The Ethics of Screening in Health Care and Medicine

The book is a Springer publication by Niklas Juth and Christian Munthe. Here’s what I wrote in my goodreads review:

“I don’t think it’s a bad book, and it’s likely one of the best on the topic. I generally liked the book and I was at three stars throughout much of it. Different ethical approaches to problems in this area are described and applied to a variety of contexts.

If you want to know why I didn’t go higher [than 3 stars], here are some relevant points of criticism: 1) Some of the concluding analysis in the last chapter was so weak that I ended up writing some relatively nasty stuff in the margin of the book. I decided at one point to subtract a star because of that stuff, but I later decided that it was probably wrong to do that, and so I ended up back at three stars. However I still find their analysis and conclusions related to DTC genetic testing poorly argued. No, not poorly argued – stupid. 2) I also very much disagree with their take on prenatal screening in general, especially when it comes to the decision-relevant parameters to be included in a discussion like that.”

If I’d gone into more detail about 2, there would be more than 2 points of criticism, but the major problem I have is with the DTC genetic testing analysis – primarily because that’s what’s coming next, and so this is where you want the analysis to be as accurate as possible as it actually may have decision-impacting consequences. Razib Khan has written more than a few posts on that topic (23andme-posts) in the past, and as I’ve read him I broadly agree with him regarding the general principles; on the other hand these guys basically seem to consider an outcome where such services are regulated out of existence to be completely fine. The basic argument they present is that they believe (for various bad reasons) that similar principles should be applied to evaluate the ethical framework of these testing measures and the ethics of screening programmes containing tests such as the ones on offer by the private companies (this I incidentally find completely mindboggling, given their repeated emphasis on the differences between voluntary patient-directed health initiatives and screening – see also below), which would mean that counseling requirements would have to be quite strict to justify them – almost certainly too strict for it to make much sense for firms to stay in countries where such requirements are imposed (RK has talked about this aspect before as well). Here’s a relevant quote:

“There is no reason not to apply exactly the same ethical requirements to DTC genetic testing as to genetic screening programmes run by hospitals and other public health institutions. If DTC genetic testing providers can meet these requirements and still make a profit, fine. If not, we may conclude that the very idea of a sustainable business of DTC genetic testing was a fantasy to begin with.”

No thought is given to the problem that the level of counseling provided to actual people using these services may well go down with the formal counseling requirements required (because people buy the services abroad instead, where requirements are laxer). Close to no thought is given to intra-industry dynamics and if one were to take the implicit sketched incentives model of major industry players in chapter 6 seriously one would surely consider it a miracle that there’s such a thing as used car salesmen, or estate agents.

But enough with the criticism, there’s a lot of good stuff as well. They talk quite a bit along the way about the Wilson and Jungner criteria, which is a natural starting point for the analysis. Stuff like that you need to know about if you want to discuss such matters, so here is one version of the criteria, and here’s the 1968 paper (it’s quite long, I haven’t read it). I should point out that the formulation of the principles in the book diverges a bit from those in the link, but it seems that many of the key principles are preserved in the linked version (the book version has 10 points rather than 9, and is a bit more detailed). The main take-away I got from this book is that it’s actually quite a bit harder to justify screening programmes than I’d have thought beforehand – not all the reasons they provide for being cautious about these things I find convincing, but some of them are clearly worth having in mind. Some quotes from the book:

“we will in the rest of this book use the following working definition of the term “screening”:
The use of medical investigation or testing methods at the initiative of health care or society for the purpose of investigating the health status of individuals with the aim of selecting some of these for possible further treatment from a large population of people that is not united by previously recognised risk or symptoms of disease.” […]

“roughly, there are three kinds of goals or values that can provide the rationale for introducing screening programmes:
1. Improvement of (physiological) health, reduction of disease or amelioration of symptoms of disease.
2. Improvement of psychological well-being or reduction of suffering.
3. Promotion of autonomy.
[…] these goals may conflict in various ways.” […]

“The most straightforward way in which an individual may be better off regarding her psychological wellbeing is by having her anxiety or uncertainty regarding her future health status removed – e.g. by having an initial assessment of risk of disease adjusted by having some test. However, the appearance of this possible benefit presupposes that the individual in question has been suspecting that she (or her offspring) is at risk for the disease(s) screened for in the first place (otherwise, there is no initial risk assessment to adjust). Since, by definition, in the case of screening, the initiative to having the test does not come from the individual herself, this cannot be presupposed here. We do not question that reducing anxiety or worry is a valid objective of health care, but this objective is primarily an argument for health care to offer medical investigations if requested by individuals, not for screening. On the contrary, screening runs the evident risk of creating more anxiety than it reduces […] Thus, in terms of well-being, screening programmes (in contrast to the offering of testing to concerned individuals approaching health care with some initial worry) must primarily be justified with reference to their potency for resulting in some kind of improvement of physiological health.”

“since screening programmes tend to target large and unprepared populations, setting up organisations for appropriate counselling is particularly complicated and challenging and usually requires substantial resources. Thus, the room for serious mistakes is larger, and meeting this challenge makes it more difficult for a screening programme to justify its costs compared to less proactive ways of organising the offering of health care services.”

“The line of reasoning underlying the idea that screening improves autonomy is roughly the following: if individuals possess the knowledge that they or their offspring have an increased risk of contracting some disease, they are in a better position to plan their lives in accordance with their own conception of a good life, to live in accordance with their own values […] Traditionally in ethics, autonomy has not been considered primarily as a value to promote, but rather as something that gives rise to moral restrictions on how we are allowed to treat each other when trying to promote other things found valuable […] according to this line of thought, there is a moral obligation to respect autonomy but not necessarily to promote it. […] a central characteristic of screening is that the initiative to the investigation is not coming from the individuals themselves, but from society or large societal institutions. This is potentially problematic already from the point of view of the idea of respecting autonomy, since that idea makes pressure to accept or abstain from medical procedures at least questionable. Respect for autonomy is usually not only taken to mean that informed consent should be obtained in order for an investigation to be justified, but also that the individual should not be pressured or subjected to more subtle manipulative efforts. […] The more that participation is taken for granted in the offer of entering a screening programme, the more pressure there is to participate. The more benefits are emphasised and the less drawbacks are mentioned, the more pressure to participate. The more active the patient has to be in order to refuse investigation, the more pressure to participate. And the more pressure to participate, the more problematic from the point of view of autonomy.35 Although pressure can be reduced in various ways, the general point remains: the very fact that the initiative comes from society always presents some problem from the point of view of autonomy. […] there are practical problems related to screening regarding the need for counselling in order to promote autonomy. […] There is, in effect, always a counter proportionality between pursuing the promotion of autonomy as a goal and economic viability in the context of screening.” […]

“there is no empirical evidence in support of the position that more prenatal testing leads to more discrimination and stigmatisation of the disabled.20 Rather, the trend in many countries seems to be the opposite: in tandem with the development of prenatal diagnosis, more resources have been invested in improving both the quality of life and the societal opportunities of disabled people.” [But this observation did not stop them from arguing as if the stigmatisation argument was valid in chapter 5, while discussing non-invasive prenatal diagnosis, even though there’s absolutely no reason to assume it’s a problem. This fact annoyed me.] […]

“Both prenatal and neonatal screening programmes […] have good chances of being efficient in terms of uptake, i.e. testing a large part of the populations of foetuses and newborns respectively. This is so, since most (prospective) parents are in contact with health care during pregnancy and birth anyway […] this is cost reducing and probably one contributing explanation of the fact that all developed countries have more or less ambitious pre- and neonatal screening programmes.” […]

“regarding these kinds of tests, there always must be a tradeoff between sensitivity and specificity – one cannot increase the one without decreasing the other. […] the case for the importance of high specificity is strengthened if there are no effective treatments […] or if existing treatments are very burdensome […] In contrast, it seems more reasonable to opt for high sensitivity if there is an effective treatment and there are severe consequences for the individual in terms of health without early diagnosis […] Traditionally, among screening specialists, high sensitivity has been seen as a more important feature of a screening programme than high specificity. […] Plausibly, the ethical assessment of how sensitivity and specificity should be balanced cannot be decided by a simple rule. Rather, such assessments would have to be done by careful consideration of each suggested screening programme in a broader context” […] as a rule, the less predictive value, the stronger the reason to abstain to use the test in a screening program.” […]

“These are very good reasons indeed not to use relative risk in the presentation of either what sort of information that may be provided by a screening test, or when informing about the actual result of a test. […] What the patient needs to know both for deciding whether or not to enter a screening programme and what to do on the basis of the test result are the absolute risk figures. In spite of this, there is a widespread tendency in the practice of screening, as well as medical testing in general, for health professionals to press more strongly on relative risk figures when the absolute risks are low or uncertain. […] Since different genes have different penetrance and expressivity […] a genetic test [in many cases] cannot even result in a precise risk-figure. Rather, if a useful test can be construed at all, the result it will present will be in terms of a range of risk, like 5–35% increased risk in comparison to the general population. If the risk-level is low to begin with, the predictive value of such a test, and thus the value as a tool for guidance for health care measures or private decision-making, is virtually non-existing. […] It is […] safe to say, that the threat to the quality of screening programmes created by the fact that the methods for testing and analysis used provide unclear information potentially affects all types of screening programmes in terms of the groups and diseases targeted.” […]

“The extent of false positives [in mammography screening] is debated […] However, Cochrane Collaboration estimates that one in ten during a 10 year period of mammography screening is a false positive. […] When treatment is initiated due to overdiagnosis, it is called overtreatment. One question in the debate regards the magnitude of overtreatment. Figures cited vary from 10% increase of overtreatment due to the initiation of mammography screening programmes,52 to 30% […] Our conclusion from all of this is that the debate regarding overdiagnosis and -treatment, while being cast in the language of scientific facts, is powered mainly by deep ethical disagreements. Positions with regard to this disagreement range from the idea that overdiagnosis and -treatment is not a great problem at all, to the notion that it is a conclusive argument against a screening programme. These positions can, of course, be upheld almost regardless of the magnitude of overtreatment. However, as pointed out, it is difficult to free oneself from the suspicion that this ethical disagreement has been allowed (by one or both parties) to leak into also what is supposed to be the purely scientific part of the discussion. […] the debate would probably gain from bringing implicit value-judgements into the open.”

“the self-reinforcing nature of screening programmes has an important consequence: unless blatantly flawed, they will probably be very difficult to roll back once in place. This has nothing to do with what is medically or ethically well-founded, but follows from the pragmatics of institutions and realpolitik. This should be kept in mind when new screening programmes are proposed. […] if we were to imagine the situation that mammography screening was proposed to be initiated today on the basis of present evidence (including the confusing scientific debate), it would not be an easy task to find it to be warranted. First, partly due to the lack of methodological consensus, the scientific basis of information is much too uncertain and difficult to interpret. Second, if counselling procedures of an appropriate nature were to be included, this would presumably seriously elevate the cost of the programme.” […]

“In 2008, The Cochrane Collaboration published a review: Screening for prostate cancer.71 The review found only two trials that were eligible for inclusion to start with. […] However, in recent years, this scientific vacuum is starting to be remedied. The most ambitious study is The European Randomized Study of Screening for Prostate Cancer (ERSPC), which was initiated in the early 1990s and includes 182,000 men from seven European countries between the ages of 50 and 74. Results published after a median of 9 years shows a reduced rate of death from prostate cancer of 20%.74 However, the absolute benefit (0.7/1000 reduction) was small and was associated with a 70% increase in prostate cancer diagnosis. In other words, 1410 men needs to be screened and 48 more cases needs to be treated in order to prevent one death, as compared to the non-screened population. […] most men tested positive with PSA are subsequently shown not to have prostate cancer; only 25–35% of men who have a follow-up biopsy are diagnosed with cancer […] new data suggest that the extent of the harms of being diagnosed with cancer have not been sufficiently appreciated in the debate. A large cohort study in USA indicates that a diagnosis of prostate cancer may increase the immediate risk of suicide and cardiovascular death. […] there is relatively large agreement that standards of informed consent in PSA screening, at least in the United States, have been poor. A recent survey among test subjects revealed that a majority (71.4%) were told about the upside of PSA testing, while only a minority (32%) were informed about potential drawbacks.”

November 16, 2013 - Posted by | Books, Cancer/oncology, Epidemiology, ethics, Genetics, Medicine, Psychology

No comments yet.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: