Why Nigerian Hospitals Abandon Their EMR (and How to Avoid It)

Very few hospitals announce that they are abandoning their EMR. It happens quietly. A nurse writes vitals on paper "just until the network is back." Pharmacy keeps its own register because the system was slow last Tuesday. Billing goes back to receipt books. Six months later the software is still installed, the subscription is still being paid, and almost nobody uses it.

This is a leadership and operations problem more than a software problem. The good news is that the causes are well understood, and most can be dealt with before you sign anything.

What the evidence shows

A 2024 study in Scientific Reports looked at the University of Medical Sciences Teaching Hospital in Ondo, a hospital with several unsuccessful attempts at implementing electronic health records. Staff understood the benefits. Yet 98.7% reported very little use of electronic documentation and information sharing between departments. The main obstacles were unreliable power, poor internet connectivity and financial constraints, along with weak ICT resources and poor interoperability. Awareness of the system's benefits also predicted how much it was used. NaturePubMed

This is one hospital and 76 respondents, and it measures under-use rather than formal abandonment. Hard national figures on abandonment are scarce. But the pattern matches wider findings. EMR projects in low- and middle-income countries commonly run into limited infrastructure, poor usability and insufficient training. Reports from Ogun State, which began an EMR rollout across six hospitals in 2023, point the same way. Usage there is described as inconsistent, with infrastructure gaps, limited training and varying digital literacy cited as causes. nihmedrxiv

Here are the seven causes, each with its fix.

1. The system was built for infrastructure the hospital doesn't have

Software that needs steady power and fast internet will fail in the first outage. When it does, staff learn that the paper process is the one that always works, and they stop trusting the system.

How to avoid it: Treat power and connectivity as part of the EMR purchase, not a separate problem. Ask what happens at 2 a.m. when the grid drops and the internet goes with it. The answer should be specific: what runs on backup power, what fails over to a second connection, and what works offline.

2. Training was an event, not a programme

A two-day training before go-live is quickly forgotten. Staff turn over, new hires arrive, and the people who missed training learn shortcuts from colleagues who are already avoiding the system. Vendors that reduce long-term drop-off do it through refresher sessions, a responsive helpdesk and in-app guidance. eLAB Assist

How to avoid it: Budget for ongoing training. Train by role (a nurse, a pharmacist and a cashier need different things). Name a trained "super user" on every shift, and build training for new hires into onboarding.

3. It was run as an IT project, not an operational change

An EMR changes who does what, in what order and with what accountability. If management treats it as something IT installs, senior clinicians opt out and everyone else follows their lead. Weak management commitment is a recurring theme in EMR failure research, and one Nigerian buyer's guide lists ignored change management with senior clinicians, poor data migration planning and weak staff training among the top reasons hospital software projects fail. Nexoris Technologies

How to avoid it: Put a senior clinician and a senior administrator in charge, with named responsibilities. Set a date after which paper is no longer accepted for a given workflow. Without a cut-off, the paper habit never ends.

4. The software doesn't fit how Nigerian hospitals work

Some systems are built for other health systems and adapted afterwards. One Nigerian vendor guide argues that foreign software that is not properly localised to Nigerian workflows and operations is a major reason for repeated EMR failure, with some hospitals going fully back to paper. That is a vendor's view, but the point stands. If billing, HMO and insurance handling, pharmacy dispensing and cash-desk flow don't match how your hospital works, staff will build workarounds. AjirMed

How to avoid it: Test with a real patient journey from registration through consultation, lab, pharmacy and payment, before you buy. Ask to see how it handles your own tariffs and payer types.

5. Departments still don't talk to each other

If lab, pharmacy and billing run on separate systems, staff re-key information, and errors and delays follow. In the Ondo study, weak standardisation and interoperability ranked among the most critical obstacles. A system that only works for registration and nothing else gets abandoned for everything else.

How to avoid it: Prefer a single connected platform over a collection of separate tools. A charge entered in the consulting room should reach billing without anyone retyping it.

6. The real costs surfaced after the contract

When hardware, power, internet and support costs weren't understood at procurement, projects tend to stall mid-implementation. A hospital that has spent its budget on licences has nothing left for the router, the UPS or the support contract. eLAB Assist

How to avoid it: Price the full three years before you commit, covering software, hardware, connectivity, power, support and training. Our earlier guide to hospital EMR costs walks through how.

7. Nobody to call when it breaks

Networks fail and printers jam. If the fix takes days, the workaround becomes the process. Support capacity is a real gap: in Abuja's public hospitals, a shortage of ICT staff to provide round-the-clock service meant network problems often led to instability and failures. A 24-hour hospital cannot rely on a support desk that closes at 5 p.m. Ijsra

How to avoid it: Get response times in writing, covering weekends and nights. Ask who is on call, how you reach them, and what the vendor does if a fault isn't fixed within the agreed time.

Warning signs that adoption is slipping

• Paper registers appear beside the screen.
• Staff say "we'll enter it later," and entries are batched at the end of the shift.
• Only registration is done in the system, while lab, pharmacy or billing are not.
• People share logins to save time.
• Downtime happens but nobody reports it.
• Usage numbers are flat or falling, but nobody has looked.

What to measure

You can't manage adoption you don't measure. Track a few numbers weekly:

• The share of patient encounters documented in the system on the same day.
• The share of bills generated from the system rather than on paper.
• Active users per department.
• Downtime hours, and how long each fault took to fix.

Review them in the same meeting where you review revenue.

A simple 90-day plan

Days 1–30: Prove it on one workflow. Start with registration, billing and one clinic. Fix problems while the stakes are small. Confirm the power and connectivity fallbacks actually work by testing them.

Days 31–60: Add departments. Bring in lab and pharmacy. Name a super user per shift. Set a paper cut-off date for the workflows already live.

Days 61–90: Lock in the habit. Review the adoption numbers. Retrain the departments that lag. Report results to leadership and agree the next phase.

Hospitals rarely abandon an EMR because clinicians dislike technology. They abandon it because the system failed when power or internet failed, training faded, support was slow, or nobody in charge insisted that paper stop. Each of these can be planned for.

This is the thinking behind how we deliver LafiaLink™. Hardware, solar power, connectivity, training and 24/7 support are part of the same package as the software, because a system that works only on good days won't survive in a real hospital. At Redwood Specialist Hospital, our founding deployment, the hospital went live in under 30 days and recorded 97%+ uptime over its first 90 days of operation.

If your EMR is already slipping, or you're choosing one and want to avoid the same fate, book a demo and we'll walk through your workflows with you.