Mental health informatics and the rhythm of community careBMJ 1995; 311 doi: https://doi.org/10.1136/bmj.311.7012.1038 (Published 21 October 1995) Cite this as: BMJ 1995;311:1038
- Gyles Glover, Senior lecturer
- Department of Public Health and Primary Care, Chelsea and Westminster Hospital, London SW10 9NH
- Dr Glover is seconded part time to the Department of Health to work on the mental health minimum data set project.
Information systems in psychiatry must be released from the asylums and updated
Clinical care in British psychiatry has been changing radically in the past two decades. Lifelong care for chronically disabled patients in institutions is giving way to networks of day hospitals, mental health centres, and small residential units. Care in hospital is now reserved largely for episodes of acute mental illness. But the information systems that record what patients, staff, and services are doing are still rooted in the asylum, and lack of appropriate data is impeding clinical work and the development of services. The Department of Health is reviewing the mental health minimum dataset (the information that NHS provider units must pass to purchasers and to the department). This offers an important chance to ensure that the right data are collected in the right way.
What information should be collected, and who needs to use it? At the point of care, staff need user friendly, accurate, and confidential information about patients' past and current problems and care, particularly when emergencies arise outside normal working hours. Electronically held current clinical summaries can also ensure that staff know their responsibilities in each patient's care.1 The Clinical Standards Advisory Group has recently emphasised the value of maintaining a register of patients' care plans.2 Providers managing clinical teams need to deploy staff and other resources around the catchment area according to need and to allocate patients in realistic numbers to appropriately skilled staff: excessive loads impede effective care and can lead to burnout. To achieve these aims efficiently and to make clinical audit a routine part of their work providers need the right information. Finally, purchasers need to know the numbers of patients being cared for, the range of clinical problems, and the outcomes of care if they are to judge the value of the services they are buying and to forecast future needs.
Nationally, the government needs to determine its mental health policies. One of its key policies, the care programme approach, specifies that people accepted by specialist mental health services should have individual care plans drawn up; these should be coordinated by key workers and based on assessment of the patient's needs for health and social care and reviewed periodically.3 The Health of the Nation (the government's health strategy for the 1990s and beyond) goes further, setting improvement in the health and social functioning of mentally ill people as the first mental health target and proposing that success must be quantified.4 The limited development of information systems is impeding each of these aims.
Three issues underlie the slow progress in developing mental health informatics. Firstly, although community care has been long anticipated, the new configurations of services have emerged only relatively recently. A decade ago, with most care still being provided in hospitals scheduled for closure, investment in computer systems seemed premature. Secondly, the developing network of small and dispersed centres of care called for wide computer networks, which until recently were expensive and unreliable.5 Thirdly, and perhaps most significantly, the minimum datasets that managers have been required to collect have had little practical use and raised little enthusiasm. Based on the approach defined by the Korner review in the early 1980s,6 the datasets treat mental health care as episodic and mainly based in hospital. The activity of each group of staff caring for outpatients is collated separately so that the extent to which they share care cannot be identified—even if the care is multidisciplinary the datasets are not.
The first two problems are now historical. The running down of most asylums scheduled for closure is well advanced and mental health services in the community are mostly in place (although facilities for social care are lagging behind). Linkages between remote computer systems are now cheap and reliable. The inappropriateness of the dataset therefore is the remaining impediment.
A dataset specifies what items of information are to be recorded and when this should occur. Two criteria determine a dataset's usefulness: it must contain the required facts, and information must be collected at convenient and logical points. It should, therefore, reflect the rhythm of staff activity. The present NHS dataset has two important omissions. Firstly, no details of care provided by local authorities or independent agencies are recorded even though these are key components in multidisciplinary community care. Secondly, standardised measures of medical and social problems—information needed for assessing both the process and outcome—are not included.
Two projects funded by the Department of Health should solve the second problem. The research unit of the Royal College of Psychiatrists is developing a scoring system (Health of the Nation Outcome Scale for routine clinical measurement of the health and social functioning of mentally ill people).7 In addition, the NHS Casemix Office is developing a new method of classifying mentally ill people by using diagnosis and problem scores, which should improve the prediction of needs for care.8 9 Both projects use the same scoring system and provide information that should be included in a current clinical summary.
Until recently there has been no satisfactory way of specifying at what point in a patient's care a dataset should be collected. The rhythm of care in the old asylums had just two peaks: the admission of people who had become mentally ill and the discharge of those who had recovered. The move into the community started in a wide variety of ways in different places and muffled the clarity of that rhythm's beat. The care programme approach offers a new rhythm. Care proceeds from review to review, with key workers monitoring what happens in between. In simple cases reviews are conducted between the patient and a single professional. For more complicated problems multidisciplinary teams meet. Either way the pattern of assessment, care plan, keyworking, and reassessment is the same. Increasingly, pioneering provider units are developing information systems that collate care plans based on the reviews.
There are good reasons for the Department of Health to follow this lead and base its minimum dataset on reviews of care. The data would be relatively easy to collect; could cover all aspects of patients' care, including those provided outside the NHS; and should give staff useful lists and up to date summaries about their own patients. If held on an appropriate computer network the dataset would give ready access to the key facts needed in emergencies. It would also provide managers and purchasers with rich data on the number of patients, the nature of their problems, the care delivered, and its outcome. Everyone needs information, but it has to be useful and usable.