Objective – To assess competing explanations for the universal preference of mothers to cradle infants on their left side and to propose a relation to hemispheric asymmetry for social attachment and communication behaviour. Methods – A review of observational, experimental, physiological, psychological, neuro-physiological, and neuro-psychological studies, including new findings on the cradling behaviour of mothers with auditory or visual impairments. Results – A significant left-cradling bias is observed in both right- and left-handed mothers which cannot adequately be explained by arguments based on handedness or closer contact to the soothing sound of the maternal heartbeat. Observations of primate behaviour have led to the suggestion that the left-cradling bias may be related to a left visual field (right hemisphere) advantage for monitoring an infant's facial expressions of distress. However, more than just monitoring, cradling subserves the mother's connection with the infant. For that reason, we have suggested that left cradling might be related to a right hemisphere specialisation for emotional communication, i.e. the speech melody, smiles, signals, and stroking which mothers use to interact with their infants. Studies of mother-infant interaction show that the sound of the mother's voice is more soothing when cradling on the left, more stimulating when cradling on the right. Cradling laterality may thus be related to emotional state and behavioural intent. There is also evidence to suggest that left cradling is linked to a special role of the right hemisphere in social attachment behaviour. This function may be disturbed in mothers with postnatal depression who show abnormal right hemisphere activity. Conclusion – Cradling embodies the symbiotic relationship between mother and infant; various lines of evidence support the suggestion that the universal preference of mothers to cradle infants on their left side is related to a right hemisphere dominance for social attachment and communication behaviour.

Closing the gap between research and clinical practice is nowadays considered a priority in outcome studies. Survey studies in community settings having as their main aim the multidimensional measure of outcome of mental health care interventions, including the use of standardised instruments administered as part of the routine clinical activities in mental health services, have recently started to be planned in various countries, but have encountered several difficulties. A naturalistic, longitudinal study aimed to assess the outcome of care provided by a community-based mental health service, the South Verona Outcome Project, has been conducted in Italy starting from the beginning of the 90's and is running since then. This paper: a) describes a series of methodological aspects of the South Verona Outcome Project, such as instruments, study design, inclusion and exclusion criteria, training of the staff, and focuses on strategies used so to ensure feasibility of the assessment and good quality of the data; b) summarises some results of the study, characteristics of feed-back provided, and outputs; c) discusses the problems faced, the impact of this approach on service provision, its limitations and future perspectives.

This paper examines two questions. Do staff and service users agree when rating the needs of those with severe mental illness? If not, what is the meaning of these differences? To put these questions in context, we shall define needs, examine one particular way in which needs can be measured in a standardised way, using the Camberwell Assessment of Need, and shall discuss three somewhat similar studies which have compared user and staff rated needs.

This paper is concerned with issues in the routine measurement of quality of life in a mental health context. It is in three parts. In the first part the first author reviews briefly, lessons from a decade of experience in the use of data produced by routine measurement using the Colorado Client Assessment Record (CCAR) in the Mental Health Centre of Boulder County (MHCBC) in Colorado, USA. In the second part, the specific issues surrounding quality of life assessment as a routine outcome measure are considered. Evidence is presented to counter some of the commonly held beliefs about QoL measurement problems. Finally, general problems that affect QoL and all other routine measures are described and analysed using a framework devised by Peterson (1989).

Three issues characterise the background to the MECCA study: A) Throughout Europe, most patients with severe forms of psychotic disorders are cared for in the community. The challenge now is to make processes in community mental health care more effective. B) There are widespread calls to implement regular outcome measurement in routine settings. This, however, is more likely to happen, if it provides a direct benefit to clinicians and patients. C) Whilst user involvement is relatively ?" easy to achieve on a political level, new mechanisms may have to be established to make the views of patients feed into individual treatment decisions. The MECCA study is a cluster randomised controlled trial following the same protocol in community mental health teams in six European countries. In the experimental group, patients' subjective quality of life, treatment satisfaction and wishes for different or additional help are assessed in key worker-patient meetings every two months and intended to inform the therapeutic dialogue and treatment decisions. The trial tests the hypothesis that the intervention – as compared to current best standard practice – will lead to a better outcome in terms of quality of life and other criteria in patients with psychotic disorders over a one year period. This more favourable outcome is assumed to be mediated through different treatment input based on more appropriate joint decisions or a more positive therapeutic relationship in line with a partnership model of care or both. Moreover, the study will hopefully reveal new insights into how therapeutic processes in community mental health care work and how they can be optimised.

Objective – To provide a relatively non-technical review of recent statistical research on the analysis and interpretation of the results of randomised controlled trials in which there are possibly all three of the following types of protocol violation: non-adherence to allocated treatment, contamination (that is, patients receiving treatments other than the one to which they were allocated) and attrition (missing outcome data). Methods – The estimation methods involve the use of potential outcomes (counterfactuals) in the definition of a causal effect of treatment and in drawing valid inferences concerning its size. Results – The methods are explained through the use of simple arithmetical expressions involving the counts from three-way contingency tables (Outcome by Treatment Received by Random Allocation). Illustration is provided through the use of a hypothetical data set. Conclusions – Recent advances in statistical methodology enable one to estimate treatment effects from the results of randomised trials in which the treatment actually received is not necessarily the one to which the patient was allocated. These methods allow one to make adjustments to allow for both non-compliance and loss to follow-up. Even for such a 'broken' randomised trial, inference concerning causal effects is safer than that from data arising from an observational study that never involved random allocation in the first place.