Publications by authors named "H I MAHON"
Article Synopsis
- Paroxysmal Nocturnal Haemoglobinuria (PNH) is a rare disorder that is difficult to diagnose due to diverse symptoms and a lack of awareness, often resulting in misdiagnosis.
- This study explores the use of a machine learning model, specifically the XGBoost algorithm, to identify undiagnosed PNH patients using electronic health records from the UK, involving 131 PNH patients and over 593,000 controls.
- The model achieved a recall rate of 27% for PNH patients, with a specificity of 99.99% for controls, and indicated that nearly 1 in 5 flagged patients may need further investigation for PNH, highlighting key symptoms like aplastic anemia and panc
Article Synopsis
- * Researchers analyzed data from the Optimum Patient Care Research Database, utilizing various modeling approaches, including logistic regression and XGBoost, to develop clinical prediction models.
- * The XGBoost model showed promising accuracy in identifying undiagnosed small intestine NETs, suggesting that these models could aid in earlier identification, though further evaluation is needed to assess their effectiveness and cost-efficiency.
Context: X-linked hypophosphatemia (XLH) is a rare genetic disorder causing renal phosphate wasting, which predicates musculoskeletal manifestations such as rickets. Diagnosis is often delayed.
Objective: To explore the recording of clinical features, and the diagnostic odyssey of children and adolescents with XLH in primary care electronic healthcare records (EHRs) in the United Kingdom.
Introduction: Hip fractures often occur in medically complex patients and can be associated with high perioperative mortality. Mortality risk assessment tools that are specific to hip fracture patients have not been extensively studied. The objective of this study is to evaluate a recently published 30-day mortality risk calculator (Hip Fracture Estimator of Mortality Amsterdam [HEMA]) in a group of patients treated at a university health system.
View Article and Find Full Text PDFArticle Synopsis
- This study assessed the implementation of MendelScan, a tool designed to find rare diseases in a UK primary care setting, addressing challenges like complexity and low physician awareness in diagnosing these conditions.
- The process involved analyzing data from 68,705 patients, using algorithms to flag potential cases based on medical records, leading to a clinical review of flagged patients to identify those needing further investigation.
- Out of the flagged cases, 75 records passed the review, resulting in 36 reports sent to GPs, with some categorized as possible diagnoses and others excluded due to existing clear diagnoses.