In May 2021, Ireland’s Health Service Executive was hit by a ransomware attack that took its entire IT infrastructure offline. Hospitals reverted to paper records. Cancer treatments were delayed. Appointments were cancelled by the tens of thousands. The recovery took months and cost an estimated $600 million. The attack vector was a phishing email opened two months earlier that had gone undetected inside the network the entire time.
That’s an extreme case. But the underlying conditions that made it possible, aging software, fragmented systems, inadequate recovery planning, overconfidence in perimeter security, exist in healthcare organizations of every size.
The Integration Problem Nobody Talks About in Sales Demos
Modern healthcare software rarely operates in isolation. An EHR connects to a billing system, which connects to a patient portal, which connects to a scheduling tool, which may connect to external labs, pharmacies, or insurance verification services. Each connection is a dependency, and each dependency is a potential point of failure.
What clinics and health systems often discover after implementation is that their software vendors have different update schedules, different security postures, and different support windows. When one system updates and breaks an integration with another, patient data can get stuck in transit, appointments can fail to sync, and billing records can become inconsistent. These aren’t theoretical risks. They’re the kinds of operational problems that surface as unexplained discrepancies in patient records or revenue that doesn’t reconcile at month end.
The organizations that manage this best tend to have someone, whether internal or contracted, whose explicit job is to monitor integration health rather than assume it’s working because nothing has visibly broken yet.
Scheduling Failures Have Clinical Consequences
Operational software failures in healthcare aren’t just IT problems. They affect care delivery in direct ways that are easy to underestimate until they happen.
Pediatric appointment no-show reduction is often discussed as a revenue and scheduling problem, which it is, but the clinical dimension is significant. A child who misses a well visit because a reminder system failed, or because a parent never received confirmation of a rescheduled appointment after a system migration, isn’t just a missed slot. Depending on what that visit was for, it could mean a delayed vaccination, a missed developmental screening, or a gap in chronic condition management. Healthcare software that handles scheduling unreliably has consequences that extend well beyond the administrative.
The same logic applies to any automated clinical workflow. Medication refill alerts that fail silently, follow-up reminders that stop sending after a software update, lab result notifications that don’t reach the ordering provider because a routing rule broke during a system upgrade. These aren’t hypothetical failure modes. They’re documented incidents that healthcare organizations deal with and often underreport because attribution is difficult.
Data Recovery Is the Risk Most Organizations Ignore Until It’s Urgent
Healthcare data has some of the highest recovery requirements of any industry. Patient records need to be available continuously, accurate to the most recent entry, and recoverable within hours if something goes wrong. The regulatory dimension adds another layer: HIPAA has specific requirements around data availability and breach notification that create real liability exposure when systems fail.
Disaster recovery in the cloud for healthcare applications is more complex than it is for most other sectors. It’s not enough to have backups. The question is whether those backups have been tested, how long actual restoration takes under real conditions, whether the recovered data is consistent across integrated systems, and who is responsible for initiating recovery when the incident happens at 2am on a weekend.
Many smaller healthcare organizations operate under the assumption that their cloud vendor’s standard availability guarantees are equivalent to a disaster recovery plan. They aren’t. Vendor SLAs cover infrastructure uptime, not data integrity across application layers, and they don’t account for scenarios where the failure originates in user error, a corrupted update, or a third-party integration rather than in the cloud infrastructure itself.
Vendor Consolidation Feels Safe and Creates Concentration Risk
There’s a reasonable argument for reducing the number of software vendors a healthcare organization works with. Fewer vendors mean fewer contracts, fewer integrations, and theoretically simpler support. The problem is that consolidating around a single major vendor creates a different kind of risk: when that vendor has a problem, the entire operation has a problem simultaneously.
The 2024 Change Healthcare cyberattack illustrated this with painful clarity. Because such a large portion of US healthcare billing ran through a single clearinghouse, the attack didn’t just affect one organization. It disrupted claims processing for hospitals, physician groups, and pharmacies across the country for weeks.
Operational resilience in healthcare software requires thinking not just about whether each individual system works, but about what happens when it doesn’t, and whether the organization can continue functioning while it’s being fixed.
Article received via email
