Clinical Informatics

American Standard Code for Information Interchange	ASCII	ISO standard, every character is an 8-bit binary
Unicode		Each character is 16-bits (most of world's language)
Hexadecimal	HEX	compact way of relaying binary information (8-bit binary ASCII -> 2-digit HEX)
Bit		single binary digit, also convenient for logical and algebraic information
Byte		8 bits
Digital Logic Gate (transistors)		Digital circuits can mimic Boolean algebra
Boolean operators		AND, OR, NOT, XOR (exclusive OR: true if either A or B is true, but not both), NAND (NOT AND: true unless both A and B are true), NOR (NOT OR: returns true only when both A and B are FALSE)
Half adder		simple digital circuit, most basic boolean algebra example using logic gates: A XOR B -> S (20) A AND B -> C(21). So CS is sum.
Transistors		fundamental building block for electronics (replaced vacuum tubes); silicon semiconductor that makes a switch with no moving parts that can be controlled with a small current. By combining them, resistors, diodes, capacitors, you can build logic gates. Usually prepackaged on integrated circuits (ICs) eg smartphone 100M, PC 1B, each is 22 nm wide
Floating-point Operations Per Second	FLOP	gigaFlop (GFLOP) is billion floating point operations per second (costs 69 trillion times less than 50 yrs ago)
Machine Code		Unique to the CPU, not intended to be human interpretable
Assembly language		Shortcut mnemonics for machine language routines; still tedious to use, not readable
High level programming languages		English-like phrases; must be compiled before execution (souce code -> object code, usually assembly language or machine code) eg C/C++, Java, VisualBasic.
Fourth-generation languages		Further abstract complex tasks (Mathematica, SPSS, MatLab, other task or domain-specific languages)
CPU Time		Performance (seconds/program) = instructions/program * cycles/instruction * seconds/cycle
Reduced Instruction Set Computer	RISC	Reduces clock cycles per instruction (shorter instructions, but more of them); instructions are fixed in length
Complex Instruction Set Computer	CISC	Reduces # of instructions per program (longer instructions, variable length, but less of them)
RISC vs CISC		RISC programs are longer and so more RAM needed to store assembly language instructions; compiler has to do more work to convert higher level languages into shorter RISC insturctions; hardware optimization: fewer transistors, performance enhancements like pipelining, hardwiring for maximum speed. These days the difference are blurred. Low-end / mobile systems tend to have RISC processors (ARM) because of better battery life but here are high-end examples (MIPS, SPARC, IBM Power PC); original X86 Pentium processors were CISC but modern X86 processors eg Intel Core i7 have features of both RISC and CISC)

Architecture		methodologic development of models, specs, and guidelines of IT systems; focuses on the system as a whole; requires governance, innovation, IT strategy, technology standards, defined data and information flows
Systems architecture		conceptual model that defines the structure, behavior and views of a system. Has three design types: Conceptual (high-level, how components and pieces fit together); Logical (conceptual + all major components and relationships and data flows); Physical.
Multi-tier architecture		data-management, application processing, presentation processes are logically separated into different tiers, eg MVC
Unified Modeling Lanaguage	UML	standard toolset for describing aspects of databases, software, and business processes. Includes graphic notation techniques to help visualize models for object-oriented software-intensive systems
UML diagrams		Class (describe object-oriented classes, attributes, methods, inheritance); Activity (show process flowchart, stepwise description of decisions, consequences); Use Case (show actors, goals dependencies); Entity-relationship (ER, describe objects and their relationships; can be used to define RDBMS logical schema)
Service-Oriented Architecture	SOA	paradigm for organizing and utilizing distributed capabilities that may be under the control of different ownership domains; provides a uniform means to offer, discover, interact with and use capabilities to produce desired effects consistent with measurable preconditions and expectations. Has a service consumer, a service provider, and a service directory.
Universal Description, Discovery and Integration	UDDI	XML-based standard for describing, publishing, and finding web services; defines specifications for a web service registry
Web Services Description Language	WSDL	describes a web service's functionality using an XML interface-based language
Simple Object Access Langugage	SOAP	lightweight protocol for the exchange of information in a decentralized, distributed environment. Defines envelope, encoding rules, communication styles

Networks

series of nodes connected by communication paths (links)

Security		Threat assessment, asset list, policy, education, technical measures
Security Layers		Physical (intrusion, fire, power, seismic protection); Network (Firewalls, WEP); Social (phishing, malware, spoofing); Software (design, updates, authentication); Data (backup, restore, redundancy)
Administrative controls		Polices and procedures designed to clearly show how the entity will comply with the act, eg Acceptable internet use policy, Password mgmt policy, Use & protection of SSN in clinical research data, BAA, Privacy procedures plan, Internal audits
Physical controls		Physical measures, policies, procedures to protect a CE's electronic information systems from natural and environmental hazzards as well as unauthorized intrusion. Badge based entry, Cameras, Dual lock systems
Technical controls		Firewalls, Encryption, IDPS, NAC, VPN, DLP, Authentication, Authorization, Audit logs, Patching, Up-to-date rules in antivirus
Authentication		any process of verifying the identiy of an entity that is the source of a request or response for information in a computing environment. Can be something you have (eg key, card), something you know (eg password, PIN), something related to who you are (eg fingerprint, voiceprint), something indicating where you are located
Authorization		What you can do (vs authentication: who you are)
Intrusion detection and prevention systems	IDPS	a network level security technical control
Network access control	NAC	a network level security technical control
Data leakage protection	DLP	a network level security technical control
Hacks		Eavesdropping/sniffing/snooping (intercepts communication between two points); Data modification; Identity / IP address spoofing; Password-based attacks; Denial of Service (DoS) / Distributed Denial of Service (DDoS) flood site with traffic so valid users cannot access); Man-in-the-middle attack (middleman captures data between two communicators, can reroute and change); Sniffer attack (software installed to read network packets and data exchanges)

Data		basis of historical record, support communication among providers, anticipate future health problems, record standard preventive measures, coding and billing, provide a legal record, support clinical research
Aspects of data		circumstances of observation, uncertainty, time, imprecision (normal variance) vs inaccuracy (bias)
Coding of clinical data		historically by a Clinical Coding Specialist (CCS); major purpose is reimbursement. Tradeoffs: stadardization of language vs freedom of expression; time to narrate vs code. Other difficulties: creating and maintaining coding systems; structuring coding systems to capture meaning
Data entry		free-form (writing/dictation/typing); structured (menu-driven); speech recognition/scribes
Structured / menu-driven data entry		via mouse, pen, typing. Benefits: data codified for easier retrieval and analysis; reduces ambiguity if language used consistent. Drawbacks: more time-consuming; requires exhaustive vocabulary; requires dedication to use by clinicians
Speech recognition		narration most commonly; many established systems are on the market that operate on front end (used by clinician, instant availability) or back end (process dictations, editing transferred from clinician to transcriptionist). Challenges : output lags behind user input; require area with minimum of background noise and where patient privacy protected; enunciation errors from mispronunciation, dictionary errors from missing terms, suffix errors from misrecognition of appropriate tenses of a word, added or deleted words, homonym errors. Recent review: report turnaround faster, human transcription more accurate slightly; macros and templates improve turnaround time, accuracy, and completeness
Data Flow Diagram	DFD	graphical respresentation of the flow of data through an information system, modeling its process aspects

Models, theories, and practices of human computer (machine) interaction (HCI)		Explanatory theories (explain behaviors through models); Predictive theories (allow system designers to predict how users will interact with the system); Generative theories (provide useful guidelines/principles to system designers)
Human Computer Interaction	HCI	The study of people and computers; combines computer science, behavioral science, psychology, design, human factors analysis, and more. Increasing awareness that HCI limitations are responsible for medical errors; cognitive overload, 'alert fatigue' is one of the 'grand challenges' facing CDS
HCI / Usability-Related Errors		Wrong Patient (user has 2 charts open, enters orders on the wrong one); Wrong Mode For Action (user tries to enter 100mg but accidentally enters 100 mg/kg); Inaccurate Data Display (lab value is truncated in a report causing user to come to incorrect conclusion); Incomplete Data Display (summary view of vital signs only shows last value per shift, user overlooks the max value within that shift); Non-Standard Measurement, Convention Or Term (weight-based medications calculated using metric units but order entry screen shows weight in English units); Reliance on User Recall (vaccine administration documentation screen requires a lot number; lot number is visible on previous screen but not on this one; users mis-type info or type in nonsense data as a result); Inadequate Feedback (user attempts to order med requiring 0.1cc precision and submits incorrect dose becuase system silently rounds dose to nearest 0.5cc; calculation not transparent to user); Corrupted Data Storage (user enters orders in discharge workflow then clicks 'next' but orders are not submitted because user did not click 'sign before proceeding to next step)
Fitt's Law		Predictive Model. Describes how much time it will take to move a mouse to a target area depending on the distance and size of target : T = a + b log2(1 + D/W)
Hick-Hyman Law / Hick's Law		Predictive Model. Predicts user response to hierarchical menus, response to finding correct option among an unfamiliar list (like a non-QWERTY keyboard); expresses user response time as a function of the number of possible responses. T = b*log2(n+1)
Goals, Operators, Methods, Selectors model	GOMS	Predictive model for observing HCI; breaks down a user's ineraction into four main components. Allows observers to measure interactions with litte effort, cost and time; does not account for user unpreditability; tasks requiring problem solving cannot be predicted by GOMS
Keystroke-Level Model	KLM	Predictive Model; a simplified version of GOMS that track time it takes to complete actions. The time to task completion is the sum of the time spent key-stroking, pointing, homing, drawing, mental operator (thinking), system response operator (waiting). Used to anticipate which functions are most amenable to shortcuts and 'hotkeys'; estimates how much time it takes to complete data input using a keyboard and mouse
Buxton's Three-State Model of Graphical Input		Descriptive Model. Exhaustive description of the states and transitions involved in using a mouse. Three stages: 1. Out of range 2. Tracking 3. Dragging
Guiard's Model of Bimanual Skill		Descriptive Model. Hands are not used equally. Each hand, because of single hand dominance, has distinct roles. The model describes, eg why left handed users are ill-served by modern 101-key keyboards. 'Acknowledge' buttons on modal dialogues are on the bottom right
Activity Theory		analysis and design for a particular work practice with concern for qualifications, work environment, division of work; analysis and design with focus on actual use and the complexity of multiuser activity; focus on development of expertise and of use in general; active user participation in design, focus on use as part of design
Information processing theory		humans dont respond to stimuli they process and analyze information they receive; ie mind is like computer
Rehabilitation Act of 1974, Section 508 ammendment		requires all electronic and information technology to be accessible to those with disabilities; applies only to public facing federal sites

Usability Evaluation		Testing (coaching, thinking-aloud, eye-tracking/click-tracking, performance); Inspection (cognitive walkthrough, heuristic evaluation); Inquiry (field observation, focus groups/interviews, surveys, usage logs)
Usability testing		How usable the software is by testing it on users. Coaching method (user asked to perform a task, allowed to ask any questions); Thinking-aloud (user attemps to complete a task and speaks aloud each step he is doing); A/B Testing; Expert Review; In-Person or 'Hallway' Testing; Remote Testing; Heat-Mapping / Eye/click-Tracking; Performance measurement (5-8 user attempts to complete a specified set of takss; evaluator measures performance statistics eg how long, how often, failure rate, recovery rate)
Mental models		explanations of a person's thought process regarding how something in the real world works; a basic way for us to understand the world around us and develop approaches to problem solving and work.
Cognitive Walkthrough		low-fidelity paper models or wireframe (incomplete mockup); will the users try to achieve the right effect eg will use know to enter weight; will user notice correct action is available; will user associate correct action with the effect to be achieved; if correct action performed will the user see progress is being made

Nielsen's Heuristics		View visibility of system status; match between a system and the real world; offer user control and freedom (undo redo exit); maintain consistency and standards; prevent errors (built-in); rely on recognition rather than recall; demonstrate flexibility and efficiency of use; have an aesthetic and minimalist design; help users recognize and recover from errors; offer help and documentation
User-centered Design	UCD	an iterative, multidisciplinary process of product design and evaluation that considers and designs to support people's tasks, skills, abilities, limitations, creativity, needs, and preferences; based on a clear understanding of users and actively involves them or their representatives in the evaluation of products (user testing) and sometimes in their design (participatory design)
Value Sensitive Design	VSD	takes into account the values of the end-users of the product, as well as those who are directly or indirectly affected by the product
NIST 2012 Recommendations		provide standards for testing and validation of EHR usability: 1. EHR application analysis (who are users, what is their work environment? what do they do? what does the interface look like? what mistakes? what evaluation has been done?) 2. EHR user interface expert review (two-person heuristic review); 3. User Interface Validation Test (performance measurement, post-testing interview)

Usability engineering		combines concepts from HCI, usability testing and standards, psychology, human factors, to create, assess, and make recommendations to improve usability
Usability components (Nielsen)		Learnability, Efficiency (speed), Memorability ('get back on the bike'), Errors, Satisfaction
Discount usability engineering (Nielsen)		Method of HCI eval that does not require large number of personnel or budget; minimum of 5 testers perform a modified verison of testing and inspection; modified think-aloud; heuristic evaluation; low-fidelity prototypes to test one process at a time, rapid iterations
Object-Action Interface	OAI	user first selects object then action; vs Action-Object Interface (AOI) the other way

Bidrectional Interface		information flows in both directions; can be symmetrical (information exchanged in both directions is the same eg ADT message) or asymmetrical (information flows in both directions, but type of information is different eg orders/results)
Foundational Systems		ADT, MPI, Registration, Materials mgmt, Workforce mgmt
Admission/Discharge/Transfer	ADT
Master Patient Index	MPI	ensures 1 record for every 1 patient
Departmental Systems		Laboratory (clinical/anatomic path, blood bank), Radiology (PACS, RIS), Pharmacy (Unit dose, Retail), Cardiology (ECG, echo, cath, PACS), Dietary, Neurology (EMG, NCV), Pulmonary (PFT), GI endoscopy
Laboratory Information System	LIS	One of the largest contributors of objective data into an EHR; larger laboratories are highly automated; all laboratories are highly regulated by federal law; interface with EHR via HL7 messages.
Picture Archiving and Communication Systems	PACS	Can accept an association from a fluoroscopic unit with PACS hostname, application entity title, and TCP
Radiology Information System	RIS	handles departmental workflow and the lifecyle of an interpretation of an image
Financial Systems		Facility Billing, Professional Fee, Financial and Strategic Decision Support
Layers of Infrastructure		Power, HVAC; Network, fixed and wireless; Point-of-care devices; EMR with interfaces; Results review, documentation, CPOE; Care delivery, clinical mission, quality
Heating, Ventilating, and Air Conditioning	HVAC
Data Center		dedicated and protected facility with specific requirements of electricity, humidity, and air conditioning. Usually one per campus per institution; houses hundreds of servers, appliances, and disk storage; placed on racks and physically measured in 'rack units'; physically protected, electrically fed
Data Center Classes		tiers reflect differences in protection against loss of services: Tier I (Basic): single path for power and cooling distribution without redundancy (99.671%); Tier II (Redundancy): single path for power and cooling distribution (99.741%); Tier III (Concurrently maintainable): multiple active power and cooling paths, but only one path active; has redundancy (99.982%); Tier IV (Fault tolerant): multiple active power and cooling distribution paths; has redundancy (99.995%)
Data Center Issues		UPS provide 30 minute of backup power; fully loaded rack may weight 2000 lbs: 100 racks require ensuring proper structural integrity; too many physical servers and other equipment --> virtualization (share hardware resources between processes); Expensive to build, hence often outsourced where space and electricity are cheaper; Redundancy in air conditioning, power, networking; Heat - fully loaded rack may consume 20 kW and require 6 tons of cooling. 100 racks require 2MW electricity and water-cooled air conditioning (chillers)
Uninterruptible Power Supply	UPS
Network Design		network switches distribute network serving a workstation at bedside <-- dual-homed (2 network interfaces) switch-routers <-- redundant backbone core switch-routers <-- WAN or server farm (may be far away).
Data closet		room where cabling ends are connected to switches and other pathways; smaller space, typically on each floor which houses networking equipment and cable ends (needs UPS)
Cable plant		topographical layout of physical cables connecting desktops to the equipment in the closets and cables interconnecting closets and the data center
Institute of Electrical and Electronics Engineers	IEEE	example of an SDO
Cabling issues		Closets may not have adequate HVAC; cables are laid, old ones are almost never taken out - a weight issue; fire codes must be followed going floors and buildings; new cables in ICU and OR require utmost caution; cables can be outdated, unable to support higher bandwidth; labor costs of cabling outweighs other hardware and software purchases; security of closets and cables are suspect; closets may be shared
Client Server		desktop client is focused on handling user interaction but server handles data requests
Thin Client		simple user device that runs application software which is connected to powerful server
Application Service Provider model	ASP	Business that provides computer services over the internet
Integrated Systems		Patient data exists in the same database used by all clinical applications (eg VistA) eg LIS as an integrated module/component of the EHR (shares tables); internal LIS is one in which LIS and EHR are both owned and managed by the same health care entity
Interfaced Systems		Data are communicated between separate applications, usually by means of an interface using HL7 protocol.
Interface Engines		aka message broker, application-level router; is a middleware application used to transform, route, clone and translate messages. A HL7 interface engine is an interface or integration engine built specifically for the healthcare industry. Can 'play back' messages when unavailable receiving system comes back online.
Point-to-Point Interfaces		number of connections when every node connected: n(n-1)/2 but with interface engine only need n connections (each to the engine)
Unidrectional interfaces		information flows in one direction only: drive workflow (eg specimen labels); faxes and printed jobs; movement of specimens down a line; can be problematic when broken
Laboratory automation		HL7 interfaces, used to communicate information between two databases; real-time for clinical care activities. For billing and datawarehousing, can send nightly
LIS interfaces		1:M EHRs, 1:M outside LIS (reference lab, outreach lab), 1:M middleware servers (automation line, point of care device, blood typing, autoverification); many lab instruments, billing systems
Automation line		Robotically operated specimen track which moves specimens from point of entry to the instrument that will perform the test: scans specimen label barcode, uses accession # from barcode to query LIS for pending orders, sends the sample to the appropriate instrument for testing
Autoverification		Process whereby computer-based algorithms automatically perform actions on a defined subset of lab results without the need for manual intervention by a medical technologist, laboratorian or pathologist. Results that fall within certain reference ranges may be automatically verified by the middleware; can do 10x as many lab tests with the same number of employees; when it isn't working, volume may quickly outpace staff's ability to keep up
Barcodes		Code used to represent alphanumeric characters, like an accession number or encounter number; can be 'read' by a barcode scanner and deocded into the original data. Advantages: reduces manual typing errors, improves speed of data entry. Strong recommendations supported by guidlines: ensure that the human-readable version of the encoded data is always printed next to the barcode. Linear barcodes are 1D (Code 39, Code 128A, 128B, 128C(double data density - numbers only)), space intensive for the amount of data encoded; damage and misprints can result in substitution errors (data decoded is not what was intended for printing) at an alarmingly high rate (1/88K); no way to know what piece of data was encoded into the barcode (MRN? MPI?). 2D barcodes represented in 2D (stacked 1D eg PDF417; Matrix eg QR code, DataMatrix, Aztec), can encode redudancy and error correction algorithms, data integrity check (128 does too but only 100 possible checksums)
Errors in Barcodes		Unrecognized character error rate (manual: 1/300 keystrokes; 1D barcode 1/90K to 37M; 2D barcode 1/1021)
Application Server		eg Citrix, connects end-user application with EMR
Methods to improve partial interoperability between HIT systems		Interfaces (eg HL7); Communicate results on paper- scan into foreign EMR; Reciprocal access; Embedded applications (take appearance of one application and put them within the EMR of the other organization); Context sharing-CCOW; Build separate application with data from both

Types of settings where systems are used		Ambulatory, surgical center, ER, OR, nursing facility, acute care facility, home, inpatient (acute care, psych, rehab, ICU: trauma/surgical, pulmonary, cardiology, neurology, neonatology, remote)
Clinical Laboratory Improvement Amendments of 1988	CLIA	overhaul of CLIA (1967); generated in reponse to public fury (false neg pap smears); pertains to every laboratory in the US. Requirements for reporting and notification of laboratory results; regulates labs not EHRs which must be certified by CMS and state
CLIA regulations		must have adequate manual or electronic system in place to ensure test results and other patient-specific data are accurately and reliably sent from the point of data entry to final report destination in a timely manner. Test report must indicate patient identification, name/address of lab, date, test, specimen source, test result, units of measurement; reference intervals or normal values must be available to authorized person who ordered test.
CLIA maintenance		every two years (or CLIA-deemed agency); college of American Pathologists accredits most labs in the US (more strict). Gen.48500: there is a procedure to verify that patient results are accurately transmitted
Patient Access Rule		Revises CLIA and HIPAA / supercedes all state laws; patient can request copy of lab results directly from lab (has 30d to provide); requirements to comply with request are the same as for health care entities; labs can refer patient to medical records if all results available and HIPAA compliant; no requirement for lab to interpret or to notify ordering provider
Transfusion systems		Regulated by the FDA; considered high risk (Class III) medical device. Guidance on software validation available
Digital Imaging		Pixel (smallest component of an image; print vs digital); Image Size (overall size of image in its final form, width by height; contains no information on image resolution; important to know if need to decide on resolution)
Image resolution		aka pixel density, commonly referenced as DPI. Gross images --> low resolution OK; microscopic images --> high resolution desired. Balance resolution against the final image size desired (small image size : low res; large : hi)
Dots per inch	DPI	number of pixels per inch per screen/print medium; DPI = 96 on average computer screen; printed DPI = 300 variable
Image Compression		Reduces the amount of memory that an image occupies by various mathematical algorithms; lossy or lossless
Lossless image compression		memory (file size) reduced but original image data can be recovered, eg PNG, GIF
Portable Network Graphics	PNG
Graphics Interchange Format	GIF
Lossy image compression		some original data in the photo is gone forever; amount of compression is proportional to the amount of loss; well designed --> substantial data reduction without obvious loss to the end-user; eg JPG
Joint Photographic Experts Group	JPEG
Tagged Image File Format	TIFF	can use either type of compression
Mutliple-image Network Graphics	MNG	can use either type of compression

Electronic health / medical records systems as the foundational tool	EMR	Evolution from department-focused to patient-focused; tab metaphor for data remains common; goal of problem-oriented medical record remains largely elusive; most visible system to clinicans and patients; target of federal incentive programs
EMR core functionalities (IOM)		health information and data, results mgmt, order entry/mgmt, CDS, e-communication and connectivity, patient support, administrative processes, reporting & population heatlh mgmt
EMR functionality 1		Message box (proprietary names vary but functionality similar); Results review (lab, path, imaging, notes); Documentation (direct entry, structured/unstructured, dictation, mixed); Order mgmt; Patient summary displays; MAR/BCMA
Medication Administration Record	MAR
Bar Code Medication Administration	BCMA
EMR functionality 2		Patient lists, schedule, rounding/handoff tools; patient monitoring review; quality metrics, dashboards; Billing; Patient support; Administrative; E-communication (with team, patients); Population health, external resources, Compliance & Decision Support (permeates all the other functionalities)
HIMSS EMR Adoption Model		commonly used: 8 stages, each has features that add to one below it; different sequences possible
Documentation Using EMRs		Satisfaction = f (time efficiency, availability/accessibility, expressivity, quality)
EMR Issues		Time spent writing notes: 4-14 min; 'electronic notes are harder to understand'; copying and pasting; note loss, notes in wrong chart, notes with wrong title, notes on wrong encounter; billing & compliance
Study on EMRs copying and pasting		1-6 scale of harm/misleading/risk. 1/10 charts contained an instance of high-risk copying; clear policies, practitioner consciousness-raising and devlopment of effective monitoring procedures are recommended
Documentation Tools		Structured (codes, needs training) vs Unstructured (narrative text, easier to learn)
Order Sets		collection of pre-configured orders; time savings; reduce errors and increase accuracy during order entry; increase completeness of orders; 'built-in' decision support and evidence driven care; reduce variability and enhance compliance with best practices. Protocol is built of orders sets is built of preconfigured orders is built from order dialog.

Telemedicine		communication of voice or appearance of patient is important to delivery of care (psych, derm, path, ENT, retinography); teleradiology, telesurgery, remote imaging/monitoring, remote ICU, remote procedures
Synchronous Teleconferencing		Dedicated hardware, broadly available tools (eg Skype, conferencing applications)
Asynchronous Telemedicine		Store & forward; email, other
Telepathology		has a specific meaning compared to all other forms of digital pathology; diagnosis which results in a report is rendered from a digital microscopic image ONLY. Static telepathology : entire image is caputred then transmitted; transmission can be in toto or in pieces (tiling). Dynamic telepathology : live video feed, with or without remote control of scanner. Hybrid telepathology do both
Telepathology regulations		vary from state to state; not the same as for teleradiology; CLIA applies; CAP has additional standards for accredidation; whole slide imaging device are FDA Class III (requires premarket approval)

Standards development approaches (Hammond)		1. ad hoc (groups come together and agree to use a common but informally developed specification eg DICOM); 2. De facto standards emerge as one earns a critical mass of adopters to make its system the standard (eg .DOC format); 3. Mandated standards by government agencies, eg US certificate of death; 4. Consensus, closed or open (preferred)
Standards development process (Hammond)		identification, conceputalization, discussion, specification, early implementation, conformance, certification
American National Standards Institute	ANSI	private nonprofit organization, developed a formal consensus process with open balloting and public review as well as an accreditation program for SDOs; coordinates US standards with international standards so American products can be used worldwide
Health IT Standards Committee	HITSC	charged with making recommendations to ONC on standards, implementation specifications, and certification criteria for electronic exchange and use of health information. Lately : increasing interoperability. 'JASON Group' of scienctists that advises US government - move to more modern API-based approaches; JASON task force developed recommendations: interoperability vision and framework --> Argonaut Project (2014, details of implementation); Standards advisory (2015) deemed ready for use
Standards Development Organization	SDO	has overall responsibility and ownership of a standard; employs a multistep process in producing standards to ensure quality, integrity, and input
American Society for Testing and Materials	ASTM	SDO responsible for many healthcare standards eg the CCR
International Organization for Standardization	ISO	International standard-setting group; ANSI is the US representative
European Committee for Standardization	CEN	CEN/TC 251 is health informatics standards body for Europe
National Institute of Standards and Technology	NIST	Information standards organization (not an SDO, exists to foster, promulgate and coordinate standards)
International Health Terminology Standards Development Organization	IHTSDO	based in Denmark. Maintains SNOMED-CT
Clinical and Laboratory Standards Institute	CLSI	Intl SDO for labs, consensus-driven (reflects equal representation from govt, industry, health care pros)
National Electrical Manufactuers Association	NEMA	maintains DICOM
National Council for Prescription Drug Programs	NCPDP	VD.0, standard for pharmacy claims transactions.
NCPDP Medicaid Subrogation Version 3.0		standardization of the Medicaid pharmacy subrogation transaction process (claim from a Medicaid agency to a payer for the purpose of seeking reimbursement from the responsible health plan for a pharmacy claim the State has paid on behalf of a Medicaid recipient.)
NCPDP Batch Standard Version 1.2		Similar to D.0 but in a batch enviroment

Standards and Interoperability Framework	S&I	standard initiative of ONC
Unique Patient Identifier	UPI	yet to come to fruition despite being mandated in original HIPAA; should be unique, nondisclosing, permanent (will never be reused), ubiquitous (everyone had one), canonical (each person has only one), invariable (will not change over time). There are probabilistic matching algorithms to link records. ?Globally unique identifier eg Record@Hospital : combine local MPI plus holding institution identifier - could be unique, permanent, nondisclosing, ubiquitous; once in place develop a 'Reocrd Locator System' to achieve linkage
Unique Device Identification	UDI	2014-2020 FDA mandate?
National Provider Identifer	NPI	10-digit numeric identifier issued by the NPS (of HHS) replaced UPIN in 2007 and is required for medical billing. 10th digit a Luhn algorithm checksum
National Provider System	NPS	Issues NPI to all US physicians
Universal Physician Identifier Number	UPIN	mandated in HIPAA
National Standard Employer Identifier	EIN
Health Plan Identifier	HPID	mandated in ACA
Other Entity Identifier	OEID	mandated in ACA
Medical Information Bus	MIB	aims to develop standards for control and linkage of information from medical devices; most implementations just transfer data, but there is capability to issue commands eg change settings of an IV pump

Transaction standards		standardized electronic exchanges involving data transfer between two parties for specific reasons
Electronic data interchange	EDI	covers transactions involving claims and encounter information, ERA, claim status, eligibility, enrollment and disenrollment, referrals and authorizations, and coordination of benefits and premium payment
Electronic Funds Transfer	EFT	under ACA physicians can require payers to use EFT. Medicare requires physicians to use EFT
Electronic payment & Remittance Advice	ERA
ASC X12 Version 5010		mandated in HIPAA to encourage electronic commerce for health claims: eligibility, claim status, EFT, ERA
Context Inspired Component Architecture	CICA

Continuity of Care Record	CCR	ASTM patient health summary standard; the most relevant and timely facts about a patient's condition; goal was for use when patient was referred, transferred, or discharged. Not compatible with existing standards so HL7 and vendors created CCD which is richer
Health Level Seven International	HL7	ANSI accredited international SDO with 2,300+ members across 500 corporations representing ~90% of IS vendors. Provides a framework for the exchange, integration, sharing, and retrieval of electronic health data ie major messaging standards.
HL7 versions		V2.1 (1987) through V2.8 (2014), syntactic-oriented, loose guidelines can be built in ad hoc fashion, backwards compatible with other V2 versions but not V3, each message type has a set of segments delimited by | consisting of a three character identifier and values. Version 3 (2005), semantic interoperability, stricter more consistent standards, plug-and-play, built upon RIM in XML, not compatible with V2.x, more expensive
Reference Information Model	RIM	Object-oriented model to define healthcare interactions based on five abstract classes : 1. Entity 2. Role 3. Participation 4. Act 5. Act relationship. All clinical, administrative, financial activities of healthcare can be expressed in 'constraints' to model
Structured Product Labeling	SPL	HL7 XML document markup standard that specifies the structure and semantics of the content of authorized published information that accompanies any licensed medicine
Clinical Document Architecture	CDA	most widely recognized HL7 version 3 component; a standard for specifying the XML-based standard structure and metadata of clinical documents; templated and object-oriented; much healthcare information is in 'documents' required for human reading, but still want computable structure. Templates are reusable, computable components of CDA documents; unstructured documents can be 'wrapped' in a CDA framework. Level 1: specification of document; Level 2: adds document types with allowable structures; Level 3 adds mark-up expressible in RIM
CDA usage		CDA defines building blocks which can be used to contain healthcare data elements that can be captured, stored, accessed, displayed and transmitted electronically for use and reuse in many formats; sets of these CDA standardized building blocks can be arranged for whatever needs exist; this approach offers tremendous flexibility; it allows for the creation of a comprehensive variety of clinical documents which share common design patterns and use a single base standard; arranging (or constraining) the CDA elements in defined ways using templates produces clinical documents
CDA Structure		Header; Body>>Section>>Narrative Block; Entry. Every document must have at least a header (sets context) and one section (contains one narrative block (human readability; rendered for viewing) and 0-many coded entries(machine-readability).
Consolidated CDA	C-CDA	document templates for common clinical notes (discharge summary (CCD), diangostic imaging report, consultation note, H&P, op note, progress note, procedure note, unstructured document), which are based on Section templates and Entry templates
Continuity of Care Document	CCD	HL7 XML specification for structuring in a static form the many data elements that are needed to record clinical encounters; based on HL7v3 and CDA. The core data set of the most relevant administrative, demographic, and clinical information facts about a patient's healthcare. Required to have : US Realm Header, Allergies, Medications, Problem, Results sections in body
Clinical Context Object Workgroup	CCOW	HL7 protocol designed to enable disparate applications to synchronize in real time and at the user-interface level; manages context of caregivers who may need to access different computer applications in process of care; oft-stated goal is 'single sign-on' across applications, network etc.
Electronic Health Record - System Functional Mode	EHR-S FM	outlines important features and functions that should be contained in an EHR system
Digital Imaging and Communications	DICOM	international medical imaging messaging standard for handling, storing, printing, and transmitting across a specified network communications protocol with a specific file format definition; maintained by NEMA
NCPDP SCRIPT	SCRIPT	NCPDP standard for electronic communications between prescriber and pharmacy; MU requires for e-prescribe
CEN ISO / IEEE 1073		standard that dictates how medical, wellness, and healthcare devices communicate with external computer systems: allows fo 'plug-n-play' use of healthcare devices.
EHR-Laboratory Interoperability and Connectivity Standards	ELINCS	goal to standardize laboratory ordering from and reporting to EHRs; relies on HL7v2 and LOINC
Blue Button Initiative		Public-private partnership to empower consumers with easy and secure access to their health records from a variety of sources in a format they can use. Began with the VA --> led to ONC launching Blue Button Toolkit to facilitate development
Substitutable Medical Apps, Resuable Technologies	SMART	an ONC SHARP project : create a platform for app development accessing store of information. Some vendors are pursuing platforms (Surescripts, Allscripts launched open architecture platform HELIOS, EPIC has developed API for accessing data and services)

terminology		set of terms representing the system of concepts of a particular subject field
synonym		different term for same concept
polysem		term that means more than one concept
dictionary		concepts plus meaning
thesaurus		groups synonyms by concept
nomenclature		system of terms elaborated according to established naming rules
vocabulary		concepts and terms in a domain
classification		system that organizes like or related entitties
semantics		insertion of meaning via relationships
term		designation of a defined concept in a special language by a linguistic expression
Nomenclatures, vocabularies, and terminologies		benefits of computerization of clinical data depend upon its 'normalization'. Clinical language is inherently vague, which is at odds with the precision of computers. Use cases: information capture, communication, knolwedge organization, information retrieval, decision support
concept		thing or idea, expressed in one or more terms

ontology		A framework for representing knowledge (web). A controlled vocabulary expressed in an ontology representation language thereby creating a formal representation (structuring of concepts and relationships between them) of definitional information of a domain in a way that allows it to support automatic information processing. The grammar contains formal constraints (e.g., specifies what it means to be a well-formed statement, assertion, query, etc.) on how terms can be used together. Like thesauri but different because there are often no preferred terms and the concepts and relationships are described in machine readable ways thereby supporting semantic interoperability. Generally adheres to either RDF or OWL XML-based standards.
taxonomy		practice and science of classification; taxonomies add structure to unstructured information to make it easy to search and filter. (tree with branches)
Predication relationship		A.1 is a certain type of A (what a statement says about its subject)
Conditional relationship		All A.1 are A
International Classification of Diseases	ICD	international standard and disease ontology published by the WHO originally for morbidity reporting; ICD-10 defined in 1990 (69k codes, 3-7 characters, first character alpha), now standard as of 10/1/15 (ICD-10-CM, maintained by the CDC) ICD-10-PCS procedure codes will be used for intpaitent procedures instead of CPT). Half of all codes related to musculoskeletal system, primarily injuries. A requirement for billing claims. ICD-9 (1975, 14k codes, 3-5 characters) limitations include NOS and NEC codes, lack of granularity
General Equivalence Mapping	GEM	enables mapping, eg from ICD-9 to ICD-10 codes
Diagnosis-Related Group	DRG	describes a bundle of services a hospital might provide; flat rate per case for inpatient hospital care to rewards hospitals for efficiency; original intent was to aggregate ICD-9 codes into groups for health services research; set of several hundred codes 'lump' hospital illnesses; adopted in 80s for prospective payment for hospitalization in Medicare;
CVX Code		codes for active and inactive vaccines in the US
National Drug Code	NDC	FDA required registration that utilizes a 10 digit, three-segment number that serves as universal identifer of the product (pharmaceutical preparation); labeler code - product code - package code
Unique Ingredient Identifier	UNII	specifies ingredients in drugs and other compounds
National Drug File Reference Terminology	NDF-RT	standardized terminology maintained by the VA, for modeling drug characteristics, including ingredients, chemical structure, formulations etc, ie ADMET (absorption, distribution, metabolism, elimination, toxicity)
RxNorm		substance-oriented standardized terminology maintained by the NLM that provides normalized names for clinical drugs and links to synonyms within First Databank, Micromedex, MediSpan, Gold Standard Drug Database, NDF-RT and Multum. The first component of UMLS; provides semantic structure for formulations and their components
Logical Observation Identifiers Names and Codes	LOINC	standard universal codes for identifying laboratory tests and clinical observations in electronic messaging; developed by the Regenstrief Institute. MU requires LOINC in messages reporting lab results, medical summaries, sending data to cancer and public health registries. Each term is assigned a unique identifier and a fully specified name with 1. component (what is measured) 2. kind of property (eg mass, substance) 3. time aspect (eg 24h collection) 4. system type (eg context or specimen type ) 5. type of scale (eg ordinal, nominal, narrative) 6. type of method (eg procedure used to make the measurement /observation)
Current Procedural Terminology	CPT-4	standardized terminology maintained by the AMA; category I: numeric codes for procedures and RVUs; category II; alphanumeric codes for tracking performance measurement 'F' eg 3725F screening for depression); cateogry III: Temporary codes for new or emerging procedures removed after 5 years
Healthcare Common Procedure Coding System	HCPCS	aka 'hickpicks', a CMS billing coding system based on the AMA CPT. Category I : same as CPT. Cateogry II: items, supplies, non-physician (ancillary) services
Common Dental Terminology	CDT	HIPAA standard for dental procedures and electronic dental claims maintained by the ADA
Systematized Nomenclature of Medicine - Clinical Terms	SNOMED CT	multi-lingual concept-oriented standard terminology, maintained by the IHTSDO (NLM is the member organization of the IHTSDO and distributes SNOMED CT). Key feature is 'multi-axial' or compositional approach allowing terms to be combined and modifiers to be added. >300k concepts, >1M terms, >1M relationships
US SNOMED CT Content Request System	USCRS	allows users to request basic changes to SNOMED CT. New concept, new synonym, new parent, change description, etc
Nursing Vocabularies		irreconcilable information models, tedious, terms not used the way clinicans express them; NANDA, NIC, NOC, Omahar System, Home Health Care Classification, International Classfication for Nursing Practice
Unified Medical Language System	UMLS	1986 NLM standard to link the different standards; integrates 58 sources.
UMLS Metathesaurus	MTH	database of information on concepts that appear in one or more controlled vocabularies (eg ICD-10-CM, LOINC, CPT). Synonymous terms from different vocabularies are given same concept identifier; each distinct term can have different lexical variants, aka strings.
Concept Unique Identifer	CUI	all terms from all vocabularies representing same notion are groujped as a concept
Term Unique Identifier	LUI	all source terms of similar form (ie differing only in lexical variation) are groups as terms
String Unique Identifer	SUI	within each term lexical variants are strings
Atomic Unique Identifier	AUI	each string is an atom from its source
UMLS Semantic Network		each concept in the metathesaurus has an associated semantic type; concepts can also be linked through semantic relationships; generic relationships between semantic types of concepts
UMLS SPECIALIST lexicon		contains information on syntactic, orthographic, and morphological information for each concept; based on metathesaurus words and terms, designed to assist in natural language processing applications
Crosswalk		a set of links between the same information in one ontology to another ontology
Clinical Element Model	CEM	stack of coded items can be ambiguous, need model for clinical elements
Clinical Information Modeling Initiative	CIMI	aims to create CEMs for clinical data, part of ONC SHARP project

Interoperability standards		Meaning, Structure, Transport, Security, Services
Interoperability		The ability of different IT systems and software applications to communicate; to exchange data accurately, effectively, and consistently; and to use the information that has been exchanged. Made possible by the implementation of standards
Levels of Interoperability		Level 1: foundational interoperability (eg mail, fax, phone) Level 2: machine-transportable (structural; information cannot be manipulated, systems can exchange data but the data does not have to be able to be interpreted by the receviing system, eg scanned document) Level 3: machine-organizable (syntactic, sender and receiver understand vocabulary eg email, proprietary format files) Level 4: machine-interpretable (semantic, structured messages with standardized and coded data that can be read and understood, eg coded results from structured notes)
Fast Healthcare Interoperability Resources	FHIR	HL7 standard to merge versions 2 and 3, CDA, with a focus on implementation; response to complexity of HL7v3 but maintaining compatibility with other HL7 standards eg RIM, CDA. Basic building block is a resource which has common way to define and represent data elements, building them from data types that define common reusable patterns of elements; common set of metadata; human-readable part; public API
Integrating the Health Enterprise	IHE	Non-federal effort (sponsored by HIMSS) that identifies and demonstrates solutions to real-world interoperability problems, usually done by organizing interoperability showcases to demonstrate solutions. Clinical and operational domains produce technical framework documents which are published resulting in a trial implementation which if successful, profile will be published in final form
Medical Device 'Plug-and-Play" Interoperability Program		focus on interoperability of network-based devices, mainly in medical settings
Continua Health Alliance		consortium of companies and organizations devoted to interoperability of personal telehealth devices
CONNECT		US government software initiative suitable for an HIE based on record locator services (query-based or pull) approach
Direct Project		US government software initiative suitable for an HIE based on a secure e-mail (query-based or pull) approach

Institutional governance of clinical information systems		best to integrate governance into existing organizational governance structures. Information system projects are best viewed as clinical rather than IT. Leadership best derived from opinion leaders rather than technophilic users. Health IT leadership has migrated up the org chart.
Medical Records Bylaws		govern electronic records: content, permissions, responsibility of MDs for entering into the medical record, standards for completion. Oversight vested in medical record or HIM committe and for preparing for review by TJC, laws, policies for completion etc
Business Associates Agreement	BAA	required by HIPAA; associate has access to PHI
Data Use Agreement	DUA	govern conditions for use and sharing of data between organizations
Data Storage Agreement	DSA	Similar to above, but parties store data for another
Service Legal Agreement	SLA	Typically contractual rather than legal agreeements for performance

Clinical information needs analysis and system selection

Create project team, Define requirements (observation, mapping, interviews, focus groups, expert interviews), Identify viable candidates, System selection

Clinical information system implementation		assess current state, process design (future state), software design/configuration, testing, training, go-live, post conversion assessment, move to ongoing support model
Information Technology Infrastructure Library	ITIL	Developed in the UK, gained recognition when Microsoft used it as the basis for MOF; deines the operational structure and skill requirements of an IT area and documents a set of operational management procedres to foster more effective management of an IT operation and infrastructure: potfolio --> services --> processes --> procedures

Unit testing		Directed at menus, templates, and other modules and subunits of the system, without regard to other system components internal or external to a given application
Application testing		All modules and subunits of the application work in connection with each other
Integration testing		Conformation that information flow between the EMR and external systems occurs as expected
Performance testing		With production loads and users, system functions within expected boundaries
Post-production testing		After conversion, all aspects of system operate as required
System testing		Tests all aspects of a given system or application, eg how well the software satisfies the stakeholders' functionality, security, performance, load, reliability, compatibility, availability, etc. requirements
Regression testing		Tests to make certain that, with the exception of the change currently being requested, all components of the software's functionality/behavior are unchanged
Backout or contingency testing		Tests the ability to back out the changes being made to a system or, if changes cannot be backed out, tests the contingency plan if modifications cause problems once implemented
Test environment		Domain (or instance) of system with configuration and data similar to production in which testing occurs
Scripts		structured simulations of workflow and system use that can reproducibly be used for testing versions of proposed production version
Conversion		aka activation or go-live: Command Center (on-site presence, all teams: technical, application, interface, user liaisons; planning duration and logistics should being early); User support team (goal is overwhelming support 'at the elbow', project and non-project team, roamers); Communication methods and plan (pagers, cell phones, scheduled status calls, shift reports); Issue management (capture and triage, documentation, communcation / closing the loop); Review downtime procedures; Have back-out or remediation plan
User support models		role of help desk, escalation procedures, onsite vs remote, superusers (may be busy)

Clinical information system maintenance		operations entails day-to-day running, maintenance, enhancement and safeguarding of the system to meet the availability and reliability requirements. Patches, upgrades from vendor, coding system updates (ICDX, DDI databases etc), interfaces with other systems, delivery (end user devices, Citrix)
Organizational strategies necessary to EHR Safety		care-process transformation, patient safety, HFE, software safety, proejct management, continuous improvement (Walker)
EMR optimization		post-implementation focus on features and functionality that have been incompletely or suboptimally adopted; review of workflow and tailored education; use of EMR to assist in meeting organizational quality, safety, and financial goals
Compliance and the law		Documentation, orders, results review and other tasks and audit trails scrutinized by compliance are increasingly accomplished using computing systems: compliance officers charged with protecting the organization and may not have full understanding of clinical computing systems functionality and workflow; copmliance and DOJ focus on cloning, upcoding, copy/paste
Clinical computing systems and the law		authentication vs authorization; nonrepudiation (user may attempt to claim they did not create document when in fact they did); audit trails (reconstruct construction of note or use of EMR), document version history; close cooperation with compliance and general counsel

Usage		Proportion of users. In ambulatory settings, CDC finds growing adoption. Hospital settings most sucessful but do not use all capabilities. Compared to other countries, US is a laggard but most other countries do not have advanced functionality
Evaluation		help experts determine what works best; allows users to cut through hype; justifies costs / helps with cost comparison
Efficacy vs Effectiveness		Efficacy is evaluation in controlled research setting; Effectiveness is evaluation in real world (aka outcomes research)
Microevaluation vs Macroevaluation		Micro: evaluation of individual components of a system; Macro: evlaution of a whole system as a 'black box'