• internet
• online drug information

Quality of care—be it at the level of the institution or the individual healthcare professional—is at the core of the modernisation of the NHS,¹ but nowhere is it more difficult to measure and assess than in the area of mental health. Unlike diabetes mellitus and, to some extent, coronary artery disease where there is good agreement on the absolute threshold for making the diagnosis and increasing consensus as to what processes constitute high quality care (e.g. retinal screening, use of aspirin), the definite diagnosis of a mental health problem and agreement on its optimum management is much less clear. This is particularly so for the large burden of mainly undifferentiated mental illness that presents in primary care. Although undifferentiated, it should not be assumed that the severity and accompanying functional impairment are not insubstantial. We are not, as some might assert, dealing with the “worried well”.

In secondary care, psychiatrists are better able to agree on the diagnosis, partly because they are more used to using internationally agreed classification systems. In addition, more research has been undertaken to examine both the process and content of care for patients with long term enduring mental illness—for example, integrated care pathways and a stepped approach to medication. With harder outcomes available for measurement, such as the number of hospital admissions, quality of care is both easier to define and to measure. However, in primary care where the various stakeholders are often unable to agree whether or not a mental illness is present, how severe it is, and the degree of impairment, consensus on management is much more difficult and developing quality indicators is an uphill struggle.

For many people the concept of quality indicators is a new one. If we agree that the range of professional practice with regard to mental illness in primary care is unacceptable and thus quality is often suboptimal, we must take steps to measure it objectively. In their recent publication Marshall et al² state that:

“Quality indicators are specific and measurable elements of practice that can be used to assess the quality of care. They are usually derived from retrospective reviews of medical records or routine information sources. Some authorities differentiate ‘quality’ from ‘activity’ or ‘performance’ indicators. The important issue is that a good quality indicator should define care that is attributable and within the control of the person who is delivering the care.”

It is the final sentence in this definition that is key. Quality is a complex construct that is likely to have a different meaning for patients, professionals, and managers. However, where objective evidence of variation exists—for example, in access, in effectiveness or in efficiency—patients and professionals are likely to agree that some process is required by which absolute and relative indices of quality can be measured. The task is to develop valid (quality) indicators for mental health care for which individual health professionals can take responsibility. What is particularly difficult in mental health is agreement on the gold standard.

The approach to developing quality indicators taken by Shield et al³ in their paper in this issue of QSHC incorporated the views of 11 different groups of stakeholders. After two rounds of a Delphi process, only one quarter of more than 300 indicators—derived from a variety of sources including published guidelines, focus groups, and the National Service Framework (NSF) for Mental Health⁴—were rated valid by all panels. Furthermore, whereas carers rated 91% of indicators as valid, GPs rated only 41% valid. Who has primacy here? The indicators on which there was consensus were grouped into 21 aspects of care, 11 relating to general practices and 10 to health authorities or primary care groups/trusts. However, if we agree with the definition of Marshall et al² that, for an indicator to be valid, it must be “attributable and within the control of the person who is delivering the care”, very few of the quality indicators derived have the health professional/patient interface at their core. Furthermore, many of the indicators could be described as generic in that they apply equally to a patient with diabetes or coronary heart disease—for example, being able to make an appointment to see a GP within 48 hours or being provided with appropriate information about one’s condition.

This leads us back to the difficulty of developing valid indicators specifically for mental health. The evidence base on which to make unequivocal statements in primary care mental health is still somewhat flimsy, but this should not stop us from making a start. The idea of measuring quality in health care is not new. Over 20 years ago Donabedian and coworkers⁵ proposed a multidimensional matrix of structure, process, and outcome to help define and then measure the quality of health care. How can we use this matrix in the field of primary care mental health? A starting point might be the common mental disorders—depression and anxiety. Thus, under “structure” one might decide that depressed patients who perceive the need could book a longer appointment with the doctor; under “process” the primary care team might agree a formulary for antidepressant prescribing; and the “outcomes” might include the number of patients who complete a full course of antidepressants. Each of these constructs can be translated into a quality indicator for which an individual has responsibility.

Mental illness is a major burden for the NHS. The NSF for mental health⁴ is but one of a number of initiatives that aim to improve quality of care. Whether this “top down” approach can drive quality at the interface between health professionals and patients is unclear. The Sainsbury Centre for Mental Health takes the quality agenda somewhat nearer to the service users and suggests a set of more focused standards for primary care organisations, some of which primary care teams should find easier to develop into quality indicators.⁶ The challenge will then be to undertake the necessary research using rigorous methodology, the results of which will allow us to make evidence based changes in the search for further quality improvement.⁷