Stream 3 National Hospital Discharge Summary

Clinical Information Project Phase 1 Report PART C Stream 3: National Hospital Discharge Summary Version: Issued: Status: 1.1 19 April 2004 Final The Clinical Information Project is a critical element of national information infrastructure (“InfoStructure”) development, and has been managed by the South Australian Department of Human Services (SA DHS) under the overall direction of the National InfoStructure Development unit. National InfoStructure Development (NID) operates as an organisational unit within the Health Information Policy Branch of the Department of Health and Ageing (DoHA), but undertakes projects of national interest and is jointly funded by DoHA and Queensland Health. Other states and territories have provided resources for specific infostructure development projects. NID is located within Queensland Health’s Information Services in Brisbane. “InfoStructure” may be defined as information infrastructure that provides shared resources and standards for healthcare agents/parties that enable information to flow in appropriately structured, identifiable (unambiguous) and secure ways. Funding for the Clinical Information Project was contributed equally from SA DHS and DoHA. Document Control Title Author Another Reader File Name Clinical Information Project – Part C Lorraine Rayson, Scott Germann Frida Cheok, Eric Browne, Linda May, Helen Murray, CIP Phase 1 Steering Committee CIP Phase 1 Report Part C-V1.1.doc Date Amendment Log Version 0.1 0.2 0.3 0.4 0.5 0.6 1.0 1.1 Reason Initial Outline Initial Draft with Scoping Survey data included. Follow-up Survey about to be sent out. Reformatted Draft Revised table format for surveys Draft for Steering Committee Draft after Quality Reviewer Comments Final Report for HealthConnect Board Amended as requested by the Customer Date 18/7/03 26/9/03 18/10/03 05/01/04 31/01/04 02/03/04 22/03/04 19/04/04 CIP Phase 1 Documents Title Overview Part A – Stream 1: Clinical Information Framework Appendix A – Stream 1: Clinical Information Framework Part B – Stream 2: Event Summaries, Lists and Views Appendix B – Stream 2: Event Summaries, Lists and Views Part C – Stream 3: National Hospital Discharge Summary Appendix C – Stream 3: National Hospital Discharge Summary Specification for a National Hospital Discharge Summary. Description Describes the context of the Clinical Information Project; processes relating to the project as a whole; outcomes; recommendations. Clinical information framework, including the Health Information Model, for determining clinical information exchange requirements for shared EHRs. Prototype priority Event Summaries (inputs into the EHR) and priority Lists and Views (outputs of the EHR). TABLE OF CONTENTS EXECUTIVE SUMMARY ............................................... III ACRONYMS AND ABBREVIATIONS....................................V 1 2 INTRODUCTION ...............................................1 BACKGROUND.................................................2 2.1 2.2 3 HealthConnect.........................................2 CIP Working Relationships with NID and Design Sections ................................................2 SCOPING AND CONSULTATION PROCESS ...................3 3.1 3.2 3.3 3.4 Approach...............................................3 Comparison Matrix ....................................3 Data Elements used in the Example Discharge Summaries .............................................4 Consultations and Surveys ............................6 4 FEEDBACK FROM CONSULTATIONS .........................9 4.1 Scoping Survey and Follow-up Survey................9 5 RECOMMENDED NATIONAL HOSPITAL DISCHARGE SUMMARY DATA SET........................................ 14 5.1 5.2 CIP Specification Structures ........................ 14 Specification of Mandatory/Recommended Data Elements ............................................. 19 6 7 8 9 IMPLEMENTATION STRATEGY.............................. 21 ISSUES ARISING ............................................. 23 WAYS TO PROGRESS THE NATIONAL HOSPITAL DISCHARGE SUMMARY.................................................... 24 RECOMMENDATIONS........................................ 25 CIP PHASE 1 REPORT PART C V1.1.DOC i ii CIP PHASE 1 REPORT PART C V1.1.DOC EXECUTIVE SUMMARY Whenever a consumer’s healthcare is transferred between one healthcare provider or facility and another, it is essential that a summary of the care provided and any proposed future management be provided in a timely fashion. The information provided must be accurate, relevant, legible and, ideally, complete (or if not a note included as to any known deficiencies). The current project looked at the situation as it was in 2003 regarding information transfer (discharge summary) of consumers leaving the in-patient hospital environment and returning to the care of a GP in the community. It did not examine the equally important transfer of information from a GP to a hospital or other healthcare facility, e.g. nursing home, nor referral from a GP to a clinical specialist or another GP. Similarly when a consumer leaves a hospital outpatient clinic or emergency department (often when not yet fully investigated or treated) or is referred back to a GP by a clinical specialist, information must be transferred. However, from the studies undertaken, it is apparent that the data elements of information required and the timeliness of its availability are similar to a hospital in-patient discharge. The quality and timeliness of current discharge summaries varies enormously from hospital to hospital and state to state. Some hospitals are using electronically assisted generation of summaries either hospital-wide or in selected departments. Some are using template driven manually generated summaries. Some are using free text only with little or no guidance to the information to be included. Whilst many summaries are now typed and thus legible, the situation still exists where the distributed summary is handwritten (and often a carbon copy and illegible). Methods of distribution also varied from a consumer carried document, a document posted, faxed or autofaxed, distributed by courier, made available electronically from a website or directly e-mailed. Many hospitals used a combination of these methods. Summaries containing a subset of information were often sent on admission, during the in-patient stay, as an interim or final on discharge, as a revised final after review by the treating consultant or at the completion of coding. Traditionally the clinical information included in the summary is determined by the hospital. This varies from hospital department (discipline) to department and from hospital to hospital. There is little uniformity of either content or presentation. Often the clinical synopsis (if present) is less than concise and information regarding the ongoing management required/proposed is inadequate. When South Australian GP’s were surveyed in 2000, they placed little value on the clinical synopsis and in-patient treatment but wanted information on discharge medication (and particularly new or altered medication), future hospital outpatient or specialist appointments, referrals to external agencies and any specific ongoing management they were expected to provide. They saw the summary as a referral for future management rather than a history of past management. Returning these views to the hospital based authors of summaries highlighted the fact that currently the one summary serves several other purposes – for example a synopsis for the hospital record of the inpatient management that occurred (particularly if elements of the history or examination were unusual, e.g. cardiac disease presenting with right rather than left sided pain), a source document that contributes to quality assurance, accreditation purposes and coding (on which their funding frequently depended) as well as a communication to other healthcare providers. Whilst these may all represent different “views” of the same information, it was apparent that the compilation of this source information which required expensive clinical input (both in time and dollars) was likely to occur only once. The project was undertaken by members of the South Australian Department of Human Services (SA DHS) who previously in 2000 had conducted consultations with members of the South Australian Divisions of General Practice (SADI) to determine the data elements they wanted to receive in a discharge summary and its presentation format. These data elements then formed the basis for the development of an electronically generated and distributed discharge summary from the Adelaidewide public hospital OACIS clinical information system. As part of this development, a registry of GPs was established by SADI and linked to the receiving doctor functionality contained within the OACIS separation summary module. For the current project, a list of these data elements was circulated to a number of interstate hospitals, in order to obtain further input. This led, via the development of a data matrix, to an initial data set that was used in the scoping consultation. The consultation involved visits to most states and territories by 2 members of the project team in late 2002 and early CIP PHASE 1 REPORT PART C V1.1.DOC iii 2003. Consultation was held with Divisions of General Practice, hospitals, HealthConnect trials and other stakeholders who attended the CIP multidisciplinary workshops. From these visits a draft data set was developed and further nationwide consultations occurred in mid 2003. The second round of consultation was conducted as face to face meetings and a survey, utilising a draft data set and a mock up discharge summary that was both e-mailed and available from a web browser. Feedback from this process has led to the development of the current data set contained in this report. The data set is divided into a header (or administrative) section and a clinical section. Details of data elements, specifications and codes are included in this report and its appendices. A recommendation is made as to what should be the mandatory and the desirable data elements in any inpatient discharge summary. Some of the desirable data elements may vary depending on the length and/or complexity of the inpatient admission. As a general principle both sets of data elements would as well be applicable to other hospital encounters such as an emergency department or outpatient clinic visit. They apply to both manually and electronically generated summaries and should lay the foundation for future hospital discharge summaries nationally. Once this report is accepted it should be initially implemented by wide circulation to both authors (hospital based clinicians) and recipients, e.g. GPs, medical specialists, allied health professionals, community nurses, and personnel in private hospitals and nursing homes. Further feedback should be sought, a revised report completed and endorsed, and then published as a standard. Modification to existing discharge summaries is an ongoing process and this report will provide guidance as to current national requirements. Implementation of these requirements should occur on a hospital by hospital basis bearing in mind workflow practices and technologies in both the authoring and distribution processes. However, once a substantially uniform data set and presentation format is present in a discharge summary and if the summary is distributed and received ideally within 48 hours of discharge, the ongoing management of consumers back in the community will be substantially improved. From the CIP consultations there is strong support for a nationally endorsed standard. iv CIP PHASE 1 REPORT PART C V1.1.DOC ACRONYMS AND ABBREVIATIONS AHIC AHMAC CDA CEN CIP DE DoHA EHR GP HCDG HDSC HIM HL7 HMR ICTC IEC ISO NDDU NHDD NHIG NID SA DHS SADI UML Australian Health Information Council Australian Health Ministers Advisory Council Clinical Document Architecture Comité Européen de Normalisation Clinical Information Project Data Element Department of Health and Ageing Electronic Health Record General Practitioners HealthConnect Data Group Health Data Standards Committee Health Information Model Health Level Version 7 Home Medicines Review Information Technology Committee International Electro-technical Commission International Organization for Standardization National Data Development Unit National Health Data Dictionary National Health Information Group National InfoStructure Development South Australian Department of Human Services South Australian Divisions of General Practice Unified Modelling Language CIP PHASE 1 REPORT PART C V1.1.DOC v 1 INTRODUCTION The Clinical Information Project (CIP) focus was the clinical information content of shared EHRs for HealthConnect and the broader EHR agenda. The clinical information content relates to information that needs to be captured, stored and viewed to support the HealthConnect objective of improved delivery of health care and better quality of care. The project comprised three major components: Part A - Stream 1: Clinical Information Framework Stream 1 of the CIP provides a generic clinical information framework to guide the development of standardised national datasets for HealthConnect and other EHR/clinical communication projects. The framework assists in the collection and representation of clinica l information that is of high quality, relevant, interoperable (structured and incorporates standards and terminologies), and cognisant of workflow implications and links to evidence-based clinical practice. It includes a health information model (HIM), a taxonomy of health event summaries, criteria for determining what constitutes a HealthConnect health event summary, guiding principles for health event summary development and a process for determining and validating the information needs of HealthConnect users (which in turn will define the information to be collected through health event summaries). Part B - Stream 2: Health Event Summaries, EHR Lists and EHR Views The aim of Stream 2 was to determine a set of health event summaries, EHR lists and EHR views considered as immediate priorities to initiate HealthConnect specifically, but which would be equally valid for other shared care EHR/clinical communication initiatives. Health event summaries are central to the HealthConnect model. The development of both a framework for specification of health event summaries and key health event summaries themselves were fundamental areas of work. In conjunction, national datasets of high priority are being specified and will be submitted for inclusion in the National Health Data Dictionary (NHDD) following Phase 2. The prototypes were determined from the clinical information requirements of the HealthConnect trials and consultations with other key stakeholders and were used as the basis for verification of the clinical information framework. Part C - Stream 3: National Hospital Discharge Summary The aim of Stream 3 was to develop a proposed specification for a national hospital discharge summary and to investigate support for a minimum common standard and strategies for implementation. It is recognised that the extent and quality of discharge summaries currently in use in Australia is varied. This Stream was a previously endorsed project to develop a national hospital discharge summary and which was incorporated within the scope of the CIP on the basis that hospital discharge summaries are a priority for exchange of clinical information between providers and an important source of information into the shared EHR. Hence, the scope of the national hospital discharge summary extends beyond information required for shared EHRs, and relates to information required by the community healthcare provider for managing the care of consumers post-discharge. The focus was on discharge of patients from hospital to General Practitioner (GP) as the first priority for action. This document describes Stream 3 and associated project findings of the CIP as shown in Figure 1. In developing a national hospital discharge summary, it is recognised that clinicians have little or no spare time and will be unwilling and unable to commit to any data collection that adds substantially to consultation time. Strategies for ongoing development & implementation Stream 1 Develop a framework for clinical information capture and representation for HealthConnect Stream 2 Develop prototype event summaries, lists, and views for the HealthConnect trials and broader HealthConnect agenda Stream 3 Develop a proposed specification for a national hospital discharge summary and strategies for implementation Governance Options Terminology Issues Communication and stakeholder engagement and consultation Figure 1 - CIP Components CIP PHASE 1 REPORT PART C V1.1.DOC 1 2 2.1 BACKGROUND HealthConnect HealthConnect is the proposed national health information network, advocated by the National Electronic Health Records Taskforce (in ‘A Health Information Network For Australia’, July 2000). The objective of the network is improved delivery of healthcare and better quality of care, consumer safety and health outcomes for all Australians through a nationally coordinated and distributed system of electronic health records. In November 2000, Health Ministers agreed to fund two years research and development work over 2001 - 2003 to test the feasibility and value of HealthConnect ahead of any decision to proceed with its implementation on a national scale. A further 2 years' research and development funding was allocated in the 2003-04 Federal budget. The CIP was established in response to the needs of HealthConnect and for clinical information interchange more broadly. It is one of the key building blocks of activity identified as necessary to the feasibility of HealthConnect. 2.2 CIP Working Relationships with NID and Design Sections To reflect the broader role of the CIP within the clinical information interchange agenda administratively the project sits within the National InfoStructure Development (NID) unit and maintains close links with the Design section of the National e-Health Systems Branch (which is the Australian government component of the HealthConnect Program Office). Both groups are within the Information and Communications Division of the Australian Government Department of Health and Ageing (DoHA). NID was established in January 2002 to accelerate the development of information management infrastructure critical to HealthConnect and other clinical information interchange capacity, and to support the coherent development of health information management infrastructure nationally. The CIP component of NID contributes to the goal of semantic interoperability of clinical information in the Australian health system that is consistent with emerging international standards and directions where possible and appropriate. The standardisation of clinical information content is critical in enabling a national system of electronic health records. The Design section (previously Projects and Implementation section) has responsibility for work on the technical aspects of HealthConnect and MediConnect. Major deliverables have been the HealthConnect Business Architecture (version 1.0) and HealthConnect Draft Systems Architecture (version 0.9). The CIP contribution to the work of the Design section is in the Data Architecture component of the Systems Architecture. This is the very heart of the architecture in that it is determining the information to be collected and delivered by the HealthConnect system. 2 CIP PHASE 1 REPORT PART C V1.1.DOC 3 3.1 SCOPING AND CONSULTATION PROCESS Approach Examples of discharge summaries were sourced from various hospitals in each state of Australia. The purpose was to determine what information is currently provided consistently to GPs. From this information the following were developed for consultation purposes: § § § § A discharge summary comparison matrix; A ‘mock-up’ example discharge summary; A scoping survey for healthcare professionals; and A scoping survey for hospitals. The ‘mock-up’ was developed to provide a visual representation of the data elements only and was not an official rendition; it was merely a tool to assist in developing a draft standard data set. The ‘mock-up example discharge summary’ contained the data elements that were commonly contained in discharge summaries currently used (as indicated by the comparison matrix) and formed the basis for developing the scoping survey for healthcare professionals. This scoping survey sought feedback from stakeholders on these data elements as well as additional items and issues. The ‘mock-up’ and scoping survey were distributed and discussed during the initial state/territory visits undertaken between November 2002 and April 2003 by Frida Cheok and Scott Germann. They were also emailed to over 100 healthcare professionals and other interested parties that had been identified by the CIP office from the state visits and other consultations. A similar scoping survey for hospitals was emailed extensively where an email address could be obtained. For the hospitals from which examples of discharge summaries were sourced, additional information was sought, in an attempt to add value to the comparison matrix, as well as provide greater insight into what is actually happening in hospitals around Australia currently. As a result of the feedback received from the consultations and scoping survey for healthcare professionals, the mock-up and the survey were revised. These were used during a second round of state/territory visits in August/September 2003. In addition, a follow-up survey was developed and distributed by email through State Divisions of General Practice aiming to obtain wider feedback and input. The follow-up survey was also distributed by email to doctors on the CIP database. From the overall feedbacks received during the consultations and from both surveys, a recommended CIP discharge summary data set has been compiled. Specification for each data element has been developed. The data elements include those considered essential by most users, plus other data elements commonly requested by (mostly GP) users. This data set is not being put forward as a minimum data set, but rather as a relatively comprehensive set for consideration. Stakeholders may elect to include all or varying amount of items from this data set when compiling their discharge summaries. However, in the interest of achieving a national standard, CIP/HealthConnect may wish to recommend a core set of data elements, i.e. a minimum discharge summary data set, that should be included in the discharge summary template. 3.2 Comparison Matrix The examples of hospital discharge summaries obtained from 19 hospitals around Australia varied considerably, both in content and how certain fields were defined. Sometimes the intended content of listed fields was not clear, even when they were assigned the same name, and decisions were made about how to handle them for the purposes of the matrix. For example where a label ‘name’ was listed on the summary, it was assumed that this included surname and given name(s), but where a ‘signature’ was listed, it was not assumed that a ‘date’ (of signing) was included. Due to time constraints and resources, it was felt that the clarification of every field was not warranted, as the purpose of collecting samples was to identify commonly used fields. Hence, the matrix may not be a totally accurate representation for particular hospitals, but it is hoped that any inaccuracies are minor and do not detract from the purpose. The project is most grateful for the willingness of CIP PHASE 1 REPORT PART C V1.1.DOC 3 hospitals to share their experiences and to consent to the use of their summaries in this document and in the development of a national hospital discharge summary data set. For the comparison matrix table see Appendix C1. Appendix C2 provides detailed information on how the example discharge summaries were presented, what data they collected and the wording used on the examples. The matrix shows that all examples include: § § § § § § § § Patient ID; Patient Name; Patient DOB; Admission Date; Discharge Date; Principal Diagnosis; Medications; and Healthcare Provider(s) Name. Other data which was regularly (≥75%) included in the examples were: § Destination on Discharge; § Procedures; § § § § § § § § § § § § § Investigations; Comorbidities; Clinical Synopsis; Future Appointments/Referrals; Presenting Location Details; Author/Sender/Signatory of the summary; Patient Sex; and Healthcare Provider(s) Address. Patient Medicare Number; Injury/Poisoning/Accident Information; Social Issues; Functional performance at discharge; and ICU/Ventilation. Those fields least likely (≤20%) to be included in the examples were: These observations and other detail in the matrices were used in developing the CIP draft standard for consultation. The hospitals that provided discharge summary examples were also asked to complete a Hospital Details Form in an effort to better understand how discharge summaries are currently produced and stored. Key findings from the feedback about system characteristics show that the main method of external distribution of discharge summaries was via fax followed by post, via patient and then email, which was normally distributed to a community-based healthcare provider within 24 hours of the patients discharge. The author(s) of the discharge summaries were predominantly single authors and in most cases the author being either an intern or registrar. The main recipients of discharge summaries were community based healthcare providers followed by Medical Records and Patients/Carers. For a copy of the hospital details form used see Appendix C3, and the feedback from these forms summarised in Appendix C7.5. 3.3 Data Elements used in the Example Discharge Summaries From the example discharge summaries, it was evident that limited consistency exists, with the wording/phrasing of field names varying markedly as well as actual content. The overall presentation and appearance of each discharge summary was also examined. In Appendix C2 the content and wording/phrasing have been collated highlighting the variations. An extract from Appendix C2 below 4 CIP PHASE 1 REPORT PART C V1.1.DOC shows specific fields/wording and the hospitals, which included that field in superscript. It can be seen that all discharge summaries included the hospital’s name but: § § § Only 41% included the telephone number; 35% the address; and 18% a fax number. Eight different naming conventions were used to identify the summary with the most popular being ‘Discharge Summary’ which was used by 47% of the hospitals. In the example Allergies/Alerts extract below, less that 60% of the discharge summaries collected this information. The example shows the wording from the discharge summaries that include both the heading and/or individual data elements. The wording ranges from: § Adverse Drug Reactions used by 2 hospitals; § Alerts & Allergies (2); § § § § § § Allergy/Allergies (5); Allergies/Adverse Reactions (1); Medication Allergies (1); Known Adverse Drug Reactions and New Adverse Drug Reactions (1); Known Drug Allergies (1); and Medication Allergies (1). The specific naming of data items on a discharge summary may not necessarily be an issue. However, should a national standard discharge summary become a reality (which there is wide support for), then decisions will need to be made for naming conventions, or at the very least clear understandings about the meaning of various labels. However, if discharge summary data are transmitted between EHRs, the issue of classification becomes critical to data integrity, and whether the data which is captured in a field, matches the definition/intention of that field. For a complete list of all the field names used, see Appendix C2. All samples received from hospitals have been de-identified by referring to them as Hx. Healthcare Facility Details § § § § § Hospital name H1, H2, H3, H4, H5, H6, H7, H8, H9, H10, H11, H12, H13, H14, H15, H16, H17; Hospital logo H1, H3, H4, H5, H6, H7, H8, H9, H14, H15; Hospital telephone number Hospital fax H1, H9, H10; Hospital address H1, H5, H9, H10, H11, H12, H15 ; H1, H9, H10, H12, H13, H15 . Naming Convention of the Summary § § § § § § § § Discharge referral to regular general practitioner Discharge Summary Discharge Summary Referral Form Full Transfer of Care Summary H9; Impatient Discharge Letter H12; Patient Discharge Report H11; Patient Separation Summary Separation summary H7. H8 H3, H4, H5, H10, H13, H14, H15, H17 H6 H2 ; ; ; ; Patient Identifier This field appeared in 100% of the discharge summary samples and was represented as: § § § § § HRN H13 ; H9 Medical Record Number MRN H5, H7; Unit No UR H6; H3, H4, H11 ; ; CIP PHASE 1 REPORT PART C V1.1.DOC 5 § § § UR Number H1, H14, H16, H17; URN H2, H8, H10; Details printed but no field name specified H12, H15. Allergies/Alerts This field appeared in 58.8% of the discharge summary samples and was represented as: § § § § § § § § § § § Adverse Drug Reactions H2, H12; Alerts & Allergies H9, H16; Allergy/Allergies H6, H7, H8, H12, H17; Allergies/Adverse Reactions H17; Known Adverse Drug Reactions H10; Known Drug Allergies H15; Medication Allergies H6; New Adverse Drug Reactions Other Information H6; Reaction H6, H17; Severity H17. H10 ; 3.4 3.4.1 Consultations and Surveys State Visits and Workshops A consultation process was undertaken in the state/territories by a series of visits and workshops. An initial dataset was prepared for use in this consultation process. This was developed with input from the discharge summaries obtained from hospitals around Australia. In late 2002 early 2003 two members of the CIP visited the states/territories presenting this dataset and seeking comments on both the content and importance or desirability of inclusion of each data element (scoping consultation). The feedback received from these consultations was extremely discerning and useful and assisted in progressing the work. A revised dataset was developed and used in follow up consultations and a scoping survey in mid to late 2003. Data elements were divided into header and clinical sections. Comments were also sought on discharge summaries in general: an example of these comments follows. Ultimately, a draft dataset was produced and this forms the basis of this report. 1. Prescriber system for when GPs want summaries; i.e. GP decides if want on admission as well as discharge, etc; 2. Don’t need provisional diagnosis unless notification of Admission; 3. Disagree with comment (from survey) on Adverse Reaction Management – don’t need; 4. Medication Profile – GP doesn’t need but would like patient to have it; 5. Should be national standard as to how medication fields populated in hospitals (pharmacy checked by MO??); 6. Past Medical History is important on Discharge Summary (one view) as hospital may collect data GP doesn’t have; 7. Group agree that don’t need Family History on Discharge Summary; i.e. not related to event; 8. Important for Procedures & Principle Diagnosis to be coded – just make it invisible if GP doesn’t want. 3.4.2 Scoping Survey for Healthcare Professionals The ‘mock-up’ example discharge summary and the scoping survey were used to seek feedback from GPs and other stakeholders at the workshops during the state/territory. Email was also used to reach over 100 healthcare professionals, including those who attended the workshops. For a copy of the scoping survey for healthcare professionals see Appendix C4.1 and for a copy of the ‘mock-up example discharge summary’ see Appendix C5.1. See Appendix C7.4 for a summary of responses and comments from the surveys. 6 CIP PHASE 1 REPORT PART C V1.1.DOC The purpose of targeting healthcare professionals for the scoping survey was: § § § To give stakeholders the opportunity to provide input and feedback into the content of a potential national hospital discharge summary; To determine the level of support for a minimum common standard in the discharge summary; and To determine the data elements which could be incorporated into a future electronic health record. The scope of the project was to develop a national hospital discharge summary focusing on patients discharged from hospital to their GP. Patients discharged to another healthcare facility, e.g. nursing home, another hospital; were out of scope. However, the opportunity was taken to collect some feedback for the type of discharges where information or relevant stakeholders presented during the course of the consultations. For example, some sample discharge summaries included such information, and whilst it was not included in the comparison matrix, the information was used to identify possible data elements listed in Part B of the scoping survey. During the consultations stakeholders were asked to refer to that section, and provide input, if they had an interest. 3.4.3 Scoping Survey for Hospitals A similar scoping survey for hospitals was developed and emailed extensively to hospitals around Australia, where email address could be obtained. The purpose of this mail out was: § § To determine the level of support for a minimum common standard in the discharge summary; and To seek information on what data is currently captured and stored electronically. Whilst the content was similar to the professionals document, this word document was presented as tables split into modules with space for comments after each module. The feedback received from this method was comprehensive and focused. For a copy of the Hospital scoping survey see Appendix C4.2 and see Appendix C7.4 for a summary of responses and comments from the surveys. 3.4.4 Follow-up Survey – Recipients of Discharge Summaries Using the feedback from the consultations and scoping survey for healthcare professionals, a followup survey was developed and sent to GPs through the Australian Divisions of General Practice and to other relevant stakeholders on the CIP database. This follow-up survey targeted recipients of discharge summaries and its purpose was: § § To give stakeholders the opportunity to provide input and feedback into the content of a potential national hospital discharge summary; and To determine the level of support for a minimum common standard in the discharge summary. The follow-up survey was designed as a form-fill type document that was posted on the web and the URL was emailed through the CIP office and the Australian Divisions of General Practice to healthcare professionals, the College of General Practitioners and other interested parties. See Appendix C6.2 for a copy of the online follow-up survey and see Appendix C7.4 for a summary of responses and comments from the surveys. 3.4.5 Divisions of General Practice Network Forum 2003 A CIP team member attended the above forum in November 2003 as part of the HealthConnect exhibition. The aim was to inform participants about the CIP project and to encourage participants to complete the follow-up survey that had been launched online several days earlier. Paper-based copies of the follow-up survey were available at the forum (see Appendix C6.1) as well as a revised ‘mock-up example discharge summary’ (see Appendix C5.2). 3.4.6 HL7 / IT14-6-6 Extensive work has been undertaken with Drs David Rowed and Stephen Chu, members of the Australian Standard IT14-6-6 Discharge-Referral Workgroup and HL7 Australia on developing CIP PHASE 1 REPORT PART C V1.1.DOC 7 messaging and implementation guides for the national discharge summary. This work has been undertaken in numerous intensive three to four day sessions throughout the project. Additional work has also been undertaken by Drs Rowed and Chu, seeking changes to the Australian Standard in an endeavour to incorporate the work undertaken by the CIP and in mapping the data elements to HL7 and CDA. Standards Australia's IT14-6-6 Discharge-Referral Technical Workgroup and HL7 Australia are currently developing HL7 V2.3.1 discharge-referral standard and implementation guide. This work is currently undergoing public review and balloting processes. Development work for a new version, v2.6, is currently underway which seeks to extend the discharge-referral standard to accommodate the requirements of the community health sector. The Comité Européen de Normalisation (CEN) has also developed the EHR Communication Standard (CEN 13606-EHRCom) which is claimed to be capable of also supporting the electronic communication of discharge summary and referral data. To enhance the credibility, and hence ensuring the acceptance of the CIP discharge summary standard, it is important that this work is harmonised with the outputs of the IT14-6-6/HL7 Australia discharge-referral and HL7 Clinical Document Architecture (CDA) standards, and if possible, the CEN 13606-EHRCom standard. The harmonisation work can also inform IT14-6-6, HL7 CDA, and CEN of enhancement needed to extend their standards to meet wider stakeholders/Australian requirements. Significant amounts of work has already been undertaken to map all the CIP discharge summary data elements to the IT14-6-6/HL7 Australia v2.3.1 and v2.6 standards. HL7 endorses the use of clinical LOINC codes as document level and section level header codes. The CIP team has investigated the use of clinical LOINC code as discharge summary section header code. A request for a set of discharge summary section header codes is currently being formulated and will be submitted to LOINC. It is anticipated that clinical LOINC codes for these section headers will be created and maintained by LOINC/Regenstrief. Mapping of the CIP discharge summary data to CDA involves iterative processes. Initial comparison of the CIP (and New Zealand health event summary data set) with the CDA Release 2 specification revealed a number of enhancement requirements for CDA. These requirements were mostly addressed through a number of meetings and conference calls. Mapping efforts conducted during January 2004 led to the identification of a number of additional issues, which are currently being addressed. It is anticipated 2-3 more mapping iterations are required to completely resolve all open issues. The CEN 13606-EHRCom standard is based on the openEHR standard. Given that the HealthConnect trail in Queensland is associated with openEHR, it is sensible and useful that the CIP discharge summary standard be harmonised with the CEN standard. This work will involve the examination of the CEN EHRCom standard, an attempt to map the CIP discharge summary data elements to the EHRCom data, to reconcile any discrepancies, and to submit enhancement requests to CEN in ways similar to the CDA experience. The outcome of this activity can also provide useful information for the Queensland trial. It is proposed that the CEN mapping and harmonisation work to begin in March/April 2004. For the copy of the discharge summary specifications and HL7/CDA mapping see Appendix C8. 8 CIP PHASE 1 REPORT PART C V1.1.DOC 4 4.1 FEEDBACK FROM CONSULTATIONS Scoping Survey and Follow-up Survey The number of responses received from the two surveys was disappointing. However, it needs to be borne in mind that these documents were also used at the face-to-face workshops and other consultations. Whilst the feedback from these workshops cannot be measured numerically by question or data element, the feedback from these face-to-face meetings was critical in determining the final proposed standard. Generally, the outcomes from the workshops supported the results from the surveys, providing greater confidence in the use of the survey results. During the Round 2 state/territory visits, the results of the scoping survey were presented at the workshops for discussion. The outcome of this feedback was used to develop the follow-up survey. The results presented below relate to the feedback from the surveys only. The reader is reminded that these were endorsed by many more individuals who attended the face-to-face meetings and GP specific and multi-disciplinary workshops in every state and territory. These workshops included hospital IT, records and clinical representatives as well community GPs, allied health workers, specialists, health department representatives and consumers. Thirty-one responses were received to the scoping survey for healthcare professionals and 37 from the follow-up survey. Of these: § § § § 33 (48.5%) responses were from recipients of discharge summaries; and 35 (51.5%) from non- recipients. Hospital (33.3%); Community (37.0%); or The responses from predominantly from healthcare professionals (83.8%) based either at: § A combination of the two (29.6%). The responses from non-healthcare professionals were mainly from other hospital or health department based people. The scoping survey for hospitals was emailed to over 90 hospitals around Australia and was in a Microsoft Word table format. Of the responses: § Twenty-seven (27), 81.5% were from rural hospitals, of which • • § 59.1% were from small hospitals 18.2% were from medium sized hospitals • 22.7% were from large hospitals; and 18.5% were from large metropolitan hospitals. The primary patient types that these hospitals treat varied: § Adults were the primary patient type in 100% of the hospitals with: • • 59.3% also treating Adult and Aged; and 33.3% treating Adult, Aged and Children. The feedback from these surveys are summarised below. For more details see Appendix C7.4. The results relate to: 68 responses from the scoping surveys and the follow-up survey of healthcare professionals; and 27 responses from the scoping survey sent to hospitals. § One of the aims of the surveys was to determine from stakeholders their view about the need for a standardised national hospital discharge summary. The feedback unanimously supports a standardised national hospital discharge summary with less than 4% ranking it as Not Desirable. The naming of the document was also asked about. Various options were suggested with the preferred naming being Discharge Summary (34.5%) followed by Discharge & Referral Summary (29.1%). Respondents were asked about the timing of discharge summaries and when they wanted to receive them: • 77.9% wanted summaries when a patient is discharged from hospital; § § CIP PHASE 1 REPORT PART C V1.1.DOC 9 • • • • § 67.6% when a patient had died; 61.8% when a patient was transferred to another healthcare facility; 60.3% when a patient discharged from E.D.; and 41.2% when a patient was admitted to hospital. Electronic messaging provides the ability to download specific items of data into clinician’s local computer systems. This could potentially update existing data fields (eg medications, allergies, procedures etc). More than 68% ranked this as essential or highly desirable with nearly 28% ranking it as optional and only 3.3% as not required. Over 83% ranked giving a patient a medication profile as essential or highly desirable. Over the counter and complementary medications being listed on a discharge summary were ranked essential by 19.7%, highly desirable by 48.2% and 32.1% as optional or not at all. Preadmission medications and during admission medications were seen as essential by 47.2% and 19.4% respectively, highly desirable by 27.8% and 41.7 and 30.5% as optional or not required. Over 56% of recipients wanted investigation details presented with a summary format in the body of the document with detailed information at the end, 33.7% only wanted the detailed information at the end of the document, while 9.6% wanted the detail in the body of the document. § § § § 4.1.1 Priority Ranking of Data Elements All rankings listed below have been grouped into two sections, one being information that would generally be contained in a header section and the other being the clinical information content. For more details on the priority ranking of data elements see Appendix C7.1.1, C7.1.2 and C7.1.3. Scoping Survey - Header Section From the scoping survey, data elements in the header section that were considered essential by recipients were: § § § Facility (Name, address & contact details); Patients Name; and Patients DOB. Other data elements that ranked greater or equal to 75% were: § Admission Date; § § § § § § § § § Discharge Date; Provider(s) Name; Destination on Discharge; Patients Address; Discharge Type (Transfer/ Discharge/Deceased); Date the Summary was Printed or Sent; Patient ID; Patients Sex; and Provider(s) Address. Scoping Survey – Clinical Content Section In the clinical content section Medication Details (Medication name including Generic Name/Strength inc strength unit/Form/Brand Name) and Medication Dose were ranked as essential by 100% of recipients. Other data elements that ranked greater or equal to 75% were: § § § § § § Description of the Principal diagnoses; Medication Frequency; Procedures; Adverse Reaction Type (Drug, Food, Miscellaneous Allergy, Miscellaneous Contraindication); Medication Treatment Duration; Investigation Procedures; 10 CIP PHASE 1 REPORT PART C V1.1.DOC § § § Dates of Future Appointments; Description of Adverse Reactions; and Medication Route. Follow-up Survey - Header Section From the follow-up survey, there were no data elements considered essential by recipients. Data elements that ranked between 75% and 99% were: § § § § § § § § § Organisation Name; Patients Name; Patients Address; Patients DOB; Admission Date; Discharge Date; Organisation Contact Details (Tel/Fax/Email); Summary Status (Interim/Final); and Patients Sex. Follow-up Survey – Clinical Content Section In the clinical content section, Discharge Medication Status (New/Changed/Ceased) and Adverse Reaction Description were ranked essential by all recipients. Other data elements that ranked greater or equal to 75% were: § § § § § § § § § § § § § § § § Description of Problems on Discharge; Clinical History (Description of Treatment, Procedures, Investigations, Duration of Treatment, Investigation Findings); Procedures (Description and Date); Discharge Medication Details (Generic Name, Strength, Form, Dose, Frequency, Treatment Duration, Route, Reason for Change); Adverse Reaction Type (Drug, Food, Miscellaneous Allergy, Miscellaneous Contra: indication); Medication Profile Details (Generic Name, Strength, Dose, Frequency, Author Name and Role, Reason for Change, Route, Form); Description and Type of Alerts; Principal Diagnosis Status (Interim/Final); Description of Principal Diagnosis; Clinical Synopsis Author Name and Role; Name and Role of the Author/Signatory of the Summary; Description of Complications; Care Team-Hospital Consultants Name; Investigations Abnormal Flag; Description of Presenting Problems; and Type of Services arranged in Care Plan. As can be seen from the above information, very few data elements were considered 100% ‘Essential’ but a considerable number were ranked highly as ‘Essential’ and “Highly Desirable’. In the header section: § § 3 of the 21 (14.3%) data elements from the scoping work and none from the follow-up survey were ranked 100% ‘Essential’ by recipients; and 6 of the 21 (28.6%) data elements from the scoping work and 4 of 24 (16.7%) data elements in the follow-up survey were ranked 100% ‘Essential’ by non-recipients. In the clinical content sections: § § 2 of the 106 (1.9%) data elements from the scoping work and 4 of 163 (2.5%) data elements in the follow-up survey were ranked 100% ‘Essential’ by recipients; and 4 of the 106 (3.8%) data elements from the scoping work and 8 of 163 (4.9%) data elements in the follow-up survey were ranked 100% ‘Essential’ by non-recipients. CIP PHASE 1 REPORT PART C V1.1.DOC 11 Non-recipients generally tended to rank data elements higher than recipients. Considering the above feedback, very few data elements specified in the draft standardised national hospital discharge summary have been recommended as mandatory. 4.1.2 Transmit Ranking of Data Elements We have used the term transmit to describe data that may be sent to an EHR, e.g. over 90% of respondents of the scoping survey thought that sending the Patient’s ID was ‘essential’ while only 11% thought that sending Family Clinical History Date was ‘essential’. Again all rankings listed below have been grouped into two sections, these being information that would generally be contained in a header section and the clinical information. For more details on the transmit ranking of data elements see Appendix C7.2.1. Scoping Survey – Header Section From the scoping survey, no data elements in the header section were considered essential by recipients. Other data elements that ranked between 75% and 99% were: § § § § § § § § § § § § Patients Name; Patients DOB; Admission Date; Discharge Date; Provider(s) Name; Patients ID; Healthcare Facility Details (Name, Address, Contact Details); Date Summary Printing and/or Sent; Discharge Type e.g. patient discharged from hospital, Transferred to another healthcare facility, Deceased no autopsy; Provider(s) Address; Patients Sex; and Status of the Discharge Summary (Interim/Final). Scoping Survey – Clinical Content Section In the clinical content section Medication Details (including Generic Name/Strength/ Form/Brand Name) were ranked as essential by 100% of recipients. Other data elements that ranked greater or equal to 75% were: § § § § § § § § Discharge Medications (Dose, Frequency, Route, Treatment Duration, Status [New, Ceased, Changed]); Procedures (Description and Date); Adverse Reaction Type (Drug, Food, Miscellaneous Allergy); Investigation Procedure; Adverse Reactions (Description and Reaction); Presenting Problems; Principal Diagnosis; and Care Teams GPs Name. 4.1.3 Data Currently Captured Electronically Scoping Survey – Header Section From the scoping survey, the only data element in the header section captured electronically by all hospitals was Destination on Discharge. Other data elements that ranked greater or equal to 75% were: § § § Facility Details (Name, Address, Contact Details); Patients Name; Patients DOB; 12 CIP PHASE 1 REPORT PART C V1.1.DOC § § § § § § § § § § Patients Sex; Provider(s) Name; Admission Date and Time; Discharge Date and Time; Patients ID; Patients Address; Patients Age; Discharge Type; Medicare Number; and Provider(s) Address and Contact Details. Scoping Survey – Clinical Content Section In the clinical content section no data elements were 100% captured. Other data elements that ranked between 75% and 99% were: § § § § § Procedures (Description, Code and Codeset); Care Teams (Primary Contacts Name and Contact Details, Consultants Name, GPs Name, Practice, Address and Contact Details); Principal Diagnoses (Description Code and Codeset); Other Conditions/Diagnoses (Description Code and Codeset); and Type of Adverse Reaction. Overall, in the header section an average of 84.5% of all the data elements are captured electronically and in the clinical content section 27.3%. Considering only at the clinical content section, the following breakdowns show that certain categories of data are captured more often than other categories: § § § § § § § § § § § § § § § Procedures 76.7%; Principal Diagnoses 67.6%; Other Conditions/Diagnoses 63.9%; Care Teams 61.1%; Presenting Problem(s) 37.0%; Adverse Reactions 23.5%; Medications 20.5%; Future Appointments 18.5%; Investigations 18.0%; Complications, Clinical Synopsis and Intended Distribution List 16.7%; Warnings 15.5%; Past Medical History 12.1%; Follow-up care plan 8.5%; Support Directives and Family Medical History 7.4%; and Risk/Preventative Factors and Lifestyle/Social Issues 5.6%. For more details on the data captured electronically by hospitals see Appendix C7.3.1 and C7.3.2. CIP PHASE 1 REPORT PART C V1.1.DOC 13 5 RECOMMENDED NATIONAL HOSPITAL DISCHARGE SUMMARY DATA SET CIP Specification Structures 5.1 In 2001, Health Online1 anticipated that the development and implementation of a national approach to electronic health records is a major undertaking that will require the components of electronic health records to be developed progressively. To that end, the HealthConnect Program Office was delegated to oversee an initial set of exploratory projects. One of the projects included the extension of the NHDD to cover the needs of electronic health records. In 2002 the Australian Health Ministers Advisory Council endorsed the NHDD as the authoritative source of standard definitions for use in clinical care delivery, including being the repository for electronic health record metadata generated by HealthConnect. To facilitate the specification of EHR data elements, CIP undertook to collaborate closely with the AIHW Knowledgebase redevelopment group, the National Data Development Unit (NDDU), to establish CIPs preliminary EHR metadata structures including a data element specification template. Ongoing consultation with the NDDU ensured that the template complies as closely as possible to the underlying Knowledgebase metadata structures which themselves were derived from the ISO/IEC 11179 international standard for the specification and standardisation of data elements. 5.1.1 Data Element Specification Template The following data element specification developed by the CIP in collaboration with the NDDU has three 'extra' fields than those required by the NHDD. Label on Output A field that allows the CIP to include what the physical Discharge Summary document may have printed on output. For example, the data element "Report Header - Patient AddressSuburb/Town/Locality" is suggested to be called "Patient Address" on output. Value Domain Source A place for the CIP to capture information about the value domain, that is, the set of permissible values for the data element. For instance, where the CIP sourced a suitable existing value domain, e.g. a predefined HL7 data table, or a value domain already specified in the NHDD. It is also a place to collect information about the consultations used by the CIP to develop suggested value domains. Information viewed as essential by the CIP for presenting to the Health Data Standards Committee (HDSC) as part of the data element approval process to show which targeted expert groups and interested parties were consulted. Without demonstrated consultation with authoritative groups, proposed new data elements and/or value domains are unlikely to be accepted by the HDSC. HealthConnect Data Group To be explained later in section 5.1.2, this field refers to the one or more HealthConnect Data Groups (HCDG) to which this element belongs. 1 Health Online, A Health Information Action Plan for Australia, Second Edition, September 2001. 14 CIP PHASE 1 REPORT PART C V1.1.DOC Data Element Name Identifying and Definitional Attributes Label on Output Identifier CIP Version Number Metadata Type Definition Synonymous Name Context Optional-CIP extra field Mandatory Optional Mandatory Mandatory Optional Optional Relational and Representational Attributes Type of Relationship Datatype Representation Class Unit of Measure Unit of Measure Precision Format Max. Character Quantity Value Domain Value Domain Source Guide for Use Derivation Rules Verification Rules Collection Methods Conditional (if composite) Mandatory (Boolean, Currency, Date/Time, Number, String) Mandatory (Average, Code, Date, Date Part, Identifier, Interval, Measure, Text, Time, Time Part, Total) Conditional (if the Representation Class is Currency or Measure) Conditional (if the Representation Class is Currency or Measure) Mandatory Conditional Conditional (if Representation Class=Code) Conditional (if Value Domain populated)-CIP extra field Optional Optional Optional Optional Administrative Attributes Submitting Organisation Steward Origin Comments HealthConnect-Clinical Information Project HealthConnect Optional Optional Data Set Specifications HealthConnect Data Group Date From Date To Optional-CIP extra field Optional Optional 5.1.2 HealthConnect Data Groups The data element specification template alone was found to be insufficient for the purposes of the CIP. To provide an intuitive and efficient data element management architecture, the CIP in collaboration with the NDDU developed a hierarchical classification structure. A higher level HCDG, based upon the existing National Minimum Data Set (NMDS) specification, was necessary to provide a mechanism to group individual data elements into either an event summary data group (highest level) or a heading data group. The Knowledgebase uses the NMDS to document the logical grouping of metadata items for mandatory data collection. However, the HCDG does not have any associated mandatory data collection requirement. The business rules associated with the HCDG are: § § A HCDG can exist at two grouping levels, as an 'event summary group' or as a 'heading group'; The HCDG contains the obligation and cardinality of the metadata items that apply to the specific purpose of that HCDG, i.e. the obligation and/or cardinality of the same metadata item may be different when used in another HCDG; A HCDG can contain other HCDGs; and Data Elements must 'belong' to at least one HCDG. § § CIP PHASE 1 REPORT PART C V1.1.DOC 15 The HCDG specification template is as follows: HCDG-Data Group/Event Summary/Heading Name Identifying and Definitional Attributes Label on Output Identifier CIP Version Number Metadata Type Description Context Optional-CIP extra field Optional Optional HealthConnect Data Group Mandatory Mandatory Metadata Items Included Metadata Item Name Mandatory Obligation Mandatory (Mandatory, Optional, Conditional) Maximum Occurrence Mandatory (Single, Multiple) Condition Mandatory (only if Obligation = Conditional) Administrative Attributes Submitting Organisation Steward Comments Data Set Specifications HealthConnect Data Group Date From Date To 16 CIP PHASE 1 REPORT PART C V1.1.DOC 5.1.3 CIP Metadata Specification Hierarchy To demonstrate the hierarchical specification structure Figure 2 uses the Discharge Summary event summary with selected headings and data elements. HCDG - Discharge Summary HCDG-Report Header HCDG-Presenting Problems HCDG-Principal Diagnosis HCDG-Medications HCDG-Adverse Reactions HCDG-Care Team HCDG-Report Header HCDG-Report Header-Patient ID HCDG-Report Header-Patient Name HCDG-Report Header-Patient Address HCDG-Report Header-Patient DOB DE-Report Header-Admission Date DE-Report Header-Reason for Discharge DE-Report Header-Patient Address-Address Line DE-Report Header-Patient Address-Suburb/Town DE-Report Header-Patient Address-State/Territory DE-Report Header-Patient Address-Postcode DE-Report Header-Patient Address-Address Type DE-Report Header-Patient Name-Given Names DE-Report Header-Patient Name-Surname DE-Report Header-Patient Name-Name Type DE-Report Header-Patient ID-Person Identifier DE-Report Header-Patient ID-Assigning Facility Figure 2 - CIP Metadata Specification Hierarchy Example The topmost level is the HCDG representing the event summary. The event summary HCDG contains the heading HCDGs and any individual data elements (not belonging to a heading) that the event summary is composed of. In turn, the heading HCDG refers to it's own component headings and individual data elements. By linking in this way, the CIP has the ability to assemble all grouped and individual data elements that belong to one event summary. CIP PHASE 1 REPORT PART C V1.1.DOC 17 5.1.4 EHR Architecture Mapping The European Standard CEN prEN 13606-1:2003 developed by the Comité Européen de Normalisation (European Committee for Standardisation) is part one of a four-part pre-standard relating to Electronic Healthcare Record Communications. The goal of the CEN 13606 standard is to specify a suitable generic representation of the information architecture required for interoperable communications between systems and services that need or provide EHR data. The standard is not intended to specify the internal architecture or database design of such systems, nor is it intended to prescribe the kinds of clinical applications that might request or contribute EHR data. The CEN 13606-1 information model comprises a set of classes and attributes and is known as the Reference Model. It is presented as a set of diagrams using the Unified Modelling Language (UML) together with formal documentation which describes each construct, and defines any associated cardinalities, data types, constraints and tem sets. The Reference Model can be reinterpreted and viewed as a set of high level 'logical building blocks' consisting of seven levels of information representation within an EHR, as illustrated in Figure 3. EHR The electronic health record for one person. Folders High-level organisation of the EHR, e.g. per episode, per clinical speciality. Set of entries comprising a clinical care session or document, e.g. test result, letter. Clinical headings/groupings reflecting the workflow and consultation process, e.g. diagnoses. Statements" representing complex or compound clinical concepts, e.g. diagnosis. Compound entries, test batteries, e.g. blood pressure, full blood count. Element entries: smallest (atomic) component of an entry with values, e.g. body weight. Date types for element values, e.g. coded terms, measurements with units. Figure 3 - Logical building blocks of the EHR Compositions Sections Entries Clusters Elements Data values 18 CIP PHASE 1 REPORT PART C V1.1.DOC The CIP specification hierarchy of the HCDG, used to categorise event summaries, headings and individual data elements, is represented in Figure 4 with examples. HCDG - Event Summary Data Group Ø Ø Ø Ø Ø Hospital Discharge Summary Initial Health Profile Pathology Imaging GP Consultation HCDG - Heading Data Group Ø Ø Ø Ø Adverse Reactions Family Clinical History Intended Recipients Clinical Synopses Individual Data Elements Ø Adverse Reactions Description Ø Sex Ø Date of Birth Figure 4 - CIP Specification Hierarchy (with examples) Mapping the CIP specification hierarchy to the seven levels of the CEN 13606-1 results in the following equivalences: HCDG (Event Summary) HCDG (Heading) Data Element ≡ ≡ ≡ Composition (any of) Section, Entry, Cluster Element 5.2 Specification of Mandatory/Recommended Data Elements From the feedback received to date, draft data element specifications have been written for the proposed national discharge summary. Very few of the data elements have been made mandatory but those that are mandatory in the Report Header section include: § § § § § § § § Patients ID; Patients Family Name; Patients DOB; Patients Sex; Providing Healthcare Facility Name; Report Status; Summary Purpose; and Admission Date. In the clinical content section, the mandatory data elements are: § Presenting Problems; § § Principal Diagnoses; Problems on Discharge; CIP PHASE 1 REPORT PART C V1.1.DOC 19 § § § § § § Procedures; Medications; Adverse Reactions; Alerts; Advance Directives; and Report Authoriser The data elements listed above in the clinical content section have been made mandatory even though they may not be relevant for every discharge summary. This was done because it was felt that they were significant data elements. For example, if no procedures were undertaken during the admission, rather than the field either not appearing on a discharge summary or it appearing with no data in it, an option of entering “None performed” should be made available. The same applies for Medications, which will have an option of “No changes”, while Adverse Reactions and Alerts will have the option of “None Known”. See Appendix C8.4 for the draft specifications. 5.2.1 CIP Recommended Standard The table below briefly summaries the obligation of each top level Discharge summary HCDG. Selected data elements within a HCDG will be mandatory. See Appendix C8.2 for a hierarchical representation of the HCDGs and C8.3 for more detailed HDCG specifications and data elements within each HCDG and C8.4 for the individual data element specifications. HealthConnect Data Group HCDG-Report Header HCDG-Presenting Problems HCDG-Principal Diagnoses HCDG-Other Conditions/Diagnoses HCDG-Complications HCDG-Problems on Discharge HCDG-Family Clinical History HCDG-Clinical History HCDG-Social Issues HCDG-Lifestyle HCDG-Risk Factors HCDG-Functional Status HCDG-Clinical Synopses HCDG-Procedures HCDG-Investigations HCDG-Medications HCDG-Adverse Reactions HCDG-Alerts HCDG-Advance Directives HCDG-Care Plan HCDG-Care Teams HCDG-Intended Recipients HCDG-Report Authoriser Obligation Mandatory Mandatory Mandatory Recommended Recommended Mandatory Optional Optional Optional Optional Optional Optional Recommended Mandatory Recommended Mandatory Mandatory Mandatory Mandatory Recommended Recommended Recommended Mandatory 20 CIP PHASE 1 REPORT PART C V1.1.DOC 6 IMPLEMENTATION STRATEGY Promulgation and Consultation Upon endorsement by the HealthConnect Board, CIP should forward the draft data element specifications to the AIHWs National Data Development Unit (NDDU) for technical review. Recommendation C1 The CIP, following endorsement by the HealthConnect Board, should forward the draft data element specifications to the AIHWs National Data Development Unit (NDDU) for technical review prior to submission to HDSC. Upon endorsement by the HealthConnect Board, CIP should forward Part C and Appendix C of the Phase 1 Report to the Information Technology Committee (ICTC) requesting they assess the feasibility of implementing the specifications within state hospital systems. Recommendation C2 The CIP, after endorsement by the HealthConnect Board, should send Part C and Appendix C of the Phase 1 Report to the ICTSC and then the ICTC, requesting they assess the feasibility of implementation within hospital systems. Upon endorsement from the HealthConnect Board, the health departments of the States and Territories should be encouraged to distribute Part C and Appendix C of the Phase 1 Report to their public hospitals for both comment and if appropriate use as an implementation guideline for a discharge summary. Distribution should also occur to the private medical sector, e.g. GPs, medical specialists, allied health professionals, community nursing, private hospitals and nursing homes, via their respective peak bodies for comment. Comments should be sought both generally on the content and recommendations of the report and specifically (from clinical and technical specialists) on the data elements (and any further stratification within the broad header and clinical sections), the detailed data specifications and the codes Recommendation C3 The CIP, after endorsement by the HealthConnect Board, should inform the relevant stakeholder bodies, e.g. AHIC, SIMC, state health departments and Quality & Safety Council regarding the proposed national discharge summary specification. Upon endorsement from the HealthConnect Board, the CIP will make the discharge summary specification available for public comment. Due to the size of the discharge summary specification, specific groups will need to be targeted to review data elements relevant to them. Recommendation C4 The discharge summary specification should be made available for public comment by the CIP after endorsement of the Phase 1 Report by the HealthConnect Board. CIP PHASE 1 REPORT PART C V1.1.DOC 21 Site implementation Site implementation is a State responsibility. Implementation of a discharge summary containing the recommended data elements and their presentation, i.e. content, format and viewing sequence; should occur progressively over several years on a hospital by hospital basis, bearing in mind the often unique workflow practices and technologies in both the authoring and distribution processes. The use of a standard data set and uniform presentation format will not only make the traditional discharge summary more readable and informative for ongoing consumer management by the recipient healthcare providers but also enable the extraction of standard data elements into a HealthConnect or other summary. Within these summaries this will allow further data manipulation to occur in order to provide a variety of views to different users and for different purposes. A standard dataset is a requirement for software developers of both authoring and recipient systems and will enable the use of HL7 messaging to directly update fields within recipient’s clinical information systems. 22 CIP PHASE 1 REPORT PART C V1.1.DOC 7 ISSUES ARISING Business Process and Workflow The development of a robust and accepted dataset is but one step in the implementation of a discharge summary. More important are the business processes at the authoring hospital (both in ‘writing’ and distributing the summary) and at the recipient healthcare professional’s workplace. Important questions that need to be answered include: § § § Who is going to enter the data? How will the data be entered? e.g. palm pilot; and, What funds will be available for both human and other resources? For standardisation to be achieved across healthcare systems, the above issues would need to be addressed. This would ensure that healthcare providers/facilities/services, e.g. software developers would have standardised guidelines by which to work. CIP PHASE 1 REPORT PART C V1.1.DOC 23 8 WAYS TO PROGRESS THE NATIONAL HOSPITAL DISCHARGE SUMMARY Data Element Value Domains All the data elements and their associated value domains, that is the set of permissible values, need to go through a process of wider development work and consultation. Value domains can be coded in three ways: 1. Uncoded (free text, numeric values), e.g. Systolic blood pressure; 2. Coded according to existing national/international standards, e.g. LOINC for pathology results; and 3. Coded according to yet to be developed HealthConnect code-sets, e.g. adverse reaction severity. As a matter of principle, uncoded free text (1) is to be avoided. Considerable effort and consultation will be required for (3). Discharge Medication Management Feedback received proposes that the national discharge summary include a section that recommends a Home Medicines Review (HMR). HMR is a service to consumers living at home in the community. The goal of HMR, as a consumer focussed service, is to maximise an individual consumer's benefit from their medication regimen, and prevent medication-related problems through a team approach, involving the consumer's general practitioner and preferred community pharmacy, with the consumer as the central focus. It may also involve other relevant members of the healthcare team, such as nurses in community practice or carers. The HMR process utilises the specific knowledge and expertise of each of the healthcare professionals involved. In collaboration with the general practitioner, a pharmacist comprehensively reviews the consumer's medication regimen in a home visit. After discussion of the visit findings and report with the pharmacist, the general practitioner and consumer agree on a medication management plan. The consumer is central in the development and implementation of this plan with their general practitioner. HMRs are entirely voluntary. Many patients leave hospital greatly confused regarding the medication they are to take and the ones they are to cease. Equally, there could be confusion surrounding directions, medication storage and compliance. Some hospitals are discussing the idea of discharging patients from hospital with 30 days worth of medication so an initiative that recommends a HMR to occur 1-2 weeks after hospital discharge could prevent many medication-related problems. Such recommendations that involve changes to current work practices need further investigation and consultation. 24 CIP PHASE 1 REPORT PART C V1.1.DOC 9 RECOMMENDATIONS Recommendation C1 (refer to Section 6) The CIP, following endorsement by the HealthConnect Board, should forward the draft data element specifications to the AIHWs National Data Development Unit (NDDU) for technical review prior to submission to HDSC. Recommendation C2 (refer to Section 6) The CIP, after endorsement by the HealthConnect Board, should send Part C and Appendix C of the Phase 1 Report to the ICTSC and then the ICTC, requesting they assess the feasibility of implementation within hospital systems. Recommendation C3 (refer to Section 6) The CIP, after endorsement by the HealthConnect Board, should inform the relevant stakeholder bodies, e.g. AHIC, SIMC, state health departments and Quality & Safety Council regarding the proposed national discharge summary specification. Recommendation C4 (refer to Section 6) The discharge summary specification should be made available for public comment by the CIP after endorsement of the Phase 1 Report by the HealthConnect Board. CIP PHASE 1 REPORT PART C V1.1.DOC 25