Blog
Hospital Information Security: Access, Audit Trails, Backups, Uptime
A plain-language guide for hospital owners and IT managers: how to approach access control, audit trails, backups and uptime, and what to ask vendors.
- mid
- hospital
If you run a hospital, four things carry most of your day-to-day information security: who can see and change records (access control), a trail of who did what (audit trails), backups you have actually restored from, and uptime planning for when systems fail. None of these needs exotic technology. Most failures come from weak habits: shared logins, backups nobody tests, and no plan for the day the system is down.
This article is a business explainer, not a technical manual. It covers what to look for, what it costs in effort and money, and where hospitals commonly go wrong.
Why this matters more for hospitals
A hospital system holds identity data, diagnoses, lab results, billing and insurance details. Unlike most businesses, a hospital also cannot simply stop work when systems go down: patients are already in beds.
Attackers know this. A Comparitech analysis reported by Perioperative Leader (formerly OR Manager) estimated that ransomware attacks on US healthcare organizations from 2018 to 2024 caused downtime averaging more than 17 days of recovery per incident. Those figures are estimates built from public reports, and they describe the US, not Indonesia. Treat them as an illustration of how long a bad incident can last, not a forecast for your hospital.
There is some better news. Sophos’ 2025 healthcare survey reported that 58% of surveyed healthcare organizations recovered within a week, up from 21% the year before. It is vendor research with its own sample, but it points to the same lesson: preparation shortens recovery.
1. Access control: the right people see the right things
Access control means each person can open only what their job needs. A billing clerk does not need to read clinical notes. A nurse on the surgical ward does not need every outpatient file.
What to look for:
- Individual accounts for everyone. No shared logins like “nurse1” used by a whole shift. If two people share a login, you cannot tell who did what.
- Roles, not one-off permissions. Define roles (registration, nurse, doctor, pharmacy, cashier, finance, IT admin) and assign people to roles. It is easier to review and easier to remove.
- Fast offboarding. When a doctor or staff member leaves, their access should end the same day. Many hospitals have a long tail of old accounts still active.
- Extra protection for admin accounts. Administrators can see and change almost everything, so they deserve stronger login steps than ordinary users.
- A plan for emergencies. Clinical staff sometimes need to reach a record they normally cannot. A “break the glass” approach lets them do so, but flags it for review afterward. Ask whether your system supports this.
Trade-off: stricter access slows people down. If the rules are too tight, staff find workarounds, such as sharing passwords, which is worse than the original problem. Design roles together with department heads, not only with IT.
2. Audit trails: knowing who did what, and when
An audit trail is a record of actions: who opened a patient file, who changed a bill, who deleted an entry, and when. It is the difference between “something looks wrong” and “we can see exactly what happened.”
Useful questions for your system or vendor:
- Are record views logged, or only edits?
- Can the log be edited or deleted by ordinary administrators?
- Can you search it by patient, by user and by date without calling the vendor?
- How long is it kept?
- Does anyone actually read it?
That last question matters most. A log nobody reviews only helps after something has gone wrong. A simple monthly routine works: pick a sample of VIP or staff-family records and check who accessed them, and review unusual after-hours activity or bulk exports.
Audit trails also help with disputes that have nothing to do with hackers: a billing disagreement, a claim that a record was altered, an internal investigation.
3. Backups: only real if you have restored from them
Everyone says they have backups. The question is whether you have ever restored from one under realistic conditions.
The principles are simple:
- Keep more than one copy, and keep at least one separated from your main network. Ransomware tries to find and encrypt or delete backups it can reach. CISA’s ransomware guide advises maintaining offline, encrypted backups of critical data and regularly testing their availability and integrity.
- Test restores on a schedule. Restore a real database to a spare machine every quarter, and time it. Write down how long it took.
- Decide how much data loss is tolerable. If you back up nightly, a failure at 4 pm loses a day of work. Is that acceptable for admissions and pharmacy? For many hospitals the answer is no, and more frequent backups cost more.
- Do not forget the rest. Imaging files, lab instruments and the billing system often sit outside the “main” backup.
Common cause of trouble: the backup server sits on the same network, with the same admin password, as everything else. When the network is compromised, so are the backups.
4. Uptime: plan for the day the system is down
Perfect uptime does not exist. What you control is how long an outage lasts and how well staff cope while it lasts.
Practical points:
- Know the cost of an hour of downtime for registration, pharmacy, lab and billing. That figure tells you how much redundancy is worth paying for.
- Hosting choice is a trade-off. An on-premise server gives control but needs power, cooling, spare parts and skilled staff. Cloud or managed hosting shifts much of that burden but adds a dependency on internet connectivity. Many Indonesian hospitals, especially outside Java, have unreliable links, so check your connection redundancy either way.
- Paper downtime procedures. Printed registration forms, medication charts and a process for entering data afterward. Rehearse it once a year, with real staff, during a quiet shift.
- Power. In practice, a large share of “system down” events starts with power or network equipment, not hackers.
- Updates and maintenance windows. Agree with your vendor when updates happen and how they are tested before reaching the live system.
Indonesian data protection law: a brief, careful note
Indonesia’s personal data protection law, Law No. 27 of 2022 (UU PDP), is relevant to hospitals because they hold sensitive personal data. According to law firm summaries, the law defines a failure of personal data protection in terms of confidentiality, integrity and availability, and requires a controller to give written notice to affected individuals and the relevant authority within 3 x 24 hours of such a failure. It also provided a two-year transition period.
Implementing regulations and the supervisory body were still being developed in the sources we reviewed, so details may have changed. Have your legal counsel verify the current obligations against the primary text and any implementing rules. No software, ours or anyone else’s, makes a hospital “compliant” on its own. What good access control, audit trails and backups do is give you the information you would need to respond.
What goes wrong most often
- Shared or never-expiring accounts.
- Backups that have never been restored.
- A single person who understands the system, with no documentation.
- Security treated as a one-time project rather than a routine.
- Staff never trained, so phishing emails still work.
- A vendor contract that does not say who is responsible for what during an incident.
Where to start
You do not need a large budget to make progress. A sensible first month:
- List every system that holds patient or financial data, and who administers each.
- Review user accounts against the current staff list.
- Do one full test restore and record the time.
- Write a one-page downtime procedure for registration and pharmacy.
- Ask your vendors what is logged, who can see the logs, and what their incident process is.
Costs vary widely with hospital size, existing infrastructure and hosting choice, so get written quotes rather than relying on rules of thumb. Much of the early work is process and discipline rather than purchases.
Talk it through
If you are choosing, replacing or customizing a hospital system and want a second opinion on roles, audit logging or recovery planning, we are happy to talk. Teknologia Solutions builds custom software and also offers our own clinic management system and hospital management system, which can integrate with Odoo. A no-charge conversation commits you to nothing, and we will tell you plainly if you do not need us.
Sources: Comparitech analysis via Perioperative Leader (formerly OR Manager) brief; Sophos State of Ransomware in Healthcare 2025 (summary); CISA #StopRansomware Guide; CMS Law-Now on Indonesia’s PDP Law. Statistics are from US-focused or vendor sources and are illustrative only. This article is general information, not legal advice; please verify figures and legal requirements against the primary sources.