Existing classification systems fail to capture the complexity of mental disorders as manifest in primary and general health care settings. They do not adequately address the problems of co-morbidity, sub-threshold disorders, cross-cultural applications or social dimensions, or acknowledge the difference between severity and impairment.

An effective classification system for primary and general health care needs to pay attention to four key elements: diagnosis, severity, chronicity and disability.

1. • Categorical diagnoses should be more stringent and precise, so that they become rarer but more significant events in primary and general health care.

2. • Measurement of severity can be achieved by combining categorical and dimensional approaches.

3. • Disability is important in its own right, as many people with current sub-threshold disorders have significant levels of impairment.

We also need to consider classification of social problems, to enable dialogue with social care.

This approach is likely to reduce unnecessary medicalisation, and enable better focus on those most in need of care. The addition of measures of severity, chronicity and impairment will encourage better targeting of interventions.

For successful adoption within primary and general health care, a classification system needs to be simple, grounded in research and reality, adaptable for specific populations and countries, useful as a teaching tool and accessible for routine data collection. CD10-PHC and WONCA's International Classification in Primary Care provide good models to build upon.

These issues are currently being considered by the WONCA Working Party on Mental Health and the APIRE Primary Care Conference Expert Group.