SNOMED CT Editorial Advisory Group

Page tree

Background

A request was received to add a concept for "Removal of a joint prosthesis due to infection".  The issues associated with this request were discussed.

Initial problem statement

We have a number of terms, both disorder and procedure that deal with "infected prosthesis". In general, prostheses themselves are not infected, but the surrounding soft (or bone) tissue adjacent to the prosthesis can become infected. This infection often does not have a demonstrable causal or temporal relationship to the procedure. Currently these are modeled with an ASSOCIATED WITH relationship:

Question:

How do we best represent the true nature of the infection? This is especially important when we deal with "Removal of prosthesis due to infection" and concepts such as "Infection of implanted cardiac device (disorder)".?

Based on previous discussions regarding "causal chain", should this be a DUE TO relationship since the infections would not have occurred if the procedure had not been done?

Initial discussions by EAG

A prosthesis can be colonized (e.g. vegetation on a prosthetic heart valve). The need to associate a procedure with these would be unnecessary and in many cases incorrect. The use of a DUE TO relationship to the device or the procedure is not appropriate.

Currently, the involved concepts inherit ASSOCIATED WITH = "Procedure" from the parent "Complication associated with device".  This relationship is not appropriate as the infection may not have any causal relationship to the procedure (which is a subtype of ASSOCIATED WITH).

There are also timing aspects that are not represented in these terms, which make them more vague.

The associated problem is the need for a definition of what is meant by "infected device". If we view the presence of the device as just another acquired body structure, then these may not be complications. The timing of the infection in relation to a procedure, may be the reason to classify something as a Complication of a procedure (i.e. within a certain number of days).

The two approaches are "close to reality", which is multi-dimensional and challenging to determine, or "simplified model" that just describes what is certain. The determination of whether something is a complication or not is often unknown. Some testing will need to be done to see the impact of applying a simplified model. If it does not meet the needs from a classification standpoint, then a more complex model will be needed.

This argues for the use of ASSOCIATED WITH = "X device" to describe the relationship for devices.

For the procedures such as "Removal of prosthesis due to infection" the possibility of the use of HAS FOCUS.

There are guidelines on the evaluation of patients prior to implantation, where pre-existing infection would cause abortion of the procedure.

Clarification on the current understanding of Complications can be found here.

Modeling approach

A recent review of prosthetic joint infections (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3993098/) indicates that the term "Prosthetic joint infection" is really a misnomer and refers to the colonization of a prosthetic joint by an infectious organism and infection of the periarticular tissue.  This was confirmed by a joint conference on periprosthetic joint infections (https://www.efort.org/wp-content/uploads/2013/10/Philadelphia_Consensus.pdf). This spread of the infection to the surrounding tissues must be taken into account in the modeling approach.  Additionally, while the majority of this type of infection occur within the first two years after implantation, there are a substantial number (up to 30%) that occur much later, thus the direct association of the implantation procedure with the infection is not supported.  However, as the colonization of the implant cannot occur with having been implanted, there is a temporal relationship (i.e. AFTER) with the Implantation procedure.  Given these considerations, the following changes to the modeling of "Prosthetic joint infection" were tested.

Prior stated form                                                                       New stated form


Prior inferred form                                                                   New inferred form

Modeling analysis

This modeling approach:

  1. removes the incorrect causal associations (i.e. DUE TO = Procedure)
  2. makes explicit the infection localization to the joint structure (which subsumes the periarticular tissues in the anatomy hierarchy)
  3. To correct the inference, the prior "Disorder of prosthetic joint" was remodeled and made fully defined.  This may need to be re-evaluated as to whether a disorder of an implnat is in itself a complication, which would again make "Prosthetic joint infection" a subtype. 
  4. The resulting classification places this term under a large number of fully defined groupers, none of which appear to be incorrect.  There are still a number of terms in this area of the temrinology that are primitive and would benefit from a review and remodeling accoring to this pattern.

Modeling "Removal of prosthetic device due to infection"

The remodeling of "Prosthetic joint infection" allows Procedure terms to be modeled using the HAS FOCUS relationship.  In order to be more specifci in the FSN, the term was changed to "Removal of joint prosthesis for periprosthetic joint infection (procedure)". The following modeling approach was taken:

Stated form                                                                                           Inferred form


Modeling analysis

This approach seems to have met the requirements of the initial request.  The pattern proposed here, i.e. use of the HAS FOCUS relationship, is an acceptable pattern for "Procedures for disorders".




  • No labels

10 Comments

  1. Bruce Goldberg

    jim, a few more comments:

    1. WRT the statement "To correct the inference, the prior "Disorder of prosthetic joint" was remodeled and made fully defined.  This may need to be re-evaluated as to whether a disorder of an implant is in itself a complication, which would again make "Prosthetic joint infection" a subtype", the remodeled concept infers 2 complication parents, 69260008 |Complication of internal prosthetic device (disorder)| and 19220005 |Complication of implant (disorder)|. As previously discussed for the 473023007 |Complication associated with device (disorder)| hierarchy, the kind of concepts this subsumes are complications due to the procedure to place the device, complications due to the presence of the device and problems with the device itself. Only the first 2 are complications, the 3rd is a finding about the device. I would say that a periprosthetic joint infection should be considered a a kind of complication. The complication is the infection of periprosthetic structures. We are not including in the model the underlying cause which is the contaminated device.
    2. WRT the statement "However, as the colonization of the implant cannot occur with having been implanted, there is a temporal relationship (i.e. AFTER) with the Implantation procedure" the article appears to challenge this: "The majority of PJIs occurring within 1 year of surgery are initiated through the introduction of microorganisms at the time of surgery. This can occur through either direct contact or aerosolized contamination of the prosthesis or periprosthetic tissue. Once in contact with the surface of the implant, microorganisms colonize the surface of the implant". 

    1. Jim Case

      Bruce,

      I am unclear what your conclusion is for item 1.  Are you saying that disorder of prosthetic joint should be restricted to condition 3. of the types of issues around prosthetic devices?  That "Disorder of prosthetic joint cannot be made fully defined? 

      WRT item 2.  Regardless of the time of inoculation, colonization occurs after the procedure has ended (it takes time, such as up to a year).  What the potential is from the procedure is the inoculation, but not the colonization.  It is the colonization that results in the periprosthetic infection, no? 

  2. Jeff Pierson

    Hi Jim and Bruce,

    I'm late to the discussion, so I hope this is helpful.

    1. I agree with Bruce that I would suggest treating a periprosthetic joint infection as a complication, even though it technically is a complication of the source of the infection.
    2. I'm concerned that in the Modeling Approach above the New inferred form tracks to postoperative infection, since "postoperative" implies the infection is temporally related to the operation. This would seem to be too narrow to me.
    3.  In order to move out from under the modeling relationship to "Complication of procedure (disorder)", have we discussed moving "Complication associated with device (disorder)" to a parent-child relationship with "Complication (disorder)"?  This would likely require a look at "Disorders of musculoskeletal implants and repairs (disorder)" as well.  Such a change would allow for modeling with or  without specified temporal relationships to a procedure or operation.

  3. Bruce Goldberg

    Jeffrey, The perioperative complication hierarchy includes preoperative and postoperative complications implies a temporal relationship to surgery but also a causal relationship to some other factor which could be the operation but does not necessarily need to be so. As an example, a periprosthetic joint infection that occurs several years after surgery but is due to hematogenous spread of a microorganism introduced by some other mechanism. I think postoperative complication is, therefore, possibly OK.  My concern is whether the infection can be established not only after the original implant procedure but during as well as per my reference above. I did have an additional attribute as part of the revised associated with hierarchy called during and/or after but this was retired because at the time I felt that a complication of a procedure would always occur during and/or after the procedure, that this would represent the default temporal relationship if one was not explicitly stated. 

    In terms of Complication associated with device (disorder), I would advocate splitting this apart to a) Complication of device insertion or implantation and b) Complication due to presence of device. There also many device malfunctions, failures, etc. which need to be moved to a new hierarchy of Device problems. I have presented this topic in Brartislava and it is undergoing further testing.

    Bruce

  4. Bruce Goldberg

    Jim, perhaps for my first comment, I may have misunderstood your intent when you stated that Disorder of prosthetic joint was redefined but may need to be reevaluated as to whether wa disorder of an implant is in itself a complication. I was just trying to assert some of the proposals that were made regarding device complications in general and that I agree that prosthetic infections are complications whether they are related to the insertion procedure or to the presence of the device. With regard to my second comment, I will defer to your superior knowledge of microbiology as to the time course of colonization although I guess there could be rare cases where a contaminated (already colonized) implant could be inserted. 

  5. Jim Case

    The latter situation you refer to would result in jail time...so I don't think we would want to cater to that.


  6. Bruce Goldberg

    Or at least a big lawsuit (smile)

  7. Paul Amos

    Hi All,

    I apologize that I am rather late to this discussion, but I find it very interesting and wonder whether with some thought and initial increase in complexity that the model could be used to cover a much wider group of implanted devices.

    From a patient management perspective I think it is important to record the underlying cause of the infection as this is often associated with the infecting microbial agent, and thus the treatment given. The classification of the 'infections' is two dimensional:

    Table 1. Classification of prosthetic joint infections

    ClassificationCharacteristic
    According to the route of infection
    		 
    Perioperative
    		Inoculation of microorganisms into the surgical surgery or immediately thereafter
    Hematogenous
    		Through blood or lymph spread from a distant focus of infection
    Contiguous
    		Contiguous spread from an adjacent focus of infection (eg, penetrating trauma, pre-existing osteomyelitis, skin and soft tissue lesions)
    According to the onset of symptoms after implantation
    		 
    Early infection (< 3 mo)
    		Predominantly acquired during implant surgery or the following 2 to 4 d and caused by highly virulent organisms
    Delayed or low-grade infection (3–24 mo)
    		Predominantly acquired during implant surgery and caused by less virulent organisms
    Late infection (> 24 mo)Predominantly caused by hematogenous seeding from remote infections

    Incorporating this type of classification into the model would allow us to apply it to other devices such as dental implants in which a number of implant failures result from either pre-existant or subsequent infection associated with an adjacent tooth. Similarly cardiac pacemaker implants suffer a similar bacterial adhesion to the device surface and subsequent production of a bacterial protective bio-film as do prosthetic joints. I am sure that a similar pathology occurs with intra-cardiac devices and catheters of all types.

    For those interested in the science behind this please see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3568104/ apologies if this has already been considered and rejected. If not then it would be both helpful and academically pleasing to have produced a model which has wide applicability to a broader range of implanted devices.

    Is this worth consideration or a complexity beyond where we wish to be at this stage in our evolution?

    Best wishes and happy New Year.

    Paul

    1. Jim Case

      Paul,

      I think one of the things we need to focus on is the ability of clinicians to record what they see as opposed to creating definitional concepts for things that do exist, but are very hard to determine in real life.  For example, all of the situations you describe above do, in fact, exist, but the ability of a clinician to determine which of these actually represent the clinical state of the patient may not (exist).

      One of the problems we have, even with your detailed description is the use of the word "predominantly".  As SNOMED logical model is dependent on always and necessarily true relationships, how would we represent things that are predominantly occurring?  The ability to qualify our confidence statements about a concept is not currently available and would need to be handled in the information model, if available. 

  8. Bruce Goldberg

    The current content in SNOMED CT involving infected devices, implants, prostheses does not reflect this degree of granularity. My inclination would be to hold off on further complicating the model until such content is requested.