Narrating uncertainties about treatment of mental health conditions.

Background: Integration of patient views in mental health service planning is in its infancy despite service provision being clearly dominated by narratives from professional consultations and medical records. We wished to clarify perceptions of uncertainty about mental health conditions from a range of provider and user perspectives (patients, carers, parents, mental health service providers) and understand the role of narratives in mental health research.

Aims: (1) To explore the utility of qualitative research methods, particularly narrative content analysis in mental health research, and (2) identify aspects of uncertainty in mental health service users and providers.

Material: Four hundred and six free text responses were considered as one element of an existing questionnaire about uncertainties about mental health treatments, collected from mental healthcare users and providers through charities, the Mental Health Research Network in Wales, health professionals and websites. Free text responses were analysed using narrative content analysis, an elaborate and rigorous research technique that involves groups of analysts working independently and together over extended group sessions.

Findings: Three main themes emerged across respondent groups: "medication and treatment options", "objectification and marginalisation of patient" and "integrity of service delivery". Within these, patients embraced the opportunity to write about their illness at length, whilst carers' and parents' main concerns were about how patients were dealing with their illnesses, the services they were getting and the side effects of treatments. Carers and patients' parents perceived themselves to be the 'go-between', carrying messages between patients and professionals, in order to enable services to function. Mental health service providers and professionals considered uncertainties surrounding medication and treatment from an 'evidence-base' perspective, concentrating on medication choices and the adoption of new approaches to care rather than patient need and expectation. Patients wanted to know what alternatives were available to the drug regimes they were on and felt their opinions were rarely listened to. As a consequence patients felt marginalised by the health systems there to support them and by society as a whole.

Conclusions: Narrative content analysis can help distil large amounts of free text data and enable their successful interpretation. Listening to patients' voices should become an integral part of routine service evaluation and may help bring patient expectation more in line with service organisation and delivery towards an optimal delivery of care.