Short Report Multiple Sclerosis Journal— Does attendance at the ECTRIMS congress impact Experimental, Translational and Clinical on therapeutic decisions in multiple sclerosis care? January-March 2019, 1–5 DOI: 10.1177/ 2055217319835226 Gustavo Saposnik , Jorge Maurino , Angel P Sempere, Maria A Terzaghi, Maria Pia Amato and ! The Author(s), 2019. Xavier Montalban Abstract Conferences traditionally play an important role in the ongoing medical education of healthcare professionals. We assessed the influence of attending the ECTRIMS congress on therapeutic decision-making in multiple sclerosis (MS) care. A non-interventional, cross-sectional study involving 96 neurologists was conducted. Treatment escalation when therapeutic goals were unmet and manage- ment errors related to tolerability and safety scenarios of MS therapies were tested using different case-scenarios. Attendance at ECTRIMS was associated with an increase likelihood of treatment esca- lation in the presence of clinical progression (cognitive decline) and radiological activity (OR 2.44; 95% CI 1.06–5.82) and lower number of management errors (OR 0.26; 95% CI 0.07–0.98). Attendance at ECTRIMS may facilitate therapeutic decisions and reduction in management errors in MS care. Keywords: Continuing medical education, management errors, behavioral economics, medical deci- sions, multiple sclerosis, ECTRIMS Date received: 11 November 2018; revised 2 February 2019; accepted: 8 February 2019 Introduction on the management of patients with MS. The aim of Correspondence to: Continuing medical education (CME) is a key part this study was to assess the influence of attending Gustavo Saposnik, Stroke Outcomes and of postgraduate training for healthcare professionals the last ECTRIMS congress on therapeutic decisions Decision Neuroscience (HCP) to gain knowledge that ensures optimal care and management errors by applying principles from Research Unit, St. Michael’s Hospital, University of and outcomes for patients.1,2 Medical conferences Toronto, 55 Queen St E, behavioral economics. Toronto, ON M5C 1R6, traditionally play an important role in the ongoing Canada. medical education of HCP, providing access to saposnikg@smh.ca breaking evidence from around the world.3,4 Methods Gustavo Saposnik, A non-interventional, cross-sectional, web-based Department of Medicine, St. Michael’s Hospital, Toronto, Making therapeutic decisions in multiple sclerosis study using the Qualtrics platform (http://qualtrics. Canada (MS) is becoming increasingly difficult due to the com) was conducted (DIScUTIR MS Study).8,9 The Li Ka Shing Knowledge Institute, St. Michael’s more complicated risk–benefit spectrum of new aim of this study was to evaluate whether neurolo- Hospital, Toronto, Canada disease-modifying therapies (DMTs).5,6 The gists’ risk preferences were associated with thera- Department of Economics, University of Zurich, European Committee for Treatment and Research peutic inertia in MS care. We implemented a novel Switzerland in Multiple Sclerosis (ECTRIMS) is a non-profit approach combining case-vignettes with the assess- Jorge Maurino, University of Buenos Aires, organization created in 1985 to promote research ment of cognitive biases through validated experi- Argentina and learning among professionals involved in the ments in behavioral economics.6,9 The application of Medical Department, Roche Farma, Madrid, Spain management of people with MS.7 At the annual these principles may help overcome those barriers Angel P Sempere, ECTRIMS congress, up to 10,000 participants have by identifying and increasing awareness about cog- Department of Neurology, the opportunity to discuss the latest scientific nitive biases or risk preferences (e.g. overconfi- Hospital General Universitario de research. However, limited information is available dence, tolerance to risk, ambiguity, etc.) that may Alicante, Spain on the impact of attending the ECTRIMS congress lead to suboptimal decisions. A post-hoc analysis Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License (http://www.creativecommons.org/ licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). Multiple Sclerosis Journal—Experimental, Translational and Clinical Maria A Terzaghi, using data from the aforementioned study was per- association between TI score and TI with indepen- Li Ka Shing Knowledge Institute, St. Michael’s formed by comparing therapeutic decisions between dent variables. All multivariable analyses were Hospital, Toronto, Canada participants who attended versus those who did not adjusted for age, level of expertise (specialty, prac- Maria Pia Amato, attend ECTRIMS (exposure). Practicing neurologists tice setting, years of practice), and MS patient NEUROFARBA Department, Neurosciences actively involved in the care of patients with MS volume/week, and reported as odds ratio (OR) and Section, University of from across Spain were invited to participate in the Florence, Italy 95% confidence interval (CI). IRCCS Fondazione Don study by the Spanish Society of Neurology Carlo Gnocchi, (Sociedad Espa~nola de Neurologıa-SEN). Florence, Italy Participants were exposed to 20 simulated MS Results Xavier Montalban, Department of Medicine, St. case-scenarios, three standardized surveys, and four A total of 96 neurologists were included in the study. Michael’s Hospital, Toronto, behavioral experiments to assess aversion to risk and Canada The main characteristics of the study population are Department of Neurology- ambiguity (unknown probability of an event). Of the shown in Table 1. The mean (SD) age was 40 Neuroimmunology, Hospital 20 simulated case-scenarios, seven scenarios were Universitari Vall dHebron, (8.5) years and 51 (53.1%) were female Barcelona, Spain designed to determine the presence of therapeutic neurologists. inertia with evidence of recurrent clinical relapses and radiological progression despite first line thera- pies. Three case scenarios were designed to assess Therapeutic inertia (TI) the appropriate management of side effects of ther- Lack of treatment escalation was detected in at least apies (e.g. transaminitis, lymphopenia, and gastroin- one case-scenario in 68.8% of participants. The testinal side effects). The remaining cases were mean (SD) TI score was 1.5 (1.0). designed to learn about physicians’ therapeutic pref- erences and are not accounted for in this analysis. The multilevel mixed-effects linear regression anal- Further details of the protocol were published else- ysis revealed that participants who attended where.8 Informed consent was obtained from all par- ECTRIMS had significantly lower TI scores (b coef- ticipants and the study was approved by the ficient 0.30, 95% CI 0.59 to 0.015; p ¼ 0.039). institutional review board of the St. Michaels The multilevel mixed-effects logistic regression Hospital (Toronto, Canada). analysis (TI as a dichotomous outcome) revealed that participants who attended ECTRIMS had 70% Study outcomes and definitions reduction (not reaching significance) in TI (OR 0.32; We assessed treatment escalation when therapeutic 95% CI 0.08–1.31). goals were unmet (e.g. clinical and radiological evi- dence of disease progression) as defined in our pre- Finally, the multivariable mixed effects model for vious studies.8,9 We completed two different case-scenarios with progressive cognitive decline analyses: (i) all case-scenarios and (ii) case- plus radiological activity revealed that attendance scenarios having a before and after cognitive testing at ECTRIMS was associated with an increased like- (e.g. a Symbol Digit Modalities Test drop from over lihood of treatment escalation (OR 2.44; 95% CI 60 to 40) showing a progressive cognitive decline 1.06–5.82). There were no differences between plus evidence of disease progression by magnetic fixed- and random-effects models. resonance imaging (e.g. at least five new/enlarging T2 lesions plus one or more gadolinium-enhancing Medical management of side effects of DMTs T1 lesions).8,10 One third of neurologists made at least one manage- The outcome of interest was therapeutic inertia (TI) ment error, whereas 18.8% made two errors out of defined as a dichotomous variable (present if identi- three case-scenarios. The multivariable mixed fied in at least two case-scenarios) and as a contin- effects model revealed the attendance to ECTRIMS uous variable (by the TI score defined according to was associated a lower number of management the number of case-scenarios where participants errors (OR 0.26; 95% CI 0.07–0.98). Figure 1 rep- exhibited inertia).9 A higher TI score indicates resents the predicted probability of management higher TI. errors by ECTRIMS attendance after adjustment for covariates (p-value for interaction ECTRIMS Management errors were tested with tolerability and attendance by management errors: 0.048). There safety scenarios of DMTs (e.g. transaminitis, lym- was no association between participants risk prefer- phopenia, and gastrointestinal side effects).11 Mixed ences (e.g. risk aversion and aversion to ambiguity) effects models were used to determine the with the outcomes of interest. 2 www.sagepub.com/msjetc Saposnik et al. Table 1. Baseline characteristics of participants. Attendees Total at ECTRIMS Non-attendees Characteristics n ¼ 96 n ¼ 56 n ¼ 40 p-value Age (mean  SD), in years 39.5  8.5 39.8  8.5 39.3  8.6 0.78 Age >40, in years 56 (58.3) 24 (42.9) 32 (57.1) 0.83 Gender, n (%) Female 51 (53.1) 32 (57.1) 19 (47.5) 0.35 MS expertise, n (%) 0.003 General neurologist 32 (33.3) 12 (21.4) 20 (50.0) MS specialist 64 (66.7) 44 (78.6) 20 (50.0) Practice setting, n (%) 0.56 Academic 69 (71.9) 39 (69.6) 30 (75.0) Community 27 (27.1) 17 (30.4) 10 (25) Years in practice, mean  SD 14.1  10 14.8  11 13.1  8 0.41 MS patients seen per week, mean  SD 20  15 22.8  21 15.2  13 0.05 Author of a peer-reviewed publication 79 (82.3) 49 (87.5) 30 (75.0) 0.11 in the last 3 years, n (%) Participants’ risk preferences Risk aversiona 26 (27.1) 17 (30.4) 9 (22.5) 0.39 Aversion to ambiguityb 26 (27.1) 15 (26.8) 11 (27.5) 0.94 Numbers between brackets represent percentages, unless otherwise specified. a Participants choose a safe amount of 120 euros or less instead of a 50/50 chance of winning 400 euros. b Participants choose the 50/50 known probability of winning 400 euros over the unknown probability of winning 400 euros. Further details are explained elsewhere.8 Discussion less effective for less severe outcomes than for CME is especially relevant due to rapidly evolving more serious ones.13 knowledge and is a required element of maintenance of certification in most countries.2,4 CME has a pos- CME has evolved from a passive, traditional didactic itive impact on physicians knowledge and perfor- approach to an interactive earner-centered approach mance.3 We found that participants who attended involving new technologies. HCP can now get faster ECTRIMS were 2.5 times more likely to escalate access to the information they need.2 Unfortunately, treatment when there was evidence of disease activ- little data are available about effective educational ity and had a significant lower TI and lower number interventions that target neurologists.1 of management errors. Our study has several limitations that deserve com- Previous studies found that didactic sessions did ment. First, we included neurologists only from not appear to be effective in changing physician Spain, limiting the generalizability of our results. performance in a review of 14 randomized con- Second, we cannot rule out the role of unmeasured trolled studies of formal educational interventions confounders (e.g. infrastructure of centers, differen- including conferences, meetings, and symposia.12 ces in previous medical education, previous partici- Later on, Forsetlund et al. examined the effects of pation in different MS/general neurology continuing education meetings on professional conferences and/or CME resources other than practice and patient outcomes.13 They reviewed ECTRIMS) and possible selection bias to explain 81 trials involving more than 11,000 HCP and our findings. Third, it is possible the presence of found that higher attendance at educational meet- residual confounding despite the adjustment for rel- ings was associated with larger improvements in evant factors and differences in baseline character- clinical practice. However, educational meetings istics. Finally, durability of the educational effect of did not appear to be effective for complex behav- attending this medical conference should be ana- iors compared to less complex behaviors as well as lyzed in future studies. www.sagepub.com/msjetc 3 Multiple Sclerosis Journal—Experimental, Translational and Clinical Non-attendees to ECTRIMS Attendees to ECTRIMS 2 2 number of errors number of errors 1 1 0 0 -1 -1 0 .5 1 1.5 0 .5 1 1.5 Predicted probability of ME Predicted probability of ME Total 2 number of errors 1 95% CI Fitted values 0 p-value for interacon (errors by ECTRIMS -1 aendance): 0.048 0 .5 1 1.5 Predicted probability of ME Figure 1. Predicted number of management errors (ME) by ECTRIMS attendance. Note differences in the slope of ME between attendees vs. non-attendees (p ¼ 0.048). Our study suggests that attendance at ECTRIMS and travel expenses for scientific meetings or participated (the most well attended CME in the specialty) is in steering committees or in advisory boards for clinical associated with improved therapeutic decisions and trials with Almirall, Bayer, Schering Pharma, Biogen, reduction in management errors, confirming the pos- Genentech, Genzyme, GSK, Merck Serono, MS itive role of CME to foster physicians’ knowledge International Federation, National Multiple Sclerosis and performance. Society, Novartis, Roche, Sanofi Genzyme, and Teva. APS received compensation for serving on scientific advi- sory boards or in speaker’s bureaus from Biogen, Bayer, Conclusion Merck, Novartis, Roche, Sanofi Genzyme, and Teva. GS is ECTRIMS and possibly the attendance at other med- supported by the Heart and Stroke Foundation Career ical conferences may play a role as a complementary Award following an open peer reviewed advertisement. strategy to optimize long-term learning of neurolo- He has also received compensation from Amgen gists that may facilitate therapeutic decisions and and Roche. reduction in management errors in MS care. Acknowledgements Funding The authors are most grateful to all physicians who par- ticipated in the study. The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of Conflict of Interests this article: This work was sponsored by the Spanish Society of Neurology (SEN) and partially funded by an The author(s) declared the following potential conflicts of operating grant from Roche Spain. interest with respect to the research, authorship, and/or publication of this article: MPA received research grants and honoraria as a speaker and is a member of advisory ORCID iD boards for Bayer, Biogen, Merck, Novartis, Sanofi Gustavo Saposnik http://orcid.org/0000-0002-5950- Genzyme, Teva, Almirall, and Roche. JM is an employee 9886 of Roche Farma, Spain. XM received speaking honoraria Jorge Maurino http://orcid.org/0000-0001-9858-3555 4 www.sagepub.com/msjetc Saposnik et al. Note 8. Saposnik G, Sempere AP, Raptis R, et al. Decision The abstract of this paper was presented at the 34 th making under uncertainty, therapeutic inertia, and Congress of the European Committee for Treatment and physicians’ risk preferences in the management of Research in Multiple Sclerosis (ECTRIMS) as an eposter multiple sclerosis (DIScUTIR MS). BMC Neurol presentation with interim findings. 2016; 16: 58. 9. Saposnik G, Sempere AP, Prefasi D, et al. Decision- making in multiple sclerosis: the role of aversion to References ambiguity for therapeutic inertia among neurologists 1. Khazanova D and Safdieh JE. Continuing medical (DIScUTIR MS). Front Neurol 2017; 8: 65. education in neurology. Semin Neurol 2018; 10. Benedict RH, DeLuca J, Phillips G, et al. Validity of 38: 479–485. the Symbol Digit Modalities Test as a cognition per- 2. Asch DA and Weinstein DF. Innovation in medical formance outcome measure for multiple sclerosis. education. N Engl J Med 2017; 371: 794–795. Mult Scler 2017; 23: 721–733. 3. Cervero RM and Gaines JK. The impact of CME on 11. Soelberg Sørensen P. Safety concerns and risk man- physician performance and patient health outcomes: agement of multiple sclerosis therapies. Acta Neurol an updated synthesis of systematic reviews. J Contin Scand 2017; 136: 168–186. Educ Health Prof 2015; 35: 131–138. 12. Davis D, Thomson O’Brien MA, Freemantle N, et al. 4. Peck C, McCall M, McLaren B, et al. Continuing Impact of formal continuing medical education: do medical education and continuing professional devel- conferences, workshops, rounds, and other traditional opment: international comparisons. BMJ 2000; continuing education activities change physician 320: 432–435. behavior or health outcomes? JAMA 1999; 5. Comi G, Radaelli M, Soelberg Sørensen P. Evolving 282: 867–874. concepts in the treatment of relapsing multiple sclero- 13. Forsetlund L, Bjørndal A, Jamtvedt G, et al. sis. Lancet 2017; 389: 1347–1356. Continuing education meetings and workshops: 6. Saposnik G and Montalban X. Therapeutic inertia in effects on professional practice and health care out- the new landscape of multiple sclerosis care. Front comes. Cochrane Database Syst Rev 2009; Neurol 2018; 9: 174. 2: CD003030. 7. ECTRIMS. Mission and goals, www.ectrims.eu/mis sion-vision-goals/ (2018, accessed November 2018). www.sagepub.com/msjetc 5