Capture, Coding and Quality Management of Information

From SystemsWiki

Jump to: navigation, search

Contents

Introduction

This Section - and the following 4 Sections of the course aim to take the student through the a conceptual framework which starts with the sources of health information, then addresses how information is captured and coded and represented followed by Sections on the processing, representation, sharing and delivery of information in a variety of settings. Two later sections then address the use to which the information may be used to improve the health system and make a difference. The last two Sections then explore the practicalities of making all this happen in reality, given the complexity of the Health Sector as revealed in the first two sections.

Major Learning Objective

The key to this section is to understand just where the information sourced in the last section is held, how it is coded and then how its quality is assured and maintained. In the next two sections a detailed review of the functionality and capabilities of the systems will be provided, while in this section the different types of systems will be enumerated and briefly discussed, followed by a more detailed discussion of terminologies and coding and concluding with some discussion of the issues of information quality, reliability and integrity.

General System Characteristics

What follows assumes the reader has reviewed Chapters 4 and 5 of the prescribed text as introductory and contextual material.

Ambulatory Care Systems

These systems provide a clinical / administrative core to support the operations of ambulatory care providers (e.g. general practitioners, office based specialists and outpatient clinics (emergency and planned care). Typically the patient record will hold the information outlined in Section 2 and provide easy access to functions that allow medication and diagnostic test ordering, electronic messaging for referrals, prescription transmission, results receipt as well as various forms of clinical reminders and clinical decision support. Typically - in the stand alone environment (e.g. General Practice) - as well as the patient record system there is billing capability provided.

Inpatient Clinical Systems

These divide rather easily into general purpose systems to allow clinicians (doctors, nurses etc.) to plan and provide routine care (which are much like the systems discussed above) and the systems that support others to undertake specialist activities (from laboratory operation to nuclear medicine and all sorts in-between).

Hospital Management Systems

These systems cover the operational aspects of the delivery of health care. Conceptually there are three major areas addressed.

Patient Administration and Scheduling Systems

These system manage patient admission, discharge, transfer as well as scheduling of activities for the patient while in hospital. Also included are such things as dietary management systems.

Business, Inventory and Staff Management and Rostering Systems

These are the traditional administrative systems typically with some tweaks to make them work well in a health institution.

Management Information and Reporting Systems

This covers the basic summary reporting systems all the way to such niceties as management information dashboards etc. The intent with such systems is to provide summary information in ways that are useful for management while at the same time permitting analysis of progressively more granular data if that is what is needed to understand what has been causing an observed trend which is of interest. It is in this area that costing systems are considered. It is important that there be relatively seamless flow of information between each of these areas to permit the sensible linking and use of patient activity, patient coding data and financial data is brought together to develop understanding of the relationships between activity, costs and so on. Integration and interfacing of these system to achieve the desired information outcomes can be very complex and difficult and does need to be carefully planned.

Minor Systems

This can cover everything from customer relationship management systems (which can handle donation management, customer satisfaction surveys, traffic management and so on. In each case the purpose is obvious in the system name.

General Computer System Characteristics

All computerised information systems can be seen as having four generic and a number of ancillary functions. The four generic functions are information capture, information storage, information processing and information output / display. To provide these functions there is used a mix of software and hardware components.

Information Capture

In terms of information capture the hardware involved can be such things as keyboards, scanners and so on. Software such as word processing is relevant here as are tools such as optical character recognition can also be used to convert scanned documents back to electronic text.

Information Storage

For information storage from the hardware perspective we can think of optical and magnetic disks, solid state memory of all sorts and so on. From the software perspective data-base managers and associated tools are clearly very important.

Information Processing

In processing hardware we think these days largely of microprocessors (using singly, in physical racks and then almost invisibly as part of various cloud computing services). Support and management of information flows in an out of the processor is a main task of the operating system software.

Information Output

For information output think things such as printers (laser, inkjet etc.) as well as software to facilitate the presentation of information (all the way from Powerpoint to managerial dashboards) Underpinning all this are, of course, are the system components that support all the functions described above - the operating system, the power supply and so on. Please see the following for a basic introduction to how these various parts all hang together in a real system. Put another way (from a colleague with a strong IT background): “Computer systems can do three things:

  • 1. They can process information, either by following commands supplied by a user working at an input device or by following pre-defined rules,
  • 2. They can store information in files or data bases, and
  • 3. They can provide access to information that is stored or which exists only temporarily.

The most critical issues are:

  • 1. The commands and rules that a computer system uses to process information,
  • 2. The structure of the information such that it can be stored, accessed and retrieved efficiently, and
  • 3. The access controls around the information in the system.

Computer systems can make access to information much easier. This is a two edged sword because, at the same time as making it easier for appropriate people to access information, it can also make it easier for in-appropriate people to access that same information. Controls that were sufficient for information that was hard to access (say in a doctor's filing cabinet), will probably not be sufficient when that same information is in a central data base that thousands of people can search rapidly. The information in a computer system may or may not reflect the external reality. Significant efforts need to be made to ensure that the information is accurate when it is created but also that it remains accurate. People change their addresses, their names, their marital status; they become ill, they become better; people are born, they die. The quality of information in a computer system can very easily degrade. Decisions made by health professionals should always be made with this in mind. Human based systems can be created on the basis of defining "normal" or "typical" processes and letting people's common sense detect anomalies and errors. Computer systems need to have comprehensive and rigorous error and exception checking built in from the start. Computer systems can do things very much faster and can do more things in a short space of time. This power must be carefully controlled, because when things go wrong they can go wrong much faster in a computer system than when done manually.” As can be seen from even this very short summary issues such as information quality and data integrity as well as data security come very quickly to mind. These will be explored in more detail in the rest of this section.

Clinical Coding and Clinical Terminologies

Textbook Reference - Chapter 9 Page 210 forward. Fundamental to the ability to process clinical information using computers is to be able to represent the various components that make up the clinical record in a way that is computer interpretable and processable. The approach that is adopted to achieve this is to associate a numerical code with a diagnosis or clinical concept.

Clinical coding

Coding / classification of clinical information is undertaken for three main reasons.

  • The first is to facilitate aggregation of statistical information for internal purposes,
  • the second is to support clinical costing (ICD-10) and
  • last for reporting to Government and internationally.

Typically the coding systems used are as follows:

  • ICD.
    • The background to this global clinical coding scheme is found here
    • This classification system is used by the World Health Organisation to produce globally comparable statistical information on disease morbidity and mortality.
    • In Australia the Commonwealth Department of Health and Ageing provides support for ICD.
    • The details are available here
    • The Australian National Casemix and Classification Centre (NCCC) website is found here
    • The NCCC is responsible for the adaption and maintenance of the Australian variants of ICD-10.

General Practice Classification Systems

  • Because of the widespread automation in the GP sector a range of general practice classification systems have evolved and been implemented over the last few decades. At base they aim to improve information management and statistical reporting but often also have other secondary objectives and purposes.

Possibly because of patchy use of these coding systems, in late 2012 the Commonwealth revised its incentive program for General Practice. This program can provide up to $50,000 per annum in payments to a practice for adoption of specific GP Computing initiatives. This is referred to as the ePIP program. Details of the program are available here

  • Of the five requirements for these payments one is put thus.

Data Records and Clinical Coding “There are no software conformance requirements for this Requirement. This requirement is a work process change associated with whichever medical vocabulary used by the vendor (such as SNOMED-CT, DOCLE, PYEFINCH and ICPC2+). Practices must ensure that where clinically relevant, they are working towards recording the majority of diagnoses for active patients electronically using a medical vocabulary that can be mapped against a nationally recognised disease classification or terminology system. Practices must provide a written policy to this effect to all GPs within the practice.” This requirement is fairly vague but clearly the intent is to - over the long term - encourage GPs to attach codes to clinical diagnoses which can be used by the system to assist with disease statistics, outcome tracking, clinical decision support and so on. In each case the code has the purpose of converting a diagnosis which may be expressed a number of different ways (e.g. Heart Failure, Congestive Heart Failure, CCF etc.) into a single numeric or alphanumeric code which is more easily managed by computers.

  • More information of each of these classification systems can be found on the following sites (not an exhaustive list).
  • Each of these classification systems vary in the way the sub-divide diagnoses and well as having different overall knowledge frameworks - albeit they are attempting to address the same problem of the variability of language representation compared with a numeric code.

Clinical terminologies

Related but not exactly the same a clinical coding systems are clinical terminologies. The key difference, as I see it, is the purpose for the creation of the code set or terminology. Clinical coding typically has administrative or statistical purposes while a terminology attempts (and attempts is the important word here) to convert written clinical concepts and diagnoses into unique numerical representations. The ambition is that the proper use of clinical terminology will facilitate clinical decision support and enable the clear, error free communication of clinical concepts thus enabling a very high level of transferability and interoperability between clinical records. Put simply the clinical terminology aims to make the meaning of any clinical concept clear and unambiguous for any purpose. However, many years of development and work have revealed that to actually develop a comprehensive, unambiguous, robust alphanumeric link between clinical concepts and the associated codes is a great deal harder than it seems. Indeed there are some who would argue that any clinical terminology will have inevitable compromises which may render them unsuitable for some purposes.

  • The best known and by far the best supported and developed clinical terminology is SNOMED-CT.

There is a useful entry on Wikipedia which is found here It is clearly beyond the scope of an introductory course to attempt to delve into the complexity and subtleties of terminologies of the scale of SNOMED CT.

  • The following paper from 2008 by Professor Alan Rector will provide any reader interested in the complexity of this area with more than enough to get started with.
  • The global organisation which maintains SNOMED-CT is found here.

Summary

In summary, for the purposes of this introductory course in Health Informatics any more detail is unnecessary beyond the student being aware such coding and classification systems exist - and that under the covers these systems can be very complex and difficult to implement with simple billing coding systems being relatively straightforward and expressive, meaningful clinical terminologies being extraordinarily complex.

Information Quality

  • A well-known saying in the technology domain is ‘Garbage In, Garbage Out’. What this means is that unless you have reliable accurate information provided to any form of computer system the information and reports the are produced will essentially be useless. For reliable information to be provided by any computer system it is necessary that the inputs are accurate and that the processing of the information is done correctly.
  • From the perspective of Health IT this broad principle leads to a range of areas that need consideration.
    • The data sources (as reviewed earlier) will typically either be from manual human data entry - where such things as error reducing interface design and clear display of entered information is vital - and information feeds from supporting systems and / or instruments such a laboratory analysers. Clearly the manually entered information is the most prone to error in these circumstances.
    • Once information is captured (correctly) obviously it is important that it be processed correctly by properly tested and validated software. This is obviously important but a less likely cause of problems in established operational systems. Another obvious issue at this point can be the selection of wrong or inappropriate codes for clinical information which will reduce the value and accuracy of the information outputs.
    • At the other end of the processing flow we also need to be aware of the of the importance of understandable and usable presentation of information and reports deriving from systems as well as the proper presentation of information as it is being used as the basis for decision making or interpretation.
  • From this brief discussion is it obvious that there are many opportunities for both a lack of accuracy as well as mis-interpretation of information as it is collected and processed. All the above also assumes perfectly functioning systems and no corruption or similar of the information held within systems.
    • It is this aspect that it is termed ‘information integrity’. Attacks on information integrity can come from a range of issues from technical failures, computer virus attack, malicious human intervention and so on.
  • The reader is encouraged to read Chapter 2 of the recommended text at this point to obtain a fuller understanding of Health Data Quality issues.

Return to HI Homepage

Questions & Comments to Geoff McDonnell
Personal tools