Bloodwork interpretation in veterinary clinical apps

Open the table with the lab results filter on

Reading the numbers versus interpreting the panel

Vendors use the phrase lab interpretation for two quite different jobs. The simpler job is capture: getting the values off a printout, a PDF or an analyzer feed and into the medical record so nobody has to retype a chemistry panel. The harder job is interpretation: looking across the values, noticing which abnormalities belong together, and suggesting what the pattern could mean for this patient.

Capture saves time and removes transcription errors. Interpretation is where clinical reasoning starts, and it is where software can mislead you if it treats each flagged value on its own. An elevated urea with a normal creatinine in a dog that just ate a high-protein meal means something different from the same urea in a dehydrated dog with dilute urine. Good interpretation reads values together and alongside the history.

The reference interval belongs to the lab that ran the sample

Every analyzer and every reference laboratory publishes its own intervals, built from its own methods and its own healthy population. A potassium value that sits inside one lab's interval can sit outside another's. Any app that interprets results should read each value against the interval printed on that specific report, not against a generic table it carries internally.

Check this during a trial. Upload a panel from your in-house analyzer and one from your send-out lab, and see whether the app flags the same borderline value differently. If it does not, it may be ignoring the report's intervals. The same caution applies to species: canine and feline intervals differ for many analytes, and so do intervals for young animals.

How results get into the app

Products tend to use one of three paths. Some connect directly to an analyzer or a reference lab portal and receive the values electronically. Some accept an uploaded PDF. Some let you photograph the printed report and read the values from the image. Each path has trade-offs:

• A direct connection is the most reliable when it exists, but it usually covers only certain lab partners or certain analyzer brands.
• A PDF upload works for most send-out labs, though it adds a step of saving and attaching the file.
• A photo works with any printed report in the room, including ones from the emergency clinic down the road, and it is the fastest when a phone is already in your hand. Glare, folds and handwritten notes on the printout are the things to test.

Trends matter as much as single values

One panel is a snapshot. Much of the real interpretive work in hospitalized and chronic patients comes from comparing panels over days or months: whether an azotemia is resolving with fluids, whether a liver enzyme is drifting upward on a long-term medication, whether a platelet count is recovering. An app that only ever looks at the most recent report misses that story. During a trial, add two or three sequential panels for the same patient and see whether the software shows the direction of change, keeps the samples in date order, and avoids comparing values from labs with different intervals as though they were the same scale.

Rechecks also expose capture errors. A value that jumps implausibly between two panels is often a misread digit or a unit mismatch rather than a clinical change, and a good tool makes that easy to spot.

What to verify before trusting an interpretation

• Spot-check the captured values against the original report, including units and decimal places.
• Confirm the app keeps the date and time of each sample, so trends across days are read in the right order.
• Look for pre-analytical cautions. Hemolysis, lipemia and delayed processing shift several common analytes, and a careful interpretation says so.
• Ask whether the interpretation changes when you add history or exam findings. If it reads the panel in isolation, its suggestions will miss context you already have.
• Open the citation for any pattern the app reports, if it offers one.

How this site marks an app as documenting lab-result reading

For lab results, an app earns a quoted cell once its maker publicly says the product reads, organizes, summarizes or interprets results. The phrase that says so is copied into the cell, linked, and stamped with the day it was read. A general statement that the app integrates with a diagnostics company does not count by itself, because it does not say what the app does with the results.

As with every column, the label records the public description, not a measured accuracy. Use the trial checks above to find out how well a given product handles your own analyzers and your own reports.