Could we measure the efficiency of prostate cancer treatment? A critical appraisal of economic evaluation studies

Estimating preferences for treatments in patients with localized prostate cancer

PURPOSE

Studies of patients' preferences for localized prostate cancer treatments have assessed radical prostatectomy and external radiation therapy, but none of them has evaluated brachytherapy. The aim of our study was to assess the preferences and willingness to pay of patients with localized prostate cancer who had been treated with radical prostatectomy, external radiation therapy, or brachytherapy, and their related urinary, sexual, and bowel side effects.

METHODS AND MATERIALS

This was an observational, prospective cohort study with follow-up until 5 years after treatment. A total of 704 patients with low or intermediate risk localized prostate cancer were consecutively recruited from 2003 to 2005. The estimation of preferences was conducted using time trade-off, standard gamble, and willingness-to-pay methods. Side effects were measured with the Expanded Prostate Index Composite (EPIC), a prostate cancer-specific questionnaire. Tobit models were constructed to assess the impact of treatment and side effects on patients' preferences. Propensity score was applied to adjust for treatment selection bias.

RESULTS

Of the 580 patients reporting preferences, 165 were treated with radical prostatectomy, 152 with external radiation therapy, and 263 with brachytherapy. Both time trade-off and standard gamble results indicated that the preferences of patients treated with brachytherapy were 0.06 utilities higher than those treated with radical prostatectomy (P=.01). Similarly, willingness-to-pay responses showed a difference of €57/month (P=.004) between these 2 treatments. Severe urinary incontinence presented an independent impact on the preferences elicited (P<.05), whereas no significant differences were found by bowel and sexual side effects.

CONCLUSIONS

Our findings indicate that urinary incontinence is the side effect with the highest impact on preferences and that brachytherapy and external radiation therapy are more valued than radical prostatectomy. These time trade-off and standard gamble preference assessments as well as the willingness-to-pay estimation could be useful to perform respectively cost-utility or cost-benefit analyses, which can guide health policy decisions.

[Cost comparison of three treatments for localized prostate cancer in Spain: radical prostatectomy, prostate brachytherapy and external 3D conformal radiotherapy]

OBJECTIVE

To compare the initial costs of the three most established treatments for clinically localized prostate cancer according to risk, age and comorbidity groups, from the healthcare provider's perspective.

METHODS

We carried out a cost comparison study in a sample of patients consecutively recruited between 2003 and 2005 from a functional unit for prostate cancer treatment in Catalonia (Spain). The use of services up to 6 months after the treatment start date was obtained from hospital databases and direct costs were estimated by micro-cost calculation. Information on the clinical characteristics of patients and treatments was collected prospectively. Costs were compared by using nonparametric tests comparing medians (Kruskall-Wallis) and a semi-logarithmic multiple regression model.

RESULTS

Among the 398 patients included, the cost difference among treatments was statistically significant: medians were € 3,229.10, € 5,369.00 and € 6,265.60, respectively, for the groups of patients treated with external 3D conformal radiotherapy, brachytherapy and radical retropublic prostatectomy, (p<0.001). In the multivariate analysis (adjusted R(2)=0.8), the average costs of brachytherapy and external radiotherapy were significantly lower than that of prostatectomy (coefficient -0.212 and -0.729, respectively).

CONCLUSIONS

Radical prostatectomy proved to be the most expensive treatment option. Overall, the estimated costs in our study were lower than those published elsewhere. Most of the costs were explained by the therapeutic option and neither comorbidity nor risk groups showed an effect on total costs independent of treatment.