Thinklytics

Specialty Pharmaceutical Company · Life Sciences · Philadelphia, PA · 14 weeks

Pharmacovigilance data pipeline automated

A specialty pharmaceutical company handled adverse event reports manually, causing delays that pushed submissions beyond the FDA’s 15-day deadline. We built an automated data pipeline that cut processing time from 18 days to 4 hours and ensured every serious adverse event was submitted on time.

Challenge

The pharmacovigilance team collected adverse event reports from six different channels: call center, field sales, medical affairs, literature monitoring, spontaneous reports, and clinical trials. Each channel used its own data format, forcing the team to manually consolidate and process the information. Despite having eight specialists dedicated to this task, delays persisted, leading to late FDA submissions and warning letters.

Approach

We created an automated pipeline to handle adverse event reports from six sources. It used an AI-assisted engine to apply MedDRA codes, identified duplicate entries, and prepared MedWatch submission packages automatically. We also set up an escalation process for serious events needing 15-day expedited reporting.

Outcome

We cut adverse event processing from 18 days to 4 hours for standard reports and 6 hours for serious cases. This overhaul pushed on-time submission compliance to full within three months. Automating data handling freed up the equivalent of five of the eight specialists’ time, shifting their focus from routine processing to signal detection.

How We Cut MedDRA Coding Time in Half and Slashed Errors with AI

Manual MedDRA coding was killing our speed on adverse event reports. So, we tried out an AI tool that scans the narratives and suggests the right MedDRA preferred terms. Boom, coding time dropped from 45 minutes to just 8 per report. Our pharmacovigilance team didn’t have to start from scratch anymore; they just reviewed the AI’s suggestions. It made a massive difference.

We set up a quick, no-fuss system to catch serious adverse events right away.

The FDA only gives you 15 days to report serious adverse events, crazy tight deadline, right? So, we built this automated triage system to take the heat off. It instantly flags the most urgent cases and fast-tracks them. Then, to keep everyone on their toes, we set reminders for days 7, 10, and 13. It’s simple, but frankly, it changed the game.

Results

  • 18 days to 4 hours Adverse event processing time
  • Full On-time FDA submission compliance
  • 5 of 8 Pharmacovigilance specialists' time freed
  • 6 sources Adverse event data sources unified

We kept missing FDA deadlines and getting warning letters. Our team of eight was overloaded but still couldn’t keep up. Once Thinklytics automated the process, we’ve hit every submission on time. Now, the team focuses on signal detection instead of data entry.

Head of Pharmacovigilance, Specialty Pharmaceutical Company

In the client's terms

The question

Which of this work should a person still be doing?

The situation

Adverse event reports arrived through six channels: call centre, field sales, medical affairs, literature monitoring, spontaneous reports and clinical trials, each in its own format. Eight specialists consolidated them by hand, and the delays had already produced late FDA submissions and warning letters.

What we found

Eight specialists were spending most of their time reformatting reports rather than looking at them. The regulatory exposure and the waste of expert attention were the same problem seen from two directions.

What it produced

Processing fell from 18 days to 4 hours for standard reports and 6 hours for serious ones. On-time FDA submission compliance reached full within three months. Five of the eight specialists moved off routine processing and onto signal detection.

The roadmap

An automated pipeline ingests all six sources, applies MedDRA coding with AI assistance, identifies duplicates and assembles MedWatch submission packages. A separate escalation path handles serious events subject to 15-day expedited reporting, because those cannot sit in a general queue.

The work behind this result

Related case studies

Frequently asked questions

How fast can adverse event processing be?

Four hours for standard reports and six for serious ones at this specialty pharmaceutical company, down from 18 days across six intake channels. On-time FDA submission compliance reached full within three months, having previously produced warning letters.

Can MedDRA coding be automated?

Assisted rather than replaced. An AI engine reads the narrative and proposes preferred terms here, which the pharmacovigilance team reviews rather than researches, taking coding from 45 minutes to 8 minutes a report.

How do you handle 15-day expedited reporting reliably?

Give serious events their own path. They are flagged at intake and routed separately here, with escalation reminders at days 7, 10 and 13 rather than one deadline nobody watches. Five of eight specialists moved off routine processing onto signal detection.