Cancer screening recommendations: an international comparison of high income countries

Cancer is the second leading cause of death worldwide [1], with 8.8 million cancer deaths in 2015 [1], and over 14 million new cancer cases diagnosed in 2012 [2]. About 30% of cancer deaths occurred in high-income countries [1]. Cancer screening programs have the potential to reduce cancer-specific and possibly all-cause mortality [3, 4]. The United States Preventive Services Task Force (USPSTF) has concluded that there is at least adequate evidence of a net benefit for screening for lung, breast, cervical, and colorectal cancers [5–8], and screening programs for cancer are widespread in other high resource countries [3, 4, 9–11]. For example, of over 31 million women eligible for breast cancer screening in European countries, 79% were invited to screen and 49% were screened [3]. Breast cancer mortality has decreased in the USA from 31.4 to 20.5 deaths per 100,000 women between 1975 and 2014, with similar trends in Europe, an effect attributable to both screening and improved treatment [12].

Cancer screening programs have largely been implemented in high-income countries with greater available resources [9, 13, 14]. However, there is considerable variation in terms of screening methods, starting age, stopping age, and screening interval between countries [3, 4, 9]. For example, the USPSTF recommends that adults aged 50 to 75 years should be screened for colorectal cancer with one of seven tests or combinations of tests (including colonoscopy), [8] while the Canadian Task Force on Preventive Health Care does not recommend colonoscopy as a screening test [15]. In Europe, even after the European Union recommendations in 2003, the implementation of the Council recommendations differs between countries [3]. We hypothesize that countries spending more on healthcare per capita will have more intensive cancer screening recommendations, screening for more cancers over a longer age range, and at a shorter screening interval (Appendix).

In the current report, we will review and compare cancer screening recommendations implemented in 21 high resource countries that in 2015 spent at least $3000 per capita on healthcare. We will review national recommendations where available, or the most relevant other guidelines for that country where national or federal recommendations such as those of the USPSTF do not exist. The cancers addressed will include breast, cervical, colorectal, prostate, skin, and lung cancer. As a point of comparison with US national guidelines from the USPSTF, we include relevant specialty society guidelines from the American College of Radiology, American College of Obstetrics and Gynecology, and others.

The goal of the study was to compare cancer screening recommendations in countries with comparable levels of healthcare spending, and to try to understand the relationship between healthcare spending and the intensity of screening recommendations. Membership in the Organization for Economic Cooperation and Development (OECD) was chosen as the initial qualification for inclusion in the study to ensure that the countries compared were similar economically. It was then decided that total health expenditure per capita was the most relevant statistic to national cancer screening recommendations. Therefore, the 35 OECD member countries were rank ordered according to total healthcare expenditure per capita using data from 2015. From this ranked list, all countries that spent over USD3000 per capita on health that year were included (n = 21) [16].

The next step was to determine which national organization’s cancer screening guidelines would be chosen as that country’s national recommendations. First, colleagues of the authors in each country of the study were consulted as local content experts. This was supplemented by searches of the internet where necessary. Once the national recommendations were determined for each country, the data were abstracted in tandem by two members of the research team, consulting with the project leader (Dr. Ebell) to resolve any discrepancies.

To create a uniform grading scheme, each screening test for each type of cancer was graded as recommended (which includes both strongly recommended and recommended), recommended selectively, recommended against, or insufficient evidence, corresponding to the categories used by the US Preventive Services Task Force. After analysis of each country’s national screening recommendations, the data were organized into tables. Finally, screening intensity was defined for breast, cervical, and colorectal cancer as the total number of lifetime screening tests recommended for an average risk person.

We have summarized national screening recommendations for the 21 countries with the highest total healthcare expenditure per capita. These countries were selected because they are likely to have adequate resources for a cancer screening program of some kind; comparison with less well-resourced settings would be unfair and would be less likely to reflect the differences in values, priorities, and the assessment of evidence that we are interested in. As physicians and researchers in the USA, we are especially interested in how our recommendations compare with other countries, and what we can learn from them.

In comparing recommendations, we are not making value judgements or endorsing a single “correct” answer regarding age at initiation, screening interval, or age at cessation of screening. These decisions likely reflect differences between countries and organizations regarding how benefits and harms are valued and balanced at a societal level, the type of evidence considered, the availability of resources and infrastructure for screening (for example, adequate trained medical personnel to perform colonoscopy), whether and how cost is considered, and different standards for assembling, evaluating, and interpreting a complex evidence base. For example, does the body making a recommendation consider only randomized controlled trials with cancer-specific mortality as the outcome, or does it also consider observational data and modeling studies? [20] Does the panel consider outcomes beyond mortality, such as disease progression, cost, stage shifts, or quality of life? Is cost explicitly considered or is the decision made solely on the balance of potential benefits and harms?

Note that in our discussion below, we consider the USPSTF to represent the US national screening recommendation, although recommendations from USA specialty and cancer societies are also presented. Also, when we use the term “more aggressive” or “more conservative” to describe screening recommendations, we refer to recommendations with a broader age range and/or a more frequent interval vs a narrower age range and/or a less frequent interval. Finally, a review of the process used by each country or organization to develop and update guidelines is beyond the scope of this article.

Guidelines for cancer screening differ between countries, with areas of commonality but also clear differences. Per capita healthcare spending among wealthy countries appears to have relatively little impact on recommendations; differences are more likely to stem from variation in how benefit and harm are evaluated, and which evidence is considered. Intellectual and financial conflicts of interest inherent to professional societies may contribute to more aggressive recommendations for breast cancer screening from these groups in the USA. A greater intensity of screening may alter the balance of benefits and harms by increasing the likelihood of direct harms of the screening test, as well as the serious harms associated with overdiagnosis [22]. In some cases, where specialty societies have adopted more rigorous methods, their recommendations have become more conservative, as when the American Urologic Association moved from a consensus to an evidence-based process [23]. Of course, one could also argue that national bodies like the USPSTF or Canadian Task Force are overly conservative since they rely largely on randomized trials and modeling for their recommendations. And, even groups using similarly rigorous methods may reach different conclusions, with USA and Canadian recommendations for colorectal cancer screening a good example.

In conclusion, we encourage the formation of independent panels of experts in each country, modeled after the USPSTF, Canadian Task Force, and the UK National Screening Committee. These panels should make independent, evidence-based recommendations regarding screening that assess the benefits, harms, and available resources in that country’s context. Consistent with IOM recommendations, the panels should include primary care physicians, patients, methodologic experts and relevant subspecialists, recommendations should be regularly updated, and panel members should be free of financial and intellectual conflict of interest to the greatest extent possible [24].