Forums » Oncology

Diagnostic Errors & The Impact of Second Opinions

  • April 1, 2022 12:41 PM EDT

    Seeking a second opinion is commonplace in human medicine. The reasons that patients seek second opinions vary, but they are a means to gaining an additional opinion on a diagnosis, proposed treatment, or prognosis from another practitioner. After having a biopsy performed and receiving a pathologist's report, patients most commonly seek second opinions for diagnoses that have high rates of morbidity and mortality, in instances where their symptoms persist, when they are hoping for a change in the diagnosis (e.g. a change from malignant to benign), or they want to confirm a diagnosis. Often, the primary goal is to find out as much as possible about their disease or condition in advance of determining their treatment plan.


    An evaluation of 6791 second opinions sought by human patients in the US1 found that people requested a second opinion most commonly in orthopedic surgery (n=1195) and medical oncology (n=588) cases for the following reasons: 

    - to help them choose between treatment options (41.3%)

    - there was no improvement in their symptoms (22.5%) 

    - they questioned the need for surgery (16%) 

    - they did not understand the diagnosis or did not trust the physician (both 6%)

    In another study of human patients,2 dissatisfaction with their doctor's communication (poor or incomplete explanations, ineffective listening, and not providing reassurance) resulted in patient-initiated second opinions.  

     

    Biopsies and Second Opinions

    Obtaining a second opinion on histopathology is a standard practice in human medicine. This is done to confirm a diagnosis and ensure that the correct treatment is instituted. Uncovering an erroneous diagnosis, after all, can profoundly impact patient management.

    Second opinion histopathology can result in one of three scenarios:

    • Complete agreement - The first and second opinions concur with each other. 

    • Partial disagreement - The second opinion differs from the first opinion, however these differences may have minimal to no impact on the overall prognosis. Examples include differences between tumor subtype, tumor grade, or in the assessment of tumor margins. These small differences do not change the overall diagnosis or prognosis for the patient. 
    • Complete disagreement - The second opinion completely disagrees with the first opinion. Examples include the tumor type (e.g. carcinoma versus a sarcoma) or aggressiveness of the tumor (e.g benign vs. malignant).


    The Impacts of Second Opinions on Patient Outcomes

    So what is the effect of a patient-initiated second opinion on outcome? It's been shown that there can be substantial changes in diagnosis, treatment, and an overall clinical impact by seeking a second opinion. 

    In that study of 6791 patients,1 patient-initiated second opinions led to recommended changes in diagnosis for nearly 15% of the individuals and alterations in treatment plans for just over 37% of the study participants. 

    Changes in Diagnosis and Treatment From Second Opinions

    • Diagnosis - change (14.8%), clarification (17%), confirmed (56.8%)
    • Treatment - change (37.4%), clarification (26.9%), confirmed (26.4)

    Clinical Impacts of Second Opinions

    • Diagnosis - major (2.7%), moderate (18.2%), minor (46.3%), none (27.9%)
    • Treatment - major (4.2%), moderate (26.5%), minor (50.1%), none (14.2%)



    Second opinions in veterinary medicine

    Although second opinions in veterinary medicine are not sought as frequently as in human medicine, the reasons for getting one are similar to those in humans. They are usually sought out for diseases that have a high morbidity and mortality rate like cancer. It's standard practice to obtain 2 opinions on histopathology samples in human medicine; This practice has been recommended in veterinary medicine as well.  

    One such recommendation came from a study published in 2009 by Regan, et. al.3 They retrospectively reviewed opinions in 430 cases (cats and dogs) to determine whether second-opinion histopathology directly affected patient care at their institution. They found diagnostic agreement in 70% of cases and a discrepancy or disagreement rate of 30% - partial diagnostic disagreement in 20%, and complete diagnostic disagreement regarding change in degree of malignancy or in cell type in 10% of cases. Overall, 17% of cases that had a second opinion histopathology review resulted in a change in treatment and/or prognosis for the patient. 

    A subsequent prospective study4 examined the impact of second-opinion histopathology on further diagnostic testing, cost and treatment of dog and cat patients with cancer. This study included the histopathologic evaluation of 52 tumors (48 from dogs and 4 from cats) over a 1-year period. There was diagnostic agreement in 52% of the cases, partial diagnostic agreement in 29% of cases, and a complete diagnostic disagreement in 19% of the cases.

    The majority of cases considered partially discordant in this study were due to a differing opinion in grade, tumor subtype or margin status. Reasons for a complete disagreement included differing opinion in cell type of origin or a change in diagnosis from benign to malignant.

    Disagreements were considered "minor" in 21% of the cases and did not impact treatment recommendations or prognosis. "Major" disagreements overall accounted for 37% of cases and did impact treatment recommendations or prognoses for those patients. Costs of staging and treatment recommendations were considerably different between first and second opinions.

    Obtaining a second opinion likely incurs additional costs for the pet owner, but in the grand scheme of things it can be considered nominal. In some cases, patients may pay more for this test and not derive a benefit. The benefit will likely be seen in those patients that would require a change in a proposed treatment or diagnostic staging plan - if, for example, a diagnosis changes from malignant to benign.

     

    Second opinion histopathology - Other considerations

    Consider the situation in which a veterinarian biopsies a lesion and receives a surprising diagnosis or one that does not match up with their clinical impression regarding that tumor's biology / behavior. An example is that of a rapidly growing mass and a corresponding biopsy report that indicates it is benign or low grade (e.g. grade I tumor). The report may not match up with the clinician's concern. The point of staging or grading a tumor is to predict how it's going to behave in terms of growth, metastasis, etc., as well as the patient's prognosis. And so if a report indicates a grade I tumor -  one that would not typically grow so quickly or be doubling in size (if that's the clinical impression) - that should prompt a clinician to seek a second opinion. If a pathologist fails to provide adequate information or comprehensive details (e.g. if tumor margins aren't recorded, or the actual degree of margin completeness is not recorded), that would be another situation to consider requesting a second histopathologic opinion.


    References:

    1. Meyer AN, Singh H, Graber ML. Evaluation of outcomes from a national patient-initiated second-opinion program. Am J Med. 2015 Oct;128(10):1138.e25-33. doi: 10.1016/j.amjmed.2015.04.020. Epub 2015 Apr 23. PMID: 25913850.

    2. Greenfield G, Shmueli L, Harvey A, et al Patient-initiated second medical consultations—patient characteristics and motivating factors, impact on care and satisfaction: a systematic review BMJ Open 2021;11:e044033. doi: 10.1136/bmjopen-2020-044033

    3. Regan RC, Rassnick KM, Balkman CE, Bailey DB, McDonough SP. Comparison of first-opinion and second-opinion histopathology from dogs and cats with cancer: 430 cases (2001-2008). Vet Comp Oncol. 2010 Mar;8(1):1-10. doi: 10.1111/j.1476-5829.2009.00203.x. PMID: 20230576.

    4. Regan RC, Rassnick KM, Malone EK, McDonough SP. A prospective evaluation of the impact of second-opinion histopathology on diagnostic testing, cost and treatment in dogs and cats with cancer. Vet Comp Oncol. 2015 Jun;13(2):106-16. doi: 10.1111/vco.12023. Epub 2013 Feb 19. PMID: 23421645.

    • 147 posts
    April 5, 2022 10:22 AM EDT

    Reading this triggered my recollection of stories from clients or conversations I've overheard between pet owners in the dog park - along the lines of ..."my vet told me he had "x" months to live, and that was over a year ago." These have never bode well for the veterinarian; the inference was that either the vet was wrong and incompetent or was trying to bilk the client to spend excessive amounts of money for additional work-up or treatment. I'm wondering how often veterinarians actually seek out a second opinion for histopath results.

    All of this underscores the takeaway from Dr. Kristin Zersen's recent discussion of the incidence of premature death in dogs with "benign" histopath findings after a splenectomy for nontraumatic hemoabdomen. Veterinarians need to be sharing this kind of information with their clients when discussing treatment options and the range of outcome possiblities.

    Coincidentally, this past week the New York Times had an article about the "medical gaslighting" of women by physicians. They referenced a 2020 publication in The Medical Journal of Australia that reported a 14% incidence rate of diagnostic errors - the publication states: "Some form of diagnostic error occurs in up to one in seven clinical encounters, and most are preventable." Another publication in JAMA Internal Medicine (2010) reported that upwards to 83.3% of diagnostic adverse events are preventable as over 96% were due to a human failure (e.g. a knowledge-based mistake or a mistake in the transfer of information). The consequences for diagnostic-related adverse events (of any type) resulted in a greater mortality rate (29.1% vs 7.4% for other types of adverse events).

    If we look to veterinary medicine and the broad topic of medical errors, this 2019 article published in Frontiers in Veterinary Science reported that 15% of medical errors resulted in patient harm, with drug and communication errors the most frequently encountered of errors in the veterinary practice setting.