Business Rules and Decision Management

What Are Business Rules and Decision Management?

To meet the demand for enterprise automation over the last few decades, computer programs have become larger and more complex.

To simplify the building and maintenance of those programs, new programming paradigms have evolved. One central method of simplification has been to extract and standardize common computing tasks from the body of application programming source code. One of the first major abstractions was to replace custom data management in programs with Data Base Management Systems (DBMS). The next big movement replaced the step-by-step embedded procedure logic in individual programs with that of Business Process Management Systems (BPMS). Next came the abstraction of business rules which automated the policies by which organizations operated with Business Rule Management Systems (BRMS). The latest evolution is the replacement of BRMS with Decision Management Systems (DMS).

Business rules systems work on the principle of rules statements – a trigger then action paradigm – typically in the If->Then->Else style and usually defined in a proprietary “language” along with decision tables. Those rules are then processed when the trigger occurs by a Business Rules Engine (BRE). The tools to define the rules, version, deploy, execute, monitor, and manage them plus the engine are normally bundled together in a Business Rules Management System (BRMS). These systems are highly technical in nature and are usually owned and maintained by IT resources. Direct business subject matter expert involvement is usually quite limited. Instead, subject matter experts (SMEs) harvest or mine (discover) business rules from within the organization and define them as requirements specifications. IT then translates those requirements into rules statements.

Decision Management Systems work on the principle of defining business decisions where the decision answers a specific business question providing the result (output) of the decision based on the values of pre-defined input variables to the decision logic – typically in the Input(s)->Decision Logic->Answer (output) style. The most commonly used decision management technique is Decision Model and Notation (DMN) an international standard published and supported by OMG, an international, open membership, not-for profit technology standards consortium. DMN focusses on decisions – higher level business asset artifacts – that are meaningful to business subject matter experts (SMEs). DMN also provides the Friendly Enough Expression Language (FEEL), a simple but powerful way to provide logical, textual, mathematical, list processing, interval, date/time and other functions business people need to make decisions.

Decision Management differs from rules management in significant ways. The most important is that rather than just a collection of rules statements, decisions are specific, reusable business logic modules usually created and maintained by business subject matter experts. The DMN standard utilizes visual models which are standardized, verifiable, and where the model serves as the documentation and as the executable source for the Decision Automation Engine. Visual modeling by SME’s eliminates one of the major sources of business rules errors, the “translation” of SME-defined rules (requirements) to IT-created rules. With DMN both the SMEs and IT personnel are working with the same decision model.

Why Are Business Rules and Decision Management Used?

Busines Rules and Decision Management systems are used to free business people and subject matter experts from needing to know and use complex IT programming languages for creating and maintaining the automated policies and logic used to make operational decisions that run their business. By abstracting these rules and decisions from conventional IT application programs and processes, rules and decision systems allow non-IT personnel to create a reusable “single version of the truth” while working directly with existing systems and IT. This abstraction facilitates not only more direct access by business people but also enhances the separation of concerns by making sure the correct people have access to the organization’s policies while providing IT access to the more technical implementation and integration requirements. Business Rules and Decision Management systems abstraction also provides ways to greatly speed up and simplify the creation of and changes to operational decisions resulting in better business agility.

While Business Rules Management Systems continue to be used, their shortcomings have become more pronounced over time. Among these shortcomings are a lack of standardization which forces customers into IT driven vendor-specific proprietary systems and requires higher cost, difficult to find technical personnel to operate. Most recently the advent of cloud computing and customer personalization requirements has disrupted most businesses causing an explosion of rules that has made rules logic more difficult to understand and maintain and that severely restrict agility. Because large business rules repositories contain many thousands of individual rules, they are notoriously difficult to validate and test so heavy step-by-step IT technology involvement is needed and that continuously widens the gap between IT and business subject matter experts.

Decision Management systems based on the DMN standard are, on the other hand, visual model driven, verifiable, standardized, are easily shared between business subject matter experts and IT personnel. A business decision is a reusable tangible business asset that can be automated directly from the visual model without IT translation. Comprehensive visual models in DMN are readable by both IT and business people, serving as a decision specification, the decision logic, the decision documentation, and the automation code – all in a single visual artifact.

Large-scale healthcare organizations are often linked businesses combining and networking physicians, acute care hospitals, long term care facilities and insurance organizations. All these organizations need secure standards-based software to help them store and retrieve data as well as standardize and centralize workflows and decision making across all parts of their business.

Healthcare is also a highly dynamic sector with constant change being generated from regulatory requirements, new procedures, new medicines, new insurance rules and rapidly evolving technologies. Every healthcare organization must be able to quickly set guidelines and policies associated with their data and tailored to their respective delivery systems and business models against which quality of service, cost containment, and patient satisfaction can be directly measured and continuously improved.

Business rules and decision management systems have been used to improve clinical guidelines, care management, prior authorization, billings and payments, off-label prescription policies, fraud management, actuarial and risk management insurance calculations, call center scripting, incorporating AI/Machine learning into decision making, complex process automation pathing and many other applications. Decision Modeling and automation, in conjunction with clinical decision support (CDS), is being used to deliver real-time clinical decision support (CDS Hooks) for providers during their patient encounters. As a result, the providers are able to consider alternative diagnoses, treatments, and potential cost-saving measures for their patients.

Other examples include hundreds of healthcare calculators like BMI, FEV/1, LACE, PEARL, DECAF, Framingham Diabetes Risk, etc. and outside agency preauthorization requirements like Medicare Home Oxygen reimbursement, conventional insurance pre-authorization for Total Knee Arthroplasty (TKA), and internal organizational policies on off-label drug use evaluation and approval. Models like these require multiple decisions, often needing specific patient data. Some of these decisions are based on preset external requirements like Medicare policies and some are specific to the decision-making organization’s policies. One size does not fit all, so the models, to be effective, must be well documented and easy to change to reflect changing policies. Business rules and decision management system solutions allow healthcare organizations to easily define and deploy these examples and many more solutions tied to evidence-based best practices.

Healthcare automation needs have become so critical and pervasive that to address these concerns head-on a community of practice, BPM+ Health™, has been created. BPM+ Health was established based on the use of open, standards-based notations including DMN (decision management) and other open IT standards which allows for all types of health organizations, professional societies, and vendors to document their care pathways and workflows so that they are sharable, discoverable, and automatable.

Trisotech is a founding member of BPM + Healthcare and Trisotech healthcare clientele include international and U.S. acute care hospitals, healthcare insurance organizations, renowned teaching hospitals, and healthcare professional organizations.

The Trisotech Decision Modeling and the Decision Automation Engine provides healthcare providers with centralized, easy-to-understand and change standard process and decision methodologies. These tools adhere to international standards for building processes and decisions and support other standards like FHIR® (HL-7®) for data storage and retrieval, CDS Hooks for real-time decision support at the point of care and PMML execution for the inclusion of AI and Machine learning into automated decision models.

Trisotech also provides nearly 1,000 free pre-built evidence-based workflow and decision models including care pathways, clinical guidelines, and healthcare calculators in the BPM+ Health standard that are being put into practice every day. These models are human-readable, machine automatable, and embeddable in most medical encounter systems. Healthcare organizations can use these models, created under the direction of Trisotech CMO John Svirbely, as they are, or quickly and easily modify them to fit the exact nature of their organization’s policies and procedures. Healthcare organizations can also, of course, create their own processes and decisions from scratch since the easily understood visual models can be built and shared by practitioners, IT and business people.