Add term for clinical studies, clinical trials, and observational trials

[cthoyt]

A question I posed on Slack: Is there an OBI class that's appropriate for clinical trials? I've found investigation (OBI:0000066) which might be a good superclass. Here's what's also in the landscape that I could find on OLS:

1. SIO has a term for clinical trial (SIO:001000), but SIO isn't an OBO ontology
2. OPMI has a term for clinical trial (OPMI:0004507?) but I'm not comfortable reusing terms from this ontology because it seems like it's a pretty use-case-specific ontology, and additionally, it's not using any modern infrastructure so I am not sure I can trust it will be maintained
3. ICO also has a term for clinical trial (ICO:0000065) but I have the same reservations. It looks like there's an overlap of the same authors in these ontologies, so I have reservations to invest in reusing terms they haven't been able to deduplicate themselves.

As someone who wants to annotate clinical trial instances with an appropriate class, I want to have this term in an ontology I can trust 🙂

Comment from Chris:

In the past, some ontologies have used this patchwork approach but it leaves us in odd positions where pieces of patchwork get abandoned and no one know where to go to get anything. OBI should claim anything under “investigation” that is biomedical and make OBI IDs.

Comment from James (on slack):

Yes, OBI 'investigation' includes clinical trials, and it's in OBI scope. We don't tend to subclass 'investigation', for whatever reason, but a good case can be made for a 'clinical trial' subclass. There are various "clinical" 'study design' terms, and many assays.

[bpeters42]

There is a lot of history on this, with several competing projects wanting to cover clinical trial work, and OBI was staying out of it in the spirit of not doing overlapping work. But given that I can't find the other projects any more (beyond what you wrote above), I do think OBI should cover this directly, and as James was saying, we have actually a substantial number of clinical trial specific terms for study designs in OBI. One of the questions back to you Charlie: What would you consider a clinical trial? There are regulatory definitions (which have changed massively over time, and are country specific); and then there is a more broad 'human health related' study definition. It would be helpful to know what you are planning to annotate.

[cthoyt]

The short answer is anything that appears in any of the 23 clinical trial registries around the world, since my motivation is to ontologize the records in them ;)

After thinking about your question a bit, we might need to clarify the difference between a clinical study and clinical trial

Here's what ClinicalTrials.gov (the highest quality registry) says a clinical study is:

A research study involving human volunteers (also called participants) that is intended to add to medical knowledge. There are two types of clinical studies: interventional studies (also called clinical trials) and observational studies.

Here's what ClinicalTrials.gov says it lists (ref):

ClinicalTrials.gov lists studies that involve people, have health-related research questions, and follow ethics review and other health authority rules and laws. The ClinicalTrials.gov database includes these types of studies:

• Clinical trials: research studies in which researchers assign participants to get one or more interventions (such as a drug, behavior, or medical device) to test what happens in people.
• Observational studies: research studies in which researchers simply collect information (called data) from participants or look at data that was already collected. The data may be about participants’ health, habits, or environments. In observational studies, researchers do not assign participants to get an intervention. If there is an intervention, participants were already using it as part of their regular health care or daily life.

Learn more about types of clinical research studies on the Learn About Studies page.

[bpeters42]

Thanks Charlie. Modeling those studies would be great, and what you picked meshes with the current NIH definitions. FDA has different ones, focusing on regulatory aspects (that's where the phases come in). So we can add as subclasses under investigation: - clinical study: An investigation in which data is generated on human subjects with the goal of advancing medical knowledge - Observational clinical study: A clinical study in which data is generated on human subjects without assigning them a specific intervention - Clinical Trial / Interventional clinical study: A clinical study in which a specific intervention is assigned to human subjects with the goal of learning more about the effects of that intervention Note that we have 'study intervention' as a type of planned process that is part of an investigation, so you would have things like: "Vaccine Trial" is_a Clinical trial and has_part some ('study intervention' and vaccination). It would be great to test out with you if OBI has everything you need to model the clinical trials you are interested in, and we should add / adjust to make that happen. Also: All of the above are suggestions that will need to be discussed with the broader OBI group, but it would be great to get your feedback now. 'Human subjects' (not necessarily volunteers)

…

On Thu, Oct 10, 2024 at 11:54 PM Charles Tapley Hoyt < ***@***.***> wrote: The short answer is anything that appears in any of the 23 clinical trial registries <https://bioregistry.io/collection/0000012> around the world, since my motivation is to ontologize the records in them ;) After thinking about your question a bit, we might need to clarify the difference between a *clinical study* and *clinical trial* Here's what ClinicalTrials.gov (the highest quality registry) says a clinical study is: A research study involving human volunteers (also called participants) that is intended to add to medical knowledge. There are two types of clinical studies: interventional studies (also called clinical trials) and observational studies. Here's what ClinicalTrials.gov says it lists (ref <https://clinicaltrials.gov/about-site/about-ctg>): ClinicalTrials.gov lists studies that involve people, have health-related research questions, and follow ethics review and other health authority rules and laws. The ClinicalTrials.gov database includes these types of studies: - *Clinical trials*: research studies in which researchers assign participants to get one or more interventions (such as a drug, behavior, or medical device) to test what happens in people. - *Observational studies*: research studies in which researchers simply collect information (called data) from participants or look at data that was already collected. The data may be about participants’ health, habits, or environments. In observational studies, researchers do not assign participants to get an intervention. If there is an intervention, participants were already using it as part of their regular health care or daily life. Learn more about types of clinical research studies on the Learn About Studies <https://clinicaltrials.gov/study-basics/learn-about-studies> page. — Reply to this email directly, view it on GitHub <#1831 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADJX2IU2UN7DUEBDF3BXHFLZ25YZ7AVCNFSM6AAAAABPWKTVFOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDIMBWGY3TCMRRHE> . You are receiving this because you commented.Message ID: ***@***.***>

-- Bjoern Peters Professor La Jolla Institute for Immunology 9420 Athena Circle La Jolla, CA 92037, USA Tel: 858/752-6914 Fax: 858/752-6987 http://www.liai.org/pages/faculty-peters

[cthoyt]

It would be great to have a first-class notion of saying what the specific intervention(s) are, as well as what the conditions that are being studied are. Note that interventions can be all sorts of things, including small molecules, biologics, vaccines, supplements. Similarly, trials usually have one or more conditions associated with them that are being studied.

There are also a pile of more specific kinds of annotations to put on, including the clinical phase as you pointed out, but these would be the most important

[bpeters42]

All the interventions you mentioned are straightforward, and would fall under 'administering substance in vivo' (http://purl.obolibrary.org/obo/OBI_0600007), which includes not only what is given, but also how (e.g. oral, injection, cream, etc.). But as say, all kinds of things can be interventions, like an exercise program, behavioral therapy, light exposure, and that all should work with the approach we have taken of just pointing out that one part of the investigation is the intervention.

The other parts we have used for clinical study modeling would be human subject enrollment (http://purl.obolibrary.org/obo/OBI_0600004) tied to things like age ranges and other eligibility criteria, and what kind of readouts (endpoints) the trial is determining, which would fall under OBI:assay. And the definition of independent arms / cohorts in a trial and cross-comparing them.

So that was a long way of saying: I believe all the parts are in OBI, but I don't think we have done a good way of demonstrating how they should be canonically linked together, and it would be great to make that more transparent / add what is missing given your use case.