IndustrialCyber

Health Management Academy details hospital cyberattacks, leading to care disruption and prolonged system outages


At least 65 cyberattacks disrupted patient care at U.S. hospitals and health systems between January 2022 and September 2026, affecting at least 305 hospitals across 36 states, the District of Columbia, Guam and the U.S. Virgin Islands, according to an analysis by The Health Management Academy. These attacks forced ambulance diversions in 26 incidents, canceled or postponed surgeries, treatments or appointments in 28, and prompted staff to switch to paper records in 58. Clinical systems remained offline for a median of 15 days among the 23 attacks with reported restoration dates, with five outages lasting at least a month.

The analysis, based on publicly reported incidents, highlights gaps in federal tracking of cyberattacks that disrupt hospital operations. The U.S. Department of Health and Human Services’ breach registry records incidents involving exposed patient data affecting at least 500 people, but does not capture operational consequences such as diverted ambulances, canceled procedures or closed clinics. The log counts 11 such attacks in 2022, 21 in 2023, 13 in 2024, 12 in 2025, and eight in 2026 through the end of September. 

Of 55 attacks in the analysis that began at least a year earlier, 12, or 22%, had no related registry filing. Meanwhile, healthcare organizations reported a record 776 data breaches to federal regulators in 2025, with hacking accounting for 81% of reported breaches between 2022 and 2025. 

“On February 19, 2026, the University of Mississippi Medical Center was hit by a cyberattack that took its Epic electronic health records system offline. Staff switched to paper charts, and all 35 of the center’s statewide clinics were forced to close until March 2, according to CNN and the Mississippi Free Press,” Andrew Reed, associate director at The Health Management Academy, wrote in a recent blog post. “The Mississippi attack was one of at least 65 cyberattacks that disrupted patient care at US hospitals and health systems between January 2022 and September 2026, according to a log compiled by The Health Management Academy from news coverage, trade publications, and hospital statements.” 

Reed added that the attacks reached at least 305 hospitals in 36 states, the District of Columbia, Guam, and the US Virgin Islands. 

The federal registry lists breaches of patient data affecting 500 or more people and has no field for diverted ambulances, canceled treatments, or closed clinics. Of the disruptive attacks that began more than a year ago, one in five has no entry in it.

“Ambulances were diverted to other hospitals in 26 of the 65 attacks, or 40%. Surgeries, treatments, or appointments were canceled or postponed in 28, or 43%, and clinics or hospital departments closed in 10. In 58 attacks, or 89%, electronic records or other clinical systems went offline and staff worked on paper,” Reed detailed. “Twenty-one of the attacks hit two or more hospitals at once. The largest, a ransomware attack on a national hospital system in May 2024, took electronic health records offline across roughly 140 hospitals.”

The count includes only attacks whose effects on patient care were reported publicly, so the full number is higher. Meanwhile, cybersecurity firm Emsisoft counted 46 hospital systems hit by ransomware in 2023 and 85 in 2024.

Reed said that care-disrupting attacks touched hospitals across 39 states and territories, with an attack on a multi-state health system counted in every state where it affected hospitals. Texas stood alone at the top with seven attacks, followed by California, Idaho, Illinois, and New York with five each. Florida, Georgia, Maryland, and Massachusetts each saw four, while Kentucky, Maine, Michigan, New Jersey, Oklahoma, and Wisconsin recorded three. Most of the remaining jurisdictions, including Alabama, Arizona, Kansas, Mississippi, Missouri, South Carolina, Tennessee, and the US Virgin Islands, had two or one.

The multi-state picture is just as telling. Texas had the most attacks on health systems spanning state lines, with three of its seven, while Florida, Illinois, Kansas, Oklahoma, Tennessee, and Wisconsin each had two. In Kansas and Tennessee, every attack involved a multi-state system. The same held for ten jurisdictions that recorded a single attack: Connecticut, the District of Columbia, Indiana, Iowa, Nebraska, New Hampshire, New Mexico, Pennsylvania, Rhode Island, and Washington. In those places, the one disruption was felt through a system that reached beyond their borders.

Some states looked very different. New York recorded five attacks, none of them on a multi-state system, and Idaho had five with only one crossing state lines, suggesting that locally contained systems carried much of the burden there. Arizona, Colorado, Guam, Louisiana, Mississippi, Missouri, Ohio, the US Virgin Islands, West Virginia, and Wyoming also saw no multi-state involvement at all.

In another incident, on April 6, 2026, the emergency room at Brockton Hospital in Massachusetts stopped accepting ambulances, which were routed to other hospitals instead. Among other disruptions, chemotherapy infusions were canceled the following day. Three days after the attack, on April 9, the Anubis ransomware group claimed it had stolen two terabytes of patient data. Brockton lifted its code black emergency status on April 15. The disruption made headlines with NBC Boston and HIPAA Journal.

For the 23 attacks in which hospitals reported a restoration date, clinical systems stayed down for a median of 15 days. Fourteen of those outages lasted at least two weeks, five lasted at least a month, and the longest ran 41 days. Patients admitted before an attack begins are also more likely to die in the hospital. University of Minnesota researchers found that in-hospital mortality rose by 34 to 38% for those patients, and that hospital volume fell by 17 to 24% in the first week.

This comes as healthcare organizations reported a record 776 data breaches to federal regulators in 2025, the third straight annual increase. Hacking accounted for 81% of the 2,878 breaches reported from 2022 through 2025.

Reed called on health systems to plan for outages that last weeks, not days. Staff fell back on paper records in 58 of the 65 attacks, and where hospitals reported a restoration date, clinical systems stayed down for a median of 15 days. Downtime plans built for that length of outage need to cover order sets, medication records, and lab and imaging requests that work without the electronic record, plus enough printed supplies and trained staff to sustain them. 

Some hospitals asked patients to bring printed medication lists, insurance cards, prescriptions, and physician orders to appointments, and those instructions can be drafted in advance as part of a downtime plan. Because systems stayed down for two weeks or longer in 14 of the 23 attacks with a reported restoration date, and for a month or more in five, recovery plans should also account for staffing paper workflows, manual billing and claims, and the backlog of postponed procedures once systems return.

Ambulance diversion, seen in 26 of the 65 attacks, pushes the strain onto neighboring hospitals. Researchers at the University of California San Diego, writing in Critical Care Explorations, found that survival with good neurological function after out-of-hospital cardiac arrest fell from 40.0 percent to 4.5 percent at two emergency departments while a neighboring four-hospital system was down after a 2021 ransomware attack. 

An earlier study of the same hospitals in JAMA Network Open found that ambulance arrivals rose from 1,741 to 2,354 and stroke code activations from 59 to 102 compared with the period before the attack. Regional agreements that spell out how diversion works during a cyberattack, including where stroke, trauma, and cardiac patients should go, give EMS agencies and receiving hospitals time to prepare for that surge.

The deeper problem is that nobody is counting. The federal registry tallies exposed patient records but has no field for diverted ambulances, canceled procedures, or days spent on paper, so no federal dataset shows how often cyberattacks interrupt care or for how long. Health systems and the companies that support them each hold a piece of that picture. A shared set of measures, such as days of clinical downtime, hours on diversion, and procedures postponed, would let organizations benchmark their readiness and recovery against peers. 

Reed pointed to the Health Management Academy’s Catalyst and Convening programs, which bring health system and industry leaders together to share data and practices, as one way for leaders to compare their downtime and diversion plans with those who have already lived through an attack. Health systems and their technology partners can also agree on what to track after an incident, so that the true cost is measured in days of downtime and hours on diversion rather than simply in the number of records exposed.



Source link