Factors leading to diagnostic delay in tuberculosis in the tropical north of Australia.

BACKGROUND: Tuberculosis (TB) incidence is decreasing in the Northern Territory (NT) but still exceeds rates elsewhere in Australia. Deaths and morbidity from advanced TB continue, with delay in diagnosis a contributor to adverse outcomes. AIMS: We aimed to describe the delay in diagnosis of TB, identify risk factors for delay and examine the associations between delay and clinical outcomes. METHODS: We conducted a historical cohort analysis which included adult inpatients diagnosed with TB at the Royal Darwin Hospital from 2010 to 2020. Patient delay was measured as time from symptom onset to first seeking care, and health system delay was quantified as time from first relevant clinical contact to diagnosis. The sum of these two periods was the total delay. Ethics approval was granted by NT HREC (2020-3852). RESULTS: Eighty-four cases were included; the median total delay was 90 days (interquartile range (IQR): 60-121), patient delay was 53 days (IQR: 30-90), and health system delay was 21 days (IQR: 12-45). Patient delay was longer among patients with extrapulmonary (median: 100 days (IQR: 90-105) compared with pulmonary TB patients (39 days (IQR: 27-54), P < 0.0001). Health system delay was longer in those aged ≥45 years (30 days (IQR: 16-51) vs younger patients (14 days (IQR: 8-30), P = 0.007) and among non-smokers (31 days (IQR: 21-55) vs 21 days (IQR: 10-40), P = 0.048). Median delay was longer among patients with non-drug-related complications of disease (P < 0.0001), those admitted to critical care (P < 0.0001), and those with respiratory failure (P = 0.001). CONCLUSION: The patient delays we report are longer than reported elsewhere in Australia. The next steps will require concerted efforts to improve community awareness of TB and strategies to strengthen health systems through better resourcing and healthcare provider support.

Professionalism and clinical short answer question marking with machine learning.

Machine learning may assist in medical student evaluation. This study involved scoring short answer questions administered at three centres. Bidirectional encoder representations from transformers were particularly effective for professionalism question scoring (accuracy ranging from 41.6% to 92.5%). In the scoring of 3-mark professionalism questions, as compared with clinical questions, machine learning had a lower classification accuracy (P < 0.05). The role of machine learning in medical professionalism evaluation warrants further investigation.

Medical student knowledge and critical appraisal of machine learning: a multicentre international cross-sectional study.

To utilise effectively tools that employ machine learning (ML) in clinical practice medical students and doctors will require a degree of understanding of ML models. To evaluate current levels of understanding, a formative examination and survey was conducted across three centres in Australia, New Zealand and the United States. Of the 245 individuals who participated in the study (response rate = 45.4%), the majority had difficulty with identifying weaknesses in model performance analysis. Further studies examining educational interventions addressing such ML topics are warranted.