Pharma R&D Today
Ideas and Insight supporting all stages of Drug Discovery & Development
Validation: The importance of search formulas in Literature Monitoring
Posted on February 29th, 2016 by Julio dos Anjos in Pharmacovigilance
A reasonable number of Market Authorization Holders (MAHs) have no ready answer to justify why they use a specific search formula for each specific product when Audit or Inspection time comes: in fact the search formula, responsible for all the inbound traffic in the Literature Monitoring process is, many times not validated, or regularly reviewed.
In this article, first of a series, we’ll discuss the challenges specific to Literature Monitoring and why the search formula is of paramount importance in the process.
The use of medical and scientific literature in PV
In the European Economic Area, the European Medicines Agency Good Vigilance Practice (GVP) Module VI mandates Market Authorization Holders (MAHs) to maintain active screening of scientific literature, for adverse drug reactions (ADR) related to their medicinal products. In the US, it’s the Code of Federal Regulations, Title 21, parts 314.80.b (regulating approval of Drugs for Human Use) and 600.80.b, (regulating approval of Biologicals). Japan, Brazil and many other markets maintain similar obligations.
As explained in a recent blog post by Joyce de Langen, there are complementary requirements for literature capture in relation to Pharmacovigilance obligations, among others, the requirement to use “published scientific literature or reports from abstracts, including information presented at scientific meetings” for the purposes of collecting “efficacy, effectiveness and safety information” as expressed in the GVP Module VII – Periodic safety update report.
GVP Module VI further qualifies that the capture of this literature can be intermediated by, although it may not be limited to, PubMed and Embase. Although neither the other GVP Modules, nor the CFR’s, mandate the use of an Abstract and Index (A&I) database, it is clear they have to be used. The A&I provided by these services is a surrogate accepted by the regulators, that must be complemented by having local affiliates perform a literature review of items not covered in these and other large A&I databases.
The nature of A&I databases
Abstracting and Indexing
[are] closely related activities in that both involve preparing a representation on the subject matter of documents. The abstractor writes a narrative description or summary of the document while the indexer describes its contents by using one or several index terms, often selected from some form of controlled vocabulary.
Lancaster, F.W. Indexing and Abstracting in Theory and Practice.
3rd edition. London: Facet Publishing, 2003
Abstracters and Indexers in fact create a summary surrogate of a document (like a thumbnail rendition of a photograph), describing what the document is about.
Some parts of the description are quite objective and will be assigned to fields, representing answers to common objectives questions like “who are the authors,” “where was the article published,” “where was the presentation delivered,” etc. Some databases choose to answer “what are the commercial names of the products discussed in the document and their manufacturers,” while others will only answer “what are the substances discussed in the document.”
The most powerful field is the one that answers: “what the document about?”. This question presents special problems for the indexer:
… because any document can, in principle, provide answers to an infinity of questions, subject analysis should establish priorities based on the specific user groups served (or specific services provided in the information ecology).[…] The best subject analysis is the one that makes the best prognosis of the future use of the document.
Hjørland, B. (2001). Towards a theory of aboutness, subject, topicality,
theme, domain, field, content…and relevance.
Journal of the American Society for Information Science and Technology, 52(9), 774-778
It is well known that both these thesauri are very different in structure and scope and this reflects, precisely, the difference in target audiences of each product.
The role of A&I databases in Literature Monitoring
A&I databases have a large role in Literature Monitoring. In fact, they make Literature Monitoring possible. Without A&I databases, the requirement would be to review the full text of all scientific and medical literature directly.
It would be impossible for any MAH to read all the literature in order to locate the specific articles or conference presentations that may be relevant for their products (either to the capture of relevant elements for the benefit/risk analysis or for ADR detection).
But, the same way no MAH will read the full corpus of scientific literature, also no one will read each and every record Embase and PubMed push out every day. Instead MAH’s create search formulas that bring out the relevant records in A&I databases, and execute the searches regularly. GVP VI mandates at least weekly, but many companies maintain a daily capture and review cycle.
If the execution of these searches, and retrieval of the results, use the email alerting system native to a search platform like (Embase.com), or are performed by daily human interaction with the platform of choice is an operational option of the MAH or its sub-contractor.
The review of the captured abstracts (Literature Monitoring or Screening) is a laborious process that can consume a lot of resources and, as a part of the Pharmacovigilance function, great attention is given to its timely operation and compliance with pre-established quality criteria or regulated standards.
Many times, however, the same level of diligence is skipped when creating the search formulas or providing for their maintenance.
The importance of search formulas in Literature Monitoring
These search formulas feed the Literature Monitoring process, which in turn feeds the MAH’s Safety Database, and, sub-sequentially, EudraVigilance and AERS. This leads some to consider that search formula creation and upkeep should be faced as a GxP activity, needing Verification and Validation as any other GxP process.
Although there is no specific regulatory mandate, as a part of the Pharmacovigilance system, the Literature Monitoring process can be (and in fact is) inspected and audited. It is a field in which a QPPV must be ready to answer questions, and it must be described in the Pharmacovigilance System Master File (PSMF).
If the Literature Monitoring process has the results of these search formulas as daily input, the formulation of the formulas (sorry for the pleonasm) should be validated, not only the regular execution of the searches and the review of their results.
In the next part of this series we’ll discuss the significance of validation in the context of search formulas for PV and what are the key indicators that one needs to achieve.
R&D Solutions for Pharma & Life SciencesWe're happy to discuss your needs and show you how Elsevier's Solution can help.
Julio dos Anjos
Professional Services Consultant
- Webinar: Using machine learning to identify adverse events from scientific literature
- Why literature is a valuable source for signal detection: From the EU point of view
- Elsevier uses machine learning to benefit pharmacovigilance
- How Big Data Transforms Reactive Drug Safety to Proactive Pharmacovigilance
- Are Drug-Drug Interactions Today’s Biggest Health Threat?