INTernational REsearch Programme on Psychosis In Diverse Settings — INTREPID II  

Psychotic Disorders Clinical Trial

Official title: Protocol for a Multi-study Programme of Research on Untreated Psychosis in India, Nigeria, and Trinidad

Status Recruiting

Clinical Trial Summary

INTREPID II aims to investigate variability in incidence, presentation, outcome, and impact of untreated psychotic disorders in three countries - India, Nigeria, and Trinidad - through four interconnected observational studies:

1. Study 1 on Incidence, Presentation, and Risk has the objective to investigate the incidence and presentation of untreated psychotic disorders in each setting and associated risk factors.

2. Study 2 on Course and Outcome has the objective to investigate two-year course and outcome of psychotic disorders and associated factors.

3. Study 3 on Help-seeking and Impact has the objective to investigate (a) help-seeking; and (b) the impact of psychotic disorders on individuals and families, using a combination of quantitative and qualitative approaches.

4. Study 4 on Physical Health has the objective to investigate the types and prevalence of physical health problems and related biological markers.

Clinical Trial Description

INTREPID II comprises four interconnected studies. As a basis for these studies, the researchers are identifying, assessing, and following, in each catchment area, population-based cohorts of cases (individuals with an untreated psychotic disorder) and controls (individuals with no history of a psychotic disorder). In each setting, the researchers will identify, assess, and follow at 2 years cohorts of 240 cases with a psychotic disorder (total, 720) and 240 matched controls (total, 720), using methods and infrastructure developed during a feasibility and pilot study, INTREPID I. The inclusion and exclusion criteria for cases are in line with those used in previous studies, including the World Health Organization (WHO) multi-country studies, and are purposefully broad to capture heterogeneity and to allow sub-analyses by duration of untreated psychosis.

The study sample includes cases, controls and relatives or caregivers:

Sample (1) Cases

To estimate incidence, the researchers aim to identify all individuals with an untreated psychotic disorder (cases) within each catchment area. Untreated is defined as never having received treatment with anti-psychotic medication for one continuous month prior to the start of the case-finding period.

In each catchment area, the researchers are using a multi-pronged approach to case identification. First, using procedures developed in INTREPID I, the researchers have established comprehensive case detection systems by mapping and seeking to engage a comprehensive set of service providers and community key informants who may encounter individuals with psychotic disorders within the catchment area. This includes the professional sector (specialist and generalist services; public, private and third sector), the folk sector (including traditional and religious services), and the popular sector (i.e. informal sources of support). Second, the researchers give providers and informants materials developed in the INTREPID I pilot work that detail, using local terms and language, the experiences and behaviours that characterise psychosis. Third, in each catchment area, researchers check with each provider and informant regularly and conduct regular checks of admissions ledgers and registers for in-patient and out-patient services (where these exist), to identify potential cases. In addition, in rural villages in Chennai and Ibadan, field workers visit village meeting points to enquire about potential cases. Potential cases are then screened for inclusion using the Screening Schedule for Psychosis, an instrument that has been widely used in epidemiological studies of psychoses. Those who screen positive and who meet inclusion criteria are approached and informed consent sought.

Case-finding began on 1 May 2018 and will conclude 30 April 2020. At the end of the case-finding period, the researchers will conduct leakage studies in each setting to identify possible cases meeting the study's inclusion criteria who may not have been identified. Each research team will systematically re-check admissions ledgers and registers for in-patient and out-patient services and complete final checks with healers and key informants.

All eligible cases identified through the incidence study are invited to participate in the programme. Rates of refusal are documented and basic data (i.e. age, gender, area of residence, sector of identification, and where possible ethnicity, religion, duration of untreated psychosis and mode of onset) is collected for those who decline to participate, or who it is not possible to interview, to assess non-response bias.

Sample (2) Controls

Age-, sex- and neighbourhood-matched controls are recruited to provide indicative population data against which to compare cases in terms of hypothesised risk factors, social outcomes, and physical health. The researchers use the Psychosis Screening Questionnaire to collect information on any current or past experiences of psychosis. In the absence of a readily accessible sampling frame to randomly select potential controls, the researchers map the ten nearest neighbouring households for each case, listing all residents in these dwellings by sex and age. All potential controls for the case (defined as the same gender and ±5 years of age) are then approached in random order, until an eligible control is identified. When no match is identified the process is repeated. This approach was successfully piloted in all settings.

Sample (3) Relatives and Caregivers

The researchers seek consent from each case to approach a close relative or caregiver to participate in the study. They then approach each designated relative to seek his/her consent. The primary purposes of including relatives are to corroborate and extend information from cases (e.g., physical health and illness), to collect information on premorbid adjustment, family history of mental disorder, and other risk factors, and to collect information on family responses to psychosis, help-seeking, and impact (burden) on family.

Follow-up

All participants will be followed at 2 years. To facilitate this, the researchers collect detailed contact information at baseline (address, telephone number, email address if applicable, service provider details) from each case and control, including details of a relative or friend who can be contacted to trace the individual. In addition, to maintain contact and minimise attrition, the researchers contact participants every six months, by telephone or in person, to confirm or update contact details. Based on the INTREPID I pilot work, the researchers expect to re-assess around 80% of cases and controls 2 years after initial identification.

Sample size

In each setting, the researchers anticipate (based on pilot findings) identifying around 300 untreated cases. Of those, given an expected refusal rate of 20% of all eligible cases (based on the INTREPID I pilot work), the researchers anticipate recruiting approximately 240 cases with a psychotic disorder (total, 720), and 240 individually matched controls (total, 720). These sample sizes are larger than most previous studies and provide good statistical power to test the study hypotheses (i.e., > 80% at p 0.05). For example: (1) with samples of around 300 incident cases in each setting, the researchers will have over 80% power to detect an incidence rate ratio of 1.5 (or greater) between two areas (e.g., urban vs. rural), if the incidence rate in the lowest risk area is 20 per 100,000; (2) with a sample of 240 cases and 240 controls in each setting, the researchers will have over 80% power to detect an odds ratio of 2.0 (or greater) in case-control comparisons when the prevalence of exposure (risk factor) is at least 15% in controls; (3) using gender as an example, with a sample of 192 cases followed at 2 years in each setting, the researchers will have 80% power (or greater) to detect a difference in the proportion of cases with a poor outcome (e.g., continuously psychotic) of 0.20 (20%) or greater, when the proportion of men with a poor outcome is 0.40 and the proportion of women is 0.20 (i.e., equivalent to an odds ratio of ~ 2.5).

Data collection

To test the hypotheses and address the research questions of INTREPID II's four studies, the researchers collect information from cases, relatives, and controls at baseline and at 2 year follow up. All measures, where necessary, have been translated into local languages and back translated to check equivalence.

All those who consent are interviewed and assessed by trained research workers using structured instruments and protocols either at home or at a local clinic. For participants who are in contact with health services, interview data are supplemented with reference to clinical notes, with participants' consent.

Interviews and assessments are conducted by researchers fluent in the local language. To ensure consistency of methods across settings, all researchers are fully trained using a mixture of online materials and exercises, with feedback, and face to face training, delivered both by the United Kingdom (UK) team and locally by senior researchers under the supervision of the country principal investigators (PIs). All PIs are experienced psychiatrists with extensive backgrounds in both national and international research. Inter-rater reliability for core instruments that require rater judgement will be tested regularly across settings using video-recorded interviews with cases and controls to ensure that the measures are applied consistently throughout the duration of the programme. Responses will be triangulated with relative reports and, where applicable, clinical records.

Reliability

All measures will be applied identically, by the same research team, for both cases and controls (where measures apply to both groups). Researchers from across the field settings rated video-taped interviews at study onset and their ratings were compared to gold standard responses developed by the PIs.

Analysis Plan

The researchers will use standard summary statistics, with indicators of spread and precision as appropriate (e.g., crude incidence rates per 100,000 person years, with 95% confidence intervals) to describe the data. They will then use appropriate regression models to compare data between and within settings (e.g., Poisson regression for incidence rates and other count data; Cox regression for time-to-event data; logistic regression (including multinomial) for categorical data [e.g., course type]; and linear regression for continuous data [e.g., GAF score, blood pressure]). In building regression models, the researchers will first fit univariable models, then test for effect modification by core variables (e.g., gender, age, setting, and time), and finally adjust for putative confounders of each hypothesised association by fitting multivariable models.

Where appropriate, the researchers will use multiple imputation to deal with missing data. In addition, or where assumptions necessary for imputation are not met, researchers will (re) conduct analyses on participants with complete data only. Where possible, analyses based on imputed data will be presented, with complete data analyses presented as sensitivity analyses in supplementary materials.

Framework Analysis will be used to analyse qualitative data, adopting an iterative process of reading and annotating transcripts to identify salient themes, which will form the basis for comparisons between and within settings. ;

Related Conditions & MeSH terms

• Mental Disorders
• Psychotic Disorders

• NCT number: NCT04336137
• Study type: Observational
• Source: King's College London
• Contact: Professor Craig Morgan
• Phone: (+44)20 7848 0351
• Email: craig.morgan@kcl.ac.uk
• Status: Recruiting
• Start date: May 1, 2018
• Completion date: June 30, 2022