CHDA key terms

Analysis that summarizes historical data to show what happened, such as counts, rates, and trends over time. It is the foundation for most routine healthcare reporting.

An approach that uses historical data and statistical models to estimate the likelihood of future outcomes, such as readmission risk or the onset of a chronic disease.

A standardized metric that uses structured data to quantify a healthcare process, outcome, or patient experience for accountability and improvement. Electronic versions are computed automatically from EHR data.

The Healthcare Effectiveness Data and Information Set, a collection of standardized performance measures used by health plans to compare quality of care across many clinical domains.

An approach that manages outcomes for a defined group of people rather than one patient at a time, using data to identify needs, target interventions, and measure results.

Comparing an organization's performance metrics against internal targets, historical results, or external peers to identify gaps and prioritize improvement work.

The process of identifying everyone who affects or is affected by an analytics project, then documenting their interests, influence, and communication needs to guide engagement.

The structured process of gathering and clarifying what users need a data product to do, using interviews, workshops, observation, and review of existing documentation.

A written description of how a specific user will interact with a system or report to reach a goal, including the triggering event, the steps, and the expected outcome.

A comparison of current performance or capability against a desired future state, used to identify missing data, process steps, or resources before work begins.

A mnemonic for writing objectives that are Specific, Measurable, Achievable, Relevant, and Time-bound, helping ensure a business need can be evaluated objectively.

The operational application where data is first captured, such as an EHR, billing system, or laboratory system, which serves as the origin for downstream analytics.

Extract, transform, load: a process that pulls data from source systems, cleans and standardizes it, and loads it into a target warehouse or mart for analysis.

Examining a dataset's structure and content to measure completeness, uniqueness, value distributions, and anomalies before the data is used or integrated.

Documenting how each field in a source system corresponds to a field in a target system or standard vocabulary, including any transformation rules that are applied.

A database that assigns one unique identifier to each patient across multiple systems, linking records that may contain different names, dates, or identifiers.

A centralized repository that integrates data from many sources, is structured for query and reporting, and typically retains historical snapshots for trend analysis.

The process of detecting and correcting inaccurate, incomplete, duplicated, or inconsistent records so that analysis results reflect the true underlying data.

An observation that lies unusually far from the rest of the data, which may signal a data-entry error, a rare event, or a distinct subgroup needing separate review.

A statistical method that accounts for differences in patient illness burden so that outcomes and costs can be compared fairly across providers or health plans.

A measure of the strength and direction of a linear relationship between two variables. It does not by itself prove that one variable causes the other.

A formal procedure that compares observed data with what would be expected under a null assumption, producing a p-value used to judge whether a result is likely due to chance.

Dividing a population into subgroups by characteristics such as age, sex, or diagnosis so that patterns and performance can be examined within each group rather than only overall.

A visual display that consolidates key metrics into a single screen or page, enabling users to monitor performance at a glance and drill into detail as needed.

The use of charts, graphs, maps, and other graphical formats to make patterns, comparisons, and trends in data easier to perceive and communicate.

A confidentiality technique that hides counts or rates based on very few individuals, preventing readers from inferring the identity of a specific person.

Comparing planned, expected, or prior-period values with actual results to explain the size and causes of differences, often highlighting drivers for management attention.

Navigating from a summarized metric to more detailed underlying data, allowing a user to investigate what is driving an observed result.

The assigned responsibility for caring for a data domain, including defining standards, resolving quality issues, and ensuring appropriate use by others.

Descriptive information about a dataset, such as field definitions, allowable values, source, ownership, and refresh dates, which helps users find, understand, and trust the data.

Documentation of where data originates, how it moves and transforms across systems, and where it is used, supporting traceability, impact analysis, and audit.

Agreed characteristics used to assess whether data is fit for use, commonly accuracy, completeness, consistency, timeliness, validity, and uniqueness, against which data can be measured and improved.

A HIPAA principle requiring that access to and disclosure of protected health information be limited to the least amount needed to accomplish the intended purpose.

Removing or altering identifiers so that protected health information can no longer be linked to an individual, using recognized methods such as safe harbor or expert determination.

A controlled list of organizational terms with agreed definitions, owners, and related synonyms, reducing ambiguity when stakeholders request or interpret data.

Policies and schedules that specify how long different data types must be kept, where they are stored, and when and how they are securely destroyed.

A range of values calculated from sample data that is likely to contain the true population value at a stated level of confidence, conveying the precision of an estimate.