- Title
- Evaluation of the pharmacovigilance systems in Singapore and South East Asia and development of an algorithm to detect signals of potential major drug safety issues
- Creator
- Chan, Cheng Leng
- Relation
- University of Newcastle Research Higher Degree Thesis
- Resource Type
- thesis
- Date
- 2020
- Description
- Research Doctorate - Doctor of Philosophy (PhD)
- Description
- Pharmacovigilance is an important aspect of health products regulation that ensures the continued safety of marketed products. The spontaneous adverse event (AE) reports submitted by healthcare professional and companies remain a critical information source for pharmacovigilance surveillance. In Singapore, the Health Sciences Authority (HSA) is responsible for the management of the spontaneous reporting system (SRS). Traditionally, individual reports are reviewed manually to detect potential safety signals. Given the rapidly growing numbers of AE reports in Singapore and other ASEAN countries, it became important to develop a signal detection algorithm that could complement manual review in a more efficient and timely manner. As there is currently no gold standard, countries such as Singapore, with a smaller SRS database would have to establish an algorithm best suit its requirements. In-depth review of the more commonly used disproportionality methodologies based on the frequentist or Bayesian methodology was conducted. The appropriate signal detection thresholds were established for the three disproportionality measures, namely the Reporting Odds Ratio, Bayesian Confidence Propagation Neural Network and Gamma Poisson Shrinker for the Singapore context. To further enhance the efficiency of signal detection, we also tested a less commonly used algorithm, i.e., the sequential proportionality risk ratio (SPRT), for continuous monitoring. As the likelihood-based SPRT model could be used for many different probability distributions, we studied its usefulness through varying the signal detection thresholds and found that SPRT with a combination of two hypothesised relative risks of 2 and 4.1 could detect signals of both common and rare AEs when applied to the Singapore SRS. This was subsequently validated in real-time application whereby SPRT when used in combination with the other 3 methodologies resulted in a more efficient utilisation of manpower resources during manual review. Since July 2016, HSA has officially incorporated the use of these tools into routine signal detection process. Overall, our studies have laid the ground work for applying such tools into Singapore’s SRS and potentially for other ASEAN countries with databases of similar sizes.
- Subject
- pharmacovigilance algorithm; signal detection; spontaneous reporting system
- Identifier
- http://hdl.handle.net/1959.13/1412980
- Identifier
- uon:36558
- Rights
- Copyright 2020 Cheng Leng Chan
- Language
- eng
- Full Text
- Hits: 904
- Visitors: 1469
- Downloads: 696
| Thumbnail | File | Description | Size | Format | |||
|---|---|---|---|---|---|---|---|
| View Details Download | ATTACHMENT01 | Thesis | 3 MB | Adobe Acrobat PDF | View Details Download | ||
| View Details Download | ATTACHMENT02 | Abstract | 516 KB | Adobe Acrobat PDF | View Details Download |