Data Driven Decision Making for Hospitals Part 4 – Defining the Problem

Understanding the type of problem you are trying to solve is an essential step for developing models of data aggregation and analysis.  As we’ll see below, the type of problem reveals the type of data that the problem solver requires to lower the risk associated with a decision.

Problems can be thought of as puzzles or mysteries. With a puzzle, all the information needed to solve the problem is evident. Often, extraneous information is present that can confound the decision maker.  The boundaries are known, and there is only one outcome.  These are viewed as structured problems.  Typical puzzles that are present in the hospital setting are activities such as scheduling staff. There is a requirement to have “x” amount of workers per shift.  The choice of personnel may be more complex but the requirement remains the same. Another is reimbursement: the conditions treated must match the respective ICD codes.

These problem types are used in teaching in the form of case studies, stories, or rules (algorithms).  The type of data necessary to solve the puzzle is usually related directly to the problem.  These attributes also lead to the use of machine and deep learning algorithms for classification or prediction.

Mysteries, on the other hand, are complex. There is often a lack of information, and there can be multiple “solutions” depending on the view of the stakeholder or problem solver.  Abstraction is a requirement in solving a mystery, an act that’s not possible for algorithms. Problems that fall into this category are strategic problems and dilemmas.  The degree of cognition necessary to solve these problems is typically much greater than for the structured problem.  The data types required to solve the mystery can be either directly related to the problem or be established through indirect relationships.  Decisions such as expansion of services are mysteries: the outcome is unknown until the decision is acted on and measured.

David Jonassen has written extensively on different problem types as well as describing cognitive capabilities required of problem solvers. [Jonassen,D.H.  ETR&D Vol. 48, 2000 pp. 63-85] There are cognitive capability traits that drive proficiency in solving different problem types. These traits include learning styles and degrees of experience, self confidence, and persistence.  Understanding the strengths and weaknesses of the problem solver can be instructive when selecting the best individual or team to address a specific problem type.

The table below lists Jonassen’s taxonomy of problem types.

TypeInputData Characteristics
LogicalPuzzleAll necessary data available to meet conditions
AlgorithmicFormula or procedureAll necessary data available to meet conditions
StoryStory with formula or procedureData constrained by story, known boundaries.
RuleConstrained system, finite rulesData available within the system.
Decision MakingLimited alternativesDerived based on stakeholder requirements.
TroubleshootingMalfunctioning system, fault isolationAll necessary data available – system constrained by design
DiagnosticComplex system with multiple faultsData available, may require sequential collection of information.
Strategic PerformanceReal time, complex, competing needsData collected based on stakeholder requirements
Case AnalysisComplex, no time elementData available, constrained by system design
Design ProblemsVague goals, few constraints, requires structuringData collected based on stakeholder requirements
DilemmasSituation with an anonymous position.Data collected based on stakeholder requirements
Taxonomy of Problem Types

In many cases, a business question can be split into multiple smaller problems of different problem types.  Understanding that the application of critical thinking to the problem will result in a variety of data types necessary to generate the knowledge for decision making.

Here’s an example. Consider the strategic problem statement described for Centerville Hospital: “Centerville Hospital needs to expand its portfolio of services”. This leads to the following  sub-questions:

  1. What are the core competencies of Centerville?
  2. What are the unmet medical needs within the core competencies?
  3.  Is there a partner available that has synergistic needs?
  4. Does the surrounding population support expansion?

The core competency question would be viewed as a diagnostic. The data is resident in the hospital and will lead to a series of clarification statements that will be substantiated by information gathered from the departments and medical staff.  This is a structured problem.

Unmet medical needs speaks to the Strategic Performance of the hospital.  There are limited resources available for development of new programs.  How do those resources benefit the stakeholders of the hospital?  Does the decision maker understand the “wants and needs” of those stakeholders?  The environment is constantly changing as regulations change and competitors launch new products.  Data would be collected both within the hospital as well as from external sources that help frame the key business drivers of the stakeholders.  The amount of information required will be dependent on how the question is framed as well as relative value or “weight” attributed to each attribute considered of value to the stakeholders.  This would be viewed as an unstructured problem

Partners and collaborations represent a potential synergy.  The mapping of core competencies between the hospital and prospective partners makes use of data available from both parties to determine synergies.   The number of available patients that fit criteria limits the decisions. The problem will make use of more abstract thinking as business opportunities are proposed based on an understanding of stakeholder needs.Understanding the problem type leads to determining what data needs to be collected.

Next, we’ll delve into the taxonomy of problem types and associated approaches to find their solutions.

NoviLens is an appliance for data driven decision making. NoviLens handles the tedium of the computing components of the process and frees up the decision maker to focus on their problem solving tasks.

Visit to learn more about NoviLens, and join our email list to get updates on new features and case studies. If you’d like to learn how NoviLens can help your organization, contact us for a 30 minute demo at