Tuesday, March 5, 2024

The Pitfalls of Statistical Fragility in Survey Research

  Survey research plays a crucial role in understanding complex issues, such as bias and work climate in medicine. However, the validity of the results heavily depends on the robustness of the statistical methods employed. A recent study by Carnes and colleagues serves as a reminder of the pitfalls of statistical fragility. Low response rates, small effect sizes, and multiple significance tests can lead to misleading conclusions. Researchers should strive for higher response rates and larger effect sizes to ensure the reliability of their findings and to drive meaningful change in their fields.

Heston TF. Statistical Fragility in Surveys. Acad Med 2024 Mar 1;99(3):240. doi: 10.1097/ACM.0000000000005584. Epub 2023 Dec 7.

Friday, February 16, 2024

Finding Hope and Connection During Hardship

This musical essay recounts my experience with homelessness and the emotional toll of isolation and uncertainty. Yet even in the darkest moments, acts of compassion from strangers ignite sparks of resilience. The message conveyed is one of shared humanity - we all have value and can make a difference through small gestures of warmth and kindness.

Citation: Heston, TF. Two Left Shoes. Zenodo 2024. https://purl.archive.org/theston/two-left-shoes

Friday, January 19, 2024

Foundations of Ethical Academic Research

This article examines core principles underlying rigorous, ethical research across disciplines. It emphasizes judicious topic selection to address knowledge gaps, exhaustive literature analysis to contextualize studies, aligning methodologies to research questions, transparent analysis and reporting, proactive ethical considerations, and effective communication for impact. Sustaining integrity as research evolves requires individual and institutional commitment to moral purpose.

Citation: Heston TF. Foundations of scholarly writing. In: An overview on business, management and economics research. BP International; 2024. p. 113-51. doi: https://zenodo.org/doi/10.5281/zenodo.10535789

PDF

Wednesday, January 10, 2024

Assessing Need for Workup of Coronary Artery Disease in Symptomatic Versus Asymptomatic Diabetics

Should diabetic patients with chest pain be worked up more aggressively for coronary artery disease than diabetics without symptoms? A retrospective study of over 1000 patients undergoing nuclear stress testing found that although anginal symptoms do predict myocardial ischemia in non-diabetics, they do not reliably indicate the presence of ischemia or ventricular dysfunction in diabetics. Therefore, the decision to initiate an ischemic evaluation in patients with diabetes cannot depend primarily on the presence or absence of symptomatic chest pain.

Citation: Heston, Thomas F. (2005). Anginal symptoms are not predictive of myocardial perfusion in diabetics. Journal of Nuclear Cardiology, 12(4), S108. https://doi.org/10.1016/j.nuclcard.2005.06.028

Wednesday, January 3, 2024

Leveraging Financial Math to Capture Volatility in Medicine

 A model used in finance to characterize unpredictable stock price changes could provide a useful framework for tackling variability in healthcare. By quantifying fluctuating volatility, the Heston Model may offer improved disease forecasting and individualized medication dosing. Before clinical adoption, rigorous empirical testing is essential. However, the mathematical versatility merits open-minded assessment, as validated applications could yield clinical dividends.

Citation: Heston TF. Quantifying uncertainty: potential medical applications of the heston model of financial stochastic volatility. Contemporary Perspective on Science, Technology and Research. 2024 Jan 1;3:92-103.

Monday, December 18, 2023

Chatbots built on ChatGPT delay mental health crisis referrals

 A new study found significant safety issues with publicly available conversational agents built on ChatGPT that are designed to provide mental health counseling. When given simulations of escalating depression and suicide risk, the chatbots frequently postponed referring users to human support until severe levels. Most failed to provide crisis resources and over 80% resumed conversations after insisting the user seek help. The findings indicate deficiencies in identifying hazardous states, jeopardizing user safety. More rigorous testing and prioritizing ethical considerations are needed before clinical implementation.

Citation: Heston TF. Safety of large language models in addressing depression. Cureus. 2023;15(12):e50729. https://doi.org/10.7759/cureus.50729

Wednesday, December 13, 2023

Standardizing PET/CT Analysis in Oncology

This article proposes a systematic methodology for interpreting and reporting PET/CT scans for oncology patients. By standardizing quality control, analysis, impression reporting, and recommendations for further imaging, the authors aim to reduce errors and provide clear, actionable information to referring physicians to guide patient care. Key elements include protocol validation, assessing fusion and artifacts, detailing tracer uptake parameters, categorizing findings by location and appearance, correlating PET and CT, comparing to prior imaging, summarizing impression and differential, and advising precise further imaging.

Citation: Heston TF, Wahl R, Jacene H. A systematic approach to PET/CT interpretation and reporting in oncology. Journal of Nuclear Medicine 2009;50(Supplement 2):1089.

 https://jnm.snmjournals.org/content/50/supplement_2/1089