A medication warning has a peculiar ambition. It wants to interrupt somebody who is already busy, tell them something they may already know, and persuade them to change what they are doing. This is a difficult social arrangement even before you put it inside a hospital computer. FDB has built a business around the knowledge behind those interruptions. Its more interesting work asks whether each interruption deserves to happen.
- The business: drug knowledge embedded in other companies’ healthcare software.
- The revealing result: an FDB customer case study reports 71% fewer hyperkalemia alerts at Lehigh Valley Health Network.
- The next move: prescription routing, patient instructions and AI-assisted drafting extend the company’s reach beyond the database.
When the warning becomes wallpaper
A warning can be medically defensible and operationally useless. Consider a software rule that spots a potentially troublesome combination of medicines. It has identified a possibility. The clinician, however, must decide what matters for the person in front of them. When possibilities arrive by the dozen, the software has delegated the sorting work to someone who was hoping it would help.
That distinction makes FDB worth understanding even if you never buy hospital software. Any product that asks for attention has to earn it. The little red notification, the fraud flag, the compliance reminder and the medication warning all compete for the same scarce resource: a person willing to stop and think.
FDB’s reported result at Lehigh Valley is striking because the gain involved subtraction. In its customer case study, hyperkalemia alert overrides fell from 81% to 38%. That measures how clinicians responded to messages. It does not establish an equivalent reduction in patient harm. The difference matters: a cleaner inbox is promising; a clinical outcome needs its own evidence.
A publisher hiding inside the software
FDB is First Databank, a Hearst Health company. Its early business included a printed drug-price periodical. Hearst acquired it in 1980. The medium has changed rather dramatically; the underlying job remains recognizable. Gather specialized information, organize it, maintain it, and put it where somebody can use it.
“We publish information,” then-president Bob Katter told Drug Topics in 2022. It is a refreshingly plain description of a business that now comes wrapped in APIs and cloud services. The publication is largely machine-readable. Hospitals, pharmacies, medical practices and insurers encounter it through software rather than by leafing through a magazine.
MedKnowledge is the core drug knowledge product. A developer licenses content and makes it part of a working application. That puts FDB upstream of the screen: a patient may recognize the hospital’s name, and the clinician may recognize the electronic record vendor, while the underlying medication knowledge comes from a different company entirely.
This is a business-to-business position with a practical advantage. One integration can put information into many users’ workflows. It also creates a constraint. Excellent content still needs the surrounding software to request the right information, interpret it correctly and display it at a useful moment. The database cannot take sole credit for the screen.
The rule needed a second draft
Lehigh Valley’s account contains a useful disappointment: the initial threshold adjustment did not reduce alert volume. Further tuning changed that. The lesson is that adding patient context is a design decision, followed by testing, rather than a ceremonial box to tick.
A laboratory value is only useful to a rule if the implementation agrees on what counts as relevant, recent and actionable. A builder copying this approach should begin with one troublesome warning category, bring clinicians into its definition, and examine how the remaining messages are handled. Counting everything the system says is easier than judging whether it said the right thing.
FDB offers two related ways into that problem. AlertSpace lets teams customize alerts. Targeted Medication Warnings brings patient context into guidance. In April 2024, the latter received Epic’s Toolbox designation, which concerns integration requirements and potential customer outcomes. It is useful evidence of compatibility, rather than a promise that every deployment produces the same result.
The scarce ingredient is the clinician’s attention.An editorial reading of FDB’s alert-fatigue work
This approach depends on usable patient data, local clinical judgment and continued review. A missing input cannot become an informed decision merely because the rule has a polished interface. Nor should an organization treat a reduction in warnings as permission to stop checking which risks remain visible.
The expensive part is keeping up
FDB sells maintained knowledge and software access, alongside contracted network services. The buyer is purchasing more than an initial pile of records. The work continues after installation: content changes, applications change, and a rule that once fitted a workflow may need another look.
Cloud Connector offers API access to medication knowledge. For a software team, the attraction is straightforward: spend less effort hosting and updating the knowledge layer. The tradeoff deserves equal attention. An external service becomes part of the application’s dependencies, and the team still owns the behavior of its finished product.
The sensible cost comparison therefore includes integration, licensing, maintenance and clinical staff time. A quoted software fee cannot tell the whole story. An internally built rule has a bill too, even when it arrives as another task on an employee’s calendar. Buyers should compare the work each option leaves behind.
A calendar with better manners
Meducation deals with the other side of the encounter: what happens when the patient takes the instructions home. Its public product listing describes personalized, plain-language plans, pictograms, font-size choices and a consolidated medication schedule. The question shifts from whether information is accurate to whether somebody can act on it.
The appeal is wonderfully unglamorous. A readable plan reduces the amount of interpretation the patient must perform. The screenshot’s morning, midday and evening symbols do some of that work before a sentence is read. A software builder outside healthcare can borrow the principle: organize the output around the next action, rather than around the structure of your database.
FDB also supplies Prizm, which organizes medical device information. The common thread is the company’s ability to make specialized knowledge usable in other systems. A device inventory and a medication calendar look different, but both punish ambiguity.
The network, the rival and the regulator
Vela, launched in 2022, moves FDB into the connections between prescribers, pharmacies and payers. In a launch-year interview, general manager Lathe Bigler said customers had asked the company to help move information as well as create it. Parker Health was identified as the network’s first customer.
That move takes FDB into competition with Surescripts. Drug knowledge has a different rival: Wolters Kluwer’s Medi-Span also supplies embedded medication data and configurable screening. Patient-specific guidance is a competitive arena; FDB’s examples, integrations and services are reasons to examine it, rather than grounds to declare the category settled.
There is an awkward historical footnote. Hearst once acquired Medi-Span. Following FTC charges, a 2001 settlement required divestiture and $19 million in disgorged profits. Drug databases may be invisible to patients, but their competitive structure has real consequences for buyers. Quiet infrastructure can still possess considerable bargaining power.
The prescription gets an AI draft
In August 2026, FDB announced Script Agent’s first commercial deployment with Tebra: encounter dialogue becomes a prescription draft for clinician review. The important boundary is the review. Generating structured medication instructions brings the technology closer to an action with consequences.
The question for a customer is whether this removes work without creating a new checking burden. For FDB, it is another test of the same expertise: put medication knowledge into a workflow, make it useful, and give the professional a manageable decision. A warning earns attention. A draft earns approval. Both have to deserve the time they ask for.