RITx: DVH600x Data Analytics and Visualization in Health Care

ACO: Accountable Care Organizations. Groups of health care providers who come together voluntarily to give coordinated high-quality care to their Medicare patients.

Acute care: Unplanned event that requires short-term care. Subacute is less critical.

ADL: Activities of Daily Living. Ability to independently perform basic functions such as dressing, grooming, using the toilet, walking/moving, eating.

ADT message: Admittance, Dishcharge or Transfer notification to providers about patient status in an emergency setting. Alerts for provider follow up.

Agile project management: Method for analysis projects involving close customer engagement, prototyping for quick feedback, and adjusting quickly to changes in direction.

AHIMA: American Health Information Management Association. Organization focused on improving health record quality and effective management of health data.

AHRQ: Agency for Healthcare Research and Quality. U.S. agency providing funding for research projects aimed and improving the quality and safety of the health care system.

AI: Artificial Intelligence. Field of computing algorithms and computer technologies applied to perform human-like tasks.

Alert fatigue: When a clinician regularly sees numerous alerts and becomes accustomed to overriding or acknowledging those alerts without much consideration. Critical alerts can be missed. Must balance frequency and necessity of alerting.

AMA: American Medical Association. A U.S. organization of physicians focused on promoting the "art and science of medicine and the betterment of public health" (source: AMA-assn.org).

Ambulatory care: See Outpatient Care.

Analytical dashboard: Type of dashboard that aims to aid decision making with less time-sensitive information. May show trending data over time.

API: Application Programming Interface. Enables software applications to communicate with one another.

Area chart: Visualization that depicts a times-series relationship along with the volume.

ARRA: American Recovery and Reinvestment Act. A 2009 stimulus bill to assist in U.S. economic stabilization by investing in infrastructure, health, and other areas.

Bar/column chart: Visualization that shows change over time, compares different categories.

Baseline: An initial known value which is used for comparison with later data.

Benchmark: Set standards for measures to evaluate organizational performance.

Benefits: Services you can receive as part of insurance coverage.

Best of breed architecture: Separate applications from separate vendors integrated into the core EHR, chosen for improved functionality. Complexities in integration. Different experiences.

BI: Business Intelligence. Software that provides a point-and-click way to build powerful data reporting.

Big data: Data where the volume, velocity, and/or variety is too large for traditional data collection, storage, and analysis methods.

BPA: Best Practice Advisory. A type of clinical decision support alert.

Bubble charts: Visualization that is useful for displaying nominal comparison or ranking relationships.

Business Analyst: Employee role. Combines domain knowledge with analysis skills to recommend business decisions based on information from reports.

Business Intelligence Developer: Employee role. Presents a clearer, humanized view of data sources to enable easier, more efficient reporting for the organization.

CAHPS: Consumer Assessment of Healthcare Providers and Systems. A standardized survey for measuring patients' perspectives on hospital care in the U.S. (communication with providers, understanding of medication instructions, coordination of needs).

CCD/CCR: Continuity of Care Document/Continuity of Care Record. Standards for creating summaries in patient records.

CCDA: Consolidated Clinical Document Architecture. Flexible markup standard that defines the structure of certain medical records.

CDC: Centers for Disease Control and Prevention (part of U.S. Department of Health and Human Services). Focuses on U.S. public health.

CDS/CDSS: Clinical Decision Support (Systems). Sends alerts and pertinent information to clinician to aid in care of patient. Examples: medication and allergy interactions, anomalies, and diagnostic alerts.

Claim/claim form: Documentation of the services provided. Includes demographic, clinical, and cost data. Uses specific code sets. Also referred to as "an encounter".

Clinical practice guidelines: Protocols and treatment frameworks (based on evidence) for clinicians about the care of patients with specific conditions.

CMS: Centers for Medicare & Medicaid Services (part of U.S. Department of Health and Human Services). Oversees many U.S. health care programs and coordinates standards.

Community care: Includes home care services, hospice care, custodial long term care.

Consumer: A person who buys or uses services. Often synonymous with "patient".

Consumerism : How individuals make choices about their health care (where and how they get it, how much they are willing to spend).

Continuous variable: Type of quantitative variable that can be any of the infinitely many values corresponding to the points on a line interval.

Control chart: Visualization that is used to study how a process changes over time (e.g. evaluating process improvement after interventions). Uses historical data to determine central line (average), upper control limit, and lower control limit.

Copay: The amount the individual consumer pays each time they receive a service.

CPOE: Computerized Physician Order Entry. Medications and treatments are entered in the EHR and transmitted to the pharmacy.

CPT: Current Procedural Terminology. Standard wording for health care providers to describe procedures, services, and technologies. Maintained by AMA.

Cross join/Cartesian join: Tables merged with no data columns in common. Each row in the first table is combined with every row in the second table, resulting in many unintended, incorrect rows.

Dashboard: Visualization that uses multiple visualization techniques to summarize the important aspects of data. Shows at-a-glance information.

Data analytics: Drawing conclusions and identifying patterns from raw data through processes of data extracting, transforming, organizing, and modeling.

Data governance: The "exercise of authority and control (planning, monitoring, and enforcement) over the management of data assets" (source: Data Management Association).

Data steward: Individuals or committees attributed to performing data governance activities.

Data/Solution Architect: Employee role. Designs data structures that model or map to business entities, relationships, and processes.

Database: Software systems that store structured data in tables. May have tools to manage and manipulate data.

DBA: Database Administrator. Employee role. Keeps database environments available to users and performing optimally.

Deductible: The amount the individual consumer pays each year before health insurance plan pays for the service.

Deep learning: AI technique to analyze images and provide quick, highly accurate classifications of them.

Demographics: The study of specific populations. Can help determine health care needs and relevant treatments.

Descriptive analytics: Answers "What happened previously?"

DICOM: Digital Imaging and Communications in Medicine. Standards for handling, storing, printing, and transmitting information in medical imaging.

Direct messaging: National standards and services for securely exchanging health care messages. National level address book of providers.

Discrete variable: Type of quantitative variable that can be only a finite or countable number of values.

Disease registry: Collects and studies information about patients with a particular disease to identify regional, demographic, or other correlative factors that can help cure or prevent the disease.

DOB: Date of birth.

Document management: Generates document for electronic signature or scans hard copies. Allows categorization of documents. Helps reduce paper/scanned copies.

DoH: Department of Health. Part of the government which focuses on population health issues.

DRG: Diagnosis Related Grouping. Standardized cost groupings based on age, gender, primary diagnosis, surgical procedures performed, and other factors. Used to facilitate payment.

Drill down report: Visualization that allows users to navigate to a different layer of data granularity. Can be used to analyze the same data through different reports.

E-health: See Patient Portal.

ECG/EKG: Electrocardiogram. A graph of the electrical activity of the heart.

ED: Emergency department (of a hospital) for critical health care concerns. Also referred to as Emergency Room (ER).

EHR: Electronic Health Record. Holistic patient record that spans care settings, enabled by data share and health information exchange.

eMar: Electronic Medication Administration Record. Guides nurses and technicians in the dispensing and administration of medication and treatments for patients.

EMR: Electronic Medical Record. Digital version of a paper-based medical record. Often available to only one provider.

Enterprise architecture: Singular EHR system offered by one vendor which encompasses every possible feature that provider/health system would need. Consistent experience and login. Centralized data.

ePHI: Electronic Protected Health Information. See PHI.

Episode of care: All the care a patient receives from an integrated system of providers for a specific illness, condition, or medical event. May be time restricted.

ETL: Extract, Transform, Load. Process of copying source data to the destination tool for analyzing.

Evidence based care/medicine: The use and application of best practices developed through clinical research.

Expert Determination (HIPAA): A complex set of determinations to declare that individual patients' data is de-identified in an analysis. Appropriate if you need those data elements for interoperability.

Explanatory visualization: Used when you already understand your data and you want to communicate a story you already know. Often of summarized data aimed at conveying a clear message.

Exploratory explanation visualization: A hybrid approach often involving explanatory style visualizations on summarized data but allows users to filter and drill into certain categories. Example: dashboard.

Exploratory visualization: Used where there is lots of data and you are not sure what’s in the data. Can help you quickly recognize patterns, trends, outliers. May have filters and drill down abilities.

Fee schedule: Standard costs used by payers to compensate providers. Base level set by CMS but insurance company may negotiate higher, if needed.

FHIR: Fast Healthcare Interoperability Resources. Standards built off of CCDA and HL7 v2 and v3.

FQHC: Federally Qualified Health Center. Provide primary care services in underserved areas, and receive payment based on ability to pay.

GAO: U.S. Government Accountability Office. Provides fact-based, nonpartisan information to Congress.

GDPR: General Data Protection Regulation. 2016 regulation that dictates standards for organizations collecting and using European Union citizens' personal data.

Genomics: Field of biology concerned with the structure, function, evolution, and mapping of genomes (complete set of DNA).

Geomapping: Visualization that overlays data on a map (e.g. where patient resides and risk level).

HAI: Healthcare Associated Infections. Infections people get when receiving health care.

HCPC: Healthcare Common Procedure Codes. Standard codes used in medical billing to describe items and services. Based on CPT but with an additional level. Maintained by CMS.

Health informatics: Use of technology and systems to collect, analyze, transmit, and share health information with the goal of improving the quality care, reducing costs, and engaging patients.

Health IT standards: Structures for message formatting, data transport, security measures, and data codification to enable different systems to share data.

Heat maps: Visualization that displays categorical data in the form of a diagram or a map, uses intensity of color to represent values.

HIE: Health Information Exchange. The process of moving data across independent systems or the organization that moves data around. Core services are community-wide clinical query portal (patient history), alerting, and direct messaging. May have data repositories and analytic engines.

High signal noise ratio: Compares the level of the important signal to level of the background noise (unimportant). Higher signal levels and lower noise levels are better.

HIMSS: Health Information and Management Systems Society. A global organization focused on better health through information and technology.

HIPAA: Health Insurance Portability and Accountability Act. A 1996 U.S. law that provides data privacy and security provisions for safeguarding medical information.

Histogram: A graphical display of data showing the distribution values. Reveals outliers, minimum and maximum, clusters or groups among values, and completness of data.

HIT: Health IT.

HITECH: Health Information Technology for Economic and Clinical Health. A 2009 U.S. law to promote the adoption and meaningful use of health information technology. Part of ARRA.

HITRUST: Training and certification for security and health information technology.

HL7: Health Level Seven International. A set of standards, formats, and definitions for exchanging and developing electronic health records.

Horizontally merge: Combines two or more data sets using a common key (columns in common). Missed records or large amounts of unintended records can occur if join is not done correctly.

ICD-10: International (Statistical) Classification of Disease, 10th revision. Standard alphanumeric codes for diagnoses, symptoms, procedures, etc. Maintained by WHO.

IHE: Integrating the Healthcare Enterprise. Promotes the coordinated use of established health care integration standards to address specific clinical needs.

IHI: Institute for Healthcare Improvement. An organization focused on improving health and health care worldwide.

Incidence of disease: Number of new cases within period of time. Estimates significance of conditions of short duration, tracks whether prevention efforts are working (decrease), reveals epidemics (spikes).

Inpatient care: For problems that require staying at the place of care for a period of time.

Interoperability: The ability of computer systems to exchange and make use of information.

IOT: Internet of things. Devices connected to the internet.

IQR/OQR: Inpatient Quality Reporting/Outpatient Quality Reporting. Programs run through CMS, includes quality measuring and reporting. Data is shared in the public domain.

IT: Information technology.

Joint commission: U.S. accreditor of health care organizations and programs.

K: Potassium.

KPI: Key Performance Indicators. Strategic values to measure and benchmark.

Latency (in data): Time to write data to storage.

Learning organization: An organization who uses data feedback to improve and is constantly learning. Supports an environment for innovation through an investment in infrastructure and tools, improved leadership, a strong team and data-oriented culture, and employee incentives. Strives to put the patient and community at the center of its mission.

Life expectancy: Statistical average time a person is expected to live.

Line chart: Visualization that shows a time-series relationship with continuous data, can help show trends. May have multiple lines.

LIS: Laboratory Information System. Manages the workflows of the lab from specimen collection to results gathering. Connects to EHR.

LOINC: Logical Observation Identifiers Names and Codes. A health IT standard.

M-Health: Using wireless communication devices to support public health and clinical practice. Patients use digital tools to collect, track, and send data to provider. May include AI-generated alerts.

MACRA: Medicare Access and CHIP Reauthorization Act. A 2015 U.S. law focused on reforming health care payments under Medicare programs though quality-based payments.

Magic Quadrant: Market research reports published by Gartner. Visualized as vendors plotted by completeness of vision and ability to execute, resulting in four quadrants (leaders, challengers, visionaries, niche players). Can help stakeholders make more informed decisions on software solutions.

Managed care: Coordinating the health care experience across functions (financing, providers, and payers) for greater efficiency and integration.

Measures/metrics: Used to quantify important aspects of health care.

MIPS: Merit-based Incentive Payment System. MU was rolled into this program in 2015 as the EHR Incentive Program. Part of MACRA.

ML: Machine Learning. The practice of engaging mathematical modeling techniques against data where the computers actually "learn" via these data.

Morbidity: Frequency of a disease condition in the population.

Mortality: Death rates.

MRI: Magnetic Resonance Imaging. A medical imaging technique.

MU: Meaningful Use. 2009 U.S. legislation that incentivized EHR adoption for eligible hospitals and eligible providers. Three stages of adoption to meet from 2011-2016. Part of HITECH.

NLP: Natural Language Processing. A set of techniques to manipulate and prepare unstructured text for further machine learning analysis. Includes: stemming, word lists, tokenization.

OCR: Optical Character Recognition. Converts scanned document images to typed characters.

ONC: Office of the National Coordinator for Health Information Technology (part of U.S. Department of Health and Human Services). Coordinates efforts to adopt health information technology and improve data exchange.

Operational dashboard: Type of dashboard that aims to answer questions on time-sensitive, critical tasks.

Order: Practitioner instructions for the treatment of a patient (can be written, electronic, verbal).

Outpatient care: For short same day care (you go home that day).

PACS: Picture Archive and Communication System. Capture diagnostic images and manage imaging workflows. Connects to EHR.

Patient identification/matching: The "process of comparing patient information in different health records to determine if the records refer to the same patient" (source GAO.gov).

Patient portal: Secure website or app where providers can share diagnostic results, progress notes, and educational materials with patients. Patients can schedule appointments, message with their provider, and order prescription refills.

Payer/payor: An organization who manages the administration of services and payment. Determines benefit packages and costs. Examples: insurance company, government.

PCP: Primary Care Provider / Primary Care Physician. Person with broadest knowledge of your health care needs. Focused on wellness and disease avoidance.

PHI: Protected Health Information. Individually identifiable health information protected under the HIPAA Privacy Rule. Examples: name, social security number, address, medical history, health insurance information.

PHM: Population Health Management. Collecting and analyzing data on segments of your patient population across their many care settings. Then using that data to improve patient health and provider financial outcomes.

PHO: Public Health Organization. Includes organizations like DoH, CDC, CMS.

Physician office: For non-critical health care concerns. Examples: routine care, some minor procedures.

Pie chart: Visualization that shows the part-to-whole relationship.

PII: Personally identifiable information. See PHI.

PMPM: Per Member Per Month. A measure to represent cost of care.

Population changes: Includes birth, death, and migration (movement between regions). Shifts how and where health services are needed.

Practice variation: Variation in how providers deliver care. Can be from lack of guidelines (warranted) or an intentional disregard of guidelines (unwarranted). Unwarranted variation leads to higher costs and lower quality.

Predictive analytics: Answers "What may happen in the future?"

Premium: Amount charged by insurer to cover anticipated level of risk for medical care. Can be cost shared between employer and individual consumer.

Prescriptive analytics: Predicts outcomes and offers recommendations.

Prevalence of disease: Total number of cases at a specific time. Estimates significance of conditions with long duration. Tracks if treatment programs are working (decrease).

Provider: A person or an organization who provides health care service for payment. Examples: doctors, hospitals.

Public health: A focus on protecting and improving the health of communities or an entire population (e.g. nations, ethnic groups).

Quadruple Aim: An extension to the Triple Aim. One common aim is to improve the work life of the health care providers. Other aims are possible depending on strategy of the organization.

Qualitative variable: Measure a quality or characteristic for each variable or data element. Visualizations should show the values that were measured and how often each value has occurred. Examples: bar/column charts, pie charts, word clouds.

Quality measures/monitoring: Common domains tracked: effectiveness, efficiency, equity, patient centeredness, safety, timeliness. Provides benchmarks, helps reduce variability in care, fuels continuous learning.

Quantitative variable: Measure a numerical quantity on each variable or data element. Visualizations should show trends over time or relationships to another variable. Examples: line charts, scatter plots, bubble charts, heat maps.

RCA: Root Cause Analysis. A method of problem solving used for identifying the root causes of faults or problems.

Regression analysis: A set of statistical processes for estimating the relationships among variables, can show if there is a correlation.

RHIO: Regional Health Information Organization. Multi-stakeholder organization created to facilitate health information exchange.

RIS: Radiology Information System. Capture diagnostic images and manage radiology department workflows. Connects to EHR.

Safe Harbor (HIPAA): 18 identifying data elements that must be removed from the analysis before distributing. Appropriate if you don't need those elements for interoperability.

Scatter plot: Visualization that shows the relationship between two sets of variables, show correlation: positive, negative, or no correlation. Can have a trend line.

SDLC: Software Development Life Cycle. Activities for requirements gathering, solution design, solution testing, and deployment for analysis projects.

SDOH: Social Determinants of Health. Economic and social conditions that influence individual and group differences in health status.

SHIEC: Strategic Health Information Exchange Collaborative. A group of community health information exchanges working on national level data exchange and alerting.

SNOMED: Systematized Nomenclature of Medicine. A health IT standard.

SOC: Service Organization Control audits. SOC2 sets criteria and vets controls of organizational data.

SQL: Structured Query Language. Programming language to perform data analysis tasks.

Stacked bar chart: Visualization that can represent the part-to-whole relationship, compare total values across several categories.

Standard deviation: Used to quantify the extent of deviation of a set of data values. May also be written as "SD".

Statistical relevance/significance: A change or difference that is not attributed to chance.

Stemming: NLP technique of breaking apart text into individual words when spaces or punctuation occurs.

Stop words: Common non-valuable words (e.g. the, an). These are removed during NLP.

Structured data: AKA discrete data. Organized data fields with defined formatting. System enforces formatting. Easier to query from data. Examples: phone numbers, zip codes, ICD10 diagnosis codes.

Supervised learning: Machine learning training where computers learn from data with already known outcomes. New data is mapped using that learning.

TEFCA: Trusted Exchange Framework and Common Agreement. Nationwide sharing initiatives.

Telemedicine: Remote clinical consultation with many specialists. Can share clinical data and images, discuss treatment plans together. Patient does not need to go to many provider locations.

Tokenization: NLP technique of breaking apart text into individual words when spaces or punctuation occurs.

Tree map: Visualization that shows hierarchical data as nested figures (usually rectangles).

Triple Aim: A framework developed by the Institute for Healthcare Improvement for optimizing health system performance. Involves reducing the cost of health care while improving the patient experience (quality, satisfaction) and the health of populations.

Unstructured data: Free text information. Unorganized and not tightly defined. Example: progress notes. Harder to analyze.

Unsupervised learning: Machine learning training where computers learn through detecting patterns in the data. New data is categorized using those patterns.

Utilization: The quantity of health care services used. A factor in determining cost and quality.

Value based reimbursement: Alternative payment methodology focused on quality and improving outcomes. Providers measure and submit performance-related quality data to CMS. Providers receive reduced payment for failure to report or meet quality benchmarks, and bonuses for submitting and meeting quality benchmarks.

Variance: Measures how far a set of numbers are spread out from their average value.

Variety (of data): Many types of data.

Velocity (of data): Rate data is created.

Vertical merge: Data rows are appended into a single data set. Relies on all columns and formatting between the data sets being in common.

Virtual medicine: Physician visits and therapy through a secure internet connection. May use video. Enables reaching more patients.

Virtuous cycle: An exchange of information between practice and research. Cycle is: identify what to study, collect and analyze data, interpret results, communicate to stakeholders, take action, repeat.

Volume (of data): Amount of data.

WHO: World Health Organization. United Nations agency focused on international public health.

Word cloud: Visual representation of text data, where the size of the word/phrase corresponds to the frequency of that word/phrase.