Scott Hasler

Diagnostic Error in Medicine: Analysis of 583 Physician-Reported Errors

BACKGROUND Missed or delayed diagnoses are a common but understudied area in patient safety research. To better understand the types, causes, and prevention of such errors, we surveyed clinicians to solicit perceived cases of missed and delayed diagnoses. METHODS A 6-item written survey was administered at 20 grand rounds presentations across the United States and by mail at 2 collaborating institutions. Respondents were asked to report 3 cases of diagnostic errors and to describe their perceived causes, seriousness, and frequency. RESULTS A total of 669 cases were reported by 310 clinicians from 22 institutions. After cases without diagnostic errors or lacking sufficient details were excluded, 583 remained. Of these, 162 errors (28%) were rated as major, 241 (41%) as moderate, and 180 (31%) as minor or insignificant. The most common missed or delayed diagnoses were pulmonary embolism (26 cases [4.5% of total]), drug reactions or overdose (26 cases [4.5%]), lung cancer (23 cases [3.9%]), colorectal cancer (19 cases [3.3%]), acute coronary syndrome (18 cases [3.1%]), breast cancer (18 cases [3.1%]), and stroke (15 cases [2.6%]). Errors occurred most frequently in the testing phase (failure to order, report, and follow-up laboratory results) (44%), followed by clinician assessment errors (failure to consider and overweighing competing diagnosis) (32%), history taking (10%), physical examination (10%), and referral or consultation errors and delays (3%). CONCLUSIONS Physicians readily recalled multiple cases of diagnostic errors and were willing to share their experiences. Using a new taxonomy tool and aggregating cases by diagnosis and error type revealed patterns of diagnostic failures that suggested areas for improvement. Systematic solicitation and analysis of such errors can identify potential preventive strategies.

Organizational and cultural changes for providing safe patient care.

Objective To describe an approach and experience with fostering a culture of patient safety. Methods (1) Organizational Change—The Department of Medicine established a patient safety committee (PSC) and charged it with reviewing adverse events. (2) Cultural Change—PSC sponsors and participants work to promote a culture of collaboration, study, learning, and prevention versus a culture of blame. (3) Collaboration—The PSC includes chief residents and members from medical informatics, nursing, pharmacy, quality assurance, risk management, and utilization management. (4) Evolution—The duties of the PSC progressed from merely learning from adverse event reports to implementing patient safety and quality improvement projects. (5) Standardization—The PSC uses standard definitions and procedures when reviewing cases of adverse events, and when conducting patient safety and quality improvement projects. Results (1) Developed an online adverse event reporting system, shortening the average report collection time by 2 days and increasing the number of adverse events reported. (2) Established a model for change using (a) safety rounds with residents, (b) e-mail safety alerts, and, in some cases, (c) decision alerts using electronic order entry. These changes in culture and capability led to improvements in care and improved financial results. Conclusions Senior management support of a culture of learning and prevention and an organizational structure that promotes collaboration has provided an environment in which patient safety initiatives can flourish by providing not only safer and higher quality patient care but also a positive financial return on investment.

An exploratory analysis of the correlation of pain scores, patient satisfaction with relief from pain, and a new measure of pain control on the total dose of opioids in pain care.

Objective: We explored the associations between opioid dose and multiple measures of pain. Study Design and Measures: Thirty-two consecutive patients admitted solely for an acute exacerbation of cancer-related pain or for surgery were followed for their entire hospital stay (115 days of pain). For each hospital day, we collected pain scores, the number of pain scores, trends in pain scores, the percentage of time patients had 100% acceptable relief from pain, and the number of times patients were asked about acceptable pain relief. Finally, we asked those who had 100% relief of pain whether they could have used more pain medicine. Linear regression models were fit to estimate the amount of variation explained (R2) in dose of medication, by each pain measurement variable. Results: Nineteen patients with cancer (74 days of pain) and 13 patients undergoing surgery (41 days of pain) were evaluated. Pain scores, the number of pain scores, trends in pain scores, and 100% acceptable relief scores poorly correlated with the use of medication in the linear regression models (R2 for all models ⩽0.2). A question about needing more pain medicine explained the greatest amount of variation in opioid dose. Conclusions: Pain and acceptable relief scores do not adequately reflect the use of medication. A prospective study is needed to further assess the value of additional measures of the adequacy of pain care.

The hospital is not your home: making safety safer (Swiss cheese is a culinary missed metaphor).

H euristics are mental shortcuts, generalizations used because our brains are not calculators of any robust consequence. The “availability heuristic,” for example, helps us assess the relative frequency of events. After observing many occurrences of some event, yet few of another, we judge the relative likelihood of those observed events. Unfortunately, the value of such generalizations is undercut by their capacity to fail us. For example, some events bring with them intense emotion; the more emotional an event is the more “available” that event becomes, subsequently inflating the event’s presumed frequency.1 Thus, generalizations may lead to inaccuracies, so, we must be careful in proposing heuristics about how to make care safer for patients. The literary equivalent to a heuristic is a metaphor. A metaphor is another generalization that frames how we think about a problem. Metaphors about safety seem especially important to get correct: we do not want an industry devoted to making care safer to follow a metaphor that leads to worse care. To be sure, there is concern that efforts to improve patient safety are falling short.2 Hospitals that perform on a continuum from well to not so well on measures that are supposed to reflect ways of getting safer, do not seem to differ much in outcomes of care.3 Some feel that the reason we have not made care safer is that the measures we use to monitor our care are poor predictors of safe actions, and propose an alternative model of advancing safety.4 However, perhaps, if we are not advancing safe care at the pace we would like, we should reexamine our deeply held metaphors about how to make care safer. Two metaphors used to rally us to an understanding of why adverse, unsafe events happen, and used also to propose ways to make care safer are the “Swiss cheese” model and that “system” is the root of unsafe acts. We think the first is an inaccurate generalization of why care is unsafe and that together these two metaphors have led to more complex rather than simple care. The Swiss cheese metaphor has been shown to be, at least, confusing. In a survey of safety experts, of the 11 response items that were compatible with the model, only 5 were endorsed.5 If there is confusion about what a metaphor is saying, the metaphor is inadequate. The Swiss cheese model suggests that in a complex system, barriers between system components prevent errors from occurring. But, like all barriers, each has a “hole,” or weakness, and, under certain circumstances, the holes align and an adverse event leaks through. This means, then, to us, that an adverse event is “bad luck,” a sort of misalignment problem. This model also suggests that adding more cheese will make alignment less likely. To be safer, all we need to do is add another barrier and make sure that its holes do not align with another barrier’s hole. This is the most problematic interpretation—and it flies in the face of safety. Safety is about finding a single process constraint to the delivery of safe care and then simplifying by either bypassing or reconstructing the poorly functioning constraint.6 It is a worn out example, but if one step works 90% of the time, adding

An Interferon-Induced Digital Vasculitis-Like Syndrome: A Case Report.

This report describes a patient with chronic hepatitis C undergoing therapy with interferon (IFN) alpha who developed bilateral ischemia of his fingers. We present a 43-year-old man with a failed renal transplant and chronic hepatitis C. He was treated with 6 months of IFN therapy with good reduction of his viral load. He presented with 2 days of pain and swelling in the second digits of both hands. Workup for extrahepatic manifestations of hepatitis C was initiated including assessment for vasculitis because of cryoglobulin- and noncryoglobulin-related causes. Extensive assessment with invasive and noninvasive vascular testing was performed. His workup for vasculitis did not reveal any specific reasons for the ischemic changes. Angiography of his fingers showed mild stenotic changes but no evidence of systemic vasculitis. IFN therapy was stopped and over several weeks his symptoms resolved. The ischemic changes were attributed to IFN therapy. The patient in this report is unique because although IFN has been historically reported to cause a variety of vascular syndromes, the reported experience in hepatitis C patients is small. In addition, the likelihood of encountering vasculitis and vasculitis-like syndromes in patients with hepatitis C is significant, and the increasing use of IFN in this population makes drug-induced vascular changes an essential consideration in this subset of patients.