From consultation to integrated health risk assessment

Don Lipsitt links the heyday of consultation- liaison psychiatry to the positive collaboration that had occurred with medical colleagues during the World War II. By then consultation-liaison psychiatry was strongly supported through National Institute of Mental Health (NIMH) programs, which later also introduced social workers and psychologists into the general hospital. This devalued somewhat the achieved professional authority of consultationliaison psychiatrists, who were mainly psychoanalysts. The uprise of social sciences led to the development of the biopsychosocial model, which became the conceptual model for consultationliaison psychiatrists. It was particularly valuable for liaison models on at-risk wards. Liaison psychiatry assumes a preventive integrated collaboration between psychiatric consultants and medical teams serving populations with a high prevalence of psychiatric disorders and distress (1). The DSM system brought psychiatry sound epidemiology, announced the decline of psychoanalysis and brought psychiatry back into the medical model. Epidemiological studies found significant discrepancies between prevalence and service delivery for patients with psychiatric disorders in the general hospital setting (2). These were the arguments for the liaison model. However, its operationalisation was lacking (3). This led to the complex paradox of being both part of the scientific medical world with an empirical diagnostic system and embracing a non-operationalised conceptual model regarded as vague (4). It drifted consultation-liaison psychiatrists into the arms of managed care, which reduced its intellectual knowledge base to 'diagnosable disease', preventing them from fulfilling an intellectual leadership role for patients at risk for psychiatric comorbidity and psychosocial distress seen in medical wards (5).

Recently we reported an empirically derived method for assessment of complexity of care, operationalised in terms of health risks and needs and based on the biopsychosocial model. From the reduced role of consultation-liaison psychiatrists as consultants taking part in a fragmented spectrum of health care provision in the medical setting, this method offers a model for a move to a role in indicator-generated integrated care for complex patients, i.e. modern liaison (6,7). The INTERMED consists of 20 variables (Table 1), rated 0-3, reflecting the risk factors with respect to somatic, psychological and social functioning, and the patients' relation with the health care system. The total score, ranging 0-60, is used as a measure of complexity of care.

Table 1.

The INTERMED

	History	Current state	Prognosis
Biological	Chronicity	Severity of symptoms	Complications and life threat
	Diagnostic dilemma	Diagnostic challenge
Psychological	Restrictions in coping	Resistance to treatment	Mental health threat
	Psychiatric dysfunctioning	Psychiatric symptoms
Social	Restrictions in integration	Residential instability	Social vulnerability
	Social dysfunctioning	Restrictions of network
Health care	Intensity of treatment	Organisation of care	Coordination
	Treatment experience	Appropriateness of referral

The reliability and validity of the INTERMED have been reported elsewhere. In a confirmative factor analysis of 1100 cases, the following underlying factors were found to contribute to the complexity of a patient: physical chronicity, psychological vulnerability, social disruption, dependency (being able to take care for oneself), diagnostic complexity and compliance (8). The interview can be conducted by a trained nurse in about 15 minutes. The scoring, which is visualized, takes another 5 minutes. Currently a semi-automatic computer generated letter describing the risks and needs in a structured way is tested. It will allow a nurse clinician to have a comprehensive report in another 10 to 15 minutes, which will support, when needed, the initiation of coordinated care. The instrument has both psychometric and clinimetric properties and acts both as traditional assessments do as well as a means of coordinated communication amongst health care providers. Therefore it allows a protocolized assessment, being highly supportive for clinical decision making by higher trained nurses who are supervised by a psychiatrist. Thereby an empirical coordination of care can be introduced in the general health sector comparable with case management in the mental health sector.

This, in combination with the new models for integrated care including coordination of care by specialist nurses (9,10) and the recognition of the discipline through subspecialisation status (11), might contribute substantially to the development of formalised integrated care for patients at risk for complexity of care seen in the general hospital setting.