What EHR Interfaces Really Decide for a Modern Clinical Laboratory
A laboratory can hold turnaround time, pass every proficiency panel, and still watch accounts walk. The testing is not usually what goes wrong. What goes wrong is the last hundred feet, after the result is verified, where it has to land in a system a physician actually opens.
The size of that gap is documented. A systematic review of ambulatory test result follow-up published in the Journal of General Internal Medicine found that failure to follow up on laboratory results varied from 6.8 percent to 62 percent across the studies reviewed (Callen and colleagues, 2011). Those are clinical safety numbers. Read from the lab side they are also a commercial statement: a meaningful share of the work you produced never got used.
What follows is what an EHR interface is worth to a laboratory in operational and revenue terms, for the people who feel the absence of one. Lab directors, outreach managers, and whoever in lab IT owns every connection.
The Term Gets Used Loosely, So Start Here
An EHR interface is a maintained electronic connection between a laboratory information system (LIS) and a provider organization's electronic health record, carrying orders in one direction and results in the other.
A unidirectional interface carries results only: the order still arrives by requisition, phone or portal. A bidirectional interface carries both, so the physician places the order in their own EHR, the order arrives structured in the LIS, and the result returns to the chart it came from.
The compendium is your test catalog, and mapping your test codes to the codes an ordering system uses is where a surprising amount of interface work lives.
The interface catalog is the set of live connections a lab is responsible for at any given moment. It is not a project. It is inventory, and inventory has a carrying cost.
Where the Absence Shows Up First
Labs without interfaces rarely call their problem an integration problem. They call it a staffing problem, a client services problem, or a fax problem.
The fax queue is usually the first symptom: results print, failed pages get re-sent, and someone spends part of every shift babysitting a machine whose only job is to move a document from a system that produced it electronically into one that will scan it back in electronically. Then there is the phone. Client services fields calls asking for results that were already sent, because a scanned page in a document inbox is not the same as a value in a flowsheet. Nobody can trend a fax.
None of this shows up as a line item. It shows up as headcount, and as the quiet assumption that this is what running an outreach book costs.
The Errors Are Real, But Not Always the Ones People Name
Manual ordering does generate errors. Handwritten or re-keyed requisitions produce wrong tests, missing diagnosis codes, absent ask-at-order-entry information, and specimens that arrive without a matching order.
Be accurate about the scale, though, because the marketing gets loose here. In a single-institution analysis of 46,270 logged laboratory errors, 97 percent sat in the preanalytical phase, the largest categories being hemolysis, insufficient quantity and clotted specimens (Clinical Chemistry, 2024). Most preanalytical failure is a collection problem, not an order entry problem. An interface will not fix a badly drawn tube.
What it does fix is narrower and still worth having: transcription between the ordering system and the LIS disappears, required fields are enforced at the point of order rather than at accessioning, and specimens stop sitting in a hold queue while someone calls the practice to ask what was wanted.
Turnaround Time Complaints Are Usually Delivery Complaints
Ask a referring practice why they are unhappy and they will say turnaround time. Ask the lab and you will hear that analytical turnaround has been stable for two years, which is usually true. Both statements hold. The physician is measuring from the moment they wanted an answer to the moment they could see one in the chart; the lab is measuring from accession to verification. The interval between verification and visibility belongs to neither party's dashboard, and in a manual workflow it can run to hours or a day.
If you want an honest turnaround number, measure from order placed to result viewable in the ordering system, not from accession to verified. Most labs have never measured that interval. The practices you serve measure nothing else.
Why Accounts Leave, and Why It Is Rarely About the Testing
Switching laboratories is disruptive for a practice. They do it anyway, usually because the friction of using you exceeds the friction of changing.
There is a mindset behind a lot of this, and it is documented. In a 2024 federal data brief on laboratory participation in health information exchange, 36 percent of health information organizations reported laboratories limiting or refusing to share electronic health information, and one of the most common reasons given was that laboratories consider their reporting obligation to end once the result is returned to the ordering provider (ONC Data Brief No. 74). That is defensible as a compliance position. As a commercial position, it is expensive, because the national reference laboratories do not hold it, and they arrive at every practice already integrated.
A practice with a live bidirectional connection to your LIS has an ordering path that takes seconds and a result that files itself. A practice without one has a printed requisition and a document inbox. When a competitor calls, one of them has a switching cost. The other has a grievance.
What a Working Interface Program Looks Like in an Outreach Lab
Three things have to be true at once, and the third is the one that gets underfunded.
Order entry has to live where the physician already works. Not in a separate portal they log into occasionally, but inside their own EHR, with your compendium, your panels, your ask-at-order-entry prompts and medical necessity checking at the moment the order is placed rather than at your accessioning bench.
Results have to return as discrete data to the same chart: individual analytes with units, reference ranges and abnormal flags, filed against the correct patient and the correct order, so the value can be trended and so a critical flag behaves like one.
The catalog has to be maintained. Every EHR vendor interprets the standard slightly differently, versions change, practices move platforms, and a connection that silently stops passing results is worse than no connection at all, because your staff assumes orders are flowing.
That third obligation is why many outreach labs buy EHR interfaces rather than build them. A small number of specialist firms sell EHR interface solutions as an ongoing managed service instead of a one-time build. Lifepoint Informatics, which has run a vendor-neutral interfacing hub since 1999, works on that model: it maintains its own database of EMR vendor specifications so a laboratory codes to one standardized format rather than to each practice's system, provides compendium management and translation so test codes map correctly in both directions, supports unidirectional and bidirectional feeds over HL7, FHIR, CCD and CCR, and assigns project managers who coordinate with the EHR vendor and the LIS vendor during a build. Its ordering module places order entry inside the physician's own EHR, with ask-at-order-entry questions, real-time insurance eligibility, rules-based order routing and ABN generation at the ordering office.
Build or buy, the maintenance obligation does not disappear. It only moves.
The Staffing Arithmetic That Rarely Makes the Business Case
Interface work competes for the same scarce people as everything else in the laboratory, which is the argument most business cases leave out.
The American Society for Clinical Pathology's 2024 Vacancy Survey, drawing on 1,027 responses covering more than 18,626 employees, reported that overall vacancy rates had declined since 2022 but remained above pre-pandemic levels, with anatomic pathology showing the highest departmental vacancy rate at 28.5 percent (AABB summary of the ASCP survey, October 2025). In that labor market, assigning a technologist to a fax queue, or a client services representative to reading results down the phone, is not a neutral cost. It is capacity you cannot replace.
So the honest framing for a finance committee is not that interfaces save money. It is that manual result distribution consumes staff time you are already struggling to buy, and the consumption grows in proportion to the outreach book you are trying to grow.
What It Adds Up To
Interfaces are plumbing. They break quietly, and nobody thanks you when they work. They also decide whether an outreach program can scale. A laboratory that absorbs a new practice in weeks, on that practice's own EHR, without adding a person to the phones, has a growth path. A laboratory that adds a fax destination for every new client has a ceiling, and it will reach that ceiling long before the bench does.
Key Takeaways
Measure turnaround from order placed to result viewable in the ordering system. The gap between verification and visibility is where most client complaints originate.
Treat live connections as inventory with a carrying cost, not as finished projects. A silent interface failure is worse than an obvious one.
Be precise about what interfaces fix. They remove transcription and enforce order completeness. They do not fix specimen collection, where most preanalytical error lives.
Order entry inside the physician's own EHR, using your compendium and your medical necessity rules, is what creates switching cost. A separate portal creates far less.
Count the staffing, then decide deliberately whether the interface catalog is maintained in-house or contracted out. The obligation is permanent either way.
