CN108122059B - Production risk identification method and automatic early warning system for pharmaceutical manufacturing enterprise - Google Patents
Production risk identification method and automatic early warning system for pharmaceutical manufacturing enterprise Download PDFInfo
- Publication number
- CN108122059B CN108122059B CN201611081205.2A CN201611081205A CN108122059B CN 108122059 B CN108122059 B CN 108122059B CN 201611081205 A CN201611081205 A CN 201611081205A CN 108122059 B CN108122059 B CN 108122059B
- Authority
- CN
- China
- Prior art keywords
- risk
- adr
- batch number
- module
- early warning
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 112
- 238000000034 method Methods 0.000 title claims abstract description 59
- 239000003814 drug Substances 0.000 claims abstract description 83
- 238000001514 detection method Methods 0.000 claims abstract description 66
- 239000011159 matrix material Substances 0.000 claims abstract description 39
- 238000012163 sequencing technique Methods 0.000 claims abstract description 7
- 206010067484 Adverse reaction Diseases 0.000 claims description 22
- 230000006838 adverse reaction Effects 0.000 claims description 22
- 238000006243 chemical reaction Methods 0.000 claims description 20
- 238000004364 calculation method Methods 0.000 claims description 17
- 238000013500 data storage Methods 0.000 claims description 15
- 230000035945 sensitivity Effects 0.000 claims description 15
- 239000002552 dosage form Substances 0.000 claims description 11
- 230000014509 gene expression Effects 0.000 claims description 9
- 238000013481 data capture Methods 0.000 claims description 8
- 239000002510 pyrogen Substances 0.000 claims description 8
- 230000001580 bacterial effect Effects 0.000 claims description 6
- 208000005422 Foreign-Body reaction Diseases 0.000 claims description 5
- PCTMTFRHKVHKIS-BMFZQQSSSA-N (1s,3r,4e,6e,8e,10e,12e,14e,16e,18s,19r,20r,21s,25r,27r,30r,31r,33s,35r,37s,38r)-3-[(2r,3s,4s,5s,6r)-4-amino-3,5-dihydroxy-6-methyloxan-2-yl]oxy-19,25,27,30,31,33,35,37-octahydroxy-18,20,21-trimethyl-23-oxo-22,39-dioxabicyclo[33.3.1]nonatriaconta-4,6,8,10 Chemical compound C1C=C2C[C@@H](OS(O)(=O)=O)CC[C@]2(C)[C@@H]2[C@@H]1[C@@H]1CC[C@H]([C@H](C)CCCC(C)C)[C@@]1(C)CC2.O[C@H]1[C@@H](N)[C@H](O)[C@@H](C)O[C@H]1O[C@H]1/C=C/C=C/C=C/C=C/C=C/C=C/C=C/[C@H](C)[C@@H](O)[C@@H](C)[C@H](C)OC(=O)C[C@H](O)C[C@H](O)CC[C@@H](O)[C@H](O)C[C@H](O)C[C@](O)(C[C@H](O)[C@H]2C(O)=O)O[C@H]2C1 PCTMTFRHKVHKIS-BMFZQQSSSA-N 0.000 claims description 4
- 230000009471 action Effects 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 3
- 239000000126 substance Substances 0.000 claims 4
- 238000007689 inspection Methods 0.000 abstract description 6
- 230000001788 irregular Effects 0.000 abstract description 4
- 238000007905 drug manufacturing Methods 0.000 abstract description 3
- 208000030453 Drug-Related Side Effects and Adverse reaction Diseases 0.000 description 163
- 206010061623 Adverse drug reaction Diseases 0.000 description 162
- 229940079593 drug Drugs 0.000 description 48
- 238000011160 research Methods 0.000 description 25
- 238000012544 monitoring process Methods 0.000 description 16
- 229940090044 injection Drugs 0.000 description 11
- 239000007924 injection Substances 0.000 description 11
- 238000002347 injection Methods 0.000 description 11
- 230000002411 adverse Effects 0.000 description 6
- 238000013528 artificial neural network Methods 0.000 description 6
- 239000000047 product Substances 0.000 description 6
- 230000002776 aggregation Effects 0.000 description 5
- 238000004220 aggregation Methods 0.000 description 5
- 230000006378 damage Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 231100000674 Phytotoxicity Toxicity 0.000 description 3
- 230000006399 behavior Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 238000009472 formulation Methods 0.000 description 3
- 230000036541 health Effects 0.000 description 3
- 239000000463 material Substances 0.000 description 3
- 239000000203 mixture Substances 0.000 description 3
- 238000010606 normalization Methods 0.000 description 3
- 230000001105 regulatory effect Effects 0.000 description 3
- WQZGKKKJIJFFOK-GASJEMHNSA-N Glucose Chemical compound OC[C@H]1OC(O)[C@H](O)[C@@H](O)[C@@H]1O WQZGKKKJIJFFOK-GASJEMHNSA-N 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 238000013480 data collection Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 229940093181 glucose injection Drugs 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 239000008354 sodium chloride injection Substances 0.000 description 2
- GSDSWSVVBLHKDQ-UHFFFAOYSA-N 9-fluoro-3-methyl-10-(4-methylpiperazin-1-yl)-7-oxo-2,3-dihydro-7H-[1,4]oxazino[2,3,4-ij]quinoline-6-carboxylic acid Chemical compound FC1=CC(C(C(C(O)=O)=C2)=O)=C3N2C(C)COC3=C1N1CCN(C)CC1 GSDSWSVVBLHKDQ-UHFFFAOYSA-N 0.000 description 1
- 206010001052 Acute respiratory distress syndrome Diseases 0.000 description 1
- 206010008531 Chills Diseases 0.000 description 1
- 208000005189 Embolism Diseases 0.000 description 1
- 206010014561 Emphysema Diseases 0.000 description 1
- 206010018691 Granuloma Diseases 0.000 description 1
- 206010037660 Pyrexia Diseases 0.000 description 1
- 208000013616 Respiratory Distress Syndrome Diseases 0.000 description 1
- 206010040047 Sepsis Diseases 0.000 description 1
- 206010040070 Septic Shock Diseases 0.000 description 1
- HJLSLZFTEKNLFI-UHFFFAOYSA-N Tinidazole Chemical compound CCS(=O)(=O)CCN1C(C)=NC=C1[N+]([O-])=O HJLSLZFTEKNLFI-UHFFFAOYSA-N 0.000 description 1
- 208000027418 Wounds and injury Diseases 0.000 description 1
- 201000000028 adult respiratory distress syndrome Diseases 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000006735 deficit Effects 0.000 description 1
- 208000009190 disseminated intravascular coagulation Diseases 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 208000021760 high fever Diseases 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- LKRWBLLGEAXPRY-UHFFFAOYSA-N metronidazole sodium Chemical compound [Na+].CC1=NC=C([N+]([O-])=O)N1CC[O-] LKRWBLLGEAXPRY-UHFFFAOYSA-N 0.000 description 1
- 229960001699 ofloxacin Drugs 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 208000001297 phlebitis Diseases 0.000 description 1
- 230000005180 public health Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000036303 septic shock Effects 0.000 description 1
- 239000000243 solution Substances 0.000 description 1
- 230000002269 spontaneous effect Effects 0.000 description 1
- 238000010561 standard procedure Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 230000001954 sterilising effect Effects 0.000 description 1
- 238000004659 sterilization and disinfection Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 229960005053 tinidazole Drugs 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000002792 vascular Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Educational Administration (AREA)
- Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The application provides a production risk identification method for a drug manufacturing enterprise and an automatic early warning system thereof, comprising the following steps: step 1, collecting and analyzing ADR report data of a target enterprise, and outputting an ADR data set; step 2, evaluating the risk degree of ADR data records with the same batch number by adopting a three-dimensional risk matrix method; step 3, sequencing the batch number risk data sets, and judging a high-risk batch number area; step 4, carrying out signal detection on the association degree between the medicine and the target ADR in the high risk time range by adopting a Bayesian signal detection method, and identifying suspected risk reasons; and 5: and outputting a high-risk time range, a high-risk batch number and variety, and possible risk reasons, and proposing corresponding recommended measures. The invention can identify the risk signals produced by the enterprises in an irregular way, judge the risk degree of the batch number, estimate the time range with higher risk and the suspected risk reason, and provide inspection suggestions for the supervision department according to the difference of the degree of the risk signals.
Description
Technical Field
The invention relates to the field of production management, in particular to a production risk identification method for a drug manufacturing enterprise and an automatic early warning system thereof.
Background
The problem of medicine safety is always the key problem of public health research, and is concerned with the life health and safety of the public. In recent years, due to the fact that production operations of some enterprises are not standard, so that 'phytotoxicity events' frequently occur, and drug production risks are more and more concerned by people. Drug production risks are mainly manifested in two ways: firstly, the risk of the production line is high, for example, the sterilization operation of the production line is improper in a period of time, so that the risk of all varieties and batch numbers output by the production line in the period of time is high; secondly, the risk of single variety, if the material feeding of a certain variety does not meet the regulations within a period of time, the production risk of the variety is higher. How to dynamically monitor the production risk of enterprises, timely discover suspected drug production enterprises with irregular operation, and timely check and control high-risk varieties and batch numbers is a problem to be urgently solved by regulatory departments.
The adverse drug reaction monitoring system is a main data source for identifying drug risk signals, in 2015, 139.8 parts of Adverse Drug Reaction (ADR) reports (hereinafter referred to as ADR reports) are received by a national adverse drug reaction monitoring network, and risk information can be mined from an adverse drug reaction database by supervision departments and production enterprises by means of an effective data mining and statistical analysis method, so that risk signal identification and risk evaluation are realized. However, at present, adverse reaction signal detection methods at home and abroad, such as a ratio report ratio method (PRR), a report ratio method (ROR), a comprehensive standard Method (MHRA), a bayesian credible interval neural network transfer method (BCPNN) and a multi-item gamma poisson distribution shrinkage estimation Method (MGPS), can only identify natural risks of drugs, cannot identify enterprise production risks which may cause 'phytotoxicity events', and a scientific and effective signal detection method is not available at home and can be used for monitoring the production risks of drug production enterprises.
In conclusion, the following problems exist in the current domestic adverse reaction monitoring: 1. the signal detection method aiming at the production risk of the enterprise is lacked, the production risk of a specific enterprise production line and variety cannot be identified, and the production line and variety of the drug production enterprise cannot be effectively supervised and checked; 2. the automatic early warning system for identifying the enterprise risk signals is lacked, abundant adverse reaction database resources cannot be efficiently utilized, dynamic monitoring and real-time monitoring of enterprise production risks are achieved, high-risk signals cannot be timely pushed to supervision departments and enterprises, and major phytotoxicity events are prevented.
Disclosure of Invention
The invention relates to a production risk identification method of a drug manufacturing enterprise and an automatic early warning system thereof, wherein the method and the system can identify risk signals of the non-standard production of the enterprise, judge the risk degree of a certain batch of drugs, estimate the production or occurrence time range with higher drug risk and suspected risk reasons, and provide corresponding inspection suggestions for a supervision department according to the difference of the degree of the risk signals.
According to one aspect of the invention, a method for identifying production risks of a pharmaceutical enterprise is provided, which comprises the following steps: step 1, collecting and analyzing ADR report data of a target enterprise, and outputting an ADR data set; each record in the ADR dataset comprises at least: report number, manufacturing enterprise, general name of medicine, batch number, ADR occurrence time, ADR expression, receiving time, severity or not, and adverse reaction state; step 2, taking a plurality of ADR data records related to the same batch number as a basic unit, and evaluating the risk degree of the ADR data records of the same batch number by adopting a three-dimensional risk matrix method, thereby obtaining the risk grade of each batch number and manufacturing a batch number risk data set; step 3, sequencing the batch number risk data sets to enable the sequenced batch numbers to approximately show the production sequence, judging high-risk batch number areas according to the risk early warning signal level standard, finding out the ADR occurrence time in each ADR data record of each same batch number in the high-risk batch number areas, selecting the earliest ADR occurrence time and the latest ADR occurrence time, and taking the ADR occurrence time period generated by the earliest ADR occurrence time and the latest ADR occurrence time as the high-risk time range of the high-risk batch number areas; step 4, a Bayesian signal detection method is adopted to carry out signal detection on the degree of association between the drug and the target ADR in the high risk time range, and suspected risk reasons are identified; and 5: and outputting a high-risk time range, high-risk batch numbers and varieties and possible risk reasons, and providing corresponding recommended measures according to the highest early warning signal level in the high-risk batch number area. The object of the present invention is a product of a certain type or a certain production line of an enterprise, and a product of the same formulation can be produced as the same production line.
In another aspect of the present invention, an automatic early warning system for production risk of pharmaceutical enterprises is provided, which includes an ADR data module, a data capture module, a three-dimensional risk matrix module, a high-risk lot number calculation area module, a bayesian detection module, and an automatic early warning module, wherein: the ADR data module comprises a data input unit and a data output unit, wherein the data input unit is used for receiving enterprise information set by a user, and the enterprise information comprises an enterprise name, a production line or a medicine universal name; the data capture module is used for reading in an ADR report of an enterprise in an ADR database according to enterprise information, acquiring a dosage form, a general name and a batch number of a medicine and ADR occurrence time and storing the dosage form, the general name and the batch number of the medicine and the ADR occurrence time as an ADR data set; the three-dimensional risk matrix module is used for receiving the ADR data set, calculating the risk level of each batch number in the ADR data set and storing the risk level in the data storage module; the high-risk recording area calculating module is used for calculating a high-risk batch number area and a high-risk time range according to the risk early warning signal level; the Bayesian detection module is used for carrying out Bayesian detection on the ADR report of the high-risk batch number area; and the automatic early warning module is used for outputting high-risk time range, high-risk batch number and variety, detection items and suggested measures of the enterprise.
The method and the system can be used for monitoring the production risk of enterprises in real time and dynamically by a drug monitoring department, and measures are taken in time once suspicious signals are found, so that the production operation behaviors of the enterprises are standardized, the occurrence of serious drug injury events is reduced, and the life health safety of people is guaranteed; in addition, an ADR database can be embedded into the system, the risk conditions of all enterprises in the ADR report library are automatically analyzed along with the continuous receiving of ADR reports in the ADR database, high-risk enterprises and high-risk time region information are pushed to an automatic early warning module system operating platform, and specific enterprises, varieties, batch numbers and detection items are subjected to supervision and inspection in a targeted manner according to output results, so that the investment of manpower, material resources and financial resources is greatly reduced, and the efficiency of supervision and inspection work is improved.
The method and the system can be used for monitoring the risks of enterprises, and the batch number risks of production lines and varieties of the enterprises can be monitored in real time by actively collecting the ADR reports of the enterprises and establishing the ADR report database, so that risk sources can be identified and traced, the nonstandard behaviors of staff can be restrained, problem sources can be found in time, the risk of drug production and circulation can be reduced, and the product quality and the brand benefit can be improved.
Compared with the prior art, the invention has the advantages that: (1) the method for detecting the drug production risk signal and the automatic early warning system (2) are provided for the first time, a risk time sequence method is created, wherein a core part of the method, namely a high risk time range division standard, is obtained through repeated data inspection, and certain innovativeness and feasibility are achieved; (3) a risk index T value in a three-dimensional risk matrix model is created for measuring the aggregation degree (4) of ADR occurrence time of batch numbers, and the target settings of unbalanced four-grid tables 'target medicine', 'other medicine', 'target ADR' and 'other ADR' in Bayesian signal detection are optimized. The optimization is mainly represented by: firstly, in order to highlight the high risk of a certain production enterprise in a period of time, the production enterprise and other production enterprises producing the same variety or the same dosage form variety are taken as comparison objects, target medicines and other medicines are changed into a certain variety or a certain production line of the target enterprise and other enterprises producing the same variety or the same dosage form production line; secondly, in order to identify the production risk, the target ADR is not limited to one or one type of ADR expression, but is an ADR expression associated with the production operation risk, such as pyrogen reaction and the like, so that Bayesian signal detection is more focused on finding a production risk signal; and taking the high-risk time range as Bayes detection time, so that Bayes signal detection is more targeted, and the existing Bayes detection has no clear requirement or regulation on the time range. (5) The three methods of the three-dimensional risk matrix model, the risk time series method and the Bayesian signal detection are combined for the first time. The three methods are different in emphasis and have respective advantages, and meanwhile, the loops are buckled and supplement each other, so that the adverse reaction monitoring system can not only discover signals, but also estimate specific risk time, risk reasons and risk lot numbers, and greatly improve the efficiency of the monitoring work.
Drawings
FIG. 1 is a block diagram of a three-dimensional risk matrix model according to one embodiment of the present invention;
FIG. 2 is a hierarchical block diagram of a three-dimensional risk matrix model according to one embodiment of the present invention;
FIG. 3 is a flow chart illustrating a method for enterprise production risk identification according to an embodiment of the present invention;
FIG. 4 is a schematic diagram illustrating an architecture of an enterprise risk automatic warning system according to an embodiment of the present invention; and
fig. 5 is a schematic structural diagram of an automatic early warning system for risk in pharmaceutical enterprise production according to an embodiment of the present invention.
As shown, specific structures and devices are labeled in the figures to clearly realize the structures of the embodiments of the present invention, but this is only an illustration and is not intended to limit the present invention to the specific structures, devices and environments, and one of ordinary skill in the art can adjust or modify the devices and environments according to specific needs, and the adjusted or modified devices and environments are still included in the scope of the appended claims.
Detailed Description
The method for identifying the production risk of the pharmaceutical enterprise and the automatic early warning system provided by the invention are described in detail below with reference to the accompanying drawings and specific embodiments.
In the following description, various aspects of the invention will be described, however, it will be apparent to those skilled in the art that the invention may be practiced with only some or all of the structures or processes of the present invention. Specific numbers, configurations and sequences are set forth in order to provide clarity of explanation, but it will be apparent that the invention may be practiced without these specific details. In other instances, well-known features have not been set forth in detail in order not to obscure the invention.
One of the purposes of the invention is to provide a method for identifying production risks and early warning signals of pharmaceutical enterprises, which is based on ADR report data of pharmaceutical enterprises and adopts a three-dimensional risk matrix method, a risk time sequence method and a Bayesian signal detection method to identify the production risks of the enterprises. The three-dimensional risk matrix method is used for evaluating the risk degree of a certain batch of medicines of a production enterprise from three dimensions of possibility, severity and sensitivity; the risk time sequence method is that a batch number risk time sequence of a production enterprise is established, and the time range of an ADR report with higher risk is estimated according to a high-risk batch number region division standard; the Bayesian signal detection method is to perform Bayesian signal detection on the target drug and the target ADR by adopting a Bayesian belief propagation neural network method based on the ADR report data of the high-risk batch number, and identify the suspected risk reason.
The invention provides a method for identifying production risks of a pharmaceutical enterprise, which comprises the following steps:
step 1, collecting and analyzing ADR report data of a target enterprise, and outputting an ADR data set; each record in the ADR dataset comprises at least: report number, manufacturing enterprise, general name of medicine, batch number, ADR occurrence time, ADR expression, receiving time, severity or not, and adverse reaction state.
Step 2, taking a plurality of ADR data records related to the same batch number as a basic unit, and evaluating the risk degree of the ADR data records of the same batch number by adopting a three-dimensional risk matrix method, thereby obtaining the risk grade of each batch number and manufacturing a batch number risk data set;
step 3, sequencing the batch number risk data sets to enable the sequenced batch numbers to approximately show the production sequence, judging high-risk batch number areas according to the risk early warning signal level standard, finding out the ADR occurrence time in each ADR data record of each same batch number in the high-risk batch number areas, selecting the earliest ADR occurrence time and the latest ADR occurrence time, and taking the ADR occurrence time period generated by the earliest ADR occurrence time and the latest ADR occurrence time as the high-risk time range of the high-risk batch number areas;
step 4, a Bayesian signal detection method is adopted to carry out signal detection on the degree of association between the drug and the target ADR in the high risk time range, and suspected risk reasons are identified;
and 5: and outputting a high-risk time range, high-risk batch numbers and varieties and possible risk reasons, and providing corresponding recommended measures according to the highest early warning signal level in the high-risk batch number area. The object of the present invention is a product of a certain type or a certain production line of an enterprise, and a product of the same formulation can be produced as the same production line.
The ADR data collection and analysis refers to the collection, normalization and classification of ADR reports relating to study subjects.
ADR refers to adverse drug reactions which are harmful reactions of qualified drugs under normal usage and dose and are unrelated to the purpose of medication.
The ADR report can be from a spontaneous report system or adverse drug reaction monitoring organizations of countries, provinces, autonomous regions and direct prefectures, the receiving range is most suitable nationwide, and the ADR report data items at least comprise: report number, name of manufacturing enterprise, universal name of medicine, batch number, ADR expression, ADR occurrence time, ADR receiving time, severity, and adverse reaction state.
The regular ADR report means that according to the standards of an ADR term library, an ADR medicine information library, a manufacturer information library and a medicine general name information library, irregular manufacturers, medicine general names and ADR expression information in the ADR report are adjusted, and the ADR report with unknown production enterprise names, medicine general names, batch numbers, ADR expressions and ADR occurrence time is screened out. The ADR report classification refers to the classification and summarization of ADR reports related to research objects by taking batch numbers as units, and an ADR data set is formed after the classification and the normalization.
The three-dimensional matrix method is to realize risk classification of batch-number drugs by establishing a three-dimensional risk matrix model (as shown in fig. 1 and 2).
The three-dimensional risk matrix model considers the risk degree from 3 dimensions of probability, severity and sensitivity. Likelihood refers to the frequency of occurrence of the risk, severity refers to the severity of the risk impairment, and sensitivity refers to the degree of risk aggregation. Each dimension is designed with a risk indicator, e.g., likelihood risk indicator can be measured by R value, severity risk indicator can be measured by SIADRAnd measuring the value, wherein the sensitivity risk index can be measured by using the T value. Each index is divided into 4 risk level ranges, for example, represented by ABCD, ABCD, and 1234, respectively, for a total of 64 risk combinations. The 64 risk combinations are classified into 4 risk classes, e.g. blue, yellow, orange and red, which in turn can be classified into class iv, iii, ii, i, with the highest risk for class i (red) and the lowest risk for class iv (blue).
TABLE 1 three-dimensional matrix Risk combinations
The probability risk index R value refers to the number of adverse reactions occurring in a certain batch of a certain drug, is a multiple of the average number of adverse reactions occurring in each batch of the product, and is also called a batch adverse reaction occurrence multiple. The R value is used for measuring the relative quantity of a certain batch of medicines in an ADR report, the production batches of the same variety of the same production enterprise are generally the same, and the R value can be used for expressing the risk probability of the batch of medicines on the premise of the same batch.
The calculation method is as follows:
r is A/B/C or A C/B.
A: the number of adverse reactions of a certain batch of medicines A is generally more than or equal to 3 times.
B: the total amount of adverse side effects of the medicine.
C: the number of batches of the medicine for all adverse reactions
Severity risk index SIADRThe values are used to measure the severity of damage reported by drug ADR, also known as an ADR severity index. SI (Standard interface)ADRThe severity of ADR performance in each ADR report is scored by formulating a severity grading standard, the SI of a certain batch of drugADRThe value is the average score reported for all ADRs for that batch of drug. The calculation method is as follows:
establishing ADR performance severity grading standards, wherein different grading standards correspond to different scores;
judging the score reported by each ADR in the batch number;
calculating average score of batch number, namely SI of batch number medicineADRThe value of the one or more of the one,
wherein S is the ADR report severity score value of the same batch number, and n is the ADR report number of the same batch number.
Table 2 severity ranking criteria example
The value of the sensitivity risk index T refers to the frequency of ADR reports occurring on the same batch number of medicines and is also called batch aggregation rate. The risk level of a lot of a drug is related to the aggregation of the time of occurrence of ADRs of that lot, and the more ADRs of a lot of a drug aggregate during the same time, the higher the risk level.
The calculation method comprises the following steps:
sorting ADR reports of the same batch of medicines from morning to evening according to ADR occurrence time;
calculating the difference value of the ADR occurrence time in two adjacent ADR reports;
thirdly, dividing the time difference into 7 types of '0 day', '1-3 days', '4-7 days', '8-15 days', '16-30 days', '31-90 days' and 'more than 90 days', and respectively assigning 7 points, 6 points, 5 points, 4 points, 3 points, 2 points and 1 point;
fourthly, counting the number x of the time difference values of each classi
Calculating T value
Where i is the time packet class, xiAs a number of time differences in the packet, fiTo assign a score to each group category, n is the ADR report number for the same lot number.
TABLE 3 time grading criteria and assigned values
And calculating the risk index number of each batch number, judging the risk combination of the batch number, and identifying the risk level of the risk combination, namely the risk level of the batch number. And recording each batch number and related information in the ADR data set into a batch number risk data set, wherein each record comprises: lot number and risk level. The batch number is not repeatedly recorded.
The division of the high risk time range refers to a time range in which a certain variety of an enterprise or a certain production line has a high risk in the whole production time sequence. The calculation method comprises the following steps: (1) sorting all records in the batch number risk data set from small to large according to batch numbers, wherein the sorted records can show the production time sequence of an enterprise; (2) because the risk degree grade of each batch number is calculated through the three-dimensional risk matrix model, the risk change trend of all batch number medicines of a production enterprise can be visually judged through the time sequence, and a batch number area with higher risk is calculated according to the risk early warning signal grade; (3) the ADR occurrence times of all ADR data records of all lots in the lot area are listed, and the time interval of the earliest ADR occurrence time and the latest ADR occurrence time is taken as a high risk time range.
The risk early warning signal level standard is set by using the sorted adjacent 3 batch numbers as a group, and dividing 9 risk early warning signal levels according to the difference of the colors of the 3 batch numbers in each group (namely, the risk levels are different), for example, the risk levels can be represented by D1, C2, C1, B3, B2, B1, A3, A2 and A1, and the risk is gradually increased from D1 to A1. According to the risk degree and the occurrence continuity, setting the region meeting the following conditions as a high-risk batch number region (1) with the early warning signal level above B2; (2) the early warning signal level appears for more than 2 times continuously, and is more than C1; (3) the early warning signal level above C2 appears for more than 3 times continuously. As shown in table 2, corresponding recommended measures are given for each early warning signal level at the same time.
TABLE 4 early warning signal partition Standard
The Bayesian signal detection is to detect ADR signals by analyzing the occurrence probability between target drugs and target ADR through a Bayesian belief propagation neural network method based on ADR report data of a high risk time range.
The Bayesian credible propagation neural network method formula is as follows:
and (4) judging the standard:
wherein:
γij＝1，αi＝βj＝1，α＝β＝2，cij＝A，ci＝A+B，cj＝A+C，N＝A+B+C+D
a represents the reported number of target adverse events of the target drug; b represents the report quantity of other adverse events of the target medicine; c represents the report number of target adverse events of other medicines; d represents the number of reports of other adverse events with other drugs.
In order to excavate the production risk of an enterprise, a target medicine is set as a certain variety or a certain production line of the target enterprise; the other medicines are medicines of the same variety produced by other enterprises or the same dosage form production lines of other enterprises; the target ADR comprises at least: (1) bacterial reactions (ADR manifested as sepsis, septic shock, disseminated intravascular coagulation, acute respiratory distress syndrome); (2) pyrogen reaction (ADR is manifested as chills, fever, shivering, high fever); (3) visible foreign body reactions (ADR manifested as phlebitis, emphysema, vascular granuloma, embolism); (4) severe reactions (adverse status in ADR report is "severe"); other ADRs are other ADR manifestations in addition to the target adverse event.
The conclusion drawn is the high risk time range, high risk batch number and variety, possible risk reasons and suggested measures for a certain variety or a certain production line of an enterprise.
The reason is probably that the Bayesian signal detection generates the detection item with the strongest signal.
High risk lot numbers are lot numbers with lot number risk grades in red, orange and yellow within this high risk time range.
The suggested measure is to provide a corresponding suggestion for the checking form of the supervision department according to the level of the risk early warning signal and the Bayesian monitoring detection result.
The invention also aims to provide a system for automatically early warning the production risk of a pharmaceutical enterprise. The automatic early warning system for the enterprise production risk at least comprises an ADR data module, a three-dimensional risk matrix module, a batch number area module for calculating high risk, a Bayesian detection module and an automatic early warning module (as shown in figures 4 and 5).
The ADR data module comprises data input, data output, data classification, data normalization, data capture and data storage. In one embodiment of the present invention, the ADR data module includes: (1) the input module is used for setting information such as the universal name, the batch number, the ADR occurrence time, the ADR performance, the adverse reaction state, whether the ADR is serious and the like of the medicine; (2) the output module is used for displaying the calculation process and result information thereof and summarizing and outputting the suggested result; (3) the ADR report library is used for storing and managing collected ADR reports (4), a manufacturer information library and a medicine universal name information library, storing the information of the manufacturers and the medicines universal names, screening ADR reports (5) of the manufacturers and the universal names, an ADR medicine information library used for storing, classifying and regulating the medicine universal names (6) in the ADR reports, an ADR expression name (7) data capture module used for storing, classifying and regulating the ADR reports, capturing the ADR reports meeting the setting requirements, and generating a new ADR report database (8) data storage module used for storing, reading and managing ADR report database files, activity database files, configuration files and history files.
The three-dimensional risk matrix module adopts a three-dimensional risk matrix model to evaluate the risk degree of all the batch-number medicines of the research enterprise from three dimensions of possibility, severity and sensitivity. The specific calculation method is partially consistent with the three-dimensional risk matrix model in the method provided by the invention.
The high-risk batch number calculating area module is used for identifying batch numbers and time ranges with higher risk in the whole production time sequence of a certain variety or a certain production line of an enterprise. The specific calculation method is consistent with the division of the high risk time range part in the method provided by the invention.
The Bayesian detection module is used for carrying out Bayesian signal detection on the target medicine and the target ADR within the high-risk time range, and the specific calculation method is consistent with the Bayesian signal detection part in the method provided by the invention.
The automatic early warning module is used for automatically early warning high-risk signals of research enterprises, and automatically outputs high-risk time ranges, high-risk batch numbers and varieties, detection items and suggested measures of the research enterprises according to the results of three-dimensional risk matrix calculation, high-risk time region division and Bayesian project detection.
The specific operation steps of one embodiment of the automatic risk early warning system for the drug production line are shown in fig. 3 and are described in detail below.
(1) Setting relevant information of research enterprises: the information setting of enterprises, production lines, general names of medicines and batch numbers is realized through an input module, and the execution steps are as follows: firstly, establishing a new research enterprise path, judging whether the research path exists, if so, returning to the reset path; if the information of the new research enterprise is created, inputting the information of the enterprise, the production line or the general name of the medicine according to the information database of the manufacturer and the information database of the general name of the medicine.
(2) ADR report data of research enterprises are captured: and reading ADR reports related to research enterprises in the ADR database into a data capturing module, classifying the ADR reports according to batch numbers, and storing the ADR reports in excel file cells. The execution steps are as follows: judging whether an enterprise related to an ADR report is a research enterprise, if so, entering the second step, otherwise, giving up grabbing, and returning to the first step for re-grabbing; incorporating the ADR report into an excel file, reading information of dosage form, general drug name and batch number in the ADR report, and classifying according to the batch number; reading from the first cell in the ADR report library, and repeating the first step and the second step until the last cell.
(3) Calculating the risk degree of the research enterprise batch number: reading the excel file generated by the data capture module into the three-dimensional risk matrix module, calculating the risk index of each batch number of the research enterprise, and judging the risk degree of the batch number according to the risk matrix model. The execution steps are as follows: reading all ADR report data of a batch number in an excel file; calculating three risk index values of the batch number according to a calculation method of the risk indexes from 3 dimensionality designed risk indexes of risk possibility, damage severity and sensitivity; judging the color area where the risk index value of the batch number is located through a three-dimensional risk matrix model, outputting the risk degree value of the batch number, and bringing the risk index value and the risk degree value into a data storage module; and fourthly, repeating the first step and the third step from the batch number of the first cell of the excel file until all batch numbers in the excel file are brought into the data storage module, and analyzing the repeated batch numbers.
(4) Calculating a high-risk batch number area: and sequencing all the lot numbers of the research enterprises in the data storage module, outputting high-risk lot number areas and ADR occurrence time ranges corresponding to the areas according to the risk early warning signal levels, and bringing the high-risk lot number areas and the ADR occurrence time ranges into the data storage module.
(5) Bayesian detection is carried out for different detection items: through a Bayesian detection module, the possible risk sources of the enterprise in the high-risk time region are deduced. The specific execution steps are as follows: reading an ADR occurrence time range of a research enterprise in a data storage module; setting the time range of Bayesian detection as the ADR occurrence time range, setting the target medicine as a research enterprise, other medicines as other enterprises producing the same variety/same dosage form variety, setting the target adverse reaction as bacterial reaction, and carrying out Bayesian signal detection according to a credible propagation neural network algorithm; thirdly, respectively changing the target medicine into a pyrogen reaction, a visible foreign body reaction or a severe reaction, and carrying out Bayesian signal detection again; fourthly, in the four-time Bayes detection, a signal is generated at least once, and the ADR generation time range, the item for generating the signal and the Bayes value are brought into the data storage module; otherwise, returning to the first step until the last ADR time range.
(6) The automatic early warning of the high-risk time area of the research enterprise is realized: and the real-time monitoring of enterprise risk signals is realized through the automatic early warning module. The specific execution steps are as follows: firstly, bringing the ADR occurrence time range, the items generating signals and the Bayes value in the data storage module into an automatic early warning module; secondly, incorporating the batch numbers with the risk degrees of red, orange and red in the high-risk batch number area and the general names of the medicines into an automatic early warning module; thirdly, according to the risk degree of the highest early warning signal in the high risk time area, bringing the corresponding supervision suggestion into an automatic early warning module; and fourthly, pushing the high-risk enterprise, the high-risk time range, the batch number, the risk reason and the supervision suggestion to an automatic early warning module system operation platform.
(7) Output of the suggested results: and the output module exports the enterprise name, the high-risk time area, the high-risk batch number, the variety and the supervision measure in the automatic early warning module in an excel form and stores the project path.
The first embodiment is as follows: risk identification of injection production line of certain enterprise in Guangdong province
The method comprises the following specific steps:
1. study subject ADR data collection and analysis: the research objects are injection production lines of a certain enterprise in Guangdong province, and the injection varieties are 6, namely metronidazole sodium chloride injection, glucose injection, tinidazole glucose injection and ofloxacin sodium chloride injection. The report of the nationwide ADR related to 6 injection varieties of a certain enterprise in Guangdong province received from 2012, 1 month and 1 day to 2015, 8 months and 30 days in the Guangdong center is 702 in total, and the related batch numbers are 504 in total.
2. And (3) carrying out risk classification on the batch number by adopting a three-dimensional matrix method: and realizing the risk classification of the batch number by establishing a three-dimensional risk matrix model.
(1) Risk index is designed from 3 dimensions of possibility, severity and sensitivity, in the example, probability is measured by multiple of batch adverse reaction incidence (R value), and ADR severityIndex (SI)ADR) Severity is measured and batch aggregation rate (T-value) is measured sensitivity.
(2) Calculating the risk index of each batch number: sequentially comparing the R value and the SI value of 3 risk index values of 504 batchesADRValues and T values are calculated. Taking one of the lot numbers 130920401 as an example, the R value was 4.69, SIADRThe value is 1, the T value is 3.67, the risk index value is placed in the three-dimensional risk matrix area, and the risk grade of the batch number is judged to be orange (the R value is more than 4.5 and the SI is more than or equal to 1 and is less than or equal to 1)ADRA value < 2, a value < 3 < T < 4.5, a risk combination Da3)
(3) Judging the risk degree of the batch number: after calculation, the risk degree of 26 of 504 batches is orange, the risk degree of 32 batches is yellow, and the risk degree of 372 batches is blue.
3. Calculating high risk time Range (i.e. calculating high risk batch number region and calculating high risk time Range)
(1) Storing all the batch numbers and the risk grades in a text form, and sequencing the batch numbers from small to large;
TABLE 5 ordering example of batch numbers of a certain enterprise in Guangdong
(2) According to the risk early warning signal level, dividing a batch number area with higher risk, and judging that the injection production line of the enterprise has 10 high-risk batch number areas; the automatic early warning signals of the regions 1-10 are respectively C1-C2-B1, C2-C2-B2, C1-C2-B3, C1-B2-B2, C2-C3-C1-B2, C2-C2-B3, B3-B2-C3, C2-C3-B3-C1, C2-C1-B2 and C1-C2-C1-B2.
TABLE 6 high-Risk batch number area 1 example for a certain enterprise in Guangdong
(3) Finding out the ADR occurrence time recorded by the ADR data of each batch number area, wherein the ADR occurrence time ranges of 10 areas are respectively from 20 days at 9 months in 2012 to 7 months in 2013, from 18 days at 5 months in 2013 to 27 months in 2013, from 16 days at 8 months in 2013 to 28 months in 2013, from 9 days at 11 months in 2013 to 17 months in 2014 4, from 1 month in 2014 1 to 12 months in 2014, from 23 days at 4 months in 2014 to 11 months in 2014 6, from 8 months in 2014 to 10 months in 2015 15, from 20 days at 12 months in 2014 to 22 months in 2015 2, from 7 days at 5 months in 2015 to 3 months in 2015 8 and from 3 months in 2015 to 7 months in 2015 to 14 months in 2015 7 months.
4. Bayesian signal detection
The Bayesian signal detection of the production line is carried out on 10 detection areas, and the result shows that the pyrogen reaction detection of the areas 9 and 10 generates signals, the visible foreign matter reaction detection of the areas 7 and 8 generates signals, and no signal is generated in the bacterial reaction and the severe reaction detection of the 10 detection areas.
5. To draw a conclusion
And (4) because the Bayesian detection of the areas 7, 8, 9 and 10 generates signals, only the specific results of the areas 7, 8, 9 and 10 are displayed, and a conclusion table is output. The conclusion table "ADR occurrence time range" is the high risk time range of each region; the "recommended action" is a recommended action corresponding to the highest warning signal level in the automatic warning signals of the regions (see table 2), and since the highest warning signal levels in the automatic warning signals of the regions 7, 8, 9, and 10 are B2, B3, B2, and B2, respectively, the corresponding recommended actions are daily check, interview, daily check, and daily check; the 'monitoring focus' is a detection item for generating signals by Bayesian detection; "proposed spot check lot numbers" are lot numbers with risk levels of red, orange, and yellow in regions 7, 8, 9, and 10; the "recommended spot check variety" is a variety name corresponding to a lot number with a risk level of red, orange, and yellow.
TABLE 7 Risk identification conclusion Table for injection production line of certain enterprise in Guangdong province
Example two: risk identification is carried out on injection production line of certain enterprise in Guangdong province by using automatic risk early warning system
The method comprises the following specific steps:
1. setting relevant information of research enterprises: the information setting of enterprises, production lines, general names of medicines and batch numbers is realized through an input module, and the execution steps are as follows:
(1) inputting a name of a certain enterprise in Guangdong province in a system input module, and establishing a research path of the certain enterprise in Guangdong province;
(2) in the system, the data item of the ' formulation ' is selected to be ' injection ', and the data item of the ' common name of the medicine is selected to be ' full selection '.
2. ADR report data of research enterprises are captured: and reading ADR reports related to research enterprises in the ADR database into a data capturing module, classifying the ADR reports according to batch numbers, and storing the ADR reports in excel file cells. The results show that: the ADR report of 5762 related to injection, 25 varieties and 2901 batch numbers are captured.
3. Calculating the risk degree of the research enterprise batch number: reading the excel file generated by the data capture module into the three-dimensional risk matrix module, calculating the risk index of each batch number of the research enterprise, and judging the risk degree of the batch number according to the risk matrix model. The results show that: of the 2901 lot numbers, 203 were orange, 419 yellow, and 2279 blue. The execution steps are as follows:
(1) the risk indexes designed from 3 dimensions of risk probability, damage severity and sensitivity are R value and SIADRInputting the three risk index calculation methods into a three-dimensional risk matrix module;
(2) reading all ADR report data of a batch number in an excel file, and respectively calculating the R value and SI of the batch numberADRA value and a T value;
(3) judging the risk combination and the color area where the R value, the SIADR value and the T value of the batch number are positioned through a three-dimensional risk matrix model, outputting the risk degree value of the batch number, and outputting the batch number and the SIADRThe value, the R value, the T value and the risk degree value are contained in a data storage module;
(4) and (4) repeating the steps (1) to (3) from the batch number of the first cell of the excel file until all batch numbers in the excel file are brought into the data storage module.
4. Dividing high risk time areas: the method comprises the steps of sequencing 209 batch numbers of an injection production line of a certain enterprise in Guangdong province in a data storage module, outputting a high-risk batch number area and an ADR occurrence time range, and bringing the high-risk batch number area and the ADR occurrence time range into the data storage module. The results show that: a total of 16 high-risk time regions are identified.
5. Bayesian detection is carried out for different detection items: and (3) carrying out Bayesian signal detection on the 16 high-risk time regions respectively according to a credible propagation neural network algorithm, wherein detection items comprise four items of bacterial reaction, pyrogen reaction, visible foreign body reaction and severe reaction. The results show that: detection of pyrogen reactions in regions 1, 2, 3, 9, 10, 11, 12, 13, 14 produces a signal; the detection of visible foreign body reactions in the regions 4, 5, 6, 7, 8, 9, 10, 15, 16 generates signals; detection of severe reactions in regions 7, 8 produces a signal.
6. Output of the suggested results: the output module exports the enterprise name, the high risk time area, the high risk batch number, the variety and the supervision measure in the automatic early warning module in an excel form, and the result is as follows:
TABLE 8 output conclusion of injection production line early warning system of certain enterprises in Guangdong
In the two embodiments, one side of the embodiment is the method provided by the invention, and the other side of the embodiment is the system provided by the invention, and the first embodiment and the second embodiment are combined, so that the method and the system can dynamically monitor the production risk of an enterprise, timely find high-risk signals suspected of irregular operation, and provide data support for supervision and inspection of a drug monitoring department, thereby greatly reducing the investment of manpower, material resources and financial resources, improving the efficiency of supervision and management work, standardizing the production operation behavior of the enterprise, and ensuring the life health safety of people.
Finally, it should be noted that the above examples are only intended to describe the technical solutions of the present invention and not to limit the technical methods, the present invention can be extended in application to other modifications, variations, applications and embodiments, and therefore all such modifications, variations, applications, embodiments are considered to be within the spirit and teaching scope of the present invention.
Claims (12)
1. A production risk identification method for a pharmaceutical enterprise comprises the following steps:
collecting and analyzing ADR report data of a target enterprise, and outputting an ADR data set, wherein each record in the ADR data set at least comprises: report number, manufacturing enterprise, general name of medicine, batch number, occurrence time, ADR expression, receiving time, severity and adverse reaction state;
step 2, taking ADR data records of the same lot number as a basic unit, and evaluating the risk degree of the ADR data records of the same lot number by adopting a three-dimensional risk matrix method, thereby obtaining the risk grade of each lot number and manufacturing a lot number risk data set;
step 3, sequencing the batch number risk data sets to enable the sequenced batch numbers to show the production sequence, judging high-risk batch number areas according to a risk early warning signal level standard, finding out ADR occurrence time in each ADR data record of each same batch number in the high-risk batch number areas, and selecting the earliest ADR occurrence time and the latest ADR occurrence time, wherein ADR occurrence time periods generated by the earliest ADR occurrence time and the latest ADR occurrence time are used as high-risk time ranges of the high-risk batch number areas;
step 4, a Bayesian signal detection method is adopted to carry out signal detection on the degree of association between the medicine and the target ADR in the high risk time range, and suspected risk reasons are identified; and
and 5: outputting the high risk time range, the high risk batch number and variety and possible risk reasons, and proposing corresponding recommended measures according to the highest early warning signal level in the high risk recording area,
wherein, the three-dimensional risk matrix method in step 2 designs risk indexes from 3 dimensions of possibility, severity and sensitivity, respectively, and calculates the risk level of each lot number in the ADR data set, including:
step 21: the probability risk index is represented by R, and the calculation formula is as follows:
R＝A*C/B
wherein, A: the number of adverse reactions of the same batch number is A is more than or equal to 3, B: total number of adverse reactions, C: the number of all batches with adverse reactions;
step 22: SI for severity risk indexADRThe calculation method is shown as follows:
establishing ADR performance severity grading standards, wherein different grading standards correspond to different scores;
judging the score reported by each ADR in the batch number; and
SIADRthe formula is the average score of the batch number:
wherein the content of the first and second substances,Sthe severity scoring value reported for each ADR in the same lot,nreporting the number of ADRs in the batch number;
step 23: the sensitivity risk index is represented by T and is calculated by the following method:
sorting ADR reports of the same batch number from morning to evening according to the occurrence time of the ADR;
calculating the difference of the ADR occurrence time in two adjacent ADR reports;
dividing the time difference into 7 types of '0 day', '1-3 days', '4-7 days', '8-15 days', '16-30 days', '31-90 days' and 'more than 90 days', and respectively assigning 7 points, 6 points, 5 points, 4 points, 3 points, 2 points and 1 point;
counting the number of time difference values of each class
(ii) a And
calculating the value of T
Wherein the content of the first and second substances,iin the form of a time packet of the type,x i is the number of time difference values in the packet,f i to assign a value to each of the packet categories,nreporting the number of ADRs in the batch number; and
step 24: combining the 3 risk indicators into a plurality of risk levels, assigning a risk level to each record in the ADR dataset.
2. The method for identifying risks in pharmaceutical industry manufacturing of claim 1, wherein step 1 comprises organizing and sorting by manufacturing company and/or dosage form in the ADR report and outputting the result as an ADR data set.
3. The method for identifying risks in pharmaceutical industry manufacturing according to claim 1, wherein the step 24 divides the 3 risk indicators into 4 levels respectively, forms 64 level combinations, and divides the 64 level combinations into 4 risk levels.
4. The pharmaceutical enterprise production risk identification method of claim 1, wherein the step 3 comprises:
step 31: sorting the batch numbers of the same production enterprise or the same dosage form of the same production enterprise in the batch number risk data set from small to large;
step 32: judging a high-risk batch number area according to the risk early warning signal level; and
step 33: and taking the time interval of the earliest occurrence time and the latest occurrence time in the high-risk batch number area as a high-risk time range.
5. The method for identifying risks in pharmaceutical enterprise manufacturing according to claim 4, wherein the step 32 is:
and grouping the sorted records in the step 31, wherein each 3 adjacent records are in a group, 9 risk early warning signal levels are set, the risk is gradually increased from D1 to A1 as represented by D1, C2, C1, B3, B2, B1, A3, A2 and A1, and the group in which the early warning signal above B2 appears, the early warning signal above C1 appears for more than 2 times continuously or the early warning signal above C2 appears for more than 3 times continuously is the high-risk batch number area.
6. The method for identifying risks in pharmaceutical industry manufacturing according to claim 4 or 5, wherein the target ADR in step 4 comprises at least a bacterial response, a pyrogen response, a visible foreign body response and a severe response.
7. The method for identifying risks in pharmaceutical enterprise manufacturing of claim 5, wherein the step 32 further comprises providing recommended actions corresponding to the pre-warning signal level.
8. The automatic early warning system for the production risk of the pharmaceutical enterprise comprises an ADR data module, a data capture module, a three-dimensional risk matrix module, a high-risk batch number area calculating module, a Bayesian detection module and an automatic early warning module, wherein the ADR data module, the data capture module, the three-dimensional risk matrix module, the high-risk batch number area calculating module, the Bayesian detection module and the automatic early warning module are arranged in the system, and the ADR data module is used for calculating the high-risk batch number area calculating module
The ADR data module comprises a data input unit and a data output unit, wherein the data input unit is used for receiving enterprise information set by a user, and the enterprise information comprises an enterprise name, a production line or a medicine universal name;
the data capturing module is connected with the data input unit and the three-dimensional risk matrix module and used for reading in an ADR report of the enterprise in an ADR database according to the enterprise information, acquiring a dosage form, a general name and a batch number of a medicine and ADR occurrence time and storing the dosage form, the general name and the batch number of the medicine and the ADR occurrence time as an ADR data set;
the three-dimensional risk matrix module is connected with the automatic early warning module and/or the batch number area module for calculating high risk, and is used for receiving the ADR data set, calculating the risk level of each batch number in the ADR data set and storing the risk level in a data storage module connected with the three-dimensional risk matrix module;
the high-risk calculating recording area module is connected with the automatic early warning module and/or the Bayesian detection module and is used for calculating a high-risk batch number area and a high-risk time range according to the risk early warning signal grade;
the Bayesian detection module is connected with the automatic early warning module and is used for carrying out Bayesian signal detection on ADR reports of high-risk batch number areas; and
the automatic early warning module is connected with the data output unit and is used for outputting high-risk time range, high-risk batch numbers and varieties, detection items and suggested measures of enterprises,
wherein the three-dimensional risk matrix module calculates a risk rating for each lot number in the ADR dataset from 3 dimensions of likelihood, severity, and sensitivity, wherein:
the probability risk index is represented by R, and the calculation formula is as follows:
R＝A*C/B
a: the number of adverse reactions of the same batch number is A is more than or equal to 3, B: total number of adverse reactions, C: the number of all batches with adverse reactions;
SI for severity risk indexADRThe calculation method is shown as follows:
establishing ADR performance severity grading standards, wherein different grading standards correspond to different scores;
judging the score reported by each ADR in the batch number; and
SIADRthe formula is the average score of the batch number:
wherein the content of the first and second substances,Sthe severity scoring value reported for each ADR in the same lot,nreporting the number of ADRs in the batch number; and
the sensitivity risk index is represented by T and is calculated by the following method:
sorting ADR reports of the same batch number from morning to evening according to the occurrence time of the ADR;
calculating the difference of the ADR occurrence time in two adjacent ADR reports;
dividing the time difference into 7 types of '0 day', '1-3 days', '4-7 days', '8-15 days', '16-30 days', '31-90 days' and 'more than 90 days', and respectively assigning 7 points, 6 points, 5 points, 4 points, 3 points, 2 points and 1 point;
counting the number of time difference values of each class
(ii) a And
calculating the value of T
Wherein the content of the first and second substances,iin the form of a time packet of the type,x i is the number of time difference values in the packet,f i to assign a value to each of the packet categories,nreport the number for ADR in the lot.
9. The automatic early warning system for the production risk of pharmaceutical enterprises according to claim 8, wherein the 3 dimensions of possibility, severity and sensitivity are divided into 4 grade ranges respectively represented by ABCD, ABCD and 1234, and the 64 grade combinations of the 3 dimensions are divided into 4 risk grades.
10. The automatic pre-warning system for risks in pharmaceutical enterprise production of claim 8, wherein the high-risk lot number area calculating module sorts the lot numbers of the same enterprise in the lot number risk data set from small to large, and calculates the high-risk lot number area and the high-risk time range according to the risk pre-warning signal level.
11. The automatic pre-warning system for risks in pharmaceutical enterprise production of claim 10, wherein the process of calculating high risk lot number areas is: and (3) taking the sorted adjacent 3 records as a group, dividing 9 risk early warning signal levels according to the difference of 3 batch number risk levels, and gradually increasing the risk from D1 to A1 by using D1, C2, C1, B3, B2, B1, A3, A2 and A1, wherein the high-risk batch number area is a batch number with the early warning signal level of more than B2, the early warning signal level of more than C1 continuously for more than 2 times or the early warning signal level of more than C2 continuously for more than 3 times.
12. The automatic early warning system of the production risk of pharmaceutical enterprises according to claim 8, wherein the Bayesian signal detection includes at least bacterial reaction, pyrogen reaction, visible foreign body reaction and severe reaction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611081205.2A CN108122059B (en) | 2016-11-30 | 2016-11-30 | Production risk identification method and automatic early warning system for pharmaceutical manufacturing enterprise |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611081205.2A CN108122059B (en) | 2016-11-30 | 2016-11-30 | Production risk identification method and automatic early warning system for pharmaceutical manufacturing enterprise |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108122059A CN108122059A (en) | 2018-06-05 |
| CN108122059B true CN108122059B (en) | 2021-12-24 |
Family
ID=62226319
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611081205.2A Expired - Fee Related CN108122059B (en) | 2016-11-30 | 2016-11-30 | Production risk identification method and automatic early warning system for pharmaceutical manufacturing enterprise |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108122059B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109273096B (en) * | 2018-09-05 | 2021-07-13 | 南京邮电大学 | Medicine risk grading evaluation method based on machine learning |
| CN111143652A (en) * | 2019-12-26 | 2020-05-12 | 广州中国科学院软件应用技术研究所 | Food and drug safety risk monitoring system |
| CN111192126A (en) * | 2019-12-27 | 2020-05-22 | 航天信息股份有限公司 | Invoice false-proof method and system based on big data analysis |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102339415A (en) * | 2010-07-22 | 2012-02-01 | 北京市药品监督管理局药品认证管理中心 | Risk evaluation system for medicine production and management enterprises |
| US8473315B1 (en) * | 2012-08-17 | 2013-06-25 | Ronald Lucchino | Detection of adverse reactions to medication using a communications network |
| CN103488890A (en) * | 2013-09-18 | 2014-01-01 | 万达信息股份有限公司 | Patient adverse drug reaction warning method and system based on Naive Bayes |
| CN103646185A (en) * | 2013-12-25 | 2014-03-19 | 顾德国 | Adverse drug reaction (ADR) report quality evaluation system and ADR report quality evaluation method |
| CN105760698A (en) * | 2016-03-18 | 2016-07-13 | 华中科技大学同济医学院附属协和医院 | Adverse drug reaction early warning and analyzing system and method |
-
2016
- 2016-11-30 CN CN201611081205.2A patent/CN108122059B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102339415A (en) * | 2010-07-22 | 2012-02-01 | 北京市药品监督管理局药品认证管理中心 | Risk evaluation system for medicine production and management enterprises |
| US8473315B1 (en) * | 2012-08-17 | 2013-06-25 | Ronald Lucchino | Detection of adverse reactions to medication using a communications network |
| CN103488890A (en) * | 2013-09-18 | 2014-01-01 | 万达信息股份有限公司 | Patient adverse drug reaction warning method and system based on Naive Bayes |
| CN103646185A (en) * | 2013-12-25 | 2014-03-19 | 顾德国 | Adverse drug reaction (ADR) report quality evaluation system and ADR report quality evaluation method |
| CN105760698A (en) * | 2016-03-18 | 2016-07-13 | 华中科技大学同济医学院附属协和医院 | Adverse drug reaction early warning and analyzing system and method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108122059A (en) | 2018-06-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110824270B (en) | Electricity stealing user identification method and device combining transformer area line loss and abnormal events | |
| Hamilton et al. | Variable decelerations: do size and shape matter? | |
| CN108122059B (en) | Production risk identification method and automatic early warning system for pharmaceutical manufacturing enterprise | |
| WO2021143268A1 (en) | Electric power information system health assessment method and system based on fuzzy inference theory | |
| CN108898311A (en) | A kind of data quality checking method towards intelligent distribution network repairing dispatching platform | |
| CN109614526A (en) | Environmental monitoring data fraud means recognition methods based on higher-dimension abnormality detection model | |
| CN109215799B (en) | Screening method for false correlation signals in combined adverse drug reaction report data | |
| CN112529327A (en) | Method for constructing fire risk prediction grade model of buildings in commercial areas | |
| CN112860769B (en) | Energy planning data management system | |
| CN111160855A (en) | Report sheet automatic auditing method, device, equipment and storage medium | |
| CN113762764A (en) | Automatic grading and early warning system and method for safety risk of imported food | |
| CN110853744A (en) | Hospital quality control management system under big data | |
| CN116307928A (en) | Object supervision management system | |
| CN110765116A (en) | Hospital honesty risk prevention and control system based on commercial intelligent system development | |
| Liu et al. | Application of hierarchical clustering in tax inspection case-selecting | |
| CN115018384A (en) | Building site security risk assessment method and system | |
| Olteanu | Cumulative sum control charts for censored reliability data | |
| CN112330095A (en) | Quality management method based on decision tree algorithm | |
| CN113642672A (en) | Feature processing method and device of medical insurance data, computer equipment and storage medium | |
| CN106355314A (en) | A Vehicle Engine Air Valve Quality Streaming Data Processing Method and Device | |
| CN117421735A (en) | Mining evaluation method based on big data vulnerability mining | |
| CN110147935B (en) | Method for establishing quality comprehensive decision model of tobacco wrapping workshop | |
| CN116703158A (en) | Risk assessment model construction method for imported food risk early warning | |
| WO2022227094A1 (en) | Data processing method and apparatus, and device and storage medium | |
| CN111143652A (en) | Food and drug safety risk monitoring system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20211224 |