Knowledge and money, not just willingness, determine who uses telemedicine.
In the rural corridors of Brandenburg, where cardiologists are scarce and patients travel far for care, a German research team has asked an old question in a new way: what separates the doctor who embraces telemedicine from the one who does not? Surveying 112 healthcare professionals and applying both classical statistics and machine learning, they found that knowledge, specialty, and gender shape adoption—while inadequate reimbursement stands as the most stubborn wall between digital promise and clinical reality. The study reminds us that technology does not spread by its own logic; it spreads through the conditions human institutions either build or neglect.
- Rural cardiology deserts in Germany make telemedicine not a convenience but a lifeline—yet fewer than 60% of surveyed professionals have adopted it.
- Two independent analytical methods converged on the same warning: the single greatest obstacle to adoption is not ignorance or resistance, but a payment system that undervalues remote care.
- Women and cardiologists led adoption, while skepticism about telemedicine's patient benefits—likely rooted in knowledge gaps rather than clinical evidence—held others back.
- A machine learning model predicted telemedicine use with 79% accuracy, giving policymakers a rare precision tool to identify who needs support and where interventions will land.
- Systematic evidence already shows telemedicine reduces mortality and costs in cardiology, yet fragmented reimbursement rules continue to override that evidence at the practice level.
- The path forward is now mapped: structured training programs and reimbursement reform are the two levers most likely to move the needle toward sustainable digital integration.
A German research team set out to understand why some heart doctors use telemedicine while others do not—a question with real stakes in a country where rural regions like Brandenburg face severe cardiologist shortages and patients sometimes travel hours for specialized care.
The team surveyed 112 healthcare professionals across Brandenburg and Berlin who treat cardiac patients. Fifty-seven percent had already adopted telemedicine; the rest had not. To understand what divided them, the researchers applied two distinct analytical frameworks: a Bayesian statistical method that accounts for model uncertainty, and an XGBoost machine learning model optimized for prediction. The convergence of their findings across both approaches gave the results unusual credibility.
Three factors consistently predicted adoption: strong knowledge of telemedicine, being a cardiologist rather than an internist or general practitioner, and being a woman. The machine learning model achieved 79% accuracy in predicting which doctors would use telemedicine—a strong performance suggesting these factors reflect genuine drivers rather than statistical noise.
But the analysis also surfaced a structural problem that no amount of training can fix alone. Both methods identified inadequate reimbursement as the single strongest barrier. German telemedicine services remain poorly compensated compared to in-person visits, and despite recent reforms, payment is fragmented across the system. More than half of those surveyed cited this as a major obstacle. Doctors who doubted telemedicine's patient benefits were also less likely to adopt it—though systematic reviews show it reduces mortality and costs in cardiology, particularly for chronic conditions like heart failure.
The study's value lies not in surprising conclusions but in quantifying them with precision. Policymakers now have actionable evidence: targeted training programs can close knowledge gaps, and reimbursement reform can remove the structural barrier that knowledge alone cannot overcome. In a healthcare system facing aging populations and rising demand for cardiac care, those two levers may determine whether telemedicine fulfills its promise or stalls at the threshold.
A team of researchers in Germany set out to answer a deceptively simple question: why do some heart doctors use telemedicine while others do not? The answer, it turns out, matters more than it might seem—especially in a country where rural regions like Brandenburg face severe shortages of cardiologists and patients sometimes travel hours for specialized care.
The researchers surveyed 112 healthcare professionals across Brandenburg and Berlin who treat heart patients. Fifty-seven percent of them—64 doctors—had already adopted telemedicine. The other 43 percent had not. To understand what separated the two groups, the team deployed two different analytical approaches: one designed to identify stable patterns of association, the other optimized to predict which doctors would most likely use digital tools. The dual method proved revealing, surfacing both what drives adoption and who is most likely to embrace it.
The findings converged on three core factors. Doctors with strong knowledge of telemedicine were significantly more likely to use it. Being a cardiologist—as opposed to an internist or general practitioner—predicted adoption. And women were more likely than men to integrate telemedicine into their practice. These three factors held up across both statistical frameworks, suggesting they represent genuine drivers rather than statistical artifacts. The machine learning model, which used a technique called XGBoost to identify patterns in the data, achieved 79 percent accuracy in predicting which doctors would use telemedicine, with an area under the receiver operating characteristic curve of 0.88—a strong performance that suggests the identified factors genuinely matter.
But the analysis also revealed a critical structural barrier. Both the statistical and machine learning approaches identified lack of reimbursement as the single strongest obstacle to adoption. German healthcare professionals reported that telemedicine services are not adequately compensated compared to in-person visits. This is not a knowledge problem or an attitude problem. It is a money problem. More than half of the cardiologists, internists, and general practitioners surveyed cited inadequate reimbursement as a major barrier, and despite recent reforms, payment for telemedical services remains fragmented and limited across the country.
The researchers also found that doctors who harbored doubts about telemedicine's patient benefits were less likely to use it—though the evidence suggests these doubts may reflect knowledge gaps rather than clinical reality. Systematic reviews have shown that telemedicine in cardiology reduces mortality, lowers healthcare costs, and improves patient satisfaction, particularly for chronic conditions like heart failure and arrhythmias. Yet many practitioners remain skeptical, especially regarding acute care scenarios. The questionnaire did not distinguish between different types of acute situations—diagnostic consultation versus emergency intervention, for instance—so the skepticism may reflect genuine uncertainty about which acute scenarios are appropriate for remote care rather than blanket opposition.
What makes this study notable is not that its conclusions feel intuitive. Of course doctors with better knowledge of a technology are more likely to use it. Of course payment matters. Rather, the strength lies in quantifying these associations with precision, using advanced statistical methods to move beyond general impressions toward actionable evidence. The researchers applied Bayesian Model Averaging, which accounts for uncertainty about which variables matter most by averaging across multiple plausible models rather than selecting a single "best" one. They then applied machine learning to capture nonlinear relationships that traditional regression might miss. The convergence of findings across these distinct approaches lends credibility to the results.
The implications are concrete. Policymakers in Germany and elsewhere could prioritize structured training programs to build telemedicine knowledge among healthcare professionals. They could reform reimbursement systems to align payment for remote care with in-person visits. They could target interventions toward groups showing lower adoption rates. And they could use similar analytical frameworks to identify leverage points in other medical specialties. The study suggests that telemedicine adoption is not inevitable—it depends on deliberate choices about education, payment, and support. In a healthcare system facing aging populations and growing demand for cardiology services, those choices matter.
Notable Quotes
Healthcare professionals who had strong knowledge of telemedicine and who perceived it as relevant for patient care were much more likely to adopt it.— Study authors
Lack of remuneration emerged as a strong barrier in both models, underscoring the need to develop reimbursement frameworks that directly address this structural obstacle.— Study authors
The Hearth Conversation Another angle on the story
Why does it matter whether doctors use telemedicine? Isn't that just a preference?
It matters because of geography and scarcity. In rural Brandenburg, patients sometimes travel hours to see a cardiologist. If doctors don't use telemedicine, those patients stay isolated. And there aren't enough cardiologists to go around.
So the study is really about access—making sure people in remote areas can get care?
Partly, yes. But it's also about understanding what actually drives adoption. The researchers found that knowledge and attitudes matter, but so does money. A doctor might want to use telemedicine, might even know how, but if they're not paid for it, they won't.
That's striking. The machine learning model could predict adoption with 79 percent accuracy. What does that tell us?
It tells us the factors are real and measurable. It's not random. If you know a doctor's knowledge level, their specialty, their gender, and their views on reimbursement, you can make a pretty good guess about whether they use telemedicine. That's useful for targeting interventions.
The study mentions that doctors were skeptical about telemedicine in acute situations. But the evidence suggests it works. Why the gap?
Probably because the questionnaire didn't distinguish between different types of acute care. A diagnostic consultation is different from an emergency requiring immediate physical intervention. Doctors may be right to be cautious about some acute scenarios while being wrong about others. The skepticism might reflect uncertainty rather than knowledge.
If you were a policymaker, what would you do with these findings?
Three things. First, invest in training—knowledge is a consistent predictor. Second, fix reimbursement immediately—it's the strongest barrier in both analyses. Third, don't assume adoption will happen naturally. Use these insights to target support where it's needed most.