The role of radiology in diagnostic error: a medical malpractice claims review

Interpretation of diagnostic studies

Of the 1325 cases in this analysis that named Radiology as the Primary Responsible Service, nearly 60% (766) were directly related to an allegation of diagnostic failure (see Figure 2). Analysis of the specific issues in these cases revealed that 48% of these cases involved a misinterpretation of a diagnostic study that led to a delay in the (correct) diagnosis and/or treatment of the patient’s condition. Of those cases classified as a misinterpretation of a diagnostic study (see Figure 3), primary malignancies and/or metastases were the most commonly missed diagnoses, with breast and lung cancer representing the most common case types. Of the 89 cases with a failed diagnosis of breast cancer, 76 involved a mammography study. In 47 cases involving lung cancer, 30 of them involved routine chest radiographs. The second largest category of cases with a misinterpretation factor (see Figure 4), involve fractures, with routine radiography the primary modality. Specific examples include:

Figure 2:

Distribution of top service categories in MPL claims by volume and dollars.

CBS n=29,777 MPL cases asserted 1/1/10–12/31/14; $5.7B total incurred losses. Total incurred includes reserves on open and payments on closed cases. Surgery includes general surgery and all surgical sub-specialties. Medicine includes general medicine and all medicine sub-specialties.

Figure 3:

Distribution of radiology events by top allegation.

CBS n=1325 MPL cases asserted 1/1/10–12/31/14 with radiology as the primary responsible service; $263M total incurred losses. Total incurred includes reserves on open and payments on closed cases.

Figure 4:

Clinical judgement issues in radiology cases.

*A case will often have multiple factors identified. CBS n=1325 MPL cases asserted 1/1/10–12/31/14 with radiology as the primary responsible service; $263M total incurred losses.

• Failure to properly interpret a mammogram resulted in a 2 year delay in diagnosing invasive ductal breast cancer with subsequent metastases and patient death.

• Misread CT resulted in a 22-month delay in diagnosing lung cancer with poor prognosis.

• Failure to diagnose multiple dislocations associated with carpal-metacarpal fractures (dorsal dislocations 2nd through 5th digits right hand) resulted in multiple unsuccessful surgical repairs and permanent loss of function.

• Failure to properly interpret CT scan (spine) resulted in delayed diagnosis of cervical fracture/cord compression and paralysis.

• Misinterpretation of CT scan (head) resulted in delayed diagnosis of subdural hematoma and death.

Communication of test results

The Institute of Medicine (IOM) report Improving Diagnosis in Health Care, published September 22, 2015, offers a new definition of diagnostic failure that includes not only “(a) the failure to establish an accurate and timely explanation of the patient’s health problem(s)”, but also “(b); the failure to communicate that explanation to the patient” [1].

Data from this study demonstrates that communication failures were noted in 23% of the radiology cases (see Figure 5). Of those, communication to the ordering provider accounted for 13% of the cases while communication to the patient was noted as a contributing factor in 10%. Specific examples include:

• Failure to communicate critical over-read of a neck radiograph, first read (by a resident) as mild swelling, but follow-up read found significant potential for airway obstruction. While the over-read was documented, its urgency was not verbally reported per the “critical result” process. The patient returned to the ED in full respiratory arrest and died of a vascular rupture before the updated report was communicated to the patient

• Failure to properly document (thus notify the ordering provider of) an incidental finding of small pulmonary nodule with recommendation for follow-up. The requested study for rib pain was negative, however, the radiologist did not mark the “abnormal result” box as a variance in the medical record, so the primary care provider was unaware of the positive finding and recommendation that should have been relayed to the patient. Three years later the patient was diagnosed with bronchogenic carcinoma with poor prognosis.

Figure 5:

Communication issues in radiology.

*A case will often have multiple factors identified. CBS n=1325 MPL cases asserted 1/1/10–12/31/14 with radiology as the primary responsible service; $263M total incurred losses.

The IOM report on diagnostic error

The IOM report on diagnosis notes in Recommendation 4 that, “Health care organizations should monitor the diagnostic process and identify, learn from, and reduce diagnostic errors and near misses….and implement procedures and practices to provide systematic feedback on diagnostic performance to individual health care professionals, care teams and clinical and organizational leaders” [1]. Additionally, recommendation 6D notes, “Professional liability insurance carriers and captive insurers should collaborate with health care professionals on opportunities to improve diagnostic performance through education, training, and practice improvement approaches and increase participation in such programs” [1]. By analyzing Medical Professional Liability (MPL) claims, we can learn from the specific patterns and trends that contribute to diagnostic failure thus appropriately allocating resources and implementing targeted interventions that address those specific issues.

Given the influential role of radiology in the diagnostic process, it is imperative to monitor the practice and provide feedback from radiologic events including what is learned from MPL claims. While issues of competencies in reading and interpreting studies may be a focus within the service of radiology, the issues of communication require much broader solutions across the healthcare team. However, the unifying issue in addressing both of these problems is raising awareness of the specific vulnerabilities that put our staff, our providers and our patients at risk. Sharing this, and similar data reports, analyses and case studies, is critical to helping providers understand the most vulnerable processes and diagnoses in hopes of increasing awareness and vigilance in these trending errors.

Radiology interpretation errors

Radiology, like all fields of medicine, requires a commitment to continued education and practice improvement. While many practice environments are unique, there are several commonalities that can be focused upon, starting with residency training. The Accreditation Council for Graduate Medical Education (ACGME) program requirements for graduate medical education necessitates a board certified radiologist to review images and sign all resident-generated reports [2]. This process not only facilitates resident education, it also aims to provide an additional level of security by allowing multiple providers the opportunity to evaluate images. Similarly, some practices allow non-radiologists to interpret imaging studies. Multiple studies have reported a broad range of discrepancy rates between radiologists and non-radiologists, however, of the noted discrepancies, many fell into categories our review found to be high-litigation diagnoses such as missed fractures and pulmonary nodules or malignancies [3], [4], [5], [6]. It is our goal to improve prompt identification of these important findings by recommending the adoption of comparable review processes whereby a radiologist “over-reads” imaging studies initially interpreted by non-radiologists.

Peer review

It is equally as important for radiology practices to internally adopt established peer review and audit procedures to detect trends and help identify opportunities to support radiologist education if unfavorable variability is encountered [7]. This recommendation aims to help reduce one of the most significant barriers of practice-based continued learning in radiology, which is inconsistent and often delayed feedback on reported findings. Continued learning and practice improvement through peer review and audit systems intend to enhance the diagnostic process with the ultimate objective to provide appropriate and timely patient care.

In addition to introspectively analyzing individual practice trends, increasing awareness of reported regional and/or national data is useful in identifying larger scale tendencies and potential vulnerabilities towards diagnostic error [8]. By bringing light to these areas, radiologists can use this information to refine image interpretation and improve patient diagnosis. For example, our review identified missed breast cancers as one of the most common diagnostic-related allegations. As anticipated, the majority of these missed cancers involved mammography, however, in a small number of cases of the unidentified breast cancer cases were missed on chest CT examinations. This finding brings light to the importance of being vigilant in the evaluation of anatomy outside the primary area of interest on imaging studies, such as the subcutaneous tissues on a chest CT where the anatomy of interest typically includes the cardiovascular and respiratory systems.

Communication errors

While the accurate interpretation of radiologic studies is key to diagnostic safety, communication of diagnostic results to the ordering provider and ultimately to the patient is equally as important. As previously noted, the IOM report on Improving Diagnosis in Health Care creates an imperative to address the communication aspects of test result management [1]. Despite the fact that our claims analysis cited communication errors less frequently than failures in diagnosis, policy improvement targeting flaws in this process has the ability to produce widespread and measurable change [9]. This can be achieved by implementing clear and well-communicated processes defining the methods and accountability for information transfer between radiologists, clinicians and patients [10].

Due to inherent differences in practice environments, unique policies may be required to suit the demands and optimize the resources of specific care settings. For example, incidental imaging findings encountered in the emergency department setting are at risk for delayed or failed follow-up for a variety of reasons, some of which include discontinuity of care and limited availability of diagnostic reports and patient records [11].

In response to this problem, Dutta et al. [12], researched the use of natural language process algorithms to incorporate additional imaging recommendations into emergency department discharge summaries. In other settings, the process of how the incidental findings are documented is a key contributor to the communication failure. Aware that clinicians reading reports may not always look beyond the result that prompted the order, some organizations have reconfigured their radiology reports to provide a more prominent display of the incidental finding. Solutions such as these align with the IOM goal to “develop and implement processes to ensure effective and timely communication between diagnostic testing health care professionals and treating health care professionals across all health care delivery settings” [1].

Practice standards and guidelines

The American College of Radiology (ACR) has established practice parameters regarding routine and nonroutine communication of imaging findings to ordering providers. When unexpected findings are encountered by the radiologist, these discoveries should be relayed to the referring clinician in a timely fashion. The practice parameter cites several situations that warrant nonroutine communication between the ordering and interpreting physicians. These conditions include findings that may require immediate or urgent intervention, conclusions that differ significantly from a preliminary report, and findings the interpreting physician feels may result in serious adverse outcomes or may worsen over time if untreated [13]. All nonroutine findings are not explicitly detailed in the practice parameter as some may be specific to certain scenarios or patient populations. It is essential for practices to define their expectations for urgent and non-routine communication to help ensure prompt and appropriate patient care [14]. Our analysis identified the outpatient setting as particularly vulnerable with missed or delayed cancer diagnoses leading to the highest number of cases resulting in severe disability and patient death. While a new diagnosis of malignancy often does not require emergent intervention, it is vital these patients be identified and integrated into a proper treatment plan as there is a greater potential for delayed management compared to a disease process with higher acuity that will result in more prompt patient return to medical care. The radiologist can help play an initial role in ensuring and expediting patient management in new cancer diagnoses by directly communicating these findings to the referring clinician.

Discrepancies between initial and final Radiology reports

Discrepancies in reported imaging findings can pose a challenge if treatment has been implemented based upon an initial radiologic interpretation that is later revised. It is the radiologist’s responsibility to inform the ordering provider if there is a significant change in imaging findings from the preliminary report. Likewise, procedures must be in place to ensure the ordering physician receives the finalized radiology report. In cases where the ordering physician may not provide long-term patient care, particularly in the Emergency Department setting, it is important that policies necessitating follow-up of discordant imaging findings are routinely followed.