Predicting future falls in older people using natural language processing of general practitioners' clinical notes.

BACKGROUND: Falls in older people are common and morbid. Prediction models can help identifying individuals at higher fall risk. Electronic health records (EHR) offer an opportunity to develop automated prediction tools that may help to identify fall-prone individuals and lower clinical workload. However, existing models primarily utilise structured EHR data and neglect information in unstructured data. Using machine learning and natural language processing (NLP), we aimed to examine the predictive performance provided by unstructured clinical notes, and their incremental performance over structured data to predict falls. METHODS: We used primary care EHR data of people aged 65 or over. We developed three logistic regression models using the least absolute shrinkage and selection operator: one using structured clinical variables (Baseline), one with topics extracted from unstructured clinical notes (Topic-based) and one by adding clinical variables to the extracted topics (Combi). Model performance was assessed in terms of discrimination using the area under the receiver operating characteristic curve (AUC), and calibration by calibration plots. We used 10-fold cross-validation to validate the approach. RESULTS: Data of 35,357 individuals were analysed, of which 4,734 experienced falls. Our NLP topic modelling technique discovered 151 topics from the unstructured clinical notes. AUCs and 95% confidence intervals of the Baseline, Topic-based and Combi models were 0.709 (0.700-0.719), 0.685 (0.676-0.694) and 0.718 (0.708-0.727), respectively. All the models showed good calibration. CONCLUSIONS: Unstructured clinical notes are an additional viable data source to develop and improve prediction models for falls compared to traditional prediction models, but the clinical relevance remains limited.

Reprocessing filtering facepiece respirators in primary care using medical autoclave: prospective, bench-to-bedside, single-centre study.

OBJECTIVE: There are widespread shortages of personal protective equipment as a result of the COVID-19 pandemic. Reprocessing filtering facepiece particle (FFP)-type respirators may provide an alternative solution in keeping healthcare professionals safe. DESIGN: Prospective, bench-to-bedside. SETTING: A primary care-based study using FFP-2 respirators without exhalation valve (3M Aura 1862+ (20 samples), Maco Pharma ZZM002 (14 samples)), FFP-2 respirators with valve (3M Aura 9322+ (six samples) and San Huei 2920V (16 samples)) and valved FFP type 3 respirators (Safe Worker 1016 (10 samples)). INTERVENTIONS: All masks were reprocessed using a medical autoclave (17 min at 121°C with 34 min total cycle time) and subsequently tested up to three times whether these respirators retained their integrity (seal check and pressure drop) and ability to filter small particles (0.3-5.0 µm) in the laboratory using a particle penetration test. RESULTS: We tested 33 respirators and 66 samples for filter capacity. All FFP-2 respirators retained their shape, whereas half of the decontaminated FFP-3 respirators showed deformities and failed the seal check. The filtering capacity of the 3M Aura 1862 was best retained after one, two and three decontamination cycles (0.3 µm: 99.3%±0.3% (new) vs 97.0±1.3, 94.2±1.3% or 94.4±1.6; p<0.001). Of the other FFP-2 respirators, the San Huei 2920 V had 95.5%±0.7% at baseline vs 92.3%±1.7% vs 90.0±0.7 after one-time and two-time decontaminations, respectively (p<0.001). The tested FFP-3 respirator (Safe Worker 1016) had a filter capacity of 96.5%±0.7% at baseline and 60.3%±5.7% after one-time decontamination (p<0.001). Breathing and pressure resistance tests indicated no relevant pressure changes between respirators that were used once, twice or thrice. CONCLUSION: This small single-centre study shows that selected FFP-2 respirators may be reprocessed for use in primary care, as the tested masks retain their shape, ability to retain particles and breathing comfort after decontamination using a medical autoclave.

Screening for hepatitis B and C in first-generation Egyptian migrants living in the Netherlands.

BACKGROUND: Egypt has high prevalence of hepatitis C virus (HCV) infection and intermediate prevalence of hepatitis B virus (HBV) infection; however, infection prevalence among Egyptian migrants is unknown. Considering the asymptomatic onset and development of disease in chronically-infected patients, many may remain undiagnosed. AIMS: To evaluate an HCV- and HBV-screening programme designed to identify undetected infections among first-generation Egyptian migrants in Amsterdam, the Netherlands. METHODS: In 2009 and 2010, viral hepatitis educational and screening sessions were established at Egyptian meeting places. Data regarding demographics and HCV risk factors were collected. Chronically infected participants were referred and followed up. Phylogenetic analyses were used to ascertain the geographic origin of infections. RESULTS: Eleven of 465 (2.4%; 95% CI = 1.3-4.2%) migrants had HCV antibodies; 10/11 were HCV RNA positive. All had genotype 4a, and strains were typical of those of Egypt and the Middle East. Older age and exposure to parenteral antischistosomal therapy (PAT) were significantly associated with HCV. Anti-HBc prevalence was 16.8% (95% CI = 13.7-20.4%); HBsAg prevalence was 1.1% (95% CI = 0.5-2.5%). All had genotype D, typical of those of the Middle East. Most (9/10 HCV; 3/5 HBV) chronic infections were newly diagnosed; four of the HCV-infected individuals started treatment. CONCLUSIONS: Anti-HCV and HBsAg prevalence among Egyptian migrants was lower compared with the general Egyptian population, but higher than the general population of Western countries. Phylogenetic analyses suggest that all infections were from the region of origin. HCV-screening programmes should target first-generation Egyptian migrants, especially those of older age and those who received PAT.