A socially assistive robot to support mental wellbeing in LGBTQ+ young people at risk of self-harm: a randomized controlled trial.

The introduction extensively cites prior research on the elevated risk of self-harm and mental health difficulties in LGBTQ+ youth, the barriers to traditional care, and the potential of digital and socially assistive interventions. It acknowledges the limitations of prior digital interventions (e.g., low engagement, weak effects) and positions Purrble as a novel approach targeting emotion regulation. The rationale linking the premise (unmet needs, emotion regulation as a transdiagnostic target) to the study objectives (evaluating Purrble's efficacy) is logically developed. The paper also addresses limitations of prior research, such as the lack of RCTs for Purrble and the need to examine subgroup differences by gender identity.

Randomization was conducted using an online generator, stratified by gender identity (cisgender vs. TGD), and allocation was concealed from the study coordinator. Blinding was not feasible given the nature of the intervention (physical device), and this is acknowledged as a limitation. A pre-specified power analysis is not explicitly reported, but the target minimum sample size was achieved. Inclusion/exclusion criteria are clearly defined. The study included a safety planning session for all participants as a standard-of-care control. No independent replication was conducted, but the paper reports attrition, sensitivity, and multiple testing corrections.

The paper reports age (mean and s.d.), gender identity (cisgender vs. TGD), sexual orientation, and race/ethnicity for both conditions in Table 1. Sex is reported as gender identity, which is appropriate for the study population. Since both sexes/genders are enrolled, a justification for single-sex is not applicable. Age and health status (baseline depression, anxiety, emotion regulation scores) are reported. Species/strain and housing conditions are not applicable for a human trial. Demographics are reported adequately.

The paper states that all study procedures were approved by the ethics committee at King's College London (RESCM-22/23-34570) and that all participants provided written informed consent. This meets the criteria for a named approving body and consent description. Regulatory compliance is implied through adherence to the CONSORT checklist and the Declaration of Helsinki is referenced in the context of the protocol publication.

The study uses a physical device (Purrble) as the intervention. It is described in the introduction and methods, but a specific model/manufacturer/catalog number is not provided. However, Purrble is a well-known device from a single manufacturer. Safety planning is described as a 'brief collaborative intervention' but specific materials or protocols are not cited or detailed. Statistical software (Qualtrics) is named, and analysis code is shared via GitHub. Antibodies, cell lines, mycoplasma, and organisms are not applicable.

The primary analysis uses two-sided ANCOVA, which is named and described. Exact p-values, effect sizes (partial eta-squared), and 95% confidence intervals are reported for primary and secondary outcomes. The statistical software (R) is identified. Data presentation includes tables with means, SDs, and Ns, and a CONSORT flow diagram. For the ANCOVA, assumptions like normality and homogeneity of variance are not explicitly verified in the text, but the use of a robust method (ANCOVA) and sensitivity analyses (e.g., outlier removal) partially address this. The self-harm analysis uses ordinal logistic regression with a proportional odds assumption check (Brant's test).

A data availability statement is present, stating that a minimally de-identified dataset is publicly available on figshare with a DOI. The full analytic dataset is available on request with a data use agreement, which is appropriate for sensitive patient data. The code is shared on GitHub. This meets the criteria for adequate data and code sharing for a clinical trial with sensitive data.

The trial is registered at ClinicalTrials.gov (NCT06025942), and the registration number is provided. The paper states it was developed in accordance with the CONSORT 2010 checklist. All pre-specified outcomes are reported, including null results for self-harm. Limitations are explicitly discussed in a dedicated section. Conclusions are generally proportional, acknowledging the need for replication and the post-hoc nature of subgroup analyses. Funding sources and competing interests are declared.