542 - Pertussis Update, Diagnostic Errors and Bias, How Certain Are You?
From the Centers for Disease Control and Prevention (CDC)
1) A Refresher on Pertussis
Pertussis has unfortunately been making an on-again-off-again resurgence over the last decade or so. As we know, it can cause a lingering “bronchitis” in adults and older children, but in infants it can be deadly. The CDC’s Advisory Committee on Immunization Practice (ACIP) has basically allowed repeat Tdap vaccination (rather than 1 dose in adulthood followed by Td vaccinations) and encouraged pregnancy vaccination and “cocoon vaccination” (vaccinating all the newborn infant’s close contacts) around expectant mothers – all with the goal of decreasing infant exposure and illness. There is an outbreak currently in North Dakota and early reports of an outbreak in the New River Valley here in Virginia.
Given this, some clinical pointers from the CDC to remember:
· Exposure to symptom onset is usually 4-10 (max 21) days.
· Pertussis is a three-phase illness:
o Coryza – 1-2 weeks – lots of congestion, mild cough. In infants – apnea.
o Paroxysmal – 4-5 weeks – paroxysmal coughing, inspiratory “whoop”, oxygen desaturation, dehydration
o Recovery – 2-3 weeks, symptoms slowly improve.
· Antibiotic treatment is best in the first two weeks of illness prior to the onset of paroxysmal coughing in order to improve symptoms. After this point, there is less benefit to antibiotics. Azithromycin is the antibiotic of first choice. (clarithromycin, erythromycin, and trimethoprim/sulfamethoxazole (for age>2 years) are alternatives).
o Azithromycin is given at a dose of 10 mg/kg/day for 5 days for children less than 6 months. Above six months, the child dose is 10 mg/kg on the first day, then 5 mg/kg/day for 4 more days (max 500 mg/dose). For adults, 500 mg at onset and 250 mg/day for 4 days is recommended.
o For infants and pregnant women, treat up to 6 weeks after cough onset, but for everyone else, treat within 3 weeks of cough onset.
· The trickiest part is differentiating URI symptoms from early pertussis. A history of contact with individuals with a severe or lingering cough, mild fever and lots of congestion are most consistent with early pertussis.
· Testing is the preferred method to confirm diagnosis prior to treatment. Most available are nasopharyngeal swabs for pertussis PCR testing.
· However, if you have clinical suspicion in an infant or other high-risk individual, it is recommended to start empiric azithromycin until the test has returned, given the benefits of early treatment.
· Use “droplet precautions” in practice when evaluating suspected patients – wash hands before and after examination, use a mask and faceshield/goggles
· Hospitalization may be required for anyone with pertussis (usually children) if their oxygenation is low, or if they are dehydrated.
· Post-exposure prophylactic antibiotics should be limited to those at high-risk of complications from pertussis (mainly infants), but it is best to coordinate this with local public health or infection control officials.
· Be sure to report positive pertussis tests to your local health department immediately. It is best that the clinician is involved in that report to provide necessary clinical details.
John’s Comments:
While treating possible high-risk pertussis cases early is a goal, be careful not to fall into overusing antibiotics. Look at the whole picture – likelihood of infection and risk from infection – to make your decision. Outbreaks of pertussis represent a failure of population levels of immunization, so make sure to get people vaccinated with Tdap as indicated whenever possible.
Reference:
· Pertussis (Whooping Cough) Clinical Information | CDC. Published August 25, 2022. Accessed April 23, 2024. Link
From the Agency for Healthcare Research and Quality (AHRQ)
2) Preventing Diagnostic Errors Through “Bias Management”
Though estimates vary, diagnostic errors in primary care are thought to be more common than we should be comfortable with. One challenge is that unlike therapeutic errors, diagnostic errors are more difficult to determine/measure. Adverse events related to misdiagnosis are more likely to be judged preventable than other types of adverse events such as medication errors. It is thought that approximately 75% of diagnostic errors have a cognitive component.
Cognitive biases are systematic cognitive dispositions or inclinations in thinking and reasoning that often do not comply with the tenets of logic, probability reasoning, and plausibility. At least 188 of these biases have been described. These intuitive and subconscious tendencies are foundational to human judgment, decision making, and the resulting behavior. Psychological frameworks consider biases as resulting from the use of (inappropriate) cognitive heuristics (shortcuts) that people apply to deal with data-limitations, information processing limitations, or lack of expertise. The two overarching cognitive components that are common within medical decision-making errors include:
· the tendency to seek only as much information as necessary to form an initial clinical impression, and
· the tendency to stick with the initial impression even as new information becomes available.
These decision-making errors are expressed through the following biases:
· Confirmation Bias – The tendency to selectively seek information that supports initial impressions.
· Overvalue Bias – The tendency to overvalue irrelevant information if it has been deliberately sought by the clinician. This bias compounds confirmation bias, because the clinician first seeks irrelevant information, then systematically overvalues this irrelevant information when it is obtained.
· Anchoring Bias – The inadequate adjustment of probabilities as new disconfirming information becomes available.
· Status Quo Bias – The tendency to stick with initial impressions as the number of new possible alternative diagnoses increases.
· Framing Effect – The tendency to be affected by how information is framed or presented. For example, by being informed of someone else’s conclusions about the information.
· Diagnostic Overshadowing Bias – The attribution of symptoms to an existing diagnosis rather than a potential co-morbid condition. This occurs most commonly in patients with psychiatric conditions.
· Implicit Bias – The unconscious attitudes, thoughts, and feelings about a particular people or group that can drive stereotypes and prejudices against that group.
· Action Bias – The tendency to act when faced with a problem even when inaction would potentially be more effective, or to act when no evident problem exists.
Given the prevalence of diagnostic errors, studies have explored interventions for prevention. Some cognitive- and system-based interventions include:
· Cognitive awareness – Includes efforts to teach trainees and clinicians about the diagnostic thinking process and methods to improve it through conscious awareness and questioning. This may also be improved through teamwork and case discussions. While clinicians are limited in their ability to self-monitor their thought processes (“Blind Spot Bias”), their colleagues might be willing and able to do so.
· System-based improvements – Structured diagnostic assessments for common clinical scenarios (e.g., chest pain, fever in an infant) can ensure that relevant findings are elicited, and common conditions considered. These can be augmented by computer-assisted decision support systems that advise or provide guidance about a particular clinical decision at the point of care.
Mark’s Comment:
The potential for bias is quite real in the practice of medicine, and the denial of it does not make it go away. We often don’t appreciate the demanding mental discipline required for effective clinical practice, even for what might appear to be common and perhaps even “mundane” issues. Adding negative emotion, either what we bring into the exam room and/or what is precipitated by our patients, only adds to the need for this discipline. Consider what biases you might be most susceptible to and strategies you are using/might use to overcome such biases.
References:
· Etchells E. Anchoring Bias With Critical Implications. AHRQ Web M+M. Published Online June 2015. Link
· Lazris A, Ross A. Lown Right Care: Diagnostic Overshadowing: When Cognitive Biases Can Harm Patients. Am Fam Physician. 2023;108(3):292-294. Link
· Doherty T and Carrol A. Believing in Overcoming Cognitive Biases. AMA J Ethics. 2020;22(9):E773-778. Link
From PeerRxMed ( www.PeerRxMed.org )
3) How Certain Are You Doctor?
“Medicine is a science of uncertainty and an art of probability.” Sir William Osler
From my formative days in medical training, the principles captured in Williams Osler’s above quote were instilled in me regarding medicine being both a “science” and an “art” and therefore the importance of having the professional humility to acknowledge that much of what we do is filled with uncertainty. Indeed, I have found myself understanding that the reverse of his quote is also true – when it comes to the practice of medicine, there is “science and art” to both uncertainty and to probability. Yet applying both the art and science of uncertainty and probability to clinical practice is not as easy as this quote might make it appear.
For example, recently I saw a woman in her 50s previously unknown to me who had been experiencing “chest pain” intermittently over the past several days. While she had some cardiac risk factors, she also had many other risk factors for both musculoskeletal pain as well as GERD, and both the history and exam were consistent with a non-cardiac source. Her EKG was normal, and a review of her chart indicated she had a negative “cardiac work-up” a year ago for similar symptoms.
Feeling convinced that this her pain was non-cardiac and that empiric treatment and “tincture of reassurance” were the appropriate next steps, I discussed my findings and recommendations with her. She listened carefully and then posed the question that whenever asked tends to pierce our clinical armor and leave us reflexively shifting into defensive mode by either going on a test-ordering spree or digging in emotionally to our position (or both): “How certain are you doctor?”
In that nanosecond of processing, somehow my mind found an alternative, wiser path, and rather than reactively feeling the need to answer her question, I responded with a question of my own. “What has you most concerned?” I asked, expecting her to say that if I couldn’t be completely sure then more testing was necessary. Instead, she answered, “I had a physician previously tell me when I had similar symptoms that if we couldn’t be absolutely sure, then we needed to do more tests. I’d prefer not to do that as it caused both great discomfort and significant expense for me.” I smiled and felt almost as if the spirit of William Osler was standing behind me, smiling as well. “Yes,” I replied, “I can understand how that might have happened. Let me tell you why I don’t think that’s necessary today.” She left satisfied with our plan, and her symptoms subsequently resolved.
We live on the edge of uncertainty and precipice of probability every day in our work. We don’t, however, often talk about how best to navigate it and the psychological toll that it can take. Because of that, it can weigh us down without our even being consciously aware of it. How about you? How do you navigate the art and science of uncertainty and probability in your clinical work? This certainly seems like a challenge (and even burden) worth exploring and sharing with a colleague, perhaps starting with your PeerRx partner. Afterall, in the world of professional uncertainty that we all navigate on a daily basis, no one should care alone.
Feel free to forward Take 3 to your colleagues. Glad to add them to the distribution list.