CME Opportunities

We are proud to offer convenient options for earning CME credits.

See all CME programs

Patient Safety
and Risk Resources


Show all


Show all


Show all

Care Delivery Models

Show all


Anesthesiology: A Claims Review Panel on Epidural Anesthesia

To address the rising number of claims related to epidural anesthesia, The Doctors Company hosted a panel on that topic.

The panel reviewed 12 cases and cited the following main categories of maloccurrence or injury: 1) nerve damage; 2) volume replacement; 3) informed consent; 4) combination anesthetics; and 5) communication with patients. The following summaries describe the categories and the cases listed within those categories.

1 Nerve Damage

Nerve injuries, including numbness and palsies, were seen in several obstetrics and gynecology cases. Ultimately, some cases were found to be inconsistent with nerve root damage from epidural placement. Examples included saphenous or peroneal nerve compression from lithotomy stirrups, femoral nerve damage from lower abdominal self-retaining retractors, and obturator nerve compression against the pelvic brim during delivery.

Surgeons, obstetricians, and even neurologists tended to assume that all nerve damage after an epidural was anesthesia related. The panelists determined, however, that physicians should refrain from documenting causation until appropriate examinations and tests are performed. Careful documentation of patient positioning also enhances defensibility in such cases, as well as the interspace level entered and the absence of paresthesias.

Case No. 1A

The anesthesiologist administered a labor epidural to a 31-year-old female. The baby was delivered vaginally without difficulty, but postpartum the patient complained of paresthesias of both lower extremities. Multiple charting entries referred to "post epidural neuropathy." A neurologist diagnosed bilateral femoral nerve injury, likely secondary to the positioning of the patient’s legs during delivery. The use and type of stirrups could not be found in the charting. Three weeks after delivery, the patient was still under the neurologist’s care, but had greatly improved.

The panel’s evaluation of this case suggests that as an anesthesiologist, you would be well advised to seek consultations by neurologists, who could assist in evaluating nerve injuries from obstetric and gynecology cases. Review of the medical records in this case indicated that appropriate care had been provided during administration of the epidural anesthesia. It was difficult, however, for the panel to ascertain if the femoral compression occurred during delivery. Defensibility of a claim improves if anesthesiologists carefully document positioning, including the use of stirrups.

Case No. 1B

The anesthesiologist administered an epidural with chlorprocaine along with nitrous oxide to a 36-year-old female undergoing an abdominal hysterectomy with bilateral salpingooophorectomy for leiomyoma. The patient’s history was significant for a positive family history of malignant hyperthermia.

Epidural insertion, induction, and intubation were uneventful. Approximately one hour into the procedure, massive bleeding was encountered because of extensive adhesions. The gynecologists worked deep in the pelvis to stop the bleeding and ultimately called for a general surgeon. The general surgeon placed additional abdominal retractors, and the bleeding was stopped. The patient remained stable throughout the surgery, despite an estimated 3,500cc blood loss.

Postoperatively, the patient developed a left foot drop and had paresthesia of the buttocks and vaginal area. An MRI ruled out an epidural hematoma. Nerve conduction studies localized the lesion as a peripheral nerve injury distal to the epidural space.

The panelists agreed that the preoperative anesthesia evaluation and postoperative follow-up care had been excellent. Documentation of the patient’s position during surgery, however, could not be found anywhere in the charting. That made it difficult to determine the actual etiology of the nerve injury. Possibilities included improper stirrup positioning or traction from the use of self-retaining retractors.

2 Volume Replacement

Several cases involved anesthesiologists either failing to recognize preoperative volume depletion or inadequately replacing intraoperative fluid loss. The vasodilation induced by epidural anesthesia is poorly tolerated by the hypovolemic patient. Preoperative evaluation should include the volume status of the patient. Signs of volume depletion may include elevated or rising hematocrit because of hemoconcentration, tachycardia, and low urine output. Preeclamptic patients may be volume-depleted intravascularly even if hypertensive and edematous.

As an anesthesiologist, you should consider using central venous pressure measurement with an epidural if volume status is uncertain; you should also include an estimate of intraoperative blood loss in the anesthesia record. In cases with expected large blood losses, you should carefully weigh indications for epidural anesthesia.

Some charts in the cases reviewed did not have adequate documentation of a preoperative anesthesia evaluation. Yet standards set by the American Society of Anesthesiologists clearly acknowledge the need for such documentation of a patient’s medical, medication, and anesthetic histories as factors in providing quality care.

Case No. 2A

The anesthesiologist insured by The Doctors Company administered a continuous epidural, and Pitocin was started in a young female primipara in labor who had a blood pressure of 138/96. Shortly thereafter, fetal heart tones decreased to 70 for three minutes, and delivery was accomplished with low forceps. Thick meconium was present at delivery. In the nursery, the newborn was found to be meconium stained, with hypoglycemia and apneic episodes. She was later diagnosed with cerebral palsy. The plaintiff’s attorney alleged the insured was negligent in his choice of continuous epidural anesthetic for a patient with preeclampsia.

The panelists determined that epidural anesthesia is acceptable in the preeclamptic patient, but that special attention must be given to adequate volume replacement. Minimal intravenous fluids had been given to this patient before and after epidural placement, and charting was not adequate to indicate that fluid volume had been properly assessed.

Case No. 2B

A 62-year-old male with renal disease who was on dialysis presented for an open reduction of a hip fracture. The anesthesiologist performed a preoperative evaluation and cleared the patient for surgery. Anesthesia plans included an epidural anesthetic combined with light general anesthesia. Following the epidural, the patient remained on his side for five minutes, during which time his blood pressure dropped to 106/62. The patient had received 200cc of albumen when he was turned supine for the general anesthetic induction. Immediately after moving, the monitoring equipment began alarming for nondetectable blood pressure and weak pulse. A Code Blue was initiated, and the patient was resuscitated. He remained in a vegetative state and later expired.

Review of this case found no mention of the patient’s preoperative volume status. A significant elevation in his hematocrit following dialysis, however, indicated possible hemoconcentration. The most likely etiology of the arrest, preceded by falling blood pressure, was believed by the panel members to be relative hypovolemia secondary to vasodilatation from the epidural. A total of 200cc of colloid probably would not have been adequate to counteract that event. The panel members reasoned that, although the choice of anesthetic technique may have been appropriate, more prompt treatment of the falling blood pressure with intravenous hydration and pressors might have improved the outcome.

Case No. 2C

Another case involved a 50-year-old female with a diagnosis of rectal carcinoma. The patient underwent a low anterior resection of the colon. The anesthesiologist administered a combined epidural and general anesthetic. The anesthesia record documented no estimated blood loss, but did show that the patient received approximately 3,000cc of crystalloid. In the recovery room, the patient’s blood pressure fell, with systolic in the 90s. She was given 250cc of albumen, with improvement in the blood pressure. In the three hours she was in recovery, she produced only 80cc of urine. The patient was given a final epidural dose of bupivacaine and was started on intravenous patient-controlled morphine. She was placed on a maintenance intravenous infusion of 125cc/hour by the surgeon. Eighteen hours postoperatively, the patient arrested and expired.

Panel members emphasized the importance of documenting blood loss. Although the exact etiology of the arrest was difficult to determine, the patient may have become progressively more hypovolemic after leaving the recovery room. Consistent central venous pressure readings would have been useful. Clear and legible PACU orders are essential.

3 Informed Consent

Complications such as headache, nerve damage, and infection should be discussed with the patient and clearly documented during the informed-consent process. Patients should understand the procedure and decide for themselves whether they wish to accept the risks. Documentation of that discussion by the anesthesiologist provides a valuable defense tool to support an informed decision. From the standpoint of a claim or a lawsuit, if it is not in the record, it did not happen. A simple consent that may be used or adapted is: "Epidural procedures and risks were discussed with the patient. He/she understands that complications are rare, but can include headaches, backaches, nerve damage, bleeding, infection, seizures, and death. All questions were answered. The patient understands and accepts."

Case No. 3A

A 54-year-old female complained of low back pain to an anesthesiologist. The patient’s history was significant because of an occipital neuralgia diagnosis that had been treated with epidural nerve blocks. An MRI revealed bulging discs at L1 and L2. She was diagnosed with lumbosacral radiculitis and treated with lumbar epidural steroid blocks with good response. Two years later, she presented again, complaining of low back pain after a bicycle accident. On the basis of a recent MRI, the anesthesiologist recommended trigger-point and facet-block injections with physiotherapy.

The patient’s low back problems continued for one year, at which time the insured performed cryoneurolysis. The patient alleged she had a debilitating back condition and that the insured caused it.

Panel review of the case indicated that an appropriate treatment plan had been developed for the patient. Adequate documentation of the informed-consent process with the patient, however, was not found in the records. Panel members emphasized the critical need to document discussions with patients. Documentation should include risks and benefits as described to the patient during the informed-consent process. The patient’s understanding and acknowledgment of that information also should be documented.

Case No. 3B

A 39-year-old married female was admitted for induction and delivery of her third child. An induction was carried out at 39 weeks to avoid shoulder dystocia, which had occurred with the delivery of her first two children. A lumbar epidural catheter was placed uneventfully. Analgesia was obtained with bupivacaine and fentanyl. The patient noted diffuse tingling in both hands before entering the delivery room.

When taken to the delivery room, the patient was completely dilated; because of the density of the epidural block, however, she was unable to push well and the epidural pump was stopped. Improved motor function in the legs returned, and a healthy infant was delivered vaginally. Total infusion time was two hours and 35 minutes.

In the next two hours postpartum, numbness and weakness of the right leg developed and subsided. Additionally, the patient noted continued numbness and weakness in both hands—more prominently on the right side. The patient remained in the hospital for 11 days, with diagnostic testing that included cerebrospinal fluid analysis, thyroid studies, CAT scan of the pelvis, and MRI of the lumbar and cervical spine.

The test results were all normal. Somatosensory evoked potentials were performed, revealing a marked conduction delay between the popliteal fossa and T12 and suggesting a source at the level of the lumbar plexus or lumbar nerve routes. There were no signs of central conduction delay. Physical therapy was initiated; some reflex, motor, and sensory functions returned to the left leg by time of discharge. The patient was unable to ambulate, however, even when aided by a walker.

The panel’s review of the medical records indicated that appropriate care had been provided, and a timely request for evaluation and consultation had been obtained. The consensus was that the patient’s condition did not fit any well-defined syndrome and that it was not possible to assign a specific etiology. Panel members believed that this case could have been defended better if the informed consent had been well documented. Although sometimes difficult in obstetrics, anesthesiologists should allow the necessary time required to provide patients with information about the risks and possible complications of epidural anesthesia.

Case No. 3C

A 69-year-old male patient was admitted, with complaints of a right lower extremity claudication, for revascularization of the right lower extremity with a femoral-femoral crossover graft. An epidural catheter was inserted by a student CRNA, with test doses and aspiration negative. The final dose was given, followed by loss of consciousness and decreased blood pressure. Intubation, IV fluids, and medications were administered. The surgery was canceled and the patient was transferred to the surgical ICU, where he was maintained on the ventilator for approximately six hours. The patient regained consciousness, was extubated, and slowly recovered sensation in the lower extremity.

The patient was awake, alert, and had no complaints the following day. He had equal strength in both lower extremities with intact sensation. The plan of treatment was continued observation and rescheduling of surgery. One day postevent, the patient was taken to the operating room, where a continuous spinal anesthetic was administered with an intrathecal catheter in place. The anesthesia and surgery were uneventful, and the patient was returned to the surgical ICU.

Postoperatively, the patient complained of numbness of the lower extremities. A neurologist evaluated the patient and indicated evidence of numbness in the legs bilaterally, but the patient was able to move all extremities. MRI was negative for any evidence of epidural hematoma, and an MRI of the brain did not show any acute infarctions.

Upon discharge, approximately one month postadmission, the patient experienced bladder and bowel incontinence and paraparesis of the lower extremities. As a result, he suffered from neurological deficits as well as cognitive problems.

Panelists were generally supportive of the care provided during the first epidural complication. The panelists were somewhat surprised, however, that a continuous spinal had been chosen for the surgical procedure after a total spinal anesthetic had occurred the day before. The patient did not seem to have any contraindications to general anesthesia. Documentation did not clearly indicate that the patient had been adequately informed of the risks with all possible anesthetics.

4 Combination Anesthetics

Four of the cases involved combined epidural and general anesthetics. As this technique has a higher complication rate, panel members noted that patients should be apprised of the risks in advance. The informed-consent process should include the possibility that the surgical procedure could be performed under a combination of general and regional block or with the use of either type of anesthesia as a separate route.

Placing the epidural block while the patient is under general anesthesia risks losing the warning signs of possible nerve trauma that pain and paresthesia provide. Anesthesiologists planning epidurals at levels above the termination of the spinal cord (e.g., thoracic blocks) should consider performing the procedure while the patient is awake.

Preoperative consultation with the surgeon is essential to determine a patient’s need for postoperative pain control. That must also be included in the informed-consent process.

Case No. 4A

A 51-year-old male with a history of rectal carcinoma was scheduled for a pneumonectomy for metastatic carcinoma to the left lung. While the patient was under general anesthesia, a thoracic epidural was placed. The surgery was uneventful except for mild hypotension treated with ephedrine and IV fluids. The patient was taken to the PACU, where an epidural infusion was started. He was awake and comfortable after surgery until 11 hours postoperatively, when he complained of pain and leg numbness and was unable to move. The epidural was discontinued and intravenous patient-controlled morphine was begun. Approximately three hours later, the patient had normal sensation with improved strength, and the epidural catheter was removed.

On discharge, the patient complained of leg numbness and weakness and walked with a cane. A neurological evaluation showed upper motor neuron signs in one leg, with a sensory deficit in both legs and the left thoracic region.

The panel members’ review of this case concluded that if there had been nerve trauma at the time of inserting the thoracic epidural, the patient would likely have demonstrated considerable pain if he were awake. Because the patient was already under general anesthesia, immediate feedback about pain or paresthesias was not possible. The necessity of an informed-consent process and its documentation in such cases is critical. The patient must have the opportunity to decide if he accepts the higher risks inherent in a combined technique. Panelists also pointed out the need for good communication with surgeons preoperatively as to whether there will be a request for an epidural for postoperative pain relief.

5 Communication with Patients

Patients with preexisting pain or neurological disorders who require epidural procedures for chronic pain can present special problems for anesthesiologists. Such patients may obtain greater satisfaction with pain control when they have a thorough understanding of the procedure. In the process of patient education, you have an opportunity to enhance physicianpatient rapport and to minimize the element of surprise that could precipitate a lawsuit.

All of your conversations with patients must be documented in the medical record. Although some complications are not preventable, a well-documented record will be your best defense.

Case No. 5A

An 87-year-old female with a history of chronic and severe thoracic back pain had been seen by an insured anesthesiologist for epidural steroid injections. The patient had received at least three epidural injections and multiple intramuscular injections of morphine without incident. Accompanied by her granddaughter, she arrived in the emergency department complaining of left shoulder blade muscle spasm. Records indicate that she received an injection. A T-9 epidural was placed for administration of Depo-Medrol 80mg, lidocaine 40mg, bupivacaine 0.25%, and Decadron 4mg. The patient was then discharged home.

Approximately two hours later, she returned to the emergency department complaining of severe mid-back pain. A telephone order from the anesthesiologist prescribed morphine 10mg and Visraril 50mg IM, which was administered by a nurse. The patient was released home in the company of her granddaughter and was not examined by a physician at that time.

Approximately 10 hours after that discharge from the emergency department, the patient’s family called the paramedics. She returned to the ER with increasing back pain, leg numbness, and urinary incontinence. The ER physician noted a sensory-motor block distal to T-7. An MRI obtained several hours later revealed extradural fluid from T-5 to the lumbar area. The patient was evaluated by a neurosurgeon, who diagnosed paraplegia secondary to a thoracic epidural hematoma. Since the patient had been flaccid for at least 12 hours, the prognosis for any functional recovery, even with surgical decompression, was small. The family decided against surgical intervention after discussing the potential risks and likely outcome.

The panel’s review of these medical records indicates that written postepidural discharge instructions should include contacting the anesthesiologist if pain, fever, or headache occurs. Epidural hematoma and infection, while rare, require immediate treatment. The anesthesiologist retains responsibility for prompt evaluation of the patient with follow up until the problems are resolved. Treatment prescribed by telephone, particularly without an examination of the patient, is outside the standard of care.

Case No. 5B

An anesthesiologist insured by The Doctors Company was called to provide a labor epidural for a 30- year-old woman in active labor. Informed consent was not documented, but the insured insisted that risks, including headaches, were discussed. The placement of the block was technically difficult and required multiple attempts. Satisfactory analgesia was obtained, and the patient delivered vaginally without complications. The patient was discharged one day postpartum.

After discharge, the anesthesiologist telephoned the patient at home and found her complaining of headaches and back pain. The insured suspected a probable dural puncture and postspinal headache, and recommended that the patient return to the emergency department at the hospital for a blood patch. The patient had no health insurance coverage and refused to go to the hospital. She later sued, however, alleging that negligent administration of the epidural caused severe headaches and back pain.

Panel members believed that the patient should have been evaluated postdelivery by the anesthesiologist and should have been made aware of the possibility of a postspinal headache. The anesthesiologist should have directed the patient to call if symptoms developed. Had the anesthesiologist been able to advise early increased fluid intake and arrange for oral pain medications, the severity of the headaches possibly could have been lessened. An arrangement might also have been made with the hospital to allow the blood patch to be performed by the anesthesiologist for a minimal fee. The panelists generally concurred that better doctor-patient communication could have prevented this lawsuit.

Case No. 5C

This case involved a 30-year-old male who injured his lower back while lifting a heavy object. An MRI revealed evidence of a herniated disc at L4-5 and L5-S1. Risks of the procedure were discussed with the patient, and verbal consent was obtained for an L4-5 epidural steroid block. The patient reported some decrease in pain, but because of continued radicular findings, a second epidural steroid block at L5-S1 was performed, without incident. Subsequent to the second epidural steroid block, the pain recurred and the patient developed pain, stiffness, and numbness in his lower back, and loss of sensation and strength in his right leg and foot.

The panel members’ review of this case illustrates the importance of anesthesiologists clearly communicating with patients about the treatment process for chronic pain. Again, as an anesthesiologist, you are well advised to use such opportunities to provide patients with a clear understanding of procedures and the inherent risks, complications, and limitations. Your discussion should also include the probability of success based on the patient’s condition.


About the Author

Ann S. Lofsky, MD (deceased), was anesthesia consultant and board member emeritus to The Doctors Company. She was a diplomate of the American Board of Anesthesiology and the American Board of Internal Medicine.


Panel Members

Panel members who reviewed these cases are all insured by The Doctors Company and are board-certified anesthesiologists: Martin Cousineau, MD, Calif.; George Herr, MD, Calif; Ann S. Lofsky, MD, Calif.; Michael Rost, MD, S. Dak.


The guidelines suggested here are not rules, do not constitute legal advice, and do not ensure a successful outcome. The ultimate decision regarding the appropriateness of any treatment must be made by each health care provider in light of all circumstances prevailing in the individual situation and in accordance with the laws of the jurisdiction in which the care is rendered.

J3216 9/99



Follow us on Twitter
Subscribe to us on YouTube
Connect with us on LinkedIn
Like us on Facebook