Why are Drugs.com side-effect incidence rates so different from manufacturer/FDA reports?

Question

Entries at Drugs.com have a section on "Side Effects... For Healthcare Professionals" which present percentages for incidence rates of side effects. However, I've found that the rates reported there are wildly at odds with those reported on manufacturer and FDA-approved drug sheets, or any other reference I can find. (Note that Drugs.com collaborates directly with the FDA, and in fact the FDA directs readers to Drugs.com for information on drugs.)

Here's a slice of one example. Researching Reclast, among the numerous side effects listed on Drugs.com, the following are included:

Gastrointestinal

Very common (10% or more): Nausea (29.1% to 46%), vomiting (14% to 32%), constipation (26.7% to 31%), diarrhea (17.4% to 24%), abdominal pain (14% to 16.3%), dyspepsia (10%)...

Musculoskeletal

Very common (10% or more): Bone pain (55%), myalgia (23%), arthralgia (21%), back pain (15%), and limb pain (14%)...

I'll compare to the prescribing information document, produced by the manufacturer Novartis and approved by the FDA, and available on both of their websites (e.g., at FDA). The following is an excerpt from Table 1.

Comparing a few of the side-effect rates noted (Drugs.com vs. Prescribing Information document):

• Nausea: 29.1-46% vs. 4.5-8.5%.
• Vomiting: 14-32% vs. 3.4-4.6%
• Diarrhea: 17.4-24% vs. 5.2-6.0%
• Bone pain: 55% vs. 3.2-5.8%
• Myalgia: 21% vs. 4.9-11.7%

The Prescribing Information document has a few other tables reporting on other studies (for osteopenia, for men, and for Paget's disease), but the incidence rates are all quite similar to those in Table 1.

On Drugs.com, individual numbers are not given specific citations, although there is a general list of references at the bottom of the page. In the case of Reclast, there are five: (1) the Product Information document from Novartis, (2) UK summary, (3) Australian APPGuide online, (4) Australian product information, and (5) Kidney Int report (specific to tubular necrosis). Having looked at all of them, the UK and Australian reports give only qualitative descriptors for incidence rates (no percentage numbers), and the APPGuide Online doesn't appear to exist anymore.

In short: I can't find any source for the incidence rates shown on Drugs.com, and the numbers shown there are dramatically different from those in the prescribing information document from Novartis and the FDA. In some cases the rates are off by up to 50 percentage points.

Furthermore, this is true for all of the larger list of side effects shown for Reclast. And if I look at any other drug I see the same situation (e.g., currently investigating osteoporosis drugs: Fosamax, Boniva, Forteo, and Prolia).

So: Why are the side-effect incidence rates shown on Drugs.com so different from the FDA-approved prescribing information report? Where are Drugs.com getting these numbers? Which one should be considered more reliable?

Answer 1

The numbers are different because the package insert rates are based on two clinical trials while the information on Drugs.com is determined by Cerner Multum, a commercial drug database.

Adverse events frequencies reported in the drug prescribing information is typically determined by initial clinical trial data. The table included in your question indicates two studies as the source of these rates.

However, the very prescribing information you linked notes:

Because clinical trials are conducted under widely varying conditions, adverse reaction rates observed in the clinical trials of a drug cannot be directly compared to rates in the clinical trials of another drug and may not reflect the rates observed in practice.

Turning to Drugs.com, editorial policy page says:

The information available on the Drugs.com website is displayed under a number of licensing agreements with various publishers. For our drug-database information, we rely on the solid reputation of our suppliers: Cerner Multum, Micromedex and American Society of Health-System Pharmacists.

Indeed at the bottom of the Drugs.com zoledronic acid (Reclast) page we find:

References relevant to chosen section (Show all references)

1. Cerner Multum, Inc. "UK Summary of Product Characteristics."

If we review the informational page about Multum on Drugs.com

Multum's content is written by full-time associates who have no affiliations with drug companies, using a combination of sources. Initial References include the Product Information/Package Insert, primary literature and information from standards groups such as the Centers for Disease Control and Prevention (CDC), the World Health Organization (WHO) and the American Academy of Pediatrics.

Physicians and pharmacists review the new information and make any necessary changes; all leaflets are reviewed by one or more practicing clinical specialists. Content is verified using secondary references that include standard authoritative medical textbooks and the Multum Expert Review Panel.

Cerner itself is the second largest vendor of electronic health record technology in the United States, and so it is also possible that adverse events may be supplemented by information gained by their customers.

Which source should be considered more reliable is difficult to say and probably opinion based.