In this study, we used the information generated by community members during an intervention design process to understand the features needed for a successful community participatory intervention to improve child health.

We conducted a concurrent mixed-methods study (November 2019–March 2020) to inform the design and evaluation of a community–facility linkage participatory intervention.

Kiyawa Local Government Area (Jigawa State, Nigeria)—population of 230 000 (n=425 villages).

Qualitative data included 12 community conversations with caregivers of children under-5 (men, older and younger women; n=9 per group), 3 focus group discussions (n=10) with ward development committee members and interviews with facility heads (n=3). Quantitative data comprised household surveys (n=3464) with compound heads (n=1803) and women (n=1661).

We analysed qualitative data with thematic network analysis and the surveys with linear regression—results were triangulated in the interpretation phase. Participants identified the following areas of focus: community health education; facility infrastructure, equipment and staff improvements; raising funds to make these changes. Community involvement, cooperation and empowerment were recognised as a strategy to improve child health, and the presence of intermediate bodies (development committees) was deemed important to improve communication and solve problems between community and facility members. The survey showed functional community relations’ dynamics, with high levels of internal cohesion (78%), efficacy in solving problems together (79%) and fairness of the local leaders (82%).

Combining the results from this study and critical theories on successful participation identified community-informed features for a contextually tailored community–facility link intervention. The need to promote a more inclusive approach to future child health interventions was highlighted. In addition to health education campaigns, the relationship between community and healthcare providers needs strengthening, and development committees were identified as an essential feature for successfully linking communities and facilities for child health.

ISRCTN39213655.

Early identification of lung cancer on chest radiographs improves patient outcomes. Artificial intelligence (AI) tools may increase diagnostic accuracy and streamline this pathway. This study evaluated the performance of commercially available AI-based software trained to identify cancerous lung nodules on chest radiographs.

This retrospective study included primary care chest radiographs acquired in a UK centre. The software evaluated each radiograph independently and outputs were compared with two reference standards: (1) the radiologist report and (2) the diagnosis of cancer by multidisciplinary team decision. Failure analysis was performed by interrogating the software marker locations on radiographs.

5722 consecutive chest radiographs were included from 5592 patients (median age 59 years, 53.8% women, 1.6% prevalence of cancer).

Compared with radiologist reports for nodule detection, the software demonstrated sensitivity 54.5% (95% CI 44.2% to 64.4%), specificity 83.2% (82.2% to 84.1%), positive predictive value (PPV) 5.5% (4.6% to 6.6%) and negative predictive value (NPV) 99.0% (98.8% to 99.2%). Compared with cancer diagnosis, the software demonstrated sensitivity 60.9% (50.1% to 70.9%), specificity 83.3% (82.3% to 84.2%), PPV 5.6% (4.8% to 6.6%) and NPV 99.2% (99.0% to 99.4%). Normal or variant anatomy was misidentified as an abnormality in 69.9% of the 943 false positive cases.

The software demonstrated considerable underperformance in this real-world patient cohort. Failure analysis suggested a lack of generalisability in the training and testing datasets as a potential factor. The low PPV carries the risk of over-investigation and limits the translation of the software to clinical practice. Our findings highlight the importance of training and testing software in representative datasets, with broader implications for the implementation of AI tools in imaging.

In population-based research, disease ascertainment algorithms can be as accurate as, and less costly than, performing supplementary clinical examinations on selected participants to confirm a diagnosis of a neurocognitive disorder (NCD), but they require cohort-specific validation. To optimise the use of the Canadian Longitudinal Study on Aging (CLSA) to understand the epidemiology and burden of NCDs, the CLSA Memory Study will validate an NCD ascertainment algorithm to identify CLSA participants with these disorders using routinely acquired study data.

Up to 600 CLSA participants with equal numbers of those likely to have no NCD, mild NCD or major NCD based on prior self-reported physician diagnosis of a memory problem or dementia, medication consumption (ie, cholinesterase inhibitors, memantine) and/or self-reported function will be recruited during the follow-up 3 CLSA evaluations (started August 2021). Participants will undergo an assessment by a study clinician who will also review an informant interview and make a preliminary determination of the presence or absence of an NCD. The clinical assessment and available CLSA data will be reviewed by a Central Review Panel who will make a final categorisation of participants as having (1) no NCD, (2) mild NCD or, (3) major NCD (according to fifth version of the Diagnostic and Statistical Manual of Mental Disorders criteria). These will be used as our gold standard diagnosis to determine if the NCD ascertainment algorithm accurately identifies CLSA participants with an NCD. Weighted Kappa statistics will be the primary measure of agreement. Sensitivity, specificity, the C-statistic and the phi coefficient will also be estimated.

Ethics approval has been received from the institutional research ethics boards for each CLSA Data Collection Site (Université de Sherbrooke, Hamilton Integrated Research Ethics Board, Dalhousie University, Nova Scotia Health Research Ethics Board, University of Manitoba, McGill University, McGill University Health Centre Research Institute, Memorial University of Newfoundland, University of Victoria, Élisabeth Bruyère Research Institute of Ottawa, University of British Columbia, Island Health (Formerly the Vancouver Island Health Authority, Simon Fraser University, Calgary Conjoint Health Research Ethics Board).

The results of this work will be disseminated to public health professionals, researchers, health professionals, administrators and policy-makers through journal publications, conference presentations, publicly available reports and presentations to stakeholder groups.