Development and application of methods for Medical Decision Making

The research within our LUMC departments is conducted within departmental research programmes. The research programme below is embedded within the department of Biomedical Data Sciences.

Aim and Focus 

The Medical Decision Making research program aims at improving quality of care, by improving not only decision-making in health care but also care processes, and by developing and evaluating methods for quality of care research.

The following major research themes can be discerned in our research: 

  • A. Modelling: decision analysis, prediction modelling, cost-effectiveness analysis, operations research. 
  • B. Patient-level decision-making: doctor-patient communication, patient preferences, shared decision-making, self-management, and eHealth
  • C. Processes and outcomes of care: patient safety, patient reported outcome measures (PROMS), quality of care, and (de-)implementation. 

Within the context of these themes both applied (clinical) and methodological research is carried out. The latter involves both the development and evaluation of research methods. 

The research program is highly interdisciplinary, and is positioned at the cross-roads of decision analysis, health services research, quality of care research, (decision) psychology and communication research, epidemiology and statistics, information technology, and medical ethics. 

The multidisciplinary nature of the research is reflected in the backgrounds of the faculty; health sciences,  (clinical) epidemiology, quality of life research, implementation research, health economics, and psychology. The group has close collaborations with many departments at the LUMC, as well as with both national and international research groups. 

Position in international content

The highly multidisciplinary nature of the research group is a key contributing factor to the group’s innovative and successful research, and is unique, both nationally and internationally. The group is recognized both within the (inter)national medical decision making and prediction modelling research community, as in the neighbouring fields such as health services research, eHealth, implementation and communication research. Practical medical or health care problems are addressed in e.g. oncology, surgery, pulmonology, diabetology and sexology. Members of our group are active in Programs of the European Union, such as Horizon2020 (e.g. MyAirCoach and Power2DM). In quality of care/patient safety research the LUMC has an important role, through members of our group, also linking to international initiatives such as Dr. Foster Global Comparators. 

Content/highlights/achievements 

The Medical Decision Making research program has three major research themes: Modelling, patient-level decision-making and processes and outcomes of care. We list our research below in more detail, with its relation to these 3 themes.

A1. Decision analysis, cost-effectiveness analysis, operations research (Van den Hout, Van den Akker - van Marle).

Medical decision making takes place in a societal context, with increasing emphasis on the relevance of non-medical outcome measures. Combining such outcome measures is facilitated by modelling, applied at the individual level (decision analysis), institutional level (operations research) and the societal level (cost-effectiveness analysis). Applied analyses are performed in collaboration with many LUMC departments, either alongside clinical trials or using mathematical models to aggregate already available data. Methodological research ranges from pragmatic trial design, measurement and valuation (of health care utilisation, productivity, quality of life, wellbeing), and analysis (disease modelling, multiple imputation), to societal appraisal.

A2. Prediction models to support decision making (Steyerberg, Stiggelbout, Sont)

In medicine, both diagnosis ('what disease is present?') and prognosis ('what will be the outcome of this disease?') involve predictions. Predictions for individuals can be made by statistical models in which patient and disease characteristics are combined. Better predictions can lead to better clinical decisions regarding diagnostic,  and treatment strategies, and finally to better outcomes. In addition, incorporation of information from monitoring patient behaviour and environmental factors by eHealth technologies in short-term prediction models might provide better support for every day decision making in order to improve patient self-management.

A number of ingredients are necessary to make better prediction models. Obviously, we need strong predictors. Many research efforts are currently focused on discovering new risk factors for developing disease and new prognostic factors for outcome of disease, including biomarkers and new imaging techniques. Second, appropriate statistical methods are needed to combine factors into prediction models. When new data become available, specific challenges lie in defining sensible strategies for improving existing models. Third, prediction models are useless if they are not implemented in clinical practice. Hereto, websites and specific Apps and eHealth applications are increasingly available. 

Prediction research requires efforts from different disciplines. First of all, medical professionals have to identify relevant questions, indicate which predictors are relevant in a specific context, and eventually implement prediction models in their daily practice. Further, new predictors will undoubtedly emerge from the basic sciences, including genomics, proteomics, research on biomarkers and sensor technologies. Biostatisticians and epidemiologists should contribute to methodological improvements in prediction modeling. Successful implementation requires support from medical informatics, medical psychology, communication science, and decision making, as available in our research group. With these joint efforts from various fields, prediction research will flourish.

B. Patient preferences, communication, shared decision-making and self-management (Pieterse, Stiggelbout, Sont)

For better decision-making and patient outcomes, doctor-patient communication is germane, also in the context of Value-Based Healthcare. Patient-centred care is the norm nowadays, which often still requires a change it both attitudes and behaviour. Most communication research has not focused on decision-making processes, which is a specialty of our group. Our research on shared decision making (SDM) is internationally recognized. This research relates to aspects of (risk) communication, training and education, patient empowerment, and ethics. Methodological research involves measurement instruments for assessing SDM or determinants or mediators of SDM (such as patient readiness for SDM and self-management). Values clarification is an important focus of attention as well. Patient portals and eHealth applications are expected to facilitate the research on SDM and self-management support by eCoaching, as a means of empowering patients both before and after the consultation, and improving personalized healthcare.

C1. Patient safety, patient reported outcome measures (PROMS) and quality of care (Marang-van de Mheen, Steyerberg, van den Akker, van Bodegom-Vos)

Information on quality of care is increasingly required and used for decisions by different stakeholders. Patients can use this information to decide e.g. in which hospital they want to be treated, insurance companies use this information to purchase care, and providers of care can utilize this information to evaluate and improve their care. Different indicators can be used to measure the aspects of quality of care, e.g. occurrence of complications, mortality or improved functioning due to surgery. These do not necessarily point into the same direction. In addition, a different weighing of the importance of each indicator, e.g. longer survival versus more pain or decreased functioning, may also affect the outcome of the decision being made by different stakeholders.

Within this research area, research efforts are focused on improving our ability to measure the concept of quality of care in a better way, by combining different outcomes that are available, or to look at new dimensions that matter to patients (e.g. improved well-being or PROMs). Related are research questions on whether we can reliably distinguish the performance on these quality of care indicators between providers, but also which factors may explain differences in quality of care between providers or between patients operated in the weekend or during the week. Research on what makes a good performing hospital, and how can we then improve the quality of care within hospitals, is still scarce and likely to advance in the years to come.

Research within the field of quality of care is done in close collaboration with medical professionals, who identify the relevant indicators to define good quality of care but also are the ones improving the quality of care in daily practice for which they require methods to signal whether they have achieved improvement. Furthermore, epidemiologists and statisticians are important for methodological improvements, and it is closely linked to implementation research which is crucial to understand how to build strategies and interventions that will be effective to improve the quality of care.

C.2. (De-)implementation (van Bodegom-Vos, Marang – van de Mheen, Sont, Elzevier)

Every day new evidence, evidence based practice guidelines and innovations (in short: innovations) become available, which in the ideal world should be integrated in existing decision making processes of doctors and their patients. These innovations can include the application of new or improved health care practices, or the abandonment of existing low value practices (i.e. de-implementation). However, these innovations frequently do not find their way to decision making processes in clinical practice. Although the steps to be taken for implementation and de-implementation are the same, de-implementation is not necessarily the other side of the same coin due to specific psychological biases (e.g. confirmation bias, loss aversion). As a consequence, people involved in implementation and de-implementation processes and the factors (at the level of the innovation itself, the individual doctor, the patient, the environment - including the social, organisational, political and financial context) that influence implementation and de-implementation processes may be different. Insight in these factors is of major importance, since these factors determine which implementation activities (such as interactive education including systematic reviews about the evidence, feedback, benchmarking), are needed for the uptake of new or improved health care practices or the abandonment of low value practices in medical decision making. 

Research in this area thus focuses on the exploration of the characteristics of agents involved in (de)implementation (what are characteristics of innovators versus laggards in implementation and de-implementation?), barriers and facilitators for (de)implementation (which factors influence implementation and de-implementation, both innovation specific and more general factors?), to test the (cost)effectiveness of interventions to accelerate (de)implementation of medical practices, and last – but not least- the differences between implementation and de-implementation. Research in this area is performed in close collaboration with health care professionals and patients in such a way that a win-win situation exists, better quality of care for patients and more insight in the (de)implementation issues to advance implementation science. Specific attention will be given to the implementation of sexual healthcare. Not only the general problems of implementation are involved but also specific issues related to the subject itself. Therefore it needs a distinctive approach in the implementation process. 

Funding and output

Our research projects are externally funded by Horizon2020, ZonMw, Dutch Cancer Society, Netherlands Lung Foundation, Kidney Foundation, Citrien Fund, and others. Research is published not only in journals in the field of Health Care Services and Research, but also in major clinical journals (e.g. JAMA, BMJ, J Clin Oncol). Members of our group have been associate editors of journals in the field (BMJ Quality & Safety, Patient Education Counselling, Medical Decision Making). PhD projects cover a range of topics, from methodological to applied, falling under the two research themes described above. Our faculty have a major role in supervising research of clinical departments as second PhD supervisor or co-supervisor. Consultation is also an important part of the work of many of our faculty, particularly for health care efficiency research (ZonMw Doelmatigheid) and implementation research. This regularly leads to invitations to write a workpackage in grant applications from national or international consortia.

The majority of our research findings result in implementation in clinical care or medical education (undergraduate, graduate, and post-graduate), and as such have a major societal impact.

Future themes 

The medical decision making group is well-positioned to perform research with a strong societal impact. Quality of care and value-based healthcare are themes that increasingly receive attention. Research will be expanded into new areas of quality of care and patient safety (incl. health care performance indicators), implementation, risk communication, and patient values clarification and decision support. A specific area is the implementation of sexual healthcare. Doctor-decision making in a multicultural setting will be a future topic that will have both scientific and societal impact, the latter especially for our partners in the Campus The Hague.  Research in these fields will profit from new or intensified collaboration within many partners, NFU Working Parties (e.g. on Performance Indicators, Shared Decision-Making). 

Methodological research will benefit from the closer collaborations that are formed by merging with the Department of Medical Statistics and by the formation of a methodological institute with the Department of Clinical Epidemiology. The group will contribute to the major themes from the National Science Agenda (NWA) and the LUMC “Personalized and value-based care”, not only through our research on prediction modelling and outcomes research, but also through research on implementation, shared decision-making and eHealth.

Cohesion within LUMC 

All topics fall within the broader theme of Value-Based Healthcare, which is the strategic theme of the LUMC in the years to come. All of our research fits the medical research profile Innovation in Health Strategy and Quality of Care. It takes place in close collaboration with many departments within the LUMC, in all four divisions, but in particular with the Departments of Pulmonology, Public Health and Primary Care, Psychiatry, Nephrology, Endocrinology, Neuro-, Orthopedic, Oncologic, Vascular, and ENT surgery, Obstetrics, Gynecology, Urology, Radiotherapy, Medical Oncology. For these and other departments the research group has an important consultation function. We also have close links with registries such as DICA and the implantation registry LROI with a focus on quality of care research.

We carry out research within the LUMC Campus The Hague, the LUMC Center for Oncology and the Leiden University Cancer Center, and the LUMC Center for Personalized Therapeutics. 

The MDM group falls under the Department of Medical Statistics and in the near future will form the Leiden Center for Quantitative Methods in collaboration with the Department of Clinical Epidemiology. 

Key publications 

  • Engelhardt EG, Garvelink MM, de Haes JH, van der Hoeven JJ, Smets EM, Pieterse AH, Stiggelbout AM. Predicting and communicating the risk of recurrence and death in women with early-stage breast cancer: a systematic review of risk prediction models. J Clin Oncol. 2014;32:238-50
  • Stiggelbout AM, Van der Weijden T, De Wit MPT, Frosch D, Légaré F, Montori VM, Trevena L, Elwyn G. Shared decision making: really putting patients at the centre of health care. BMJ 2012 Jan 27; 344: e256. doi: 10.1136/bmj
  • Honkoop PJ, Loijmans RJ, Termeer EH, Snoeck-Stroband JB, van den Hout WB, Bakker MJ, Assendelft WJ, ter Riet G, Sterk PJ, Schermer TR, Sont JK; Asthma Control Cost-Utility Randomized Trial Evaluation (ACCURATE) Study Group. Symptom- and fraction of exhaled nitric oxide-driven strategies for asthma control: A cluster-randomized trial in primary care. J Allergy Clin Immunol. 2015 Mar;135(3):682-8.e11. doi: 10.1016/j.jaci.2014.07.016
  • van Laar JM, Farge D, Sont JK et al. Autologous hematopoietic stem cell transplantation vs intravenous pulse cyclophosphamide in diffuse cutaneous systemic sclerosis: a randomized clinical trial. JAMA. 2014 Jun 25;311(24):2490-8. doi: 10.1001/jama.2014.6368.
  • Vickers AJ, Van Calster B, Steyerberg EW. Net benefit approaches to the evaluation of prediction models, molecular markers, and diagnostic tests. BMJ. 2016 Jan 25;352:i6. doi: 10.1136/bmj.i6.
  • Fischer C, Lingsma HF, Marang-van de Mheen PJ, Kringos DS, Klazinga NS, Steyerberg EW. Is the readmission rate a valid quality indicator? A review of the evidence. PLoS One. 2014 Nov 7;9(11):e112282. doi: 10.1371/journal.pone.0112282. 
  • van Bodegom-Vos L, Davidoff F, Marang-van de Mheen PJ. Implementation and de-implementation: two sides of the same coin? BMJ Qual Saf. 2016 Aug 10. pii: bmjqs-2016-005473. doi: 10.1136/bmjqs-2016-005473.
  • de Groot IB, Otten W, Dijs-Elsinga J, Smeets HJ, Kievit J, Marang-van de Mheen PJ. Choosing between hospitals: the influence of the experiences of other patients. Med Decis Making 2012;36:764-78.
  • Van den Hout W. The Value of Productivity in Health Policy. Appl Health Econ Health Policy. 2015 Aug;13(4):311-3. 
  • de Vries EF, Rabelink TJ, van den Hout WB. Modelling the Cost-Effectiveness of Delaying End-Stage Renal Disease. Nephron. 2016;133(2):89-97.