Signal Detection in PV
What is A “Signal”
A signal in context of drug safety is a potential and established indicator of new adverse reaction. It is any new possible causal link between the adverse reaction and suspected medicine which was previously unknown or not completely documented.
Note- A signal is not a confirmatory finding but it is a hypothesis-generating situation that must be validated or disapproved.
Signal Management Process
Signal management process which is a part of pharmacovigilance refers to set of activities perform to determine whether there are new risks associated with an active substance or whether the risk has changed. Signal detection is one of the activities of signal management process.
Signal is a potential and established indicator offer new ADR, the evidence in a signal is only an early indication as it may change substantially overtime as more data accumulates.
Signal Detection in Pharmacovigilance
Pharmacovigilance deals with continuous monitoring of medicinal product to identify new potential risks and to develop risk minimization action plans to prevent or mitigate the risk associated with the medicinal product.
It is the act of looking for or identifying signals using even data from any source. These sources can be;
- Unsolicited sources- spontaneous reports.
- Solicited sources- organized collection of data (clinical trials etc.)
- Regulatory authorities (RA).
These collected data are carefully studied and analyzed to identify patterns which indicate the new safety information and also if the new information changes the benefit risk ratio associated with the use of a medicinal product.
Signals have qualitative and quantitative aspects, Qualitative analysis (Individual case review) of spontaneous reports, preclinical findings, experience with other similar products in class, new drug or food interactions etc.
- Quantitative analysis (Data mining algorithm) through data mining and other statistical activities.
Signal Prioritization
It helps in identifying signals with important public health impact or that may affect benefit risk balance of medicinal product. These kind of signals needs urgent attention.
The number of reports needed to provide sufficient evidence for a signal may differ depending on the nature of the effect, quality of report and possible evidence from other sources.
Signal assessment of relatively small number of cases
A single good report may be a valid signal when there are relatively small number of cases. Pharmacovigilance experts can review the data and can focus on clinical characteristics of individual cases.
Traditional PV Methods:
- Individual case review
- Aggregate analysis
- Periodic reports
Signal assessment of a huge number of cases
When the amount of data is large statistical and epidemiological methods would be useful. A huge amount of data however doesn’t preclude the importance of clinical judgment and assessment. So, in addition to the traditional hands-on review of spontaneous cases and other safety information by trained medical professionals data mining may also be carried out. This is a process of applying sophisticated statistical algorithms to last safety database to determine whether certain adverse event is being reported for a medicine with a greater frequency than expected based on our statistical model.
Data Mining Algorithms:
- Proportional Reporting Ratio (PPR)
- Multi-Item Gamma Poisson Shrinker (MGPS)
- Bayesian Confidence Propagation Neural Network (BCPNN).
Signal Management Process
Recommended Actions:
- Setting up a study to investigate the particular signal.
- Taking action related to medicine such as withdrawal, change in product information etc.
- Continuing review on the issue.
- Finding signal unreal and suspending the issue.
- Issuing a warning to health professionals in the form of a Direct healthcare professional communication.
Importance of Signal detection;
- Signal detection is one of the most important objectives of pharmacovigilance. The whole process of risk evaluation depends on effective detection of signals.
- Early signal detection can help in identifying potential risks associated with marketed drugs and the identified risks can be managed effectively which in turn helps the company to protect their brand and provide consumers improved product.
- The continuous monitoring of risk profile of a medicinal product is a legal obligation.
