Delayed Diagnosis in the Setting of Virtual Care: Remembering the Physical Examination
A 71-year-old frail, non-ambulatory woman a with history of multiple sclerosis and neurogenic bladder presented to the emergency department (ED) of an academic medical center with fever of two days duration, sweating, and dry cough. Labs showed leukocytosis (up to 11,100 cells per µL) and hyponatremia (128 mmol/L). She was admitted for work-up including a COVID-19 test, viral panel, urinalysis, and blood cultures – all of which were negative. She was deemed high-risk for COVID-19, but a second test was also negative. No source of infection was found, and she was discharged after three days. Neither the ED staff nor the internal medicine team documented any rectal or genital examinations, and no additional testing or imaging was performed after the infectious disease evaluation remained negative.
The next day, her husband called her primary care office because she was feeling worse with rapid breathing, sweating, and diarrhea. During a video visit that afternoon, the symptoms were attributed to viral illness. The physician also noted that the patient described a skin breakdown on her “backside,” but the affected area was not visualized by the physician. The plan was for a home health nurse to evaluate the skin breakdown later in the week.
Thirty-six hours later, she was brought to the hospital via ambulance due to her husband’s concern about her altered mental status. On arrival, she was in septic shock with systolic blood pressure in the 80’s. The physical examination triggered concern about a necrotizing soft tissue infection. In the operating room, a necrotizing infection that tracked from a perirectal horseshoe abscess, through the perineum, up onto the anterior abdominal wall was found. She underwent an extensive debridement. After further discussion with her family, she was transitioned to comfort care in alignment with her goals. She died the day after surgery.
By Wesley Valdes, DO and Garth Utter, MD, MSc
This case appears to involve a presentation of Fournier’s Gangrene secondary to a perirectal abscess. Fournier’s Gangrene is a rare complication of infection that is accompanied by a high mortality rate, 20-40% in some case series reported in the literature and as high as 88% in others1 (although the published case series are small, most involving less than 100 patients who were predominantly male1). Fournier’s Gangrene is typically attributed to a necrotizing infection with a rapid and fulminant rate of spread. Early detection is difficult due to the lack of superficial evidence on the skin until late stages.2 Risk factors include diabetes mellitus, obesity, chronic renal insufficiency, and immune suppression.2
Patients with multiple sclerosis have been observed to have several immunological abnormalities involving perivascular T- and B-lymphocytes, activation of T-lymphocytes, intrathecal immunoglobulin production, and the presence of immunoglobulins, complement proteins, and cytokines in inflammatory plaques.3 Such abnormalities may have been contributory in this case. In addition, because patients with multiple sclerosis can experience neurological deficits due to loss of myelin and axons, this pathology often resulting in loss of vision, paralysis, sensory loss, ataxia, adverse brainstem signs, psychiatric disorders, and dementia,3 any sensory loss or neurological deficits in the perineal region of this patient may have suppressed her symptoms or made them lack specificity. While the severity or progression of the patient’s multiple sclerosis is not specified, the description of the patient in this case as being non-ambulatory suggests she had an Expanded Disability Status Score (EDSS) of 7.5 or higher, supporting consideration of the possibility that neurological deficits had masked symptoms and limited her ability to localize pain.
Given that this patient presented with constitutional symptoms of fever, diaphoresis, and dry cough during the COVID-19 pandemic, it was important to rule out that diagnosis. Evaluation for other respiratory viruses was also essential. Viral illnesses can exacerbate the signs and symptoms of multiple sclerosis, as phenomenon known as “pseudo-exacerbation.”3 However, once SARS-CoV-2 was ruled out by high-sensitivity polymerase chain reaction testing of multiple specimens from the posterior nasopharynx, there should have been efforts to pursue an alternate diagnosis, particularly given numerous alternative diagnoses in an older, non-ambulatory patient with multiple complications from chronic diseases.
Additionally, clinicians should be well versed in the complications that elderly, immunocompromised patients face after any hospitalization. Common examples include pneumonia, Clostridium difficile infection, pressure injuries, delirium, and sequelae from nosocomial exposure to other pathogens. Therefore, the rapid breathing, diarrhea, and sweating experienced by the patient in this case on the day after discharge should have been considered possible complications of her hospital stay, especially in a patient with multiple sclerosis. Unfortunately, the presumed respiratory infection and underlying comorbid condition distracted attention away from a festering abscess in the rectal region, likely delaying proper evaluation and readmission of the patient. It is not known whether the skin breakdown was emphasized to the physician, so it is possible that the patient or her caregiver minimized the severity of the problem. However, the mention of diarrhea in an incontinent patient with skin breakdown on the “backside” should have raised an alarm, as wound infections often develop in this situation.
Diagnostic Biases and the Limitations of Virtual Care
Two important circumstances affected the outcomes in this case. First, the patient’s age and co-morbid conditions put her at high risk for infectious complications. Although details about her overall condition are not available, her mobility and peripheral sensation were reduced by her multiple sclerosis, while the risk of infection was increased by her neurogenic bladder and immunomodulator therapy. She was at risk for skin breakdown and pulmonary atelectasis secondary to her immobility.
Second, the current pandemic has required significant modifications to normal processes both in the hospital environment and in ambulatory care to protect patients and staff from the virus. It has also biased clinicians to focus on diagnosing COVID-19, a phenomenon described as “availability bias” because it leads clinicians to overdiagnose conditions that are relatively available in their memories and to underdiagnose conditions with which they have little experience. In a typical hospitalization, nurses would perform a complete skin assessment at least once daily. The fact that this patient presented with respiratory symptoms may have made it seem less important and more challenging for the staff to do a complete skin exam while she was in the hospital. It is likely that the physician staff were focused on ruling out respiratory viruses, and the nursing staff were focused on preventing viral transmission, rather than conducting routine skin assessments. Additionally, the patient had a chronic neurogenic bladder; if she had a suprapubic catheter in place, direct visualization of the perineal area would have been less likely. As a result, a developing skin infection could have been missed by the nursing staff. Alternatively, nurses may have noted mild skin changes but may have been so focused on respiratory issues that they did not report the skin findings to the treating physicians.
Recent policy changes have led to rapidly expanded use of video visits, which bring their own challenges, particularly related to the physical examination. The use of a video visit for wound care is possible as skin and wound tissue can be adequately evaluated, especially in a “patient-assisted” exam.4 However, positioning the patient to visualize the skin in the perineal area may prove challenging and visits may be limited by poor lighting or difficulty positioning the camera or mobile device for adequate visualization of the area of concern.5 Fear of loss of privacy and social norms may also interfere with effective video examination of the perineal area. Finally, the value and benefit of smell and touch in the exam process is lost in a video visit. Certain portions of the physical examination cannot be performed online; however, there are tools such as remote stethoscopes and pulse oximeters that are available as well as sensors on mobile devices that can supplement direct visualization, if the necessary infrastructure support is available.6,7
In this case, a follow-up visit using video technology was appropriate, as the patient had been discharged just the day prior. However, the patient’s worsening condition probably justified a follow-up video or in-person visit the next day to assess how she was doing and to decide whether the plan established during the first video visit should be altered. Other options for this patient would have been to send her to the ED for further evaluation and potential admission, given her risk for complications. Consensus-based triage protocols may be helpful for physicians’ office or clinic staff to recommend the most suitable type of follow-up visit for each patient, based on their symptoms and underlying conditions. One such effort identified the following concerns as inappropriate for video visits: chest pain, shortness of breath, ear pain or hearing changes, abdominal symptoms or pain, and leg swelling.8
In this case, it appears that the physicians and nurses anchored on the patient’s originally presenting symptoms, which suggested a respiratory problem, and did not pay sufficient attention to the new problems of diaphoresis, diarrhea, and skin breakdown. Anchoring bias is an error in human cognition when a person prioritizes what is readily imaginable over other information that should be considered, especially after new findings appear or new test results become available. As clinicians are still learning about COVID-19 and its manifestations, there may be a tendency to attribute any respiratory symptoms to COVID-19 until proven otherwise. In this case, the patient tested negative twice for SARS-CoV-2, which both clinicians and caregivers probably viewed as a relief. Taking false reassurance, they failed to reconsider her differential diagnosis given her worsening condition and the new symptoms that were reported during the video visit.
Differential Diagnosis: Back to the Basics
There are certain basic processes that remain critical to adequately diagnosing patients, regardless of the setting of care. Gathering a patient’s history is still the most important aspect of the clinical diagnostic process. The physical examination, laboratory testing, imaging, and other investigations are meant to support that process by removing diagnoses from the “differential diagnosis” list. High-risk patients or, arguably, any patient sick enough to be admitted to a hospital, warrant complete physical examinations. While invasive parts of the physical examination, such as a digital rectal or vaginal bimanual exam, may be often refused by patients or deferred by clinicians, the visual inspection for skin integrity and cutaneous manifestations of underlying diseases should not be avoided. Significant cutaneous problems are still often overlooked, during both in-person care and virtual care.
Perianal abscesses are the most common type of anorectal abscesses. About 90% of anorectal abscesses are caused by non-specific obstruction and subsequent infection of the glandular crypts of the rectum or anus.9 Pain in the anal area (dull, sharp, aching, or throbbing) is common, and that pain may or may not be associated with bowel movements. A digital rectal examination is useful to rule out other causes of perianal pain, such as hemorrhoids, and to assess for fluctuance, dolor, rubor and calor suggesting an abscess.9 When the physical exam is refused due to pain or other reasons, magnetic resonance imaging (MRI) is preferred to computed tomography (CT) as CT may miss small abscesses in immunocompromised patients. However, CT is more readily available in ambulatory care, especially after usual work hours. Even if the presence of an abscess cannot be confirmed by imaging, it is reasonable to evaluate perianal pain and tenderness in an acutely ill patient with examination under anesthesia in the operating room. Abscesses are the most common cause of such symptoms, but other causes may also be amenable to operative treatment, such as fistulotomy for a fistula-in-ano or lateral internal sphincterotomy for a severe anal fissure.
In retrospect, this type of work-up would have been useful, but there was no mention of perianal pain by the patient in this case. The extension of the necrotizing process through the perineum and anterior abdominal wall was unexpected, as the perianal space is continuous with the fat of the buttock.9 The case does not specify the location of the perirectal abscess (perianal, intersphincteric, ischiorectal, or supralevator) but the path of extension typically follows the pathway of least anatomical resistance, in a posterior direction.9
If the history-gathering process does not result in adequate development of a differential diagnosis, the physical examination, laboratory tests, and imaging tests are unlikely to identify the patient’s underlying condition, as these procedures are chosen to help rule in or rule out diagnoses on the “differential diagnosis” list. In this case, it does not appear that consideration was ever given to skin infection, despite the patient’s high-risk due to immobility and immune compromise. Given the rarity of Fournier’s Gangrene and its lower prevalence among women than among men, this complication was likely not considered on the “differential diagnosis,” especially given the patient’s initial presentation with fever, cough, and diaphoresis. The concern of “skin breakdown on the backside” was only mentioned in the video call one day following discharge. Without more information, it is difficult to discern whether the caregiver was, in fact, referring to perianal tissue or, as is more commonly meant when using the term “backside,” the sacral and fleshy areas of the buttock. In any case, none of these areas was visualized during the video visit and physical examination was deferred to a home health nurse who was expected to visit the patient’s home later in the week. Limited information from the patient about her emerging symptoms and failure to visualize the area of concern led to delayed diagnosis in this case. It is hard to know whether the outcome in this case would have been altered by an in-person visit, but it is important for clinicians to consider alternative methods for gathering the information needed to make a prompt diagnosis. In this context, protocols or guidelines for when video visits are upgraded to in-person visits may be helpful.10
Engaging patients after discharge from the hospital offers opportunities to understand their clinical progress, whether toward resolution or decline. While there are certainly limits to telephone and online solutions compared to in-person experiences, the benefits of virtual care are increasingly recognized, especially during a pandemic. However, a broad “differential diagnosis” is important in all settings of care, especially when a patient develops new symptoms or deterioration in the overall condition. It is also important to have alternatives to virtual care when clinicians feel uncomfortable or unable to perform the necessary components of the physical examination during an online visit. Given the morbidity and high mortality rate associated with infectious complications of multiple sclerosis, a more thorough and timely physical examination may not have changed the ultimate outcome. However, high-touch post discharge monitoring of high-risk patients, whether through enhanced video visits or in-person care, offers opportunities to refine the “differential diagnosis” and potentially implement effective interventions at an earlier stage.
- In all patients with persistent signs and symptoms of illness, it is important to keep the differential diagnosis open to new data and to data that do not trend as expected.
- Chronically ill patients can develop new pathologies that may share symptomatology with a pre-existing active or resolving condition. In this case, the presumed upper respiratory infection and underlying comorbid conditions confounded the early discovery of a festering abscess in the rectal region.
- Practitioners should remember that cognitive biases, such as anchoring bias, can prevent full consideration of appropriate differential diagnoses.
- The practical difficulty that many patients experience in returning to ambulatory clinics for follow-up (even before the current pandemic) presents the opportunity for using new models of care, communication methods, and data-gathering tools. These options all should be considered, as well as use of remote monitoring equipment that allows for more frequent and real-time assessments of high-risk patients while they are at home.
- High-risk patients or, arguably, any patient sick enough to be admitted to a hospital, benefit from complete physical examinations. While invasive parts of the examination, such as a digital rectal or vaginal bimanual exam, may often be refused by patients or deferred by clinicians, visual inspection for skin integrity should not be avoided. Such inspections can be performed during both in-person care and virtual care.
Wesley Valdes, DO
Telehealth and Wound Specialist
Garth Utter, MD MSc
Professor, Department of Surgery
UC Davis Health
- Sorensen MD, Krieger JN. Fournier's Gangrene: Epidemiology and Outcomes in the General US Population. Urol Int. 2016;97(3):249-259. doi:10.1159/000445695
- Wähmann M, Wähmann M, Schütz F, et al. Severe Fournier's gangrene-a conjoint challenge of gynaecology and plastic surgery. J Surg Case Rep. 2017;2017(12):rjx239. Published 2017 Dec 8. doi:10.1093/jscr/rjx239
- Dimitri Agamanolis, M., 2020. Neuropathology. [online] Neuropathology-web.org. Available at: https://neuropathology-web.org/chapter6/chapter6aMs.html
- Benziger CP, Huffman MD, Sweis RN, Stone NJ. The Telehealth Ten: A Guide for a Patient-Assisted Virtual Physical Examination. Am J Med. 2021;134(1):48-51.
- National Telehealth Technology Assessment Resource Center (TTAC). 2020. Clinician’s Guide To Video Platforms. Available at: https://telehealthtechnology.org/toolkit/clinicians-guide-to-video-platforms/
- Carver S, Norouzi N, Bromberg S, et al. Time-frequency visualization of alcohol withdrawal tremors. Conf Proc IEEE Eng Med Biol Soc. 2014;2014:5474-5477. doi:10.1109/EMBC.2014.6944865
- Eulerian Video Magnification [video]. Youtube. https://www.youtube.com/watch?v=ONZcjs1PjmkPublished May 23, 2012.
- Croymans D, Hurst I, Han M. Telehealth: The Right Care, at the Right Time, via the Right Medium. NEJM Catal Innov Care Deliv. 2020;10.1056/CAT.20.0564. Published 2020 Dec 30.
- Sigmon DF, Emmanuel B, Tuma F. Perianal Abscess. In: StatPearls. Treasure Island (FL): StatPearls Publishing; June 27, 2020.
- Reeves JJ, Ayers JW, Longhurst CA. Telehealth in the COVID-19 Era: A Balancing Act to Avoid Harm. J Med Internet Res. 2021 Feb 1;23(2):e24785.