About this Journal

Article Abstract

Online ISSN: 1099-176X    Print ISSN: 1091-4358
The Journal of Mental Health Policy and Economics
Volume 3, Issue 4, 2000. Pages: 175-186

Published Online: 22 Aug 2001

Copyright © 2001 John Wiley & Sons, Ltd.

Research Article
Using epidemiological data to model efficiency in reducing the burden of depression*
Gavin Andrews, M.D. 1 2 *, Kristy Sanderson 1 2, Justine Corry 1 2, Helen M. Lapsley 1 3
1WHO Collaborating Centre in Epidemiology and Health Policy, St. Vincent's Hospital, Sydney, Australia
2School of Psychiatry, University of New South Wales, Australia
3School of Health Services Management, University of New South Wales, Australia
email: Gavin Andrews (gavina@crufad.unsw.edu.au)

*Correspondence to Gavin Andrews, UNSW Psychiatry at St. Vincent's Hospital, 299 Forbes Street, Darlinghurst, NSW 2010, Australia

*Source of funding: This study was part of a project funded by the Australian National Health and Medical Research Council (No. 113807)

Funded by:
 Australian National Health and Medical Research Council; Grant Number: 113807


The Global Burden of Disease study has suggested that mental disorders are the leading cause of disability burden in the world. This study takes the leading cause of mental disorder burden, depression, and trials an approach for defining the present and optimal efficiency of treatment in an Australian setting.

Aims of the Study:
To examine epidemiological and service use data for depression to trial an approach for modelling (i) the burden that is currently averted from current care, (ii) the burden that is potentially avertable from a hypothetical regime of optimal care, (iii) the efficiency or cost-effectiveness of both current and optimal services for depression and (iv) the potential of current knowledge for reducing burden due to depression, by applying the WHO five-step method for priorities for investment in health research and development.

Effectiveness and efficiency were calculated in disability adjusted life years (DALYs) averted by adjusting the disability weight for people who received efficacious treatment. Data on service use and treatment outcome were obtained from a variety of secondary sources, including the Australian National Survey of Mental Health and Wellbeing, and efficacy of individual treatments from published meta-analyses expressed in effect sizes. Direct costs were estimated from published sources.

Fifty-five percent of people with depression had had some contact with either primary care or specialist services. Effective coverage of depression was low, with only 32% of cases receiving efficacious treatment that could have lessened their severity (averted disability). In contrast, a proposed model of optimal care for the population management of depression provided increased treatment contacts and a better outcome. In terms of efficiency, optimal care dominated current care, with more health gain for less expenditure (28 632 DALYs were averted at a cost of AUD295 million with optimal care, versus 19 297 DALYs averted at a cost of AUD720 million with current care). However, despite the existence of efficacious technologies for treating depression, only 13% of the burden was averted from present active treatment, primarily because of the low effective coverage. Potentially avertable burden is nearly three times this, if effective treatments can be delivered in appropriate amounts to all those who need it.

This paper reports a method to calculate the burden currently averted from cross-sectional survey data, and to calculate the burden likely to be averted from an optimal programme estimated from randomized controlled trial data. The approach taken here makes a number of assumptions: that people are accurate in reporting their service use, that effect sizes are a suitable basis for modelling improvements in disability and that the method used to translate effect sizes to disability weight change is valid. The robustness of these assumptions is discussed. Nonetheless it would appear that while optimal care could do more than present services to reduce the burden of depression, current technologies for treating depression are insufficient.

Implications for Health Care Provision and Use:
There is an urgent need to educate both clinicians (primary and specialist) and the general public in the effective treatments that are available for depression.

Implications for Health Policies:
Over and above implementing treatments of known efficacy, more powerful technologies are needed for the prevention and treatment of depression.

Implications for Further Research:
Modelling burden averted from a variety of secondary sources can introduce bias at many levels. Future research should examine the validity of approaches that model reductions in disability burden. A powerful treatment to relieve depression and prevent relapse is needed. Copyright 2000 John Wiley & Sons, Ltd.

Received: 17 August 2000; Accepted: 27 December 2000