As medicine evolved as a science in modern times, evidence-based practice assumed a prominent role in day-to-day patient care. Today, for most common and well-known diseases, physicians are expected to prescribe cost-effective exams, to use safe and effective drugs and to follow up patients managing risk correctly. [1]

On one hand, evidence-based medicine de-emphasizes intuition and unsystematic clinical experience. On the other hand, clinical pathways play an important role to translate best-practices into physicians' daily practice. [2] However, even the widely accepted best-practices are not completely followed in healthcare settings. [3]

Research barriers, lack of resources, lack of time, inadequate skills, and inadequate access, lack of knowledge and financial barriers [4] are found to be the most common barriers to evidence-based medicine. In this article we explore another one: best practices are translated into clinical pathways in a random fashion.

Multiple languages to represent protocols

Clinical pathways are mostly created in a form of diagrams, text documents or tables. [5] It’s true that the mentioned approach allows capturing a wide range of relevant information, but on the other hand, it compromises understanding.

Recently, researchers started to apply generic modeling languages, such as UML activity diagrams, Event-driven process chain (EPC) or Business Process Modeling Notation (BPMN), as well as domain specific process modeling languages, in order to formalize the representation of clinical pathways. However, none of these languages sufficiently covers the requirements of clinical pathway models, and the choice of a suitable modeling technique remains a problem. [5,6]

Best practices are translated into clinical pathways in a random fashion

Imagine driving in a highway where each county has its own road and traffic signs. How difficult would it be to find your way safely? The same is true for cinical pathways, it is a challenge today to find a line of reasoning clear enough to make you feel safe and confident with your patients care journey. That’s exactly what happens if one (be that a physician, a hospital, medical society, etc.) designs an algorithm with total degrees of freedom. Talk about reinventing the wheel...

Apart from being paradoxical, standardizing medicine without a universal and intuitive language leads to several problems:

1) Complexity

Disease or specialty-specific diagrams can be complex to represent due to multiple artifacts invented, even digitally. The more ways to represent a particular action, the more complex an algorithm turns.

2) Ambiguity

Differently designed algorithms have the risk of being ambiguous, driving to different interpretations, depending on each professional, which can result in patient harm.

3) Specialty-specific

Having specific notations for each specialty blocks algorithm matching, makes it impossible to represent and follow the entire patient's journey, affecting patient-centered care/integrated care.

4) Time-consuming learning curve

Newcomers must be constantly learning new representations of protocols.

5) Lack of collaboration

All the above-mentioned problems create silos and isolate potential contributors to these algorithms.

Problems with process-oriented notations

1) Business Process Modeling Notation

BPMN is an established standard for business process modelling. It is, in the broadest sense, a notation which enables the description and a relatively easy graphical imaging of complex processes. Mainly, its use is focused on industrial, services, and generic businesses. However, in recent years, it has begun to optimize for clinical processes.

Previous studies found that BPMN is sufficiently suitable for planned modelling and imaging of clinical pathways. Authors wrote: “The application in medicine is new, and transfer from the industrial process management is in principle possible.” However, they also underline additional issues:

  • It requires high amounts of manpower and time to be implemented; [7]
  • The integration of a modelling language represents a sensible approach for the development of new hospital information systems; [7]
  • Lack of available technological resources; [8]
  • Poor training in BPMN; [8]

2) Unified Modeling Language

UML is a standardized modeling language consisting of an integrated set of diagrams, developed to help system and software developers to specify, visualize, construct, and document software systems. It comprises a set of tools for documenting the analysis of a system, generally used to describe and evaluate the functioning of complex systems.

There are clear benefits, especially in terms of clarity of communication and repeatability, if a standardized and rigorous notation is employed broadly to represent healthcare processes. However, the application of UML to the modelling of healthcare systems is not as prevalent as in other application domains. There are some studies on UML applied to represent clinical guidelines and protocols, but no evaluation of the benefits was reported. [9]

In general, research find that: [10]

  • It is difficult to learn UML and its notations;
  • There is more than one way to represent something;
  • There are many types of relationships and several diagramming techniques;
  • UML can be confusing since it has many concepts and constructs;
  • A particular diagram might become overwhelming or overcomplicated;
  • It takes a long time to manage and maintain UML diagrams;

The solution: A new open notation for Medicine

In other industries, notations have already been developed and disseminated but that not true in Medicine. Or rather, that was not true in Medicine, until UpHill let a medical notation open to everyone.

From endless medical languages to UpHill Notation
We have already mentioned some major disadvantages of having multiple ways to represent medical algorithms and our approach to this problem – UpHill Notation. In this article, we detail how it works.

Making pathways easier to implement is our priority. Find more about our process to design and implement clinical pathways.

References:

[1] Guyatt G, Cairns J, Churchill D, et al. Evidence-Based Medicine: A New Approach to Teaching the Practice of Medicine. JAMA. 1992;268(17):2420–2425. doi:10.1001/jama.1992.03490170092032

[2] Rotter T, de Jong RB, Lacko SE, et al. Clinical pathways as a quality strategy. In: Busse R, Klazinga N, Panteli D, et al., editors. Improving healthcare quality in Europe: Characteristics, effectiveness and implementation of different strategies [Internet]. Copenhagen (Denmark): European Observatory on Health Systems and Policies; 2019. (Health Policy Series, No. 53.) 12. Available from: https://www.ncbi.nlm.nih.gov/books/NBK549262/

[3] Braithwaite Jeffrey. Changing how we think about healthcare improvement BMJ 2018; 361 :k2014

[4] Sadeghi-Bazargani H, Tabrizi JS, Azami-Aghdash S. Barriers to evidence-based medicine: a systematic review. J Eval Clin Pract. 2014 Dec;20(6):793–802.

[5] Shitkova M, Taratukhin V, Becker J. Towards a Methodology and a Tool for Modeling Clinical Pathways. Procedia Computer Science. 2015;63:205‑12. DOI: 10.1016/j.procs.2015.08.335.

[6] S. Bernardi, C. Mahulea and J. Albareda, “Toward a decision support system for the clinical pathways assessment,” Discrete Event Dynamic Systems: Theory and Applications, 29(1):91–125, March 2019. DOI: 10.1007/s10626-019-00279-9.

[7] Scheuerlein H, Rauchfuss F, Dittmar Y, Molle R, Lehmann T, Pienkos N, Settmacher U. New methods for clinical pathways-Business Process Modeling Notation (BPMN) and Tangible Business Process Modeling (t.BPM). Langenbecks Arch Surg. 2012 Jun;397(5):755-61. doi: 10.1007/s00423-012-0914-z. Epub 2012 Feb 24. PMID: 22362053.

[8] Ramón Fernández A, Ruiz Fernández D, Sabuco García Y. Business Process Management for optimizing clinical processes: A systematic literature review. Health Informatics Journal. 2020;26(2):1305-1320. doi:10.1177/1460458219877092

[9] Vasilakis, C., Lecnzarowicz, D., & Lee, C. (2008). Application of Unified Modelling Language (UML) to the Modelling of Health Care Systems: An Introduction and Literature Survey. International Journal of Healthcare Information Systems and Informatics (IJHISI), 3(4), 39-52. doi:10.4018/jhisi.2008100103

[10] Siau, Keng & Loo, Poi-Peng. (2006). Identifying difficulties in learning UML. IS Management. 23. 43-51. 10.1201/1078.10580530/46108.23.3.20060601/93706.5.