National guidelines and the gap between At risk and At increased

SNOMED Groups Translation User Group
1

The decision made to translate “At increased risk” as “At risk”, was made partly as understand with input from the nurses´ organization as there was a possibility that risk could be interpreted as decreased or increased.

Now a local stakeholder has objected as there is a difference between At risk and At increased risk. The International version has both, one as preferred and one as allowed, while Swedish and other languages have translated the FSN/Preferred term “At increased risk” as “At risk”. There is no “At increased risk” in Swedish.

For some conditions, the procedures differs depending on if the level is increased or not, for example for suicide risk.

The modelling for some of the concepts do have the interpretation “increased” , like At increased risk for suicide, so the context explains that this is the increased level. The finding is Increased level. (SCTID: 225444004). But the translation is “i riskzonen för…” (At risk of)

There is another SCTID for At low risk for suicide: 394687007.

However, the finding “At risk for suicide”, does not have a concept of its own and At increased risk, which in Sweden require a different procedure, is not visible unless you look at the modelling.

We have a similar problem when it comes to the risk of a fall, where a high risk can be further increased by, for example, certain prescriptions, and require different procedures.

Not all concepts that in English has the FSN At increased risk, are included in the ICNP, but all have in Swedish received the translation “at risk”, and none have “at increased risk”.

Is this an issue that we should bring to the table for new considerations, in accordance with the Swewdish stakeholder’s wishes?

1 Like
2

Unfortunately, we only got this reply:

”Turns out that there have been lengthy discussion on this question in the past. The concepts therefore will remain as they are. I am Attaching the https://conf.spaces.snomed.org/wiki/spaces/Nursing/pages/133990082/At+Risk+Content±+Input+on+Meaning+of+Concepts https://conf.spaces.snomed.org/wiki/spaces/editorialag/pages/133988654/2022-04-05+SNOMED+Editorial+Advisory+Group+Meeting “

3

Hi all members of TUG. We were doing some work on the term abuse (really tricky to translate) and as I was updating a lot of concepts I noticed “child abuse” and the reference to At increased risk as usP, with At risk as usA 704659007 | Child at increased risk of abuse (finding) |.

For us, and most likely for you, suspected or identified child sexual abuse is covered by laws and regulations regarding documentation and actions taken. There are levels here as well.

I am convinced that the previous decision to use only “at risk” in the description and model it only as “increased risk” does not apply to all concepts.

Not popular to raise this issue again, but is there a need to make a rule for when the current strategy should apply and when it is not applicable?

4

What is the clinical interpretation of “at risk”?

Risk is fundamentally a probability of an event occurring over time, and everyone technically holds a ‘general’ or absolute risk, but this baseline susceptibility is not clinically meaningful to document.

The critical distinction lies in relative risk: the patient’s probability compared to the population as a whole. In clinical documentation, the finding of ‘At risk’ implies the presence of a specific risk factor, and by definition, a risk factor is something that elevates probability above the population baseline. Therefore, ‘At risk’ and ‘At increased risk’ are semantically and mathematically equivalent in this context; if the presence of a factor did not result in an increase over the absolute risk, it would not be a risk factor.

This logic is currently supported by the existing SNOMED CT hierarchy, where qualitative assessments like ‘High risk’ and ‘Very high risk’ are modeled as subtypes of ‘Increased risk,’ confirming that ‘At increased risk’ is the correct parent concept for any patient identified as having a probability elevated above the norm.

If the confusion is due to the continued presence of “at risk” descriptions associated with “at increased risk” concepts, then these can be inacitvated.

1 Like
5

Hi @jcase if the FSN for all At increased risk could be replaced with At risk (as your reasoning is that it is the same level for both), this could clear up a lot of confusion.

The point I have been trying to make, is that all concepts with At increased risk are not the same and do not carry the same implications. For example when a child is involved, like:

716565009 | At increased risk of child sexual exploitation (finding) |

704659007 | Child at increased risk of abuse (finding) |

For these, health care personnel are obliged to contact other autorities to alert them about the child at risk.

In Sweden, the first call would be to the social services for child at risk (not general risk, but a specific risk identified by health care personnel). This is not optional, but mandatory in accordance with the Swedish Social Welfare Act (section 14).

But an increased risk, and in urgent cases, the point of contact could instead be the police authorities or the proesecutor. The Privacy and secrecy legislation act (section 21) voids the obligation to keep secrecy for healthcare personell in certain cases, like for children under the age of 18 when the crime against them would be punishable with a prison sentence.

These “at risk”-situations are not only about the need of health care, and the decision to contact the social services or police should be documented. The Snomed CT-codes I highlighted should be relevant.

As we have discussed, the modelling states that these are all increased risks and that is considered to be from a general risk (that is never documented as I understand it), but for the examples we have mentioned earlier as well, there is a need to be able to code also an increase. The example our doctors provided was when an elderly person with documented at risk in regards to falling, is prescribed a medication with the side effect of dizziness, and thus the risk is increased and further measures must be taken.

So could there be a point to look at all At risk/At increased risk concepts and for most update the FSN to At risk, but for some create a separate At risk and keep the At increased risk - but without “At risk” as a usA?

6

To clarify: Not all children would be at risk för sexual abuse, but all elderly would be at risk for fall. So the general “everyone is at risk” would only apply to certain concepts, not all.

7

In Belgian French and Dutch, the increased part was left out on purpose.

The reasoning is that even if “increased” can be modeled through Has interpretation > Increased, it remains a subjective and/or relative notion.

Subjective: because whereas physicians/nurses could consider that all patients are at risk of… and therefore only want to register situations when the risk is increased, others might think that the normal situation is no risk, and find it useful to register both At risk of and perhaps At increased risk of, and perhaps also At decreased and Not at risk of if the risk is considered gone*. In these cases, I think it would be advisable to rely instead on standardized assessment scales (e.g.,the Morse fall scale).

Relative: because the increased notion introduces a comparison that is void, since it is unclear what we’re comparing the risk against. Is the patient’s risk of falling higher than it was before for that patient? Or higher than the average patient? The same would apply for potential Decreased risk of…

Maybe the stakeholder in question can indicate increased/decreased as a separate field if using FHIR standards.

Sidenote: FSNs of these concepts are ambiguous when they refer to a person (elder, child). Without access to its parents and/or definition, which is the reality for most SCT users in real-life settings, At increased risk of child sexual exploitation (finding) is likely to be interpreted as to mean that the patient has a risk of exploiting a child. Another example is 706874009 |At increased risk of elder neglect (finding)|.

*As in 1157152000 |Not at risk of elopement from healthcare setting (situation)|

1 Like
8

Totally agree with the ambiguity issue @plammertyn! We actually had to correct some translations as they could be interpreted as a child commiting sexual abuse instead of being the victim of abuse.

9

Q: Understandability: Do we understand the meaning of the concepts given their FSNs, relationships, attributes (and text definitions)? In other words: Do we understand the distinction between two or more concepts?

A: Finding of risk level is the parent of Finding of increased risk level and the parent of Finding of low risk level.
In regard to the background provided by ICNP and SINT:s terminologist this should be understood as all concepts with At increased risk should be interpreted as from the perspective of a general risk.
However, nothing in the logical model conveys this clearly, and the text description should stand on its own but a usP At increased risk with a usA At risk has most times been translated as only At risk, with some translators not getting the memo and translating is as At increased risk.
Some queries remain like is the low risk level also in comparison to the general risk level? Or is it the result of safety procedures when the patient had increased risk levels?

Q; We don’t assess issues regarding the target langauge expressivity, only the semantics of the concepts in the ontology.

A: Is At risk an acceptable synonym to At increased risk on a semantic level? Here most would argue not. Only a definition could have explained that the perspective “from a general level” makes the two equal. And how many coders read the definition?

Q: If the semantics are unclear, it is a content issue of SNOMED CT. If the semantics is clear, we proceed to the next question:

A: SINT argues that the semantics is clear. Even if I do not agree, we can move on to the queries relating to non-content issues.

Q*: Is the distinction (i.e. semantics) clinical useful and relevant?*

A: We need an increased risk level in documentation as added factors will increase the risk level.

Q: We, the translators, understand the meaning of the concepts, but they are not deemed as useful and relevant for the clinicians.

A: I would say that the terminologists could understand, but the clinicians would rather have a starting point At risk, and concepts to stress that the risk has increased or decreased or is unchanged.

Q: This might be explained by the logical structure of SNOMED CT, i.e. the distinction between concepts as clinical artefacts and ontological artefacts. The concepts might not be relevant for a clinicians, but maybe for other user groups?

A: Highly relevant to document risk level, but maybe not from a philosophical point of view that there is a general risk level.

Q: What is the use case of including the concepts? Is this rather an implementation issue than a translation issue?

A: In implementation cases, our users in Sweden object to At increased risk being translated as At risk.

Q: Language granularity: We understand the difference between concepts, and they are deemed as useful and relevant, but the distinction is not expressed by different terms in the target language.

A: The problem is not with the term “increased”, but a definition could have helped here. The easiest solution, however, would have been to not add “At risk” as a synonym.

Q: How to proceed in this case is an issue of translation, not content in SNOMED CT. Different approaches or strategies can be applied, including neologisms

A: Translation-wise we can make the decision to translate the meaning of the FSN and keep increased, but the problem arises if different languages take different approaches to this, especially in light of the EHDS where the languages will be switched in the interface in order for health care personnel to read an patient summary originating in a different languages.

10

Hi @klindve, as discussed in today’s TUG meeting, this topic could be discussed in Vienna, along with your proposed solution.