Applied Evidence

Diagnostic challenges in primary care: Identifying and avoiding cognitive bias

Author and Disclosure Information

 

References

A full history and physical exam revealed a myriad of gastrointestinal (GI) complaints, such as diarrhea. But the PCP recalled a recent roundtable discussion on debiasing techniques specifically related to gynecologic disorders, including OC. Therefore, he decided to include OC in the differential diagnosis—something he would not routinely have done given the preponderance of GI symptoms. Despite the patient’s reluctance and time constraints, the PCP ordered a transvaginal ultrasound. Findings from the ultrasound study revealed stage II OC, which carries a good prognosis. The patient is currently undergoing treatment and was last reported as doing well.

Early signs of ovarian cancer can be chalked up to a “GI issue”

OC is the second most common gynecologic cancer21 and the fifth leading cause of cancer-related death22 in US women. Compared to other cancers, the prognosis for localized ­early-stage OC is surprisingly good, with a 5-year survival rate approaching 93%.23 However, most disease is detected in later stages, and the 5-year survival rate drops to a low of 29%.24

There remains no established screening protocol for OC. Fewer than a quarter of all cases are diagnosed in stage I, and detection of OC relies heavily on the physician’s ability to decipher vague symptomatology that overlaps with other, more common maladies. This poses an obvious diagnostic challenge and, not surprisingly, a high level of susceptibility to cognitive bias.

Most PCPs correctly identified bloating as a key symptom of ovarian cancer; however, they weren’t as good at identifying less common symptoms, such as inability to finish a meal.

More than 90% of patients with OC present with some combination of the following symptoms prior to diagnosis: abdominal (77%), GI (70%), pain (58%), constitutional (50%), urinary (34%), and pelvic (26%).25 The 3 most common isolated symptoms in patients with OC are abdominal bloating, decrease in appetite, and frank abdominal pain.26 Patients with biopsy-confirmed OC experience these symptoms an average of 6 months prior to actual diagnosis.27

Knowledge gaps play a role. Studies assessing the ability of health care providers to identify presenting symptoms of OC reveal specific knowledge gaps. For instance, in a survey by Gajjar and colleagues,28 most PCPs correctly identified bloating as a key symptom of OC; however, they weren’t as good at identifying less common symptoms, such as inability to finish a meal and early satiety. Moreover, survey participants misinterpreted or missed GI symptoms as an important manifestation of early OC disease.28 These specific knowledge gaps combine with physician errors in thinking, further obscuring and extending the diagnostic process. The point prevalence for OC is relatively low, and many PCPs only encounter a few cases during their entire career.29 This low pre-test probability may also fuel the delay in diagnosis.

Watch for these forms of bias. Since nonspecific symptoms of early-stage OC resemble those of other more benign conditions, a form of anchoring error known as multiple alternatives bias can arise. In this scenario, clinicians investigate only 1 potential plausible diagnosis and remain focused on that single, often faulty, conclusion. This persists despite other equally plausible alternatives that arise as the investigation proceeds.28

Affective error may also play a role in missed or delayed diagnosis. For example, a physician would prefer to diagnose and treat a common GI illness than consider OC. Another distortion involves outcome bias wherein the physician gives more significance to benign conditions such as irritable bowel syndrome because they have a more favorable outcome and clear treatment path. Physicians also favor these benign conditions because they encounter them more frequently than OC in the clinic setting. (This is known as availability bias.) Outcome bias and multiple alternatives bias can result in noninvestigation of symptoms and inefficient or improper management, leading to a delay in arriving at the correct diagnosis or anchoring on a plausible but incorrect diagnosis.

Continue to: An incorrect initial diagnostic...

Pages

Recommended Reading

The vegan diet is popular but not automatically healthy
MDedge Family Medicine
Artificial sweeteners: A modifiable cancer risk?
MDedge Family Medicine
Sleep deprivation sends fat to the belly
MDedge Family Medicine
Anaphylaxis risk with IV iron low, but varies with formulation
MDedge Family Medicine
Does evidence support benefits of omega-3 fatty acids?
MDedge Family Medicine
Avocados linked to lower cardiovascular risk
MDedge Family Medicine
No link between cell phones and brain tumors in large U.K. study
MDedge Family Medicine
New HF guidelines feature ‘quad’ therapy, tweaked terminology
MDedge Family Medicine
CDC recommends hep B vaccination for most adults
MDedge Family Medicine
Study: Physical fitness in children linked with concentration, quality of life
MDedge Family Medicine