
This afternoon I tried to change the password on my Apple account and couldn’t get in. After 30 minutes on the phone with Apple Support, the advisor said he realized Apple was having server connectivity issues. Around the same time, our consulting business’s email provider was struggling with network problems. The dreaded pop-up message read:
Error – An error occurred inside the server which prevented it from fulfilling the request.
Three hours later and no change. For a consulting firm, that means delayed emails and a few missed connections. It’s inconvenient, but nobody’s safety depends on it.
In a healthcare setting, the same kind of failure is a different matter. And this week, it’s happening to almost everyone.
A Week of Failures
Microsoft 365 went down on November 12 and stayed down into a second day, taking Outlook, Teams, and Copilot offline in dozens of countries. Microsoft traced the problem to a network capacity upgrade, not an attack. The next day, the FAA halted flights nationwide after a network problem. On Monday, a software bug at JPMorgan Chase delayed transactions for customers of several banks. Today, Apple’s iCloud sign-in, App Store, and Apple Pay had problems, and the National Security Agency reportedly had a partial outage on an internal network. As of this writing, officials and companies have blamed technical faults rather than cyberattacks, and reporting on the week shows no sign the failures are connected.
That last point is the lesson. None of these organizations needed a hacker to cause serious disruption. A routine upgrade or a software bug was enough.
What It Looked Like in Healthcare
During the Microsoft outage, Becker’s Hospital Review reported that University of Iowa Health Care lost email and voicemail across all its campuses and pointed patients to its patient portal. Henry Ford Health lost access to Outlook and Teams. Atlantic Health’s chief information officer said the disruption reached beyond email into patient care and the supply chain. In the United Kingdom, hospital staff reportedly fell back on paper records, and local reports described delayed prescriptions after GP surgeries and pharmacies lost email.
Many core clinical systems kept running, and some organizations reported little patient impact. But the communication channels around those systems, the way staff reach patients, pharmacies, and each other, went dark. Patients with urgent needs were told to call, and phone lines filled up.
HIPAA Requires Contingency Planning
Most people think of HIPAA as protecting privacy, but the Security Rule protects three things: the confidentiality, integrity, and availability of electronic protected health information (ePHI). Losing access to PHI when you need it is a security problem, even if no one steals anything.
That’s why the Security Rule requires a contingency plan. Under the Administrative Safeguards, you must have:
- A data backup plan (required)
- A disaster recovery plan (required)
- An emergency mode operation plan, meaning a way to keep protecting PHI and operating while systems are down (required)
- Testing and revision procedures (addressable)
- An applications and data criticality analysis, meaning a ranking of which systems matter most (addressable)
The Technical Safeguards also require emergency access procedures so authorized staff can reach ePHI in an emergency.
While the regulations currently classify the last two on the bullet list above as “addressable,” not “required,” this distinction is losing its importance. Addressable never meant “optional.” It meant you must evaluate each item for your organization and either implement it or document an equivalent alternative. However, as we’ve discussed before, the proposed Security Rule overhaul eliminates “addressable” and makes all the safeguards “required.”
The Privacy Rule matters also. It requires reasonable safeguards for PHI at all times, and your obligations don’t pause during an outage. Patients still have access rights, and PHI still has to be protected. In fact, outages create privacy risk. When the normal channel fails, staff improvise by texting patient details, using personal email, or opening unapproved apps. Your plan should say what’s allowed so improvisation doesn’t become a breach.
Plan for Vendor Failures, Not Just Disasters
When we wrote about contingency planning in 2020, we focused on fires, hurricanes, and floods. Those risks haven’t gone away, but this week’s failures show another category: the vendors we depend on. Email, cloud-based records, telehealth platforms, sign-in services, billing, and clearinghouses can all fail through no fault of your own. Ask which vendors would stop your operations if they went down, and what you would do in the meantime.
What a Good Outage Plan Covers
- Rank your critical systems. Decide which systems you can’t operate without, how long you could manage without each, and what to restore first.
- Set up backup communication channels. Keep a current phone list for staff, a way to reach patients if email fails, and a secure, approved fallback for messaging.
- Prepare paper backup procedures. Keep printed schedules, key forms, and medication information accessible, often called “downtime procedures,” along with steps for adding paper entries back into the record afterward.
- Make sure staff can reach records in an emergency. Document who can access what, and how, if the normal system is down.
- Back up your data and test the restore. Keep copies that an attack or outage can’t reach, and practice restoring from them.
- Prevent account lockouts from stopping you. My own morning is the example. Have more than one administrator, current recovery contacts, and backup sign-in methods stored securely, so one locked account doesn’t halt your operations.
- Know your vendors’ commitments. Keep an escalation contact list, review vendor availability commitments, and identify alternatives for essential services.
- Define safe workarounds. Spell out which tools staff may use when normal systems fail, and which they may not.
Test It, Train on It, Document It
A plan nobody has practiced is a guess. At least once a year, walk your team through a realistic scenario, such as “email is down for two days” or “our records system is unavailable.” Note what broke, fix it, and update the plan. Train your workforce so they know their roles, keep copies that leadership can access even if your systems are down, and document everything, including tests, training, and revisions.
The Failures Will Come Again
We can’t prevent Microsoft, Apple, or anyone else from having a bad day. We can decide in advance how we’ll keep patients safe and PHI protected when they do. For a step-by-step guide, see our post on how to create a contingency plan for HIPAA compliance, and use our HIPAA compliance checklist to find out whether your plan has gaps.
The HIPAA E-Tool® includes a contingency planning template that helps you build a plan tailored to your own organization, from backup and recovery to emergency operations, and documents your work so you can show it to a regulator if asked.

