Equivalent signal mining method for adverse reaction signals of srs combined medication

Abstract

The invention discloses an equivalent signal mining method oriented to SRS combined drug adverse reaction signals, including acquiring data through the spontaneous reporting system SRS of drug adverse events, and obtaining the combined drug adverse reaction signal set X={D 1 ,D 2 ,...,D x},D x To determine the drug combination as an adverse reaction signal; give priority to the signal with the smallest size of the drug combination set, and solve the largest equivalence group D of D G , combined with D G The largest equivalence group with intersection to the largest equivalence group D G ; then will not belong to D G And the signal equivalent to D is merged into D G , the D G Join X G , output the largest set of equivalence groups X G . x G Each maximum equivalence group in is the signal set with pairwise equivalence relationship between the adverse reaction signals of combined drug use. Select a signal from the maximum equivalence group to represent the signal set (that is, the maximum equivalence group), and participate in other data Analysis is beneficial to reduce the workload of the overall analysis.

Description

technical field [0001] The invention relates to the technical field of signal mining, in particular to an equivalent signal mining method for SRS combined drug adverse reaction signals, where SRS stands for Spontaneous Reporting System of Drug Adverse Events. Background technique [0002] Limited by factors such as the scope of the experiment, research objects, and time, it is impossible to complete all combined drug interaction studies in the pre-marketing stage. Many potential drug interactions are only discovered after the drug is used on a large scale and for a long time after it is marketed. Spontaneous Reporting System (SRS) provides an important data source for post-market drug interaction signal mining. For example, the US FDA Adverse Event Reporting System (FDA Adverse Event Reporting System, FAERS), etc., truly reflects the complexity of drug safety in real life. According to the survey, 60% of the adverse event reports in the patient’s drug use are more than one, ...

AI Technical Summary

Problems solved by technology

The above methods have their own advantages, but on the whole, there are common problems of large signal clues and low accuracy rate (4%), which cannot really take advantage of big data. This defect will be more prominent when mining high-order drug combination adverse reaction signals

This is because the drug combination mode in real life is very complex, and the known disequilibrium measurement method assumes that the drugs are independent and irrelevant when evaluating the risk of signal adverse reactions, and only evaluates the drug combination itself without considering the impact of other drug combinations. Other drugs (combinations) that cause a positive signal to be routinely used may also be evaluated as a false positive signal

Among them, a very typical problem is that several combination drug adverse reaction signals that appear to express differently come from the same group of patient reports in the SRS database

For medical researchers to carry out subsequent signal analysis, the existence of such signals will lead to two problems: (1) for signals from the same patient subgroup, independent data analysis will cause duplication of labor; From the perspective of data analysis, these signals are confounding factors, in other words, the number of positive signals is far less than the number of signals

At present, there is no relevant research work on how to describe and mine such signals in a standardized way.

Images