Automate search and deduplication in MLM-AI with Monitors

The medical and scientific literature is an important source of information regarding safety issues of medicinal products and pharmaceutical companies are required to monitor the literature periodically for events of interest affecting their products.

Literature screening of safety events is typically done by searching literature databases using keywords covering relevant products or therapeutic areas. Doing this activity manually is time consuming and subject to inefficiencies such as duplications. Identifying and keeping a record of key words, terms and phrases usually requires skills of an experienced professional. Lastly, it is repetitive: literature monitoring should be done periodically for products during its lifecycle, from clinical development to market.

Biologit MLM-AI enables users to automate literature search strategy process by configuring Monitors that continuously search relevant databases and screen articles according to a search criteria. By creating Monitors:

  • Users define a search strategy containing multiple relevant terms,

  • The search process is automated (weekly, on a target weekday) or can be performed ad-hoc,

  • Results are automatically generated, tagged using AI for faster screening, and de-duplicated.

Creating Monitors in MLM-AI

Creating a Monitor is a one-time configuration activity to begin generating results in MLM-AI, All configuration changes to Monitors are preserved in an auditable record.

Defining Product Search

A product search is defined by a list of relevant terms, searched as an OR-query. The configuration screen allows auto-expansion of a name into its synonyms (brand and generic names):

Setting up the Monitor

MLM-AI automatically ingests data from different global scientific repositories. Having a Monitor search on multiple repositories automatically is as simple as selecting them from the list:

Periodic monitoring is a frequent requirement in high-volume literature screening workflows. A Monitor allows for results to be generated on a schedule:

Monitors also define how screening activity will take place. In particular, what options users will avail of when selecting special situations and exclusions during screening. The lists are seeded with best practice values and can be customized for each Monitor.

Finally, Monitors also enable automatic filtering of incoming results according to tags. Auto-exclude settings can be selected by the user depending on the desired level of automation

Saving the monitor will automatically schedule results for the next weekday. Monitors can also be disabled whenever required. Here's the summary view for the monitor we just created:

The Results

Now everything is set for results to start flowing: Results are automatically searched, de-duplicated and presented on the scheduled weekday, or they can be run on demand, and ready for screening.

Learn More