NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Cancer (UK). Prostate Cancer: Diagnosis and Treatment. Cardiff (UK): National Collaborating Centre for Cancer (UK); 2008 Feb. (NICE Clinical Guidelines, No. 58.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

Cover of Prostate Cancer

Prostate Cancer: Diagnosis and Treatment.

Show details

Appendix 3An Economic Evaluation of Radical Prostatectomy Versus Alternative Treatment Options for Clinically Localised Prostate Cancer


The aim of this study was to assess the cost-effectiveness of a number of different treatment options for clinically localised prostate cancer.

Existing Economic Evidence

The systematic literature review identified 5 relevant studies. One of these studies (Horwitz et al. 1999) compared 3D conformal radiation therapy with conventional techniques, in a US setting, but was only available as an abstract. The most recent study, by Konski et al. 2006, was also performed in a US setting, and compared 3D conformal radiotherapy with intensity modulated radiotherapy (IMRT). The main limitation with this study was that differences in treatment effect were estimated using non-randomised studies, and few details of the literature search used to identify the non-randomised studies were provided. That is, people receiving IMRT were assumed to have a 2% lower probability of biochemical failure each year compared to people receiving 3D conformal radiotherapy, but the evidence base to support this notion is weak. The remaining two studies were both performed in the UK (Hummel et al. 2003; Calvert et al. 2003). Hummel et al. (2003) assessed the costs and effects of a number of different treatment options, including active surveillance and radical prostatectomy, from an National Health Service (NHS) cost perspective. However, a core assumption within the analysis was that the treatment options did not differ in terms of slowing the progression of the underlying prostate cancer. Differences in treatment effect were therefore only estimated in terms of expected side-effect profiles, although none of the evidence was derived from randomised trials. While the baseline estimates suggested brachytherapy was cost-effective compared to active surveillance and radical prostatectomy, the authors concluded that this finding was not robust given the significant uncertainty surrounding the relative side-effects of brachytherapy (and other treatments).

The economic evaluation by Calvert et al. (2003) compared policies of watchful waiting with radical prostatectomy in 60-year-old men with Gleason scores of 5–71. Costs were considered from a NHS perspective and survival was adjusted for changes in health-related quality-of-life in terms of the underlying prostate cancer and adverse effects of treatment such as incontinence and impotence. The results of the analysis suggested that watchful waiting was less costly and more effective than radical prostatectomy (that is, it produced more Quality-Adjusted Life-Years [QALYs]). However, it should be noted the number of QALYs gained per patient was almost equivalent suggesting that gains in survival attributable to radical prostatectomy were more than offset by increases in the incidence of post-operative complications.

The evaluation by Buron et al. (2007) compared the costs and benefits of (interstitial) brachytherapy with radical prostatectomy for men with a mean Gleason score of approximately 6. The evaluation was performed from a (French) societal perspective using data for almost 550 patients treated in French hospitals collected between 2001 and 2002. The results suggested that the mean societal costs of the two treatment options were similar (Euros 8,000–8,700) but that side-effect profiles, and hence health-related quality-of-life scores, differed. More specifically, impotence and urinary incontinence were more pronounced after radical prostatectomy, whereas urinary frequency, urgency and urination pain were more prevalent following brachytherapy. However, there were a number of significant limitations with the analysis: 1) changes in health-related quality-of-life were not measured using a utility-based instrument (meaning it is unclear which, if either treatment, was to be preferred on quality-of-life grounds); 2) patients in the study were not randomised to the treatment options and 3) the treatment options were assumed to be clinically equivalent in terms of the progression of the underlying prostate cancer.

In terms of developing the understanding of the cost-effectiveness of the treatment options for men with localised prostate cancer, there are arguably two main limitations with the existing literature. Firstly, only the evaluation by Hummel et al. (2003) attempted to assess the cost-effectiveness of more than two treatment options. Secondly, none of the studies incorporates information from the more recently published randomised control trial (RCT) that compares radical prostatectomy versus watchful waiting (Bill-Axelson et al. 2005).


The primary aim of this study was to perform an economic evaluation of watchful waiting versus radical prostatectomy using the 10 year RCT published by Bill-Axelson et al. (2005). In the absence of suitable RCT data, a secondary objective was to estimate how effective other therapies (brachytherapy, standard external beam radiotherapy, intensity modulated radiotherapy, high intensity focused ultrasound HIFU and cryotherapy) would need to be in order to be considered cost-effective compared by conducting a threshold analysis on the number of additional QALYs that were required to achieve certain willingness to pay thresholds for a gain value of one additional QALY.


The economic evaluation was based on a Markov model and performed from a NHS cost perspective. Markov models divide a patients’ possible prognosis into a series of discrete health states. Costs and benefits are assigned to each health state and transition probabilities define the movement (as a consequence of disease progression and treatment) of an individual between these health states over a particular time frame (cycle length). The costs and benefits of comparative treatments are then estimated on the basis of the length of time individuals spend in each health state.

The original and preferred model structure was to base the economic evaluation on a three-state Markov model (clinically localised disease, metastatic disease and dead), in line with Calvert et al. (2003). However, the RCT evidence published in Bill-Axelson et al. (2005) did not allow an estimate to be made of the probability of death given metastatic disease. Therefore, a Markov model with only two health states was constructed; alive and dead. The possibility of patients’ progressing from clinically localised disease to metastatic disease was contained within the health state ‘alive’ (Figure A3.1). This approach represents a mathematical means of staying true to the observed trial (Bill-Axelson et al. 2005) while at the same time allowing for disease progression in terms of developing more advanced prostate cancer. An alternative approach would have been to use the three-state Markov model as described above, using estimates of the probability of death given metastatic disease from alternative published sources. However, as the RCT was considered to represent the highest quality data source, this approach was considered to be less appropriate.

Figure A3.1. Schematic/Programming of Markov Model Showing Life-Years Gained As the Outcome Measure.

Figure A3.1

Schematic/Programming of Markov Model Showing Life-Years Gained As the Outcome Measure.

The model’s cycle length was yearly (as the progression of prostate cancer in the model cohort of patients was considered to be relatively slow), and the time horizon for the analysis was 20-years, by which time, the overwhelming majority of hypothetical patients had died. In the base case (the scenario which was considered to be the most likely given all the available evidence and necessary assumptions), hypothetical patients were assumed to have a mean age of 65 years and a modal Gleason score of 5–6, in line with Bill-Axelson et al. (2005).

Each cycle, patients allocated to receive watchful waiting or radical prostatectomy had an annual probability of 1) continuing to have localised disease/be cured 2) developing metastatic disease, 3) dying from natural causes or 4) dying from prostate cancer. All patients who developed metastatic disease were assumed to receive hormonal therapy until death. Patients who were allocated to receive radical prostatectomy were assumed to receive surgery on entry to the model. All patients were assumed to receive two prostate specific antigen (PSA) tests per year on an outpatient basis until death.

Three baseline results were generated:

  • Cost per additional life-year gained
  • Cost per QALY gained (side-effects excluded)
  • Cost per QALY gained (side-effects included)2.

Transition Probabilities and Treatment Effects

The baseline annual probability of death from prostate cancer for the watchful waiting strategy was taken from Bill-Axelson et al. (2005). Standard regression techniques were used to estimate a Weibull function3 from the published 10-year Kaplan-Meier disease-specific survival curve (Figure A3.2). To this was added the annual probability of death from other causes, taken directly from the UK Government’s Actuarial Department ( The annual probability of developing metastatic disease was also estimated from Bill-Axelson et al. (2005) by again fitting a Weibull function. However, as a consequence of using a two rather than three-state model, the probability of developing metastatic disease was assumed to be cumulative, and as such, represented at any single point in time, the proportion of patients who were in the health state ‘alive’ but living with metastatic disease.

Figure A3.2. Reported and extrapolated disease-specific survival curves and hazard functions derived from Bill-Axelson et al. (2005).

Figure A3.2

Reported and extrapolated disease-specific survival curves and hazard functions derived from Bill-Axelson et al. (2005). RP, Radical Prostatectomy; WW, Watchful Waiting

The survival curves are analogous to Kaplan-Meier survival curves. However, the hazard functions relate to the annual probability of death, which increases with increasing time. In both instances, the first 10-years relate to the observed data, whereas years 11–20 relate to the extrapolation

The effectiveness of radical prostatectomy was modelled by adjusting the baseline probabilities of death from prostate cancer and metastatic disease by the associated relative risks, as published in Bill-Axelson et al. (2005) 0.56 (95%CI 0.36–0.88) (Figure A3.1) and 0.6 (95%CI 0.42–0.86) respectively.

A number of side effects are possible as a result of treatment for prostate cancer. Indeed, the choice of treatment is often based on the anticipated side-effect profiles given the presenting patient, and is therefore an important concern.

In an ideal scenario, the disutility (reduction in health-related quality-of-life) associated with side effects would be derived from randomised studies comparing the relevant treatment options using an appropriate utility-based instrument. A next best solution would be to calculate the proportion of patients in each arm of a RCT that experienced each side effect and to estimate the overall level of disutility by linking this information to relevant published utility weights.

In the context of this modelling exercise, Bill-Axelson et al. (2005) did report a selection of side-effects for both the watchful waiting and radical prostatectomy arms. However, utilities were not measured within the trial and specific utility weights were not available for the majority of the reported outcomes (e.g. pain during intercourse).

The main quality of life conclusions from the RCT were published by Steineck et al. (over 4 rather than the full 10 years). The authors concluded that erectile dysfunction (80% versus 45%) and urinary leakage (49% versus 21%) were more common in the radical prostatectomy treatment arm whereas urinary obstruction was more common in the watchful waiting arm (44% versus 28%). Levels of bowel function, anxiety, depression and well being were all reported as being similar across the trial arms. Therefore the following and only assumptions were included in the model with respect to reductions in health related quality-of-life as a result of side-effects: 35% more people receiving radical prostatectomy experienced erectile dysfunction and 28% more people experienced urinary leakage compared to watchful waiting. It was also assumed that 16% more people in the watchful waiting arm experienced urinary obstruction compared to those receiving radical prostatectomy. In the main baseline scenario, the side effects were assumed to occur at the beginning of the model and to be permanent. Sensitivity analysis was used to test the robustness of the results to these and other assumptions.

Health-Related Quality-of-Life (HRQoL)/Utility Weights

The systematic literature review revealed that there have been a reasonable number of HRQoL studies involving men with prostate cancer. However, relatively few have reported utilities, which are required to incorporate HRQoL into economic evaluations in order to estimate Quality-Adjusted Life-Years (QALYs). Therefore, it was assumed that men aged 65 years with localised disease had levels of health equivalent to the general population. Using the UK EQ-5D dataset (Dolan P, 1997), this is equivalent to a utility4 value of 0.785. The utility value associated with metastatic disease was taken from Cowen et al. (1999) as 0.42. Cowen et al. (1999) also reported a number of utility scores with respect to treatment-related side-effects for localised prostate cancer; a mean of 0.69 for impotence (taken herein to be equivalent to sexual dysfunction) and 0.57 for incontinence (taken herein to represent both urinary obstruction and leakage)6.

Further simplifying assumptions were required to operationalise the model with respect to incorporating reductions in health-related quality-of-life as a consequence of side effects. Specifically, a disutility weight was calculated for the three possible side effects by subtracting the side-effect specific utility from the utility value for localised disease:

Disutility for impotence = 0.78 – 0.69 = 0.09

Disutility for urinary obstruction/leakage = 0.78 – 0.57 = 0.21

The disutility weights were also assumed to be additive, meaning for example, that a man with localised disease, with impotence and urinary obstruction experienced a utility of 0.48 (0.78 – 0.09 – 0.21). Whereas, for a man with metastatic disease with impotence but no urinary obstruction, the utility value was 0.33 (0.42 – 0.09).


Costs were only considered from a NHS’s perspective. The costs of treatment and PSA testing were taken from published sources, mostly Hummel et al. (2003), Calvert et al. (2003) and the NHS Cost Index (Table A3.1). The costs of complications associated with treatments for localised prostate cancer have not been well documented, therefore the following assumptions were made. For urinary obstruction, all men were assumed to receive a transurethral resection of the prostate (TURP). An annual cost of treating incontinence was also included, although it is noted that the study from which this value was taken relates to men with severe urinary storage problems and was not prostate-cancer specific; no published costs for urinary problems in men with prostate cancer could be identified.

Table A3.1. Unit cost estimates.

Table A3.1

Unit cost estimates.

Where necessary, costs were inflated to 2006 prices using the Hospital and Community Health Services (HCHS) Pay and Prices Index.


In the base case analysis, costs and health outcomes were both discounted at 3.5% per annum in line with NICE recommendations (NICE 2004).

Sensitivity Analysis

A number of one-way sensitivity analyses (where one input variable is changed, the model re-run and a revised incremental cost effectiveness ratio (ICER) calculated) were undertaken to highlight the variables that were the most important in terms of determining the cost-effectiveness of treatment.

Threshold analysis was also undertaken to determine how effective, in terms of additional QALYs, other therapies (brachytherapy, standard external beam radiotherapy, intensity modulated radiotherapy, HIFU and cryotherapy) would need to be, to be considered cost-effective compared to watchful waiting. Threshold analysis is undertaken by fixing the threshold willingness to pay for an extra unit of health outcome, and determining the size of health benefit survival required to produce an ICER equal to this willingness to pay value7. NICE does not have an absolute level indicating cost-effectiveness. However, NICE’s method document suggests that technologies with ICERs above £30,000 per additional QALY are unlikely to be considered cost-effective in the absence of ‘robust’ evidence (NICE 2007). Therefore, £30,000 per additional QALY was taken to represent the threshold willingness to pay.


The baseline results are shown in Table A3.2. The results show that radical prostatectomy costs approximately £4400 more than watchful waiting, but that radical prostatectomy produces an average discounted increase in life expectancy of 0.5 years. This is equivalent to an ICER of approximately £9000 per life-year gained. When no post-operative complications were assumed, radical prostatectomy was also associated with approximately 0.5 extra QALYs, with an associated ICER of £7918. However, when treatment related side effects were assumed to occur, as described in the methods section, radical prostatectomy was ‘dominated’ by watchful waiting (the main baseline result). That is, radical prostatectomy was more costly and less effective than watchful waiting.

Table A3.2. Baseline incremental cost-effectiveness ratios.

Table A3.2

Baseline incremental cost-effectiveness ratios.

The figure in bold represents the main baseline result. In this instance, RP is more costly and less effective than WW, thus it is ‘dominated’.

Sensitivity Analysis

Sensitivity analysis was performed with respect to the scenario that assumed the possibility of side effects (i.e. the main baseline result). Analysis showed that the baseline ICER was not sensitive to changes regarding, the costs of watchful waiting or the costs of metastatic disease. However, the ICER was found to be extremely sensitive to differing assumptions regarding the possible side effects associated with radical prostatectomy and watchful waiting. For example, when the additional proportion of people undergoing watchful waiting who experienced urinary obstruction was assumed to increase to 40% (from 16%), the ICER was found to be £20,155 per QALY if radical prostatectomy was used instead of watchful waiting. Thus, radical prostatectomy under this assumption appears to be a lot more cost-effective than under the baseline assumptions. The ICER was similarly sensitive to the probability of urinary leakage.

For example, when the probability of urinary leakage following radical prostatectomy was assumed to be 9%, the ICER equalled £30,000 per additional QALY. However, because the disutility associated with impotence was relatively small (0.09) compared to the disutility associated with urinary problems (both 0.21), the baseline results were not so sensitive to the probability of people becoming impotent post-surgery.

The side effect data from the Bill-Axelson et al. (2005) are only published in detail after a mean follow-up period of 4-years. When it was assumed that all treatment related side effects resolved after 4 years, the main baseline ICER was £33,926 if radical prostatectomy was used instead of watchful waiting.

One-way sensitivity analysis also showed that the baseline ICERs were relatively sensitive to the cost of radical prostatectomy. However, only when the cost reduced to under £1000 per patient (equivalent to 18% of its original costs), was it judged to be cost-effective compared to watchful waiting at the £30,000 per QALY gained level.

The baseline model did not include the possibility of patients developing hormone-refractory prostate cancer. However, as a proxy, a threshold analysis was undertaken to demonstrate how costly treatment for hormone-refractory prostate cancer would need to be for radical prostatec-tomy to be cost-effective (at the £30,000 per QALY gained level) compared to watchful waiting. This value was found to be approximately £30,000 per year. Considering the costs quoted in a recent NICE Assessment Report for using docetaxel in combination with a steroid, a cost of £30,000 per year is highly unlikely (

The baseline ICER was shown to be sensitive to the relative risk of survival. However, only when the relative risk was reduced to approximately 0.04 (from 0.56), was radical prostatec-tomy cost-effective at the £30,000 per QALY gained level. Given the lower 95% confidence interval reported by Bill-Axelson et al. (2005) of 0.36, this scenario is considered to be unlikely.

No sub-group specific relative risk of survival was reported by Bill-Axelson et al. (2005) for people with more advanced disease (higher Gleason scores), as it was not found to be a significant predictor of disease-specific mortality. However, disease-specific mortality was shown to differ by age. One-way sensitivity analysis showed that expected costs and QALYs for the two different treatment options differed markedly when different starting ages were assumed. However, in all instances, radical prostatectomy remained the dominated option.

In the absence of suitable RCT data, an estimate was made of the relative risk of disease-related survival that would be required for men with Gleason scores above 6. This was attempted by assuming men with Gleason scores above 6 had double the baseline risk of cancer related death compared with those enrolled in the Bill-Axelson RCT (Bill-Axelson et al. 2005). To achieve a threshold willingness-to-pay per QALY gained of £30,000, a relative risk of approximately 0.4 was required. When the baseline risk was quadrupled, this relative risk increased to approximately 0.59, which is above the original baseline relative risk as reported by Bill-Axelson et al. (2005).

Threshold analysis was also conducted in order to calculate how many QALYs the various other therapies (brachytherapy, standard external beam radiotherapy, intensity modulated radiotherapy, HIFU and cryotherapy) would need to produce in order to be cost-effective8.

The original intention was to perform this analysis in relation to the expected costs and QALYs of treating men with radical prostatectomy. However, since in the main baseline result, radical prostatectomy was dominated by watchful waiting, this would have been nonsensical, as it is not considered to be an economically relevant option in the first instance. Therefore, threshold QALYs were calculated in relation to watchful waiting (using a threshold willingness-to-pay of £30,000 per additional QALY).

The results from the threshold analysis showed that relatively modest gains in QALYs are required over 20 years if any of the listed treatments are to be considered cost-effective (Table A3.3). For example, external beam radiotherapy cost an additional £2103 than watchful waiting (£8288–6185), meaning that 0.07 QALYs are required to make it cost-effective compared to watchful waiting, over a 20 year period. For IMRT, the most costly option at £14688, the equivalent value was 0.29 QALYs, or an additional 4.3 months of perfect health over 20 years.

Table A3.3. Results from the threshold analysis over a 20 year period compared to watchful waiting.

Table A3.3

Results from the threshold analysis over a 20 year period compared to watchful waiting.


The primary aim of this study was to perform an economic evaluation of watchful waiting versus radical prostatectomy using the 10 year RCT published by Bill-Axelson et al. (2005) (in men with Gleason scores of 5–6). The results suggest that the cost-effectiveness of radical prostatectomy is highly dependent on the choice of health outcomes included in the analysis. If only patient survival is considered, then radical prostatectomy is arguably cost-effective. However, when quality-of-life considerations with respect to both the underlying prostate cancer and treatment-related side effects are included, watchful waiting becomes the dominant option. These results are in line with conclusions drawn by Calvert et al. (2003) The sensitivity analysis, however, showed that the results were not robust to certain assumptions, specifically surrounding the health-related effects and treatment-related side-effects; a conclusion also drawn by Hummel et al. (2003). Importantly, the results suggest that the cost-effectiveness of radical prostatectomy (and all treatments for that matter) is more dependent on the side-effect profiles than the relative risk of disease progression. Therefore, in order to be able to draw firmer conclusions regarding the cost-effectiveness of radical prostatectomy, more needs to be known about the relative probabilities of the side-effects, their duration and impact on HRQoL (it is anticipated that the ongoing MAPS study will provide more information in these issues as will the ProtecT study

In the absence of RCT data, threshold analyses were undertaken to calculate how many additional QALYs other therapies (brachytherapy, standard external beam radiotherapy, intensity modulated radiotherapy, HIFU and cryotherapy) would need to produce in order to be cost-effective at a £30,000 per additional QALY level. Radical prostatectomy was ruled out as an option, therefore these QALY gains were calculated with respect to watchful waiting. The results suggest that relatively modest improvements are required for these treatments to be cost-effective. For example, external beam radiotherapy only needed to generate an extra 0.07 QALYs over a 20 year period compared to watchful waiting for it to be considered cost-effective. This is equivalent to approximately one extra month of perfect health. For IMRT, the most costly option, the equivalent figure was 3.4 months. Thus while the absence of RCTs prevents a robust economic evaluation of these ‘newer’ treatments, it is possible to conclude that the scope for them to cost-effectiveness is relatively large. Indeed, it is feasible that they could be cost-effective even if it is proved that their greatest impact is on improving the side effects more commonly associated with the ‘older’ treatments. In the mean time, decision-makers will need to judge how likely it is that these QALY gains will be realised.

There are a number of limitations with this economic evaluation. Firstly, the cost-effectiveness of active surveillance has not been estimated. This is partly because active surveillance has not been subject to a RCT but also because modelling its cost-effectiveness would require a much more complicated model. Assuming that PSA testing is the favoured method of monitoring for progressive disease, PSA levels would themselves need to be modelled, pre and post treatment, rather than cancer stages as has been performed herein. However, the relative effect of treatment on PSA would still be uncertain given the absence of RCT data. Therefore, even if it could be concluded that radical prostatectomy is cost-effective compared with watchful waiting, it is unclear whether it is cost-effective compared with a policy of active surveillance. Similarly, it is also unclear how cost-effective watchful waiting would be compared to active surveillance. Ultimately, however, the cost-effectiveness of active surveillance is likely to depend on a combination of the proportion of men who develop progressive disease, the ability to accurately detect progressive disease and treatment efficacy in men with progressive disease.

A second limitation was that a robust sub-group analysis was not performed for men with differing Gleason scores. This is typically performed using a sub-group specific relative risk of disease progression derived from RCTs and using a sub-group specific relative risk of death. However, this information was not available, and indeed was reported by Bill Axelson et al. (2005) not to be statistically significant at the 5% level in a pre-planned sub-group analysis. However, as an indicator to cost-effectiveness, the baseline risks of death were doubled and quadrupled for men with Gleason scores of >6, in order to ascertain how effective treatment should be in terms of preventing deaths in order to be cost-effective. The results showed that when the baseline risk of prostate-specific death was quadrupled, and a relative risk akin to the value reported by Bill-Axelson et al. (2005) was assumed, radical prostatectomy was cost-effective at the £30,000 per QALY gained level. However, it is unclear how plausible a relative risk estimate this is in the absence of RCT data in this patient group.

The major conclusion that can be drawn from this evaluation is that the cost-effectiveness of all the modelled treatment options for men with clinically localised prostate cancer is highly dependent on the side effects (and therefore reductions in HRQoL) associated with each of the treatments. Indeed, the baseline assumptions suggest that radical prostatectomy should not be an option for people with Gleason scores of <6 because of its associated post-operative complications. However, different assumptions regarding side effect profiles dramatically altered the findings. Thus, future studies that attempt to quantify these relative side-effect profiles would help to produce more accurate estimates of cost-effectiveness.


  1. Bill-Axelson A, et al. Radical prostatectomy versus watchful waiting in early prostate cancer. New England Journal of Medicine. 2005;352(19):1977–1984. [PubMed: 15888698]
  2. Buron C, Le Vue B, Cosset J-M, Pommier P, Peiffert D, Delannes M, Flam T, Guerief S, Salem N, Chauvenic L, Livartowski A. Brachytherapy versus prostatectomy in the localized prostate cancer: results of a French multicenter prospective medico-economic study. International Journal of Radiation Oncology, Biology, Physics. 2007;67(3):812–822. [PubMed: 17293235]
  3. Calvert NW, et al. Effectiveness and cost-effectiveness of prognostic markers in prostate cancer. British Journal of Cancer. 2003;88(1):31–35. [PMC free article: PMC2376796] [PubMed: 12556955]
  4. Cowen ME, et al. The value or utility of prostate cancer states. Journal of Urology. 1999;155:376.
  5. Dolan P. Modeling valuations for EuroQol health states. Medical Care. 1997;35:095–1108. [PubMed: 9366889]
  6. Horwitz EM, Hanlon AL. The cost effectiveness of 3D conformal radiation therapy compared with conventional techniques for patients with clinically localized prostate cancer. International Journal of Radiation Oncology, Biology, Physics. 1999;45(5):1219–125. [PubMed: 10613316]
  7. Hummel S, et al. Clinical and cost-effectiveness of new and emerging technologies for early localised prostate cancer: A systematic review. Health Technology Assessment. 2003;7(33) [PubMed: 14609482]
  8. Konski A, et al. Using decision analysis to determine the cost-effectiveness of intensity-modulated radiation therapy in the treatment of intermediate risk prostate cancer. International Journal of Radiation Oncology, Biology, Physics. 2006;66(2):408–415. [PubMed: 16887291]
  9. National Institute for Clinical Excellence. Guidance for manufacturers and sponsors. London: National Institute for Clinical Excellence; 2004.
  10. National Institute for Health and Clinical Excellence. The guidelines manual. London: National Institute for Health and Clinical Excellence; 2007.



Calvert et al. (2003) did include a third treatment option, a selection-based management option using DNA-ploidy as a marker of disease progression. However, as this option was considered to be experimental, it is not expanded upon in this paper.


The latter scenario was taken to represent the main baseline result.


A Weibull function is a mathematical method used to estimate the probability of an event happening over time given the observed data. In this instance, it has been used to estimate the probability of death each year.


Utility values of 0 and 1 are taken to equal death and perfect health respectively. States of health between death and perfect health are therefore taken to have utility values somewhere between these two points.


A number of utility values representing clinically localised prostate cancer were available, however, they were not adjudged to differ significantly from 0.78 and were not always UK specific.


Cowen et al. (1999) derived these values in 31 individuals using the time-trade off method.


An incremental cost-effectiveness ratio (ICER) is calculated by dividing the difference in health benefits (in this instance, additional life-years or QALYs) between the different treatment options, into the difference in costs.


The main assumption underpinning this analysis is that these treatments have been assumed to be equally effective as radical prostatectomy in terms of slowing the progression of the underlying cancer. Thus, any results are contingent on this assumption.

Copyright © 2008, National Collaborating Centre for Cancer.

No part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licenses issued by the Copyright Licensing Agency in the UK. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

Bookshelf ID: NBK49535
PubReader format: click here to try


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (6.6M)
  • Disable Glossary Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...