Following a high-profile article calling for a change in guidelines to make it easier for African Americans to be eligible for lung cancer screening, Prof Sasieni re-examines what is equitable when it comes to group-specific guidelines on cancer screening. He argues that it is the chances of those offered screening benefitting that we should aim to equalise, not the chances of an affected individual being offered screening.
A recent headline in Healio declared “USPSTF lung cancer screening guidelines ‘woefully inadequate’ for African American smokers”. Whilst EurekAlert! reported on the same JAMA oncology article saying “Health disparity for blacks exists within lung screening guidelines”. Even the JAMA oncology article concludes: “race-specific adjustment of pack-year criteria in lung cancer screening guidelines would result in more equitable screening for African American smokers”. Since the USPSTF (U.S. Preventive Services Task Force) are currently reviewing their guidelines of who should be eligible for lung cancer screening, this article is certainly timely. Unfortunately, I think the conclusion it draws is wrong!
It is surely true that guidelines could be altered so as to reduce health inequalities affecting African Americans but adjusting eligibility criteria so as to include the same proportion of cancer patients regardless of the racial group they are from is not necessarily equitable. Eligibility should be based on balancing benefits and harms in a cost-effective manner. In order to estimate the likely benefits it is necessary to establish the proportion of eligible individuals who will get lung cancer. One can then decide whether it is appropriate to spend more for the same benefit for a particular racial group within the context of distributive justice or affirmative action.
A case of being overly sensitive about sensitivity?
The JAMA Oncology paper focuses on the “sensitivity of the USPSTF guidelines” to lung cancer diagnosed during follow-up. They found that 32% of African American smokers who developed lung cancer would have been eligible for screening compared with 56% of White smokers who developed lung cancer. That sounds inequitable. But think what would happen if we used age instead of race. Had the study included a large population of smokers aged 30-44 then none of those who developed lung cancer would have been eligible for screening. That sounds like age-discrimination. Except that the risk of lung cancer in people aged 30-44 is less than a tenth of what it is in older people (aged 65-69). In order to include the same proportion of young people who develop lung cancer over the next decade one would need to screen at least ten times as many to find each cancer.
Consider another example. Preventive surgery (double mastectomy) is offered to women (e.g. Angelina Jolie) with a very high risk of breast cancer. Ashkenazi women are more likely to have inherited genetic mutations (in BRCA1 and BRCA2 genes) that put them at very high risk of breast cancer. Consequently, the relative reduction in breast cancer from the policy of offering surgery to mutation carriers is greater among Ashkenazi women than among women from other ethnic or racial groups. To be more equitable should we be offering preventive surgery to Hispanic and Asian women with a much lower risk of breast cancer (any woman with one or more first or second degree relative who had breast cancer, say) in order to have the same relative benefit in these ethnic groups? That would be daft. It is necessary to consider the benefits and harms from the perspective of each woman offered surgery. The harms are largely the same regardless of race. But the benefits depend critically on the underlying risk of breast cancer. Surgery might be a good option for someone with an 85% chance of getting breast cancer. It is unlikely to be a good option if the risk is just 15%.
What is the risk of lung cancer in those eligible for screening?
Most models for identifying individuals for lung cancer screening are based on the risk of lung cancer over the next 5 or 6 years. There is a lack of consensus for selecting ever-smokers at highest risk in these models due to different risk-factor effects between risk-models. One recent paper comparing 9 risk models argued that a screening eligibility risk-threshold of 2.0% lung-cancer risk over 5 years is a sensible compromise.
If we look instead at the proportions of eligible people who developed cancer in the JAMA Oncology paper the figure show that 4.5% of African Americans and 5.4% of Whites were eligible. And among those who are not eligible, the proportions developing cancer were extremely similar 2.0% of African American smokers and 1.9% of White smokers. On that basis the USPSTF guidelines are not inequitable towards African Americans.
The paper recommends relaxing the guidelines so that African Americans would be eligible with just 20 pack-years of smoking (equivalent to 20 years of smoking one pack a day, or 40 years of smoking half a pack a day) instead of 30 pack-years as required for Whites. Among African Americans with between 20 and 30 pack-years, 3.5% developed lung cancer; and among those with fewer than 20 pack-years 1.8% developed cancer. These figures offer limited support for reconsidering the pack-years required to be eligible for screening. Similar values for White Americans are not provided, so it is not possible to say whether it should also be reconsidered for Whites too.
The bottom line? Decisions regarding who should be eligible for screening should be based on the chances of benefitting from screening and that depends on the incidence of cancer, the stage distribution of cancer, and the benefits of early diagnosis. The relative risk reduction in a population should not be the basis for determining equity.
The views expressed are those of the author. Posting of the blog does not signify that the Cancer Prevention Group endorse those views or opinions.
Share this page