • Malkowicz S.B.
      • et al.
      Predicting prostate specific antigen outcome preoperatively in the prostate specific antigen era.
      ] and number of comorbidities (0, 1, 2+). Patient function after prostate cancer treatment was measured by the Expanded Prostate Cancer Index (EPIC) composite score (a well-validated measure of health-related quality of life among patients with prostate cancer) [
      • Wei J.
      • Dunn R.
      • Litwin M.
      • et al.
      Development and validation of the expanded prostate cancer index composite (EPIC) for comprehensive assessment of health-related quality of life in men with prostate cancer.
      ]. EPIC urinary, bowel, and sexual function scores ranged from 0 to 100, with 100 reflecting the best possible health with no symptoms. The EPIC instrument was administered at baseline prior to treatment and at 12-month follow-up. Changes in EPIC scores measured decline among prostate cancer patients accounting for initial function, which is a proxy for severity of side effects. Changes in scores from the EPIC instrument at baseline and at 12-month survey were classified as no/mild decline and moderate/severe decline on the basis of previously established grouping criteria [
      • Ososba D.
      • Rodrigues G.
      • Myles J.
      • et al.
      Interpreting the significance of changes in health-related quality-of-life scores.
      ]. Decline in scores ranging from 0.1 to 10 points in urinary irritation were considered mild, and changes above 10 points were considered moderate/severe. Declines in scores from 0.1 to 25 points were considered mild for urinary incontinence and sexual function and above 25 points were considered moderate/severe [
      • Ososba D.
      • Rodrigues G.
      • Myles J.
      • et al.
      Interpreting the significance of changes in health-related quality-of-life scores.
      ].

      Data analysis

      We used data from the baseline and 12-month follow-up surveys from 75 couples (patients and their spouse/partner) (Fig. 1). First, we described population characteristics for patients and family members. We also used chi-square tests to compare the analyzed sample with the full study sample at baseline to examine characteristics associated with the probability of responding to WTP questions (potential self-selection). Second, we presented the observed distribution of WTP among patients and family members in Figure 2 in the format of cumulative proportion and the observed mean and median WTP for patients and family members in the study sample. We then standardized WTP amounts by household income. Prior studies have demonstrated that income significantly affects people's WTP [
      • Lang H.C.
      Willingness to pay for lung cancer treatment.
      ,
      • Jonas D.E.
      • Russell L.B.
      • Chou J.
      • et al.
      Willingness-to-pay to avoid the time spent and discomfort associated with screening colonoscopy.
      ,
      • Ramsey S.D.
      • Sullivan S.D.
      • Psaty B.M.
      • et al.
      Willingness to pay for antihypertensive care: evidence from a staff-model HMO.
      ]; thus, we scaled responders with higher and lower household incomes to the median annual income observed in the sample. This was accomplished by dividing an individual's WTP by his annual income (e.g., percentage of income willing to pay) and multiplying that amount by the median household income.
      Figure thumbnail gr2
      Fig. 2Willingness to pay for prostate cancer treatment without side effects among patients and their family members at 1 year after diagnosis. WTP, willingness to pay.
      We used Spearman's correlations to examine WTP amounts between patients and their family members. Generalized linear models using a gamma distribution and log link were used to examine the association of patient covariates including age and change in urinary and sexual function scores. A gamma distribution with a log link was selected to account for skewness in the distribution of WTP amounts [
      • Manning W.G.
      • Basu A.
      • Mullahy J.
      Generalized modeling approaches to risk adjustment of skewed outcomes data.
      ]. Additional covariates included in the model were patient's age, patient's race, insurance, household income, responder's education, responder's rating of importance for avoiding side effects in the initial treatment decision, responder's count of comorbidities, changes in patient's EPIC scores, and site. The model results we present are estimated income-adjusted mean WTP amounts and 95% confidence intervals for each variable rather than the gamma model's coefficients to assist the readers in interpreting relationships between response levels in the study variables and WTP amounts. These estimated WTP amounts were calculated by using the margins function in Stata by setting all other covariates to the mean level. Correlation in the error term of WTP among patients and family members from the same household was incorporated into the analyses by clustering on family ID. We applied bootstrapping to estimate the standard errors (SEs) with 200 replications to account for the impact of a small sample size. Analyses were conducted by using Stata version 11.2 [
      StataCorp
      Statistical Software: Release 7.0.
      ]. Given a small sample of 75 couples and potential overfitting, we explored using different model specifications to examine the robustness of the results and assumptions. We also did rigorous model selection by using Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) and examined the impact of outliers.

      Sensitivity analysis

      To test whether the certainty of response affected the results, we ran the analysis by using only the highest “Yes, definitely” bid as the dependent variable (WTP). In addition, we reanalyzed the data by using a WTP value that was the midpoint between the respondents' highest accepted bid and the next highest bid. Our using respondents' “Yes, probably” bids is somewhat conservative because WTP will presumably be somewhere in the range between the highest bid that is accepted and the next highest bid.

      Results

      Population characteristics

      Subjects who responded to the WTP questions were mostly white and had an education of college graduate or higher compared with those who did not respond to the WTP questions. No differences in other characteristics were found between the sample of responders and the original study sample (data not shown). The sociodemographic characteristics and health status of 75 couples who met the inclusion criteria for the analysis are shown in Table 1. Patients were primarily white (83.6%), graduated from college (68.0%), and had an annual household income of more than $75,000 (65.3%). Similarly, 75% of family members were white and 57.3% graduated from college. The majority of patients (92.0%) and family members (77.3%) reported that side effects were an important factor in the treatment decision. Most (78.4%) patients were diagnosed with a cancer for the first time, and 73.3% of them had one or more comorbidities. Many patients experienced some urinary or sexual side effects at 1 year after treatment, 82.4% for sexual function, 93.3% for urinary incontinent function, and 94.7% for urinary irritation function.
      Table 1Description of sample characteristics of prostate cancer patients and family member pairs analyzed at 12-mo follow-up.
      Characteristicn (%)
      Patients (n = 75)Family members (n = 74)
      Age, y
       <602330.673141.33
       60–693749.333648.00
       70+1520.0079.33
      Race
       White6183.565475.00
       Nonwhite1216.441825.00
      Education
       High school or less56.67810.67
       Some college1722.672432.00
       College graduate2634.673040.00
       Graduate degree2533.331317.33
      Household income ($ per year)
       <40,000912.00912.00
       40,000–74,9991317.331317.33
       75,000+4965.334965.33
      Avoidance of side effects in treatment decision
       Yes6992.005877.33
       No68.001722.67
      Insurance
       Medicare2635.62n/an/a
       Private (non-Medicare)3446.58n/an/a
       VA/military1216.44n/an/a
       Other/unknown11.37n/an/a
      First-time diagnosis of cancer
       Yes5878.38n/an/a
       No1621.62n/an/a
      Surgery received
       Yes6080.00n/an/a
       No1520.00n/an/a
      Disease classification
       Low risk3952.00n/an/a
       Moderate/high risk3648.00n/an/a
      Number of comorbidities
       02026.67n/an/a
       13445.33n/an/a
       2+2128.00n/an/a
      Change in EPIC domain summary scores (12-mo vs. baseline)
       Urinary irritation function
        No change/mild decrease45.33n/an/a
        Moderate/severe decrease7194.67n/an/a
       Urinary incontinent function
        No change/mild decrease56.67n/an/a
        Moderate/severe decrease7093.33n/an/a
       Sexual function
        No change/mild decrease1217.65n/an/a
        Moderate/severe decrease5682.35n/an/a
      Survey site
        USC4965.334965.33
        UTHSCSA1216.001216.00
        MUSC1418.671418.67
      Note. Couples who did not return the survey or did not answer the WTP question were excluded from 12-mo analysis.
      Family members reported their own age, race, education, and marital status. Other characteristics about patient's health status at baseline (e.g., risk classification and number of comorbidities) were compared between family members who responded at baseline and those who responded to the WTP question at 12-mo follow-up.
      EPIC, Expanded Prostate Cancer Index; MUSC, Medical University of South Carolina; n/a, not applicable; USC, University of South Carolina; UTHSCSA, University of Texas Health Science Center at San Antonio; VA, veterans affairs; WTP, willingness to pay.

      Observed and income-adjusted WTP

      The WTP values varied from approximately 1% to 60% of respondents' median annual household income. Maximum WTP values showed a bimodal distribution, with 63.4% of patients accepting bids less than $250 and 18.3% accepting the $1500 bid. Among family members, 39.3% accepted bids less than $250 and 36.9% accepted the $1500 bid. From the frequency distributions of the WTP values, we mapped demand curves for this hypothetical treatment decision for patients and family members. Figure 2 displays the curve with price (WTP) on the vertical axis, against the cumulative proportion of the sample willing to pay a sum up to any given price. The demand curve for family members is higher than that for patients. Among 75 couples who were included in the analysis, the observed mean and median WTP per month were $563.7 (SD $553.2) and $250 for patients and $804.3 (SD 598.5) and $500 for family members, respectively. Income-adjusted mean WTP per month was $400.8 (SE 54.3) for patients and $650.2 (SE 72.2) for family members. The unadjusted WTP amounts for patients and family members in the same household were significantly correlated (Pearson ρ0.30; P = 0.01).

      Population characteristics and WTP

      The income-adjusted WTP per month varied by the sociodemographic characteristics and medical characteristics (Table 2). Notably, patients reported significantly lower WTP amounts ($588; SE $66) than did their family members ($819; SE $74; P = 0.015). White patients reported higher WTP amounts than did nonwhite patients. Patients who reported at baseline that side effects were important in the initial treatment decision (P = 0.01) and patients from California reported higher WTP amounts than did patients from other study locations. Patients who experienced moderate to severe declines in urinary and sexual function from baseline to a 12 month follow-up reported slightly higher WTP amounts than did patients who did not experience functional declines, but the differences were not statistically significant. Race and the importance of side effects in the initial decision were not associated with different WTP amounts among family members. Family members of patients who reported declines in sexual function did report higher mean WTP amounts ($714) than did family members of patients who did not report declines in sexual function ($429), P = 0.047. Declines in patient's urinary function were not associated with differences in family member's WTP amounts. Family members of patients from Texas reported higher WTP amounts than did family members of patients from other study locations.
      Table 2Willingness to pay for new prostate cancer treatment without side effects among patients and family members at 1 year after diagnosis.
      CharacteristicPatient WTP ($ per month)Family member WTP ($ per month)
      Income-adjusted mean95% Confidence intervalIncome-adjusted mean95% Confidence interval
      Total
      Statistically significant at the 1% level.
      400.75294.26– 507.23650.15508.69–791.62
      Age, y
       <60421.58261.43– 581.73703.57469.00–938.14
       60–69394.67285.54–503.81624.64472.32–776.96
       70+369.48185.35– 553.61554.56274.24–834.88
      Race
       White443.79320.14– 567.43
      Statistically significant at the 1% level.
      623.92463.05–784.78
       Nonwhite239.86110.73–368.99807.72495.28–1120.17
      Education
       Some college or less364.73212.87– 516.60426.08270.60–581.57
       College graduate and above421.37286.20– 556.54373.01229.45–516.57
      Avoidance of side effects in treatment decision
       Yes417.56304.87– 530.24
      Statistically significant at the 1% level.
      688.66523.19–854.13
       No197.3523.44–371.26506.08291.62–720.54
      Disease classification
       Low risk457.48306.29– 608.67702.14497.81–906.47
       Moderate/high risk330.41210.22– 450.61595.97419.33–772.62
      Comorbidities
       No other health problems475.51256.47– 694.54547.00327.73–766.26
       1 other health problem405.14295.70– 514.57628.49487.35–769.63
       2+ other health problems345.18205.80– 484.57722.13499.56–944.70
      Surgery received
       Yes409.0289.3–528.6589.1442.3–735.9
       No367.5165.0–569.9836.9478.6–1195.2
      Change in EPIC domain summary scores (12-mo vs. baseline)
       Urinary irritation
        No change/mild decrease258.410–519.23352.3767.33–637.41
        Moderate/severe decrease416.79300.54–533.03668.34519.30–817.38
       Urinary incontinent
        No change/mild decrease224.6426.12–423.16508.15135.87–880.43
        Moderate/severe decrease424.89307.22–542.56659.95511.93–807.98
       Sexual function
        No change/mild decrease399.58168.15– 631.01429.60234.46–624.74
      Statistically significant at the 5% level.
        Moderate/severe decrease426.20293.45– 558.94714.18542.33–886.03
      Survey site
       USC500.07343.61– 656.53
      Statistically significant at the 1% level.
      517.29357.21–677.37
       UTHSCSA288.07134.83–441.30
      Statistically significant at the 5% level.
      982.56519.88, 1445.24
      Statistically significant at the 5% level.
       MUSC225.29107.33– 343.26585.95324.91–846.99
      EPIC, Expanded Prostate Cancer Index; MUSC, Medical University of South Carolina; USC, University of South Carolina; UTHSCSA, University of Texas Health Science Center at San Antonio; WTP, willingness to pay.
      low asterisk Statistically significant at the 1% level.
      Statistically significant at the 5% level.
      Figure 3 displays the demand curve of both patients and family members associated with declines in sexual function. The figure displays WTP amount on the vertical axis against the cumulative proportion of the sample willing to pay an amount up to any given price. Separate lines are displayed for patients and family members who experienced moderate/severe declines in sexual function and responders who did not.
      Figure thumbnail gr3
      Fig. 3Willingness to pay for prostate cancer treatment without side effects among patients and their family members by level of sexual side effects at 1 year after diagnosis. WTP, willingness to pay.

      Sensitivity analysis

      Different model specifications, including log-linear models of WTP amounts instead of the gamma distribution, provided consistent results. Using only the “Yes, definitely” response to define the highest amount of WTP resulted in slightly lower WTP amounts. The observed mean WTP among patients was $342.8 per month (SD 425.7), and the income-adjusted mean WTP was $393.7 per month (SE 179.4). Similarly among family members, WTP results were lower using “Yes, definitely” amounts with an observed mean WTP of $638.4 per month (SD 592.6) and income-adjusted mean WTP of $633.0 per month (SD 77.0). Results are similar using the midpoint between the respondent's highest accepted bid and the next highest bid as the WTP.

      Discussion

      One year after treatment, prostate cancer patients and their family members indicated fairly high WTP amounts for a hypothetical treatment without side effects, with a mean WTP of $588 per month for patients and $819 per month for family members. This suggests that both patients and family members are experiencing tangible burdens associated with the treatment they originally selected and would be willing to pay considerable amounts out of pocket to have a treatment that did not cause the side-effect burdens they are experiencing.
      Only a handful of studies have previously attempted to quantify the burden of prostate cancer treatment by using WTP methods [
      • Byrne M.M.
      • Ashton C.M.
      • Brody B.
      • et al.
      Willingness to pay values for prostate cancer and related treatment effects.
      ,
      • Diefenbach M.A.
      • Schnoll R.A.
      • Miller S.M.
      • et al.
      Genetic testing for prostate cancer: willingness and predictors of interest.
      ,
      • Sennfält K.
      • Carlsson P.
      • Sandblom G.
      • et al.
      The estimated economic value of the welfare loss due to prostate cancer pain in a defined population.
      ]. We highlight that our finding that patients and family members who indicated avoiding side effects was important in the initial decision reported higher WTP amounts at 12 months suggests face validity of the WTP approach. In addition, although the differences were not statistically significantly different, all patients who reported moderate or severe declines in urinary and sexual function (and their partners) reported higher WTP amounts than did patients who did not experience declines in function.
      The observation that family members were generally willing to pay higher amounts than their husbands is quite interesting. Prior studies have reported that patients' health states in prostate cancer are associated with lower quality of life among partners [
      • Basu A.
      • Dale W.
      • Elstein A.
      • et al.
      A time tradeoff method for eliciting partner's quality of life due to patient's health states in prostate cancer.
      ]. Although husbands' and wives' WTP amounts were strongly correlated, our finding that wives report higher WTP amounts may suggest that family members are quite sensitive to the burden their husbands are experiencing. Patients and their spouses may have differing perceptions, especially the ones related to the impact of sexual functioning on survivorship [
      • Rivers B.M.
      • August E.M.
      • Gwede C.K.
      • et al.
      Psychosocial issues related to sexual functioning among African-American prostate cancer survivors and their spouses.
      ,
      • Ezer H.
      • Chachamovich J.L.
      • Chachamovich E.
      Do men and their wives see it the same way? Congruence within couples during the first year of prostate cancer.
      ]. Thus, wives may be placing more value on the burdens their husbands are experiencing, especially the losses in sexual function that they may be aware of more than other burdens.
      To our knowledge, this is the first study to focus on the burden from side effects associated with prostate cancer treatment for both patients and family members in their WTP for treatment. WTP studies have been widely used to evaluate perceived benefit from treatment/intervention (e.g., cancer screening, vaccination, and hypothetical new drug) [
      • Lang H.C.
      Willingness to pay for lung cancer treatment.
      ,
      • Martin M.Y.
      • Pisu M.
      • Oster R.A.
      • et al.
      Racial variation in willingness to trade financial resources for life-prolonging cancer treatment.
      ,
      • Frew E.
      • Wolstenholme J.L.
      • Whynes D.K.
      Willingness-to-pay for colorectal cancer screening.
      ]. Most have focused on patient's WTP and did not include others potentially impacted by treatment such as a spouse.
      There were several limitations with our study. First, the results estimated in this study may not be representative of those in the general population with localized prostate cancer because of a small sample size and because patients and family members were convenient samples recruited in three study clinics. There may be nonresponse bias because most of the respondent patients (78.4%) were diagnosed as having cancer for the first time in their life. Second, eligibility for the analyzed sample was based on the completion of the baseline survey and the WTP question in the 12-month follow-up by both the patient and his family member. Subjects who responded to the WTP questions were more likely to be white with college graduate or higher education level compared with the overall population identified for the baseline survey. The WTP item was pretested in a small sample of prostate cancer survivors participating in a prostate cancer support group and may not have been easily understood by all subjects. Third, WTP is sensitive to income, and income was measured only in broad categories in this study. Fourth, there are several known limitations to WTP assessment derived under hypothetical circumstances. For example, no information is available on current out-of-pocket health care or other monthly costs, and so the concept of monthly out-of-pocket costs may not have been well understood by study participants. There may be starting-point bias (anchoring bias) and range bias from the WTP question design that could not be accounted for using self-administered survey among a small sample [
      • Frew E.J.
      • Wolstenholme J.L.
      • Whynes D.K.
      Comparing willingness-to-pay: bidding game format versus open-ended and payment scale formats.
      ]. Zero was included in the list of bids; thus, starting-point bias is likely to be conservative. It is possible, however, that if we had included a higher or lower range than $0 to $1500, we may have obtained different results. We note that the self-administered WTP assessment is an established method with practical appeal because it incorporates indirect and intangible value of treatment and can be used directly for program evaluation and has been used successfully in other contexts [
      • Ramsey S.D.
      • Sullivan S.D.
      • Psaty B.M.
      • et al.
      Willingness to pay for antihypertensive care: evidence from a staff-model HMO.
      ,
      • Johannesson M.
      Theory and Methods of Economic Evaluation of Health Care.
      ].

      Conclusions

      Local-stage prostate cancer patients and their family members indicated a fairly high WTP for a hypothetical treatment that cures disease without side effects, suggesting a high perceived burden associated with the presence of side effects at 1 year after treatment. Family members are generally willing to pay higher amounts than patients, further highlighting the burden of this disease and its associated side effects on caregivers. Patients who experienced more side-effect burden expressed higher WTP amounts. WTP may be an alternative approach to assessing the burden of prostate cancer treatment to quality-of-life or utility measures.
      This study estimated the value of avoiding side effects from prostate cancer treatment and offers an alternative to utility measures and provides valuable information for patients, family members, providers, and policymakers in considering the burden of prostate cancer. Our findings suggest that to fully assess the harms and benefit of prostate cancer treatments for a society, it is necessary to include both the patient and affected family member perspectives.
      Source of financial support: This study was supported by Cooperative Agreement Number SIP 25-04 1-U48-DP-000050 from the Centers for Disease Control and Prevention, Prevention Research Centers Program, through the University of Washington Health Promotion Research Center. The findings and conclusions in this study are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.

      References

        • U.S. Cancer Statistics Working Group
        United States Cancer Statistics: 1999–2005 Incidence and Mortality Web-Based Report.
        U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, and National Cancer Institute, Atlanta, GA2009 ([Accessed May 5, 2009])
        • Ahmedin J.
        • Thun M.J.
        • Ries L.A.G.
        • et al.
        Annual report to the nation on the status of cancer, 1975–2005, featuring trends in lung cancer, tobacco use, and tobacco control.
        J Natl Cancer Inst. 2008; 100: 1672-1694
        • Li C.
        • Ekwueme D.U.
        Years of potential life lost caused by prostate cancer deaths in the United States—projection from 2004 through 2050.
        Cancer Epidemiol. 2010; 34: 368-372
        • Brown M.L.
        • Riley G.F.
        • Schussler N.
        • et al.
        Estimating health care costs related to cancer treatment from SEER-Medicare data.
        Med Care. 2002; 40: IV-104-17
        • Roehrborn C.G.
        • Albertsen P.
        • Stokes M.E.
        • et al.
        First-year costs of treating prostate cancer: estimates from SEER-Medicare data.
        Prostate Cancer Prostatic Dis. 2009; 12: 355-360
        • Wilt T.J.
        • Shamliyan T.
        • Taylor B.
        • et al.
        Comparative-Effectiveness of Therapies for Clinically Localized Prostate Cancer.
        Agency for Healthcare Research and Quality, Rockville, MDFebruary 2008 ([Accessed June 8, 2011])
        • Sanda M.G.
        • Dunn R.L.
        • Mcihalski J.
        • et al.
        Quality of life and satisfaction with outcome among prostate-cancer survivors.
        N Engl J Med. 2008; 358: 1250-1261
        • Lang H.C.
        Willingness to pay for lung cancer treatment.
        Value Health. 2010; 13: 743-749
        • Jonas D.E.
        • Russell L.B.
        • Chou J.
        • et al.
        Willingness-to-pay to avoid the time spent and discomfort associated with screening colonoscopy.
        Health Econ. 2010; 19: 1193-1211
        • Milligan M.A.
        • Bohara A.K.
        • Pagán J.A.
        Assessing willingness to pay for cancer prevention.
        Int J Health Care Finance Econ. 2010; 10: 301-314
        • Brown D.S.
        • Johnson F.R.
        • Poulos C.
        • et al.
        Mothers' preferences and willingness to pay for vaccinating daughters against human papillomavirus.
        Vaccine. 2010; 28: 1702-1708
        • Anderson G.
        • Black C.
        • Dunn E.
        • et al.
        Willingness to pay to shorten waiting time for cataract surgery.
        Health Aff (Millwood). 1997; 16: 181-190
        • Malin J.L.
        Wrestling with the high price of cancer care: should we control costs by individuals' ability to pay or society's willingness to pay?.
        J Clin Oncol. 2010; 28: 3212-3214
        • Ramsey S.D.
        • Sullivan S.D.
        • Psaty B.M.
        • et al.
        Willingness to pay for antihypertensive care: evidence from a staff-model HMO.
        Soc Sci Med. 1997; 44: 1911-1917
        • Byrne M.M.
        • Ashton C.M.
        • Brody B.
        • et al.
        Willingness to pay values for prostate cancer and related treatment effects.
        Abstr Book Assoc Health Serv Res Meet. 1998; 15: 174-175
        • Diefenbach M.A.
        • Schnoll R.A.
        • Miller S.M.
        • et al.
        Genetic testing for prostate cancer: willingness and predictors of interest.
        Cancer Pract. 2000; 8: 82-86
        • Sennfält K.
        • Carlsson P.
        • Sandblom G.
        • et al.
        The estimated economic value of the welfare loss due to prostate cancer pain in a defined population.
        Acta Oncol. 2004; 43: 290-296
        • Schumm K.
        • Skea Z.
        • McKee L.
        • et al.
        ‘They’re doing surgery on two people': a meta-ethography of the influences on couples' treatment decision making for prostate cancer.
        Health Expect. 2010; 13: 335-349
        • Jones R.A.
        • Steeves R.
        • Williams I.
        Family and friend interactions among African-American men deciding whether or not to have a prostate cancer screening.
        Urol Nurs. 2010; 30 (166): 189-193
        • Zeliadt S.B.
        • Penson D.F.
        • Moinpour C.M.
        • et al.
        Provider and partner interactions in the treatment decision-making process for newly diagnosed localized prostate cancer.
        BJU Int. 2011; 108: 851-856
        • Zeliadt S.B.
        • Moinpour C.M.
        • Blough D.K.
        • et al.
        Preliminary treatment considerations among men with newly diagnosed prostate cancer.
        Am J Manag Care. 2010; 16: e121-e130
        • Blumenschein K.
        • Johannesson M.
        • Yokoyama K.K.
        • Freeman P.R.
        Hypothetical versus real willingness to pay in the health care sector: results from a field experiment.
        J Health Econ. 2001; 20: 441-457
        • D'Amico A.V.
        • Whittington R.
        • Malkowicz S.B.
        • et al.
        Predicting prostate specific antigen outcome preoperatively in the prostate specific antigen era.
        J Urol. 2001; 166: 2185-2188
        • Wei J.
        • Dunn R.
        • Litwin M.
        • et al.
        Development and validation of the expanded prostate cancer index composite (EPIC) for comprehensive assessment of health-related quality of life in men with prostate cancer.
        Urology. 2000; 56: 899-905
        • Ososba D.
        • Rodrigues G.
        • Myles J.
        • et al.
        Interpreting the significance of changes in health-related quality-of-life scores.
        J Clin Oncol. 1998; 16: 139-144
        • Manning W.G.
        • Basu A.
        • Mullahy J.
        Generalized modeling approaches to risk adjustment of skewed outcomes data.
        J Health Econ. 2005; 24: 465-488
        • StataCorp
        Statistical Software: Release 7.0.
        Stata, College Station, TX2001
        • Basu A.
        • Dale W.
        • Elstein A.
        • et al.
        A time tradeoff method for eliciting partner's quality of life due to patient's health states in prostate cancer.
        Med Decis Making. 2010; 30: 355-365
        • Rivers B.M.
        • August E.M.
        • Gwede C.K.
        • et al.
        Psychosocial issues related to sexual functioning among African-American prostate cancer survivors and their spouses.
        Psychooncology. 2011; 20: 106-110
        • Ezer H.
        • Chachamovich J.L.
        • Chachamovich E.
        Do men and their wives see it the same way?.
        Psychooncology. 2011; 20: 155-164
        • Martin M.Y.
        • Pisu M.
        • Oster R.A.
        • et al.
        Racial variation in willingness to trade financial resources for life-prolonging cancer treatment.
        Cancer. 2011; 117: 3476-3484
        • Frew E.
        • Wolstenholme J.L.
        • Whynes D.K.
        Willingness-to-pay for colorectal cancer screening.
        Eur J Cancer. 2001; 37: 1746-1751
        • Frew E.J.
        • Wolstenholme J.L.
        • Whynes D.K.
        Comparing willingness-to-pay: bidding game format versus open-ended and payment scale formats.
        Health Policy. 2004; 68: 289-298
        • Johannesson M.
        Theory and Methods of Economic Evaluation of Health Care.
        Springer-Verlag, New York1996