Is screening different to case finding in high risk groups
The NSC provides advice to the 4 UK Governments on screening policy
Often “screening” is done under the premise of “ahhh but its not population screening, its case finding in populations at high risk”.
I was asked to try to define the difference between “screening” and “case finding”
I failed to find much of a difference. Here is why
All screening programmes do harm; some do good as well, and, of these, some do more good than harm at reasonable cost[i]. Even when there is good evidence of a favorable benefit of a programme there is no a priori guarantee this will be seen in practice. This underscores the importance of complete implementation and ongoing quality assurance.
Currently the four UK Governments ask UK NSC to provide advice on a proposal to offer population screening. Following a request, the UK NSC systematically assesses the evidence for screening and make a recommendation. The advice of the NSC is used to inform a policy recommendation and implementation of programmes. This is based on a robust process and systematic consideration of the proposal against set criteria[ii], a modified version of the Wilson Jungner criteria.
Other bodies such as NICE may provide an analysis and interpretation of the clinical and economic evidence, but may stop short of contextualising the evidence in a policy recommendation and implementation.
Many (from lay to expert) are unclear about the harms of population screening, and often over estimate the benefits and minimise the harm. The over estimation of benefit is best expressed through the three biases that exist in interpretation of the evidence[iii]. Lead time bias is the overestimation of survival duration among screen-detected cases (relative to those detected by signs and symptoms) when survival is measured from diagnosis. Length bias is the overestimation of survival duration among screening-detected cases caused by the relative excess of slowly progressing cases. Finally overdiagnosis bias is the overestimation of survival duration among screen-detected cases caused by inclusion of pseudodisease—subclinical disease that would not become overt before the patient dies of other causes. Furthermore, it is not infrequent to hear people give a label to an initiative of “case finding” and think this is somehow different to “screening”.
The line between the two concepts of screening and case finding is very gray. It is an important question
Many have interpreted the NICE guidance on diabetes prevention and risk assessment [iv]as a recommendation to screen for diabetes in high risk populations, this is set against the NSC recommendations[v] (which also considered primary evidence in high risk populations)[vi]. The fact that NICE have recommended this may be argued as strong enough evidence for the normal NSC tests to not be applied.
Confusion between screening and case finding
A main source of confusion in the everyday use of the term ‘screening’ is that it can mean any of the following terms that are often conflated, used interchangeably and differentially by different stakeholders:
- A test offered opportunistically to one person
- A test offered systematically to a group of people or a whole population
- A set of loosely linked activities encompassing tests and interventions that roughly comprise a screening programme
- A rigorously quality-assured and evidence-based screening programme encompassing all necessary steps for achievement of risk reduction.
We believe that the the lack of distinction between the terms “case finding” and “screening” matters. Screening programmes are introduced only after a national and very robust assessment of evidence and contextualisation of this into a process of robust implementation and QA. This is a high bar to cross and arguably protects the screened against harm and the wider population of having to bear the opportunity cost of wasteful commitment of resources on low value or harmful screening.
Screening is considered by many to be ethically different from clinical practice. It can be considered a positive offer by the state, to people not seeking help – thus is there a higher ethical standard with respect to more benefit than harm, and as little waste as possible to those screened (harm and overdiagnosis potential, and in terms of opportunity cost borne by others who may be denied care as a result of investment in screening.
One way of avoiding having to cross this bar is to term a proposal “case finding” in high risk groups”. There is no such process to assure the harm benefit ratio, value and acceptable opportunity cost of “case finding”; nor is there any garuntee of robust implementation or QA. There is greater potential for harm and waste, yet the same ethical and clinical issues apply equally.
The UK NSC defines screening as:
“a process of identifying apparently healthy people who may be at increased risk of a disease or condition. They can then be offered information, further tests and appropriate treatment to reduce their risk and/or any complications arising from the disease or condition”.
Other definitions are available (e.g Raffle / Gray[vii], Last[viii]) and there are other definitions of the term across time and in different parts of the world. Case-finding is rather more “difficult to define as it tends to be used rather vaguely. It can mean finding cases in known high risk individuals.” (Raffle Glossary). Last defines case finding as a concept either applying to control of infectious disease – seeking persons who may have been exposed (often referred to as contact tracing) or one of early clinical detection of disease (or risk) in persons using health services for other reasons (for e.g. opportunistic blood pressure checking as part of a visit for another reason).
It can be argued that all attempts to “find” asymptomatic disease carry the same risks of harm (over / under-diagnosis) and waste. Thus it follows that, given the slightly different ethical basis (?) that the same methodology should be applied to assessing the merits of “case finding” as per “screening”.
Drawing distinctions between “screening” and “case finding”.
Here we try to find some distinctions between the terms, using the NSC criteria as a guide.
|Population to be offered a test
High risk case finding is offered to a population considered at “high risk” of something, based most often on some aspect of clinical features, age, gender, family history.
The purpose of this is purportedly to focus down on those most likely to be “found” and thus test as few as possible – avoidance of harm and waste?
The same issues and considerations might be said about “population screening” – a population to invite is carefully selected on the basis of doing maximum good and minimum harm.
Not having a pre-defined population might be argued to lead to ad hoc and opportunistic testing.
The ability to define a population at risk is key – but this applies to both concepts.
|Scale of testing across a population
Does size of population matter.
In any geographically defined population does the scale of the population being tested make a difference to the standards of evidence that should be applied to a proposal? Could one make a case that “screening” about whole or large population and case finding about v carefully selected pops on basis of some construct of risk? Given the above point we don’t think so.
|The condition and the treatment
Does the seriousness of the condition matter? We think this is not a sound ground on which to differentiate screening and case finding.
One might say that “screening” a more appropriate term when significant investment in infrastructure or significant service model change is needed (for example Abdominal Aortic Aneurysm, Diabetic Retinopathy) and “case finding” more appropriate where something can be implemented readily into what is already there? (for example diabetes). However the same issues of risk, harm, over diagnosis and poor value apply, scaled to the population that is being offered a test.
|Call and recall vs ad hoc, opportunistic and sporadic
Is the presence of some form of population register of at risk population with call and recall the defining feature? Possibly.
|Awareness of those tested
those being “found” either through case finding or screening are not seeking help, and unaware of their “condition”.
Therefore this argument does not apply to diagnostic or prognostic testing in clinical practice.
|It is taken as read that both case finding and screening should use a valid test. With appropriate sensitivity and specificity to the task at hand
Not doing so might be argued as unethical.
The NSC test requires – Good q evidence of effectiveness and cost effectiveness of the treatment. This is usually expressed as RCT evidence of reduction in morbidity or mortality comparing those screened and found early compared to those found in routine clinical care.
It is hard to see a justification that case finding should be introduced where there is not a clinically and cost effective treatment available.
|Optimised current treatment
It is not unusual for the UK NSC to recommend against a proposed programme on the grounds that treatment of the current population is not optimised (AF –stroke prevention; Diabetes – key care processes are contemporary examples).
The rationale for this criterion is that there is more good that can be done, often at less cost, by optimising the currently known population; and that adding more may detract from the treatment of the current population, especially if more resources are not made available for treatment. This might be considered unethical
It is easy to argue that the same criterion should apply to case finding.
|Focus on test vs test and intervention vs programme
Historically in the UK, population screening has focused on system and system outcomes, not just test performance, or test performance contextualised in population prevalence or incidence. This may be a distinguishing feature.
A fundamental requirement of UK NSC recommended screening programmes is a well organised QA programme.
Case finding might not have the same requirements, and implementation may be less organised. The consequences of this might be important.
Thus it is hard not to argue that high risk case finding should have robust QA arrangements in place.
|Implications of getting it wrong|
|Clinically, financially (opportunity cost borne by others), reputational.
For a proposal implement high risk case finding (which does not pass the UK NSC criteria) what are the potential implications of a decision which subsequently turns out to not deliver the gain predicted.
There may be occasions where it is politically expedient for UK NSC to not consider a proposal, for a number of reasons.
|Who benefits / who takes risks
Benefits to individuals, population as a whole
Risks to the person tested / not tested or others
Two eg of are things considered “high risk case finding”
Infection contact tracing – others benefit, those who won’t be infected downstream
FH cascade testing those “found” can be risk managed
Conclusion and summary points
We find it difficult to draw any clear distinction between the terms “case finding” and “screening”.
Thus given the special ethical issues associated with inviting those that don’t know they may have a disease for a “test”, combined with the potential for harm and waste we argue that all such proposals should be required to pass the same set of tests as per consideration for a national screening programme. Also “case finding” should be required to implement robust QA alongside the testing component.
The UK NSC can’t consider everything, it doesn’t have the resources to do so. There are other bodies, with differing brief. The English Quality and Outcomes framework for example essentially financially incentivises screening (that may not satisfy the NSC criteria).
Others may have different views and may suggest the above statement is essentially an artificial set of barriers to limit progress.
We are aware of a number of examples of screening programmes masquerading as “case finding” (H Checks, Chlamydia) that would not , or have not passed the NSC test. We are also aware of many localised examples of such phenomena where a deliberate decision has been made to screen in spite of a negative NSC recommendation, or in circumstances where there was no rec (Manchester / Liverpool CT Lung Ca). Such programmes may be beneficial for some, but there is also harm, waste and over diagnosis. Sometimes this is not a question of whether a proposed programme will benefit some people, but it is a question of whether the good outweighs the harm and whether the opportunity cost is worth it and more valuable things could not be done with the same resources.
It follows that if it “looks like screening, smells like screening it probably is screening” (Mackie to STC) and thus a proposal to undertake something under the banner of “case finding” should also need to pass the same set of criteria.
over screening leads to over diagnosis. This can cause harm in itself
over diagnosis leads to over treatment
over treatment can cause harm and waste – iatrogenic harm to well individuals wrongly labelled as “patients” and treated as such; waste in the form of opportunity cost and resources being diverted away from better treatment to those we know are poorly already.
Should ad hoc random testing that often masquerades as “high risk case finding” be re badged as “fishing”
watch “one and the same” – https://vimeo.com/89803387 Brilliant
[ii] NSC to insert
[v] NSC rec and ref to (Prof Waugh paper)
[vi] Simmons R K, Echouffo-Tcheugui JB, Sharp SJ et al. Screening for type 2 diabetes and population mortality over 10 years (ADDITION-Cambridge): a cluster-randomised controlled trial. Lancet. 2012 Nov 17;380(9855):1741-8. doi: 10.1016/S0140-6736(12)61422-6. Epub 2012 Oct 4.
[vii] Raffle Gray
[viii] Last Dictionary of Epidemiology