19.07.05 New test in lab Flashcards
What is verification
Confirmation, through the provision of objective evidence, that specified requirements have been fulfilled (doing test correctly). Comparison with existing performance specification.
What is validation
Confirmation, through the provision of objective evidence, that the requirements for a specific intended use or application have been fulfilled (doing the correct test). Define performance specification (accuracy, limitations, controls).
What framework is used in the process of implementing a test
ACCE
What does ACCE stand for
Analytical validation, clinical validation, clinical utility and consideration of ethical, legal and social implications of the test.
Factors considered when implementing new tests
- Is there a clinical need
- is there a clinical demand for new service
- New service have clinical utility
- Is the test legal/ethical
- testing strategy
- How would the new service impact current working practice
- budget considerations
- staff skills, education and training
- Equipment and lab considerations
Is there a clinical need for the service? Things to look for
- Incidence and prevalence of a condition
- Nature of disease, severe enough to require genetic diagnosis
- Predisposition to a disease (carrier risk, reproductive counselling)
- Expected workload
- Is the service already provided by an alternative accredited lab with sufficient capacity
Factors in deciding if there’s a clinical demand for the new service?
- Who is requesting service. —-Physicians, counsellors, geneticists, patient lead groups
- Where would referrals come from
- Is a new service necessary if another lab offers
- Would it be a definitive diagnostic test or a screening test to rule out a condition (e.g. Fragile X and dev delay.)
Factors in deciding if the service have clinical utility
- Would testing affect treatment or identify an effective drug regimen.
- Would it provide a diagnosis for a currently unidentified genetic disease, where no test is currently available
- Identifies low penetrance alleles (HD)
- Do environmental factors have a contribution to phenotype.
- Used for other family members (predictive, prenatal testing)
Factors to decide if the test is legal/ethical
-Particularly with regard to prenatal/predictive testing
-Reporting incidental findings, data sharing and storage
-Sexing for family balancing
Paternity testing
Factors to determine technical testing strategy
- Biochemical, cytogenetic, molecular.
- Single gene or multigenic
- hotspots, common path vars
- Alternative tests: biochemical, histological, clinical. Will other tests be required (methylation)
- Analytical validity (reliable, reproducible, effective, specific, sensitive)
- Use of controls
- Uncertainty of measurement
- Would it detect intermediate outcomes (mosaics, premutations)or incidental findings.
- What would be tested- DNA, RNA, chromosomes. Is the material easy to obtain.
- test samples- easy to obtain, extract. Appropriate for test. e.g. Pallister-killian not seen in blood.
Impact on current working practices
- Increased workload
- Would staff need training
- Time required to set up service (validation, purchasing new equipment, training etc)
Considerations if a test is to replace a current one
- Does it improve TAT, is it more sensitive/specific
- Does TAT fit with clinical need (short enough to act clinically)
- How much DNA required- less than previous test
- Are there national guidelines, EQA schemes
- Has the test been externally validated. How to internally validate/verify
- What controls are required
- Will it impact another service, i.e. does it need to be confirmed by another test.. e.g. array CGH reducing karyotyping. MLPA to validation del/dups found by NGS
What budget considerations are there
- Feasible service cost, could it be offered at a competitive price/TAT and cover costs.
- Is funding available to accommodate introduction of new test
- Start up cost
- Recurring costs of on-going testing, maintenance of equipment
- Is a business case needed
- Is pharmaceutical company funding available, if result indicates treatment of a drug
- billing procedure for service users
Implications on staff skills, education and training
- Determine workload and staffing requirements. i.e. expected number of samples, amount of work per sample.
- Do staff already have adequate training. Is internal/external training required. Internal/external support for troubleshooting and clinical advice. Enough competent staff to run service (staffing structure). Can you recruit new scientists with appropriate exp and skills.
- IT bioinformatic support, data storage requirements. LIMS functionality.
- Health and safety aspects. Implementing SOPs, internal QC, risk assessments, COSHH.
- Test development in house or require expensive kits/ equipment
Equipment and laboratory considerations
- determine the best tests for service, including hands on time, throughput, sensitivity and cost.
- Requirements for new equipment. Leased/ bought outright. Will capital funding be required from trust.
- Start up and maintenance/running costs of equipment.
- If new equipment required can it be used for other services (current and in future).
- Can current equipment be used for service. Will equipment cover demand
- Does a cheaper test exist. Can testing be automated.
- Suitable space in current premises to carry out new test. New space required or current space re-arranged.
Population factors
- incidence of disease in total population.
- Does incidence warrant new service (e.g. if very rare)
- Will service be national or regional. If national can sample arrive quickly enough.
- Other considerations: reduced penetrance, variable expressivity, pre-mutation, recessive carrier.
- Ethnicity of the population. Do certain populations have higher prevalence (e.g. Ashkenazi Jews- Tays Sachs 1/27 carriers)
- Are certain variants seen in all population types or in certain ethnic populations.
- Broad population spectrum testing/ validation, that takes into account ethnicity.
- Other factors like socioeconomic, environmental, co-morbidities, age, gender that may affect accuracy of testing.
Background work
- Disease: clinical aspects (acceptance criteria agreed with clinicians), inheritance pattern, genetic causes.
- Technical: equipment/reagents required, controls, design assay.
Validation work
- Adjust protocols if required.
- Write SOPs
- Screen anonymised samples (positive)
- test control samples in parallel (old vs new). Demonstrate accuracy, repeatability, reproducibility, robustness.
- Audits to assess new test (examination, vertical and horizontal). Look for quality improvements, non-compliances.
- Review figures. Expected vs observed numbers of referrals. Pick up rate as expected.
- Relevant EQA schemes available? If not can you cross-lab test
- Meet ISO requirements?
Implement service
- Staff are appropriately trained
- SOPs authorised
- Managing workloads- batching samples for testing
- Design worksheets, sample database
- Design report templates. Riders, cut-off values etc.
- Advertise service: lab website, UKGTN, ACGS, BSMG, conferences.
- Audit of requests, results.
- Submit variants onto public databases
- Write gene dossier or BP guidelines.
Feedback from users
- Surveys, user group meeting
- Act on feedback to optimise service
On going verification
- Periodic review of trends when in routine use.
- Recording process variations (errors/incidents).
