150,000 Medical Diagnostic Errors Estimated Per Year

Study Finds Doctor/Patient Communications Largely to Blame

By Teresa McLoughlin Rice, Esq.

Over 100 years ago the esteemed physician, William Ostler, advised medical students to listen to the patient and the patient would give them the diagnosis. A recent study has confirmed the enduring nature of this advice. Undertaken by Dr. Hardeep Singh, MD, MPH and colleagues, and published in the March edition of JAMA Internal Medicine, the study Types and Origins of Diagnostic Errors in Primary Care Settings, concluded that in confirmed cases of diagnostic errors in primary care settings (generally defined as missed, delayed or wrong diagnoses) a staggering 78.9% were due, in part, to mistakes made in the doctor-patient clinical encounter. Given the time constraints inherent in the practice of modern day primary care medicine, and the fact that in the workers’ compensation system, there are no actual face-to-face clinical encounters with doctor and patients for purpose of utilization review and independent medical review, this finding has generated much discussion. As noted by Dr. Singh, the office visit “is where the doctor-patient dialogue is so important, and that dialogue is getting shorter and shorter over time as we spend more time with the computer and more time doing administrative tasks. In general, we’re talking less to patients, and those skill sets and techniques of getting the history and examination of the patient are going a bit downward.” (O’Reilly, Kevin B. (2013, Mar 11) Although the rate of misdiagnosis is low, doctors are advised how to minimize the impact of briefer patient visits to boost the quality of care. (“O’Reilly”)

The Scope of the Problem

The scope of the problem was underscored in an Invited Commentary to the study, Measuring Diagnostic Errors in Primary Care, The First Step on a Path Forward, written by Dr. Newman-Toker and Dr. Makary. Extrapolating the results of the Singh study, they estimated that “more than 150,000 patients per year in the United States might have undergone preventable misdiagnosis-related harm.” Given these statistics, the Singh study is noteworthy as it looks at which diseases are most often missed and explores “why” and “how” this occurs. The ultimate goal of the study was to understand the scope and cause of the problem in order to best determine how the situation may be remedied. Although some relevant data previously existed, it was derived mainly from studies of malpractice claims, peer reviews or autopsy reports. Each of these, according to Singh, presented with its own biases. For example, misdiagnosis of cancer was considered to be among the most common diagnostic error, but this was arguably due to its overrepresentation in studies of malpractice claims. To counter these biases, Singh’s study instead reviewed the medical records of diagnostic errors found at two sites (a VA Hospital and a large integrated private health care system) through the lens of electronic health record-based triggers. The triggers were of two discrete types: (a) an unplanned hospitalization within 14 days of a primary care visit; or (b) a second (or more) primary care visit, emergency room or urgent care visit within 14 days. After reviewing electronic medical records of 212,165 visits at both sites, the study found instances of “diagnostic error” – in other words, a missed opportunity to make an earlier diagnosis – in 190 cases.

In nearly all these cases of misdiagnosis (96%) there was a “clear relationship” between the admission or second visit and the patient’s presentation on the initial visit. In fact, in nearly two-thirds of the cases (67.4%) the chief presenting symptom was “directly related” to the missed diagnosis, yet in nearly half (48.9%) of these cases there was inadequate scrutiny and analysis of the symptom. So why was this happening? In particular, what was going wrong in the doctor/patient interview such that 8 in 10 cases of misdiagnosis were found to originate there? According to the study’s results, the top two culprits were breakdowns in ordering diagnostic tests (57.4%) and failure to obtain an accurate medical history (56.3%). These were closely followed by an error in the physical examination (47.4%). Another 15% of errors were occasioned by a failure on the part of the physician to review previous documentation.

Probing further, the study found that most errors were “cognitive” – in other words related to data-gathering and synthesis problems. Of interest, where doctors “copied and pasted” previous progress notes into their reports (a relatively small 7.4% of cases), this contributed to more than a third of errors in cases in which a misdiagnosis was recorded.

It should be noted that although the lion’s share of misdiagnosis was attributable in part to some breakdown in the doctor/patient interview, the study also found errors in referrals (19.5%), patient-related factors (16.3%), follow-up and tracking of diagnostic information (14.7%), and performance and interpretation of diagnostic testing (13.7%).

Most Common Conditions Misdiagnosed

The second prong of the study focused on which specific conditions were misdiagnosed. It found that a wide range of conditions was impacted. Most are conditions common in a primary care setting. They include:

> Pneumonia (6.7%)

> Congestive heart failure (5.7%)

> Acute renal failure (5.3%)

> Cancer (5.8%)

> Urinary tract infection (4.8%)

As the study cogently notes, while pneumonia and congestive heart failure were the most commonly missed, they accounted for less than 13% of errors. Put another way, a wide variety of conditions are misdiagnosed and in fact run the gamut from asthma to psychiatric disorder to hypertension. While it is true that the study found diagnostic errors in relatively few instances (0.1% of cases reviewed), the study also made the sobering finding that 14% of the misdiagnoses had the potential to lead to immediate or inevitable death, and 19% had the potential to lead to serious permanent damage.

Limitations of Study

Dr. Singh did note some limitations to the study. It was retrospective in nature and “hindsight bias” therefore cannot be discounted. Nor were the researchers able to “debrief” the physicians, an exercise which could have yielded additional information. Additionally, the study captured only acute conditions that presented within 14 days of the initial visit. Other conditions, such as cancer, which would emerge after a longer period of time, are likely underrepresented. Nevertheless, the study notes that this is an area ripe for further analysis and further notes that it is important to focus on all errors uncovered so that steps may be taken to prevent future harm.

The Debate Over How to Remedy This Problem

Confirming that a problem exists, and shedding some light on its root causes are only the first steps – the next is to address remedies. Eschewing the term “error” in favor of “missed opportunities in care”, the study recommends further follow-up in the areas of why practitioners may not adequately search for data and how best to improve clinical skills and cognitive process. Dr. Singh rejects the notion that “disease-specific” efforts to reduce diagnostic errors would be sufficient, and posits instead that since such a wide range of diseases are misdiagnosed, honing in on just a few will not adequately address the scope of the problem. Others note that this approach may be difficult in today’s primary care practice. As the invited commentary notes “if we are to improve diagnostic accuracy without dramatically reducing efficiency, there may be no substitute for problem-specific solutions – in other words, for having knowledge and experience linked directly to presenting symptoms and diseases.”

The Singh study has generated a lot of discussion, which in turn has led to suggestions from others as to how misdiagnoses can be reduced. There seems to be a growing consensus that “situational” or “contextual” awareness is of utmost importance. In other words, the physician should be aware of a patient’s individual background (employment, financial status, “health literacy” etc.) and use this information to inform the diagnosis. (Patient Safety Solutions, August 10, 2010, It’s Not Always About the Evidence (“Patient Safety Solutions”)) This was borne out by the Singh study, which showed noticeable differences in the most common missed diagnoses between the VA Hospital and the integrated private health care system. Dr. Singh et al attributed these differences to “local contexts” and “patient and practitioner populations.” Even so, at both locations, the most common source of error was still the doctor/patient encounter.

There is no question that the healthcare model is evolving. Dr. Singh et al urge that in this age of increasing reliance on technology and team-based care, basic skills such as gathering medical history, performing an adequate physical examination and synthesizing the data cannot be ignored. They recommend that further thought be given as to how technology can be leveraged and woven into traditional skills to improve rapport and communication at the most basic doctor/patient level. Other experts caution that “as we strive to standardize care, we must not sacrifice the need to individualize care for each patient.” (See, Patient Safety Solutions) Technology is certainly an asset towards this goal. For example, and as noted in the invited commentary to the study, electronic health record systems could be enabled to monitor diagnostic performance, practitioners could be required to record symptoms rather than just diagnoses in the electronic health records, and simulation software could be used for students to complete self-tutorials for specific problem types.

Nonetheless some experts posit that to boost “situational awareness”, technology alone may not be enough. They urge that communication skills be improved by a variety of different approaches including:

> Training medical students in interview techniques so that they know how to engage the patient and listen without interrupting while he explains to the doctor why he is there. (Institute for HealthCare Communication, July 2013, Communication Matters, Taking Time to Save Time – and Lives!)

> Promoting multiple avenues of communication between the doctor and patient. One suggestion is to involve their whole team – a patient may confide something to a nurse or clerical assistant that was not confided to the doctor. (See Patient Safety Solutions)

> Ensure that the doctor is aware of a patient’s financial situation – it is no good prescribing medication if the patient is not filling it because of cost concerns. (See Patient Safety Solutions)

The Singh study is a powerful reminder that William Ostler’s advice is as relevant today as it was a century ago, and that even as technology is used to promote efficiencies, the basis human skill of communication remains indispensable.

© Copyright 2013 LexisNexis. All rights reserved.

___________________________________________________________________

Special Discount Price $79*; Books shipping now to customers!

Keep track of how the workers' comp landscape is changing with this 400+ page compendium. Here's what you get:

• A 50 state survey at a glance of workers' comp-related legislation, including selected drug bills, with commentary from 27 defense attorneys, 16 claimant's attorneys, and National expert Thomas A. Robinson, stafff writer for Larson's Workers' Compensation Law
• In-depth analysis and insight on key issues, including exclusive remedy, medical marijuana, opt outs, Affordable Care Act & much more
• Larson's Spotlight on interesting cases for 2013, written by Thomas A. Robinson

View the brochure & table of contents.

View sample pages.

Order online or contact Christine Hyatt at ph. 937-247-8166, or Email: Christine.E.Hyatt@lexisnexis.com.

PROMO CODE: WCEIA

*Price does not include sales tax, shipping or handling. Price subject to change without notice. Discount cannot be combined with other offers. Expires 12/31/2013.

For more information about LexisNexis products and solutions connect with us through our corporate site