More than half of US amputations are related to long-term health challenges such as diabetes and vascular disease. For those individuals, the earliest warning signs of limb loss can appear years before amputation becomes necessary. The red flags are often scattered throughout a person’s patient chart, in the form of drug prescriptions, specialist referrals, test results, and other records. Look backward after an amputation, these dots are easy to connect. But healthcare practitioners don’t always see the pattern until it’s too late.
Researchers in Denmark took a deep look at those patterns to see if they can help doctors recognize at-risk patients before symptoms turn serious. The study looked retroactively at the healthcare histories of more than 2,500 people who underwent lower extremity amputation due to diabetes and other vascular diseases. Those records were compared to equivalent histories from nearly 25,000 matched individuals who didn’t require amputation. The result, published last December in BMJ Open, is one of the most detailed portraits yet of the years leading up to dysvascular limb loss. And it contains at least one finding that will surprise even people who know this territory well.
Denmark is an ideal country for this study because its healthcare system maintains comprehensive health registries. Every prescription, checkup, clinical procedure, and outpatient visit flows into centralized databases linked to a unique patient ID number. That kind of detailed, lifelong tracking data is rare, and it allowed the researchers to look back many years before each amputation and ask: Are there any indicators that consistently presage limb loss? If so, how far back can we see them?
The researchers divided the pre-amputation window into three risk periods: Immediate (two years before surgery), Early (the two- to five-year range), and Long-term (five to ten years). They logged every healthcare data point across all three windows, along with a demographic markers such as age, sex, marital status, smoker status, and other indicators.
The amputees’ data was compared to equivalent data from two control groups. The first comprised the general Danish population, matched by age, sex, and location. The second, more targeted control group included people who already had diabetes or peripheral artery disease (PAD), and therefore faced elevated risk of limb loss. This cohort was matched by age, sex, and how long they’d been living with dysvascular conditions. The second control group enabled the researchers to zoom in on high-risk patients and ask what distinguishes those who eventually lose a limb from those who don’t.
The Antibiotic No One Was Watching
The single strongest predictor of amputation, across every time window and both control groups, had nothing to do with A1C thresholds, vascular scans, or circulatory tests. It was a prescription for a commonplace antibiotic called dicloxacillin. This workhorse medication is a go-to for primary-care doctors to treat patients with many kinds of routine infections. It’s particularly effective against Staphylococcus aureus, a bacterial species that commonly attacks soft tissue wounds such as foot ulcers.
Given that association, it’s not too surprising that patients who are prescribed dicloxacillin have an elevated risk of eventual limb loss. What’s startling is the strength of the connection. In the Long-term risk window—five to ten years before amputation—dysvascular amputees were nearly three times more likely to get a dicloxacillin prescription than members of the general population. And they were roughly twice as likely to need dicloxacillin than their close peers in control group #2, ie people diagnosed with diabetes or peripheral artery disease.
On average, people who eventually underwent major amputation received their first dicloxacillin prescription nearly eight years before the surgery. This warning light starts flashing on the dashboard nearly a decade before the situation becomes dire—plenty of time to intensify monitoring, change diet and exercise, and take other noninvasive steps to keep an unhealthy limb from deteriorating to the point of amputation. But nobody has recognized it as a warning light before.
The researchers are careful to separate correlation from causation. There are many possible reasons to prescribe dicloxacillin, and not all of them implicate limb loss. But a reasonable theory is that dicloxacillin prescriptions often reflect early management of chronic wounds and skin infections associated with damaged circulation. Now that this correlation is known, doctors can take quicker, more assertive action for patients with seemingly “minor” infections. If a patient’s infection requires dicloxacillin, it might portend major problems down the line, so preventive treatment should become a priority.
The dicloxacillin finding doesn’t stand alone. Opioid prescriptions are also strongly associated with later amputation across all three time windows, as is regular use of over-the-counter pain killers such as acetaminophen (ie, Tylenol). The researchers read these findings as signals of significant underlying pain—the kind associated with poor circulation and chronic wounds. Moving to the Immediate risk window, the study found that in the six months leading up to surgery, 73 percent of those who eventually had an amputation were on five or more medications. By comparison, just 30 percent of the general population control group fell into that category, along with just 41 percent of the more targeted control group (ie, people with diabetes or PAD).
Social circumstances also provided some telling clues to future limb-loss probability. People who eventually underwent amputation were significantly more likely to live alone compared to the general population. They were more often divorced or never married. Intriguingly, people who regularly visited a dentist were less likely to be in the amputation group across all three risk windows. That probably has less to do with dental health per se than with overall preventive healthcare habits. People who schedule routine teeth cleanings are more likely to schedule routine checks of their blood pressure, foot circulation, and other vascular risk factors.
These social indicators all align with research showing that limb-loss rates are disproportionately high among socioeconomically disadvantaged populations. People in these communities have less generous insurance, lower health literacy, lower trust in the medical system, and more difficulty navigating the healthcare system when they do seek treatment.
What This Means for Early Warning
Primary-care practitioners are often the first, and often the most trusted, point of contact for people who eventually lose a limb. By the time those patients reach a vascular surgeon, the options are often limited. Reliable triggers for earlier referral might give clinicians a broader range of treatment options and better odds of preventing amputation. The Danish study suggests that those triggers exist in data that’s readily available in everyone’s health records. Each individual data point might not sound an alarm by itself. But taken together, tracked over years, they blare out the incessant message that someone’s legs are under serious stress and need urgent, attentive care.
The authors are careful to note the limitations in their data. The Danish registries don’t include smoking status, body weight, alcohol use, or physical activity, all of which may influence amputation rates. In addition, their baseline numbers are from 2017 and 2018, which may not perfectly reflect current clinical patterns.
Even so, this kind of research strongly suggests that thousands of limbs might be saved simply by making better use of information that already exists. It doesn’t necessarily require whiz-bang technology, intrusive monitoring, or other high-cost, high-effort interventions. The story of limb loss unfolds over years, and the chapters are written in ordinary health records. The question is whether the people reading those records know how to follow the plot.
Paper is online in the December edition of BMJ Open.
