The first cardiology visit may not need a waiting room

A study of 5,735 newly referred cardiology patients found that beginning specialist care by telephone or video was not associated with more short-term emergency visits, hospitalisations or slower cardiovascular diagnosis than beginning with an in-person appointment.

November 19, 2026
Editorial
Virtual cardiology could expand how patients enter specialist care, but the evidence comes from an integrated health system able to connect remote assessment with diagnostics, records and in-person escalation.Studio Romantic / Shutterstock.com

IPM Take

Telemedicine has already proved that follow-up care can happen without putting every patient in a consulting room.

The harder question has always been what happens at the beginning.

A patient with palpitations, exertional symptoms or possible heart failure has not yet been diagnosed. A cardiologist may need an ECG, imaging, laboratory data or physical examination. The intuitive response is therefore that the first specialist encounter should happen in person and virtual care can come later.

New data from Kaiser Permanente Northern California complicate that assumption.

Among 5,735 adults referred from primary care with previously undiagnosed symptoms suggesting atrial fibrillation, coronary disease or heart failure, 70.9% began cardiology care through telephone or video. After statistical adjustment, those patients were no more likely to be hospitalised within 30 days and were not diagnosed more slowly than patients who attended in person. They actually had fewer repeat cardiology visits within 90 days.

That does not mean every first cardiology appointment should become virtual.

Patients selected their appointment modality, the study was observational, outcomes were short term and Kaiser Permanente provides a level of integrated records, testing and follow-up that many fragmented health systems do not.

That last point may be the most important.

Telemedicine is not simply a screen with a cardiologist on it. It works as part of a delivery system.

The policy opportunity is therefore not replacing cardiology clinics with video calls. It is using virtual entry points to expand specialist capacity while preserving rapid access to testing and in-person examination for patients who need them.

If that model holds outside integrated systems, the scarce resource in cardiology may increasingly be specialist expertise rather than specialist rooms.

Executive Summary

A cohort study published in JAMA Network Open evaluated 5,735 adults referred from primary care to Kaiser Permanente Northern California cardiology services between June 2020 and May 2022.

Patients had cardiovascular symptoms prompting a first specialist referral but did not have a recent established cardiovascular diagnosis or recent cardiology care.

The suspected conditions were:

56.4% atrial fibrillation, representing 3,238 patients;

28.2% coronary artery disease, representing 1,615 patients; and

15.4% heart failure, representing 882 patients.

Overall, 4,064 patients, or 70.9%, had their initial cardiology consultation by telephone or video, while 1,671 attended in person.

After adjustment for characteristics including age, sex, race and ethnicity, hypertension, comorbidity, socioeconomic measures, digital access and previous video-visit experience, telemedicine was associated with a 23% lower odds of having another cardiology appointment within 90 days compared with an initial in-person consultation.

There was no significant difference between telemedicine and in-person care in 30-day hospitalisation or in the likelihood of receiving a cardiovascular diagnosis within 60 days.

Across the full cohort, telemedicine did not increase subsequent emergency-department use. Among the subgroup referred for suspected coronary artery disease, virtual care was associated with lower odds of an emergency-department visit within 30 days, but that result should not be interpreted as evidence that telemedicine itself prevents emergency visits.

The study was observational rather than randomised. Patients chose between telemedicine and in-person appointments, creating the possibility that clinicians or patients selected in-person care when symptoms appeared more concerning.

Telephone and video consultations were also analysed together rather than separately.

The results therefore support telemedicine as a plausible first-contact option within an integrated cardiology pathway, not as evidence that remote assessment can replace physical examination or in-person specialist care for all patients.

Why it matters

  • HTA bodies: Telemedicine should increasingly be assessed as a care pathway rather than a standalone technology. Its value depends on whether remote assessment connects efficiently with diagnostics, electronic records, escalation and follow-up. Evaluation should therefore consider diagnostic timeliness, downstream utilisation, patient outcomes and total pathway costs rather than simply comparing the price of a video visit with an office appointment.
  • Payers: Virtual first-contact specialist care could reduce geographic and logistical barriers while allowing cardiology capacity to reach more patients. Payment models will need to avoid two opposite problems: reimbursing virtual encounters that simply generate an additional in-person visit, or underpaying virtual specialist assessment despite evidence that it can substitute for some conventional visits.
  • Industry / innovation partners: The opportunity lies less in building another video platform than in integrating virtual consultation with ECGs, ambulatory rhythm monitoring, imaging, laboratory data and decision support. The most useful digital cardiology systems will be those that help clinicians determine who can remain virtual and who needs rapid physical examination or testing.

Cardiology has a capacity problem that cannot be solved simply by building more consulting rooms.

Populations are ageing. Cardiovascular disease remains one of the largest sources of morbidity worldwide. More patients are being identified with atrial fibrillation, heart failure and coronary risk, while sophisticated imaging and increasingly complex preventive therapies are expanding what cardiologists can do once patients reach them.

The bottleneck is often reaching them in the first place.

Telemedicine offers an obvious response, but specialist cardiology has always presented a more difficult test than routine virtual follow-up.

Once a patient has an established diagnosis, much of a follow-up visit can revolve around symptoms, medication tolerance, blood-pressure readings and existing investigation results.

A first referral is different.

The cardiologist does not yet know what the diagnosis is.

Someone reporting palpitations may have atrial fibrillation, benign ectopy or something unrelated to the heart. Breathlessness could represent heart failure, lung disease, anaemia, deconditioning or several other conditions. Chest discomfort can range from benign musculoskeletal symptoms to obstructive coronary disease.

That uncertainty creates an intuitive argument for physical presence.

The new Kaiser Permanente study asks whether the assumption holds in practice.

Researchers examined adult members of Kaiser Permanente Northern California who completed a first cardiology referral between June 2020 and May 2022 after presenting to primary care with undifferentiated symptoms suggestive of heart failure, coronary artery disease or atrial fibrillation.

Patients with relevant recent cardiovascular diagnoses, cardiology appointments or hospitalisations were excluded, allowing the investigators to focus on people genuinely entering specialist cardiovascular care rather than established patients switching the format of a follow-up appointment.

The cohort included 5,735 adults with a mean age of 61.3 years.

More than half had symptoms suggesting atrial fibrillation, 28.2% were referred for possible coronary artery disease and 15.4% for possible heart failure.

Most did not begin with a conventional office appointment.

Just over seven in ten had their first cardiology consultation by either video or telephone.

That alone is important.

Telemedicine in this setting was not a small experimental service offered to a handful of selected patients. It had become the dominant route into cardiology care during the study period.

The clinical question was what happened next.

If virtual first visits missed important disease, one might expect patients to return more frequently, present to emergency departments, require hospitalisation or experience delays before cardiovascular diagnoses were established.

The study did not find that pattern.

After statistical adjustment, patients beginning through telemedicine had lower odds of requiring another cardiology appointment within 90 days.

There was no significant difference in 30-day hospitalisation.

There was also no evidence that patients receiving virtual consultations took longer to receive a cardiovascular diagnosis during the following 60 days.

That last result addresses one of the most important concerns around virtual specialist assessment.

Convenience means little if it delays diagnosis.

For the conditions examined, the researchers found no such signal.

Patients with suspected coronary artery disease who began virtually also had lower adjusted odds of visiting an emergency department within 30 days than those seen initially in person.

That finding should be interpreted cautiously.

It would be tempting to argue that telemedicine somehow prevented emergency care.

The study cannot establish that.

Patients were not randomly assigned to virtual or physical consultation, and the people who selected or were directed toward in-person appointments may have differed in ways that made subsequent emergency care more likely.

The same limitation affects the entire study.

Appointment type was chosen in routine practice.

Researchers statistically adjusted for numerous measurable differences, including age, sex, hypertension, overall comorbidity, previous experience with video visits, access to the patient portal, language, socioeconomic characteristics and broadband availability.

Adjustment improves comparability.

It does not create randomisation.

A clinician concerned about a patient’s presentation may have preferred an in-person appointment. A patient who felt particularly unwell may have made the same choice. Some of those factors may not be completely visible in electronic records.

The researchers appropriately describe the results as associations.

The findings therefore do not prove that virtual and in-person first cardiology consultations are clinically equivalent under every circumstance.

They provide reassurance that a telemedicine-first strategy did not produce an obvious short-term safety or efficiency penalty within this particular healthcare system.

And the system matters enormously.

Kaiser Permanente Northern California is an integrated delivery network.

The cardiologist conducting a virtual consultation is not operating in isolation from the rest of the patient’s care. Electronic records, primary care, specialist services, testing and follow-up exist within a connected organisation.

That infrastructure changes what a virtual consultation can accomplish.

If a cardiologist decides that palpitations require ambulatory rhythm monitoring, the diagnostic pathway can continue.

If possible heart failure requires echocardiography, the patient can be referred for imaging.

If symptoms suggest that physical examination or urgent evaluation is necessary, the virtual encounter can become a triage point rather than the endpoint of care.

This is fundamentally different from interpreting telemedicine as a substitute for diagnostics.

The screen does not need to perform the echocardiogram.

It needs to connect the right patient to the right test.

That distinction could become crucial as health systems rethink specialist access.

Traditional specialist care assumes that clinical expertise and physical location have to arrive together.

Every patient travels to the specialist because the specialist is located in the clinic.

Digital care separates those two resources.

Expertise can travel electronically while physical examination and diagnostic testing remain local.

That architecture could be particularly important for rural and underserved communities where cardiologist availability is limited.

A patient might obtain specialist assessment remotely while undergoing an ECG, laboratory testing or imaging closer to home.

But the Kaiser study does not establish that this model will perform equally well in such settings.

Its patients were members of one integrated Northern California system, and most had strong digital connectivity. More than 95% across the major referral groups had mobile portal access.

The infrastructure that makes telemedicine successful may therefore be precisely what is missing in populations where expanded specialist access is most urgently needed.

That is the implementation paradox.

Virtual medicine can overcome distance.

It cannot automatically overcome fragmented records, limited diagnostic capacity or poor broadband access.

A patient in a remote area may be able to reach a cardiologist by video while still waiting weeks for an echocardiogram.

In that scenario, the consultation has become faster without the pathway becoming faster.

Health policy therefore needs to stop evaluating telemedicine as if it were simply another appointment format.

The relevant unit is the complete diagnostic pathway.

How long does it take from primary-care concern to specialist review?

How quickly can necessary tests be completed?

How many encounters are required?

Does a virtual first visit reduce or increase duplicated care?

How often does the cardiologist decide that an in-person examination is necessary anyway?

Are high-risk patients escalated reliably?

And what happens to clinical outcomes?

The Kaiser study answers some of those questions.

Patients beginning virtually did not require more repeat specialist visits. In fact, adjusted repeat cardiology utilisation was lower.

Diagnosis was not slower.

Emergency and hospital utilisation did not increase overall.

Those are encouraging signals because one concern around telemedicine has been that virtual convenience might generate additional healthcare activity rather than substitute for it.

If every video call merely leads to an office visit, the health system has added an encounter rather than expanded capacity.

That was not the pattern observed here.

But more research is needed before concluding that virtual-first cardiology reduces costs.

The study did not provide a comprehensive economic evaluation, and repeat cardiology appointments are only one component of downstream utilisation.

Virtual patients might use different combinations of laboratory tests, rhythm monitoring or imaging.

The economic question is therefore more complex than counting appointments.

Patient preferences also matter.

Seventy percent using telemedicine demonstrates uptake, not necessarily preference.

The study period covered 2020 to 2022, when the COVID-19 pandemic profoundly altered both availability and willingness to attend healthcare settings in person. The authors acknowledge the limited post-pandemic observation period as a limitation.

Telemedicine use after emergency pandemic conditions may evolve differently.

Another issue is the decision to combine telephone and video visits.

A telephone consultation and a video consultation do not provide identical information.

Video can allow clinicians to observe appearance, breathing pattern or other visual clues that a telephone call cannot provide. Neither replaces auscultation, palpation or other elements of physical examination.

The researchers combined the modalities because the number of telephone visits was too small for stable separate analysis.

That was methodologically reasonable.

It means the study cannot establish whether telephone-only and video-first cardiology perform equally well.

Future implementation needs that distinction.

There is also no reason to turn the findings into an argument against physical examination.

The more intelligent use of telemedicine is selective.

Some patients will clearly need face-to-face evaluation from the start. Others may be appropriate for virtual specialist review followed by targeted testing. Still others may need no additional cardiology follow-up after the specialist reviews their symptoms and existing data.

A high-performing system needs to identify which is which.

That is where telemedicine starts becoming a capacity-management tool rather than a convenience feature.

The implications extend beyond cardiology.

Healthcare systems frequently allocate specialist appointments according to a binary model: the patient either receives an office appointment or remains outside specialist care.

Virtual consultation creates a third possibility.

Specialist expertise can be inserted earlier into the pathway without automatically consuming the same physical infrastructure as conventional outpatient care.

That could allow specialists to review more newly referred patients, redirect inappropriate referrals, order targeted investigations earlier or determine which patients genuinely need physical assessment.

For cardiology, where diagnostic information often comes from ECGs, rhythm monitors, laboratory testing and imaging, that model has particular potential.

But the threshold for evidence should remain high.

A dermatology image or medication review can often be assessed remotely with relatively little loss of clinically relevant information.

Cardiovascular symptoms can represent time-sensitive disease.

Any system that pushes virtual access simply because it is cheaper or easier risks missing the point.

The objective is not maximum telemedicine.

It is appropriate telemedicine.

That means the policy metric should not be the percentage of appointments delivered virtually.

A system could boast an impressive virtual-care rate while delivering poor care.

The more meaningful metrics are diagnostic delay, emergency utilisation, hospitalisation, treatment initiation, patient experience, equity and total resource use.

The new study performs well against several of those early measures.

Its broader lesson may be that the most important prerequisite for successful telecardiology is not video technology at all.

It is integration.

The cardiologist needs access to the clinical record. The patient needs somewhere to obtain testing. Primary and specialist care need to communicate. Patients who require physical examination need to move into in-person pathways without restarting the referral process.

When those pieces exist, the first specialist conversation may not need to happen in the same room.

That has important implications for workforce policy.

Healthcare systems frequently respond to specialist shortages by asking how to produce more specialists.

That remains necessary.

But capacity also depends on how existing specialist time is used.

If some new cardiology referrals can be assessed safely and efficiently without occupying a conventional consultation slot, physical clinic capacity can be concentrated on patients who actually require it.

The model could also reduce travel and time costs for patients, although the study did not directly quantify those benefits.

Those gains are particularly relevant for people managing employment, caregiving responsibilities, disability or long travel distances.

Yet virtual access can create new inequalities when digital literacy, language, devices or broadband become prerequisites for specialist care.

The Kaiser cohort had high levels of portal access, and most patients spoke English as their primary language.

A system attempting to replicate these findings cannot assume the same digital readiness.

Telephone options may help, but the evidence comparing telephone with video for first cardiology assessment remains limited.

That leads back to the most important limitation of the research.

The study tells us what happened inside a mature integrated system.

It does not prove that the same results will appear when the model is exported.

Fragmented systems may need to build the infrastructure first.

Otherwise telemedicine risks becoming a digital front door leading into the same disconnected hallway.

The policy opportunity is therefore larger than virtual visits.

It is redesigning specialist referral around what the patient needs rather than where the specialist traditionally sits.

The office remains essential.

It may simply no longer need to be the compulsory first stop.

If virtual cardiology can identify which patients need testing, which require in-person assessment and which can be managed without another specialist visit, the question is no longer whether telemedicine can replace the clinic. It is whether every patient ever needed to start there.

Source & Evidence