Conectar
by Cole Etherington, Amelia Palumbo, Kelly Holloway, Samantha Meyer, Maximillian Labrecque, Kyle Rubini, Risa Shorr, Vivian Welch, Emily Gibson, Terrie Foster, Jennie Haw, Elisabeth Vesnaver, Manavi T. Maharshi, Sheila F. O’Brien, Paul MacPherson, Joyce Dogba, Tony Steed, Mindy Goldman, Justin Presseau
IntroductionThe growing demand for plasma protein products has caused concern in many countries who largely rely on importing plasma products produced from plasma collected in the United States and Europe. Optimizing recruitment and retention of a diversity of plasma donors is therefore important for supporting national donation systems that can reliably meet the most critical needs of health services. This series of three systematic reviews aims to synthesize the known barriers and enablers to source plasma donation from the qualitative and survey-based literature and identify which strategies that have shown to be effective in promoting increased intention to, and actual donation of, source plasma.
Methods and analysisPrimary studies involving source or convalescent plasma donation via plasmapheresis will be included. The search strategy will capture all potentially relevant studies to each of the three reviews, creating a database of plasma donation literature. Study designs will be subsequently identified in the screening process to facilitate analysis according to the unique inclusion criteria of each review (i.e., qualitative, survey, and experimental designs). The search will be conducted in the electronic databases SCOPUS, MEDLINE, EMBASE, PsycINFO, and CINAHL without date or language restrictions. Studies will be screened, and data will be extracted, in duplicate by two independent reviewers with disagreements resolved through consensus. Reviews 1 and 2 will draw on the Theoretical Domains Framework and Intersectionality, while Review 3 will be informed by Behaviour Change Intervention Ontologies. Directed content analysis and framework analysis (Review 1), and descriptive and inferential syntheses (Reviews 2 and 3), will be used, including meta-analyses if appropriate.
DiscussionThis series of related reviews will serve to provide a foundation of what is known from the published literature about barriers and enablers to, and strategies for promoting, plasma donation worldwide.
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.
FreshRSS 