Vital Signs

Data profiling is a technique for examining data to check qualities such as accuracy and completeness and to uncover errors in how a source is organized. It is also called data discovery. This post explains the HIMSS Dictionary definition and reviews four sources: a systematic review of EHR data quality assessment, a review of the clinical data quality life cycle, an OMOP mapping project for tuberculosis records in Cameroon, and a method for converting HL7 CDA documents to OMOP.

Data provenance is the ability to trace and verify how data was created, how it moved among databases, and how it was altered over its life cycle. This post explains the HIMSS Dictionary definition and reviews five sources: a systematic review of provenance in healthcare, a scoping review of provenance in biomedical research, a case on understanding EHR data through system flow mapping, a study of provenance and completeness across three data partners, and an analysis of two Kaggle datasets with no clear provenance.

OLAP, or online analytical processing, is a category of tools that help people analyze multidimensional queries, often on top of a data warehouse. This post explains the HIMSS Dictionary definition and how OLAP differs from transaction processing, then reviews six sources: a scoping review of clinical data warehouses, an OLAP-GIS tool for community health assessment, a national outpatient data warehouse, a community health priority tool, a spatial OLAP application for climate-related health vulnerability, and a transfer system built on an OLAP cube.

Advanced analytics is the use of sophisticated techniques, such as machine learning, forecasting, and text mining, to find deeper insights, make predictions, or generate recommendations beyond traditional business intelligence. This post explains the HIMSS Dictionary definition and reviews five sources: a six-part operating model for healthcare analytics, a commentary on governance, a paper on equity in primary care, a multicenter pharmacy study across 11 facilities, and a machine learning study of drug diversion across 10 hospitals.

Data linkage is the merging of information from two or more sources to consolidate facts, and also the technique of connecting records within and between data sets using demographic data such as name, date of birth, address, sex, and medical record number. This post explains the term and reviews five studies: a systematic review of privacy-preserving linkage in US real-world data, a bias framework built on a cohort of more than 131 million people, a Korean claims linkage quality study, a comparison of three methods on 2,247 patients with multiple sclerosis, and a Bayesian de-identified matching method.

A severity system estimates how likely a disease is to progress regardless of treatment, using diagnosis codes such as ICD codes, additional clinical information, or both. This post explains the idea and reviews five studies: APR-DRG severity subclasses that outperformed the Charlson index in 63,681 surgical visits, hourly ICU scores across 561,478 adult ICU patients, an automated VA ICU model built from routine data, sepsis risk adjustment with EHR data, and a study of how many diagnoses are present on admission.

Electronic clinical quality improvement, or eCQI, is the use of common standards and shared technologies to monitor and analyze the quality of health care and patient outcomes. This post explains how eCQMs fit within eCQI and reviews three recent development and testing studies: three neurocritical care measures tested at three sites, a pulmonary embolism measure whose accuracy depended on standardized coding across 38 emergency departments, and two pneumonia antibiotic measures tested in 109 VA systems and two academic centers.

NEMSIS, the National EMS Information System, is the framework for collecting, storing, and sharing standardized emergency medical services data from states across the country. This post explains what it is, who runs it, and what three 2026 studies using NEMSIS data found: national ambulance response times across more than 32 million activations, rural-urban differences in EMS care across more than 45 million responses, and how often weight was documented for children who received weight-based medications.

Data validation is the checking of source data for accuracy and quality before the data are used, imported, or processed. This post explains common check types, how a harmonized framework groups data quality into conformance, completeness, and plausibility, what a systematic review of 29 studies found about the factors behind medical record data quality, and why checks matter most when data move between systems.

Data analytics is the systematic use of data to drive fact-based decisions for planning, management, measurement, and learning. This post explains the four levels of analytics, what a 2025 systematic review found about nurse managers, what a 2024 scoping review found about public health policy tools, and why data quality and governance decide whether analytics helps.

A clinical data warehouse (CDW) groups data from diverse sources about a health system so it can be analyzed for secondary purposes such as quality measurement and research. This post explains how a CDW differs from the EHR, what a Paris pilot found when it automated a stroke quality indicator for more than 6,000 records, what a national rare eye disease warehouse needed to meet strict data protection rules, and why coding alone can misclassify patients.

A registry is a directory-like system that manages data about one kind of entity so that records can be found by predefined search parameters. This post explains how registries inside an EHR differ from disease and clinical registries, what a study of robotic process automation in two cancer registries found, what a 1,076-record COPD reporting analysis showed about documentation gaps, and how automated multi-source registries are being built.

A dashboard is a screen or web page that presents a current, mostly visual snapshot of key facts on progress and performance. This post explains how clinical and executive dashboards differ, what a scoping review found about clinician preferences, what a nursing-leader pilot on controlled substance reconciliation showed, and why data quality and workflow decide whether a dashboard changes anything.

GIGO, garbage in, garbage out, is the principle that the quality of any output depends on the quality of the input. This post explains how it applies to EHR documentation, reporting, research, and AI, what a 2026 systematic review found about factors shaping inpatient record data quality, and what nurses and informaticists can do to protect accurate data.

SQL is the standard language for querying and managing data in relational databases, the engine behind many EHR reports and research data warehouses. This post explains what SQL does, how clinical analysts use it, what two recent studies of AI text-to-SQL tools found, and why SQL injection prevention and least privilege protect patient data.

A digital twin is a virtual representation of a real-world entity or system that can be static, like a simulation, or dynamic, linked to live data. This post explains how twins are built, what reviews and an emergency department study found about hospital use, what the National Academies report says about rigor, and how nurses and informaticists can judge claims.

ETL, short for extract, transform, load, is the process of pulling data from one database, converting it into the form a target needs, and writing it into that target. This post explains the three steps, how ETL differs from real-time interfaces, how the OHDSI community describes designing, mapping, implementing, and quality-checking an ETL, what two studies found when converting EHR data, and where nurses and clinical experts fit.

Atomic data are data elements at the lowest level of detail, captured once at the source and available to be combined, summarized, and reused many ways. This post explains how atomic data differ from free text and summary data, how standards such as FHIR Observation and USCDI depend on them, what research says about structuring the EHR for secondary use, and why nurses and informaticists must agree on what each data point means.

Master data management is the practice of linking an organization’s critical data to a single trusted master record that serves as a common reference across systems. This post explains patient, provider, location, and item master data, the role of governance and data quality, how MDM relates to the master patient index, and why MDM can also mean mobile device management.

A clinical data repository is a structured, centralized store of patient-specific clinical data collected from many applications. This post explains how a CDR differs from an EHR and a data warehouse, the role of terminology and FHIR, governance and data quality, research use, and what nurses should know about the data they create.