TL;DR: A hospital lab integrates with one EMR. A reference lab integrates with every client it signs, to systems it does not control and cannot prioritize. That makes per-interface cost and build time a growth constraint rather than an IT detail. Five questions separate vendors who have run this at scale from vendors who will learn on your client accounts.
- Ask the cost of the tenth interface, not the first. The first is priced into the deal.
- Ask who owns the interface after go-live, and whether that is the same party who built it.
- Ask how long from signed client to live interface, and what sits outside the vendor’s control.
- Ask who answers at 2am when one client’s results stop flowing, because that is a client SLA and not an internal ticket.
- Ask who owns the interface specifications if you change LIS.
- HL7 v2 still carries the order and result traffic. FHIR is the direction, not yet the replacement.
If you run informatics at a reference or independent pathology lab, integration is not a phase of your LIS project. It is a recurring cost that repeats every time sales closes a deal. That is a different problem from the one most integration content addresses, and it changes which vendor answers matter.
What is LIS to EMR Integration?
It means running a separate connection to each client, carrying orders in and results back out. One LIS, many EMRs, each configured locally by an organization that is your customer rather than your colleague. The technical work is ordinary. The multiplication is what makes it hard.
How Does EMR Integration Differ for a Lab Serving Many Clients?
A hospital lab builds one interface, to an EMR owned by its own institution, with internal escalation when it breaks. A reference lab builds one per client, to EMRs owned by other organizations, with no internal escalation anywhere. The first is a project. The second is an operating model.
What is the Difference Between an Interface and an Integration?
An interface is one connection carrying one kind of message. An integration is the whole working relationship between two systems. Vendors use the words interchangeably in proposals, and pricing is usually per interface, so ask which meaning appears in the quote.
What Does HL7 v2 Actually Do?
HL7 v2 defines the message format carrying most clinical data between systems today, including lab orders and results. It is text-based and heavily customized site by site. That local variation is why each client interface takes weeks rather than hours: the standard permits enough deviation that every implementation is partly bespoke.
Where Does FHIR Fit?
FHIR is HL7’s modern API-based standard and the direction the industry is moving. HL7’s overview and ONC’s interoperability program both treat it that way. What it is not yet is a replacement for production order and result interfaces. Ask a vendor claiming FHIR support which resources they implement and whether lab orders are among them.
What is an HL7 Integration Engine, and Do You Need One?
An HL7 integration engine is middleware that routes, transforms, and monitors messages between systems. Some LIS platforms include the capability, others assume you will buy an engine separately. Ask which, because a separate engine is a licence cost, a second vendor, and another party in the escalation chain.
Does EMR Integration Happen Through an API?
Increasingly, but not exclusively. An EHR integration API can carry demographics, documents, and some clinical data, while order and result traffic still travels over HL7 v2 interfaces at most sites. Ask a vendor which of your flows would be API-based and which would not, and treat a claim of full API integration as something to verify.
What Are the Most Common EHR Integration Challenges?
Four failures, and all of them scale with your client count rather than your case volume.
Why Does Each New Client Cost More Than Expected?
Because nothing carries over. Each client configures its EMR locally, so the mapping work restarts every time. A vendor who quotes one interface price is quoting the easiest one. If your growth plan is ten new client accounts a year, per-interface cost is your growth cost, and it belongs in the LIS decision rather than in next year’s budget.
Why Do Orders Arrive Incomplete?
Because your client’s EMR was configured for clinical lab ordering, not anatomic pathology. Fields your lab requires may not exist upstream, and you cannot add them. The result is orders that pass validation and still cannot be accessioned without a phone call to someone who does not work for you.
Why Do Results Land in the Wrong Place in a Client Chart?
Because narrative pathology reports, synoptic data, and images do not map cleanly into chart fields designed for numeric values. Ask a prospective vendor to show you how a signed pathology report renders in an ordering clinician’s inbox at a client site. A web-based LIS does not solve this by itself. When a report renders badly, your client’s clinicians blame your lab, not their EMR.
What Happens When a Client Upgrades Their EMR?
Interfaces break, often quietly, and you find out from the client. A hospital EMR upgrade can change a segment your interface depends on with no notice to an outside lab. The question is not whether this happens. It is who monitors for it across every client connection, and who is contractually on the hook for the fix.
What Should You Ask an LIS Vendor About EMR Integration?
Five questions. Ask them during evaluation, in writing, and keep the answers.
What Does Each Additional Interface Cost?
Ask for a schedule, not a range, and ask specifically about the tenth. This single number determines whether signing a small client account is profitable or whether it costs you money to serve. It is the most important answer in the entire evaluation and the one most often left vague.
Who Builds the Interface, and Who Owns It After Go-live?
Two different answers, and vendors often supply only the first. A vendor who builds the interface then hands ownership to a third-party integrator has moved your escalation path outside your contract, which matters more when the outage belongs to a client.
How Long From Signed Client to Live Interface?
Ask in weeks and ask what sits outside the vendor’s control. This number is your time to revenue on every new account. A credible answer names the client’s integration queue, test environment availability, and sign-off cycles. A vendor who quotes a duration without naming a dependency has not done many.
Who Answers at 2am When a Client’s Results Stop Flowing?
Ask for the escalation path, the contracted response time, and whether monitoring is proactive or waits for the client to call you. A pathology result that does not reach a clinician is a patient safety issue and a contract breach at once. Your client will hold you responsible for a vendor’s support hours.
What Happens to Your Interfaces if You Switch LIS?
Ask who owns the interface specifications and mappings. If the vendor does, your next migration rebuilds every client connection from nothing, and with fifty clients that cost can exceed the software. Portability is a contract term, not a technical one, so it has to be negotiated now.
How Do You Handle Epic EHR Integration?
Through HL7 interfaces built against that client’s Epic instance, scheduled by the client’s integration team. The technical work is well understood. The constraint is that you are an outside vendor in someone else’s queue, with no standing to escalate.
That is the part worth probing. Ask a prospective vendor how many Epic client sites they have live, and what their queue position actually meant in calendar time. A vendor with real experience answers in months and does not sound defensive. One who quotes weeks has either never done it or is describing a test environment.
What EMR Integration Services Belong in the Contract?
Four things, in the contract rather than the proposal: per-interface pricing as a schedule including volume tiers, named ownership for build and for post-go-live support, a contracted response time for interface failures with an after-hours path, and your right to the interface specifications and mappings if you leave.
Everything else in an integration conversation is a judgement call. These four cost real money later if they stay verbal.
See What Integration Looks Like Across Every Client Account
NovoPath has built laboratory information systems for pathology labs since 1993, for independent, reference, molecular, and veterinary laboratories. Our anatomic pathology LIS integrates natively with EMRs, lab instruments, and third-party applications, and we own our interfaces rather than routing you to an integrator. Labs running many client hospital accounts operate all of them from one platform.
Request a demo of NovoPath’s anatomic pathology LIS and bring this checklist. We would rather answer the five questions than talk around them.
