Unemployment and prostate cancer mortality in the OECD, 1990–2009

Bookmark and Share
Published: 20 Oct 2015
Views: 919
Rating:
Save
Johnathan Watkins - Institute for Mathematical and Molecular Biomedicine, King’s College London, London, UK

Dr Watkins discusses the first study to systematically explore the consequences that changes in unemployment have had in generating excess deaths due to prostate cancer.

He talks about the wide variety of variables involved and how doctors can use the knowledge generated by the study to change their clinical practice, as well as the implications for policy. 

Read the paper here.

 

Unemployment and prostate cancer mortality in the OECD, 1990–2009

Johnathan Watkins - Institute for Mathematical and Molecular Biomedicine, King’s College London, London, UK


Can you explain the rationale behind this study, and what your findings were?


Despite a body of evidence suggesting that macroeconomic indicators such as unemployment are associated with poorer health outcomes, there hasn’t really been in-depth investigation into whether unemployment also affects chronic diseases or mortality from chronic diseases such as cancer. So we chose to investigate the relationship between unemployment and mortality from a treatable cancer, namely prostate cancer, and we found that in OECD countries a rise in unemployment was associated significantly with increases in prostate cancer mortality rates. This held, this association held, despite accounting for various potential confounders.


There are a lot of variables involved – how did you identify and control for variables in this study?


We first of all sat down and identified different categories of variable that may either modify or mediate the relationship between unemployment and prostate cancer mortality, for example, urbanisation, other economic factors, out of pocket spending etc. We then sourced the data for these variables from various places, including the World Bank’s database, and we then incorporated these data into our statistical models and re-ran the analyses.


How does unemployment affect prostate cancer patients?


There are at least one of two ways in which unemployment could affect prostate cancer mortality rates. First, patients who live in areas where unemployment or underemployment is widespread are far less likely to be diagnosed promptly. As a result when they do eventually get diagnosed they’re more likely to be diagnosed with late stage disease with a concomitant worse prognosis. Second, such patients are also likely to adhere less well to the treatment regimes that their doctors put them on, as such we’re gaining worse survival rates are likely to happen. However, before really confirming these findings we need to look at or analyse cohort data rather than national aggregate data in order to really examine and test these hypotheses.


Can doctors use this knowledge to change clinical practice?


One of the ways in which doctors may use this knowledge is to take a more active role in the management of patients who are unemployed or who live in areas where unemployment is rife. However, really the really big gains from this work will probably come at the policy level rather than at the individual patient level.


What are the policy/political implications of this paper?


In addition to the obvious economic and political benefits of schemes or policies that prevent further job losses or back to work schemes, they may have the added benefit of reducing mortality from chronic diseases such as prostate cancer. Second, screening programmes or early detection programmes could be rolled out in areas, or more vigorously in areas, where unemployment or underemployment is widespread although of course the actual methods of early detection are obviously up for debate.


What is the take-home message?


First that macroeconomic indicators such as unemployment can have quite a drastic effect on health outcomes such as prostate cancer mortality. Second, policies that try to revive the unemployment situation may also have indirect consequences for health outcomes.