Screening

Screening

Definitions

·        Screening: 

o  The presumptive identification of unrecognised/preclinical disease or defects by the application of tests, examinations or other processes that can be applied rapidly (and cheaply)

o  Sorts people into high and low risk groups for further diagnosis of high risk.

o  Is NOT diagnostic on its own

·        Mass/population Screening: systematic screening of populations

·        Opportunistic screening: Non-systematic, when the opportunity arises

·        Selective Screening: systematic screening of high risk groups

·        Screening test: a test performed without a clinical indication

·        Objectives of screening:

o   ¯Mortality/morbidity from disease on the individual

o   Limit the impact of disease on a community

o   Identify compensatable disability (eg poor eyesight in kids)

Criteria for Screening Programmes

·        Is the disease an important health problem (incidence, impact, preventability)

·        Is a suitable screening test available:

o   Acceptable, simple,

o   High sensitivity, as specific as possible 

o   PPV: probability that a person with a positive test does have the disease, depends on sensitivity, specificity AND prevalence

o   NPV: probability that a person with a negative test does not have the disease

o   Yield (proportion of cases of the disease accurately identified by a screening test),

o   Repeatability (depends on variation in method, subject variation, observer variation)

·        Is the natural history of the disease well understood:

o   A recognisable latent or early symptomatic stage 

o   The length of the asymptomatic stage determines screening frequency. Is this long enough to make the screening interval reasonable?

·        Does screening lead to interventions that improve quality of life:

o   Does early intervention offer benefits over later intervention 

o   Accepted treatment, proven effectiveness. Ideally want an RCT that demonstrate screening verses no screening improves mortality/morbidity 

·        Is there an appropriate infrastructure available to provide screening and follow-up services:

o   Are there pilot studies demonstrating how it should work?

o   Is there local and national support? 

o   Are the services accessible (in terms of geographic, cultural barriers, cost), does the system have sufficient capacity, and is there appropriate quality control processes in place 

·        Is the screening programme cost effective?

Screening Test

·        Eg PSA for prostate cancer, intra-ocular pressure for glaucoma, etc 

·        For a screening test, you want a test that maximises sensitivity: maximise true positives (minimises false negatives), so that you identify all diseased cases. The downside is an increasing rate of false positives who have unnecessary further investigation 

·        A highly specific test would maximise true negatives (ie minimise positives, so would not further test anyone unnecessarily), but at the cost of increasing false negatives – who are the people you actually want to detect 

·        See also Topic: Evaluation of Diagnostic Tests

Biases

·         Lead-time bias: interval from detection to point where diagnosis would have been made without screening. Depends on length of pre-clinical phase, frequency of testing, and the test sensitivity 

·        Length bias: Cases with a disease with a longer natural history are more likely to be detected by a screening programme. But these cases also have a better prognosis. Thus screening leads to a better prognosis, regardless of whether screening itself confers any benefits

·        Selection bias: selection, referral or volunteer bias results in a selected subset of the population being screened

Screening Programmes in NZ

·        National screening programmes:

o  Neonates: inborn errors of metabolism – Guthrie Card

o  Cervical Cancer

o  Vision/hearing testing at school entry (erratic)

o  Mammography

·        Controversial and not currently recommended population screening programmes:

o  Prostate (PSA)

o  Colorectal cancer

o  Otitis media with effusion

·        Current screening pilots: Hepatitis B

·        Opportunistic screening

o  Antenatal screening

o  Blood pressure

o  Cholesterol

o  Blood glucose

o  HIV

o  Osteoporosis

o  Glaucoma, etc

·        Deciding to implement a screening programme: 

o  The decision to implement a population based screening programme is complex, must be justified on the basis of standard WHO criteria and supported by research evidence 

o  The rules to do with population health are NOT those of an individual clinician (ie just because you would screen an asymptomatic man for prostrate cancer is not a reason to implement a national programme) 

o  Potential to do harm at a population level is considerable („first do no harm‟)

o  We are „imposing‟ something – need sound evidence

Ethical Considerations

·        Costs and benefits: 

o  Costs should include adequate support, counselling, etc. Benefits should include quality of life (but subjective) 

o  Many harms are personal – false alarm, false reassurance.  Difficult to account for

·        Justice:

o  Distribution: benefits accrue to a few and are large; harms fall on many and are minor.  Is this fair? 

o  Inconvenience borne by many to benefit the few – but this also benefits the group (social welfare function) 

o  Collective gains depend on high levels of individual participation

·        Autonomy:

o  Motivation: altruism only effective if participants well informed/educated

o  Imposition: Opt-out strategies – trade-off between recruitment level and maximal choice

o  Results: safeguards on third party disclosure

·        Opportunistic screening:

o  Cost and benefit usually borne by the same individual

o  Offered responsively rather than proactively

o  Appropriate treatment or other follow-up available

·        For Prostate Screening, see Prostate Cancer Screening. Prepared for Public Health test.