The Humanitarian Health Evidence Review

INTERVENTION APPROACHES

Scaling up: There is a need for more evidence on the effectiveness of low intensity and low cost interventions in non-specialised health care and community settings. Research on the following was suggested:

• The effectiveness and feasibility of delivering care through different cadres of health workers, including with intermediate training (e.g. BSc level or less) and community health workers. This should include potential risks and trade-offs.
• Evidence on how services can be most effectively integrated through existing health services/systems, particularly at the primary health care level.
• Effectiveness of different training methods, particularly how much supervision is needed to achieve effective task shifting.
• Effectiveness and feasibility of interventions for groups, in addition to individuals.
• Effectiveness and feasibility of e-mental health interventions.
• Effectiveness and feasibility of interventions addressing issues underlying multiple disorders, rather than specialist interventions on a single disorder (e.g. PTSD). Such an approach could be maximised by applying components from different treatments methods to address co-morbidity (i.e. component- or module-based therapy, see below).

User and community-orientated services: Mental health users, family members and communities can be involved more in preventive and treatment interventions. Research on the effectiveness of the following interventions was suggested:

• Evidence for interventions aimed at strengthening participation of affected communities in humanitarian settings, a key principle in current guidelines.
• Where appropriate, using parents and natural support systems (rather than external counsellors), including parent management training.
• Strengthening social support and coping mechanisms at family and community levels.
• School-based interventions for improving pupil mental health outcomes, including looking at the the burden on teachers (if teachers are involved).

Interventions with other sectors: There is a need for more evidence on how mental health and psychosocial support interventions can be better integrated with other sectors. Examples include evidence on the effectiveness of mental health and psychosocial support interventions with the following sectors:

• Research on preventive interventions that address major determinants of mental health in humanitarian settings, including interventions targeting ongoing violence (particularly against women), poverty, and social exclusion.
• The education sector (e.g. school-based interventions to improve pupil mental health outcomes).
• The nutrition sector (e.g. mental health interventions with mothers to also improve maternal and infant nutritional outcomes).
• Protection and welfare (e.g. to prevent punitive parental violence against children to improve child mental health and behaviour outcomes; e.g. to improve parental mental health to reduce parental violence against children; e.g. improve mental health outcomes among survivors of sexual violence).
• Communicable disease (e.g. how improved mental health (e.g. reducing harmful alcohol use) may reduce risky behaviour and communicable disease transmission).

TYPES OF INTERVENTION TO BE RESEARCHED

The following two complementary approaches were highlighted.

• Psychosocial interventions: The lack of evidence on the effectiveness of psychosocial (preventive) interventions in particular was frequently reported (although the additional complexities of evaluating such interventions were recognised). Cited examples include stronger evidence needed on psychological interventions such as psychological first aid, generic counselling, psycho-education, social interventions such as addressing violence and social exclusion, and childhood interventions.
• Modular transdiagnostic approaches for psychological interventions: For people with mental disorders, it was suggested that rather than providing a single treatment for a specific disorder (e.g. those with a more proven evidence base such as CBT and IPT), it could be beneficial to apply simplified individual modules from within these different treatments in order to simultaneously respond to a range of symptoms and disorders (rather than just a single disorder). Evidence would therefore be required on the effectiveness of this modular approach.

MENTAL HEALTH AND PSYCHOSOCIAL SUPPORT OUTCOMES TO BE RESEARCHED

• Multiple outcomes: It was widely recognised that evidence on the effectiveness of interventions treating a range of different outcomes (e.g. PTSD, depression, anxiety or alcohol disorder) was required given the variety of disorders commonly observed with crisis-affected populations, the high levels of co-morbidity, and the greater cost-effectiveness of treating multiple disorders.
• Severe disorders: There remains a clear need to strengthen the evidence-base of common mental disorders, but in addition there is a need for a much stronger evidence-base on treating severe mental disorders (e.g. psychosis, schizophrenia, severe depression) given their increased burden and that they commonly take up the majority of clinic load (albeit context specific). It was noted that while the efficacy of some interventions for severe disorders is well proven from more stable settings (e.g. use of specific drugs), the broader effectiveness of such interventions in emergency settings has not been proven. In addition, the use of supplementary psychosocial interventions for severe disorders has not been adequately tested or proven (e.g. the use of anti-stigma campaigns in communities, community-based rehabilitation and inclusion approaches).
• Functioning: The need for more measurement of functioning as an outcome for mental health interventions was frequently noted. The additional need to ensure that functioning measures were culturally appropriate was also raised.
• Substance misuse: The need for further research on interventions addressing harmful alcohol use and drug taking was frequently raised, including brief interventions.
• Determinants of mental health: given that humanitarian settings are often chronic, it is key that interventions also address the ongoing determinants of mental health (e.g. violence against women and children, socioeconomic adversity, social exclusion). These interventions are popular in practice, but have not been rigorously evaluated.

STUDY DESIGNS TO BE USED

• Randomised control trials (RCTs): Despite the obvious challenges of conducting RCTs in humanitarian settings, the need for RCTs was widely recognised, but also that RCTs (and other study designs such as quasi-experimental designs with comparison groups) should collect data over a much longer period than has previously been the case in order to be able to track longer-term intervention effects (where ethically appropriate).
• Mixed-methods studies: The need for combined quantitative and qualitative studies was commonly reported. For example, to better understand local explanations for mental health disorders and causes, issues of access, the cultural acceptability and appropriateness of interventions and their implementation, and study measures.
• Greater use of routine data: Greater analysis and publication of high quality routine facility-based data on mental disorder outcomes over time, including the use of clinical audit data and also case-study approaches to observe treatment effects and review service delivery models. This includes incorporating mental health in routine surveillance systems as soon as possible in humanitarian crises.
• Feasibility studies: The need for greater evidence on feasibility of interventions was frequently raised, particularly in comparing different interventions (e.g. by being linked to an RCT) and in scaling-up interventions and for task-shifting interventions. Aspects of feasibility should include economic (see below), social and cultural (e.g. how acceptable), political, technical and operational.
• Economic studies: There was a widely expressed need for more studies on economic aspects of interventions, in particular for cost-effectiveness analysis to be included in intervention studies. Other research related to the need to understand the overall economic costs and benefits in interventions involving scaling-up services through the health system. Gaining a better understanding of the economic costs of poor mental health (and subsequent potential gains through improved mental health) was also highlighted.

PARTICULAR STUDY POPULATIONS OF INTEREST

• Children and adolescents: The need for evidence on interventions for child mental health, behavioural problems and development was raised, including the use of brief family interventions, peer education, and appropriate community mechanism and resources.
• Older populations: Particular evidence on interventions addressing dementia and old age problems.
• Survivors of sexual and other forms of gender-based violence: For example, evidence on the effectiveness of mental health interventions for survivors of sexual and gender-based violence (e.g. intimate partner violence); evidence on interventions to reduce anger and violence among men

KEY ETHICAL ISSUES

• Adverse effects: More evidence is required on the adverse effects of mental health interventions. For example, culturally insensitive interventions; lack of sustainability; poor or (unintentionally) abusive practice due to limited training, capacity, monitoring and supervision.
• Independent RCTs: There is a need for more independently led trials, rather than being led by proponents of the particular mental health intervention being trailed.
• Quality: There is a need to improve the quality of research in order to ensure its appropriateness, and the accuracy, validity and reliability of results of the interventions.