Background

The rapidly evolving therapeutic landscape of multiple sclerosis (MS) can make treatment decisions challenging. Novel tools using artificial intelligence (AI) can provide estimations of MS disease progression, which may aid MS therapeutic decisions. However, whether neurologists are willing to utilize information provided by AI-based models when making therapeutic decisions is unknown.

Objectives

To assess whether neurologists rely on clinical judgment (CJ) or quantitative/ qualitative estimations of disease progression provided by hypothetical AI-based models (assuming these models can reliably identify patients at high vs. low risk of disease progression) in simulated MS case scenarios.

Methods

Overall, 231 neurologists with expertise in MS from 20 countries were randomized to receive qualitative (high/low) or quantitative (85-90% vs. 15-20%) information regarding the likelihood of disease progression. Participants were presented with simulated MS case scenarios, and initially made 7 treatment decisions based on the clinical information using CJ. After randomization, participants made 10 treatment decisions using CJ and estimations of disease progression provided by AI models. We evaluated concordance and discordance of therapeutic decisions based on CJ and AI. The primary outcome was the proportion of “optimal” treatment decisions defined as treatment escalation when there was evidence of disease progression or continuing the same treatment when clinically stable. Mixed models were used to determine the effect of randomization group, case risk level, and CJ/AI. Clinicaltrials.gov #NCT04035720

Results

Of 300 neurologists invited to participate, 231 (77.0%) completed the study. Study participants had a mean age (SD) of 44 (±10) years. Of 2310 responses, 1702 (73.7%) were classified as optimal. Optimal decisions were more common for the high-risk vs. low-risk CJ group (84.5% vs 57.6%; p<0.001). There were no differences in the estimated odds of optimal responses between the quantitative vs. qualitative groups (OR 1.09; 95%CI 0.86, 1.39) after adjustment for pre-intervention responses. The estimated odds of optimal decisions for the high-risk vs low-risk CJ group was 2.96 (95%CI: 2.47, 3.56 ) after adjusting for group, pre-intervention responses, and AI-based estimations. For low-risk CJ cases, additional input by AI-based estimations was associated with a lower likelihood of optimal responses; being worse for high-risk vs. low-risk AI estimations (OR 0.235; 95%CI: 0.16, 0.340) adjusting for covariables.

Conclusions

Neurologists were more likely to make optimal treatment choices for high-risk simulated scenarios. The addition of hypothetical information provided by AI-based models- did not improve treatment decisions for low-risk cases. These results provide a framework for understanding therapeutic decision-making in MS neurologists, who are more reliant on their own CJ over AI-based tools.

Background

Therapeutic inertia (TI) is a worldwide phenomenon affecting physicians who manage patients with chronic conditions. Previous studies in Multiple Sclerosis (MS) showed TI affects 60 to 90% of neurologists and up to 25% of daily treatment decisions.

Objectives

To determine the most important factors and levels of attributes associated with treatment escalation in an international sample of neurologists with expertise in the management of patients with MS.

Methods

We conducted an international study comprised of 300 neurologists with expertise in MS from 20 countries (Europe: 59.4%, Asia/Australia: 18.3%, America: 22.3%). Participants were presented with 12 pairs of simulated MS patient profiles reflective of case scenarios encountered in clinical practice. Patient profiles included information on age, sex, previous MS history of relapses, MRI findings, desire for pregnancy, and other relevant details. We used disaggregated discrete choice experiments (a conjoint analysis), which is a standard technique used in economic research to estimate the weight of factors and attributes (e.g. categories) affecting physicians’ decisions when considering treatment selection by asking respondents to choose between pairs of options. In our study, participants were asked to select the ideal candidate (Patient A, B or neither) for treatment escalation (from first-line to second-line therapies- eg. Fingolimod, Cladribine, Monoclonal antibodies).

Results

Of 300 neurologists invited to participate, 229 (76.3%) completed the study. The mean age (SD) of study participants was 44 (±10) years. The mean (SD) number of MS patients seen per week by each neurologist was 18 (±16).

The top 3 factors (relative importance) associated with treatment escalation were: previous relapses (20%), EDSS (18%), and MRI activity (13%). Patient demographics and desire for pregnancy had a modest influence (<3%) in treatment escalation.

Participants were 13% less likely to escalate treatment for patients with EDSS >7.0 (compared to EDSS <6.0), whereas symptom severity during most recent relapse and higher number of MRI lesions at 1 year were each associated with 6% higher likelihood of treatment escalation.

We observed differences in the weight of factors associated with treatment escalation between MS specialists and non-specialists and participants practicing in European vs. non-European countries.

Conclusions

This is the first study applying a conjoint design to assess factors associated with treatment escalation and therapeutic inertia in neurologists caring for people living with MS. Our results provide critical information on factors influencing neurologists’ treatment decisions and should be applied to continuing medical education strategies.