Explore the Pomegranate Health Podcast library
Browse all episodes below, starting with the most recent releases.
Latest episodes
Ep157: AI models for discharge letters and clinical coding
Large language models could assist with more prompt and more comprehensive discharge letters ensuring safe transition of patients to community care. Meanwhile, standardisation of clinical documentation could improve the accuracy of medical billing in both public and private practice.
Ep156: AI scribes in hospital
Ambient scribes promise to slash the time practitioners spend on clinical documentation but there are risks to be aware of. As they are being trialled in Australian hospitals, researchers have recognised that the efficiency gains differ from the clinic, to the ward round to the ED, and not all specialty consults benefit to the same degree.
Ep100: Conversations with ChatGPT
Natural language processor models could save hours of time spent writing clinical notes and searching through preexisting ones. But there are problematic aspects to networks on the scale of the astounding ChatGPT.
Ep99: When AI goes wrong
Uncertainty around the medicolegal aspects of AI-assisted care is of the main reasons that practitioners report discomfort about the use of this technology. It's a question that hasn’t been well tested in the courts but there is evidence about the types of adverse events that result.
Ep97: The governance of AI
The inexplainability of deep learning models creates discomfort for some clinicians and regulators. But AI-based clinical interventions can still be tested to the standards of evidence-based medicine we are accustomed to.
Ep96: The ergonomics of AI
AI-assisted medicine can help overcome some of the natural limits of human cognition. But it all depends on how seamlessly the machine learning devices fit in with decision-making in the clinical workflow.
Ep95: Machine Learning 101
Artificial intelligence can help interpretation of diagnostic images and perform very nuanced risk stratification based on medical records. But machine learning algorithms must be trained on good quality data to avoid error and bias from being introduced.
Ep92: Data-driven practice improvement
The field known as Practice Analytics seeks to provide clinicians with a bird’s eye view of their case load and performance. This can draw attention to cases that stood out from the trend and help reflection and practice improvement.
[IMJ On-Air] Making sense of HACs
Hospital-acquired complications are assumed to be preventable and to provide some metric of quality of care. But HACs may be more strongly associated with patient-related factors than they are with deviation from best practice.