Measurement Of Consumer Outcome In Mental Health - 1994

Download The Entire Report Here

Executive Summary: Measurement Of Consumer Outcome

This report provides a review of the issues, literature, instruments available, and implementation strategies to facilitate the development of consumer outcome measurement for Commonwealth and State funded mental health services.

Preamble

Consumer outcome is defined as the effect on a patient's health status attributable to an intervention by a health professional or health service.
The objectives of the National Mental Health Policy that pertain to consumer outcome are:

1. Objective 30 "To institute regular reviews of outcomes of services provided to persons with serious mental health problems and mental disorders as a central component of mental health service delivery" and;
2. Objective 32 "To encourage the development of national outcome standards for mental health services, and systems for assessing whether services are meeting these standards."

The central role of a health system is defined as the treatment of persons who are ill. The central role of good clinical care in this process is emphasised, and the routine measurement of outcome is seen as an integral part of good clinical care.

The potential consumers are the 25 to 30% of the population who will meet criteria for a mental disorder in any year. The actual consumers, or patients of the health services, are the one third of the potential consumers who actually seek treatment from the medical sector in any year. Three quarters of those who do seek treatment are seen by general practitioners with the remainder being shared between private psychiatrists and the public sector mental health and addictive services. This report is focussed on outcome measurement in consumers with serious mental disorders. Three percent of the population meet criteria for a serious mental disorder in any year. Their principal diagnoses are schizophrenia (10%), mood disorder (44%), anxiety disorder (25%), or substance use disorder (21%). Currently, only half get treatment.

Outcome measurements are of interest to five groups: consumers, carers, clinicians, local health service administrators and State and Federal health service administrators. Privacy considerations are paramount. Administrators should only have access to deidentified or grouped data, carers should only have access to their consumer's data with the consumer's permission, clinicians should have access to data about patients in their care, and consumers to their own records.

The repeated measurement of diagnoses, symptoms, disability and risk factors using standardised measures is regarded as likely to improve treatment as well as providing the information necessary for identifying the outcome of treatment. Measurement, provided it does not detract from treatment time, does not interfere with therapy. Information on outcome will be of considerable value to health planners and administrators at state and national levels.
The presumption that changes in outcome can be automatically attributed to a treatment can be considered to be true for routine clinical practice, but this presumption would require considerable supporting data if changes in outcome were to be considered evidence of the efficacy of a treatment or the efficacy of a clinic or service.

Outcome data will be the first clinical data routinely available to administrators. First, because aggregated disability data will establish the load on the service and second, because the changes in these disability measures will provide an outcome performance indicator for individual clinics and clinicians. If such a system is going to comply with privacy requirements then only deidentified or aggregated data should be sent from a patient's clinical record to administration or head office.

Population outcome indicators, while measures of health, are not favoured because they would not allow the cause of any change to be attributed. A multidimensional measure of individual outcome could serve to guide clinical practice and, when aggregated with measures from other consumers, serve to guide the delivery of services.

Consumer outcome measures are reviewed in regard to their applicability, acceptability, practicality, reliability, validity and sensitivity to change. The advantages and disadvantages of questionnaires and interviews, completed by clinicians or consumers are delineated. Measures which are suitable for use in routine clinical practice are likely to be brief, low cost, multidimensional measures which require minimal training for their administration, scoring and interpretation. Other more complex measures are reviewed only if they are likely to be of value for collateral studies on outcome.

Methodology

1. Fifty three per cent of 377 providers, non-government organisations and carer and consumer groups responded to a survey, either by completing the questionnaire or by offering views. All persons known to be expert in the measurement of disability and outcome, both in Australia and overseas, were contacted for advice. Peak Bodies were consulted and their opinions given special weight.
2. A keyword search on Medline returned 350 references which were reviewed for possible outcome measures. Outcome measures were also identified from speciality texts, citation lists and other sources. Ninety five measures were reviewed by two independent raters in relation to their function and psychometric properties.

Results Of The Consultation Process

1. The replies to the survey show that disability and quality of life is viewed as more important areas of outcome than either symptoms or satisfaction with services. Consulting the consumer, either by questionnaire or interview deemed better than either clinician opinion or carer rating. Respondents thought that the principal use of such data is to improve service efficacy.
2. The professional bodies representing psychiatrists, mental health nurses, clinical psychologists and occupational therapists recognise the importance of the measurement of consumer outcome, but express reservations as to how it might be implemented.
3. The views of the participants at the National Workshop on Consumer Outcome in Mental Health were diverse, but there is consensus about the need to measure outcome, especially with a reliable multidimensional measure.

Results: Review Of Existing Generic Measures

1. The reviews of symptom measures are restricted to those applicable to broad groups of patients. Eighteen such measures are reviewed.
2. Level of functioning is an important target to measure and 20 measures of functioning are reviewed, with one, the Role Functioning Scale being recommended.
3. Seventeen quality of life measures are reviewed but none meet the criteria for a consumer outcome measure suitable for routine use.
4. Five measures of family burden are reviewed. None are recommended for routine use but some would be valuable in studies focussed on family burden.
5. Eight measures of satisfaction with services are reviewed but all are too narrow in focus to serve as the sole consumer outcome measure.
6. Twenty-seven multidimensional measures met criteria for review. Five met minimum criteria but one, the Life Skills Profile, is not recommended on the grounds of cost. Four are recommended for use as routine measures of consumer outcome. These four are the Behaviour and Symptom Identification Scale (BASIS), Health of the Nation Outcome Scales (HoNOS), Medical Outcomes Study Short Form (SF36), and the Mental Health Inventory (MHI).
7. [NB. The K10 is now recommended as being superior to the MHI, the SF12 as superior to the SF36]

Recommendations

1. Field test all five measures and explore the level to which they are applicable, acceptable, practical, reliable, valid and sensitive to change in the appropriate consumer groups and service settings.
2. Establish pilot sites in each state to implement the measurement of consumer outcomes.
3. Develop training packages for consumers, clinicians and administrators as to the meaning of the information being produced by the measures of consumer outcome.
4. Establish a mechanism to coordinate the above activities and facilitate the nationwide implementation of routine consumer outcome measurement.

Edited by Gavin Andrews MD, UNSW, 2007
©2007 CRUfAD