Conectar
by Miranda R. Chilver, Richard A. Burns, Ferdi Botha, Peter Butterworth
Self-report measures are widely used in mental health research and may use different recall periods depending on the purpose of the assessment. A range of studies aiming to monitor changes in mental health over the course of the COVID-19 pandemic opted to shorten recall periods to increase sensitivity to change over time compared to standard, longer recall periods. However, many of these studies lack pre-pandemic data using the same recall period and may rely on pre-existing data using standard recall periods as a reference point for assessing the impact of the pandemic on mental health. The aim of this study was to assess whether comparing scores on the same questionnaire with a different recall period is valid. A nationally representative sample of 327 participants in Australia completed a 7-day and 30-day version of the six-item Kessler Psychological Distress Scale (K6) and a single-item measure of psychological distress (TTPN item) developed for the Taking the Pulse of the Nation survey. Linear mixed models and mixed logistic regression models were used to assess whether altering the recall period systematically changed response patterns within subjects. No substantive recall period effects were found for the K6 or the TTPN, although there was a trend towards higher K6 scores when asked about the past 30 days compared to the past 7 days (b = 1.00, 95% CI: -0.18, 2.17). This may have been driven by the “feeling nervous” item which was rated higher using the 30-day compared to the 7-day recall period. Neither the K6 nor the TTPN item were significantly affected by the recall period when reduced to a binary variable of likely severe mental illness. The results indicate that altering the recall period of psychological distress measures does not substantively alter the score distribution in the general population of Australian adults.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 