Other core technical concepts

Health and care records, coding and curation

What are health and care records, coding and curation?

A health or care record is a collection of dedicated information about a person which documents evidence of the person's health or care journey. It is unique to the individual person.

Coding means applying unique and precise ‘codes’ to various aspects of care. Codes are of two types:

    • Classifications - give specific codes to groups of illnesses, symptoms, or procedures. For example, ‘I50.0’ is the ICD-10 code for ‘congestive heart failure’. 
    • Terminologies -  give specific codes for each individual illness, symptom, procedure or medicine.SNOMED-CT is a ‘terminology system’ - it has codes for each and every illness, event, symptom, procedure, test, organism, substance and medicine. For example, ‘15629591000119103’ is the SNOMED-CT code for ‘congestive heart failure stage B due to ischemic cardiomyopathy’ 

Curation of health and care record data involves improving data to ensure that it meets quality standards. For example, it may involve removing errors and inconsistencies, checking for missing values, ensuring consistency in coding and representing the data in a consistent format - e.g. a comma delimited file.

Good quality coding and curation of health and care record data is important to make the data valid and useful for driving decision support systems.

How do I provide evidence of competency in this area? 

Can you...

  • Explain in simple terms to others the concepts of health and care records, data coding and curation?
  • Explain why these are so important to decision support? 

Blooms level 2:  Understand

DDAT Framework roles: Data scientist, Data engineer, Business analyst

Core statistical concepts

What are core statistical concepts in decision support?

Statistics is the science that allows us to collect, analyse, interpret, present, and organize data. It provides a  set of tools for understanding patterns and trends, and making inferences and predictions based on data.

Statistics are inseparable from the reliance on data which drives decision support. They play a crucial role in decision support technology, particularly within the realm of machine learning. 

Some of the ways in which statistics underpin decision support include:

  • Providing the methodologies and principles for creating models in machine learning and other forms of decision support. For instance, linear regression models look at the relationship between two or more variables to enable decision support tools to inform decision support models. Predictive analytics uses historical data to forecast future trends and events and underpins many risk assessment and screening tools.
  • Interpreting results. Statistical concepts allow us to interpret the results generated by machine learning models. Measures such as p-value, confidence intervals, R-squared, and others provide us with a way to measure the machine learning model’s performance.
  • Validating models. Statistical techniques are essential for validating and refining decision support models - for example, techniques like hypothesis testing and cross-validation help us to quantify the performance of models.

How do I provide evidence of competency in this area? 

Can you...

  • List some of the core statistical methods which are key to decision support and explain why they are important?

Blooms level 2:  Understand

DDAT Framework roles: Data analyst, Data scientist, Data engineer

Reasoning methods, Boolean logic and fuzzy logic

What are reasoning methods, Boolean logic and fuzzy logic?

Decision support systems perform reasoning on a knowledge base using different logical schemes.

Reasoning methods include:

    • Inductive reasoning- or "bottom up" reasoning  involves starting with specific observations and using them to draw a conclusion.  
    • Deductive reasoning or "top down" reasoning is a logical approach where you progress from general ideas to specific conclusions. 
    • Abductive reasoning starts with a conclusion or observation. It identifies relevant facts and evidence that to generate possible explanations or hypotheses that could account for the observation. Abductions are made after an event has taken place, and used to hypothesize what probably happened.

Boolean logic is the basis of modern computer systems.

    • It is a mathematical way to figure out the truth of an expression using the simple concept of true or false.  The three basic Boolean operators AND, OR, NOT - underpin mathematical sets and database logic. They are used to either narrow or broaden a set of results.
    • Boolean logic plays a crucial role in driving many decision support systems other digital systems. It manipulates binary values (0 and 1) based on Boolean algebra principles, to perform logical operations and make decisions based on input conditions

Fuzzy logic is an approach to computing based on "degrees of truth" rather than the strict true or false (0 or 1) Boolean logic. 

    • Fuzzy logic includes 0 and 1 as extreme cases of truth but with various intermediate degrees of truth. In this respect it is similar to human reasoning and cognition.
  •  
    • Fuzzy logic is well-suited for building decision support technology for decisions without clear certainties and uncertainties, or with imprecise data - such as with natural language processing.

Henry, 2023-last update, Medium. Available: https://medium.com/towards-data-science/on-ai-and-types-of-reasoning-fc6980295158 [Nov 3, 2023]           

Klein 2022-last update, Codecademy. Available: https://www.codecademy.com/resources/blog/what-is-boolean-logic/#:~:text=Boolean%20logic%20is%20a%20type,AND%2C%20OR%2C%20and%20NOT. [Nov 3, 2023]

How do I provide evidence of competency in this area? 

Can you...

  • Explain reasoning methods, Boolean and fuzzy logic in simple terms to others?
  • Outline why these methods are important to decision support?

Blooms level 2:  Understand

DDAT Framework roles: Data analyst, Data scientist, Data engineer

Decision support delivery mechanisms

What are decision support delivery mechanisms?

Digital knowledge resources and computer-executable knowledge can be embedded in different types of delivery systems - for example:

  • Standalone applications such as websites and mobile apps
  • Integration into external systems, including electronic care record systems, chatbots, robots and many other systems. 

How do I provide evidence of competency in this area? 

Can you...

  • Describe and provide examples of a range of ways in which decision support can be delivered to the end-user, both as standalone applications and integrated with external information systems ?

Blooms level 2:  Understand

 DDAT Framework roles: Data scientist, Data engineer