**3. Clinical vignette #1**

structure of altered case reports will be provided, which constitutes one way of identifying error in pathology. **Figure 1** highlights significant sources of error in both of these processes [13, 14]. It is of paramount importance for providers to understand the limitations of research in the current literature regarding the preponderance/magnitude of potential and actual error that exists in pathology (as well as the common failure modes in such settings)

**Figure 1.** Relative frequency of errors occurring during practical/systemic and interpretive/diagnostic processes. (**A**) Reasons for clinical lab error prior to delivery of sample for interpretation. These errors are not differentiated between pre- and post-verification [13]. (**B**) Data for cases of medical negligence resulting in practice considered below the standard of care. Clinical pathology refers to laboratory error, practical error refers to system errors, miscellaneous surgical pathology errors refer to claims which show no pattern in specimen diagnostic criteria and are considered random, and other repetitive pattern errors include sarcomas, lymphoma, lung, gastric, fine need aspirates, prostate, bladder, and nongynecologic cytology errors; 57% of claims are from practical errors, melanoma, breast, Papanicolaou,

(**Figure 2**).

144 Vignettes in Patient Safety - Volume 3

and gynecologic samples [13, 14].

Shahar et al. [16] described a 47-year-old man who presented to the emergency room after reports of progressive right lower extremity weakness. Relevant history included 40-packyear of tobacco abuse as well as upper-limb dysmetria. Magnetic resonance imaging (MRI) revealed distinct ring-enhancing lesions in the left frontal lobe of the brain, which were biopsied and reported as metastatic small cell lung carcinoma. The patient received radiation treatment for increasing right lower extremity weakness, headache, and blurred vision.

The patient demonstrated worsening lethargy and headache prompting a brain computed tomography (CT) that showed an enlarging mass with midline shift. Histopathological examination suggested glioblastoma with no evidence of metastatic carcinoma. Despite suspicion of a possible rare "collision tumor" (a tumor specimen from a single patient in which pathology reports do not coincide), DNA sequencing of the two biopsies was performed to determine if the tumor was monoclonal. Several genotypic and microsatellite analyses revealed that the samples did not originate from the same patient.

In this particular case, the patient and clinicians were fortunate enough to identify the sampling error early in treatment, which allowed for an appropriate adjustment of the treatment (for the correct diagnosis of glioblastoma).

somewhere during the process of securing specimen(s) for the initial biopsy sample and review, an error occurred resulting in patient-specimen mismatch. Labeling error is not a phenomenon unique to pathology, but can occur in any process leading up to a report generation for therapy or prognosis. Labeling errors may occur when specimens are labeled with incorrect patient name/identification, or accession number, but may also be related to the sample's origin (e.g., lower versus upper extremity); time (e.g., two procedures by two different surgical teams); or location (e.g., endoscopy suite versus operating room). The possibility of labeling errors also exists within the analytical framework whether in regard to the pathologist(s) or in the context of report retrieval/delivery [21, 22]. About 0.25% of cases are subject to a labeling error during the pre-analytical phase, with a majority of errors (73%) being associated with patient name [21]. Implementation of safety measures such as open communication with the patient and formalized checklists incorporated in transfer of information/specimens from the operating room to the laboratory and vice versa have shown significant efficacy in reducing labeling errors (as in aviation) [23]. Root cause analysis performed in CV1 ultimately determined that the error occurred during the initial specimen

Patient Safety Issues in Pathology: From Mislabeled Specimens to Interpretation Errors

http://dx.doi.org/10.5772/intechopen.79634

147

Due to the potential for substantial downstream impact of erroneous labeling and thus the generation of incorrect pathology/laboratory reports, these events warrant an expanded discussion (**Figure 1**). It has been reported that specimen labeling errors tend to be evenly distributed among the processes of accessioning, gross pathology processing, and tissue cutting, with some additional errors being identified in subsequent steps of processing [24]. Approximately 1.3% of these errors affect patient care [24]. The many "moving parts" within the pathology/specimen processing workflow may be subject to significant risk of errors and "near misses" [25]. The emergence of adverse or "never events" in patient care typically involves multiple breakdowns in both systemic and individual processes (the phenomenon known as the "Swiss cheese model") [17, 26, 27]. Failure to recognize errors in multiple successive steps of specimen preparation and interpretation can result in significant errors and resultant patient harm, as demonstrated in CV1. A proactive and critical review of processes

The inherent complexity of multi-step processes is implicated in the genesis of pathology errors. Lack of adequate coordination and/or communication is often cited in this context. Lapses in communication are among the most common sources of medical error, with over 20% of cases identifying communication errors as directly contributing to wrong site, wrong procedure, and wrong patient surgical procedures [29], and there are numerous calls for improvements in this area throughout all specialties [8, 17–19, 30]. CV1 highlights a breakdown in communication, and the importance of cross checks and verifications used for initial error rectification. Every critical communication carries a risk of error, but at the same time, it presents an opportunity for detection of error. For example, preoperative checklists and surgical "time-outs" have been shown to make operative care safer [31, 32]. A similar framework for preventing "never events" may also be effective in reducing pathology labeling errors [33]. Moreover, the initial errors that may have occurred during initial specimen processing in CV1 may have been compounded by other errors, including potential oversight issues from downstream employees who were under time constraints/heavy workloads thereby failing to

processing stage due to a clerical mistake [16].

may aid in reducing the incidence of such events [28].
