首页|University Hospital Reports Findings in Machine Learning (Development and evalua tion of a model to identify publications on the clinical impact of pharmacist in terventions)
University Hospital Reports Findings in Machine Learning (Development and evalua tion of a model to identify publications on the clinical impact of pharmacist in terventions)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
New research on Machine Learning is th e subject of a report. According to news reporting from Montreal, Canada, by New sRx journalists, research stated, "Pharmacists are increasingly involved in pati ent care. Pharmacy practice research helps them identify the most clinically mea ningful interventions." The news correspondents obtained a quote from the research from University Hospi tal, "However, the lack of a widely accepted controlled vocabulary in this field hinders the discovery of this literature. To compare the performance of a machi ne learning model with manual literature searches in identifying potentially rel evant publications on the clinical impact of pharmacist interventions. To descri be the dataset that was built. An automated PubMed search was performed weekly s tarting in November 2021. Titles and abstracts were retrieved and independently evaluated by two reviewers to select potentially relevant publications on the cl inical impact of pharmacists. A Cohen's kappa score was calculated. Data was col lected during an 11-month period to train a machine learning model. It was evalu ated prospectively during a 5-month period (predictions were collected without b eing shown to the reviewers). The performance of the model was compared with man ual searches (positive predictive value [PPV] and sensitivity). A transformers-based model was selected. During the prospectiv e evaluation period, 114/1631 (7 %) publications met selection crit eria. If the model had been used, 1273/1631 (78 %) would not have n eeded review. Only 3/114 (3 %) would have been incorrectly excluded . The model showed a PPV of 0.310 and a sensitivity of 0.974. The best manual se arch showed a PPV of 0.046 and a sensitivity of 0.711. On December 12, 2023, the dataset contained 8607 publications, of which 544 (6 %) met the cr iteria. The kappa between reviewers was 0.786. The dataset and the model were us ed to develop a website and a newsletter to share publications.A machine learning model was developed and performs better than manual PubMed searches to identify potentially relevant publications. It represents a conside rable workload reduction."
MontrealCanadaNorth and Central Amer icaCyborgsEmerging TechnologiesMachine Learning
NETL
NSTL
