Shabeer Syed - Thesis submitted in partial fulfilment of the Degree of Doctor of Clinical Psychology (DClinPsych)

Abstracts Integrated Thesis (includes three research items): Critical review of the literature (CRL) Trauma-focused cognitive behavioural therapies for posttraumatic stress disorder among adults with multiple-incident traumas compared with single-incident traumas: a systematic review and meta-analysis Objective: Trauma-focused cognitive behavioural therapies (TF-CBT) are considered the first-line treatment for posttraumatic stress disorder (PTSD) following different traumatic experiences. Whilst there is compelling evidence for TF-CBT for PTSD following single-incident traumas, the effectiveness of TF-CBT for PTSD following multiple-incident traumas is less clear. We conducted a systematic review and meta-analysis to evaluate the pooled effects of TF-CBT for multiple-incident PTSD relative to the effects of TF-CBT for single-incident PTSD. Methods: We searched 18 electronic databases between Jan 1990 and Mar 2022 for RCTs comparing any TF-CBT for PTSD among adults exposed to multiple-incident traumas (defined as intimate partner violence; IPV; child maltreatment; CM; systemic violence and combat exposures) or single-incident traumas, relative to treatment as usual (TAU). The primary outcome was PTSD symptoms as per a validated self-report or clinician administered PTSD instrument immediately post-intervention. We used random-effects meta-analyses to pool standardised mean differences (Hedges’ g) of TF-CBT against TAU stratified by trauma types. Results: We included 144 RCTs (11,949 individuals), involving three types of single-incident traumas and four types of multiple-incident traumas, including systemic violence and refugees, CM or IPV and military combat-related exposures. We found no significant differences in the pooled effect of TF-CBT against TAU in RCTs involving multiple-incident traumas (Hedges’ g, -0.90, 95% CI -1.07, -0.73; 65 RCTs) compared with the pooled effect from RCTs involving single-incident traumas (Hedges’ g, -0.70, 95% CI -0.85, -0.54; 67 RCTs). Estimates remained similar when adjusting for baseline severity of PTSD symptoms, comparing specific traumas, and limiting estimates to active comparator groups. TF-CBT produced smaller treatment effects in RCTs involving multiple-incident traumas with comorbid serious mental illnesses (e.g., psychosis and personality disorders; Hedges’ g, -0.37 95% CI -0.57, -0.17) relative to RCTs with less comorbidity. Conclusions: TF-CBT showed comparable efficacy for reducing PTSD symptoms in RCTs involving repeated and sustained traumatic events as in RCTs of PTSD from single-incident traumas. However, treatment approaches varied substantially between groups of trauma types and estimates should be considered with caution. Further research is needed to evaluate the effectiveness of TF-CBT for multiple-incident PTSD related to IPV and populations across inpatient and forensic settings. Service improvement project (SIP) Predicting length of stays, delayed discharges, and readmissions in adult psychiatric inpatient units in one NHS trust: a mixed-methods study Background: Despite repeated NHS initiatives to reduce the length of stays (LOS) of adult acute psychiatric admissions in England, the factors influencing LOS and delayed discharges are poorly understood. Aims: To identify and estimate key factors that may influence LOS and delayed discharges among adult psychiatric admissions in one NHS trust using a mixed-method design of staff interviews and risk prediction models. Methods: In phase one, we thematically analysed data from 17 semi-structured interviews with inpatient staff, focusing on barriers and facilitators influencing LOS. In phase two, we analysed data on 5351 acute admissions (aged 18-65 years old) from 7 adult psychiatric inpatient units in one NHS trust covering two counties (2015-2021) using adjusted multivariable regression models. The primary outcome was locally defined delayed discharges (LOS >56 days). Secondary outcomes included adjusted mean differences in LOS and readmissions within one year of discharge. Results: In phase one, consistently identified barriers included a need for inpatient social workers, increased recognition of discharge barriers with shared responsibility between professions to overcome barriers, difficulties in forming consensus within clinical teams about discharge criteria for patients with complex needs, and more effective information sharing systems with community services and placement providers. Consistently reported facilitators for reducing LOS included improved gatekeeping criteria and goals upon admission with clearly assigned staff roles, and improved communication with services outside the hospital. In phase two, the adjusted mean LOS of adult acute admissions between 2016 and 2021 was 58.4 days (95 % CI 50.9 days-65.5 days), or 54.2 days (95% CI 44.5 days-63.9 days) excluding the psychiatric intensive care unit. Of these, 27.4% (1476/5351; 95% CI 24.4%-30.4%) had a LOS above 56 days and the national average. The strongest predictors of delayed discharges and one-year readmissions were positive COVID-19 tests, older age, black ethnicity, homelessness, a diagnosis of bipolar or psychosis (affective manic type), and patients with risk concerns of self-neglect, non-compliance with care and aggression. Faster allocation of care coordinators upon admission and admission to a ward with lower bed capacity (e.g. higher bed occupancy) were associated with significantly lower LOS and reduced delays. Interpretation: In this study of one NHS Trust, only one in four adult psychiatric admissions had a LOS above the national average of 51 days for adult mental health admissions. This performance could likely be further improved by early identification of homeless patients, faster allocation of care coordinators, and increased support to make risk positive discharge decisions of patients with complex needs. Interactive risk prediction models, key findings and trust specific recommendations are provided at the end (see executive summary) and available online (https://ohft.github.io/lengthofstay/). Theoretically driven research project (TDRP) Early adverse childhood experiences, emotion dysregulation and subsequent risk of chronic pain in children and adolescents: a population-based birth cohort study Objectives: Children with adverse childhood experiences (ACEs) face an increased risk of developing chronic pain (CP), yet longitudinal evidence on the factors underlying this relationship remains limited. We assessed the association between early adverse childhood experiences (ACEs) and subsequent risk of CP, and whether this relationship was mediated by emotion regulation difficulties (ERDs). Design and participants: Population-based birth cohort of 179,207 mothers and children with linked electronic health records (1998-2018) from the Clinical Practice Research Datalink GOLD (primary care), Hospital Episode Statistics (secondary care), and the Office for National Statistics mortality register. We followed the cohort from 1-year before birth (mothers only) up to 16-years post-birth. Primary outcome: First recorded CP in children between 7-years and 16-years post-birth, identified by an algorithm capturing recorded diagnoses of CP and continuous pain episodes. Main exposures: ACEs and ERDs in children. ACEs were defined as child maltreatment, maternal intimate partner violence, maternal mental health problems, and adverse family environments recorded one year before birth (maternal records) up to three years post-birth. ERDs were defined as emotional and behavioural presentations in children between four and six complete years post-birth. We used Cox-proportional models to test the hypotheses that ACEs increase the risk of CP, and that this risk was mediated by ERDs and common non-communicable child diseases (NCDs). NCDs were hypothesised intermediate confounders and analysed as co-joint mediators with ERDs. Results: During a median follow-up of 11-years, children with any ACEs or ERDs had almost a two-fold higher risk of CP between 7 and 16-years old relative to children without ACEs or ERDs. We found a consistent dose-response relationship between the number of ACEs, the number of NCDs, ERDs and increased risk of CP. ERDs in children between 4- and 6-years post-birth mediated 10% of the association between ACEs and CP (mediated proportion range: 9%-11%; p Conclusions: ACEs and ERDs are associated with an increased risk of CP among children presenting to primary care during early child development. ERDs may play an important causal role in the development of CP. Further longitudinal research should evaluate specific ERD-related mechanisms for early intervention in children with recurring pain.