Death by 1,000 Clicks: Where Electronic Health Records Went Wrong

Electronic health records were supposed to make health care safer, smarter, and less dependent on the mysterious handwriting of a physician who had clearly been chased by a bear while signing a prescription.

In many ways, they succeeded. Digital records can make medication lists easier to review, allow faster access to test results, support care coordination, and give patients more direct access to their own information. But somewhere between the promise of a modern medical record and the reality of 14 pop-up warnings, three password resets, and a note template the size of a small novel, the electronic health record became one of health care’s most frustrating tools.

The problem is not that clinicians dislike technology. Most doctors, nurses, pharmacists, and medical assistants use technology constantly and often appreciate what a well-designed system can do. The problem is that too many electronic health records, or EHRs, were built around billing rules, reporting requirements, legal protection, and institutional workflows rather than the natural rhythm of patient care.

The result is a system where clinicians can spend enormous amounts of time clicking, documenting, sorting messages, responding to alerts, and hunting for information that should be easy to find. It is not one catastrophic failure. It is death by 1,000 clicks.

The Original Promise of Electronic Health Records

Electronic health records did not begin as a villain in a hospital-themed science-fiction movie. They were introduced to solve real problems. Paper charts could be lost, misplaced, illegible, incomplete, or located in a filing cabinet that seemed to exist in another zip code.

In theory, EHR systems could improve health care by making patient information available when and where it was needed. A clinician could review medication lists, allergies, prior laboratory results, imaging reports, and specialist notes without waiting for a fax machine to wheeze back to life.

Digital systems also created opportunities for safety improvements. Medication alerts can flag potentially dangerous drug interactions. Order sets can guide clinicians through common treatment pathways. Electronic prescribing can reduce handwriting-related mistakes. Patient portals can help people review results, request refills, and communicate with care teams.

Those benefits are real. The mistake was assuming that digitizing health care automatically meant improving health care.

A computer system can make a good workflow better. It can also make a bad workflow faster, louder, and infinitely more capable of producing pop-up windows.

Where Electronic Health Records Went Wrong

1. They Were Designed Around Documentation, Not Conversation

At the center of modern clinical frustration is a simple mismatch: patients want a clinician’s attention, while the EHR often demands the clinician’s hands, eyes, and keyboard.

During a typical office visit, a doctor may need to listen to symptoms, ask follow-up questions, review medications, examine the patient, explain a diagnosis, order tests, document medical decision-making, update preventive care records, reconcile outside information, and complete billing-related documentation.

That is a lot to do while trying to maintain eye contact and avoid making the patient feel like they are competing with a monitor.

Many clinicians describe the uncomfortable experience of turning away from a patient to complete a note in real time. The doctor is physically present but mentally dividing attention between the person in the room and the digital checklist on the screen. It is difficult to create trust when a conversation keeps getting interrupted by a reminder that the patient may be due for a vaccine, a screening test, a quality measure, or an annual discussion about something they may have already discussed three times.

The EHR did not invent administrative burden. It simply gave administrative burden a keyboard shortcut.

2. Billing and Compliance Turned Notes Into Novels

Medical notes used to exist primarily to communicate clinical reasoning. A good note explained what happened, what the clinician found, what they thought was going on, and what the plan would be.

Over time, documentation became tied to billing, coding, quality reporting, legal defense, prior authorization, public health reporting, and performance measurement. Each requirement may have a reasonable purpose on its own. Together, they can create a chart note with the elegance of a tax return assembled during a power outage.

Templates, copied text, auto-populated fields, and checkboxes were meant to save time. Sometimes they do. But they can also create “note bloat,” where a record contains pages of repetitive information while the actual clinical story is buried somewhere between a normal review of systems and a paragraph copied forward from 2019.

A bloated note creates new risks. Important changes can be harder to spot. Clinicians reviewing the chart may need to sift through outdated details, duplicated medication lists, and auto-generated language that sounds impressive but says very little.

The best medical record is not the longest one. It is the one that helps the next person make the right decision.

3. Too Many Alerts Created Alert Fatigue

Clinical decision support is supposed to help clinicians avoid mistakes. For example, an EHR may warn about a possible drug interaction, an unusually high medication dose, a duplicate order, or a missing laboratory result.

That sounds sensible until the same clinician receives dozens or hundreds of alerts that are low-value, repetitive, poorly timed, or clinically irrelevant.

When every alert behaves as if it is a five-alarm fire, users learn to treat most alerts like a smoke detector that goes off every time someone makes toast. They click through them quickly because they have to keep moving.

This is known as alert fatigue. It is not a character flaw. It is a predictable human response to excessive noise.

The danger is obvious: when clinicians become accustomed to overriding alerts, an important warning may eventually blend into the background. Health IT safety experts have repeatedly emphasized that alert systems need better prioritization, better clinical relevance, and stronger human-factors design.

The answer is not necessarily more alerts. It is fewer alerts that matter.

4. The Inbox Became a Second Full-Time Job

For many clinicians, the most exhausting part of the EHR is not writing visit notes. It is the in-basket.

An EHR inbox can receive patient messages, refill requests, laboratory results, imaging reports, pharmacy questions, specialist updates, insurance forms, prior authorization requests, staff messages, hospital notifications, and automatically generated reminders.

Some messages require real clinical judgment. A patient reporting chest pain, worsening shortness of breath, or a severe medication reaction should not be handled like a routine password reset. But many systems send large volumes of administrative messages directly to physicians even when another trained team member could handle the first step.

This creates a strange version of modern medicine in which a doctor finishes a full day of patient care, then opens a digital mailbox that has somehow reproduced overnight.

Research on physician well-being has found meaningful connections between high volumes of EHR messaging, after-hours electronic work, and burnout. The issue is not that patients should lose access to care teams. The issue is that access needs a thoughtful system behind it.

Patients deserve timely answers. Clinicians deserve an inbox that does not follow them into dinner, vacations, and the last 12 minutes of a family movie.

5. Interoperability Has Often Been More Promise Than Reality

One of the biggest promises of digital health was interoperability: the ability for different systems to exchange and use patient information.

In practice, many clinicians still face fragmented records. A patient may have information scattered across hospital systems, specialty offices, pharmacies, imaging centers, laboratories, rehabilitation facilities, and insurance portals. Some records arrive as structured data. Others arrive as scanned PDFs. Others appear as a 97-page document with no useful search function and an attachment title like “FINAL_FINAL2_REVISED.pdf.”

When systems do not communicate well, clinicians spend valuable time reconstructing the patient story. They may have to re-enter medication lists, chase old test results, manually review outside records, or ask patients to repeat information they have already provided elsewhere.

Interoperability has improved in important ways, especially through national policy efforts and data-sharing standards. But exchanging data is not the same thing as making data useful. A clinician does not need another mountain of information. They need the right information, clearly displayed, at the right moment.

6. EHR Implementation Often Ignored Real Clinical Workflow

An EHR can be technically functional and still be miserable to use.

Many implementation failures happen because systems are configured without enough input from the people who actually use them. A hospital may focus heavily on launch dates, financial targets, regulatory compliance, and training completion. Those things matter. But they do not replace careful observation of how nurses administer medications, how physicians review results, how medical assistants prepare charts, or how pharmacists resolve medication questions.

When an EHR workflow adds several unnecessary clicks to a task completed hundreds of times per day, the burden compounds quickly. A few seconds lost during one prescription refill may not sound dramatic. Multiply that by thousands of refills, messages, orders, and notes, and the lost time becomes enormous.

This is why human-centered design matters. Health care software should be tested in the real environment where it will be used, not just demonstrated in a conference room where nobody is interrupted by a phone call, a patient emergency, or a nurse asking whether a medication order is correct.

The Patient Safety Problem Nobody Can Click Away

It would be tempting to frame EHR frustration as merely a workplace satisfaction issue. That would be a mistake.

Usability problems can affect patient safety. Poorly designed order-entry screens, confusing medication displays, delayed result follow-up, duplicate orders, wrong-patient selection, unclear alerts, and fragmented information can all create opportunities for error.

Consider a hypothetical but realistic scenario. A physician opens a chart to renew a medication, but the patient’s name appears in a similar-looking list near another patient with a nearly identical name. The system may technically provide warnings, yet if the layout is cluttered or the workflow is rushed, the risk remains.

Another example is the overlooked test result. A clinician may receive important laboratory or imaging information alongside dozens of routine notifications. If the system does not prioritize urgent findings clearly, a clinically meaningful result can get buried under a digital avalanche.

The solution is not to abandon electronic records. Paper charts had safety problems of their own. The solution is to treat EHR usability as a patient safety issue, not as a minor inconvenience for people who should simply “learn to use the system better.”

How Health Systems Can Fix the Click Burden

Measure the Work, Not Just the Login Time

Health systems should use EHR audit logs and workflow data to understand where time is going. It is not enough to know that a clinician logged in for several hours. Leaders need to know how much time is spent on documentation, chart review, message management, order entry, refill work, prior authorization, and after-hours tasks.

Measurement should lead to action. If a team discovers that physicians are spending large amounts of time routing messages that staff could safely manage, the solution is not another wellness webinar. The solution is redesigning the message workflow.

Build Team-Based Inbox Management

Not every message needs to land on a physician’s desk first.

Medical assistants, nurses, pharmacists, referral coordinators, and administrative staff can often handle appropriate parts of refill requests, scheduling, insurance paperwork, preventive care outreach, and routine patient questions. Clinicians should remain responsible for clinical decisions, but they should not be the universal switchboard operator for every task in the health system.

Clear protocols, training, escalation rules, and accountability are essential. Team-based care is not about pushing work away. It is about matching work to the right person.

Reduce Low-Value Documentation

Health systems, payers, regulators, and professional organizations should continue simplifying documentation requirements. The record should contain enough information to support safe care, clear communication, and appropriate reimbursement. It should not require clinicians to produce a novel every time they diagnose seasonal allergies.

Templates should be reviewed regularly. Auto-populated content should be limited when it adds clutter without clinical value. Copy-forward practices should be controlled so outdated information does not quietly become immortal.

Make Alerts Smarter and Fewer

Organizations should evaluate which alerts are routinely overridden and why. A warning that is ignored 98 percent of the time may be poorly designed, poorly targeted, or simply unnecessary.

High-risk alerts should stand out. Low-risk reminders should be less disruptive. The goal is to preserve clinician attention for situations where the system can genuinely prevent harm.

Demand Better Vendor Accountability

Hospitals and medical groups should treat usability as a core procurement requirement. Before signing a major EHR contract, organizations should ask practical questions:

• How many clicks does a common workflow require?
• Can clinicians quickly find urgent results and medication changes?
• How does the system support team inbox management?
• Can the organization customize alerts safely?
• What data can be exchanged with outside systems?
• How will the vendor respond to reported safety and usability problems?

Software vendors should be evaluated not only on features, but also on whether those features reduce cognitive burden. A system with 400 capabilities is not automatically better than one with 40 capabilities that people can actually use.

Use AI Carefully, Not as a Digital Bandage

Ambient documentation tools, voice recognition, message triage systems, and AI-generated draft responses may help reduce repetitive work. These tools can be useful when they are accurate, transparent, well-governed, and easy to review.

But artificial intelligence should not become an excuse to preserve broken workflows. A system that creates poor notes faster is still creating poor notes. A tool that drafts inbox responses is helpful only if clinicians can review those responses safely and if the underlying message volume is manageable.

The goal should be to eliminate unnecessary work, not merely automate the misery.

What Better Electronic Health Records Should Look Like

A better EHR would feel less like a compliance machine and more like a well-organized clinical assistant.

It would show the most relevant information first. It would make urgent results easy to identify. It would avoid interrupting clinicians for low-value alerts. It would support shared work across the care team. It would allow information to move between organizations without turning every outside record into a scavenger hunt.

Most importantly, it would protect the human relationship at the center of medicine.

A patient should not leave an appointment thinking, “The doctor seemed nice, but their computer had a lot going on.”

Technology should create more room for attention, trust, and thoughtful decision-making. When it does the opposite, it needs to be redesigned.

Experiences From the Front Lines of EHR Click Burden

The following experiences are composite, field-informed examples based on common themes reported by clinicians, health systems, patient-safety researchers, and professional organizations. They are not accounts of any single individual.

The Primary Care Physician Who Finishes the Day Twice

Picture a family physician seeing patients from morning through late afternoon. The schedule is packed: diabetes follow-ups, blood pressure checks, respiratory infections, medication questions, preventive visits, a patient with new fatigue, and another who needs forms completed for work.

During each appointment, the physician is trying to be present. But the EHR requires medication reconciliation, preventive care prompts, diagnosis coding, problem list updates, order placement, note completion, and follow-up instructions. The visit may last 20 minutes. The digital work often lasts longer.

At the end of the day, the physician’s waiting room is empty. The workday is not.

There are refill requests, laboratory results, patient portal messages, insurance forms, medication prior authorizations, and automated reminders. Some tasks are clinically urgent. Others are administrative. Most arrive in the same digital place, competing for the same limited attention.

The physician goes home, eats dinner, helps with homework, and later opens the laptop “just for 20 minutes.” Two hours pass. The EHR has quietly turned into an unpaid evening shift.

The Nurse Who Knows the Workaround Better Than the Workflow

A hospital nurse may use the EHR for medication administration, vital signs, care plans, patient education, handoffs, orders, documentation, and communication with the care team. Every step matters. Every step is also time-sensitive.

In an ideal world, the system supports safe care. In the real world, a nurse may encounter duplicate alerts, screens that hide the most useful information, or a workflow requiring multiple clicks to document something straightforward.

Over time, experienced staff learn shortcuts and workarounds. They know which screens contain the useful information, which alerts are usually noise, and which fields must be completed to prevent the system from refusing to move forward.

That experience is valuable, but it is also a warning sign. When safe care depends on memorizing system quirks, the technology is asking too much of the user.

The Specialist Who Receives a Flood of Outside Records

A specialist often receives referrals with years of prior records attached. In theory, this is helpful. In practice, the information may arrive as scanned documents, duplicate notes, fragmented medication lists, and reports from different systems that do not fit neatly together.

The specialist may spend the first part of a consultation reconstructing the patient’s story: Which medication is current? Was that scan done last year or last month? Did the patient already have the laboratory test? What did the prior specialist recommend?

The patient may assume the doctor already knows everything because “it is all in the computer.” The doctor may wish that were true.

This gap between available information and usable information is one of the biggest failures of fragmented electronic health records. Data can exist without being organized. It can be exchanged without being clinically useful. It can be technically interoperable while still being practically exhausting.

The Patient Who Thinks a Portal Message Is Instant Care

Patient portals have improved access in meaningful ways. Patients can request refills, ask questions, review results, and avoid unnecessary phone calls. That is a major improvement over waiting on hold while listening to the same 11 seconds of elevator music for 43 minutes.

But portals can also create unrealistic expectations. A patient may send a detailed message late at night and expect an immediate clinical response. Another may attach several photographs, list multiple symptoms, and ask for treatment, all within a message system designed for communication rather than a full diagnostic visit.

Care teams need clear standards for what portal messages can safely address, how quickly patients should expect a reply, and when a visit is needed. Otherwise, the portal becomes a hidden source of uncompensated clinical work, and clinicians become trapped between patient expectations and an inbox that never sleeps.

The Health System Leader Who Learns That “Go-Live” Is Not Success

From an executive perspective, launching a new EHR can look like a major achievement. The system went live. Users were trained. Orders are being entered. Billing is functioning. The dashboards are green.

But go-live is not the finish line. It is the beginning of a long process of listening, measuring, refining, and fixing.

Leaders who succeed treat clinicians as design partners, not as end users who need more training whenever something goes wrong. They review inbox burden, after-hours work, alert override patterns, turnaround time for results, and clinician feedback. They ask where the clicks are multiplying and whether each click serves patient care.

The healthiest EHR culture is one where a nurse, physician, pharmacist, or medical assistant can say, “This workflow is unsafe or wasteful,” and the organization responds with curiosity instead of defensiveness.

That is how electronic health records stop being a thousand small frustrations and start becoming what they were always supposed to be: tools that make it easier to care for people.

Note: This article is for general educational purposes and is not clinical, legal, regulatory, or health IT implementation advice.