(Exams) Questions to remember

Question 1

Ontologies and SOAP notes

Two reasons why the use of ontologies improves the value of the data in SOAP notes?
One reason why forcing the use of ontologic terms in SOAP notes is unacceptable?

Answer 1

The 2 reasons:
- Disambiguation of descriptions.
- Makes SOAP notes easier to parse for automatic analysis.

The 1 reason:
- Doctors like to use drawings with their text, sometimes they don’t know the exact medical term.

Question 2

E-health and medications

Two reasons for new health products based on apps coming so fast to the market compared to other health products such as medications?

Answer 2

• New apps don’t have too many regulations.
• Only a download away, some of them without any cost.

Question 3

Genomic vs. Phenotypic vs. Exposomic

For each type of data say which high level data category can be obtained from that type.
- SOAP notes.
- DNA sequence data.
- Medical imaging data.
- RNA expression data.
- DNA methylation data.
- Gut metagenomic data.
- Blood test data.
- Family disease data.

Answer 3

• SOAP notes: P, E.
• DNA sequence data: G.
• Medical imaging data: P.
• RNA expression data: P.
• DNA methylation data: P.
• Gut metagenomic data: E.
• Blood test data: P, E.
• Family disease data: G.

Question 4

SNP and its expression

A given SNP in the coding region of a gene in the human genome is known to be associated with a disease.
Which of the molecular biology processes listed below can explain why an individual with the SNP does not have the disease?
- Recombination.
- Replication.
- DNA methylation.
- Histone modification.
- Transcription.
- Splicing.
- RNA interference.
- Post-translational modification.

Answer 4

• DNA methylation (silencing of the gene).
• Histone modification (silencing of the gene).
• Transcription (if that gene does not strongly disrupt the regulatory network).
• Splicing (remove introns and ligation of exons; that exon is not included).
• RNA interference (silencing of the gene).
• Post-translational modification (modification of proteins; if that SNP has a minor effect on the protein, the protein has mechanisms to recover from that).

Question 5

CDSS into EHRs

Two examples of integration challenges of a CDSS into EHRs and provide a way to overcome these constraints.

Answer 5

• Costs of implementation (cost-effectiveness analysis).
• Integration (interoperability).

Question 6

Testing hypothesis

It is hypothesised that genetic variations may also play a role in the onset of severe COVID-19. How would you go about testing this hypothesis? Propose one approach, detailing whom you would test, what type of data you would capture, how you would capture it, and what results you would expect if the hypothesis is true.

Answer 6

• Sequence the genomes of COVID-19 patients, divided by prognosis (light vs. severe).
• Compare the genomes through sequence alignments.
• Use gene association studies to see if any variant is related with COVID-19 prognosis.

Diagnosis - “Your diagnosis is the flu”.
Prognosis - “Your prognosis is a full recovery after lots of rest”.

Question 7

CLAMP

Explain why the recognition of clinical entities within textual sources is often insufficient for a correct interpretation of the data. What could be done?

Answer 7

Entities are not associated to unambiguous entries in a clinical terminology, such as SNOMED CT.
Associate these different mentions to a same identifier.

Question 8

Verifying biology mechanisms

In the following 3 biology mechanisms, name 1 experimental technique that can be used to verify that same mechanism.
- DNA methylation.
- Histone modification.
- RNA interference.

Answer 8

• DNA methylation (2nd or 3rd generation sequencing).
• Histone modification (chromatin immunoprecipitation + mass spectrometry).
• RNA interference (RNA sequencing).

Question 9

Interoperability

What is system interoperability? Assume 4 systems, A is compatible with B, B with C and C with D, are they interoperable?
What is data interoperability? Give 2 things to use to assure data interoperability in the datasets.

Answer 9

System interoperability
- Compliance to an open standard.
Only if A, B, C, D comply with the same common specification we can say that they are interoperable.

Data interoperability
- Each computer system has knowledge of the other system data exchange formats.

To assure this:
1. Common vocabularies (unique and persistent id’s).
2. Well-defined framework to describe and structure (meta)data.

Question 10

CDSSs

True or false.
1. These systems replace the clinician in deciding what is the best treatment.
2. Care interruptions is a drawback of CDSS.
3. CDSS contribute to health costs reduction.
4. CDSS can improve patient safety by reducing prescribing and dosing errors or by providing drug allergy and contraindications warnings.
5. Free-text notes from the EHRs can not be used as data source for CDSS.
6. The knowledge based CDSS are grounded on the use of Bayesian probability.
7. In passive delivery systems the user has no control of the decision.
8. The performance of the system is directly related to the rate of generated alerts.
9. Knowledge based systems can transform guidelines into point of care systems.

Answer 10

1. F.
2. T.
3. T.
4. T.
5. F.
6. T.
7. F.
8. F.
9. T.