developments

DEVELOPMENTS

NV MA D u t c h

A s s o c i a t i o n

f o r

M e d i c a l

R e c o r d s

A d m i n i s t r a t i o n

BACK UP ON KODAK M I C R O F I L BREAK THE Y E A R 2 0 0 0 BARRIER.

Millennium-bug jitters. Technology obsolescence. Costly migrations. These days, choosing storage media can be risky business. But not when your source documents and computer files are archived on Kodak Microfilms. For a fraction of a penny per image, you get back-up that can be retrieved and read or scanned into the next century and beyond. No wonder so many insurance companies, financial institutions, and government agencies are rediscovering the true value of microfilm. Whether you use a service bureau or archive your own records, be sure to ask for Kodak Microfilms and processing with Kodak chemicals. They can't be beaten for consistent u quality and clarity. And our 9k free quality assurance If programs help make sure _ things stay that way for < a minimum of 500 years.* Plus, you're automatically protected by the Kodak Disaster Recovery Program at no extra charge.**

il': • !

t

! !

Make sure your documents are secure. Archive them on Kodak Microfilm, the world's leading brand. For more information visit www.kodak.com/go/docimaging. And put your future on a roll. * Certified life expectancy when processed and stored under specified conditions. ** When processed in Kodak chemicals al a Kodak-authorized processing facility. Please ask for details.

BUSINESS IMAGING SYSTEMS niMitv. 1WS KodA i-

information CIP-lnformation Royal Library, The Hague Health Information Developments in the Netherlands [et al.; red. W.A. Dekker] Amsterdam, The Dutch Association for Medical Records Administration Robert A. Stegwee With. lit. ref. ISBN 90 - 802225-5-0 Keyword: Health Information Development Organized by: M.J.G.M. Dekker W.A. Dekker J. Runnenberg The Organizers like to thank: Academic Medical Center, Amsterdam Robert-Fleury Foundation, Leidschendam Medical Center Alkmaar, Alkmaar University Hospital Vrije Universiteit, Amsterdam Moret Ernst & Young, Utrecht PMC Advice, Apeldoom TAB Products, Amsterdam Kodak, Hagestein AGFA, Rijswijk SIG, Utrecht Care for Care, Woerden Ondersteuningsgroep gezondheidszorg, Nijmegen Orga Info, Woerden Synergie Consultancy, Den Haag Lifeline Networks, Capelle a/d Ussel Innovit, Weert RINO, Utrecht Marcelis & van der Lee BV, Alkmaar Main sponsor: Jalema, Reuver Their support was of great value to this production. 1999 NVMA - The Dutch Association for Medical Records Administration M.J.G.M. Dekker/W.A. Dekker/J. Runnenberg Lay-out and composition:

Han Runnenberg, Hoorn

Cover design and paintings: Ad Werner, Amsterdam designer, architect and artist Concept Stones:

Aat Veldhoen, Amsterdam artist

Lithography:

Leander Werner, Amsterdam graphic designer

Prepress / Press:

Marcelis & van der Lee BV, Alkmaar

All rights reserved. This work may not be translated or copied in whole or in part without permission of the publisher. HEALTH INFORMATION DEVELOPMENTS IN THE NETHERLANDS - APRIL 1999

Health Information Developments in the Netherlands SPECIAL ISSUE OF THE DUTCH ASSOCIATION FOR MEDICAL RECORDS ADMINISTRATION

Journal Advisory Board Jacques van der Palen, Robert A. Stegwee

APRIL 1999

Contents 3

Preface

J.J.N. van der Palen

4

Health care information standardization efforts in the Netherlands

R.A. Stegwee

NTMA-staff

10 Niaz-accreditation in Dutch hospitals

J.C.F. van Oijen P.A.E. Sillevis Smitt sr. E.M.S.J. van Gennip

17 Clinical protocol support with web-technology

M.C. Vissers A. Hasman

22 The quality control system trial project Accreditation in a central medical filing department

S. Schermer Voest

28 Zouga development and implementation of an Intranet based epr architecture

M.G. Ros B.J. de Vries

35 The psychiatric case register Rotterdam region

A.I. Wierdsma C.J. Dieperink G.T. Koopmans

42 A short review of the development of nursing standards

J.Th. Kedzierski

46 Developing an indicator system for a health care centre

U. Nabitz P. Nanitsos T. Verkooyen

52 Developing of a task-based reminder system for critical care environments

P.A. de Clercq J.A. Blom A. Hasman H.H.M. Korsten

63 Medical protocols supporting in-hospital cardiological care

C.M.B. Duwel D.A.C.M. Kruijssen

69 Protocols and guidelines in a hospital Intranet

J.W.van der Slikke W.A. Schats

75 Edi and the Internet

A.U.N. Jadoenathmisier

79 Selections of items for the nursing minimum data Set

W.T.F. Goossen T. Feuth P.J.M.M. Epping M. Cousijn

Managing Editor

Han Runnenberg Associate Editor

Ans van Manen

Design Ad Werner

Production Matthieu Dekker Wybe Dekker Han Runnenberg

Advertising Representative Matthieu Dekker Robert-Fleury Stichting 2260 AK Leidschendam telefoon +31704441022 telefax +31704441097 email: [email protected]

NVMA Office Adresses Matthieu Dekker Leidsestraat 86 2182 DR Hillegom, the Netherlands telefoon +31704441022 telefax +31252531995 Wybe Dekker Meibergdreef 9 1105 AZ Amsterdam, the Netherlands telefoon +31205662501 telefax +31206919854 email: [email protected]

NVMA's World Wide Web home page http://home.wxs.nl/~nvma

HEALTH INFORMATION DEVELOPMENTS IN THE NETHERLANDS • APRIL 1999

preface J.J.N. VAN DER PALEN Rl President of the Dutch Association for Medical Records Administration

Dear readers It is with great pleasure that the Dutch Association for Medical Records Administration (NVMA) presents a second selection of the developments in the field of care administration and Electronic Patient Records. In the Netherlands, the demands made on the quality of health care continue to increase, requiring the deployment of the latest technology within diagnostics and therapy but particularly also in the field of information processing and communication. Technological innovations only become truly fruitful when they are correctly introduced into the health care institutions. All care personnel and auxiliary staff within the health care system must be able to apply the new technology well. That requires not only good education and information but also attention for the implementation projects. Future users must become involved in design of the structure within which the innovations are introduced. This implementation problem is by no means unique to the Netherlands. On the contrary, all countries are wrestling with the same issues, which is why it is useful to exchange experience on projects. Our magazine wishes to help you in that process by sketching a picture of the developments within the Netherlands, whether in the initiating, design and construction phases, or already in implementation and evaluation. You will read practical stories, written by people from your profession. The NVMA is proud of the initiative and efforts taken by the editors of this magazine, and is grateful for the sponsoring by all the companies who are featured in this magazine or elsewhere. We hope to provide you with a good picture of the developments in the Netherlands and, at the same time, invite you to request further information on any subjects relevant or interesting to you. The editors will be pleased to point you in the direction of the right addresses. I wish you every success in your work within the health care system.

HEALTH INFORMATION DEVELOPMENTS IN THE NETHERLANDS - APRIL 1999

J.J.N. van der Palen

Right page: Acrylic on canvas 100x80 cm

Overview

health care information standardization efforts in the netherlands ROBERT A. STEGWEE, Moret Ernst & Young Management Consultants

Robert A. Stegwee

As most of the work in health care organizations is information-based, health care should be regarded as an information-intensive industry. In order for information to make sense in the communication between health care professionals, these professionals have to agree upon the content and meaning of information. The basis for this mutual agreement is formed during the staffs initial or continued training. As such, professionals have been educated in the terms and procedures to use when communicating with each other. However, increased interdisciplinary collaboration, regulations on patient data registration, advanced epidemiological research, and the extended use of information and communication technology by professionals in health care has led to a myriad of different vocabulary and communication processes. In order for effective communication to take place within one health care organization or throughout a health care network, a certain degree of standardization has to be agreed upon. This contribution provides a frame of reference for different types of standardization initiatives in the Netherlands. The individual contributions in this special issue of the Dutch Association for Medical Records Administration illustrate a specific category of these standardization efforts.

A broad view of standardization Within the context of this contribution, standardization is viewed as a process to develop and implement a standard. A standard can be defined as 'something established by authority, custom or general consent as a model or example.' [1] As such, standards can govern a multitude of objects and activities, and can be specified at a number of different levels. For example, the accreditation of health care organizations, as described in [2] and [3] relies upon a set of standards with respect to the kinds of rules and regulations that have to be present to effectively monitor the quality of care within (parts of) the organization. In addition to these standards, the way in which the accreditation process is carried out is also subject to standardization. A next step in the standardization process incorporates the specific protocols and guidelines developed within the organization for the actual delivery of care, illustrated in [4], The way in which such guidelines are to be authorized and distributed is described in [5]. Part of these guidelines may include standards with respect to the registration of specific patient data and the coding schemes to be used for such a registration. These coding schemes then tie in to standardization efforts carried out with a more epidemiological background, such as the mental health care case registry [6], Also, patient data standards are used as input to • HEALTH INFORMATION DEVELOPMENTS IN THE NETHERLANDS - APRIL 1999

•p****^^^

either context sensitive protocol systems, such as ProtoVIEW, described in [7], or knowledge based decision support systems, as illustrated in [8]. These knowledge based decision support systems themselves specify standards which apply to the provision of care to the patient. Other uses of standardized patient data can be found in specific indicators to monitor patient care performance and quality [9]. Finally, standardization on a technological level will facilitate the electronic interchange of health care information within the hospital [10] or between health care providers [11].

Focus of standardization efforts One of the difficulties in standardization is the need to standardize on a number of different levels. Not all standard development organizations will be able to develop or govern a standard for all necessary levels, as required by a practical application within the specified objective. Often, they will rely on standards developed by other standard development organizations or will postpone the specific standardization of elements to the implementation within the health care organization itself. It is helpful to focus on a specific level of standardization, relevant to the objective of participants involved. Such a differentiation between levels of standardization is very common in information technology, see e.g. the ISO standard for Open Systems Interconnection (the ISO/OSI model) [12]. Key to such a layered approach to standardization is a clear definition of the different layers. Based upon the contributions to this issue, a preliminary frame of reference is developed for the standardization initiatives in the Netherlands.

of standardization. This can be a mixture of four basic elements: the data, the process, the technology, and the systematization (a description of the related set of standards needed to achieve a given goal). The primary aim of the standardization effort provides us with a third dimension. Broadly speaking, the aim of a standardization effort can be characterized as: provision of care, management of care, medical research, and epidemiological study. Finally, the field of application of the standard leads to the fourth dimension of standardization. For example, a standard may relate to a specific group of patients (diagnosis related), to a specific aspect of management (e.g. quality control), or a specific professional group (e.g. radiology). Table 1 lists a number of sample standards in health care (referring mostly to contributions to this issue), with a description of their position on each of the four dimensions mentioned. Within a specific practical application, a number of standards will usually have to be applied in order to achieve a fully operational situation. For example, in order for a hospital to comply with the PACE accreditation standard for Medical Records Administration, one has to implement several (possibly local) process and data standards. Process standards (i.e. procedures) govern, among others, the request, delivery and return of medical records. Data standards are necessary for unique patient identification, for medical record request (minimal required dataset), and for monitoring the process, e.g. date and time of request, delivery and return.

Combining standards

Dimensions of standardization An obvious dimension of standardization is the organizational scope of the standard: is it a local standard within a specific part of the organization or, at the other extreme, is it an international standard. Between these extremes one can differentiate between institutional, regional and national standards. A second dimension can be found in the object

When several standards are needed, it is necessary to combine these standards wisely. Some standardization efforts will run into challenges for standardizing objects that are not within the scope of the project itself, but are necessary for practical application. In such a case, some people will invent their own local standard for tackling this problem. For example, let's take a look at a national effort to •

Table 1: Examples of standardization efforts Standard

Organizational scope

Object of standardization

Primary aim

Field of application

PACE

national

systematization

management of care

quality control

ZOUGA

institutional

technology

provision of care

point of care

REBUS

regional

data

epidemiological

mental health care

HL7

international

data and process

provision and management of care

data management

Clinical protocols

local

process

provision of care

patient groups


standardize the clinical protocol for the treatment of myocardial infarct. Part of such a clinical protocol may prescribe the need for certain blood tests to be administered. If these tests are not specified according to the LOINC terminology (Logical Observation Identifier Names and Codes), the validity of the protocol for other institutions is limited. Also, a future interface of the protocol to an automated test ordering system will be hampered, requiring additional specification of the protocol. Naturally, there is a certain amount of 'the chicken and the egg' dilemma in this situation. If all sorts of new standards are considered prerequisite for a project, not much progress will be made. The example clarifies the use of the reference framework presented above. The project at hand (the specification of a clinical protocol) can be characterized as a national standardization effort, with the process as its object of standardization, aimed at the provision of care, and confined to a specific patient group. In the course of the project, a standard is needed (to specify the blood tests) which is clearly outside the specified scope of the project. A standard for coding blood tests is a subset of a standard pertaining to data (the object) with respect to laboratory test orders (the field of application). When choosing an appropriate standard for representing the necessary blood tests, it is helpful to look for standards that already have been developed for the issue at hand. In this case, one might consider three different standards: - the local lab test order codes of the hospital; - the billing codes for lab tests (a financial standard in the Netherlands), or; - the LOINC vocabulary. The choice between these standards depends on the fit between the characterization of the project and the characterization of the standard. The local lab test order codes are not suitable, as they do not conform with the national scope of the project. Although defined at a national level, the billing codes for lab tests are aimed at the management of care, not at the provision of care, and as such they might lack the specifics needed for the clinical protocol. LOINC seems to be well suited, as it is aimed at the provision of care and has a (inter)national scope. However, LOINC has not been widely adopted in the Netherlands, so the project faces some difficulty here. They might settle for generic descriptions of blood tests (as used in textbooks) and leave the specific conversion to local lab test order codes to the implementation within each individual hospital. In general, however, looking for additional standards which match the criteria of a given project, may help in determining a useful combination of standards to be used.

Conflicting standards A source of conflict within or among standards can usually be traced to an overlap in scope or aim of the standardization effort. An interesting example is the product definition project in The

Netherlands ('producttypering'). The project is intended to arrive at a national definition of products within hospitals. However, the primary aim of the project seems to be unclear. The use of diagnosis related groups (DRG's) has been considered, but the participants preferred a more recognizable product definition from a medical point of view. Some medical specialists wanted to categorize their patients according to their complaints, whereas others would like to stay close to the final diagnosis of the patient. In view of the relationship between products and the financial budget of the hospital, hospital staff would rather identify groups of patients with a comparable coststructure. The role of products in the operational management of the hospital is likely to trigger yet another line of discussion: comparable care profiles for the patients within one patient group (product). From a medical point of view, a product definition should relay important information with respect to the provision of care (which illnesses or complaints has a patient been treated for). However, other participants in the project aim at the management of care, although with different fields of application (cost and operational planning). At the moment, the only reliable product registration is based on clinical care and is coded using ICD-9 (International Classification of Diseases). From an epidemiological point of view this is very worthwhile, but the support for this registration within individual hospitals is very limited. This example shows the necessity of clarifying your choices on the dimensions of a standardization effort beforehand.

Resolving conflicts Some conflicts between standards cannot be resolved. However, it is worthwhile to try and tie different standards together. This can be done in a number of ways. First and foremost, standards should refer to other standards when the object of standardization clearly lies outside the bounds of the standardization effort. For example, the HL7 standard (Health Level 7, referring to the seventh layer of the ISO/OSI model) is used to facilitate the electronic information interchange between computer based information systems in health care. As such, it has to standardize the data that has to be exchanged (messages) at specific moments in time (events). In table 1, HL7 has been categorized as an international standardization effort with respect to data and process, aimed at the provision and management of care, and with data management as its prime field of application. Wherever possible, the standard refers to other standards, such as ICD or LOINC, in the definition of particular data elements. Also, there is no mention of lower level protocols for the technical implementation of the standard: HL7 works well in local area networks, over the Internet, or by exchanging tapes or floppy disks, taking into account the various restrictions of these media. With respect to the process standards within HL7, little or no reference is made to other standards. Due to the nature of HL7. such •



process standards would have to deal with broader organizational issues, and few internationally accepted standards exist for organizational processes within a health care institution. However, when designing process standards within a hospital, such as a patient admission procedure, one should take into account the events defined by HL7, in order to avoid conflicts during the implementation of the process standard using computer based information systems. A second way to resolve conflicts between standards is to provide a translation mechanism. For example, local test order codes or LOINC may well be translated into national billing codes, provided there is a manyto-one relationship between these codes. A simple translation table will do the job in this case. The same is probably true for product definitions and ICD codes. In other cases more sophisticated (possibly rule-based) translators may provide the conversion from one datastandard to another. As a matter of fact, designers of standards should take into account the relationships between different standards, in order to provide reconcilable conflicts between standards. Currently, a considerable amount of energy is committed to international efforts to reconcile different standards.

Coordination of standardization As standards and standardization projects are becoming more prominent, the need to coordinate these efforts is being felt. Probably due to the extensive use of information and communication technology in health care, combined with the pressure to reduce administrative cost, the overlap and conflict between standards is surfacing. Standards have been developed bottom-up, with a specific goal in mind, but the use of the same information is spreading well beyond the original goal of the standard. Hence, new standards are being developed, sometimes conflicting with older ones. In some cases, standards are developed without any knowledge of the existence of other standards. The key national organizations within health care in the Netherlands have decided to coordinate their standardization efforts with respect to information exchange. A central organization (CSIZ) serves as a forum for different standard developing organizations and provides a directory of current standards and their developers. In addition, CSIZ facilitates discussions on a national roadmap for standardization of information in health care. It is stipulated in the mission statement that the CSIZ does not itself develop new standards; this is left to existing standard developing organizations or newly established ones. Therefore, it depends upon the individual standard development organizations to what extent their standards will combine or conflict with each other.

Conclusion Promising results can be reported from various standardization efforts in The

Netherlands. The range of projects is varying from standards for the national accreditation of hospitals to local technical standards for information interchange. Specific challenges can be found in the role of different standards within a single project. Considering that the standards for accreditation rely heavily upon the specification of protocols and procedures with respect to the processes carried out in health care, one can expect an increased interest in these process standards. Hence, the way in which information and communication technology can facilitate the formulation, distribution, and use of protocols becomes an important issue. Solutions range from intranet applications, via context sensitive protocol viewers, to the incorporation of protocol support in electronic patient record systems. Based upon this protocol support, data can be gathered specifically aimed at decision support for clinical health care. In turn, this leads to new standards for coding different data elements in the electronic patient record. Fortunately, this structured coding also enables a wider range of medical research, based upon the recorded data. Ideally, information in the area of management and epidemiological study can be distilled from the electronic patient record, provided the medical standards combine well with standards in other fields of application. We come full circle when we realize that performance and quality indicators are needed to exercise adequate control, and that the use of such indicators is required by the accreditation standards. Indeed, coordination between various standardization initiatives is needed, as well as extensive education to this effect. The users of standards and the standard development organizations will have to be made well aware of the need for such coordination. Steps are taken to accomplish these tasks within The Netherlands. However, we are all at the mercy of the individual professionals in individual health care institutions: will they be able to put the standard into practice? [I] [2] [3] [4] [5] [6] [7] [8] [9] [10] II1] [12]

Webster's New Collegiate Dictionary. The contribution by van Oijen, Sillevis Smitt sr, and van Gennip. The contribution by Schermer Voest. The contribution by Kedzierski. The contribution by Van der Slikker and Schats. The contribution by Wierdsma. The contribution by Vissers and Hasman. The contribution by De Clercq, Blom, Hasman and Korsten. The contribution by Nabitz, Nanitsos and Verkooyen The contribution by Ros and De Vries. The contribution by Jadoenathmisier. For a description refer to the literature on computer networks, such as Computer Networks by A. S. Tanenbaum.

Correspondence Robert A. Stegwee Moret Ernst & Young Management Consultants email: [email protected].


NIAZ

accreditation in dutch hospitals J.C.F. VAN OIJEN, TNO prevention and health P.A.E. SILLEVIS SMITT SR., Chairman Netherlands institute for accreditation of hospitals E.M.S.J. VAN GENNIP, Director Netherlands institute for accreditation of hospitals

m

NIAZ tuul voor Accraditahe /an ZiBk«nhuizftn i a

In January 1999 the Netherlands Institute for Accreditation of Hospitals (NIAZ) has started. This institute is set up by the Netherlands Hospital Association, the Netherlands Association of University Hospitals, the Netherlands Organization of Medical Specialists and PilotprojectACcrEditation (PACE) Foundation. The aim of NIAZ is to allow hospitals to request (voluntary) an accreditation of their organization by peer review surveyors. During the last ten years the process of accreditation is developed. In this period of time a hospital wide NIAZ standard Quality System and 35 NIAZ-PACE standards of organizational units in hospitals were developed together with I 7 Dutch hospitals. Also peer review surveyors carried out pilot-accreditations in 19 hospitals. Introduction The Netherlands Institute for Accreditation of Hospitals (NIAZ) is launched at January 1999. Dutch Hospitals can request NIAZ for an accreditation. NIAZ makes use of a peer review system. This means that the surveyors, who are working in other hospitals, visit the hospital that wants voluntary to submit an NIAZaccreditation. In this article is explained how NIAZ perform the accreditation process. First of all some (background) information is given, which answers questions like 'what is the legal context in the Netherlands', 'what is the aim and organizational structure of NIAZ?', 'how are the standards developed?', 'what kind of standards are available?'. Also NIAZ accreditation is described from an informational point of view. Questions like: 'what kind of information is used in an accreditation?', 'what is the result of an accreditation?', 'how is managed that the information for the accreditation is useful for the surveyors to work with?', 'what are the experiences in pilotaccreditation from this point of view?' will be answered.

Legal context For a better understanding of the approach of NIAZ, some knowledge of the legal context in the Netherlands is useful. The two most important laws related to quality in health care in this respect are the Law on Professional Practice (BIG) and the law on Quality in Health Care Institutes. The aim of the BIG (1994) is to control and enhance the quality of the health care delivery by the individual professionals working in the medical field. It regulates titleprotection and lists titles of medical professions that are protected by law. It enforces mechanisms such as re-registration, disciplinary law and peer-review and makes professional bodies responsible for adequate training and conduct of their members. The law

thus enforces self-regulation within certain borders. The law on Quality in Health Care Institutes is endorsed April 1996 and concerns health care institutes. The law formulates four requirements: - health care institutes should provide responsible care; - the organization of health care institutes should be such that it allows the delivery of responsible care; - health care institutes should systematically monitor, control and improve the quality of care; - health care institutes should account themselves for their quality management in a public annual quality report. In conclusion: the Dutch law has clearly defined the responsibility of the health care organizations for the quality of the organizational aspects, and the responsibility of the health care professionals for the professional quality of the health care they provide. However, the Dutch law does not specify a given model or a specific set of standards to be applied. Instead, it is based on the principle of self-regulation. Standards or models to be applied and mechanisms for control should be developed by the sector itself. Therefore, NIAZ is working with its own standards, developed specific for the Netherlands. NIAZ has accepted the standards, which are developed in the course of the Pilotproject ACcrEditation (PACE).

The Netherlands Institute for Accreditation of Hospitals The development of the accreditation system for hospitals in the Netherlands started in 1989. The PACE foundation - a co-operation of 17 general and university hospitals - developed a system of 35 standards and accompanying I


accreditation guides (the questionnaires of the standards). The standards cover the various organizational units, which are common within hospitals, i.e. Radiology, operating Room, Daycare etcetera The standards are based on those of the 'Canadian Council of Health Facilities Accreditation' (1988) and the ISO 9000 series. The standards were not translated, but developed specific for the Dutch situation, based on the critical points identified within the Dutch hospitals, by the people in the hospitals themselves. The Health Insurance Community financially supported the development of the standards.

Advisory Council

1

The NIAZ, established and operational in 1999, offers Dutch hospitals the possibility of evaluation of the quality of their organization by external peer review. The aim is to assess whether the structure and conditions for quality care are available. An accreditation can be requested for a hospital as a whole or for parts of the hospital. The peer review surveyors are trained by NIAZ and are recruited from the hospital management. The accreditation is done on the basis of information supplied by the hospitals, including a self-evaluation report of the hospital as a whole and self-evaluations by the departments. In accreditations, the NIAZ-PACE standards are used as a frame of reference. Hospitals can request for a NIAZ accreditation, no matter which system they apply in quality management. The NIAZ institute is the only one of this kind in the Netherlands and is established and financially supported by the Netherlands Association of General Hospitals, The Netherlands Association of University Hospitals and the Organization of Medical Specialists. TNO Prevention and Health (the Dutch Organization for applied research) has coordinated the development of standards and accreditation system for PACE since 1989, supported the Precouncil and now supports NIAZ. The Director of the NIAZ is delegated from TNO PG. The organizational structure of NIAZ is shown in figure 1. For the first year 1999, one hospital has requested a hospital wide accreditation by NIAZ.

—

Board of Appeal

Director

—

Panel of Experts

Execute Accreditation Procedure

Quality Declaration Committee Fig. 1: Organizational structure of NIAZ

Standards In 1997 the Netherlands Association of General Hospitals, the Netherlands Association of University Hospitals, the Netherlands Organization of Medical Specialists and the PACE Foundation decided to set up a Netherlands Institute for Accreditation of Hospitals. The three organizations, together with the PACE foundation, set up a Precouncil, which prepared the foundation of NIAZ. In the course of this preparation, one additional, hospitalwide standard was developed (Quality System) and two more pilot-accreditations have been carried out in 1998.

Foundation NIAZ Board

and accreditation

guides

The NIAZ 'umbrella' standard Quality System In addition to the standards of organizational units/departments in hospitals, an 'umbrella' (hospitalwide) NIAZ standard 'Quality system' has been developed in 1998. In this standard, TQM concepts (derived from the EFQM model) are included. Within the accreditation system, the standard Quality System should describe the minimal requirements for accreditation. In 1999 a project has started to refine the NIAZ standard Quality System, as the minimum requirements for accreditation are yet not specific enough. The contents of the NIAZ standard Quality System is: - management and organization of hospitals (mission, vision and strategy, quality management); - management of cure and care; - management of personnel, quality education of personnel, devices and supplies; - quality assessment, quality improvement and quality assurance; . associate with internal and external clients, handling of complaints; . participation in internal and external survey programs; . management of quality documents, instructions and protocols. NIAZ-PACE standards of organizational units in hospitals Evaluation by NIAZ is based on a set of 36 standards, one hospital wide NIAZ standard Quality System and 35 NIAZ-PACE standards of organizational units in hospital (table 1). The standard Quality System should describe the minimum requirements and the 35 NIAZ-PACE standards are used as a frame of reference. The NIAZ standards are developed 'bottomup'. First, in one department, in one hospital, critical points were identified and a draft standard is developed. The standards only describe what should be organized, not how this should be done. This draft was put forward to similar departments in ten other hospitals. Next the draft standard was open for comments on a national level for all that wish to comment (including patient organizations, insurance companies and national societies). Quality


•

©RGA/NFO •

ADVIES EN MANAGEMENT BV

Orga-lnfo Advies en Management vertaalt samen met de klant integrate vraagstukken naar werkbare en samenhangende operationele-, (be)sturings-, en informatieprocessen. Orga-lnfo is goed bekend met de specifieke problematiek op het gebied van:

inrichtingzorginformatiesystemen organisatie van de informatievoorziening proces- en ketenanalyse en inrichting implementatie van informatiesystemen

Orga-lnfo is een onafhankelijk bureau met ruime kennis van en ervaring in de werkgebieden van overheid en gezondheidszorg, alsmede bij bedrijven in de segmenten van zakelijke dienstverlening (o.a. verzekeringen), handel en industrie. Orga-lnfo heeft 80 adviseurs in dienst en is gecertificeerd lid van de Raad van Organisatie-adviesbureaus (ROA).

CSt

Voor meer informatie: Orga-lnfo Advies en Managementbv Vijzelmolenlaan 3 Tel: 0348 - 436262 3447 GX Woerden Fax: 0348 - 436236 Postbus 505 E-mail: [email protected] 3440 AM Woerden Internet: www.orgainfo.nl

officers of TNO PG co-ordinated the process. After the standards had been formalized, they were translated into questionnaires: the accreditation guides. Also this translation was done in close co-operation with people within hospitals. Recently, the answering possibilities of the accreditation guides have been upgraded according to the Deming cycle. The current answering possibilities are: 1 plan: planned, not arranged; 2 plan - Do: planned, arranged, not tested; 3 plan - Do - Check: planned, arranged, tested, not improved; 4 plan - Do - Check - Act: planned, arranged, tested, improved.

NIAZ accreditation procedure Based on the experiences in the 19 pilotaccreditations from 1996-1998 the process of accreditation has been evaluated and improved. In the largest pilot-accreditation, in the University Hospital Maastricht in 1997, a team of six surveyors assessed 50 departments (25% of the total) in this hospital during five days. From 1999, NIAZ performs accreditations of hospitals on their own request. The NIAZ-accreditation in hospitals consists of two phases. In Table 2 a rough survey of the steps of an accreditation is given.

In the accreditation the two main players are the hospital and the surveyors. During the accreditation they have to work and Frequently hospitals use the standards or their communicate a lot together. In Figure 3 is accreditation guides themselves as instruments shown how the work/input of the to develop their quality system, but this hospital and surveyors interact. is not mandatory. On the left is shown what the surveyors do within The contents of the NIAZthe process of PACE standards are Quality system uniform, they all accreditation and describe criteria on the right the concerning: Management and Organization work/input of - management hospital. and organization; The - process standards control; and - devices and (electronic) supplies; accreditation guides are - knowledge open to the and skills; public. For - quality this reason the assurance; standards and the accreditation guides are The relation between available on Internet these five topics is (www.niaz.nl). Also the shown in Figure 2. working procedure of an accreditation, which is Fig. 2: Topics in the standard and described in table 2, is available how they relate to each other and the management of quality on Internet. •

Table 1 NIAZ-PACE standards of organizational units in hospitals Labor and Environment Blood Issue in hospitals Admission Department Central Medical Archives Central Sterilization Civil Service Daycare Dietetics Economic Administration Department Ergotherapy Physiotherapy Mental Care Hygiene and Infection Prevention Computerization and Automation Purchase and Logistic Instrumentation Service Intensive Care Clinical Department: Treatment- and Nursing Department

Laboratory (Medical) Welfare Work Medical Staff Neonatal Care Nuclear Medical Science Research Department: Function and Nursing Operating Room Patient Information Service Personnel and Organization Outpatient Department Board of Directors/Management Radiology Urgent Medical Aid / Assistance Technical Affaires Nursing Service: Nursing- and Staff Department Dietary Advice Department Hospital Pharmacy


in healthcare Using simply just one approach to solve all IT problems quickly and successfully is impossible. Moret Ernst & Young Healthcare Industry Group therefore uses an integrated approach in which various IT-experts work in close co-operation with specialised auditors, lawyers, tax advisers and management consultants. Projects include computerized patient records, hospital information systems and digital radiology. Together with the Systems Development & Technology advisory group they provide optimal solutions for our Healthcare clients. This has proven to be a way that works. For more information please contact Robert A. Stegwee, e-mail: [email protected], telephone +31 30 258 87 52. You Can e x p e c t m o r e f r o m US.

=U MORET ERNST & YOUNG

•'Computerization enables us to focus attention on the patient. At

directly sent on to the operating room by means of a wireless connection. As

present, I feel so powerless when a patient has to wait this long before he can

soon as the patient is brought in, we can begin. It can make the difference

have an x-ray or scan taken. He is often kept in suspense for weeks. Soon, it

between life and death."

will only take one day in most cases to help a patienl. And thanks to carelully devised computerization, the hospital organization is also better than

Computerization offers many challenges, resulting in a better

ever before. We will probably have more possibilities from a medical point

atmosphere at work, more efficiency and higher standards. The people from

of view as well in the near future, but health care in particular will improve

Care lor Care - doctors, nurses, paramedics, managers and, last but not least,

in quality because data will be more accessible, whether the patient knows

information systems analysts - are familiar with the logistics and dynamics of

the hospital or not. In emergency cases for instance, all basic examinations,

health care by practical experience. Therefore, we would like to give you our

blood sample, E.C.G. etc., can be taken in the ambulance. The data can be

view on your computerization. Built on the future - whatever it may bring.

Care for Care, efficient in computerization, experts in health care. Zaagmolenlaan 4 • 3447 GS Woerden, the Netherlands. Tel. [31 348] 489300 • Fax [31 348] 489301 • E-mail: [email protected]

L CARE S O I 1 W A R E 5.' t ( ) \ M I. I A \ C V

Table 2 NIAZ accreditation process (in steps) The following steps are carried out in the NIAZ-accreditation in a hospital 1 A hospital voluntary enrolls itself at NIAZ for an accreditation. The appointments between NIAZ and the hospital are described in an accreditation contract. 2 The hospital writes a self-evaluation report based on the standard Quality System and the accompanying questionnaire. 3 Surveyors judge the hospital at first sight based on the self-evaluation report (PHASE 1). They determine which sample of the organizational units will be evaluated in depth. They consider whether NIAZ can recognize certificates obtained by other organizations for organizational units of the hospitals. 4 The selected organizational units of the hospital fill in the relevant electronic accreditation guides. 5 The surveyors investigate the quality system of the hospital in practice by analyzing the score of the electronic accreditation guides, by analyzing quality documents and by a visit (survey) of the organizational units, which they have selected. During this visit they speak with people in the department and inspect documents on the shop floor (PHASE 2). 6 The surveyors provide a report. 7 After the Quality Declaration Committee of NIAZ has read the report, they advice the Board of NIAZ whether or not to grant an accreditation certificate. 8 The Board of NIAZ decides to give the accreditation certificate or not. The Board, based on the advice of the Quality Declaration Committee determines the period of validity of the accreditation certificate (in general three years). The Board can also decide that the hospital has to implement specified improvements within a period of time, before the accreditation certificate can be given.

possible stopmom ents stop

Registration f

Surveyors

\ r

Self Evaluation Report

Check

stop Preliminary Examination

Preparation Audit

stop Preparation Audit Audit Reporting Follow-up Fig. 3: Interaction between the work of the surveyors (left) and the hospital (right)

Conclusion After a thorough preparation of almost ten years, the NIAZ is launched. With this institute, the Dutch hospital directors and the medical specialists have chosen for peer review as an important instrument for quality assurance. The national involvement of these major players provides a solid basis for the Dutch accreditation system. While realizing that an important milestone has been reached, development should continue, in order to assure that NIAZ can continue to support good hospital care in the Netherlands. -


emergency department. All interns who regularly worked F.r- WdewM To P w . \ T O l ' On Hit Web '»•» *«•* in the Accident & a D u s ©o ©is IU L*J LKJ 2 a a sjaai m CUCD @QDSI a a usi Emergency (A & E) department of the Fraclure of Radial Head University Hospital in MftchanU i of injory: Ajriiv^j^hlhr ra Jial . ' ^ ^ n s o n*&v t»? &acr>tfcJ by dir-:i Maastricht (as part of n o l i m c Judi a; a fall i t W'^w on the side *f the Jratfcui, with the radial h*ad imkjng the capiteStuai asked to participate in ite oi p*m BJ the eft>ew and 6»i« may t.* loc noimnatifinTha this study. In total , Modemtfi may tK>l t>? ^parent tnlil th« hff3^ t; ro!^rcd ijndtr ihc c^cvmnm£ timenb (gtntly pTonat^ a twelve interns, divided itie ta-iai head The fau^f 'ji f*oiiaOi-ii atid sufmatoii n ui ually reitrictci into six pairs of two, Kramplf radhn head rrarturp fl participated in the study. After a brief introduction into ProtoVIEW each pair of interns was asked toadfracturwvnD dii>w m the staadv4 Ixitfii mi to answer eleven (fiipiiub

13043255510 Sir*h

Fl

«1

Unit. Ph»». I D . .

End

1 1 1 2 2 2 11 Day468 0 2 4 2 4 6 8 0 2

The combination of 9 low potassium value (< 3 mmol/IJ and digoxin can be potentially hazardous for the patient.

Monitor

Lab SOP T!SS

Provide rationale * 7bG>*Uw*#'"'

Add Medication

Stop Medication

Change Medication

Info

Pumpdose

Figure 2. When a guideline is not followed, the reminder system generates and displays a reminder, shown overlaying the user interface of the PDMS.

Figure 1. A guideline's expression in the KBE. The Potassium pop-up window shows particular properties of the potassium laboratory test.

This particular expression evaluates to true when a laboratory test, carried out less than 2 days ago, returns a potassium concentration that is lower than 3 mmol/l and the medication Digoxin has been prescribed as well. As a result, a reminder will be issued. The expression consists of instantiated objects from the ontology class tree, which are selected from the tree and dragged to the right window (this operation implements the AND-operator). It is also possible to edit a term's properties, depending on its representation in the real world. Finally, propositions that consist of multiple terms can be created by dragging a term from the ontology tree atop an existing proposition in the right window (implementing the OR-operator). Implemented guidelines are validated by testing them on a large patient data set of previously admitted patients. ICU physicians then evaluate the outcome of this validation. Based on this evaluation, the guidelines may be updated. When the ICU physicians have approved a guideline, it is exported to the reminder system to be used in daily practice. Using implemented guidelines, the reminder system checks for inconsistencies between the guidelines and the patient data stored in the PDMS. When an inconsistency is found, the corresponding reminder is generated and displayed by means of a pop-up window, overlaying the user interface of the PDMS (figure 2).

In order to validate the helpfulness and correctness of implemented guidelines, the pop-up window also allows PDMS users to classify the reminder as 'correct' (the reminder was accurate in this case) or 'incorrect' (it was a false alarm) by examining the corresponding data of the guideline in the PDMS. When it is not possible to classify a reminder, using the patient data, the reminder is labeled 'inconclusive'. Also, the user may provide a rationale to explain his or her evaluation of the reminder. Abstracting guidelines to task specifications Besides their use to provide decision support to the ICU health care workers, guidelines can also be viewed as instances of task specifications. Although guidelines are independent symbol-level modules, various guidelines may have similar characteristics. For example, guidelines that handle drug-drug interactions (undesirable combinations of drugs) have the same format (figure 3). Name: A Combination of aBeta-Blockerand Amiodaron is undesirable. Author: Korsten Type: Reminder Category: Undesirable drug combinations Validation: Test Explanation: See literature reference: [Zagola GP. and MacGregor D, The Critical Care Drug Handbook 2nd Ed. pp. 195.1997] Message: A Combination of a Beta-Blocker and Amiodaron is undesirable. Logical expression: Beta-Blocker AND Amiodaron Name: Undesirable combination of antibiotics. Author: Korsten Type: Reminder Category: Undesirable medicine combinations Validation: Test Explanation: Literature reference: [Zagola GP. and MacGregor D, The Critical Care Drug Handbook 2nd Ed.pp. 165,1997J Message: Undesirable combination of antibiotics. Logical expression: Clavulanic AND Cefuroxim

Figure 3. A drug-drug interaction guideline, taken from the KBE's knowledge base. Beta-Blockers is a class of drugs, defined in the KBE by means of an intermediate guideline.


•

Although these guidelines describe two different drug-drug interactions, their general form is:

*

Ubetolol

Atenolol

Esmolol

Anti-Hypertensive

_ ., "«

D

If drug A is started; And drug B is present; Then report that drug A and drug B are interactions. In which: A = a newly prescribed drug B = an already administered drug

Digoxin

Figure 5. Medical terms and relations, used by the drug contraindications task.

This generalized rule is referred to as a guideline template. Using other guidelines from the KBE's knowledge base (figure 4), we can expand this template into an even more general one that not only includes drug-drug interactions but also drug-disease interactions and drug-laboratory interactions (referred to as drug contraindications). Name: HOCM and pericarditis are contraindications forDigoxin Author: Roos Type: Reminder Category: Digoxin Validation: Test Explanation: See: [Drug Therapy in Cardiothoracic Surgery, v Zwieten en Eijsman, pp 40,1997] Message: HOCM and pericarditis are contraindications forDigoxin Logical expression: (HOCM OR Pericarditis) AND Digoxine

Figure 4. A drug contraindication guideline.

As a result, the template that describes drug contraindications (including drug-drug interactions) is defined as: If drug A is started; And finding B is present; Then report that finding B is a contraindication for drug A. In which: A = a newly prescribed drug B = an administered drug, established disease or performed laboratory test Using this bottom-up procedure, guidelines are generalized to templates, using corresponding properties of the implemented guidelines. Although templates are more general than individual guidelines, the reviewing and medical knowledge are still intertwined. In order to acquire a task specification, the medical knowledge can be removed [7]:

Figure 5 shows a number of terms and relations, used by the drug contraindications task. It contains the required medical facts, such as known drugs and diseases, but also relations between facts, such as the is contraindication relation (e.g. 'Amiodaron is a contraindication for any Beta-Blocker') and the is a relation, which creates a hierarchical tree of classes. Note that, regarding this task, the is contraindication relation between two drugs is symmetrical. We can combine this structure with medical knowledge from other tasks and the ontology that was already developed for the application domain (e.g. the ontology, we developed for our ICU). In conclusion, the guideline abstraction process not only results in a number of abstract task specifications but also in an extension of the ontology of the application domain. Whenever a panel of physicians specifies a task, it is implemented into the reminder system as well as into the KBE. Then, physicians are able to revise and correct the terms and relations of the defined task structures. Implementing tasks in the reminder system By means of combining task specifications, medical knowledge and patient data that is stored into the EPR system, Van der Lei [7] describes a model for a task-based reminder system. This model, called the critiquing model, is shown in figure 6. EPR

For any drug that is started; Get the known contraindications of that drug; For all known contraindications of that drug; Determine whether that contraindication is present; If that contraindication is present; Report the contraindication

Tasks-

Domain ontology

Furthermore, the separated medical knowledge that is used by this task is defined as a collection of medical entities and the relations among them (figure 5). Critique

Rgure 6

T h e c r i t i q u i n g m odel.


•

In this model, patient data that is stored into the PDMS can cause certain events, called pertinent events (e.g. 'starting a drug'). Based on these events, relevant tasks are executed. For example, the 'starting a drug' event will execute the drug contraindications task. One could see the pertinent events as abstract notions that identify classes of possible actions or decisions of health care workers. Whenever a task is executed, it fetches the necessary medical knowledge from the medical fact base (e.g. all known contraindications of the started drug) and patient data from the PDMS (e.g. all administered drugs) and issues a reminder when the task procedure returns true. This is done by generating critiquing statements, which are instances of the abstract templates. An example is the critiquing statement 'Amiodaron is a contraindication for any Beta-Blocker', which is an instance of the drug contraindication template. Using the notions of events, task procedures and critiquing statements, the drug contraindications task specification becomes: applies to: any drug that is started critiquing statement: existence of a contraindication procedure: Get the known contraindications of that drug; For all known contraindications of that drug; Determine whether that contraindication is present Whenever a new task is specified, it is implemented into the reminder system according to the critiquing model. Also, the PDMS is updated whenever a new event class is specified. For example, whenever a physician prescribes a new drug for a given patient, the PDMS activates the reminder system with a 'starting a drug' event. The PDMS also supplies additional parameters to the reminder system such as the patient's ID and the name of the started drug. This event causes the drug contraindications task to be executed, as the 'starting a drug' event is linked to the drug contraindications task (however, this is not necessarily the only task that is executed, as one event may trigger multiple tasks). The drug contraindications task then fetches all known contraindications for the started drug from the medical fact base and queries the PDMS to determine whether one of them is present. Whenever a contraindication exists, the reminder system generates a critiquing statement. Similar to the generation of reminders (figure 2), critiquing statements are issued by means of pop-up windows. Also, users are able to validate issued critiquing statements by classifying them as 'correct', 'incorrect' or 'inconclusive'. To create a reminder system that is suitable not only for experts but also for novice users (the reminder system is also meant as a learning system for novice health care workers), detailed information about every template instance can be specified in the KBE. Examples are a more

detailed explanation, literature references, references to text-documents and URLs. Revising medical knowledge The manner in which physicians revise and correct medical knowledge in the KBE depends on the nature of the task (e.g., what are its terms and relations). In all cases, the user must specify an event that will execute a given task. The term that represents an event is selected from the ontology window and dragged onto the event list of the KBE's user interface. Next, other terms can be linked to the event similarly: they are selected from the ontology window and added to a corresponding list. The number and format of these lists depends on the nature of the task. For example, the KBE represents the drug contraindications task by means of an event list and a contraindication list, where the contents of the contraindication list depends on the selected event in the event list (figure 7).

£ie £d» Gutdeltne categories ^_J Consistency Database Q2 Intermediate guideline _ j Medicine U nursing C j Para-medical Tasks Hi! Drug monitoring HH Drug Cortrandicatron'— HH Admission preparation a t o n av l

" I i ICU ontology _. I Adrriission CJ Allergy C j Bloodgas CJ Calculations C j Clime • C j Comorbidity • _ J Complication C j Decubit • C j Disease _Cardro-surgical j | Aneutysm cor I Aorta Asc JArlRadialisCi. JCA£G JHOCM JMVP 1MVR

J

±S

Drug Contraindications Descnpfon: This task, which warn-; against drug contraindcatkms. has the following format Fot any dtug that is started: Get the known contraindications of that drug. For ad known contraindications of that drug; Determine whether that contraindication is present. if that contraindication is present: This task is executed by events of the following type: Medication

_±lAI(owed contraindication types: I • Medicafon —S-Laboraloriumtest Contraindications:

Events: Medication: Labetalol Medication: Cefuioxim

Medication: Amtodaon

[This contiaindication is iigrnficant

Medication. DtpyfidamoJ Medication- Esmolol Medication- Losaflan Medication: AHopunnol Medication- Haloperidof

alure Reference |v Zwreten en Eijsman. 199/J. More information about this contraindication can be found in IIHI Me.ll,n.•

OK

Figure 7. Revising medical knowledge, used by the drug contraindications task.

It is also possible to edit properties of events and terms (again, depending on the nature of the task) as well as to provide a more detailed explanation. Results At present, the physicians have entered 75 guidelines into the reminder system's knowledge base by means of the KBE. The contents of the guidelines varies from relatively simple, such as 'a patient, admitted outside normal working hours, is usually marked as an emergency' to more complex, such as the detection of drug contraindications and side


3!

Centrale

RINO Groep

The Centrale RINO Croep offers you a professional expertise comprising a mixture of training, education and advice. In this connection the combination of detailed knowledge together with a profound understanding of the Health Care organisational area, holds a distinct advantage. The professionalism that has been built up over the years within the Centrale RINO Croep could be of assistance to your organisation whenever employee training is

The Centrale RINO Croep

is the largest domestic

required, educational advice becomes desirable or whenever your organisation finds itself in a period of rapid change.

educational institute within the Mental Health

Course to qualify as Assistant in the Care Administration This 2 year course will focus on the further education of medical administrative

Care field in the

personnel (MBO+ level) in General and Mental Health Care together with a deeper understanding of professionalism in daily practice. This combination of actual

Netherlands. Since 1983

knowledge and close attention to daily practice will form the basis for this modular course.

it has been active in the

Several subjects fall within the framework of this course; among these Medical/Psychiatric Terminology, Health & Medical Law, HealthCare Organisation

educational areas of

and Administration, Information Systems, Registration Systems, Planning and Dossier formation.

OSA-program me

training, organising special courses and

The availability of specialised Courses for Secretaries, Administrative Assistants, Receptionists, Telephonists and other Support Services in all the Mental and

post-matriculation and

Health Care and allied fields. The OSA Programme consists of various components. Among these: Relating

Postdoctoral Professional

to Aggression, New Spelling, Reception and Customer directed Telephoning Skills, Time Management, Mail Handling, Professional Correspondence.

Training.

Information: For supplementary information about our activities, please contact Mr. Hilair Balsters (Course to qualify as Assistant in the Care Administration) or Mw. Jeannet Kool (OSA-programme). Centrale RINO Croep • P.O. Box 347 . 3500 AH UTRECHT • Netherlands Telephone: (00) 31 30 230 6110 • Telefax: (00) 31 30 230 6139 E-mail: [email protected] • Internet: http://www.crg.nl

effects or reporting the absence of certain monitoring requirements. From these guidelines, we have specified and implemented various tasks, such as the drug contraindications task, monitoring tasks that embody the notion that prescribing a drug or establishing a diagnosis may require certain observations (e.g. a laboratory test) and preparation tasks that warn when a minimumworkup requirement is not found in the PDMS. Besides these tasks that were extracted from implemented guidelines, we have also implemented tasks that were originally developed for the field of hypertension [7], In addition to the ICU physicians of the Catharina Hospital, a national group of ICU physicians has agreed to use our procedure for the development of standard ICU protocols for use in the Netherlands. Furthermore, other critical care domains such as traumatology and pre-operative screening have shown interest to develop and implement protocols by means of this method. We are currently preparing an evaluation of the correctness and helpfulness of extracted tasks in the ICU.

Discussion Although the KBE and the reminder system are in practical use only for a few months now, the first results are promising. Physicians are able to enter guidelines without assistance of a knowledge engineer. Also, when they have entered enough instances of a template, they are able to extract and describe a task's general features. After implementation, they are also capable of revising a task's medical knowledge by defining its terms and relations. An important issue is the attitude of the physicians and the nursing staff towards the use of decision support systems [20], During the last two decades, the majority of expert consultation systems utilized the 'Greek Oracle' approach. These systems, such as INTERNIST-I [21], typically expect a health care worker to enter information about a patient, after which the system produces a number of conclusions and recommendations such as a list of possible diseases, a set of suggested tests or a treatment plan. Experience revealed, however, that the system's users could become annoyed by this approach, because the user's role often is diminished to that of a passive observer or even a 'slave' of the computer. Reminder systems, on the other hand, utilize the so-called critiquing approach. Critiquing systems are silent whenever the computer judges the user's (planned) behavior to be satisfactory given the case data, but offer a critique of the behavior should the user (propose to) take an action that is not consistent with the system's knowledge base. As a result, the additional workload on health care workers is kept to a minimum. An additional requirement however, is that the system must operate in real time to be accepted by its users, especially in critical care environments [5, 6], We developed the

reminder system therefore by means of the SIMPLEXYS toolbox, a collection of tools that assist in the design of real time expert systems [22], As a result, the response time of the reminder system is very short: executing the drug contraindications task on a Pentium-166 PC takes about one second. The assumption that the separation of reviewing knowledge and medical knowledge facilitates sharing of knowledge seems to be justified: various tasks that were originally developed for the field of hypertension are implemented for use in Intensive Care. However, further research must determine if these tasks are also transferable to other clinical domains. Automated knowledge acquisition tools and decision support systems are able to improve the quality of care as long as the physicians and nursing staff are willing to use them. The KBE as well as the reminder system were therefore developed in close collaboration with and as specified by the physicians and nursing staff of our ICU. After a short training period, health care workers find the user interface of the KBE as well as the reminder system useful and sufficiently 'intuitive'. Our findings and results convince us that the bottom-up strategy, described in this paper, combined with automated knowledge acquisition tools enable medical specialists themselves to improve the quality of the knowledge base, the EPR system, and hopefully, of the care in critical care environments.

Address for correspondence Paul A. de Clercq Eindhoven University of Technology Room EH 3.08 P.O. Box 513 5600 MB Eindhoven The Netherlands tel.: +31 40 - 2474794 fax: +31 40 - 246 65 08 email: [email protected]


References [1] Safran C. Rind DM. Davis RB. Sands DZ. Carabalto E. Rippel K. Wang O. Rury C. Makadon HJ. Cotton DJ. A clinical trial of a knowledge-based medical record. Medinfo 1995:8:2:1076-80. [2] Ornstein SM. Garr DR. Jenkins RG. Musham C. Hamaden G. Lancaster C. Implementation and evaluation of a computer-based preventive services system. Fam Med 1995:27:260-6. [3] Effective Health Care. Implementing Clinical Practice Guidelines: Can guidelines be used to improve clinical practice? Effective Health Care 1994:1(8). [4] East TD. Morris AH. Wallace CJ. ClemmerTP. OrmeJFJr. Weaver LK; Henderson S. Sittig DF. A strategy for development of computerized critical care decision support systems. Int J Clin Monit Comput 1991;8(4):263-9. [b] Gill H. Ludwigs U. Matell G. Rudowski R. Shahsavar N. Str'm C. Wigertz O. Integrating knowledge-based technology into computer aided ventilation systems. int J Clin Monit Comput 1990:7(1):1 6. [6] Latfey TJ. Cox PA. Schmidt JL. Kao SM. Read JY. Real- lime KnowledgeBased Systems. Al Magazine 1988:9:1:27-45 [7] Van der Lei J. Muscn MA. A model tor critiquing based on automated medical records. Comput Biomed Res 1991:24:344-78. [8] Newell A. The knowledge level. Artificial Intelligence. 1982; 18:87-127. (91 Eriksson H. Musen MA. Conceptual models for automatic generation of knowledge-ac(}uisi!ion tools. Current Developments in Knowledge Acquisition - EKAW '92. 6th European Knowledge Acquisition Workshop Proceedings 1992:14-36. [10] Musen MA. Dimensions of Knowledge Sharing and Reuse. Comput Biomed Res 1992:25:435-67. [ 11] East TD. Henderson S. Pace NL Morris AH. Brunner JX. Knowledge engineering using retrospective review of data: a useful technique or merely data dredging? Int J Clin Monit Comput 1991;8(4):259-62. [12] Eriksson H. Musen MA. Metalools for knowledge acquisition. IEEE Software 1993:10:3:23-9. [131 Davis R. LenatD. Knowledge-Based Systems in Artificial Intelligence. New York: McGraw-Hill. 1980: 229-490. [14] Musen MA. Fagan L. Combs DM. Shortliffe EH. Use of a domain model to drive an interactive knowledge-editing tool. Int J Man-Mach Stud 1987:28:10521. [15] Metnitz PGH. LenzK. Patient data management systems in intensive care the situation in Europe. Intensive Care Med 1995:21:703-715. [16] McDonald CJ. Hui SL. Smith DM. Tierney WM. Cohen SJ. Weinberger M. Reminders to physicians from an introspective computer medical record. A two-year randomized trial. Ann Intern Med 1984, 100:130-8. [171IOM. Clinical Practice Guidelines: Directions fora new program. MJ Field andKNLohr, eds. Washington, DC: National Academy Press, 1990. [18] Hripcsak G. Rationale for the Arden Syntax. Comput Biomed Res W94:27(4):29t-324. [19]Gruber TR. A translation approach to portable ontologies. Knowledge Acquisition 1993;5(2):199-220. [20] Shortliffe EH. Testing Reality: The Introduction of Decision-Support Technologies for Physicians. Meth Inform Med 1989:28:1-5. [21] Miller RA. Masarie FE Jr. The demise of the "Greek Oracle" model for medical diagnostic systems. Meth inform Med 1990:29:1-2. [22] Blom JA. Temporal logics and real time expert systems. Comput Methods Programs Biomed 1996;51;35-49.

THIS BROCHURE IS PRINTED BY Marcelis & van der Lee Alkmaar - Holland phone+31 (0)72 5126497 fax +31 (0)72 5154822 e-mail: [email protected]

medical protocols supporting in-hospital cardiological care Caspar M.B. Duwel, MD, PhD, Dick A.C.M. Kruijssen, cardiologist This study was performed at the hospital 'Zuiderziekenhuis' in Rotterdam and was part of the information project PROCAS (PRofiles Of CAre System). It is one of the Fourth Frame AIM-projects (Advanced Informatics in Medicine) whose objective is to improve the efficiency and quality of medical treatment by information technology. The target of PROCAS is to make explicit profiles of care, with regard to a group of patients with similar conditions. It involves the linking of a specific health problem of a patient (group) with a specific package of care. Eight medical protocols supporting logistics of cardiological care are developed. The designed protocols contain mandatory and optional investigations or interventions supporting cardiological care. Introduction Ordering diagnostic and therapeutic interventions may be found in medical dossiers described as short remarks such as 'routine admission' or 'routine heart catheterization'. Usually the elements of these routines are part of common knowledge or 'verbal history' of physicians, nurses and paramedical staff. In this article medical protocols which are common practice at the cardiology department are described. As part of an information project cardiological in-hospital protocols have been defined. A start has been made to explicit clinical guidelines. Levels of authorisation have not been mentioned.

Methodology Diagnostic and therapeutic orders of interventions may be given at the first aid, coronary care unit or cardiological unit. In order to evaluate the order registration and routines, 225 medical dossiers of the most frequent diagnosed cardiological patients groups have been evaluated and interviews have been taken from cardiologists, physicians in training, head nurses, coronary care nurses, secretary- and doctor assistants. The verbal information of the routines and the study of the medical dossiers have been structured in 'mandatory' and optional items of investigation. Draft documents have been discussed on a individual basis. Remarks were being collected and evaluated during a medical staff meeting.

Results The following medical protocols, of which 1-3 are listed in the Appendix, have been developed: 1 admission protocol for chest pain; 2 myocardial infarction schedule; 3 unstable angina pectoris or atypical chest pain; 4 pacemaker implantation; 5 pacemaker replacement; 6 cardioversion; 7 heart catheterization and/or electrophysiological investigation; 8 post-open heart surgery. 1 HEALTH INFORMATION DEVELOPMENTS IN THE NETHERLANDS - APRIL 1999

Dick Kruijssen (left) and Caspar Duwel

Right page: Oil on canvas 100x80 cm

Admission protocol for chest pain The admission protocol may be executed at the First Aid or during an elective (planned) admission. At the First Aid patients presenting chest pain are subject to immediate execution of the protocol. An electrocardiography (ECG) will be made. If ECG-deviations are being detected, urgent biochemical laboratory tests will be implemented. Depending on additional qualifications of the chest pain (which include: duration, frequency, severity and associated symptoms) a short intravenous catheter and / or a chest X-ray will be implemented. Positive blood tests (increased enzymes) will lead to execution of the entire admission protocol.

Myocardial infarction schedule The myocardial infarction schedule is an episode protocol of seven days. The first day includes the admission protocol, monitoring heart rate, blood-pressure, ECG, and haematological and biochemical laboratory tests. However, Swan-Ganz-, intra-arterial- or urine catheter, blood gas analyses, echocardiography, heart catheterisation, PTCA (Percutaneus Transluminal Coronary Angioplasty), and haematological and biochemical laboratory tests are optional parts of the protocol and will be implemented if necessary i.e. with an appropriate indication. If patients show complications the protocol will be extended. If necessary 'blanc days' may exist i.e. days without investigations or interventions. The mobilization plan of patients is not included.

Unstable angina pectoris or atypical chest pain To observe patients in-hospital for a short period the protocol unstable angina pectoris is being executed. This includes the admission protocol and eventually two days of investigations.

Cardioversion, pacemaker replacement or -implantation In these short protocols (not shown in the Appendix) the conditions have been described for (planned) admissions with subsequent intervention to correct an abnormal heart rhythm. The protocols may be part of an elective admission but may also be executed during a patient's stay.

Heart catheterization, open heart surgery These protocols (not shown in the Appendix) describe the checklists for interventions and investigations before and during (elective) admissions. The hospital performs open heart surgery. Postoperatively, patients dismissed from the recovery room, start the protocol on day one at the intensive care unit. The rehabilitation plan is not included.

mandatory items. The optional parts reflect the idea that the protocols are not rigid 'cookbook recipes' and therefore do not affect the flexibility of care. Depending on an active health problem, a patient may change from one protocol to another, for example from the unstable angina pectoris protocol to the myocardial infarction schedule or vice versa. The mandatory part of the protocols reflect the minimal amount of needed investigations. They will act as a reminder or checklist in the management of patients streams, as part of scheduling and supporting the logistics of care. In the perspective of medical training, the mandatory part may guarantee an adequate supply of clinical data, informing the cardiologist in duty. Computer-based registration of medical data will sometimes be mentioned as an Electronic Patient Dossier (EPD). The definition of an EPD and the best way to computerize information about a hospital or outpatient clinic are discussed a lot. Although 'turnkey' EPDsoftware implementations are being offered to hospitals, computerization takes place step by step. Interests find expression in computersupport of partial care-chains such as scheduling patients' appointments. Ordering diagnostic and therapeutic interventions may be a next step. In the future computerized protocols as shown in the Appendix, communicating with medical- and nursing dossiers, with agendas or laboratory working lists, will reduce the redundancy or incorrect of data-entry and will efficiently support the logistics of care. It could decrease the time and workload of physicians, nurses and paramedical staff. Computer-supported care may therefore increase production and information-intensity as encountered in commercial companies. The connection of medical protocols with health problems or working diagnoses may provide a basis for optimizing the standards of care. The eight described protocols were revised regularly and updated in managing patients with similar conditions. In the future the clinically collected patient data may be part of an integrated (quality) management system for clinicians as well as for hospital managers. Most of this article was published in Gordon and Christensen fed): Health Telematics for Clinical Guidelines and Protocols, vol 16; IOS Press Ohmsha 1995, Amsterdam - Oxford - Tokyo, Washington; ISBN 90.5199.1991 Caspar M.B. Duwel joins the consultant company Orga-lnfo in Woerden and Dick ACM. Kruijssen works as a tutor at the Academic Hospital of the Erasmus University of Rotterdam, both in the Netherlands.

Address for correspondence Discussion The developed medical protocols are part of everyday practice: they are explicit operational procedures. Both for acute (emergency care and coronary care unit) as well as for not-acute management the large part of them consist of

C.M.B. Duwel, MD, PhD, Paramaribostraat 140-1 1058 VP Amsterdam, the Netherlands

«


A more cost effective way of filing

Jalema File Management frees valuable archive space Innovation is the basis of all Jalema products and services. Jalema File Management is another example of how imaginative thinking results in real solutions for the problems contemporary medical institutions have to deal with. JFM provides necessary space One of the main chalenges medical institutions are facing these days is how the archive space can be used as effeciently and effectively as possible. The duty to retain medical files leads inevitably to an ever increasing demand on space available. Often the 'sleeping' files take up the valuable space needed for the active files. However big the institution, there is an end to the space that can be made available for the archives. At this point Jalema File Management comes in. Jalema File Management is in fact an annexe to your files. Adjacent to our head office in Reuver, the Netherlands, we have founded a thouroughly secured building (Fort Knox at the Maas) that provides a safe environment for medical files from medical institutions all over Europe. These files do no longer take up valuable space in the premises of the organisation to which they belong, but are readily accessible at all times, through lighteningfast ISDN lines. Free computer and software with JFM For clients that would like to make use of the electronic application service, we provide a free PC, software, printer and ISDN-modem. After a request for a specific file has been sent to Jalema, we are able to scan the file within 1,5 hour and put it at your disposal via the ISDN-modem. You can view this in the computer, or have it printed automatically. The electronic files you will receive are of course perfectly suited to be filed in a DIS (Document Info System) environment. All of this has a positive influence on the organisation costs, because in fact these costs (looking up files,

Why JFM is of interest to your organisation: • It's cost effective • Secure and discrete storage • It saves valuable space • High quality file management • Organisation Costs turn into Direct Costs • Continuity in information supply • Highly efficient conversion from physical to electronic filing through 'scanning on demand' • It is more economic than scanning complete files or putting them on microfilm • Integration in existing DIS environment possible.

walking distances, chatting with collegues) become purchase costs. The exploitation of your archive will become more economic with JFM. JFM at a glance: • Individual filing: no storage in boxes; each and every file readily accessible at all times • We do not store, we manage your files • Lateral filing in racks • Registration (via barcode) of each individual file • JFM-agreement specifies exactly which services are rendered. Part of the agreement is a complete list of all individual files that are taken in by Jalema • A very heavily secured building: - Burglar alarm - Admittance contrail system - Video surveillance system - Fire Alarm System - Fire Extinguishing System • Files accessible on-line • Cost effective service • Destruction of files at your request, protocol of destruction will be delivered by Jalema • JFM-service is ISO-9001 certified • Extra statistical information Jalema File Management: a real solution It is our goal to offer real solutions for information management at hospitals, GP's practices and care establishments. In health care we have earned an impeccable reputation over tens of years in the whole of Europe. We are convinced that with Jalema File Management we have added another cornerstone supporting the long-term relationship we maintain with hundreds of hospitals and medical establishments across the whole of Europe.

JFM Building

Intake/Registration

Verification

Scanning

Protocol admission CLINICAL GUIDELINES This protocol will be carried out: 1 at non-elective admission at the First Aid 2 at elective admission at the Cardiology Ward 3 at a diagnostic-therapeutic path at the ward: each determination / investigation is optional - Patients who have received a thrombolytic therapy during transport in the ambulance should be transferred to the CCU immediately. ECG (rest) Venflon catheter (optional) Chest X-Ray: Anterior Posterior and Left Lateral (optional)

DAY 1 after 6.00 hrs CCU: Biochemistry: Sodium (optional) Potassium (optional) CPK / Troponin ALAT ASAT LDH ECG (rest) Prothrombin time DAY 1 after 12.00 hrs CCU: Haematology: Biochemistry: Haemoglobin (optional) Sodium (optional) Haematocrit (optional) Potassium (optional) CPK / Troponin ALAT ASAT LDH Creatinin (optional) ECG (rest) Prothrombin time

DAY 4 at 8.00 hr CCU: Biochemistry: (or ward) CPK/Troponin ALAT ASAT LDH INR (optional) Chest X- Ray: Anterior Posterior plus Left Lateral Patient transferred to the ward: 1 Telemetry (optional) 2 Ergometric test (optional) 3 Holter ECG (optional) 4 Tl or Tc-SPECT (optional) 5 MUGA (optional)

Biochemistry: Urine: Haematology: Sodium pH Haemoglobin Haematocrit Potassium Protein Glucose Sedimentation rate Urea Creatinine Sediment Leucocytes Alkaline Phosphatase Thrombocytes CPK/Troponin Differential count ALAT ASAT LDH Bilirubin total Bilirubin 1 min (if Bilirubin total has been increased) Total protein Albumin Glucose INR (optional)

DAY 5 at 8.00 hr (ward or CCU): Haematology: Biochemistry: Urine: Sedimentation rate Sodium pH Haemoglobin Potassium Protein CLINICAL GUIDELINES Haematocrit Creatinine Glucose Transfer to the ward: with haemodynamically Leucocytes CPK/Troponin Sediment stable patients and no CPK / Troponin ALAT increments ASAT LDH DAY 2 at 8.00 hr CCU: Alkaline Phosphatase Haematology: Biochemistry: Glucose (fasted) Sedimentation rate Sodium ECG (rest) Haematocrit Potassium Haemoglobin Creatinine 1 telemetry (optional) Leucocytes CPK / Troponin 2 Bicycle ergometric test: age > 65 years (optional) ALAT ASAT 3 Holter EKG (optional) LDH 4 Tc-Sestamibi /Thallium-SPECT (optional) Alkaline Phosphatase ECG (rest) Prothrombin time 5 MUGA (optional)

CLINICAL GUIDELINES - Intravenous catheter: if acute treatments are likely. - Chest X-Ray, which has been carried out in the last 3-6 months will only be repeated if complaints are substantial.

DAY 2 at 20.00 hr CCU: Biochemistry: CPK / Troponin ALAT ASAT LDH Prothrombin time

Protocol myocardial infarction schedule CLINICAL GUIDELINES

DAY 3 at 8.00 hr CCU (or ward): Haematology: Biochemistry: Sedimentation Rate Sodium Haemoglobin Potassium Haematocrit Creatinin Leucocytes CPK / Troponin ALAT ASAT LDH ECG (rest) Alkaline Phosphatase INR (optional) Patient transferred to the ward: 1 Telemetry (optional) 2 Ergometric test (optional) 3 Holter ECG (optional) 4 Tl or Tc-SPECT (optional) 5 MUGA (optional)

This protocol will be carried out in case of emergency- or non-elective admissions for acute treatment. - This protocol starts at the CCU. - Decision: CCU or Ward based on intake, ECG (and CPK / Troponin). - Patients with a thrombolysis therapy during transport in an ambulance should be transferred to the CCU directly.

Protocol admission DAY 1 CCU: (= day of admission) Monitoring: ECG, heart rate, blood pressure after 1-2 hrs CCU:ECG sinus rhythm recording (optional) ECG (rest) 1 Swan-Ganz catheter (+ pace port) (optional) 2 Intra-arterial catheter (optional) 3 Urine catheter (optional) 4 Blood gas analysis (optional) 5 Echo 1 Doppler (optional) 6 Heart Catheterization (optional) 7 PTCA (optional) 8 Telemetry (optional at the ward)

CLINICAL GUIDELINES Haemodynamically stable patients with small infarctions without complications will be transferred to the ward. Patients with large or relapsed infarctions, eldery patients, or patients with PTCA / complications will stay longer at the CCU.


DAY 6 at 8.00 hr: Blanc day ECG (exercise-) at an age of < 65 years (optional) DAY 7:Blanc day or DISCHARGE CLINICAL GUIDELINES The schedule of 6-7- days involves patients with a small infarction without complications. With a large or a relapsed infarction, with the elderly patients, or patients with complications the average stay is about 12 days. DAY 8 at 8.00 hr: Haematology: Biochemistry: Sedimentationrate Sodium Haemoglobin Potassium Haematocrit Creatinin CPK / Troponin ALAT ASAT LDH Alkaline Phosphatase INR (optional)

Dtk

dfe u

uwe

g de informatie tvoor het beheren rmatie

Europese Beurs & Congres RAI Congrescentrum Amsterdam van 8 t/m 10 juni 1999

De interne en externe communicatie, de inkoop en de verkoop, de ontwikkeling en fabricage van producten, de klantenservice, het voldoen aan de wettelijke voorschriften en natuurlijk het delen van de aanwezige kennis... Stuk voor stuk onderwerpen waarbij informatie een belangrijke rol speelt. Goed informatiebeheer kan deze bedrijfsprocessen dan ook aanzienlijk verbeteren. Het resultaat? Verlaging van

de kosten en verhoging van de winst. Informatiebeheer staat of valt echter met de juiste managementoplossingen. Op de IMC '99 kunt u zelf zien hoe een complete serie oplossingen in de praktijk werkt. En welke oplossingen voor uw specifieke situatie het meest geschikt zijn. U maakt kennis met nieuwe en toekomstige technieken. Reden genoeg dus om een bezoek te brengen aan de IMC '99.

WILT U ZICH NU AL AANMELDEN, NEEM DAN CONTACT OP: Web: http://www.iimc.org/

EXCALIBIR1

h

E-mail: [email protected] Telefoori: +44.1962.868.333 Fax: +44.1962.868.111

SiDoc

UNISYS

Hans van der Slikke

protocols and guidelines in a hospital intranet J.W. HANS VAN DER SLIKKE, MD PhD, Obstetrician and Gynaecologist WINNIE A. SCHATS, Librarian In this article we describe how in a community-hospital in the Netherlands (350 beds) an Intranet appeared to be the solution for problems in the field of patient information and protocols/guidelines. The management and logistics are described and some pitfalls are mentioned. Introduction When stocktaking was done about (paper) protocols and guidelines, used in our hospital 'de Heel' in Zaandam, The Netherlands, it became clear that there were many problems: - it was not always clear if a written guideline for a procedure existed; - it was not always clear where it was; - it was not always clear when the guideline had been updated; - it was not always clear who was the responsible author of the guideline. This situation was not acceptable: there could arise even dangerous situations by using outdated guidelines. For this paper we will use the word protocol for description of a procedure (often for nurses, but also on the administrative department). We

use the words (clinical) guidelines (CGL's) where flowcharts for treatment were meant: most of those were written and used by doctors. Protocols and guidelines in a database So we had the problem, described in the first alinea: many out-dated protocols and brochures. We tried to solve the problem by making a standard format for protocols and CGL's. Then we asked the responsible nurse/clinician to rewrite and/or re-edit the protocols if necessary. Also they had to answer several questions: - Is there an (inter)national CGL for this procedure? - Is there an advice or guideline or are there standards from your professional society? I


They were stimulated to use such national guidelines if available. A database was made with the following fields: - aim of the protocol; - name of the procedure/topic; - department(s); - for whom was it written (nurse, clinician, administration); - name of the author; - date of authorisation; - date of expiration. The database was rapidly growing, but only part of the problem was solved: The database was available on a stand-alone computer in the library only: It took a lot of trouble trying to find the proper protocol. Of course this was not acceptable during busy practice.

Patient information A year before we had the same experience with patient information leaflets and brochures. In this field we had faced still another problem: in the clinic/doctors office we kept many stacks of self-printed materials, waiting for the patient. And if a little change was necessary (a phonenumber changed) the whole stack had to be thrown away and reprinted. Of course this was not only an expensive solution but also gave a lot of extra work.

Hospital Intranet Heel Wide Web As the name of our Hospital is 'de Heel' the hospital Intranet was named 'Heel Wide Web1'. (Fig. 1) A script was developed in which each patientinfo has the same layout. By putting the brochures on the net we could restrict ourselves to printing the brochure during the visit only. This way the stacks weren't necessary any more. There were extra advantages: we could let the patients use a computer with touch-screen, were they could

Figure 1: Opening screen Heel Wide Web gdA

/

View

,'

£o

gommurs&alor

3

^

Help

J*-

Reload Home Seaich ,Jf Bookmaks .£ Location: [wc^viwar

4^

•-•

6ude

d? Ssajwfy

Welkom op het Heel Wide Web Op he! Heel Wide Web kunt u alleriei informatie vinden Klik op een van de knoppen memaast of typ hieronder een woord in waar u op wilt zoeken en k!ik op 'zoek' binnen het Heel Wide ,,, ,

Documert: Done Ap j ^jposirak IN MICIC

Immediately after the implementation of the system we did a randomised trial with 40 women, who entered for a sterilisation by laparoscopy. Twenty women received the standard patient information: after the doctor explained the procedure, they received a brochure with the same information (paper-only group). In the 'video-group' during the information a video-clip was shown (duration 35 seconds) where was shown how a little silastic band was placed around the Fallopian Tube. Both groups were asked to complete two forms at home, one the day after the visit to the doctor's office and one after the sterilisation. We received back 37 sets of forms. We missed two from the PO-group and one from the video group: one woman changed her mind and resigned from sterilisation. Two women didn't react at all. As for the questions concerning knowledge the answers in the two groups weren't different. However the question if they could imagine what was really happening during the operation was answered positive by all women from the video-group and by 3/18 of the paper-only group. 17/18 answered 'yes' on the question if they would have liked to see a little video-clip showing the actual procedure. After this positive experience, we changed our policy and decided to use the 'Heel Wide Web' for the guidelines and procedures as well.

In between one of us had experience with developing clinical guidelines and protocols on the Internet. This was within the website of the NVOG (Netherlands Society for Obstetrics & Gynaecology, http://www.nvog.net). The first draft is published on the website within a secure intranet-section. The author follows a fixed format and sets a deadline. All peers can contribute until the deadline. With their remarks and additions the author edits the second draft. This circle can be followed several times. In our hospital the authors of a protocol could use such national standards, when available. However these standards often need some editing in order to 'customise' them to the local situation. In 'de Heel' a little HTML-editor has been developed. The aim of this editor is to give the author a pre-set form, in which the protocols can be written in HTML-format, without people having to learn this language. (Fig. 3) •

II

Heel Wide Web

-,,.,„ r Zoek op

Patient compliance/satisfaction

Making and authorising the (national) guidelines

» Heel Wide Web . Netscape File

see the information for themselves and see little video-clips as well. For the outpatient clinic we used these touch-screen monitors: now it was possible for the patients to search the system without needing a keyboard. (Fig.2) This appeared to be a very adequate and versatile multimedia patient information system: because there was only one folder for each topic centrally stored, we always disposed off the most recent version. On the other hand it was available as well in the outpatient clinic, as in the doctor's office or the wards.

•£• 1 afcSnaglt/32 Capture

||£%Heel Wide Web -—

14 03


Sterilisatie met behulp van de lapaioscoop DE ORGANEN C* DE BUIK

Figure 2: Touch screen design for patient use without a keyboard.

we studied the problems and pitfalls that were encountered in introducing an Intranet system for guidelines and protocols. An online-questionnaire was used to find out how the users of the Intranet perceive the system, both in terms of user friendliness and in content. The results will be used to improve the Intranet, before it will be implemented hospital-wide. Second half of 1999 all (nursing) departments of the hospital should have access to the protocol-database via the Intranet. (By that time in our hospital the Windows 3.11 environment will be replaced by Windows NT, with also upgraded computers, with a faster processor and more memory). From this hospital-wide upgrade we expect to gain momentum for the use of the Intranet.

Lack of compliance to the CGL

Figure 3: The protocol-editor for Heel Wide Web.

The protocol editor is a simple tool developed by our consultants 'Care for Care' in Woerden (NL). The use can be learned in 10 minutes.

Publishing the guidelines There's one person that co-ordinates this database. In our hospital this is the librarian, who is also the HWW-webmaster (WAS.) Anyone in the hospital can propose protocol, but only the webmaster can put the protocol on the Intranet. In this way it is guaranteed that only those protocols that have followed the proper procedure are made publicly available. Also a certain degree of uniformity is guaranteed. So after authorisation the guideline is put into the database. The author can indicate the life span of a certain protocol. This date is incorporated in the database. Three months before the expiry date a message is generated for the author, asking him if he wants to update the protocol. The text of the protocol concerned then is attached to this message. On the Intranet one can use a search engine, with which one can search protocols on keyword, department or author.

Implementation: a pilot first Before implementing the system, a pilot was done in two nursing departments. In this pilot

One of the problems with guidelines in general is the way physicians or nurses do not comply with them. When a guideline is new, people incline to consult it. After one or two consultations one thinks to know the protocol and never looks again... even not when the protocol has been updated. The compliance with the guideline can become very poor in those cases. We tried to solve this by giving the people the occasion to subscribe on a certain document: every time this document is edited/updated, the subscriber finds a message in his mailbox that there has been an update for this concerning procedure.

The ultimate goal There are many arguments to suppose that in the near future the Internet-browser will be the common Electronic Patient Record (EPR) interface for nurses and physicians. Modern web-technology makes it possible to extract the data from the 'old' clinical databases, using an 'intermediate layer' or a communication server. This interface has many advantages, of which one of the most important is the transparency: there is one layout with much functionality. The same interface as well in the hospital as at home! The ultimate goal is connecting the clinical guidelines with EPR. And here the knife will cut on both sides: Because there will be a context related hyperlink between the patient record and the protocol database in the Intranet-server, the doctor facing the EPR will find the context-related guideline as soon as he chooses the 'trigger'-code. Finally the inputscreen could be customised for the specific guideline: then the doctor is really 'guided' through the case. Even more fascinating is the idea that by recording the doctors' decisions and by giving the doctor feedback from these decisions (and the outcome!) the quality of the protocols can be even enhanced. Even new guidelines can be developed from this selfbuilding knowledge base. *


1

In Dutch 'Heel' as a verb means 'heal', but as an adjective 'very'. Protocols and Guidelines in a Hospital Intranet Correspondence email: [email protected]

Loopt u verloren in uw eigen archief ? Verdrinkt u wel eens in de computerlistings ? Agfa biedt u met het ALD-4000 systeem een efficiente oplossing. Gedaan met de stapels stoffige dozen, geen schuiven meer vol onoverzichtelijke hangmappen. Met ALD-4000 verandert uw labyrint in een handig computersysteem. In een handomdraai raadpleegt u het archief of bewaart u belangrijke documenten. Dankzij de enorme opslagcapaciteit van het ALD-4000 systeem kunnen alle soorten ingescande documenten en computeroutput (COLD) op een CD-schijfje bewaard worden. Het opzoeken en opslaan van de gegevens verloopt makkelijk en razend snel. Het ALD-4000 systeem werkt dus als een kompas : even geniaal als eenvoudig. Het is een onmisbaar instrument voor wie de juiste koers wil varen.

1

documen tmanagement

I I

Agfa - Gevaert B.V. Micrografie en Documentsystemen Polakweg 10-11/ 2288 GG Rijswijk / Tel. (070) 41 31 211 / Fax (070) 41 31 587

AGFA als u d a c h t a l l e s g e z i e n t e

hebben.

VIZI

virtual integration of information for care and cure In Europe and in the United States, there is great interest in the electronic patient dossier (EPD). In the Netherlands too, a large number of excellent initiatives have been taken in this area. This month, a project has been launched in the Netherlands under the name of VIZI: Virtual Integration of Information for Care and Cure. The umbrella organizations of the care providers have commissioned SIG Health Care Information to produce an Action Plan for VIZI. The initial phases of this plan will get underway this month. Mission and objectives The problem that VIZI has been designed to tackle can best be illustrated with a number of examples: - A patient arrives at his family doctor with the news that he has been diagnosed at the hospital as suffering from a serious illness but his GP has received no information about this. - Immediately after admission to hospital, a patient is prescribed certain medicines. However, the hospital is unaware that during previous patient/care provider contact, it emerged that these drugs do not have the desired effect on this particular patient. - A patient in a confused mental state is admitted to Intensive Care and forgets to inform the department that he is allergic to a medicine being administered to him by IC staff. - A specialist demands that an expensive and intensive test be carried out for a second time because he does not trust and/or understand the data available from previous tests. These examples illustrate insufficient quality in the provision of information: the information comes too late, cannot be trusted or is simply not available at the location where the patient is presently undergoing treatment. This has a direct negative influence on the quality of care: effectiveness and efficiency suffer as a result. Looking to the future, we see the prospect of much shorter hospital stays, while care and cure are pushed still further in the direction of primary health care and home care, and an increase in patient transfers to other care sectors. Under these circumstances, the status quo in terms of information simply cannot be maintained. The provision of information has to improve and become more integrated. In spite of all this, it is not true to say that there are no effective applications in the hands of individual care providers. However, there is often a lack of operational communication functions and clear use of terminology by care providers both within and between the various echelons (home care, specialists, patients etc). Existing applications lead too much of an isolated existence (island automation).

The aim of VIZI is to make a contribution towards solving these problems by building bridges. The VIZI mission statement reads as follows: that, in the interests of high-quality care, the necessary information should be available to the authorized persons in any contact between care provider and patient, in a standardized yet flexible form and in a wellsecured environment, regardless of the time and place at which the information is required. It goes without saying that the EPD has a crucial role to play in this respect. There has been no attempt within VIZI to provide a strict definition for the EPD. Instead a combination of existing applications which contribute towards the mission will come together to form a new cornerstone for EPD. A virtual EPD that is. The most important functions of the EPD are first that it can bring together information divided between various places in the service of patient/care provider contact (data and the associated data management remain at the location where the responsibility for management resides) and second that it supports the specific care process of the care providers (e.g. migration measurement for hip implants at the orthopaedist, children's growth curves for the paediatrician). A pragmatic approach forms the basic principle for VIZI: progress is made in small steps, use is made of the available resources and the theoretical framework is developed empirically from practical experience (de facto standards). What is needed in addition to a collective mission in order to solve the problems outlined above? - sufficient administrative power - a control instrument - a starting point with an initial benchmark Administrative power The unique situation with regard to VIZI is that the collective care providers in the 'Cure Sector', namely hospitals, specialists and general practitioners have joined forces in order to ensure the success of this mission. The umbrella organizations for the hospitals (Association of Hospitals in the Netherlands,

NVZ), specialists (Order of Medical Specialists) and general practitioners (National Association of General Practitioners, LHV) have joined forces to set up a steering group in which they take responsibility for the results of the initial benchmark: a reference model with a number of 'new specified guide applications'. The government, in the shape of the Ministry of Health, Welfare and Sport, is now also taking measures in the form of further direction and financing. The extent to which the health care sector (the fourth production factor) is lagging behind other sectors with regard to the supply of information is now making itself felt.

motto is 'start simply and only expand when the project and the users are ready for it'. As a result of this line of thinking, the 'Cure Sector' was chosen as the point of departure. Initial delineation (see diagram) The communication between medical specialist and: - nurses - laboratory, image and function rooms - fellow specialists within and outside the hospital and the general practitioner - hospital and high-street pharmacies - care supporting hospital functions such as patient logistics and financial management and communication between general practitioner and - hospital pharmacy - laboratory, image and function rooms

A control instrument A so-called 'reference model' is being developed as a control instrument. This is a general structure within which the many

The EPD (as process support instrument) is defined by specialist medical and nursing care. Other relations High-street pharmacy (for example between high-street pharmacy and general Care practitioner, or among supporting non-hospital care hospital functions providers) can be incorporated into VIZI at a later stage. An initial benchmark On 12 February 1999, the Ministry of Health, Welfare and Sport gave Fellcw specialists the go-ahead for the in same hospital first two activities of the VIZI Action Plan. Assuming that the complete plan can be implemented, a reference model will be available this time next year including detailed specifications for a number of guide applications. Depending upon further financing, these applications will be made operational in 2000. The umbrella organizations for the care providers will then make every effort to introduce the reference model and the guide applications at grass-roots level within their own sectors.

Initial delineation for VIZI General practitioner A

r

•

£

L

Hospital pharmacy

Laboratory function image

—

•»

Specialist/ patient contact

»

*

*

Nursing

Fellwv specialists in other hospital

existing applications (and those yet to be developed) are to be positioned so that standardization and communication can be realized step-by-step. The reference model, within which law and technology-related aspects can also be included, is intended to prevent the wheel having to be reinvented again and again. It also supplies functional assessment criteria for use in the process of choosing between suppliers' offers. The reference model is developed by integrating and updating existing models on the basis of a number of successful applications which act as a guide and can serve as the cornerstone for the virtual EPD and the reference model.

»

•»

M.J. Jonker, VIZI Project Leader SIG Health Care Information, Utrecht, the Netherlands Correspondence Dorenda van Seumeren

A starting point

Tel. 030 - 2345.597

This mission cannot be achieved in one fell swoop: this would mean that the entire care chain (some patients consult over 20 different care providers in the space of a single year) would have to be involved from the very beginning. In the development of the reference model and the choice of the guide applications, VIZI has opted for a growth scenario. The

email: [email protected]

"

edi and the internet DRS. A.U.N. JADOENATHMISIER Consultant and projectmanager, LifeLine Networks Currently, the majority of electronic communication in the Dutch health care system is based on EDI messaging. EDI is the automatic, electronic exchange of structured and normalized messages (based on EDIFACT standards) between computers of different organizations. Electronic messaging is mainly used to replace paper messages sent by institutions (hospitals, laboratories, etcetera) to general practitioners. The messages exchanged generally refer to patient admission and discharge, or test results (laboratory and radiology). The number of messages exchanged between hospitals and insurance companies (e.g. messages for checking insurance rights and accounting), and between general practitioners and pharmacists (prescription messages) is increasing gradually. Nevertheless, the number of messages exchanged electronically still presents only a fraction of the total number of messages exchanged. Moreover, the use of electronic messaging differs substantially between different regions of the Netherlands.One of the reasons institutions are hesitant to use electronic data change is their reluctance to invest in technology that is perceived as antiquated, on the other hand, the new possibilities that Internet can offer are being discussed more and more, especially with regard to the electronic exchange of messages. This raises the question, 'what does this imply for EDI?'. Is EDI still needed for exchange of information between applications when there is the Internet? EDI in comparison to the Internet VAN Currently most organizations use a Value Added Network (VAN) for their EDI messaging. VANs are exploited commercially by third parties. These provide a network infrastructure, technical support, a helpdesk, and assistance in solving EDI and communication problems. In addition to this, they offer services like configuring software, security, archiving of data, possibilities for audit trails (tracking-andtracing), etcetera LifeLine Networks BV is a provider of a VAN. Internet The Internet is a worldwide network and has over 100 million connections. It has already existed for more than 15 years, but has become known to a larger audience only recently. The introduction of graphically oriented browsers has increased the popularity of the Internet. These browsers present information in a simple, user-friendly and intuitive manner using standard, platformindependent Internet techniques (HTML, JAVA and TCP/IP). A consequence of the success of the Internet is the availability of Internet software at low cost. It is easier and cheaper to connect to the Internet, than to connect to a VAN. However, there is a significant difference between the unstructured communication of individuals on the Internet, and the structured exchange of data between computer applications using EDI. Although it is not likely that Internet mail will replace EDI messaging, the Internet could be used as an infrastructure for EDI, since EDI is

platform independent. However, the use of the Internet for EDI has some drawbacks compared to EDI communication through a closed VAN. In the following paragraphs these drawbacks will be explored in more detail. Responsibility The Internet consists of a collection of interconnected networks and it is not controlled centrally. Those who use e-mail regularly may have experienced that e-mail messages may occasionally get lost. Since there is no central control, no one can be held responsible for this. There are also no guarantees with respect to delivery times. It may take a few minutes for a message to get to its destination, but it may also take a few hours or even days. This depends on the availability of the Internet provider and on the volume of the other Internet traffic. The reliability of the Internet is determined by the weakest link in the chain of Internet Providers used for relaying the message. In contrast, a VAN is centrally controlled, it can be scaled to the number of users and traffic volume, messages can not get lost and guarantees for delivery times are offered. Security The Internet is a shared public infrastructure and as such the system is not highly secure. In fact, it is possible to read messages of others. Security measures on a VAN are more intensive. Sending messages through the Internet is often compared to sending a postcard through the postal services while sending a message through a VAN can be compared to sending a registered letter. I



Reliability On the Internet, the identity of a sender cannot be established, and people can deny ever having received or sent a message. In a VAN information related to the sender is automatically chained to the messages themselves. Therefore, it is possible to establish the route a message has followed from the moment it was sent, and whether it has been received. Messages can also be reproduced from the archives.

the organisation, phone directories, protocols, procedures, etcetera can be provided on an Intranet.

EDI within VPN When in the future the availability and reliability of the public Internet can be guaranteed, the infrastructure of the Internet can be used for sending and receiving messages. This can be achieved by creating so-called Virtual Private Networks (VPN). This essentially means realizing a VAN on the Internet.

EDI on the Internet Solutions are being provided to secure the content of (e-mail) messages on the Internet through, for example, the use of encryption techniques. These solutions require additional software, such as the popular encryption program PGP (Pretty Good Privacy). However, guarantees on availability and delivery times are not expected in the near future. Furthermore, it is not possible to perform extended audit-trails on messages sent via the Internet. This does not imply that the Internet and EDI cannot co-exist together. The techniques and standards used on the Internet are applicable to EDI messaging within a VAN. EDI using TCP/IP A disadvantage of EDI is the relatively large investments that have to be made acquiring special communication software. This is partly due to the fact that the traditional EDI networks are based on the X.400 communications protocol and special software is needed to use this protocol. By using Internet standards such as TCP/IP (a connection protocol), SMTP (Simple Mail Transfer Protocol) and MIME (Multipurpose Internet Mail Extensions), it will be easier to connect to the network because these protocols are supported by Windows 95 and most software distributors. Using TCP/IP, EDI messages can be exchanged without losing the added value of a VAN. Essentially, the best of both worlds, i.e. simple connection and standard protocols on one side, and authentication and reliability on the other hand, are combined into one. An additional advantage of TCP/IP is the potential use of WebEDI. WebEDI can be used when the software needed to make EDI messages is not available or the number of messages sent is too low to invest in it. Using WebEDI, pre-defined forms can be filled out and sent through the network using a browser (such as Netscape). These forms are centrally converted to EDI messages so the receiver can automatically process them.

A prerequisite for a VPN is the guaranteed security of the communication on the Internet. To achieve this, more and more solutions are becoming available on the market. Examples are Microsoft's PPTP (Point to Point Tunneling Protocol) and Digital's Alta Vista Tunnel. LifeLine has selected a platform-independent solution based on the standardized Ipsec tunneling protocol, using TimeStep products. Witrh these products, a virtual private network is formed by creating secure connections, also known as tunnels, using the insecure channels of the Internet. At the beginning of a tunnel, data is encrypted and at the end the data is decrypted. Each tunnel uses separate encryption keys which are agreed upon at initialization when setting up the connection. This means that even when someone can compromise the tunnel, one can only see unreadable encrypted data. Therefore VPNs solutions can provide for secure and safe communication on the Internet.

Conclusions It is expected that in the near future standard Internet techniques will be used for EDI messaging meaning that the EDIFACT standard of structured messages will continue to be used. The first step towards using Internet techniques will be the transition from the X.400 protocol towards TCP/IP. With the availability of new security techniques, VANs will increasingly use Internet structures and create Virtual Private Networks rather than build their own physical networks. As a result of these developments, techniques will be simplified. It is expected more messages will be exchanged and in the end investment and exploitation costs for the users will gradually decrease, resulting in lower costs per message. -

An Intranet is created when Internet standards are applied within an existing VAN. An Intranet is a separate network based on the Internet technology which is protected from the outside world and controlled by the provider. Furthermore, such an Intranet can also be used for communicating all kinds of information to those connected. This is particularly suitable for information that is used as reference material. For example, general information on HEALTH INFORMATION DEVELOPMENTS IN THE NETHERLANDS - APRIL 1999

Avoid that co-operation turns into a game of tug-of-war! Are you at present actively optimising the tuning and co-operation within your institution? Are you perhaps looking for co-operation with other institutions? You want this co-operation to be successful for your clients, for you and your potential partners, and you do not want it to end in a game of tug-ofwar? Make an appeal to the expertise of the 'Ondersteuningsgroep Gezondheidszorg' [Support Group Health Care]. Our consultants have vast experience in finding solutions for a healthy co-operation and the development and implementation of such solutions. The services we offer aim at co-operation within the care industry in the field of: • • • • • • •

'transmural care' (trans institutional care) total quality management business process re-designing of care processes information provision logistics and chain-care 'product typing' and management of care internal and external communication

We combine these services with skills such as methodical guidance, project management, process guidance, coaching and interim management. Among our clients are hospitals, domiciliary care, nursing homes and homes for the elderly, regional and nation-wide platforms. In order to be able to serve you optimally we ourselves too emphasise internal co-operation. The consultants of the 'Ondersteuningsgroep Gezondheidszorg' are physicians, qualified nurses, nursing scientists, business administration experts or communications experts. They combine their knowledge in order to be able to present the best possible multidisciplinary solution fitting your situation. Together with you our consultants view your situation in a slightly broader perspective. This allows us to create space for healthy advise. Our co-operation however does not stop here. Besides a core team formed by our own consultants the 'Ondersteuningsgroep Gezondheidszorg' also has a network of experts in employment with care institutions: people who are involved, people who are familiar with the issues in the care industry and its daily practise. Because of their knowledge and experience and our approach we can deploy these experts on your projects. If you require more information: Our consultants are keen to answer the questions concerning your organisation and are looking forward to tell you more about the services we render.

Ondersteuningsgroep Gezondheidszorg Weg door Jonkerbos 100 6532 SZ Nijmegen The Netherlands phone: (024) 3658656 fax: (024) 3658886 E-mail: [email protected] Internet: http//:www.og.nl

selection of items for the nursing minimum data set W.T.F. GOOSSEN, T. FEUTH, P.J.M.M. EPPING, M. COUSIJN

Introduction Currently, there is no systematic collection of nursing care data available in the Netherlands, while pressure from the profession, health care at large and society is growing to legitimate the contribution of nursing and its costs (2,3). A powerful yet limited set of nursing data could be advantageous for nursing to make visible the care delivered, and this is often promoted as a nursing minimum data set (NMDS) (8,1,7). A NMDS has been defined by Werley and others as: 'a minimum data set of items of information with uniform definitions and categories concerning the specific dimension of nursing, which meets the information needs of multiple data users in the health care system' (8). The systematic collection of nursing care data allows to describe patients' needs for nursing care, and to visualise the delivery of nursing care. This collection of nursing data should further allow for comparison, quality assurance, management, research and funding of nursing care (8,1,10,7). A comparison of NMDS systems in use and under development revealed that two countries currently use an NMDS, and that several others are developing such a systematic collection of nursing data. The initiative for a NMDS started in the USA, and currently different uses of the US NMDS have been reported (8,9,6). Since January 1, 1988, all Belgian general hospitals are required by law to collect data for an NMDS, four times a year (the 'Minimale Verpleegkundige Gegevens (MVG)/ Resume Infirmier Minimum (RIM)') (7). Other countries developing NMDS include Canada, Australia and Switzerland, among others (4). The Nursing Minimum Data Set for the Netherlands project develops a method to describe the diversity and complexity of patients and their problems for different patient populations, and the variability of patient related nursing care activities. Based on these nursing activities it must be possible to calculate the nursing workload and to determine general indicators for the quality of nursing care by means of benchmark information. The purpose of this contribution is to explain how the categories and data items were identified for inclusion in the Nursing Minimum Data Set for the Netherlands.

Methodology Eight hospitals, with a total of 16 wards, participated in the study. A multi-method explorative approach was used to establish the NMDSN items. This included two rounds of interviews with nurses, head nurses and nurse managers from each participating ward. Further methods involved document analysis, and

drawing up lists of most frequently occurring patient problems, interventions and results of care. The categories and items were compared with literature about minimum data sets, patient problems, nursing interventions and nursing outcomes. Further, these data items were compared with existing Dutch systems to measure complexity of care and patient classification that are currently used for workload measurement. During the selection and development process, several draft versions were sent to the hospitals to obtain feedback in order to achieve consensus about the categories and items.

Results A set of categories and items is established and can be related to medical, demographic, and hospital data. The categories and nursing data items that emerged after analysis and grouping of the material include: - hospital related items: hospital, ward, speciality, type of nursing delivery system, date of data collection; - patient demographics items: sex, year of birth, admission and discharge dates, unique patient code and age; - medical condition items: this includes the medical diagnose(s) and complications, stability and predictability of the health situation and life threatening situations; - nursing process items: including the phases of the nursing process, except carrying out care activities; - patient problems: a collection of the most frequently occurring patient problems including information needs, communication, uncertainty, fear, stress, ADL and self care, vitals, pain, risks, elimination, breathing, food intake, bedsores, sleep, among others; - nursing interventions: including instruction, psychosocial support, assisting with feeding, breathing, ADL, self care, disorientation, mouth care, prevention of bedsores and risks, checking vital signs, giving medications, infusion, wound care, sampling, preparation and aftercare for tests and operations, among others; - results of nursing care: Including patient satisfaction about information given, pain alleviation, and care in general. In addition, the number of patient falls is included; - complexity of care. In the above categories, factors that determine complexity of nursing care are included. Additional items about complexity of care include: a) the number of patient problems and nursing interventions, and b) two visual analogue scales to score the complexity, and to score whether the •


amount of care that could be given was appropriate. Every item could be traced in the existing documentation systems, the interviews, the literature, and /or in the lists of patient problems and interventions. The lists from the hospitals with results of nursing care proved to be very heterogeneous at this stage, and gave no more information than the lists of problems and interventions. Therefore it was decided to use some outcome indicators from the American Nurses Association (Pollard, Andres, Dobson, 1996).

Conclusion A set of categories and nursing care data items is established and can be related to hospital, demographic and medical data. The content validity is supported by its consistency with the literature, by findings from practice documentation, and by the judgement of potential users. The nursing outcomes that have been included could be traced in existing documentation systems, but this category needs further development in the future. The collection of the data set and its value for the purpose will be tested in practice in 9 hospitals. During one week, nurses will daily complete the NMDSN as a separate list for every patient. Then, it will be possible to determine whether the NMDSN meets the expectations.

[8]

Werley HH, Lang NM (1988). Identification of the Nursing Minimum Data Set. New York: Springer. [9] Werley HH, Devine EC, Zorn CR, Ryan P, Westra B (1991). Nursing Minimum Data: abstraction tool for standardized, comparable, essential data. American Journal of Public Health, 18, 4, 421-426. [10] Zielstorff RD, Hudgings Cl, Grobe SJ (1993). Next-Generation Nursing Information Systems, Essential characteristics for Professional Practice. Washington, American Nurses Publishing.

Correspondence Address W. Goossen, Co-ordinator Nursing Minimum Data Set the Netherlands Project, School of nursing PO box 1080 8900 CB Leeuwarden the Netherlands