The Doctors Company analyzed 596 claims (written demands for payment) against radiologists that closed from 2013 through 2018. We divided the claims into two subspecialty groups, diagnostic (542) and interventional (54) radiology claims.
Regardless of the outcome, all radiology claims against our members that closed within this time frame were included in this analysis. This approach helps us better understand what motivates patients to pursue claims and gain a broader overview of the system failures and processes that result in patient harm.
This study, reinforced by expert insights and relevant case examples, focuses on the following areas:
- Most common case types.
- Most common patient injuries.
- Injury severity.
- Factors contributing to patient injury.
- Strategies for mitigating risk.
Our approach to studying radiology malpractice claims began by reviewing the types of cases to capture the key clinical areas.
We then looked at patients’ injuries to understand the full scope of harm. Physician experts for both the plaintiffs/patients and the defendants/physicians reviewed claims and conducted medical record reviews. Our clinical analysts drew from these sources to gain an accurate and unbiased understanding of actual patient injuries.
We identified factors that led to patients’ injuries, and physician reviewers evaluated each claim to determine whether the standard of care was met. Contributing factor categories included clinical judgment, technical skill, patient behaviors, communication, clinical systems, clinical environments, and documentation.
Our team studies all aspects of the claim and identifies risk mitigation strategies that physicians can use to decrease the risk of injury, thereby improving the quality of care.
The three most common case types in diagnostic radiology made up 85 percent of those claims. The three most frequently identified case types in interventional radiology made up 87 percent of those claims.
|Top 7 Case Types||Diagnostic Radiology Claims||Percent of Diagnostic Radiology Claims||Interventional Radiology Claims||Percent of Interventional Radiology Claims|
|Diagnosis related (failure, delay, wrong)||364||67%||1||2%|
|Improper performance of treatment or procedure||26||5%||32||59%|
|Improper management of treatment course||18||3%||9||17%|
|Retained foreign body—medical treatment||5||1%||6||11%|
|Failure to ensure safety—falls||7||1%||0||0%|
|Wrong or unnecessary treatment or procedure||5||1%||1||2%|
Rate of paid claims:
The most common diagnoses associated with diagnosis-related claims (failure, delay, or wrong) were:
The following conditions each represent 1 percent or more of diagnostic radiology claims: subarachnoid hemorrhage, malignant neoplasm of colon, malignant neoplasm of pancreas, cerebral thrombosis with infarction, acute cerebrovascular accident (CVA), cerebral aneurysm, pelvis fracture, ankle fracture, and intracranial abscess.
Of the top 14 missed, delayed, or wrong diagnoses, 26 percent involved malignant neoplasms; 5 percent involved thrombi, hemorrhages, or abscesses in the cranium; and 4 percent were bone fractures. There were 161 other diagnoses, each constituting of fewer than 1 percent of diagnostic radiology claims.
Misinterpretation of Diagnostic Studies
This factor was present in 78 percent of cases. The top five diagnostic studies most frequently misinterpreted make up 98 percent of the diagnostic studies that were misinterpreted.
Misinterpretation of diagnostic study by study type:
CT scan conditions that were misread included:
- Basilar or cerebral artery occlusion/thrombosis or aneurysm (10%).
- Subarachnoid, subdural, or intracerebral hemorrhage (9%).
- Acute appendicitis (8%).
- Neoplasm of kidney or renal pelvis (5%).
- Neoplasm of stomach, small intestine, or colon (4%).
- Malignant neoplasm of head, neck, face, or thyroid gland (2%).
- Malignant neoplasm of bronchus or lung (2%).
The top conditions misinterpreted from x-ray radiographs included:
- Malignant neoplasm of bronchus or lung (19%).
- Fracture of radius and/or ulna or navicular of wrist (9%).
- Fracture of neck of femur (6%).
- Fracture of ankle or calcaneus (6%).
- Fracture of metatarsal bones (5%).
- Fracture of pelvis (4%).
- Fracture of open base thumb or phalanges (3%).
- Fracture of clavicle (2%).
The top conditions misinterpreted from MRIs included:
- Spinal epidural abscess (16%).
- Acute CVA, cerebral aneurysm, or cerebral thrombosis (10%).
- Malignant neoplasm of brain (8%).
- Meningitis or meningoencephalitis (4%).
- Malignant neoplasm of pelvic bones (4%).
The top conditions misinterpreted from ultrasound included:
- Malignant neoplasm of breast (31%).
- Phlebitis/thrombophlebitis (13%).
- Torsion of testis (13%).
- Malignant neoplasm of liver (6%).
- Pulmonary embolism (6%).
- Ectopic pregnancy (6%).
Communication Among Providers
This was the second-most common contributing factor in diagnosis-related cases, appearing in 18 percent of claims. Most of these cases involved a break in communication between or among clinicians. Examples included ancillary findings such as lesions or nodules on the lungs or livers of ER patients with trauma or acute illness that were not addressed due to the focus on the emergency. In most cases, it was determined that the radiologist should have called the ER physician or the attending physician to make sure that there would be follow-up to these troubling findings.
Some of these cases were hampered by inadequate information about patients and the rationale for ordering radiological studies. It is important for radiologists to receive information that helps them arrive at interpretations that are consistent with clinical findings. If findings are not what was expected, the radiologist and ordering physician should collaborate on next steps. Radiologists need to be integrated into the treatment team to provide the best possible care through accurate diagnoses and treatment.
Communication was hampered in some situations where one of the clinicians spoke English as a second language. In a case of acute appendicitis, the radiologist called the ER physician with his findings. The ER physician misunderstood the radiologist, thinking he had said that the patient’s appendix had been removed. This delayed the diagnosis, resulting in a ruptured appendix. All communications between clinicians should include read-back to confirm that information is correctly conveyed and understood.
The third-most common factor contributing to patient injury in diagnosis-related claims was patient behaviors, appearing in 14 percent of claims. In several cases, patients were told of suspected malignancies or brain aneurysms, and they failed to attend appointments or schedule follow-up studies. In one case, it was suspected that the patient did not arrange for additional studies because they had no insurance. In other cases, patients ignored symptoms of possible stroke, minimizing the chance that clinical interventions would be successful.
There were 71 cases of improper inspection and maintenance of equipment. In all cases, CT scanners were new or received updates from the manufacturer. The hospitals and radiologists depended on the manufacturer to select the settings, which were later determined to be too high. In most cases, patients received excessive doses of radiation with repeat studies over the course of a year.
When patients were notified of the errors, they claimed hair loss and fear of cancer or cataracts from the excessive exposure. The FDA attributed the initial errors to the manufacturer but stated that the hospitals and radiologists were responsible for checking each CT scanner for radiation emitted before doing a scan. There were also recommendations for dosing protocols and quality control procedures.
There can be circumstances when protocols are lost during routine maintenance or when equipment needs to be rebooted. Policies and procedures should outline the process for reestablishing and confirming that the correct protocols are in place.
The factors that contributed to patient harm in these cases were inadequate staff training, need for a policy or protocol, and failure to inspect equipment.
Examples included alleged rupture of breast implants from mammograms or needle biopsies. Patients reported nerve damage from insertion of needle into wrist, or tissue damage with extravasation of contrast media. Patients also complained of infections following injections of contrast media into hips and knees, and perforation of colon during administration of contrast media. In other cases, x-rays were misread due to poor positioning or inadequate compression of breast tissue, resulting in poor quality radiographs.
The factors that contributed to patient injury were known complications not due to substandard care, poor technique, inadequate communication between the clinician and patient, and insufficient or lack of documentation.
The most common injuries associated with misinterpretation of a diagnostic study included:
*Open or closed
Note that patients may suffer more than one injury, such as hemorrhage and death, so the total is greater than 100 percent.
A 49-year-old patient presented for a routine screening mammogram. The patient had a history of smoking and a family history of breast cancer. The radiologist who read the mammogram noted that the study was not complete due to density of breast tissue, and recommended additional studies. He recommended additional imaging to evaluate a density in the right upper breast.
One week later, a second mammogram was performed. The second radiologist noted very dense tissue but no evidence of malignancy. No follow-up was recommended.
One year later, the patient presented to her gynecologist with a complaint of a breast lump. The gynecologist ordered bilateral mammograms and right breast ultrasound. The third radiologist reported mildly asymmetrical breasts with numerous small cysts. He also noted increased density in the upper outer quadrant of the right breast but did not identify any suspicious nodules. The radiologist advised follow-up studies in one year.
The patient did not return for three years. The gynecologist and radiology department sent letters recommending follow-up studies, but the patient did not respond.
The patient then presented to urgent care with severe back pain, shortness of breath, and chest pain with deep breathing. A chest x-ray noted clear lungs, but identified osteolytic lesions on the spine. A biopsy showed metastatic breast carcinoma. The patient expired two years later.
After a claim was filed against the radiologists and gynecologist, plaintiff’s experts opined that the original mammogram was abnormal enough to prompt an MRI and biopsy. The second year’s mammogram was more suspicious of abnormality and included a complaint of breast lump. They also stated that if diagnosed and treated in the first year, the cancer would not have metastasized.
Defense experts determined that the second and third radiologists should have recommended MRI and biopsy, and the gynecologist should have ordered an MRI or biopsy following the second mammogram. The primary concern was the failure to order a biopsy when the patient presented with a breast lump. The dense breast tissue obscured the ability to identify tumors.
The defendants settled this case.
A 45-year-old male presented to the emergency department with burning chest pain and a history of hypertension. The patient was experiencing shortness of breath, diaphoresis, dizziness, and anxiety.
His EKG was borderline with nonspecific T wave abnormality. His chest x-ray showed no definite abnormality. Troponins were not diagnostic for myocardial infarction, but his D-dimmer was elevated.
A CT of the chest was read as normal with no pulmonary emboli. The patient was admitted to the hospital with an order for a routine cardiac consult. An EKG showed tachycardia with ST depressions. The patient experienced nausea and pain (8/10) that was uncontrolled with a nitroglycerin patch. A nurse gave morphine but did not notify a physician. Oxygen saturation was not evaluated.
The day-shift nurse documented chest pain (10/10), tachycardia at 130 bpm and oxygen saturation at 74 percent on room air. She started the patient on 100 percent oxygen, and contacted the patient’s primary care physician.
The physician heard a loud murmur. He ordered heparin bolus and drip, and contacted the cardiologist requesting an immediate evaluation. An echocardiogram showed an enlargement of the aortic root. The radiologist was contacted and asked to reevaluate the patient’s chest CT scan. The radiologist then noted that the aortic root was 4.8 cm and the descending aorta was 3.2 cm. The patient was diagnosed with an aortic dissection.
The patient was transported to a tertiary care center, where he expired from hypoxic encephalopathy and type A aortic dissection. The family filed a claim.
The plaintiff’s experts criticized the care provided by the ER physician. They stated that he should have checked blood pressure in both arms, and should have ordered an echocardiogram. The radiologist was criticized for not considering dissection and recommending additional studies. The night nurse should have notified physicians of the patient’s changing status.
Defense experts supported the care provided by the ER physician, stating that it was reasonable to depend on the radiologist’s interpretation of the CT. They also opined that the workup was appropriate for the patient’s symptoms, which were atypical for dissection. The transfer was timely once a correct diagnosis was made.
The radiologist settled the claim.
A 49-year-old female presented to the hospital for an ultrasound-guided breast biopsy. Some risks of the procedure were discussed with the patient, not including pneumothorax, which occurs in less than 2 percent of these procedures.
The patient received local anesthetic and the lesion was localized using ultrasound. A core biopsy was acquired and sent to pathology.
A short time after the procedure, the patient began to complain of dyspnea and chest pain. A chest x-ray showed a moderate-sized pneumothorax, so the patient was admitted to the hospital. A few days later, the pneumothorax resolved, and the patient was discharged with no further sequelae.
In the claim filed by the patient, plaintiff’s experts opined that the radiologist fell below the standard of care by proceeding with the procedure despite the patient’s obvious anxiety. Her anxiety increased the risk that the patient would move during the procedure. They also said that the radiologist must not have monitored the biopsy needle during the procedure, as the needle had to move beyond the tumor to puncture the lung.
Defense experts supported the care, stating that pneumothorax is a known risk. Also, the radiologist quickly recognized the signs of a pneumothorax, and took steps to protect the patient. However, since pneumothorax is a known risk and was not mentioned during the informed consent discussion, the case was settled for a small amount.
The most common case type in interventional radiology was related to performance of treatment or procedure, appearing in 59 percent of claims. Examples of procedures included biopsies of kidney, liver, breast, pulmonary nodules, lymph nodes, and bone marrow. They included embolization of AV malformation, uterine arteries, and cerebral artery aneurysms. Other procedures were angioplasty, draining abscesses, epidural steroid injections, and placement of IVC filters.
This factor appeared in 76 percent of claims. Many of the claims and suits were brought by patients who had invasive procedures and suffered poor outcomes. In most cases, the injury was known to the patient as a risk of the procedure. Only 11 percent of these cases were identified as poor technique or incorrect body site and resulted in indemnity. In most cases (65 percent), experts found that the correct procedure was performed appropriately, even when patients had undesirable outcomes from these high-risk procedures.
Communication Between Patient and Provider
This factor appeared in 24 percent of claims. Communication issues in interventional radiology most often were related to inadequate consent for invasive procedures or inadequate consent for other treatment options. Documentation in these cases is essential to demonstrate that the patient gave consent, even if the patient could not recall the conversation.
Insufficient or Lack of Documentation
This factor appeared in 16 percent of claims. In some cases, no notes were made regarding changes in position of central line catheters. In another case, the interventional radiologist canceled the patient’s procedure after attempting unsuccessfully to insert contrast media into the patient’s colon. The patient complained of pain and later was found to have a perforated colon. The interventional radiologist had not documented attempts to infuse contrast media, delaying diagnosis of the complication. Other cases of inadequate documentation involved informed consent for treatment when patients alleged that they had not been told of the risks. Radiologists had difficulty defending their care without that documentation.
Selection and Management of Therapy
Selection and management of invasive procedures appeared in 16 percent of claims. In one case, a patient had several nodules that were suspected of being malignant. The nodule selected for biopsy was close to a major blood vessel, and this resulted in a life-ending hemorrhage. It was determined that biopsy of one of the other nodules would have been less risky.
This was the second-most common case type for interventional radiologists, appearing in 17 percent of claims. These cases were closely related to invasive procedures, but were about the management of the procedure and management of the patient following the procedure.
Some procedures resulted in hemothorax from central line placement, thoracentesis, or lung biopsies. Some resulted in thrombi like DVTs from injuries to blood vessels. Insertion of percutaneous gastric tubes resulted in sepsis, and nerve ablation caused damage to the wrong location.
In these cases, patients suffered complications from the procedure. Immediate care was managed by the radiologist. In all but one of these cases, experts determined that the care provided by the interventional radiologist was appropriate. Only one case resulted in an indemnity payment.
The factors that contributed to patient injury were identified as known complications. Other factors were lack of communication among clinicians regarding a patient’s condition, and information lost in the transition of care.
The third-most common case type, appearing in 11 percent of claims, involved retention of pieces of equipment or catheters used in invasive procedures. Part of a Hickman catheter broke off, causing an abscess. A portion of a microcatheter broke off and remained in a patient’s brain, causing stroke. A portion of a urethral stent was retained in a patient’s bladder and later retrieved without complications. One of the tines from a radiofrequency ablation (RFA) probe broke off and had to be removed surgically following a procedure to address a suspected hemangioma.
In most cases, these complications were considered risks of the procedure that were known to the patient. Information was shared during the informed consent discussion. These complications were not considered substandard care. In the two cases that resulted in indemnity payments, one was a retained suture that held a nephrostomy tube in place. It resulted in an abscess. The other was a small fragment from a piece of equipment that had to be retrieved. It was determined that the equipment malfunctioned due to operator error; the broken piece was visible on films, and could have been retrieved before it caused ongoing pain.
The most common patient injuries in interventional radiology procedures included:
Patients may suffer more than one injury, so the total percentages add up to greater than 100 percent.
A 79-year-old patient with a history of COPD, chronic atrial fibrillation, hypertension, and peripheral vascular disease presented to the interventional radiologist for fluoroscopic-guided intrathecal contrast injection of lumbar spine for a CT-myelogram. The radiologist learned the day of the procedure that the patient was taking Xarelto for chronic atrial fibrillation, but decided to proceed. He reasoned that the patient had extensive vascular problems that would exacerbate the risk of clotting and atrial fibrillation, which would increase the risk of stroke, if the patient was to discontinue Xarelto.
The risks, benefits, and alternatives of the procedure were discussed with the patient. The radiologist also talked with the patient about the risks of proceeding because the patient was on an anticoagulant. There was no documentation of this discussion.
Following the procedure, the patient complained of pain at the injection site. He developed internal bleeding in the lumbar area and aseptic meningitis. He now has limited strength in his legs, difficulty with speech, and other neurological sequelae.
Expert reviewers were critical of the lack of documentation of the informed consent discussion. They said that the black box warning to be off Xarelto for 24 hours prior to the procedure was not followed. They also stated that the radiologist should have called the prescribing physician to discuss the patient’s situation or should have consulted with a cardiologist or vascular surgeon before making the decision to proceed with the procedure. The case was settled.
A 55-year-old female with a large uterine fibroid was scheduled for bilateral uterine artery embolization (UAE). The procedure was planned to reduce the bleeding during the planned hysterectomy.
Laboratory studies showed a slightly elevated prothrombin time of 15.6 seconds (normal is 10.8–14.7 seconds) and INR of 1.20 (normal is 0.8–1.29).
The radiologist had difficulty deploying the sheath into the blood vessel. She changed to a stiffer guidewire, but accidentally lacerated the blood vessel. A vascular surgeon repaired the artery.
After embolization, the patient was transferred to the operating room for her hysterectomy. Before surgery, the patient had complained of pain in her thigh, which was thought to be secondary to compression on the femoral artery following the arterial catheterization.
In the operating room, during preparation for surgery, nurses noted that the patient’s right foot was cold, and pulses were not palpable even with Doppler. After conferring with the radiologist and vascular surgeon, the hysterectomy was canceled.
An arteriogram showed focal stenosis secondary to dissection of the distal external iliac artery. Dissection may have been caused by a closure device and possible peripheral artery disease. The decreased flow rate was thought to have been the result of spasm or atheromatous plaques dislodged from the intimal injury from the stent of the external iliac artery. Physicians ordered an angiogram and started the patient on heparin.
The patient did well until the heparin levels became subtherapeutic, which is a rare complication. The patient had thrombi in both anterior and posterior tibial arteries. Stents were placed, with some improvement in circulation, but ischemia in the patient’s foot resulted in poor healing of toes and possible gangrene. The patient experienced partial amputation of her foot.
Plaintiff’s experts testified that a delay in addressing the blood clots in the patient’s lower leg was the cause of ischemia and foot amputation. They stated that t-PA should have been used to lyse the clots. Heparin only keeps clots from forming. They also questioned using a closing device on an obese patient.
Some defense experts supported the care, stating that the physicians identified the problem and took steps to mitigate the harm. The informed consent covered relevant potential problems. UAE was reasonable for removing a pelvic mass considering the potential for bleeding during a hysterectomy.
They also stated that when the problem was recognized, an angiogram was ordered. The patient did well until Heparin became subtherapeutic.
Other defense experts believed that the patient had blue toe syndrome, and that the physicians should have inserted a microcatheter to give t-PA directly to the location of the thrombi. Due to mixed reviews, the case was settled.
- Diagnosis-related (failure, delay, wrong) cases were the most common for diagnostic radiologists. Malignant neoplasms of the breast, malignant neoplasms of the lung, and spinal epidural abscesses were the most frequently identified conditions in these cases. Have a defined process for identifying and analyzing diagnostic errors. Focus education on problematic diagnoses and encourage second reviews for scans and films with uncertain findings.
- Track inspections of equipment and monitor updates and settings. Update policies and protocols as needed. Have a process for following up on equipment-related incidents.
- Encourage physicians and other clinicians at all levels to provide clinically relevant information when radiological studies are ordered. The American College of Radiology’s “ACR Practice Parameter for Communication of Diagnostic Imaging Findings” recommends that ordering clinicians include a working diagnosis, pertinent clinical symptoms, and specific questions to be answered.
- Radiologists should be integrated into the healthcare team. If findings are not consistent with clinical information provided by the ordering physician, radiologists should have a conversation with them to analyze the situation and determine next steps in the diagnostic process.
- Order repeat studies when views are not of good quality or positioning blocks the full view. For instance, when mammograms or CT studies of the breast have poor image quality, or the areas of concern are not fully visible.
- Maintain policies on communication of ancillary findings that require follow-up and on emergent or critical imaging interpretations. If the ordering physician cannot be reached, consider contacting the patient directly. Document date, time, information provided, and the person with whom the communication occurred.
- Create or update the process for notifying the appropriate clinicians regarding discrepancies between preliminary readings and final reports. Document these communications. If the ordering physician cannot be reached, consider contacting the patient directly.
- When reporting findings to other clinicians, ask the receiving clinician to repeat back the message to ensure accurate transmission of findings, including patient identification.
- Review and update as needed the practice’s or organization’s policies on managing medical emergencies in radiology. Conduct drills to enhance the ability to quickly respond to these emergencies.
- Informed consent is essential for any invasive procedure. Document that the patient accepted the risks of the procedure. This is necessary for defense of claims when patients say that they were unaware of the risks that resulted in harm.
- Sixty-five percent of claims alleging improper performance of a procedure were found by physician reviewers to have been performed correctly. Patients who experience unwanted outcomes from invasive procedures sometimes file claims when they don’t understand the cause of their injury. Radiologists should make opportunities to talk to patients or their families about what happened. Patients may remain unhappy with the outcome, but they will have an accurate understanding of the cause. Sharing information helps to build trust between patients and their clinicians.
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 healthcare provider considering the circumstances of the individual situation and in accordance with the laws of the jurisdiction in which the care is rendered.