Using the Natural Language Processing System Medical Named Entity Recognition-Japanese to Analyze Pharmaceutical Care Records: Natural Language Processing Analysis.

BACKGROUND: Large language models have propelled recent advances in artificial intelligence technology, facilitating the extraction of medical information from unstructured data such as medical records. Although named entity recognition (NER) is used to extract data from physicians' records, it has yet to be widely applied to pharmaceutical care records. OBJECTIVE: In this study, we aimed to investigate the feasibility of automatic extraction of the information regarding patients' diseases and symptoms from pharmaceutical care records. The verification was performed using Medical Named Entity Recognition-Japanese (MedNER-J), a Japanese disease-extraction system designed for physicians' records. METHODS: MedNER-J was applied to subjective, objective, assessment, and plan data from the care records of 49 patients who received cefazolin sodium injection at Keio University Hospital between April 2018 and March 2019. The performance of MedNER-J was evaluated in terms of precision, recall, and F1-score. RESULTS: The F1-scores of NER for subjective, objective, assessment, and plan data were 0.46, 0.70, 0.76, and 0.35, respectively. In NER and positive-negative classification, the F1-scores were 0.28, 0.39, 0.64, and 0.077, respectively. The F1-scores of NER for objective (0.70) and assessment data (0.76) were higher than those for subjective and plan data, which supported the superiority of NER performance for objective and assessment data. This might be because objective and assessment data contained many technical terms, similar to the training data for MedNER-J. Meanwhile, the F1-score of NER and positive-negative classification was high for assessment data alone (F1-score=0.64), which was attributed to the similarity of its description format and contents to those of the training data. CONCLUSIONS: MedNER-J successfully read pharmaceutical care records and showed the best performance for assessment data. However, challenges remain in analyzing records other than assessment data. Therefore, it will be necessary to reinforce the training data for subjective data in order to apply the system to pharmaceutical care records.

Prediction of theophylline clearance in CCl4-treated rats using in vivo CYP1A2 and CYP3A2 contents assessed with the PKCYP test.

We previously established a method to predict the drug metabolism capacity of injured liver based on pharmacokinetic estimation of the amount of cytochrome P450 (CYP) in vivo (PKCYP test), by introducing the apparent liver-to-blood free concentration gradient in vivo (qg) as a parameter. Here we show that the amount of CYP3A2 in CCl(4)-treated rats can be estimated appropriately by applying the PKCYP test using midazolam (MDZ) as a probe, assuming that the qg value in control rats does not change. We applied the results to predict the clearance of theophylline as a model drug with a physiologically based pharmacokinetic model. Male Sprague-Dawley rats were pretreated with CCl4, and the amount of CYP (A-CYP(vivo)) was quantified by Western blotting. The qg value of MDZ was determined in control rats and used to estimate the amounts of CYP3A2 in CCl4-treated rats; the result agreed well with the observed values. The qg value of CYP3A2 estimated with MDZ as a probe was used together with our previously reported value for CYP1A2 (theophylline metabolism in the liver is known to be almost entirely mediated by CYP3A2 and CYP1A2) to predict the total body clearance (CL(tot)) of theophylline in CCl4-treated rats. The predicted CL(tot) was about one-third of the observed value, which was considered acceptable. The time-course of theophylline concentration in serum simulated with a physiologically-based pharmacokinetic model agreed well with the observed values. Thus, the PKCYP test using MDZ as a probe can be used to predict the amount of CYP3A2 and the CL(tot) of theophylline in CCl4-treated rats.

Application of the PKCYP test to predict caffeine clearance mediated by CYP1A2 in a rat acute liver injury model.

We previously established a method for assessing in vivo drug-metabolizing capacity by pharmacokinetic estimation of the quantity of cytochrome P450 (CYP) in vivo (PKCYP test), in which an apparent liver-to-blood free concentration gradient in vivo (qg) is introduced (Matsunaga et al., Jpn. J. Hosp. Pharm., 26: 492-504 (2000)). This method was applied to estimate the amount of CYP2C11 in rats treated with carbon tetrachloride (CCl(4)-treated rats). In this study, we estimated the amount of CYP1A2 in CCl(4)-treated rats by using acetanilide and caffeine as a probe and a model drug, respectively. In CCl(4)-treated rats, the total body clearance (CL(tot)) of acetanilide and caffeine was about one-fifth and one-eighth of that in control rats, respectively. In CCl(4)-treated rats, the amount of CYP1A2 was predicted as 0.60+/-0.06 nmol/kg from the clearance of acetanilide mediated by CYP1A2. Moreover, the clearance of caffeine mediated by CYP1A2 in CCl(4)-treated rats was estimated as 0.47+/-0.05 mL/min/kg by using the predicted amount of CYP1A2. The observed value was 0.44+/-0.03 mL/min/kg, and the predicted value was within the 95% confidence interval of the observed value. In conclusion, we have demonstrated that the PKCYP test can also be applied for estimating the amount of CYP1A2 in CCl(4)-treated rats.