Addressing Overdiagnosis of Prostate cancer: drivers, trends, and patient information.

  • Thanya I. Pathirana

Student thesis: Doctoral Thesis


The issue of overdiagnosis has drawn much attention over recent decades. Although several important drivers and possible solutions for overdiagnosis have been previously described in literature, there is a lack of a coherent map of drivers matched to potential solutions. Overdiagnosis is arguably the most significant adverse effect associated with cancer screening, as this means people are diagnosed with, and usually treated for, a cancer that would otherwise have not caused any symptoms or problems.

Of the possible cancer overdiagnoses, prostate cancer overdiagnosis is probably the most important public health concern in many developed countries, including Australia. To better explore the likelihood and the possible extent of prostate cancer overdiagnosis in Australia, a detailed analysis of prostate cancer incidence, mortality, and screening rates is required.

To facilitate shared decision making about testing, decision aids are valuable tools that may also prevent overdiagnosis. Previous studies have examined how overdiagnosis is reported in patient decision aids for prostate cancer screening but have not provided a clear picture of all information on overdiagnosis included in prostate cancer screening decision aids.

This thesis aimed to (i) identify and map the drivers of overdiagnosis to potential solutions (ii) examine evidence of overdiagnosis in prostate cancer by analysis of the trends in PSA screening, prostate cancer incidence and mortality in Australia (iii) quantify the risk of overdiagnosis associated with prostate cancer screening in Australia using a novel lifetime risk approach, and (iv) review whether, and how, overdiagnosis was described in patient decision aids for prostate cancer screening.

Methods and Results:
Four studies were conducted to investigate nine specific research questions.
The first study was a comprehensive review and systematic analysis to map drivers and potential solutions of overdiagnosis across five inter-related domains: culture, health system, industry and technology, healthcare professionals, and patients and the public. Common beliefs of “more is better”,
“early is better”, intolerance of uncertainty and error, expanding disease definitions, financial incentives, conflicts of interest in guideline panels, development, promotion, and use of increasingly sensitive tests, and medicolegal concerns regarding missing or delaying a diagnosis were among the
key drivers identified. Raising public awareness through information and education campaigns, reforming incentives for professionals and healthcare organisations to reward the quality over quantity of care, avoiding conflict of interests in guideline panels, rigorous evaluation and increased regulation of new and existing diagnostic technology for their impact on health outcomes, revising current medical education curricula to include overdiagnosis and overuse, promoting shared decision making, and consideration of conservative management options such as watchful waiting or active surveillance where appropriate were among the key solutions recommended.

The second study sought to better understand prostate cancer overdiagnosis in the Australian context, by undertaking a critical analysis of trends in Australia in prostate cancer screening, incidence and mortality, prostate biopsy, trans-urethral resections of the prostate and treatment of benign prostatic hypertrophy. Using data from Australian Institute of Health and Welfare, Medicare Benefits Schedule Statistics and Pharmaceutical Benefits Schedule Statistics, we calculated standardised time trends in prostate specific antigen (PSA) tests, prostate biopsies, treatment of benign prostatic hypertrophy and prostate cancer incidence and mortality in Australia in men aged 45-74, 75-84, and 85+ years. PSA testing increased from its introduction in 1989 to a peak in 2008. It then declined in men aged 45-84 years. Prostate biopsies and cancer incidence fell from 1995 to 2000 in parallel with decrease in trans-urethral resections of the prostate and latterly changes in pharmaceutical management of benign prostatic hypertrophy. After 2000, changes in biopsies and cancer incidence paralleled changes in PSA screening in men 45-84 years, while in men ≥85 years biopsy rates stabilised and incidence fell. Prostate cancer mortality in men aged 45-74 years remained low throughout. Mortality in men 75-84 years gradually increased until the mid 1990s, then gradually decreased. Mortality in men ≥85 years increased until the mid 1990s, then stabilised.

The third study quantified the risk of overdiagnosis associated with prostate cancer screening in Australia based on a novel lifetime risk approach using publicly available population data. The study found that the lifetime risk of being diagnosed with prostate cancer in Australian men increased from 6.1% in 1982 (1 in 17) to 19.6% in 2012 (1 in 5). Using 2012 competing mortality rates, the lifetime risk in 1982 was 11.5% (95% CI 11.0% to 12.0%). The excess lifetime risk of prostate cancer in 2012 (adjusted for changing competing mortality) was 8.2% (95% CI 7.6% to 8.7%) (1 in 13). This increase in the lifetime risk suggests 41% of all prostate cancers diagnosed in in Australia in 2012 may have been overdiagnosed.

The final study was a comprehensive review of all publicly available decision aids for prostate cancer screening published in English language. All decision aids included general information about potential benefits and harms of screening for prostate cancer. Around half presented the potential benefits of PSA screening before the harms (n=22, 54%). A minority of the decision aids presented
information on competing risks of mortality (n=16, 39%), and very few included information of the prevalence pool of undiagnosed indolent prostate cancer in the general population (n=2, 5%). Although only a minority (n=9, 21%) of the decision aids used the word “overdiagnosis”, most decision aids (n=38, 92%) included valid information about overdiagnosis. Only 7 (17%) decision
aids included information on the extent of the overdiagnosis related to prostate cancer screening, ranging from 3%-66%. In 5 (12%) decision aids, the estimates were in the range of 20-50%.

Conclusions and Implications
Collectively, these studies highlighted three important elements of overdiagnosis of prostate cancer in Australia - drivers, trends, and patient information. Among the multiple drivers identified in study 1, screening was a key driver of overdiagnosis. Along with prostate cancer screening trends, analysis of incidence and mortality trends in study 2 established evidence of prostate cancer overdiagnosis in Australia based on the large increases in prostate cancer incidence with minimal changes in its mortality suggesting a high degree of overdiagnosis. The novel lifetime risk method described in the third study estimated the extent of prostate cancer overdiagnosis in Australia to be 41% of all prostate cancers diagnosed in 2012. This novel method may be used in conjunction with other existing methods to triangulate the likely extent of cancer overdiagnosis in populations where data from randomised controlled trials are not available, and it can be easily applied using freely available statistical software and publicly available administrative datasets. Exploring one of the key solutions for addressing overdiagnosis identified in study 1 - promoting shared decision making through patient decision aids - the fourth study concluded that most patient decision aids for prostate cancer screening
lacked important information on overdiagnosis and that specific guidance is needed on how to effectively communicate the risks of overdiagnosis in decision aids, including appropriate content and terminology.

Key recommendations arising from this thesis that may help address cancer overdiagnosis in Australia more broadly are:
a) Enhance awareness and understanding of overdiagnosis by health professionals, policy makers, and the public through a variety of educational processes
b) Monitor epidemiological trends in testing, cancer incidence, mortality, and treatment rates after the introduction of new cancer screening tests into clinical practice, to identify potential overdiagnosis early
c) Monitor longer term trends in cancer overdiagnosis in populations using the novel lifetime risk method, combined with other methods to triangulate the likely extent
d) Provide clear information on harms of screening, including overdiagnosis, in cancer screening decision aids along with important contextual information to help patients understand what cancer overdiagnosis means.
Date of Award15 Jun 2022
Original languageEnglish
SupervisorPaul Glasziou (Supervisor), Catherine Bell (Supervisor) & Jennifer Doust (Supervisor)

Cite this