Turning a blind eye on second opinions

By Warren R. Heymann, MD, FAAD
June 21, 2023
Vol. 5, No. 25

Whenever I encounter patients who ask for, or are being seen for, second opinions, I silently chuckle at Henny Youngman’s timeless one-liner about the subject (you can Google it). The fact is that second opinions are no joke. There are innumerable reasons why patients seek them ranging from diagnostic dilemmas to exploring therapeutic options.

I encourage my patients to get second opinions when I am unsure of the diagnosis or my therapeutic interventions have been inadequate. Some patients worry about offending me. Dr. Anees Chagpar states: “People are generally very nice and they don’t want to hurt the first doctor’s feelings.” (1) When I sense this, I immediately dispel it — I tell patients that all I want is for them to get better and the last thing I care about is my ego.

I am honored that colleagues will often refer patients to me for my opinion. (Interestingly, I have patients contact me from around the world seeking my advice because of a particular DWI&I column I wrote — I sheepishly inform them that I am not an expert on most of these topics and appropriately refer them elsewhere.) I have basic rules for complex consultations: 1) they must be performed with ample time allotted; 2) patients should bring their records; 3) I review these records after I perform my history and physical examination — I do not want to be biased by prior opinions. (Another rule — when I have a patient who tells me; “I’ve been to the Mayo Clinic, Cleveland Clinic, Lahey Clinic, Massachusetts General Hospital, Yale, and Penn — they could not help me — I heard you were great — you’re my last hope,” it is essential to set appropriate expectations. I will say: “If so many excellent doctors have found your case difficult, in all likelihood, I will as well. All I can promise is that I’ll give you my best effort.”)

Given the fast pace of most dermatology practices, diagnostic and therapeutic considerations are made quickly. According to dual process theory, clinical decision-making may be System 1 (intuitive, heuristic, and unconscious) or System 2 (analytical and logical conscious thinking). With experience, based on pattern recognition, most of us get through the day utilizing System 1; we will shift to System 2 when befuddled. Errors in clinical judgment are often a function of cognitive bias. In a survey of 387 emergency room physicians, the most common cognitive biases were overconfidence bias (the tendency to have an inaccurate and false opinion about one’s self), confirmation bias (the tendency to tweak the information to fit one’s hypothesis), availability bias (the tendency to instinctively think of what comes to mind easily as being more representative than in actuality), and anchoring bias (the tendency to adhere to the first idea without considering other possibilities). (2)

---

I cannot overemphasize the importance of attempting to reach your own conclusions first. I say attempting because it is not always easy. I will specifically ask the patients not to tell me what other doctors told them, “Do you mean that you don’t want to know that Dr. Ro thought this was subacute cutaneous lupus?” “Yes, Ms. La, that is exactly what I did not want to know!” There is evidence that knowledge of an initial diagnosis can bias the consulting practitioner who is providing the second opinion, especially when the providers have an established relationship. (3)

In pathology, inaccurate diagnoses may lead to inappropriate therapeutic interventions (or non-intervention). In a review of archived records of the Royal College of Pathologists of Australasia Quality Assurance Program (2015-2017), the rate of inaccurate diagnoses (assessed as a major discordance) ranged from 3% to 9% among different specimen groups, with the highest mean percentage of inaccurate diagnoses in gynecology, dermatopathology, and gastroenterology. (4) Pathologists with higher self-awareness (“metacognitive sensitivity”) are more likely to request additional tests or second opinions for their inaccurate diagnoses. (5) Peck et al note that second opinions reduce the incidence and necessity for amendments of original biopsy reports. They also assert: “While second opinions in challenging cases may be referred internally or externally, the latter has been particularly beneficial in detecting, and therefore preventing, major diagnostic errors. External referrals have been shown to have a five times greater sensitivity for detecting major discrepancies compared with internal reviews by peers. External second opinions (i.e., referrals) afford laboratories the distinct advantages of securing additional and specialist peer input, while removing any potential internal diagnostic bias or conflict of opinion between colleagues.” (4)

Clinical-pathological correlation is the optimal road to diagnostic success in dermatopathology. When I read a slide, initially I do not want to know the clinician’s impression — I will try to ascertain what I can; in cases that are not clear-cut, knowing the clinical impression will assist in histological diagnosis. When I am uncertain of the diagnosis, I will seek a second opinion. Elmore et al evaluated the extent to which dermatopathologists’ diagnoses are influenced by prior diagnostic information from another dermatopathologist. A total of 149 dermatopathologists provided 5,322 interpretations of melanocytic lesions. Participants were more likely to increase the severity of their diagnosis when the prior diagnosis was of greater severity compared with when no prior diagnosis was provided (RR, 1.52); likewise, participants gave less severe diagnoses when prior diagnoses were of lesser severity (RR, 1.38). Trends were similar among dermatopathologists who had previously stated they were “not at all influenced” by prior diagnoses. Prior diagnoses also swayed dermatopathologists away from correct diagnoses. The authors concluded that “in this randomized controlled trial, despite the preference of most dermatopathologists to receive prior diagnoses when providing second opinions, this information swayed them away from a correct diagnosis to an incorrect diagnosis.” The implication is clear — to obtain an independent second opinion, dermatopathologists should be blinded first. (6)

Point to Remember: When providing second opinions as a dermatologist or dermatopathologist, you may see a clearer picture if you are blinded first.

Our experts’ viewpoints

David A. Wetter, MD, FAAD
Professor of Dermatology
Mayo Clinic
Rochester, Minnesota

When I was a senior dermatology resident at Mayo Clinic, I saw a patient who had traveled a great distance to seek care for their complex multidisciplinary concerns. As part of their itinerary, the Mayo general internist had internally referred the patient to dermatology for a second opinion regarding their generalized, recalcitrant skin eruption associated with severe pruritus and dysesthesias. The patient was receiving outstanding care at a renowned tertiary referral center and under the guidance of an internationally recognized dermatologist. But because I was practicing at Mayo Clinic (and despite my inexperience and relative obscurity in the world of dermatology), the patient placed their complete trust in me and asked me to provide guidance to their local dermatologist regarding diagnostic and therapeutic recommendations. (The patient did not realize that as I briefly stepped out of the room during our visit, to seek answers for their questions, I was feverishly perusing a major dermatologic textbook in which their local dermatologist was a primary author.) This impressed upon me the innate trust and respect that patients had for me, not based on anything I had done to earn or deserve it, but solely because of the high reputation of Mayo Clinic (which had been built by generations of supremely talented physicians who had preceded me in the clinical practice). This experience is one of many that have taught me what an awesome privilege (and responsibility) it is to perform “second opinion” consultations at a quaternary-care center. I am immensely grateful to my patients, colleagues, and trainees who implicitly motivate me to constantly learn more, so that I can (hopefully) be a better dermatologist as my career continues. Even though some days I feel like I may not have any “expert” knowledge to offer, I hope that by being mindfully present during the visit, empathetically communicating with patients and their families, and listening carefully to their unique stories will allow patients (and their referring clinicians) to be pleased (most of the time) that I was a part of their care team. (7-9)

Margot S. Peters, MD, FAAD
Professor of Dermatology and Laboratory Medicine & Pathology
Mayo Clinic
Rochester, Minnesota

Many years ago, my first look on entering the room of a gowned patient was of her partially exposed back, revealing the serpiginous tracks of cutaneous larva migrans — one of the rare times I not only could help a patient who presented after several prior opinions but make the diagnosis blind to all information except the clinical morphology. Her painfully pruritic eruption had responded to therapy by the time slides from a skin biopsy arrived, which showed mixed dermal inflammation including eosinophils, a pattern “consistent with” cutaneous larva migrans but also many other inflammatory dermatoses. This story is not intended to minimize the value of blinded dermatopathology interpretation but to highlight the sometimes-underappreciated limitations of pathology as the gold standard of diagnostic truth.

As dermatologists, we have at our disposal the entire gross pathology, the complete exam, in addition to clinical history and laboratory data, but depending on factors such as size and site, sections from a punch or shave biopsy may or may not be representative of an inflammatory or neoplastic dermatosis — a form of blindness in dermatopathology practice that may influence the accuracy of diagnosis underpinning clinical management. Initial blinded review is thus just one and often not a critical part of primary interpretation or second dermatopathology opinions. An extreme but not rare example of biopsy-related blindness and limits of blinded review is Fleming’s published vignette on an ulcerated specimen submitted without clinical information, considered consistent with a traumatized nevus, with second opinion agreement — two weeks later, he was informed that the 1-mm specimen was from a 2-cm black plaque. (10) Blinded or nonblinded second opinion consults may help guard against misses and misjudgments, and demonstrate due diligence, but do not necessarily overcome challenges posed by the biopsy specimen, range of histopathological patterns per diagnosis, overlapping features among sometimes disparate disorders, diagnoses of low interobserver concordance, and the pathologist’s sense of pressure to provide a specific diagnosis on nonspecific histopathology or assumptions about how the report will be interpreted by the clinician. The gold standard of pathology is thus not 24 karats but involves a subjective of process with range of sensitivity and specificity depending on the diagnostic category and specimen at hand. Even with an optimal biopsy specimen, interpretation of melanocytic proliferations is fraught with problems best captured by the phrase “one dermatopathologist’s moderately atypical nevus may be another’s melanoma.” (11) In a large relatively blind study (only patient age and biopsy site and type provided), prompted by the first author’s experience as a patient who received 3 interpretations from benign to invasive melanoma on a skin biopsy, Elmore and colleagues found diagnostic accuracy rates of 92% for class I (nevus/mild atypia), 25% for moderate atypia (II), 40% for severe atypia/melanoma in situ (III), 43% for early-invasive melanoma (IV), and 72% for invasive melanoma > T1b (class V) — and intraobserverreproducibility of only 35% to 63% for class II-IV lesions. (12) A more recent study found 65.5% mean accuracy of single reviewer diagnosed melanocytic lesions, with nonblinded and blinded second opinions providing “modest” (69.9% and 69.2%, respectively) but hardly inspiring improvements. (13) At least in part because melanocytic lesions constitute about 25% of all skin biopsies (14), busy dermatopathologists and dermatologists adopt various approaches to deal with the fallibility of these diagnoses, e.g., moderate atypia means benign, thus of no consequence, or moderate atypia requires re-excision. Reports on nonmelanocytic pathology, such as a study finding only 35.6% agreement among 3 blinded dermatopathologists in diagnosis of nonmalignant skin ulcers (15), also highlight limitations of the gold standard.

Although often of unmeasurable value, initial blinded review is the pathologist’s only chance to formulate conclusions based on histopathology alone. Our dermatopathology group is fortunate to have adopted a digital pathology practice, which facilitates blinded view without knowledge even of the patient’s age or biopsy site, because the digital slide images launch before the software containing information on patient demographics, anatomic site of the specimen, and clinical impression — avoiding both patient information bias and previous opinion bias. But this purity of review quickly takes a back seat to the importance of considering all available information. Clinical-pathological correlation is indeed not only the optimal but sometimes the only road to diagnostic success in dermatopathology. Most gold jewelry is 14 karats, 58% gold, perhaps the average agreement rate among experts in the diagnosis of challenging abnormalities, which constitute most requests for second opinions. We need eyes and minds wide open, empathetic consideration of consequences for the patient, and a good dose of humility in rendering this gold diagnostic standard and using it to formulate management decisions.

1. Katella K. Can a second opinion make a difference? Yale Medicine. January 5, 2020. www.Yale medicine.org

2. K, Harada T, Watari T. Cognitive biases encountered by physicians in the emergency room. BMC Emerg Med. 2022 Aug 26;22(1):148. doi: 10.1186/s12873-022-00708-3.

3. Halasy M, Shafrin J. When Should You Trust Your Doctor? Establishing a Theoretical Model to Evaluate the Value of Second Opinion Visits. Mayo Clin Proc Innov Qual Outcomes. 2021 Apr 8;5(2):502-510. doi: 10.1016/j.mayocpiqo.2021.01.014. PMID: 33997646; PMCID: PMC8105538.

4. Peck M, Moffat D, Latham B, Badrick T. Review of diagnostic error in anatomical pathology and the role and value of second opinions in error prevention. J Clin Pathol. 2018 Nov;71(11):995-1000. doi: 10.1136/jclinpath-2018-205226. Epub 2018 Aug 1. PMID: 30068638.

5. Clayton DA, Eguchi MM, Kerr KF, Miyoshi K, Brunyé TT, Drew T, Weaver DL, Elmore JG. Are Pathologists Self-Aware of Their Diagnostic Accuracy? Metacognition and the Diagnostic Process in Pathology. Med Decis Making. 2022 Sep 20:272989X221126528. doi: 10.1177/0272989X221126528. Epub ahead of print. PMID: 36124966.

6. Elmore JG, Eguchi MM, Barnhill RL, Reisch LM, Elder DE, Piepkorn MW, Brunyé TT, Radick AC, Shucard HL, Knezevich SR, Kerr KF. Effect of Prior Diagnoses on Dermatopathologists' Interpretations of Melanocytic Lesions: A Randomized Controlled Trial. JAMA Dermatol. 2022 Sep 1;158(9):1040-1047. doi: 10.1001/jamadermatol.2022.2932. PMID: 35947391.

7. Prose NS. Some words that matter. Arch Dermatol. 2003; 139:21-22.

8. Nguyen TV, Hong J, Prose NS. Compassionate care: Enhancing physician-patient communication and education in dermatology. Part I: Patient-centered communication. J Am Acad Dermatol. 2013; 68:353. e1-8.

9. Zulman DM, Haverfield MC, Shaw JG, et al. Practices to foster physician presence and connection with patients in the clinical encounter. JAMA. 2020; 323:70-81.

10. Fleming MG. Pigmented lesion pathology: what you should expect from your pathologist, and what your pathologist should expect from you. Clin Plast Surg. 2010; 37(1):1-20. doi: 10.1016/j.cps.2009.07.003. PMID: 19914454.

11. Elston D, McNiff J, Maize J Sr. Histologically dysplastic nevi that extend to a specimen border. J Am Acad Dermatol. 2013; 68(4):682-683. doi: 10.1016/j.jaad.2012.10.061. PMID: 23522414.

12. Elmore JG, Barnhill RL, Elder DE, Longton GM, Pepe MS, Reisch LM, Carney PA, Titus LJ, Nelson HD, Onega T, Tosteson ANA, Weinstock MA, Knezevich SR, Piepkorn MW. Pathologists’ diagnosis of invasive melanoma and melanocytic proliferations: observer accuracy and reproducibility study. BMJ 2017; 357: j2813. doi: 10.1136/bmj.j2813.

13. Kerr KF, Longton GM, Reisch LM, Radick AC, Eguchi MM, Shucard HL, Pepe MS, Piepkorn MW, Elder DE, Barnhill RL, Elmore JG. Histopathological diagnosis of cutaneous melanocytic lesions: blinded and nonblinded second opinions offer similar improvement in diagnostic accuracy. Clin Exp Dermatol. 2022; 47(9):1658-1665. doi: 10.1111/ced.15219. Epub 2022 Jun 22. PMID: 35426450.

14. Lott JP, Boudreau DM, Barnhill RL, et al. Population-based analysis of histologically confirmed melanocytic proliferations using natural language processing. JAMA Dermatol. 2018; 154 (1):24-29. doi:10.1001/jamadermatol.2017.4060.

15. Hammer P, Latour E, Bohnett MC, McKenzie F, Korcheva VB, Mengden S, White KP, Ortega-Loayza AG. The utility and challenges of histopathologic evaluation in the diagnosis of nonmalignant skin ulcers. Wound Repair Regen. 2020; 28(2):219-223. doi: 10.1111/wrr.12780. Epub 2019 Nov 20. PMID: 31705777.

---