« Previous | Contents | Next »
Listen
SCOTTISH DIABETES CORE DATASET
INTRODUCTIONRecent years have seen major developments in diabetes services in Scotland. This record of innovation and action, combined with a growing recognition of the pressures being caused by the increasing prevalence of diabetes, led the Scottish Executive to highlight diabetes as a national priority in the Scottish Health Plan -
Our National Health: a plan for action, a plan for change. The
Scottish Diabetes Framework, published in April 2002, drew together the various strands of activity and set out a series of action points and milestones. A critical part of the
Scottish Diabetes Framework is the development and implementation of a robust clinical management system to support diabetes care. This is being taken forward as the Scottish Care Information Diabetes Collaboration (SCI-DC). One of the core principles of information systems is compatibility and transferability of data. This necessitates agreement on the dataset - what data items should be collected and how these items should be defined. In September 2000 CRAG published,
A Report by the Working Group on IT to Support Shared Care in Diabetes. A key deliverable of the working group, chaired by Dr Ewan Crawford, was to specify a core dataset suitable for sharing clinical information for the care of patients with diabetes, building on the SIGN minimum dataset and other published work. The report made clear that the dataset was only a starting point and that further development would be required to make it useable in clinical practice. This document presents a further iteration of the diabetes dataset. Full account has been taken of the concerted effort now being made to improve the IM&T infrastructure for diabetes care in Scotland. Purpose of the Core Dataset The principal purpose of this dataset is for it to be implemented as part of the SCI-Diabetes Collaboration work. The details of this work have been published elsewhere (available on www.DiabetesInScotland.org) but the SCI-DC programme will offer diabetes IM&T solutions to regions based upon two elements - SCI-DC clinical and SCI-DC network. Although it is expected that most areas will adopt the SCI-DC approach, this is not mandatory so publication of the core dataset is essential for the suppliers of other systems to ensure compatibility, and for the development of disease specific screens within primary care information systems. This will be the first detailed direct care dataset standard to be defined and implemented in Scotland. |
Concept of Core Record
The aim is to create a core electronic record that resides on a central server (for each NHS Board area). Clinical staff with appropriate permission will be able to access the diabetes records of patients in their care wherever this care is delivered and view some or all of the contained information (see
Concept of Views of the Data below). The core record will also be analysed to provide dynamic information for local audit and benchmarking as well as the details required for the annual Scottish Diabetes Survey.
By definition, this approach will generate a diabetes register that can also be used to furnish a call-recall system for diabetic eye screening.

Figure 1: The Scottish Diabetes Core Dataset and the architecture of diabetes data
Figure 1 makes it clear that the Scottish Diabetes Core Dataset does not, at present, incorporate all of the diabetes data that may be held on local systems which contribute information to the core record. This distinction is less obvious when the 'local system' actually constitutes browser access to the central record, but this architecture does not preclude there still being 'private' fields with only a subset being shared.
Thus, whether specialist data are held centrally or on systems in hospital or in primary care is irrelevant. What matters is the contribution to the core dataset and how this interaction is managed.
It is self evident that this dataset should not be construed as the entirety of data required for the care of a patient with diabetes. The test that was applied to proposed data items was, 'is this information essential to all those who may have to care for the patient without access to further information?'.
An obvious omission is the detail of Drug Therapy. Currently, this detail will be recorded on local systems but there is agreement that central recording should conform to the drug dictionary established for GPASS which is derived from the eVadis drug database supplied and maintained by ISD. The ultimate relationship to the evolving UK Clinical Products Reference Service (UKCPRS) remains to be seen.
Concept of Views of the Data
While this document details the elements of the Scottish Diabetes Core Dataset which potentially may be shared by health professionals, it is vital to understand that the actual data access that any individual obtains will be tailored and based upon function-related criteria (and patient preference). For example, a podiatrist would not be expected to have access to pregnancy-related information. This document is not about systems but suffice to say that the access control infrastructure will be managed at a local level.
Quite deliberately, no attempt has been made to apportion tasks to professional groups in relation to the population of the dataset. It is essential that local circumstances dictate best practice.
The inclusion of some sensitive data elements may cause concern. However, they need not be completed if patients would prefer this information to be maintained locally. These items are included in the dataset to allow sharing where this is deemed appropriate.
Summary of Changes
The current iteration has been coded (Read v2) to ensure compatibility with primary care systems. This is essential if these systems (principally GPASS) are going to be able to contribute to the dataset and/or incorporate elements from the dataset. Current national IM&T strategy suggests that there will be a transition to SNOMED-CT but this is not yet available in operational systems.
The original CRAG dataset comprised some 129 items while this version has 130. However, there have been significant changes to the Eye Screening section to reflect the work of the Health Technology Board for Scotland in preparation for a national diabetes eye screening programme. The pregnancy, childhood and dietetic sections have also been expanded but many of the items suggested by the subgroups set up to assist with the review process were felt to be too specialised for inclusion in the Scottish Diabetes Core Dataset. These items will, however, be incorporated within the SCI-DC development programme.
Next Steps
The immediate task is to roll out the SCI-DC elements across Scotland and to work with the SCI and GPASS teams and other suppliers (including CDSS) to facilitate contribution to this common dataset from all places from where diabetes care is delivered.
For NHS Boards, the Scottish Executive has made it clear in a recent circular - Developing services for people with diabetes, HDL (2002)81 - that all diabetes data collection systems in Scotland are expected, over time, to be compatible with this core dataset.
Developing the Dataset
It is self evident that diabetes does not occur in isolation and patients frequently may have other health problems e.g. cardiovascular disease. This poses real but surmountable challenges in relation to the evolution of integrated records - obviously data should be reusable so that once an item is collected, it is (potentially) available elsewhere.
In terms of dataset development, further additions and refinements will be made in the light of experience and to accommodate future data demands. Discussions with ISD are ongoing about how best to promulgate and maintain the diabetes dataset (and indeed other clinical datasets). In the meantime, SCI-DC will continue to provide the focus for dataset development. Comments about the dataset should be submitted to
Dr Lorna Ramsay, Clinical Advisor, Data Intelligence Group, ISD Scotland, Trinity Park House, South Trinity Road, Edinburgh EH5 3SQ.
It is anticipated that this will be the last version of the dataset to be published in this way. In future, it is likely that the dataset will be made available only in electronic form.
« Previous | Contents | Next »