CIDO-ontology / cido

CIDO: Coronavirus Infectious Disease Ontology
https://obofoundry.org/ontology/cido.html
Creative Commons Attribution 4.0 International
20 stars 11 forks source link

Difference between COVID-19-related clinical finding and diagnosis? #6

Open yongqunh opened 4 years ago

yongqunh commented 4 years ago

Clinical finding in OGMS: Term IRI: http://purl.obolibrary.org/obo/OGMS_0000014 Definition: A representation that is either the output of a clinical history taking or a physical examination or an image finding, or some combination thereof. subclasses:

diagnosis Term IRI: http://purl.obolibrary.org/obo/OGMS_0000073 Definition: The representation of a conclusion of a diagnostic process.

Asiyah: clinical finding (based on symptoms or ) is input of a diagnosis.
'diagnostic process': has_input some 'clinical finding' has_output some/only 'diagnosis' (Bill: if use only, diagnostic process does not have to have output diagnosis. If use some, you always get some diagnosis. However, if diagnosis process has another output, it violates the rule.) Bill: sometimes we get no diagnosis or wrong diagnosis. Oliver: then no diagnosis or wrong diagnosis is itself a diagnosis.

yongqunh commented 4 years ago

diagnostic process Term IRI: http://purl.obolibrary.org/obo/OGMS_0000104 Definition: A health care process that involves the interpretation of a clinical picture from a given patient (input) and the assertion to the effect that the patient has a disease, disorder, or syndrome of a certain type, or none of these (output).

clinical picture Term IRI: http://purl.obolibrary.org/obo/OGMS_0000016 Definition: A representation of clinically significant bodily components, dispositions, and/or bodily processes of a human being that is inferred from relevant clinical findings.

Based on the above definitions: 'diagnostic process': has_input some 'clinical picture' has_output some/only 'diagnosis' Oliver note: for RT-PCR method, we do not have a clinical picture. 'clinical picture' appears a very clear term. Asiyah: we probably don't need to implement the clinical picture as another layer. Oliver: what does it mean by "inferred from relevant clinical findings"? Asiyah: if SARS-CoV-2 RT-PCR test is positive, it does not mean that the patient is for sure positive since it can be false positive.

yongqunh commented 4 years ago

Sivaram: Clinical picture is a combination of certain clinical findings usually for a disease. So it is a collection of clinical findings.
For example, a patient with pneumonia typically presents with a clinical picture of cough, fever, breathless, chest pain, sputum, and consolidation on chest X-ray.

Asiyah: Radiologists often see diffuse cloudiness in the X-Ray image for COVID-19 patients. In this case, we may not need clinical picture. We can directly use clinical findings.

Sivaram: yes. we can say that. we don't have to use clinical picture. but we can say that COVID-19 patients present with clinical picture of the following symptoms: cough, fever, ....

Asiyah: The clinical picture can be mild, intermediate, severe. This is disease stage, so it is clinical picture.

Sivaram: Each disease stage presents with a specific clinical picture. If it is at an asymptomatic disease stage, it's a clinical picture where the patient is not exhibiting any clinical symptom or sign. These asymptomatic patients are detected by lab tests (like PCR test antigen positive, but not antibody test positive).

Bill: This is good on details. Oliver: I also agree.

yongqunh commented 4 years ago

Bill: One clinical picture can be associated with different diseases. Like ear disease.

Sivaram: True. That's why the output of a diagnostic process is a differential diagnosis. You want to go with most common diagnosis that fits the pattern. After a differential diagnosis, we may want to confirm the diagnosis using different methods.

Different kinds of diagnosis: initial diagnosis, final diagnosis, working diagnosis, wrong diagnosis.
"Wrong diagnosis" is problematic. Based on a diagnosis, do some treatment. If the treatment improves the outcome, it proves it's right diagnosis. If not, it proves it's a wrong diagnosis. Think diagnosis as more a hypothesis that needs to be proven.

Asiyah: So we can create an AI diagnosis system. A diagnosis may just be a possible diagnosis.

https://en.wikipedia.org/wiki/Differential_diagnosis

linikujp commented 4 years ago

OGMS had a issue tracker on updating the definition for clinical picture. https://github.com/OGMS/ogms/issues/138