When talking to Directors of Pharmacy about some of their near term goals, an increasing number are
coming back to us asking for ways to effectively identify drug induced conditions. With 70% of Americans taking prescription drugs and 20% taking five or more, it is extremely difficult for a hospital’s Pharmacy Department to be able to link an admission to a side effect on the drug’s label, and nearly impossible to link it to a side effect that is NOT on the label.

Given how important it is to quickly ascertain this information to aid in diagnoses, it surprises me that a true standardized solution has yet to be implemented in many hospitals.

As such, here are 3 hacks that we’ve worked through with our clients to help with this pressing issue.

  1. Use statistics to identify drug reported side effects. The FDA and pharmacovigilance departments inside pharmaceutical companies have a lot data about drugs and the side effects of those drugs as they have been reported into the FDA Adverse Event Reporting System (FAERS). They use that data to calculate Reporting Odds Ratios (ROR). ROR is a disproportionality measures commonly used by drug safety professionals to help identify drug-associated AEs that are reported more frequently than expected. The methods compare the observed reporting frequencies of a specific AE for a specific drug to the expected frequency of the AE (based upon all drugs and all AEs in FARES). Using ROR to alert pharmacists to an instance where an admission’s condition could be drug related is a great way to cut through the noise and only look at what is important.
  3. Have the ability to see what side effects are occurring that are not yet on the drugs label.It’s easy to find the side effects that were identified in clinical trials, and those that the manufacturer and FDA have negotiated to put on the label once the drug comes to market- just look at the prescribing information. But what if the side effect wasn’t identified in clinical trials? Or if the side effect has yet to come under FDA scrutiny? To do this you need to have access to post market surveillance data, such as that found in FAERS and then be able to match that up to the labeling information. With this information in hand you can then use ROR to determine where an admission’s condition could be drug related as mentioned in hack 1.
  5. Interpret the likeliness of it occurring with your admission using estimated incidence rates. ROR gives us sense of relative frequency and an idea if the condition could be drug related. Labeling information can give us insight into side effects not yet being communicated out by FDA and the manufacturer. But how can we can see the actual odds of this occurring? It gets a little tricky here, but it’s possible. If you have a solid data source for patient utilization of the medications in question, you can use that as a denominator with AE case reports from FAERS as the numerator to get an estimated incidence rate. Is this a perfect way of looking at the data? No. Is it an estimation that a skilled pharmacist can use when trying to determine the likelihood a condition is drug induced? Yes.

Now that you have these 3 great drug safety pharmacy hacks, you probably have some ideas on how you can start implementing them in your facility. What you will find is that the information needed probably doesn’t exist within your current systems. It’s hard to come by and making the calculations takes a skilled informatics professional that knows the data inside and out. Luckily for you, you don’t need to spend countless hours setting up a system to identify drug induced conditions. We’ve done it for you.

Sign up for a complimentary trial of AdverseEvents Explorer to see for yourself.


Jim Davis

Jim Davis

Executive Vice President

Jim Davis LinkedinJim Davis Twitter


Jim Davis

Written by Jim Davis

As Executive Vice President, Jim is responsible for the commercialization strategy for Advera Health Analytics.