Disease classification for American College of Rheumatology / European Alliance of Associations for Rheumatology
For complex illnesses like rheumatic and musculoskeletal diseases it can be very difficult for clinicians to determine whether or not a given patient is suffering from the disease in question or something else with similar symptoms.
To help them decide, clinicians use decision-support tools consisting of “disease classification” criteria and weights representing their relative importance to classify patients according to whether they have the disease or not.
Since 2008, international teams of experts in specific rheumatic and musculoskeletal diseases have used 1000minds to create a wide range of disease-classification tools.
This large international body of work has been delivered mostly under the auspices of the American College of Rheumatology (ACR) and the European Alliance of Associations for Rheumatology, formerly known as the European League Against Rheumatism (EULAR).
20+ disease-classification tools
300+ clinical expert participants
10,000+ ACR and EULAR members
Complex diseases are often “mosaics” of closely related conditions sharing common features. This characteristic makes it difficult for clinicians to determine if patients are suffering from the disease in question, and to what extent, or something else with similar symptoms.
Since 2008, experts in specific rheumatic and musculoskeletal diseases (mostly members of ACR and EULAR) use 1000minds to, in effect, codify their specialized knowledge and experience by creating disease-classification tools they can have confidence in.
The tools’ validity and reliability have resulted in more accurate classifications of patients with complex diseases.
About American College of Rheumatology / European Alliance of Associations for Rheumatology
Founded in 1934 and 1947 respectively, ACR and EULAR are large global organizations dedicated to reducing the suffering and overall burden associated with rheumatic and musculoskeletal diseases.
Headquartered in Atlanta, USA, ACR serves over 7700 doctors and other health professionals and scientists worldwide who work in rheumatology.
Headquartered in Zürich, Switzerland, EULAR (formerly “European League Against Rheumatism”) represents people with arthritis/rheumatism, health professionals and scientific societies of rheumatology from European countries.
The challenge
When conducting a randomized clinical trial (RCT) into a treatment’s effectiveness, it’s important to be confident that the patients included in the RCT actually have the disease under investigation.
Otherwise, if researchers can’t be sure that all the patents in the RCT have the disease, then how can they accurately measure how well the patients responded to the treatment, i.e. if, in fact, some patients weren’t suffering from the disease? They can’t!
Also, if clinicians are able to agree on the main variables (or criteria) indicating the disease’s presence – instead of a “grab bag” of every conceivably possible variable – then the cost of collecting the relevant data in the RCT will be significantly reduced.
But what are the “right” classification criteria for a disease? And what is their relative importance (“weight”) vis-à-vis each other?
The challenge facing researchers is to create disease-classification tools (criteria and weights) that are valid and reliable, so they can be used with confidence.
The solution
For each disease considered, 1000minds enables international teams of clinical experts to determine valid and reliable criteria and weights for classifying patients.
Participants’ preferences are elicited and quantified by 1000minds asking them a series of simple trade-off questions.
1000minds tools support a comprehensive and iterative process for, first, identifying the classification criteria, and then weighting them (reflecting the experts’ specialized knowledge and experience).
The results
Clinicians are able to trust in the accuracy of the disease-classification tools created using 1000minds because of the software’s validity and reliability, including the award-winning PAPRIKA method for determining weights on the classification criteria.
The ability to accurately classify patients according to whether they have a particular disease or not gives clinicians greater confidence in the disease classification process.
Conclusion
The availability of tools to support clinicians to classify patients with complex diseases (e.g. rheumatic and musculoskeletal diseases, but also others) leads to better decision-making, especially with respect to RCTs (randomized clinical trials).
Basing an RCT on patients who researchers are confident have the disease in question increases the RCT’s power and lowers the cost of data collection.
This large international body of work into creating disease-classification tools (since 2008) is documented in more than 50 peer-reviewed publications, as below.
Peer-reviewed publications
About the use of 1000minds to create disease-classification tools.
H de Lautour, W Taylor, A Adebajo et al (2016), “Development of preliminary remission criteria for gout using Delphi and 1000minds® consensus exercises”, Arthritis Care & Research 68, 667-72
S Johnson, R Naden, J Fransen et al (2014), “Multicriteria decision analysis methods with 1000minds for developing systemic sclerosis classification criteria”, Journal of Clinical Epidemiology 67, 706-14
Another 35+ peer-reviewed articles are available.
