Monday, 28 December 2015

ClinicSpeak: treatment algorithms

Is it time to let AI (artificial intelligence) take-over clinical decision-making? #MSBlog #MSResearch #ClinicSpeak

"Can we manage MS using a simple algorithm or flowchart? The review paper below suggests this is possible. I have always been reluctant to formulate algorithms; in my opinion there are simply too many personal decisions that need to be considered when choosing a treatment strategy and/or DMT to make a easy to use algorithm possible. May be I am wrong? Do you think there is a need for an app, powered by an algorithm, to help you make a decision regarding DMTs? I suspect if the app was intelligent enough it could perform as well, if not better, than a neurologist who may have certain biases when it comes to prescribing DMTs."


Epub: Ingwersen et al. Advances in and Algorithms for the Treatment of Relapsing-Remitting Multiple Sclerosis. Neurotherapeutics. 2015 Dec 23.

Treatment options in relapsing-remitting multiple sclerosis have increased considerably in recent years; currently, a dozen different preparations of disease-modifying therapies are available and some more are expected to be marketed soon. For the treating neurologist this broad therapeutic repertoire not only greatly improves individualized management of the disease, but also makes choices more complex and difficult. A number of factors must be considered, including disease activity and severity, safety profile, and patient preference. We here discuss the currently existing options and suggest treatment algorithms for managing relapsing-remitting multiple sclerosis.

CoI: multiple

5 comments:

  1. Why do you capitalise app? You sound like my dad (who is 70).

    ReplyDelete
  2. While I surely do not have the MS knowledge you folks do I do have more than many folks. I have encountered now well in excess of 1500 people enduring MS. As a software engineer I would state that such an application in my humble opinion could be downright dangerous.

    There are simply too many variables involved with MS not to mention peoples own lifestyles and/or other medications they take. These are personal decisions in a complex disease with many many external factors involved.

    Such decisions and even related advice need be left up to medical health care professionals, not software.

    Of course, none of the meds could be prescribed by an app. But even from a suggestive standpoint its a no no IMHO. Legal implications aside its simply too complex.


    ReplyDelete
  3. GPs use a formula to decide if a patient should take statins. Unfortunately, it doesn't take into account family history. Every tool has its uses, but it needs to be used with caution.

    ReplyDelete
  4. The MS Trust took over responsibility for the MS Decisions website and relaunched it in September 2015.

    We tested the potential for a decision tool which would ask a series of questions and feed responses into an algorithm which would then suggest a shortlist of drugs. Many factors go into someone’s choice about which DMD is best for them, and there are differences in priorities and perspective between the neuro and the person with MS. We concluded that it's near impossible to capture all of this in a decision aid without making it very complex and difficult to future-proof as new drugs become available. We also felt that an algorithm would not be able to take good account of a person’s individual values, preferences and attitudes.

    With help from people with MS, MS nurses and neuros we’ve designed a decision aid which helps people filter and select drugs using different criteria (such as how you take the drug, type of MS it's for), compare up to three DMDs with side-by-side summaries and read more detailed info on individual drugs. You can access the decision aid here: https://www.mstrust.org.uk/understanding-ms/ms-symptoms-and-treatments/ms-decisions/decision-aid. MS Decisions also includes a guide to making your choice and a list of FAQs.

    The intention is to help you compare the DMDs in a simple, intuitive way. We’d welcome any feedback on our approach: email infoteam@mstrust.org.uk.

    ReplyDelete

Please note that all comments are moderated and any personal or marketing-related submissions will not be shown.