Basically, this chapter introduces the field of medical informatics. It describes how and why it emerged, some past accomplishments, and what can be expected in the future.

The emergence of medical informatics as a new discipline is due in large part to advances in computing and communications technology; to an increasing awareness that the knowledge base of medicine is essentially unmanageable by traditional paper-based methods; and to a growing conviction that the process of informed decisions making is as important as the collection of facts on which clinical decisions or research plans are based. (pg 20)

Historical perspective: (pg 20-26)

The trend from mainframe computers to microcomputers and personal computers.

1890: first practical application of automatic computing relevant to medicine

1940’s: electronic digital computers began to appear

1950’s: general purpose computers began to appear in the marketplace

• One early activity in medical computing was an attempt to construct systems that would assist a physician in decision-making

1970’s: -introduction of HIS applications

-medical-computing activity broadened in scope and accelerated with the appearance of minicomputers. These machines made it possible for individual departments to acquire their own dedicated computers and to develop their own application systems.

early 1980’s: everything radically changed when PC’s became available. This change broadened the base of computing and gave rise to a new software industry.

Future:

• the vision of integrated medical-computing environment of the future
• the concept to EPR

 Key Terminology:

• medical care:

consists of making a diagnosis, choosing a treatment, managing therapy, making decisions, monitoring a patient, and preventing disease. The fabric of medical care is a continuum in which these elements are tightly interwoven. (pg 28)

• information science:

is used to referred to the broad range of issues related to the management of both paper-based and electronically stored information. (pg 19)

• medical informatics:
• a scientific field that deals with the storage, retrieval, and optimal use of biomedical information, data, and knowledge for problem solving and decision making. (pg 20)
• the study of how computers are used in medicine to ease the burdens of information processing, and the means by which new technology promises to change the delivery of health care. (pg 32)

relationship to medical science (pg 26)

• medical informatics is entwined with the substance of medical science. It determines and analyzes these structure of medical information and knowledge, whereas medical science is constrained by that structure.
• medical informatics is the interface between computer science and medical science.

relationship to biomedical engineering (pg 28)

• biomedical engineering focus on research and development of instrumentation and towards the development of medical devices, whereas the medical informatics focus on information. Yet, there are no strict boundaries between the two fields. The topics of medical informatics clearly have an overlap with biomedical engineering, however, the two fields have difference in emphasis. In biomedical engineering, the emphasis is on medical devices; in medical informatics the emphasis is on medical information and knowledge, and their management using computers.

relationship to computer science (pg 30)

• medical informatics draws from all the concepts of computer science - development of hardware, software and computer science theory.

• there are several global forces that are affecting medical computing and that will determine the extend to which computers are assimilated into medical practice:

1. new development in computer hardware and software

2. increase in the number of professionals who have been trained in both clinical medicine and computer science.

3. ongoing change in health care financing designed to control the rate of growth of medical expenditures (pg 33)

• integrated computer systems potentially provide the means to capture data for detailed cost accounting, to analyze the relation of costs of care to the benefits of that care, to evaluate the quality of care provided and to identify areas of inefficiency. (pg 34)
• designers of medical systems must satisfactorily address the logistical and engineering questions before computers can be fully integrated into medical practice. e.g. are the computer terminals conveniently located? can the users interact easily and intuitively with the computers? etc. (pg 34)

Data are central to all medical care because they are crucial to the process of decision making. In fact, all medical care activities involve gathering, analyzing or using data. There are a number of forms of data present in health care delivery. These range from concrete numeric data values such as laboratory test results to interpretive information about a patient’s family or economic setting, or the subjective impression of disease severity that an experienced physician will often obtain within a few seconds of entering a patient’s room. No physician disputes the importance of such observations in decision making during patient assessment, yet the precise roles of these data and the corresponding decision criteria are so poorly understood that it is difficult to record the data in ways that convey their full meaning, even from one physician to another.

1. What are the Types of Medical Data?

There is a broad range of data types in the practice of medicine and the allied health services. Narrative data account for a large component of the information that is gathered in the care of patients. For example, the patient’s description of his present illness, including responses to focused questions from the physician, generally is gathered verbally and is recorded as test in the medical record. Many data in medicine take on discrete numeric values. These include such parameters as laboratory test results or vital signs. Subsequently, the issue of precision becomes important. Other forms of data include analog data (e.g. ECG) and visual images – either acquired from machines or sketched by the physician.

2. Who Collects the Data?

Physicians are key players in the process of data collection and interpretation. They converse with the patient, examine the patient, and generally decided what additional data to collect. Nurses also play a central role in making observations and recording them for future reference. Because nurses spend more time with patients than physicians do, especially in the hospital setting, nurses often build caring relationships with patients that uncover information and insights that contribute to proper diagnosis, to understanding of pertinent psychosocial issues, or to proper planning of therapy or discharge management. A variety of other health care workers contribute to the data collection process )e.g. office staff and admissions personnel, physical or respiratory therapists, laboratory personnel, radiology technicians and radiologist, pharmacists...). Clearly, most people employed in health care settings gather patient data, record them, and make use of them in their work.

2. Uses of Medical Data

Medical data may be needed to support the proper care of the patient from whom they were obtained, or may contribute to the good of society through the aggregation and analysis of data regarding populations of individuals. One problem with traditional data recording techniques has been that the paper record has worked reasonably well to support the proper care of individual patients, but has made clinical research across populations of patients extremely cumbersome. Computer based record keeping offers major advantages in the regard.

1. For the Basis for the Historical Record

Medical records are intended to provide a detailed compilation of information about individual patients: patient history, symptoms, physical signs, lab results, etc. It is the careful observation and recording by skilled health care personnel that has always provided the foundation for generating new knowledge about patient care.

2. Support Communication Among Providers

A central function of structured data collection and recording in health care settings is to assist in providing coordinated care to a patient over time. In the world of modern medicine, the emergence of subspecialization and the increasing provision of care by teams of health professionals have placed new emphasis on the central role of the medical record. Now the record not only contains observations by a physician for reference on the next visit, but also serves as a communication mechanism among physicians and other health care personnel such as physical or respiratory therapists, nursing staff, radiology technicians, social workers, or discharge planners.

3. Anticipate Future Health Problems

Quality medical care involves the education of patients about the ways in which their environments and lifestyles can contribute to or reduce the risk of future development of disease. Medical data therefore are important in screening for risk factors, following patients’ risk profiles over time, and providing a basis for specific patient education pr preventative interventions, such as diet, medications, or exercise.

4. Record Standard Preventative Measures

The medical record also serves as a source of data on interventions that have been performed to prevent common or serious disorders (e.g. immunizations).

5. Identify Deviations from Expected Trends

Data are often useful in medical care only when viewed as part of a continuum over time. An example of this issue is the routine monitoring of children for normal growth and development by paediatricians.

6. Provide a Legal Record

Medical data is the foundation for a legal record to which the courts can refer if necessary. The medical record is a legal document; most of the clinical information that is recorded must be signed by the responsible individual.

7. Support Clinical Research

It is only by formally analyzing data collected from large numbers of patients that researchers can develop and validate new clinical knowledge of general applicability. Thus, another use of medical data is to support clinical research through the aggregation and statistical analysis of observations gathered from populations of patients.

3. Weaknesses of the Traditional Medical Record System

• the patient’s chart may be unavailable when the health care professional needs it (e.g. in use by someone else at another location)
• once the chart is in hand, it may still be difficult to find the information required (i.e. poor organization of the chart)
• once the health care professional has located the data, he may find it difficult to read them
• when a chart is unavailable, the health care professional still must provide medical care. This providers make do without the past data, basing their decisions instead on what the patient can tell them and on what their examination reveals.
• when a patient with chronic or frequent diseases is seen over months or years, their records typically grow sufficiently large that they must be broken up into multiple volumes

2. Redundancy and Inefficiency

Health professionals have developed a variety of techniques that provide redundant recording to match alternate modes of access. For example, the result of a radiologic study typically is entered on a standard radiology reporting form, which is filed in the portion of the chart labeled "X-ray". For complicated procedures, the same data often are summarized in a brief note by the radiologist in the narrative part of the chart, which she enters at the time of the study because she knows that the formal report will not make it back to the chart for 1-2 days.

3. Influence on Clinical Research

It is arduous to sit with stacks of patients’ charts, extracting data and formatting them for structured statistical analysis, and the process is vulnerable to transcription errors. There are retrospective chart reviews (investigate a question that was not a subject of study at the time the data were collected) and prospective chart reviews (the clinical hypothesis is known in advance and the research protocol is designed specifically to collect future data that are relevant to the question under consideration).

4. The Passive Nature of Paper Records

A manual archival system is inherently passive; the chars sit waiting for something to be done with them. Automated record systems introduce new opportunities for dynamic responses to the data that are recorded in them.

4. The Structure of Medical Data

Medicine is remarkable for its failure to develop a standardized vocabulary and nomenclature, and many observers believe that a true specific basis for the field will be impossible until this problem is addressed. The debate has been accentuated by the introduction of computers for data management because such machines tend to demand conformity to data standards and definitions.

1. Coding Systems

Because of the need to know about health trends for the population, and to recognize epidemics on their early stages, there is a variety of health reporting requirements for hospitals and practitioners. Much of this reporting involves the coding of all discharge diagnoses for hospitalized patients, plus coding for certain procedures that were performed during the hospital stay. For such data to be useful, the codes must be well defined as well as uniformly applied and accepted.

2. The Data to Knowledge Spectrum

Data = single observational point that characterizes a relationship

Knowledge = the formal or informal analysis of data

information = both organized data and knowledge

A database is a collection of individual observation without any summarizing analysis. A knowledge base, on the other hand, is a collection of facts, heuristics, and models that can be used for problem solving.

5. Strategies of Medical Data Selection and Use

There can be marked interpersonal differences in both style and problem solving that account for variations in the way practitioners collect and record data for the same patient under the same circumstances. An example of this is the difference between the taking of the first medical history, performing of the physical examination, and writing of a report by a novice medical student, and the similar process undertaken by a seasoned clinician examining the same patient.

1. The Hypothetico-Deductive Approach

This approach is effectively summarized by the diagram on page 62.

Note that the hypothesis-directed process of data collection, diagnosis, and treatment is inherently knowledge-based. It is dependent not only on a significant fact base that permits proper interpretation of data and selection of appropriate follow-up questions and tests, but also on the effective use of heuristics techniques that characterize individual expertise.

2. The Relationship Between Data and Hypothesis

6.1 What is the Medical Record? (pg. 182-185)

6.1.1 Purposes of a Medical Record

Purposes are divided into 3 classes of functions:

1. facilitates patient care

• main purpose
• summarizes a patient’s history and documents the observations, diagnostic conclusions, and plans of health-care professional
• serves as external memory for health-care providers to refer to at later time
• means of communication between health-care providers
• primary mechanism for ensuring continuity of care during a patient’s hospitalization
• Ambulatory Medical Record
• helps ensure continuity of care across outpatient visits
• enable to view data collected over time - track problems and diseases

1. financial and legal record

• used to determine if the patient received proper care
• contains info on health-care provider’s actions & justifications for those actions
• provides basis for Quality Assurance and accreditation procedures - allows to judge appropriateness of care
• legal requirements
• record must be resistant to change
• available for 7 years after last visit (time to retain records will vary)
• financial uses
• info used to classify patients into DRGs (basis of Medicare payments)
• third party payers check to make sure all charges are valid (procedure was actually done, it is recorded in the chart)
• hospitals use charts to check that all possible procedures were charged for

1. aids clinical research

• retrospective studies based on medical-record reviews to determine medical relationships
• very important in epidemiology

6.1.2 Use of a Computer-Stored Medical Record

Advantages of computer-stored Medical Record over a paper medical record

• easily accessed via nearby terminals - quick retrieval of data needed
• permits remote access
• permits simultaneous access
• more legible and better organized reports
• not handwritten
• impose structure on data
• improve completeness and quality of entered data through validity checks
• interactive systems can prompt for additional information
• can help data entry (e.g. manage flow of input reports or verifying that needed reports have been completed)
• report content on variety of media (from video display terminals to paper)
• same information can be reported in a number of different formats
• reports can be tailored in content and format to purpose of the report (reduces need for redundant data recording)
• can make decisions automatically about the data it captures and reports
• can ask for missing data
• can analyze data to assist care providers

1. informational scope - does it contain data from only 1 or 2 sources or does it include lots of info sources (e.g. lab, imaging, physician, pharmacy, etc.)
1. duration of use - a record that has accumulated patient data over 5 years will be more valuable than one that contains records of only those visits made during 1 specific month.
1. representation of data - if data is coded or just narrative text entry. Coded data is standardized and has consistent use of medical terminology, therefore can aggregate and summarize data provided by different sources or the same source at different times. An unstructured on-line record cannot aid users actively in decision making or medical research.
1. geographic dispersion of terminals that access the system - a system uses by many people and accessible from only a few sites will be less valuable than if it is accessible form lots of terminals around the hospital or even from private offices and physician’s homes.

• requires larger initial investment because of hardware, software, and training costs
• delays between the installation of the computer-stored record and the appreciation of its full benefits - when most active patients have electronic records of substance
• importance of maintaining confidentiality of EPR (electronic patient record) adds to work of maintaining system and to the cost of managing it
• potential for subtle and catastrophic failure
• need to keep alternative manual procedures in case of failure
• implement and maintain backup procedures
• transcription errors when data entered by clerk from handwritten form
• faulty software could alter even correctly entered data
• problem of capturing physician data
• physicians often use a large amount of data to make 1 small decision
• cost of entering all data that a physician needs to make a decision can be high relative to the decision assistance provided by the system
• speech recognition could reduce this problem

6.2 Historical Development of Medical-Record Systems (pg. 185-187)

6.3 Fundamental Issues for Computer-Stored Medical-Record Systems (pg. 187-211)

6.3.1 Data Entry

1. data capture

• if the scope of data capture is restricted to the variables under control of the organization then data capture is trivial
• if need data from other organizations then need to negotiate exchange of data (e.g. of physician’s office needing info from a hospital about her patient’s recent hospitalization (pg 187))

1. data input

• burdensome because of personnel time required
• people must interpret data as well as enter them into the computer
• data may be entered in free-text, coded or in a combined form
• major advantages of coding
• data classified and standardized which facilitates research, billing, and selective retrieval
• lets computer "understand" data ® can process more intelligently
• usually able to be stored more compactly
• if codes are few, data can be entered in a few key strokes
• major disadvantages of coding
• cost of translating source text into valid codes
• time to classify source text by code class (especially if text is unusual and seems to fit no class)
• time to train personnel & time to maintain code dictionary
• potential for coding errors (difficult to catch because no redundancy as in text)
• the more coding the greater the interpretation time; the more free-text the greater the typing time - can balance by using combination of codes and free-text
• coding can be difficult because of legibility problems - problems can be solved by using standard forms
• availability of electronic data from other systems (like lab) eliminates the need to retype (still would have to code and/or normalize data from different systems)

1. the patient’s reports (i.e. medical history, or present symptoms)
1. physical findings
1. differential diagnosis
2. treatment plan

6.3.2 Alternative Methods For Displaying Information

• presents data almost instantaneously
• can change dynamically in response to users
• quiet and require no tending for ribbon changes or paper changes

• presents more info on 1 page than will fit on 1 screen
• formatting capabilities of printers usually better than VDTs
• people can carry paper around, annotate it, and use without training
• people can read off paper faster than off a screen

1. presentation graphics - typical line or bar graphs (e.g. daily ICU reports from HELP system, fig. 6.11, page 203)
2. diagrammatic reports - combine diagrams with number or graphs (see fig. 6.12, pg 204)
3. continuous curves and pictures - ECG strips, radiographic images and photographs.

1. predefined or canned notes - common phrases or paragraphs that can be inserted into dictated or written narratives with a few key strokes. Used in results reporting from radiology and results of EEGs, spirometry and surgical pathology exams.
2. mail merge - for sending standard letters to many people (e.g. patient appointment reminders)

6.3.3 Query and Surveillance Systems

• clinical care - reminders have increased use of preventative care in eligible patients. When used for order entry (order tests or treatments), surveillance systems can improve cost-effectiveness and quality of patient care. Query systems are useful for conducting ad hoc searches.
• clinical research - query systems can identify all patients that qualify for the study (meet entrance criteria). Surveillance systems can support the execution of the study (track patients through visits, follow the steps of a clinical trial as described in the study protocol to ensure that treatments are given and measurements obtained when required)
• retrospective studies - systems can identify case studies and corresponding controls and can help with the analysis (comparison of the 2 groups). Computer records are likely to be more complete especially if direct data entry is from other systems. Good for research on physician’s practice patterns, the efficacy of tests and treatments and the toxicity of drugs.
• administration - system can provide cost info (based on DRGs) (see chapter 19). Administrators must be able to monitor physicians’ use of resources of various classes of patients, and to provide appropriate feedback. Query systems can provide info about the relationships among diagnosis, indices of severity of illness and resource consumption.

6.4 Examples of Automated Ambulatory Medical Record Systems (pg. 211-215)

An automated ambulatory medical record is a key feature of both a HIS and an automated ambulatory medical-record system (AAMRS) - both support patient care and administrative functions. Most AAMRS contain modules for medical records, finance and admin. info and getting reports. 4 examples of AAMRSs:

1. COSTAR (pg 212/213)

• 2 design goals - 1) to facilitate patient care by improving the availability and organization of the medical record and 2) to support the function of an ambulatory care practice by providing administrative, managerial, and financial support functions.
• modules for
• system security and data integrity (required)
• patient registration (required)
• appointment scheduling
• billing and financial reporting
• collection and storage of medical records (required)
• management reporting
• can operate as a fully automated medical record system
• system prints summary medical record and blank encounter form for each patient prior to visit
• visit data collected on encounter form
• data entered into system from encounter form
• system produces standard output reports
• supports medical query language

1. The Regenstrief Medical Record System (RMRS) (pg 213/214)

• part of larger system which handles appointment scheduling and charge capture
• has a reminder system that actively reviews patient data and produces reminder notes for the physician based on 1400 encoded protocol rules.
• schedules all appointments and produces 3 documents prior to visit
• quality assurance report - recommendations of preventative procedures to do
• flowsheet summary - time-ordered summary of the clinical database
• patient encounter form - form based on patient data and physician supplied rules
• physician fills out patient encounter form
• data entered into form from encounter form
• RMRS supplements rather than replaces the paper file
• uses CARE (medical query language)

1. The Medical Record (TMR) (pg 214)

• design goal to eliminate paper record
• emphasizes capture and storage of patient care data
• also provides appointment scheduling and billing
• maintains complete list of diagnoses and procedures and a time-orientated record of subjective and physical findings, lab data, meds and procedures for each patient
• generates a patient-specific encounter form prior to visit
• physician uses encounter form to review and collect new data
• data entered from encounter form
• data can be viewed on VDT in 1 of 3 orientations: problem, time or encounter

1. Summary Time Oriented Record (STOR) (pg 214-215)

• provides 2 areas of clinical services 1) the computer based storage and retrieval of records, 2) the on-line display of patient info in response to queries
• creates common inpatient and outpatient database gathering data from departmental systems via LAN
• system reports flowsheets, graphs, problem and therapy lists and registration data, results reporting (of tests), discharge diagnoses and meds and ECG reports
• prints patient specific encounter form
• physician updates encounter form during visit
• data entered into system from encounter form
• flexible in that it is an arbitrary depth hierarchy of coded or uncoded items

6.5 The Future of Automated Medical Record Systems (pg. 215-217)

The purpose of a hospital information system (HIS) is to manage the information that health professionals need to perform their jobs effectively and efficiently.

HISs and ambulatory medical-record systems (see chapter 6) both facilitate communication, integrate information, and coordinate action among multiple health-care professionals; in some cases, they also aid in patient monitoring. In addition, they assist in the organization and storage of information, and they perform some record-keeping functions. The degree to which specific features are emphasized, differ between ambulatory medical-record systems and HIS.

1. Information Requirements

The most important function of an HIS is to provide communication among the many health-care workers who cooperate in caring for patients, and to organize and present patient-specific data so that the staff can more easily interpret and use those data in decision making.

Hospital’s Information needs is classify in 3 broad categories:

Health-care workers require detailed and up-to-date factual information to perform the daily tasks that keep a hospital running.

Hospital personnel also require information to make short and long-term decisions about patient care and hospital management. An HIS can help plans by aggregating, analyzing, and summarizing the information relevant to decision making.

The need to maintain records for future reference or analysis (i.e. Medical record).

It is important to know that hospital planers can analyze the data to detect trends in resource utilization and can make predictions about demand, they can expand or contract hospital services to an appropriate level.

2. The cost of Information

Information management in hospitals is a costly activity. The collection, storage, retrieval, analysis, and dissemination of clinical and administrative information necessary to support a hospital’s daily operations, to meet external and internal requirements for documentation, and to support short-term and strategic planning are important and time-consuming aspects of the jobs of hospital workers.

3. The function of a Hospital Information Systems

A careful designed computer-based system can increase the productivity and effectiveness of health professionals and potentially can provide a means to decrease a hospital’s labor costs.

Friedman and Martin proposed a functional model for an HIS that consists of components that perform six distinct functions:

1. Core Systems

Perform the basic centralized functions of hospital operation, such as patient scheduling, admission, and discharge. The census is maintained by the admission-discharge-transfer (ADT) component of an HIS, which updates the census whenever a patient is admitted to the hospital, discharged form the hospital, or transferred to a new bed. Using a computer-based system, health-care workers can examine and update the centralized patient database form remote locations via terminals. When an HIS is extended to the pharmacy, laboratory, and other ancillary departments, the core system can provide a common reference based for use by these systems as well.

2. Business and financial systems

Assist with traditional financial functions, such as managing the payroll and accounts receivable. The majority of these data-processing tasks are well structured, labor-intensive, and repetitious—ideal applications for computer. Data-processing techniques developed to meet the needs of general industry were adapted easily for use in hospitals. Hence, they were first to be automated.

3. Communications and networking system

Allow the integration of all components of an HIS. An HIS can facilitate communication between the ward and the ancillary departments. Automated order entry and results reporting are two important functions provided by a computer-based HIS. Health professionals can use the HIS to communicate with ancillary departments electronically, elimination the easily misplaced paper slips and thus minimizing delays in conveying orders. The information then is available online, where it is accessible to health professionals who wish to review a patient’s drug profile or previous laboratory-test results.

4. Departmental-management systems.

Support the informational needs of individual departments within the hospital (i.e. laboratory, pharmacy, and radiology departments). If those departments systems are integrated with the rest of an HIS they can use data collected in other parts of the hospital as inputs for their own internal functions.

5. Medical-documentation Systems

Perform the functions of the standard medical record in collecting, organizing, storing, and presenting the clinical information used to manage the care of individual patients (see chapter 2 and 6).

Medical-documentation systems also can assist in hospital wide activities, such as infection control, discharge planning, quality assurance, and utilization review. Medical-documentation systems assist in collection and preliminary analysis of data that help hospital administrators to assess patient outcomes, quality of worker’ performance, and compliance to hospital policies and standard procedures. Administrators then can use this information to identify potential problems and to evaluate the effectiveness of existing hospital policies.

6. Medical-Support systems

Assist directly clinical personnel in data interpretation and decision making. Once the clinical components of an HIS are well developed, clinical-support systems can use the information stored there to monitor patients and to issue alerts, to make diagnostic suggestions, and to provide limited therapy advice.

4. Integration of Hospital Information

When computers are used, the hospital’s information processing often is performed on separate computers. These computers may also be managed separately, thereby avoiding conflicts about priorities in services and investment.

The lack of integration of data form diverse sources creates a host of problems. If clinical and administrative data are stored on separate systems, then data of mutual interest must be copied either from the source documents into both systems of form one system to the other. In addition to the expense of redundant data entry and data maintenance incurred by this approach, the consistency of information tends to be poor because data may be undated in one place and not in the other, or information may be copied incorrectly. Inconsistencies in the databases can result in lost charges, inappropriate patient management, and inappropriate resource-allocation decisions.

The integration of data myriad sources in the hospital produces a rich database for decision making. If these data collectively are accessible to application programs, extremely sophisticated applications are possible (i.e. infection-control program).

The bottom line, the objectives of high-quality and cost-effective health care cannot be satisfied if multiple computer systems operate in isolation.

2. Alternative Architectures for Hospital Information Systems

There are 3 alternative models for an integrated HIS: Central, modular, and distributed models. The choice of architecture affects the choice of hardware and the design of software for storing, accessing, and transmitting data. In general, we have seen an evolutionary trend in the development of HISs from central, to modular, to distributed systems (and probably back to centralize now…Kathleen note).

1. Central Systems ~1965-1973 (see page 228 for figure 7.1)

A single, comprehensive, or central system that meant to best meet a hospital’s information needs. A system in which a single, large computer performs all information processing and manages all the data files using application-independent file-management programs. The Technicon Medical Information System (TMIS) is an example of a central system (see section 7.3.1 for detail information on that system)

Advantages and disadvantages of the Central Systems

2. Modular Systems ~1973-1983 (see figure 7.2 on page 230)

Only one or a few machines are dedicated to the hospitals. Distinct software application modules carry out specific tasks, and common framework, which is specified initially, defines the interfaces that will allow data to be shared among the modules. May tasks may be performed by freestanding systems. The IBM Patient Care System (PCS) is an example of a modular approach (see page 231 for details).

Advantages and disadvantages of modular systems:

3. Distributed Systems

In a distributed system, an HIS consists of a federation of independent computers that have been tailored for specific application areas. The computers operate autonomously and share data (sometimes things like printers) by exchanging information over a local-area network (LAN) using a standard protocol for communication.

4. Hospital Information Systems: Present and Future

Hospitals may use locally developed and maintained programs running on hospital-owned or leased computers, they may purchase vendor-supplied and vendor-supported computer systems, or they may contact with outside service organizations.

Use of large, shared systems, operating form remote locations, is declining, and most companies no longer provide hospital services from centralized locations.

More hospital move to in-house data processing and purchase commercial systems.

In 1986, ~80% of hospitals used a computer to assist in financial data processing and ~70% used one to help with admissions and patient registration, only about one-quarter of hospitals had a system that allowed personnel on the wards to access the computer interactively and thus to enter orders and to review test results.

Hospitals still find that vendor systems are inflexible and do not completely satisfy their needs. Insufficient technical support for both end users and in-house data-processing personnel is a common problem.

An HIS is expensive to purchase and maintain and there has been no clear demonstration that a comprehensive HIS can reduce cost; without accompanying changes in the organizational and managerial structure of the hospital, many of the benefits of an HIS cannot be realized.

There is a growing recognition that information is the basis for every decision made in the hospital and that the integration of data collected throughout the institution is critical if a hospital is to function effectively in today’s competitive health-care environment.

Demands for data integration and better capability to access and analyze information are producing the following trends in HIS development:

Hospital personnel using PCs that are linked into a LAN will have greater access to data and greater flexibility in analyzing and using data for decision support.

These point-of-care systems potentially can increase the productivity of nurses by reducing the amount of paperwork, can provide more accurate and timely access to the detailed clinical information for better cost-management and quality-assurance activities, and even can improve the quality of care.

In conclusion, to accommodate their special needs, many of these hospitals will have to develop and accommodate their special needs, many of these hospitals will have to develop and maintain their own systems; they will not be able to rely heavily on vendors or on contributions from other, similar hospitals.

Smaller institutions will depend largely on turnkey systems provided by vendors and modified only nominally, if at all. Because the simple billing, order entry, and reporting functions are now mature, those will be the primary functions available to the small hospitals.

PS I will think at what we have been taught so far…which differ from this conclusion…I believe!!!! Just a thought.

Nursing was characterized by nurturance, but a nurturance vested with authority and guided by knowledge.

Nurses typically are the primary users of HISs; features such as the order entry, results reporting, and automated charting of many HISs are used most often by nurses performing the documentation and communication aspects of their jobs.

Thus, NISs and HISs are integrally related, a NISs have specialized components to help with the unique tasks of nursing.

11. Clinical Practice

Nursing’s mission is complementary to but different form medicine’s (see figure 8.1, page 246…IMP). The nurses assess patients’ needs, plan patient care, administer treatments, educate, and provide emotional support.

For nursing, the process of diagnosis consists of looking for indicators of how the person responds to an actual or potential health problem, and of the person’s ability to provide independently for her own care.

The provision of computer-based support for the nursing process is complicated by 2 factors:

1. The fledgling state of nursing science

2. the complexity of the process itself.

Unlike medicine, nursing has no generally accepted taxonomies of nursing diagnoses, nursing objectives, or nursing interventions. Accordingly, it is hard to identify what the best nurses are doing to be effective, so that their behavior can be replicated by other nurses, and can by tracked and supported by computer-based systems.

Although the nursing care of each client is planned, implemented, and evaluated by means of a deliberative process, nurses are handicapped in carrying out this process by the lack of a standard taxonomy of data, diagnoses, objectives, interventions, and outcomes, and by the absence of explicit rules for decision making at each stage of care and for moving form one stage to the next. As a result, nursing-care planning often is unsystematic and intuitive; statements of diagnoses, objectives, and interventions usually are so varied as to make comparisons difficult; and nurses receive little if any feedback on the effectiveness of care. The complexity of the situations with which nurses must deal is also a problem in implementing a NIS.

The necessity of keeping their own records and of communicating with other professionals results in nurses’ spending up to 40 percent of their time on paperwork. The amount of information confronting nurses who must make decisions about the care of individual patients, an must coordinate the care of many patients in concert with the entire health-care team, far exceeds the limits of human capacities for information processing.

2. Nursing Administration

Nurse administrators must identify the knowledge, skills, time and resources necessary to provide high-quality, cost-effective nursing for each of their clients, and they must plan how to provide those human and material resources to all clients, 24 hours per day, 7 days per week.

Nurse administrators must consider the different levels of education, the knowledge, skills, and experience of their personnel. As well as the constraints on the personnel’s availability and they must match those variables to the needs of clients to provide necessary services 3 shifts per day, every day.

After 1960, assign nurses to units on the basis of the workload anticipated for the assigned shift. Has been criticized because it is based on the average times nurses spent in the past to care for a patients with certain needs. It is not known whether those average times represent the optimum for either productivity or quality of care.

What they want: A clinical-practice data linking patient needs or problems with objectives, interventions, and outcomes of care. Then it is possible to determine which interventions are most effective an efficient, and to use the times required for productive, high-quality care in computing the nursing hours needed on a unit.

Quality of health services can be measured along any of 3 dimensions: structure (number of nursing personnel, the working conditions etc.) process, and evaluation of outcomes.

The quality of nursing care look at all 3 dimensions.

3. Nursing Research and Development

Clinical nursing records are so unstandardized, disorganized, incomplete, and difficult to retrieve as to be virtually useless as a resource for research. Standardizing the taxonomies and organization of clinical nursing records and providing for ready retrieval of data would offer important advantages for the development of nursing knowledge.

2. A historical Perspective: Development of Computer Systems for Nursing

Although these systems offer no direct support to decision making, they organize clinical data more systematically that do traditional paper-based systems, and thus significantly reduce the time nurses spent processing information. The problem of providing the standardized, reliable clinical data for research, however, has not yet been resolved.

1. Systems for Clinical Practice

This section gives you example of systems. Please skim through! (page 255 to 257)

2. System for Nursing Administration

It is interesting to note that, in spite of the wide availability systems for computing nurse staffing needs, there continues to be great dissatisfaction in many areas with most nurse-staffing methodologies.

3. Systems for Nursing Education

Nursing education perhaps has been the area where the development of computer applications has been most dramatic, and prolific. Until 1980, development of CAI (computer aided system>>>ask questions to student and the student give the answer) for nursing was limited. There was a lack of sufficient rewards or nursing faculty members to invest the considerable time and work necessary to design and program CAI.

3. Fundamental Issues for Nursing Information systems

The primary functions of an NIS are to assist nurses in record keeping and in decision support. Nurses and systems designers must create systems that use the information put into them, that transform raw data into more useful forms, and that propose clinical inferences for the nurse’s consideration. For such systems to become possible, nurses must define the content of each step of the nursing process in clear and consistent language, and must tell how the elements fit together within each step and between steps.

1. Taxonomies of nursing phenomena

There is general agreement that nursing care is delivered via a problem-solving process that includes collecting data, formulating diagnoses, setting objectives, choosing and implementing interventions, and evaluating outcomes.

Nurses diagnosis refer to nurses identifying problems (or resources) that were the particular focus of nursing practice. The diagnoses they derive contain three elements: a statement of the problem itself, a statement of etiology in the form of a list of possible causes, and a list of defining characteristics indicating that the problem is present.

2. Computer-Aided Planning and Documenting of Care

The developers of NISs have adopted standard vocabularies so that computers can assist in patient care. The description of patient’s condition is not only a summary but also a point of departure.

Standardized care plans are undoubtedly useful. They save time in terms of compilation and write-up and they provide consistency over time, and reduce the probability of errors. It is important though that nurses have the opportunity to add to or to change the standard care plans.

3. Evaluation Care

Two tasks constitute the evaluation procedure: 1) the evaluation of the patient’s condition, and 2) the evaluation of the nursing care rendered. An NIS should permit both. An NIS should also accommodate multiple time points and allow clinical updating. The clinical updating is important procedure in nursing inference, because it maximizes accuracy, detail, and sophistication. Quality assurance can be built into the NIS subroutines for auditing the nursing-care plans and documentation of care.

4. Current Nursing Information Systems

Examples of information systems for nursing are TMIS (technicon system p.267), IBM’s Patient Care System (PCS see section 7.2.2 and p.267), the COMMES (Creighton On-Line Multiple Medical Expert System…see page 268), and the ULTICARE system (see page 268)

2. Systems for Nursing Administration

Examples of systems for Nursing Administration include NPAQ, GRASP, COSTAR, CASH, and PIRS. (One small paragraph on page 268-269).

3. Systems for Nursing Education

Examples of systems for Nursing Education include NEMAS (nursing education modular authoring system) which helps nursing instructors to write patient simulations that will teach specific cognitive skills in the nursing process.

5. A look to the Future

The changes in computing technology that are shaping the development of HISs affect the ways and the extent to which nurses use computers in their work. The information systems support nursing decision making and the professional aspects of nursing practice, rather than just the clerical aspects. A effective NIS systems must not merely emulate what nurses do, but rather must complement the nursing process.

External pressures are forcing nurses to describe more explicitly the nature of their work, and to justify the costs of that work. There is a growing interest among nurses and hospital administrators n defining standard nursing taxonomies, developing uniform treatment procedures, and studying ways to make the nursing staffing more responsive the level of patient needs in the hospital. We need to develop a deeper understanding of what is entailed in nursing care.

• primary functions of the clinical laboratories are to acquire, validate, interpret, and communicate the information derived from analyzing patient specimens.
• The quality of laboratory services depends not only on the accuracy and precision of the test results, but also on the timeliness of test completion and the availability of the results.
• Implementation of DRG based reimbursement for Medicare-insured hospital patients forced clinical laboratories to become more efficient, thus producing additional incentives to use computers. (note this is American)
• Laboratory Information Systems (LIS) support fundamental functions in both data processing and laboratory management.
• Help to analyze the primary data, to store and distribute test results, to monitor testing quality, to document laboratory procedures, and to provide information used by managers in controlling inventory, monitoring workflow, and assessing laboratory productivity.

1. ORGANIZATION OF THE CLINICAL LABORATORY

• There are various specialized laboratories:
• clinical chemistry labs
• tests such as plasma, serum, whole blood, urine, cerebrospinal fluid
• chemical reaction by mixing reagent with sample specimen
• Common tests measure the concentrations of electrolytes, blood gases, enzymes, lipids, carbohydrates, proteins, physiologic waste, hormones, drugs, poisons.
• hematology
• tests to characterize the cellular elements of the blood (red and white blood cells, platelets)
• CBC (Complete blood count) - a panel of tests to determine numbers, sizes, and hemoglobin content of the cells.
• microbiology
• identifies the infectious agents (bacteria, fungi, parasites, viruses) present in patient specimens.
• Examined under a microscope, cultured in isolation, or subjected to biochemical analysis
• This info used by physicians to prescribe effective therapeutic regimens
• surgical pathology
• examines all specimens removed during surgery to diagnose the type and extent of the diseases present
• typically very thin sections are cut from the tissue and are examined microscopically
• cytology
• examines specimens for malignant cell
• immunology
• measures antigens and antibodies in body fluids to diagnose infectious diseases and immunological disorders
• blood bank
• storing and distributing blood products
• tests donor and recipient bloods to ensure that the blood types are compatible and that only biologically safe products are released for transfusion
• optimum organization of the laboratory is determined by the type and size of the medical installation, the location, and kinds of tests performed.

2. INFORMATION FLOW IN THE CLINICAL LABORATORY

• Clinical laboratory refers to any type of lab in this context
• It responds to external demands for services
• Extralaboratory cycle: specimens and test requests are delivered to the laboratory, and test results are communicated to requesting physicians.
• Intralaboratory cycle: the analysis cycle - consists of steps to label each specimen with a unique accession number, to divide the specimen into aliquots (small portions), to deliver the aliquots to the appropriate workstations, to perform tests, and to prepare the report of test results.
• This coincides with a diagram in the text and what was done in class (p.278 of text)

2. HISTORICAL DEVELOPMENT OF LISs

• clerical duties accounted for 15 - 30% of lab personnel work time in 50s
• workload was low and the timeliness of results not crucial
• Late 50s: labs overburdened: tests were ordered but never completed, specimens misidentified, erroneous results were generated, reports misplaced.
• Automated lab instruments greatly improved the ability of the clinical labs to perform the tests in-house. (60s)
• generated large numbers of data, however, and overwhelmed the data interpretation processes within lab.
• 1970 Du Pont introduced the first computer controlled lab instrument, the Automated Clinical Analyzer (ACA).
• Currently, numerous instruments use microprocessors to control many phases of operation
• Tangible benefits of computers in Labs:
• decreased turnaround time from test request to reported result
• increased accuracy of test result and marked reductions in the numbers of transcription errors and misplaced results
• improved quality control and better monitoring of equipment
• more efficient storage and more timely retrieval of data used in research and teaching
• increased productivity per laboratory worker
• Greater availability of information used for administrative and managerial purposes

3. FUNDAMENTAL FUNCTIONS OF LISs

• paper slips sometimes were misplaced and were vulnerable to transcription errors and misinterpretation of handwritten results
• computers greatly improved the ability of labs to deliver results rapidly and accurately by allowing remote printing of info via devices located on the wards

2. PATIENT AND SPECIMEN IDENTIFICATION

• LISs assist in the identification process by producing specimen-collection lists by room number and by preprinting specimen labels

3. DATA PROCESSING AND RECORD KEEPING

• LISs tabulate test requests and prepare specimen collection lists
• computer generated worklists then help technicians to prepare specimens for analysis at individual workstations
• LISs also maintain tracking logs that allow lab personnel to determine when and by whom each step of the testing process was performed

4. DATA ACQUISITION

• test results must be entered into the LIS and displayed in a format that is meaningful to the healthcare professionals caring for patients
• The characteristics of the data determine the mechanisms used by the LIS for data acquisition and report generation
• Computer evaluates raw digital data to determine the final numeric result
• Results must be associated with individual specimens
• There is a one to one correspondence between specimen numbers and results
• LIS may acquire and integrate info from multiple instruments, store online, and produce reports of lab test results
• Characteristics of the testing process make data entry problematic and necessitate a flexible system for report generation.

5. REPORT GENERATION

• LISs can produce reports in a variety of formats that can help health professionals in interpreting results
• Often flag abnormal values
• printed in a time oriented flowsheet
• values could be displayed graphically to facilitate detection of subtle trends

1. QUALITY CONTROL

• purpose is to monitor the quality of work performed
• PRODUCTION ANALYTIC RUNS = comparing the measured and expected results derived from the standard samples, lab personnel can determine whether instruments are operating correctly
• Statistical analysis of control samples analyzed over time can help to define the accuracy and precision of test results, and can distinguish trends in patient data from problems with instruments or reagents
• ABSOLUTE LIMIT CHECK = someone reviews all test results outside certain limits to verify the accuracy of the value.
• DELTA CHECKS = identify instances in which the current result differs from the same patient’s previous test result.
• Important criterion for accreditation is documented adherence to strict quality control protocols
• Labs must maintain detailed records of all specimens that they receive and all tests that they perform.
• LISs can perform may of the record keeping tasks associated with QC activities

1. MANAGERIAL REPORTING

• LISs can perform various data collection and data-analysis tasks that help laboratory managers to evaluate current laboratory procedures and to suggest changes that will improve the laboratory’s efficiency
• LISs can collect and analyze info on test requests, including data such as specimen receipt time and test completion time
• LIS also can produce reports that help lab managers to evaluate departmental productivity
• TIME MOTION STUDIES can determine the average amount of time required to perform various test procedures within lab.
• LISs can keep records of the current inventory of blood products, the sources, and the expiration dates

4. CONTEMPORARY LISs

• Citation markets a family of LAN based, menu driven systems designed to support the fundamental information management tasks of the clinical labs
• supports real time data acquisition from automated testing instruments; record keeping; and report generation for chemistry, hematology, and urinalysis tests
• provides subsystems for : microbiology, blood bank, surgical and anatomical pathology, and business manager subsystem

1. PATHNET

• marked by Cerner Corporation in 1979
• expanded rapidly to become one of the major laboratory software vendors
• emphasis on management decision support
• wide selection of system options allows individual buyers to choose the configuration of services that best matches the lab’s operational needs

5. EXPERT SYSTEMS IN THE CLINICAL LABORATORIES

• computer based tools have also been developed to support the professional functions of the physicians and technologists who work in these settings

51. PATHFINDER/INTELLIPATH

• developed an expert system called Pathfinder, designed to assist surgical pathologists in interpreting the findings noted on microscopic examination of lymph-node tissue
• explored automated approaches to hypothetico deductive reasoning, using probability theory to assess belief in the likelihood of alternative diagnoses and decision theory to guide the questioning process
• designed to assist primarily with the clinical
• integrates an expert reasoning system for pathology diagnosis with a videodisc library of histological slides
• Has the ability to form a differential diagnosis of plausible diseases based on the histological features that have been entered into the system.
• uses the subjective probabilities of experts to infer the probability that each disease is responsible for the evidence that has been reported and orders the diseases by the relative likelihood that each is present

2. RED

• expert system helps techs in the blood bank to interpret the info derived from antibody-identification tests
• interpretation of the data is complicated, especially when multiple antibodies are present
• problem solving approach is similar to the hypothetico deductive approach used by physicians in making diagnoses
• uses knowledge gained from expert techs to classify antibodies in 4 categories: present, likely present, possibly present, and absent.

6. THE FUTURE OF LISs

• There are at least 3 ways in which LISs can help to reduce costs, while still maintaining the quality of services
• 1) Labour Costs
• 2) Computer based info management can reduce lab test costs by minimizing the number of tests performed
• 3) LISs can contribute to the decentralization of the clinical laboratories, thereby reducing transportation time and costs
• availability of instruments such as these will encourage the geographical dispersion of lab testing, especially where there is a high demand for testing and a need for rapid turnaround
• future LISs will be capable of assisting physicians in both the ordering and interpretation of lab tests
• Computer based system could reduce the risk of clinicians ordering tests inappropriately and of their misinterpreting lab data when caring for patients
• As LISs become integrated into HISs, and therefore have access to non-lab data, they will be able to provide more direct interpretive assistance
• In the future, we expect that knowledge based decision support systems will be increasingly incorporated into LISs
• LIS of future will be a distributed computer network in which microcomputers in instruments communicate with more powerful processors that collect data, generate reports, perform data correlation, and execute medical interpretive functions
• EXPANDED LISs WILL SHARE LAB DATA WITH CMPTR SYSTMS PERFORMING OTHER PATIENT CARE APPLICATIONS, AND WILL COLLECT DATA FROM A VARIETY OF SOURCES TO PERFORM DETAILED INTERPRETIVE FUNCTIONS

1. OVERVIEW OF PHARMACY PRACTICE

• pharmacy systems manage info related to drugs and to the use of drugs in patient care
• one of the earliest patient-care services to benefit from the development of medical info science and from the intro. of computers into healthcare settings
• drug preparation functions by manufacturer shifted role of the pharmacist; modern pharmacy practice comprises a predominantly patient-oriented, distributive set of functions that is characterized collectively as drug use control
• DRUG USE CONTROL defined as "the sum total of knowledge, understanding, judgments, procedures, skills, controls, and ethics that assures optimal safety and efficacy in the distribution and use of medication
• control of use of drugs implies surveillance from the moment the substance enters the healthcare system until the time that substance is eliminated from the body or is destroyed.
• Steps of the drug use process:
• Diagnose
• Determine drug history
• Prescribe
• Select product
• Dispense
• Counsel
• Administer
• Monitor
• Pharmacology = study of the properties, uses, and actions of drugs
• pharmacists, physicians, nurses, other health-care staff, patients cooperate to complete the steps of the drug use process and to ensure that drug therapies are beneficial, rather than detrimental, to the patients’ health
• Practice of pharmacy encompasses a large number of health care related functions performed in a variety of settings
• 4 types of pharmacies explained here:

1. INPATIENT PHARMACY SERVICES

• fundamental task of providing (dispensing) drug products in response to physicians’ orders for drug therapy
• "getting the right drug to the right patient at the right time"
• on receiving an order, pharmacist verifies it for accuracy and evaluates it in the context of what is known about the patient’s current drug therapy
• identify potential drug-drug and drug-food interactions, allergies, and other drug sensitivities that may adversely affect the patient
• The order as dispensed is recorded in the patients MEDICATION PROFILE, and drug delivered to the nursing station for administration to the patient
• There are two distinct approaches to the drug-distribution function:
• multiple doses delivered in a labeled container to the nursing station for the patients entire course of therapy
• Unit dose dispensing- each drug is packaged, labeled and distributed to nursing stations in its unit of use = medications for current administration period
• Additional responsibilities: acquisition of drug products, provision of information for billing, and maintenance of certain legally required records, such as those related to dispensing and administering narcotics
• Also compiles and manages the use of the drug formulary = document specifying which drugs are to be kept in the pharmacy’s inventory, and that details any restrictions on the use of these drugs

2. COMMUNITY (OUTPATIENT) PHARMACY SERVICES

• essentially the same functions as does the inpatient pharmacy, but must perform its tasks in a more open and complex environment in which it has much less control over the drug-use process
• in inpatient setting, the number of patients is limited to the number of beds (fewer than 500); community pharmacy provides drugs to 3000 to 5000 clients at any one time
• pharmacist conducts professional evaluations of drug therapies with respect to drug interactions and contraindications
• Most pharmacies now maintain medication profiles for their patrons
• community pharmacists educate and counsel clientele
• They are the most assessable of health professionals, and enjoy a high degree of public trust
• often are asked about general health concerns, thus serve as an entry point into the healthcare system
• Tasks include ordering products, managing inventory, billing medical-insurance companies, maintaining client charge accounts, tracking product sales, and preparing financial summaries

3. CLINICAL PHARMACY SERVICES

• relatively new area of practice
• focuses on assessing patients’ drug therapies than on the traditional dispensing roles
• work closely with physicians to determine the optimal drug therapies for patients, and often assist in monitoring the effects of such therapies
• also evaluate patients medical histories and review lab test results and other clinical data to detect and prevent therapeutic problems

4. DRUG INFORMATION SERVICES

• evaluation and dissemination of comprehensive drug information to the institutions’ staff and patients has been identified as an important aspect of pharmaceutical services
• answer drug related questions posed by the hospital staff (primarily physicians and nurses)
• pharmacists participate in the activities of the pharmacy and therapeutics committees by evaluating new drugs, developing drug-use policies, and maintaining drug formularies

2. HISTORICAL PERSPECTIVE - EVOLUTION OF PHARMACY AND DRUG INFORMATION SYSTEMS

• computers first introduced to hospital pharmacies in the mid 60s
• By 1975, approx. 7% of all hospitals had some form of computer service in pharmacy
• late 70s and early 80s, smaller hospitals began to install their own computer systems (standalone systems or were linked electronically to a central hospital computer)
• major improvements in communications capabilities brought about by LAN provided the means for pharmacy computer systems to share info and to communicate electronically with other hospital computer systems
• use of computers in community pharmacies lagged behind that in hospital pharmacies b/c of high initial and maintenance costs and the need for links to larger mainframe systems
• Today, pharmacy systems are among the most widely used computer systems in medical care
• B/C they provide a range of business, operational, and clinical functions that greatly improve the efficiency and cost-effectiveness of pharmacy practice, they have become a fundamental tool for the practice of pharmacy in many hospital and community pharmacies

3. FUNDAMENTAL CONCEPTS FOR PHARMACY SYSTEMS

• maintenance of complete, accurate, and up to date records concerning all aspects of drug therapy is requisite for high quality patient care
• Cost control strategies in the pharmacy focus on maximizing the efficiency of pharmacy personnel and on controlling tightly the acquisition, storage, and distribution of drugs
• also can support departmental decisions making by helping managers to collect and analyze pharmacy costs, worker productivity measures, and drug utilization information
• Dramatic increase in the amount of info available concerning drugs and their effects requires that personnel in drug information services regularly consult and review a wide variety of references

1. FUNCTIONS OF A PHARMACY INFORMATION SYSTEM

• pharmacy computers increase the accuracy and efficiency of drug distribution and perform routine medication monitoring and screening for drug interactions.
• functions aided by computer systems:
• online order entry
• pharmacist review
• medication profile update
• printed labels
• drug dispensing reports
• medication administration reports
• inventory maintenance and automatic drug reorder
• drug use review reports
• controlled drug reports
• Many hospitals provide some features to support business and managerial functions
• may generate bills for patients or third party payers and maintain the pharmacy productivity, sales volume, prescription activity, and drug usage
• many systems allow pharmacies to update their drug pricing and drug-drug interaction databases via a telephone line connection to a centrally maintained database

2. DRUG RELATED DECISION SUPPORT

• pharmacy systems play an important role in the provision of safe and effective therapy by maintaining accessible, legible, and up to date medication records
• more sophisticated systems can verify that appropriate monitoring tests are being ordered and can detect clinically significant changes in lab test results that may be related to drug use
• into of lab techniques for determining drug concentration levels in the blood and the development of pharmacokinetic models got interpreting the clinical significance of these levels allowed researchers to devise a new approach to dosing based on measured drug concentrations
• models that represent the drug levels over time as a function of the amount and timing of drug doses; the route of administration; drug specific pharmacokinetic parameters; patient factors; and concurrent therapies and disease factors that influence the drug’s effects
• Bayesian forecasting models that adjust the parameters of the pharmacokinetic model based on the results of drug level tests

3. CHARACTERISTICS OF PHARMACY INFORMATION

• advantage of pharmacy info systems (PIS) is that the info with which they deal generally is straightforward
• Drug names, for example, are easier to encode than are disease states
• basic functions of pharmacy practice are well-defined processes with definite inputs, and they constitute a stepwise progression of activities that results in a definable product

4. MODERN SYSTEMS TO SUPPORT PHARMACY PRACTICE

• support most of the functions discussed in Section 10.3.1
• some also support patient care functions, such as allergy screening and drug-dose calculation; research functions, such as creation and maintenance of historical databases; and financial management functions, such as cost accounting

2. CLINICAL PHARMACY SYSTEMS

• most exciting systems from a clinical perspective are those that directly help pharmacists and physicians to prescribe drugs and to monitor patients receiving drug therapies
• researchers in Montreal Children’s Hospital have developed a decision support system to assist in prescribing TPN for pediatric patients
• system interactively guides physicians through the prescription process and uses encoded rules to custom tailor its recommendations based on the patients age, weight, length of therapy, and previous prescriptions
• MEDIPHOR system developed early 70s was one of first computer systems to use a pharmacy’s medication records to provide online drug interaction surveillance
• HELP system, in addition to warning physicians of potential drug-drug interactions and adverse side effects, it also identifies abnormal chemistry levels, concurrent diseases, and other patient conditions that should be considered when particular drugs are prescribed

3. DRUG INFORMATION SERVICES

• Bibliographic retrieval systems are the primary computer based info systems used by drug info services
• These systems facilitate the provision of drug info services by allowing rapid access to the large body of drug related medical literature
• Other Bibliographic databases that contain pharmacological info:
• international pharmaceutical abstracts (IPA)
• excerpta medica
• BIOSIS Previews
• Pharmaceutical news index (PNI)
• knowledge bases can store expert summaries of the articles and related conclusions

1. A LOOK AT THE FUTURE OF PHARMACY SYSTEMS

• Pharmacy practice is in the midst of redefinition and transformation
• some pharmacists and pharmacologists have begun to call for expansion of the professional role of pharmacists to embrace greater responsibility in patient care and in the medication use process
• Changes in healthcare financing and delivery place pressure on health institutions, including pharmacies, to minimize costs while maintaining high quality patient care
• Likely that computer systems will assume many of the dispensing and distribution functions of the pharmacy as well
• computer routinely will screen orders for potential adverse interactions, flag complicated or unusual therapies for review by pharmacists, count out tablets and capsules automatically, and package them for delivery to patients
• Use of bar codes in the pharmacy will remove the last major technological obstacle to automated drug dispensing by computer controlled dispensing machines
• Bar codes even can reduce drug-administration errors-health professionals can verify that the correct drug is given to the correct patient by scanning both the bar coded patient identification bracelet and the bar coded drug label prior to administering a drug
• Advances in networking and communications technology also affect pharmacy practice.
• Automation of the drug order-entry process will allow quicker and more accurate filling of orders, but also provides the opportunity to give immediate feedback to physicians about their orders.
• Implementation of point - of care systems using bedside terminals to collect critical info, such as vital signs and drug administration data, will produce more comprehensive medical databases and will facilitate sophisticated, patient specific drug-therapy monitoring
• Online access to detailed clinical data and the application of artificial intelligence (AI) techniques to drug therapy monitoring will improve significantly the ability of such monitoring systems to detect and make recommendations concerning potential problems with drug therapy
• Integration of videodisc technology with computers will soon change the ways in which patients can learn about the drugs they are taking
• Future holds increased automation of pharmacy practice
• use of computer based systems to perform drug distribution, record keeping, and routine monitoring functions will simplify many time consuming tasks, and thus will allow pharmacists to emphasize the patient care aspects of their practices
• Computer based systems will better equip pharmacy services in both inpatient and community settings to provide high quality patient care in the face of growing pressure to control healthcare costs

• The primary function of a radiology department is the acquisition and analysis of medical images
• Activities of the department can be divided into four categories. Application of computer to support (or enhance) these tasks is immense. (pg 325)

1. Image generation:

• process of producing images using a variety of imaging modalities such as X-ray, computed tomography, and ultrasound.

application of computers.... (pg. 325 and 347)

• use of digital radiography, where images are digitized and data is recorded in computer memory rather than on film
• once the images are digitized, it’s easier to increase visualization of low concentrations of contrast material by using image enhancing techniques to visualize the image better
• can be manipulated for display or analysis in ways not possible with film-based images
• computed tomography(CT) provides a precise mapping of the internal structures of the body in 3-D space and also greatly improves contrast resolution.
• images produced from ultrasound can be stored in digital memory

2. Image analysis:

• interpret medical images

application of computers...

• Using digital imaging, radiology personnel can use computers to manipulate the images to allow better visualization of diseases. For example, adjusting the gray scale of an image to display those parts that you want to see better
• using image enhancement procedures to view images better. Example, digital filtering techniques to remove signal noise, enhance edges, sharpen blurry images, quantify images (e.g. make the heart bigger)

3. Image management:

• the storage and retrieval of the images

application of computers....

• digital acquisition of images reduces the physical space requirements, material cost and manual labor of tradition film-handling tasks.
• images can be transferred over communication networks.

4. Information management:

• involves:
• maintenance of the film library,
• scheduling of examinations
• registrations of patients
• [performance of examinations
• review and analysis of studies
• transcription of dictated reports
• billing for services

application of computers...

• computer based Radiology information systems to handle information-management tasks

Fundamental concepts for computer-based radiology systems: (pg 337)

• any image can be stored in a computer by converting it from an analog to a digital representation, or by generating it directly in digital form. It can be communicated over networks, stored in databases, and displayed on graphical monitors.
• all images can be characterized by several parameters of image quality:

1. spatial resolution: related to sharpness
2. contrast resolution: related to intensity
3. temporal resolution: measure of time needed to create an image

• ability to generate functional images
• although, computers cannot perform pattern recognition as well as humans do, still they some support in these subtasks of pattern recognition:
• global processing: computers can enhance an image for improving human visualization e.g. gray scaling a image. Humans cannot distinguish more than about 100 shades of gray but computers can
• segmentation: extracting regions of the overall image
• feature detection: extracting useful parameters of the regions e.g. volume of the heart
• classification: determining the type of object found e.g. classifying tumors etc.
• data compression abilities to store the images without taking too much space

Historical perspective: (pg 332-336)

Basically gives some history regarding development of medical-imaging modalities, automated interpretation of medical images and computer-based information-management systems.

1960’s: first applications of computer to radiology-information management developed

• radiology systems integrated with the overall HIS

Terminology: (pg 337)

digital image: represented in a computer by a 2-D array of numbers. Each element of the array represents the intensity of a pixel.

pixel: a small square are of a picture

voxel: a volume element

NOTE most of this was covered in the June 17 & 18 classes

12.1 What is Patient Monitoring? (pages 367-369)

1. patients with unstable physiological regulatory systems
2. patients with suspected life threatening conditions
3. patients with a high risk status
4. patients in a critical physiological state

1. acquire physiological data
2. communicate data from distant labs to the ICU
3. store, organize and report data
4. integrate and correlate data form multiple sources
5. function as a decision-making tool that health professionals may use in the care of critically ill patients

12.2 Historical Perspective (pages 369-373)

12.3 Data Acquisition and Signal Processing (pages 373-383)

Microcomputers in monitors has revolutionized the acquisition, display and processing of physiological data. Most monitors uses at least one microcomputer.

Some signals are already electrical (eg. heart signals recorded by ECG) and just need to be amplified and the analog signal converted to digital. Digital data is then processed and the results displayed. The sampling rate is important when digitizing a signal (see fig. 12.7). It is important to sample at a rate which captures all the features without distortion.

• digital computer’s ability to store patient waveform info such as the ECG permits pattern recognition and feature extraction. Classification of ECG arrhythmias is done through waveform templates.
• signal quality can be monitored and maintained
• system can acquire physiological signals more efficiently by converting them to digital form early in the processing cycle. Waveform processing can then be done on the microcomputer.
• transmission of digitized physiological waveform signals is easier and more reliable. Digital transmission is noise-free.
• selected data can be retained easily if they are digitized.
• measured variables can be charted over prolonged periods to aid with detection of life-threatening trends
• alarms from bedside monitors are now much "smarter" and raise fewer false alarms. Digital systems are not susceptible to signal artifacts as analog systems are ® fewer false alarms. Uses more than one signal to verify (eg. blood pressure can be measured by ECG or arterial blood pressure)
• systems can be updated easily - only software programs need to be updated

Computer based ECG arrhythmia monitoring systems were quick to be accepted. Computers generally use 16 bit architecture, waveform analysis and real-time cross-correlation techniques to classify rhythm abnormalities. The machine retains the tracings in memory for future review.

Modern algorithms for processing ECG rhythms take sampled data and extract features (the beginning and end of each of the P, QRS, and T sections). The algorithm compares the slope (used to detect the start of each waveset) with a threshold value. If the threshold value is exceeded, a trigger response signals the presence of a waveform edge.

The P wave is relatively difficult to detect. Lack of ability to reliably detect the P wave is a serious limitation of some machines. Noise in the signal is another problem. Patient movement can cause signal distortions triggering false alarms. The Marquette 7700 monitor has 4 leads and if it finds signal artifacts, it displays the word "noise" and waits 2.5 seconds before continuing to monitor. This cuts down on false alarms. The system compares each waveform to templates and generates an alarm if an abnormal rhythm is identified.

12.4 Information Management in the ICU (pages 383-391)

12.5 Currant Issues in Patient Monitoring (pages 391-398)

1. all ICUs should have arrhythmia monitoring
2. invasive monitoring should be performed safely
3. generated data should be correct
4. derived data should be interpreted properly
5. therapy should be employed safely
6. access to lab data should be rapid and comprehensive
7. entreal (tube-feeding) and IV nutritional support should be available
8. titrated therapeutic interventions that use infusion pumps should be available

• The development of inexpensive hardware and easy-to-use software, it is now possible to transfer some of this information-processing burden to machine.
• Offices are approaching the ideal of a completely paperless office.
• Computer applications can be applied to: office-practice setting, the organization of private practices, the types of services provided in office settings, the ways that ambulatory-care services are purchased, and the interactions between community practices and other segments of the health-care industry combine to make outpatient medicine a unique and rapidly changing environment.
• These physicians provide a variety of services, ranging from routine care for self-limiting diseases to long-term treatment for chronic illnesses.
• More recently, outpatient facilities, such as emergency-care centers and same-day surgery clinics, have begun to provide services that previously were available only in hospitals.
• Outpatient care is moving towards greater specialization.
• Multiple health professionals often co operate in caring for individual patients.
• Data sharing and communication among the various health professional involved in a patient’s care are crucial.
• To care for their patients effectively, physicians must be able to access the clinical information collected during hospitalization.
• Health-care professionals must collect, process, and record a large amount of information in a short period.
• Outpatient visits are relatively inexpensive, data entry and storage are high-overhead activities.
• A complete record of past medical problems and treatments is necessary to ensure continuity of care for patients over time.
• The retrieval of historical medical data is one of the critical information-management tasks in outpatient care.
• Incentive to physicians to provide more services in outpatient rather than inpatient settings.
• Outpatient care will continue to gain prominence in response to the cost-containment measures that have been imposed on hospital care.
• Office personnel locate, retrieve, and update patients’ medical records.
• Nurses collect and records baseline data, such as weight, temperature, and blood pressure, as well as much of the past medical history. Physicians gather additional data, form opinions about the patients’ problems, and devise treatment plans. This information becomes part of the visit note, which physicians either enter into the record directly or dictate for later transcription.
• Health-care workers need to ensure that the practice remains profitable.
• Several technologic, sociologic, and economic trends favour the development of computer systems for use by physicians in private practice. Rapid increase in the volume and complexity of medical knowledge, an aging population and the associated increase in chronic disease, and increasing documentary requirements by third-party payers create growing internal and external demands for gathering information. Furthermore, the trends away from general practitioners toward specialists, and away from solo practice toward partnerships and group practices, create a need for more effective storage and communication of information. The accessibility and legibility of the record also become crucial as more health professionals share medical data.
• Vendors must design their systems to fit key characteristics of physicians’ practices.
• Physicians will adopt a new technology only if it improves the quality or reduces the cost of the services they provide. Also may reject even a system with substantial benefits if using the system would increase the total amount of time spent on information-management tasks, or if setting up and learning the system would be excessively time consuming.
• Systems must be simple to learn and use.
• Computer systems must be practical, inexpensive, and easy to use.
• Health professionals manage 4 types of information:
• The medical information describing history of illness, clinical findings, diagnoses, and treatments for individual patients.
• The scientific information applied by physicians while caring for patients
• The administrative and accounting information involved in scheduling appointments, managing patient accounts, paying bills, collecting revenue, and performing the other tasks necessary to operate a small business
• The financial-management and practice-planning information that physicians use to assess the status of their practices, including the magnitude of revenues and expenses, the numbers and types of procedures they perform, and the demographic composition of the patient populations they serve.
• The patient’s past medical history is pertinent to most evaluations. This historical information describes the patient’s family and social history, as well as past medical events that are relevant to the physicians understanding and treatment of current problems.
• Medical practice requires the management of general information as well as of patient-specific information.
• ICD9-CM - The ninth International Classification of Disease- Clinical Modification for encoding medical diagnoses is used frequently. Originally, the codes were developed to classify the diagnoses assigned to patients in hospital care; however, this system has been extended to incorporate the symptoms and non-specific complaints with which patients in ambulatory care commonly present.
• Appointment scheduling is critical to the smooth operation of an office practice. An overly ambitious schedule, however, can seriously detract from the quality of patient care by restricting the time a physician spends on each case.
• A schedule must be flexible enough to accommodate unexpected events without drastically increasing the time patients spend in the waiting room.
• Collecting data needed to analyze the financial health of the practice, to evaluate the efficiency of the practice’s operation, and to predict future demands for resources.
• Financial reports and expenditure reports and income analyses provide a basis for judging the net productivity of a practice’s various activities.
• Some useful measures are the number of active patients; the number of new patients; the distribution of patients by age, by gender, and occupation; the frequency of each procedure performed; the revenue and expenses of services by volume; and the source and volume of referrals. Comparison of current and past measures may expose important trends that can affect practice.
• Information-management process:
• Patient calls and schedules an appointment.
• Sent to examination or consultation room where the physician and other health-care personnel collect the bulk of the medical information.
• The physician documents in the medical record the important findings, diagnoses, and treatments prescribed.
• Charge slip is given to staff and account is updated.
• Schedule a follow-up appointment, or a bill for service is issued.
• Laboratory-test can be done or ask questions about progress.
• Office sends monthly billing statements to patients that have outstanding balances.
• Paperbased system:
• Moderately inexpensive and easy to use, Require little training, and Flexible.
• Disadvantages: Patient charts can be lost or misplaced. Charts can be incorrectly filed or hidden.
• The static organization and inflexible display of information are other shortcomings of the traditional paper chart.
• Besides being labour-intensive, reliance on transcription and manual calculation increases the likelihood of bookkeeping errors.
• Powerful and affordable microcomputers made it financially feasible for even solo practices to own computer systems.
• General-purpose PC software assist with many of the information-management tasks. (p 424) Have also been used to generate bills and insurance invoices in small practices. Preparation of letters and reports, access online databases of biomedical literature using communication programs and commercial networks, schedule patient visits using personal time-management software, keep track of inventories using spreadsheet programs, accounting programs can be used to manage the payroll and accounts-receivable transactions in the office.
• Drawbacks of a general-purpose software is that it must be custom-tailored to meet users’ requirements. Also may not be powerful enough to fulfill specialized requirements.
• These systems can provide greater versatility and convenience, once the system has been custom-tailored. They are, however, more difficult to learn and more expensive.
• Development of microcomputer technology expanded the functionality of the standalone systems that first appeared in the previous decade.
• Systems fully integrate the basic financial functions, and also patient scheduling and inventory control.
• Special-purpose systems tend to be relatively inexpensive and easy to use.
• Types of information-processing tasks: (p 425)
• Appointment scheduling – electronic appointment books are available to assist office personnel in scheduling appointments.
• Patient surveillance and recall – Office personnel can run the surveillance program periodically to identify patients for recall.
• History taking --
• Risk assessment – help physicians to evaluate a patient’s degree of health risk based on individual characteristics and the results of diagnostic tests.
• Decision support – provides a fast and convenient means for physicians to reference diagnostic and treatment information for hundreds of medical conditions.
• Medication lists and prescription writing – assists physicians in the task of prescribing and writing prescriptions.
• Patient education – produce printed instructions for patients; summarization of drug-use instructions and describe drug effects and side effects in easily understandable language.
• Bibliographic retrieval – aids in searching the medical literature
• Communications – facilitates communication among health professionals by providing access to electronic bulletin board for the exchange of information and ideas related to general medical knowledge, to specific issues, and to special interest groups.
• Computer-aided instruction –allows physicians to gain experience with a variety of different patient problems.
• Business Systems – performs most of the business-related tasks;
• Graphical interface technology simplifies interactions between the operator and the machine. GUI provides easy access to data and generally cause only modest reductions in speed of data retrieval.
• The complexity of the software and the necessary hardware considerable increase the operating and maintenance costs of these systems.

• Computer systems will improve as new technology is developed and existing technology is applied more effectively. The efficiency of available mechanisms for data entry currently limits the role of computers in doctors’ offices. Accessing medical information often is more difficult when data are stored electronically than it is when traditional paper-based record systems are used. Coding schemes for recording medical information will allow for more efficient entry, storage, and retrieval of clinical information. The current trend is away form solo practice and toward partnerships and larger group practices. The standardization of record keeping enforced by a computer-based system will help to minimize the difficulties in data access and communication that arise as more and more physicians join group practices. Development of a system for electronic publishing, computer-aided instruction for patients, clinical consultation, and other potential applications for outpatient care.

• Medical informatics refers broadly to issues in the storage, management, retrieval, and use of medical information.
• For many years, they relied on personal libraries of books, journals, and articles to which health professionals could turn when they needed information. It soon became clear, however, that the sheer volume of written material made the biomedical literature unusable without special auxiliary methods—indexes and, in recent years, computer systems with an ability to search the literature for pertinent references.
• Written indexes => Computer-based literature management.
• Direct personal communication with peers and experts frequently is the easiest way to gain assistance in problem solving.
• Gatekeeper: a person who maintains a high level of external communication and, through contacts, keeps colleagues informed of new developments.
• Traditional guides to the literature are printed indexes, catalogues, and abstracts that identify the subjects and authors of articles and books relevant to the reader’s interest.
• If is important that the bibliographic apparatus not only list what is available, but also indicate where items are located.
• Computer technology is being used to facilitate rapid and convenient access to biomedical literature.
• Computer technology as a means of dealing with a huge and rapidly growing files of data and information in biomedicine.
• MEDLINE, the National Library of Medicine’s major bibliographic computer-based database, contains more than 900,000 references to the recent journal literature; the medical literature continues to be a vital link connecting research, education, and practice.
• Online bibliographic searching: is much faster, it is more cost-efficient, and it allows for selective retrieval of items most relevant to user requirements. Also comprehensive, complex searches that would be impossible to do using printed indexes.
• Organization and systematic means for access, allow information saved at the NLM to be usable.
• Dr. John Shaw Billings developed the Index-Catalogue of the library of the Surgeon General’s Office; it included both books and periodicals. Terminated 1961.
• Index Medicus provided both subject and author listings of journal articles, and soon became an indispensable guide to the world’s medical literature. 1927-1931 known as Quarterly Cumulative Index Medicus.
• Manual card system was replaced by punched cards; punched cards enabled the use of sorters and collators for alphabetical filing. Index Medicus was produced for the first time by the NLM computer-based Medical Literature Analysis and Retrieval System (MEDLARS); Achieved automated publication of Index Medicus, but also made possible the development of a computer-based search and retrieval service.
• MeSH was developed to provide the categorized lists of headings and subheadings (and cross-references). The MeSH vocabulary is routinely updated to reflect the constantly evolving vocabulary of biomedicine.
• Unified Medical Language System (UMLS), based on MeSH, is directed toward making the myriad of other biomedical classification systems invisible while providing a single logical path to broad range of biomedical information sources.
• MEDLARS went online and became MEDLINE in October 1971; This system provided in-depth indexing with an enlarged vocabulary, as well as online retrieval in real time.
• MEDLINE have immediate access to a database of about 6 million citations.
• MEDLINE is an interactive system; the user identifies the journal-article references needed by carrying on a dialogue with the computer.
• Other Biomedical Database Systems:
• BIOSIS - scientific journals covering world-wide literature on research in the life sciences, as well as books, monographs, and meeting and conference proceedings.
• Excerpta Medica - a broad-based information-retrieval service that indexes and abstracts biomedical and clinical literature; an online database.
• PSYCHINFO - Covers the world’s literature in psychology and related disciplines in the behavioural sciences.
• CA SEARCH - Provides international coverage of the chemical literature from more than 14,000 scientific and technical journals, as well as from conference proceedings, technical reports, books, and patents.
• SCISEARCH/SOCIAL SCISEARCH- Online counterparts of Science Citation Index and Social Science Citation Index. The data they contain indicate the frequency with which published articles are cited, and they identify articles that are potentially related to one another because they cite the same reference.
• Telecommunication networks that make it technologically and financially feasible to link users’ terminals to remote computer centers or to host systems containing the databases.
• Host systems maybe owned and operated by the same organization that produces the database contents. Host systems are frequently commercial enterprises built specifically for the purpose of vending a database service to users.
• Vendor Systems typically contain a multitude of databases that are produced by a variety of organizations.
• Search Intermediary, specialist/librarian who interprets the information requirements of the search requester and translating these needs into the particular command language and vocabulary of the search system being used.
• The ready availability and proliferation of relatively low-cost microcomputers and mass-storage devices- in combination with the introduction of user-friendly access- has had a revolutionary effect on bibliographic-retrieval systems and practices, as has had on many other facets of the workplace and marketplace.
• Database vendors have responded to the increased interest in end-user searching in two ways: by training users; and by designing systems that, while retaining as much as possible of the power of command-language systems, can be used by people who have no special training.
• User-friendly interface is designed for use by individuals who are not computer specialists who have had only limited training, and who might reject a system that is non-intuitive or complicated, or that uses unfamiliar jargon to communicate. In the case of bibliographic-retrieval systems, such interfaces may be simplified versions of their conventional counterparts, or may be a brand new search systems. These interfaces generally use a few basic commands that are a subset of the more powerful and versatile conventional search systems used by the intermediaries, and they have human-engineering features that make them relatively non-threatening to persons unfamiliar with computers.
• End-user searching is more convenient than is intermediary searching, and it allows end users to capitalize on their special knowledge of the subject matter.
• Bibliographic database consists of a set of records, each of which uniquely characterizes a document in terms of author, publication source, and subject matter.
• The most important element in this characterization involves the assignment of index terms, or descriptors, which help to define the content of a document.
• Indexing plays a crucial role in this process; effective assignment of index terms links documents that are related, discriminates broadly among subsets of documents within the entire file, and discriminates finely among individual documents.
• Such systems promise not only to improve retrieval relevance, but also to decrease the cost of database production.
• Artificial Intelligence (AI), such as natural-language interfaces and conceptual indexing techniques that semantically structure the contents of a database, have the potential to achieve this objective to an even greater extent.
• Three ways in which AI methods will find useful applications in the field of literature management. P461
• "Intelligent" retrieval assistance.
• Expert advise systems based on literature references.
• Programs that "read" the literature – natural language understanding.
• User-friendly interfaces is intended to make computers generally more accessible to people who have no special training.
• The focus of AI tends to focus on how a computer can be programmed to "understand" natural-language commands and queries in much the same way as a well-informed human would.
• Conceptual indexing organizes an information base by the meaning of the entries; the system actually understands the information it contains.
• Traditional measures of system performance such as precision and recall should show the corresponding improvement.
• Precision - ratio of relevant to irrelevant records retrieved.
• Recall - ratio of relevant records retrieved to all relevant records in a database.
• The use of an intelligent system that is capable of reading and understanding the items to be added can potentially achieve greater indexing consistency and fewer errors.

The goal of clinical research is to advance the state of medical science -- to add to the base of knowledge that guides health professionals in caring for individual patients -- and thus, ultimately, to improve the practice of medicine. In general, clinical researchers must be content with observing interventions and the resulting outcomes as physicians try alternative therapies to help patients regain health. The two central tasks of clinical research are data management and data analysis.

Clinical research typically comprises the following steps:

1. problem formulation
2. hypothesis generation
3. experimentation
4. model development
5. hypothesis testing
6. evaluation and dissemination

In prospective studies, researchers define the experimental conditions before collecting any data. They specify the parameters of the study, define methods for enrolling subjects, design protocols for delivering alternative treatments, and define the criteria for measuring outcomes. In retrospective studies, clinical researched analyze existing datasets, rather than collecting new observations. Often, they conduct chart reviews to glean research data from clinical records.

The most desirable form of clinical research is the randomized clinical trial (RCT), a type of prospective experiment in which patients are randomly assigned to alternate groups, and are treated according to a study protocol. RCTs are attractive because they are designed to avoid experimental bias -- differences in experimental outcome that are caused by factors other than the experimental therapy. Well designed and properly conducted RCTs produce the least bias when testing a hypothesis -- they are the gold standard for clinical research. However, they are time consuming and expensive to conduct.

Retrospective studies provide an alternative to RCTs. If a suitable database exists, researchers can avoid the long and expensive process of enrolling qualified patients and collecting data.

The larger the study, the greater the need for computers to assist in the research process. If there are more subjects, there are more data to collect, manage, and analyze. Information systems with a research orientation may include direct links to statistical-analysis programs. Statistical applications typically provide a variety of capabilities, including descriptive statistics, graphical presentations, and analytic statistical techniques. Researchers use descriptive statistics to search for unexpected events and to support the generation of hypotheses. Graphical presentations extend the capabilities of descriptive statistics, and they can elucidate hypotheses about trends and correlations among variables. Analytic methods are used primarily to verify hypotheses.

Three types of elements are necessary to describe a clinical event adequately for research purposes: 1)variables that identify which events are being studies, 2)variables that describe the study intervention and the experimental outcome, and 3) potentially confounding variables that could affect the experimental outcome.

The choice of identifying information must be specific enough to identify each observation uniquely within a set of data. Time is a critical element for the identification of clinical data. Many clinical studies are longitudinal -- therefore, they timing of the event must be recorded with the data that describe that event.

Data are represented in a variety of forms in a clinical database. Some measurements are best stored as numeric values. Other information may be encoded according to a classification scheme or may be stored as free test. The choice of representation determines the types of operations that the computer can perform on the data. Further, the choice of representation depends on a data element’s intended use.

The quality of data can be assessed on three dimensions: correctness, completeness and consistency. Correctness refers to the accuracy with which events are records and entered into the computer. Completeness refers to the prevalence of omitted events and missing observations. Consistency refers to uniformity in the meaning of data encodings within the database.

Multiple institutions may join to create databases to be pooled and to participate in cooperative clinical trials. Although multi-institutional studies solve the difficulty of gathering sufficient data, they aggravate problems of data inconsistency. Research centers must define a common vocabulary to characterize patients for collaborative cross-center studies. One way to build a large database is to pool smaller databases from several independent studies. Although pooling increases the overall sample size, data quality tends to suffer. The individual studies may have used different criteria for inclusion of patients in the experiment. Also, since the specific data sets were collected to satisfy different objectives, some studies may be missing values for one of more variables of interest.

y = b₁x₁ + b₂x₂ + ... + E

Researchers encounter a variety of problems when building and testing models. Among the most common are the following: omitted variables, incorrect model specification, large unexplained variance, and correlation among independent variables.

3. History of Clinical Research Systems

This section gives a long and excessively detailed history of some American clinical research systems.

As computer based medical records become more common, the body of clinical data readily available for research grows. Currently, some institutions are adapting their medical record systems to support research as well as patient care.

4. Current Clinical Research Systems

Computer systems that support clinical research offer three basic options: 1) databases developed in-house, 2) database available commercially, and 3)specialized systems for clinical research. Do-it-yourself systems offer researchers the most flexibility; the systems can be custom tailored to meet the goals of the research project. General commercial DBMSs require little programming and adequately support data entry, retrieval, and long term storage.

5. The Future of Clinical Research Systems

There is a growing demand for the collection and analysis of large bodies of clinical data. The existence of online databases will facilitate research on new approaches to analyzing clinical data. To address this increasing demand, researchers are developing methods for automating both the summarization of medical records and the discovery and validation of new medical relationships.

1. Automated Abstraction from Clinical Data

Because a single disease can be associated with multiple abnormal findings, many clinical data are correlated. Thus, much redundant information exists, even in incomplete medical records. The automated abstraction of higher level conceptual entities – disease syndromes – from low level data (such as laboratory test values and physical findings) could help to reduce the problem of missing data, which is so common in clinical research. There are three objectives of research in automated abstraction:

1. to deal with data overload by reducing many data items to a much smaller number of items of information
2. to compute automatically the intensity and certainty associated with each finding
3. to mitigate problems caused by missing data

In summary, technological advances in computer hardware and software have decreased the costs of data storage and processing, and thus have reduced the barriers to long term storage of rich clinical data. The development of new techniques for analyzing longitudinal data also encourages the creation and maintenance of large clinical databases. In the future, such databases will become increasingly valuable resources as researchers rely on computers to capture, store, and analyze large numbers of clinical data.