Ending Epidural Failures

Published: October 21, 2026

By Lauren Liebhaber

7 min read

Epidurals are considered the gold standard for pain management during surgery and childbirth. They also fail in roughly 1 in 4 patients who receive them.

During an epidural, doctors thread a catheter—a small flexible tube that delivers numbing medication—into the epidural space near the spinal cord so pain signals are blocked during and after surgery. But the procedure only works if the catheter is placed in exactly the right location.

When epidurals fail, patients can wake up from invasive procedures in severe pain. Doctors may fill this gap in pain control with higher doses of opioids, which carry risks ranging from physical discomfort and slowed recovery to addiction.

More than a century has passed since the first epidural was performed on a battlefield in Morocco. Despite widespread advances in surgery and pain management since that time, the failure rate has remained stubbornly high and is widely considered an unavoidable limitation of a technically demanding, precise procedure.

But not everyone is willing to accept that the failure rate was unmovable. At Dartmouth Health’s Dartmouth Hitchcock Medical Center (DHMC) in Lebanon, New Hampshire, anesthesiologist Brian Sites, MD, MED ’96 RES ’02, director of the Acute Pain Medicine Service, and his team approached it as a fixable problem.

“We asked whether a 25% failure rate was acceptable and decided it wasn’t,” Sites says.

Over the past decade, the Acute Pain Medicine Service at DHMC has reduced epidural failure rates to around 1% to guide catheter placement. They didn’t invent a new piece of technology. Instead, Sites and his team found a novel use for X-ray video generated by fluoroscopy machines—equipment already found in most hospitals—resulting in a practical technique for placing a catheter in exactly the right spot; one that hospitals around the world can adopt to drastically improve patient pain management.

“Pain is so personal,” Sites says. “I love the idea that we can mitigate suffering with technical procedures and medical interventions so patients can recover better and be with their families, and not be fearful of surgery. This has been one of the most dramatic improvements in patient care in my career.”

Pain Is Personal. So Was the Solution.

Sites’ innovation traces its roots to his days as a resident at DHMC from 2001–2002. At the time, he and his colleagues were trying to improve a type of regional anesthesia called peripheral nerve blocks, which also had high failure rates.

Peripheral nerve blocks use injections to numb a specific nerve group in a particular part of the body before surgery, such as the shoulder, knee, or foot. Epidurals, by comparison, deliver numbing medication into the space around the spinal cord where it can block pain from larger regions of the body like the abdomen or pelvis.

The younger Sites had a conversation with his father, Vincent Sites, MD, MED ’67, a radiologist at Lahey Clinic in Burlington, Massachusetts, and fellow graduate of Dartmouth Medical School. His father asked whether he had ever considered using ultrasound imaging to guide nerve block placement.

“My dad didn’t know anything about regional anesthesia, but he knew that ultrasound could see some of these nerves,” Sites says of his father, who passed away in 2025. As a radiologist, “my dad influenced me to think about image guidance and not accept blind techniques.”

That idea, borne of a father-son chat, transformed how peripheral nerve blocks are placed. Within a decade of Sites and his colleagues developing the ultrasound-guided technique, it had become standard practice across the United States.

Later, when David Glass, then chair of anesthesiology at DHMC, asked the department a simple but catalyzing question—How can we get the benefits of epidurals that we know are possible if they’re not working to their full extent?—Sites drew on his earlier experience and conversation with his father.

If image guidance could dramatically improve one regional anesthesia procedure, Sites wondered if it could do the same for epidurals.

That led Sites and his team to explore fluoroscopy machines since other specialties already used them for spine-related procedures.

“We were already comfortable thinking about image guidance in related settings,” Sites says. “The question became whether we could do this safely, efficiently, and consistently, and whether it would actually improve reliability enough to justify the effort.”

Why Epidurals Fail

While epidurals may be most commonly known for their role in childbirth, they are also used in many major surgeries, including procedures involving the lungs, stomach, and intestines. When placed correctly, “epidurals provide profound pain control” and dramatically improve recovery, according to Sites, who is also a professor of anesthesiology and of orthopaedics at the Geisel School of Medicine at Dartmouth, and serves as the editor-in-chief of the journal Regional Anesthesia & Pain Medicine. Patients experience fewer complications and can even leave the hospital sooner.

But epidurals can fail in one of two ways, Sites says. The first and most common is a technical failure, which happens when the catheter is placed in the wrong spot. The second is a management failure: Even if the catheter is correctly positioned, it may move or stop working, requiring careful monitoring.

Traditionally, epidurals are placed using the “loss-of-resistance” technique, which relies on physical markers and the clinician’s sense of touch. With this technique, an anesthesiologist guides the needle by feel, pushing through thick layers of tissue until they feel a subtle “give” that signals entry into the less-dense epidural space, which lies between the vertebrae and the outer lining of the spinal cord.

With this so-called “blind” technique, it is more challenging to account for the physical variability from one patient to the next. Even a slightly misplaced catheter could mean it’s too far from the target nerves for the medication to fully block pain signals. But when it comes to epidurals, close isn’t good enough to deliver adequate pain relief.

“It’s either in the correct location or it isn’t,” Sites says. With traditional methods, and without the benefit of real-time imaging, clinicians don’t realize a catheter is misplaced right away. Anesthesiologists have to wait up to 30 minutes before testing whether the patient is experiencing the expected numbness and pain relief.

High failure rates have caused “a lack of enthusiasm for epidurals,” leading some doctors to abandon the procedure and rely more heavily on opioid pain management instead, Sites says, which comes with its own set of significant challenges, including, most notably, a high risk of opioid use disorder.

As a field, “we’ve kind of accepted that it’s okay to treat without an epidural because the failure rates are so high, instead of addressing the fundamental problem.”

According to Sites, successful epidurals can substantially reduce opioid utilization after major surgery. In communities disproportionately impacted by opioid use disorder, like those throughout New Hampshire, improving non-opioid pain management options offers a chance to reduce unnecessary opioid exposure.

How it Works

With fluoroscopy-guided placement, patients are brought into a dedicated procedure room within the main operating room and sedated. Fluoroscopy uses a continuous beam of X-rays to show movement in the body. This allows anesthesiologists to see exactly where they are positioning the catheter.

After placement, they inject contrast dye to confirm, via live video, that the numbing medication injected will spread exactly where intended. If the catheter isn’t correctly positioned, they can adjust it immediately, before adding the medication.

Sites and his team have been performing image-guided epidurals since 2011, when the outpatient surgery center at DHMC opened, including a dedicated space for the procedure.

“From there, we moved forward, conducted research, published results, and demonstrated improved success rates compared to blind techniques,” Sites says. “We never looked back.”

Building a New Standard of Care

Despite its effectiveness, fluoroscopy-guided epidural placement has not been widely adopted. “We are not aware of another center that has implemented it as standard practice in the way we have,” Sites says. The primary barrier is not technological availability, but organizational capacity, according to Sites.

Fluoroscopy machines, also known as C-arms, are already present in most hospitals. But implementing the technique requires dedicated staff, additional procedure time, and institutional logistical support.

“We built a team of nurses, fellows, and residents, and we dedicated anesthesiologists specifically to placing these catheters correctly,” Sites says. In a dedicated space in the outpatient surgery center, the Acute Pain Medicine Service at DHMC was able to use the equipment they already had and reorganize staffing and workflows to prioritize imaging-guided placement.

A Dramatic Difference in Recovery

Correct epidural placement can influence more than pain levels.

Patients with effective epidurals often breathe more easily after surgery, reducing their risk of pneumonia. They can move sooner, accelerating recovery and reducing the risk of complications such as blood clots.

Sites and his team at DHMC have found that epidurals can shorten hospital stays, reducing costs and complications compared with alternative pain management strategies.

In a 2025 study co-authored by Sites, study participants who received epidurals were 45% less likely to have prolonged hospital stays, with abdominal surgery patients leaving about 1.4 days sooner and experiencing fewer readmissions.

While image-guided epidurals are currently only performed at DHMC, Sites plans to expand this approach and training to all Dartmouth Health hospitals.

“When a patient has a functioning epidural, they’re comfortable, talking with family, breathing well, and recovering faster,” Sites says. “The difference is dramatic.”