Method for identifying production risk of drug production enterprise, and automatic early warning system

Abstract

The invention provides a method for identifying a production risk of a drug production enterprise, and an automatic early warning system. The method comprises the steps of 1, collecting and analyzingADR report data of a target enterprise, and outputting an ADR data set; 2, assessing a risk degree of an ADR data record of a same batch number by adopting a three-dimensional risk matrix method; 3, sorting a batch number risk data set, and judging a high-risk batch number region; 4, by adopting a Bayes signal detection method, performing signal detection on an association degree between a drug ina high-risk time range and a target ADR, and identifying suspected risk reasons; and 5, outputting the high-risk time range, high-risk batch numbers and varieties, and possible risk reasons, and proposing corresponding suggestions and measures. The risk signals of enterprise non-standard production can be identified; the risk degrees of the batch numbers can be judged; the relatively high-risk time range and the suspected risk reasons can be estimated; and according to different risk signal degrees, inspection suggestions are provided for supervision departments.

Description

technical field [0001] The invention relates to the field of production management, in particular to a method for identifying production risks of pharmaceutical production enterprises and an automatic early warning system thereof. Background technique [0002] Drug safety has always been a key issue in public health research, which is related to public life, health and safety. In recent years, due to the non-standard production operation of some enterprises, "drug injury incidents" have occurred frequently, and the risk of drug production has attracted more and more widespread attention. Drug production risks are mainly manifested in two aspects: one is the risk of the production line, such as improper sterilization operation of the production line within a certain period of time, resulting in higher risks for all varieties and batch numbers output from the production line during this period; the other is the risk of a single variety, such as a period of time The feeding of...

AI Technical Summary

Problems solved by technology

However, the current domestic and foreign adverse reaction signal detection methods, such as proportional reporting odds ratio (PRR), reporting odds ratio (ROR), comprehensive standard method (MHRA), Bayesian confidence interval neural network transfer method (BCPNN) and The multiple gamma Poisson distribution shrinkage estimation method (MGPS) can only identify the natural risks of drugs, but cannot identify the production risks of enterprises that may lead to "drug injury incidents". There is no scientific and effective signal detection method in China that can be used to Monitoring the production risks of drug manufacturers

[0004] To sum up, the current domestic adverse reaction monitoring has the following problems: 1. There is a lack of a signal detection method for the production risk of the enterprise, which cannot identify the production risk of the production line and variety of the specific enterprise, and cannot target the production line and variety of the drug manufacturer. Carry out effective supervision and inspection; 2. Lack of an automatic early warning system to identify enterprise risk signals, unable to efficiently use abundant adverse reaction database resources, realize dynamic monitoring and real-time monitoring of enterprise production risks, and fail to timely report to regulatory authorities and enterprises Send high-risk signals to prevent major drug accidents

Images