Fiona Warren

Towards evidence-based practice in medical training: making evaluations more meaningful

Context  The evaluation of training is problematic and the evidence base inconclusive. This situation may arise for 2 main reasons: training is not understood as a complex intervention and, related to this, the evaluation methods applied are often overly simplistic.

Postnatal debriefing: have we thrown the baby out with the bath water?

Postnatal debriefing is offered by 78% of maternity services in the UK despite little evidence from randomised controlled trials (RCTs) that it is effective. RCTs in this area have applied debriefing as a prophylactic to all or high risk women, rather than as a treatment for women who request it. This pragmatic trial therefore evaluated existing postnatal debriefing services that provide debriefing as a treatment for women who request it. Forty-six women who met criterion A for posttraumatic stress disorder (PTSD) and requested debriefing 1.3-72.2 months (median 16 weeks) postpartum completed measures of depression, PTSD, support and negative appraisals of the birth before and one month after debriefing. Women were compared with others who gave birth in the same hospitals during the same time period (n=34), who met criterion A for PTSD but had not requested debriefing. Results showed PTSD symptoms reduced over time in both groups but greater decreases were observed in women who attended debriefing. Debriefing also led to reduction in negative appraisals but did not affect symptoms of depression. Therefore, results suggest providing debriefing as a treatment to women who request or are referred to it may help to reduce symptoms of PTSD.

The properties of self‐report research measures: Beyond psychometrics

Self-report measures pertinent for personality disorder are widely used and many are available. Their relative merits are usually assessed on nomothetic psychometrics and acceptability to users is neglected. We report reactions of lay, patient and professional groups to the Personality Diagnostic Questionnaire (PDQ-IV); Millon Clinical Multiaxial Inventory (MCMI-III); the Borderline Syndrome Index (BSI); Rosenbergs Self-Esteem Scale (RSE) and the Social Functioning Questionnaire (SFQ). These were sent to 148 professionals, ex-patients and lay people for comment. Thirty-six per cent were returned. Pattern-coding by three raters revealed problematic themes across all measures, including inappropriate length, vague items and language, cultural assumptions and slang, state-bias and response-set. Measures can be depressing and upsetting for some participants (both patients and non-patients), hence administration of measures should be sensitive. Treatment may make people more self-aware, which may compromise validity for outcome research. This evaluation raises issues and concerns, which are missed in traditional psychometric evaluation.

What do the experts by experience tell us about 'what works' in high secure forensic inpatient hospital services?

The evidence base for ‘what works’ for patients detained in high secure hospitals has predominantly been established from a clinical perspective, with the voices of those at the centre of care, the expert by experience, absent. Neglecting this voice renders an important source of information for evidence-based practice inaccessible to outcome evaluators. Twelve high secure patients considered ‘ready’ to be discharged were interviewed to explore what in their view had helped or hindered their progress to this stage. Thematic analysis steps were applied to interviews and eight themes were generated that represented valued elements of high security: temporary suspension of responsibility, collaboration in care, learning from others, supportive alliances, specific interventions (medical and psychotherapeutic), a safe environment and opportunities for work. Narratives demonstrated the complementary and unique contribution of the patient experience in informing ‘what works’, and are discussed alongside existing theories relevant to promoting clinical change and risk reduction.

A Critical Analysis of Clinical Evidence from High Secure Forensic Inpatient Services

High-secure forensic inpatient hospitals serve important functions in the detention and rehabilitation of people with serious mental health needs who present as a risk to society. Establishing whether services are effective in restoring mental health and reducing risk is an important task, but one that has not been systematically undertaken. A systematic review of outcome evidence from this setting was therefore conducted to establish its scope and quality. Evidence was found from 22 studies for nine different interventions targeting multiple and overlapping outcomes, which included recidivism, mental health, aggression, social functioning and quality of life. Studies were commonly assessed as being at a potentially high risk of bias from validity threats. The ‘best’ available evidence was for medication, psychoeducation and third-wave cognitive-behavioral interventions. Clinical implications are tentatively offered given the narrow focus of the review on high security. Directions for the high-secure services research agenda are more definitively proposed.

Cognitive behaviour therapy training in a developing country: a pilot study in Tanzania

Cognitive behaviour therapy (CBT) has been found to be effective in treating mental health problems in the UK, but little has been done to evaluate the potential of CBT in developing countries. This paper aims to discuss the development and implementation of a CBT training course for clinicians working in Tanzanias main psychiatric hospital in the capital city, Dodoma. A 12-session training course in CBT was delivered to nine clinicians. An outcome evaluation was conducted using multiple measures and methods, taken before and after the training. Information on cultural adaptations of the training was obtained. All participants completed the course, but there were several obstacles to full completion of the evaluation measures. Despite this, there were significant improvements in clinicians’ basic understanding of CBT concepts, and their ability to apply the CBT model to formulate and recommend treatment strategies in response to a clinical case. Qualitative information indicated the potential of developing CBT training and implementation further. As a pilot study, this investigation shows the promise that CBT holds for mental health services in Tanzania. Further research into the training and clinical effectiveness of CBT in Tanzania is indicated.

Bloodletting, bulimia nervosa and borderline personality disorder

Some recent case-reports in the psychiatric literature have presented patients who bloodlet and suggested that theirs is a rare pathological behaviour which is part of the purgative psychopathology associated with bulimia nervosa. Other reports have construed bloodletting as a self-mutilative behaviour and indicative of severe personality disorder. We review the literature to ascertain whether it supports either of the above claims or whether there are patients who conform to both or neither stereotype. We present two further cases, which highlight the importance of (1) eliciting Axis I and Axis II psychopathology in the assessment of all patients and (2) eliciting the patient’s view of the motivation and meaning of the behaviour. There is strong evidence to suggest that bloodletting is not solely a feature of eating disorders and that it may be more generally related to personality disorder, where, even in the absence of Axis-I pathology, it may be an indicator of severity of psychopathology. * c 1998 John Wiley

Essential elements of treatment and care in high secure forensic inpatient services: an expert consensus study

Purpose – The evidence base for what works with forensic patients in high-security inpatient settings has typically focused on outcome research and not included clinical expertise from practice-based experience, which is an important facet of evidence-based practice. The purpose of this paper is to establish whether experts with clinical and/or research experience in this setting could reach consensus on elements of high-security hospital services that would be essential to the rehabilitation of forensic patients. Design/methodology/approach – A three-round Delphi survey was conducted to achieve this aim. Experts were invited to rate agreement with elements of practice and interventions derived from existing research evidence and patient perspectives on what worked. Experts were also invited to propose elements of hospital treatment based on their individual knowledge and experience. Findings – In the first round 54 experts reached consensus on 27 (out of 39) elements that included physical (e.g. use of CCTV), procedural (e.g. managing restricted items) and relational practices (e.g. promoting therapeutic alliances), and to a lesser extent-specific medical, psychological and social interventions. In total, 16 additional elements were also proposed by experts. In round 2 experts (n=45) were unable to reach a consensus on how essential each of the described practices were. In round 3 (n=35), where group consensus feedback from round 2 was provided, consensus was still not reached. Research limitations/implications – Patient case complexity, interventions with overlapping outcomes and a chequered evidence base history for this population are offered as explanations for this finding alongside limitations with the Delphi method. Practical implications – Based on the consensus for essential elements derived from research evidence and patient experience, high-secure hospital services might consider those practices and interventions that experts agreed were therapeutic options for reducing risk of offending, improving interpersonal skills and therapeutic interactions with patients, and mental health restoration. Originality/value – The study triangulates what works research evidence from this type of forensic setting and is the first to use a Delphi survey in an attempt to collate this information.