Stroke Census Drop and Resource Utilization Changes with COVID19

Where I work, Tampa General Hospital, is a tertiary care facility that has witnessed a robust increase in the number of stroke patients every year from about 800 in 2011 to 1,300 in 2019. This trend reflects a constant influx of retirees to Florida and also relates to our reputation as a Comprehensive Stroke Center, receiving many transfers and referrals from 13 surrounding counties to help diagnose and treat complex strokes. Because the initial epicenter of the COVID-19 pandemic in the US in February to March 2020 was in the northeast, we didn’t see many cases of COVID-19 here on the west coast of Florida. Unexpectedly, we did see a precipitous decline in our daily stroke census. Typically, on any given day, our census has 15 to 25 ischemic or hemorrhagic strokes for us to round on–either in the wards or in our 32-bed Neurosciences Intensive Care Unit. During the early months of the pandemic, however, our census averaged only 5 to 6 per day, and some days we had only 1 or 2 stroke consultation requests in the entire hospital!

Admittedly, a smaller patient load had its upsides. Our email and electronic medical record (EMR) inboxes were totally cleaned out, our administrative paperwork was “signed, sealed and delivered,” and various stroke research projects that had been marinating on the back burner were finally being cooked and seasoned for publication. As this low census persisted, however, we realized this was no fluke and that, startlingly, people were staying at home with strokes (and heart attacks) instead of seeking proper medical attention. We surmised that the “Stay at Home” order was being adhered to perhaps too strictly to include emergency medical conditions and therefore keeping scared citizens away from the hospital. Thus, I rallied my colleagues to write up a paper for the journal The Neurohospitalist,1 after we quickly calculated the degree of stroke volume drop in 2020 compared with the average we had seen in the same months of 2017, 2018, and 2019. We found that people who were showing up to the hospital with stroke seemed to have more severe symptoms and require more extensive services such as neurointerventional clot retrieval to extract the thrombus from the brain. We postulated this increasingly severe case-mix was could be explained by the following: 1) patients with mild stroke or transient ischemic attack (TIA) could more easily dismiss their symptoms than someone with complete hemiplegia or profound aphasia (family would be more insistent, perhaps, to call 911); 2) COVID-19 was provoking a more severe variety of stroke because of a virus-induced intensification of hypercoaguability, arrhythmia, vasculitis, vasospasm, or other cryptic factor; or 3) mild strokes or TIAs may have worsened at home then eventually presented to the emergency room with exacerbating neurologic deficits that were too overwhelming for caretakers at home.

Numerically, we reported an initial stroke volume decrease of 15%; this was similar to the 23% reduction published by the stroke team at Jefferson University Hospital in Philadelphia.2 Acutely, the rate of mechanical thrombectomy at Tampa General during this initial period went up by about 8%1 (and increased by a whopping 50% at Jefferson).2 Although both of our hospital systems saw a rise in stroke interventional procedures, the degree of change could be explained by regional differences of viral prevalence. The Jefferson team also reported that 25% of all stroke patients who needed thrombectomy were positive for COVID-19,2 whereas we had close to 0% at that time. In June, however, as Florida began to reopen communities and businesses, our local COVID-19 infection rates started to rise; we were concerned we might be trading places with the Northeast as a new viral hotspot, especially because our stroke census started to climb concurrently.

Presentation of Stroke with COVID-19

In the first half of the pandemic, we did not see people with COVID-19 whose strokes were related to or possibly caused by COVID-19, unlike reports from the Northeast. In July and August, however, we did start to see more individuals who were infected with COVID-19 presenting with strokes–as well as strokes occurring in patients already hospitalized with COVID-19 and cardiopulmonary symptoms. Generally, these individuals had multiple other comorbidities (particularly diabetes), were already in critical condition (often on vasopressors), and had strokes that unfortunately were a precursor to their demise. These patients were relatively younger than the typical patient with stroke before the pandemic—in their 40s and 50s, including an individual with diabetes who had a cerebral vein thrombosis and venous infarct at age 39.

The question of whether (and how) COVID-19 infection increases risk or severity of stroke remains up for debate, and reports of strokes while hospitalized for the virus range from 0.9% to 2.4%. A study from the University of Pennsylvania in Philadelphia reported that out of 844 patients hospitalized with COVID-19, only 2.4% had a stroke.3 This rate is very similar to the 2.3% stroke rate reported from Wuhan, China in 214 people hospitalized with COVID-19.4 In both of these studies, those who had strokes were more likely to have comorbid obesity and hypertension. In the Wuhan cohort, most strokes occurred in individuals with more severe COVID-19 symptoms.

A study from New York reported ischemic stroke in only 0.9% of 3,556 individuals hospitalized with COVID-19–these patients were either admitted for stroke and found to have COVID-19 or admitted for COVID-19 and subsequently had a stroke. They were younger and had more severe strokes compared with people without COVID-19 who had strokes in the same time period. In contrast to the Wuhan and Philadelphia studies, the New York patients with stroke and COVID-19 were less likely to have hypertension or a history of prior stroke or TIAs than those with stroke before the pandemic.5 The time to stroke diagnosis in the Penn study was about 3 weeks on average, whereas the Wuhan and New York time to stroke diagnosis were median 1 to 2 days and 10 days, respectively.

Of course, the differences seen between New York, Philadelphia, and Wuhan may also relate to differing demographic factors, regional rates of COVID-19 infection, and severity during the various timeframes studied. COVID-19 is still a new disease–evolving and even possibly mutating right now–and our understanding of how it relates to stroke risk, incidence, and illness severity is still an unfolding story that likely will remain challenging to interpret for a while.

Covidophobia and Collateral Damage

The amount of fear during this pandemic is quite varied. Some people have a healthy dose of caution, and yet will still appropriately show up to the emergency room with stroke symptoms. Others have explicitly told us that they decided to “stay at home with my stroke, hoping it will get better on its own.” The likelihood of a stroke fully reversing itself at home is not zero, but close to zero; strokes do not just miraculously improve to normal, with the exception of TIAs, and, even in the case of TIAs, going to the hospital is considered necessary to see if symptoms return and identify the cause of the TIA (essentially a “would-be stroke”) in order to prevent a bigger stroke down the road. For serious medical emergencies, including stroke, we have spent years effectively developing and finessing excellent systems-of-care to expedite services for hyperacute therapies. Nosocomephobia (fear of hospitals) is not uncommon–and certainly this virus may have intensified this to an irrational fear of immediately contracting COVID-19 just by walking in the hospital’s front door. Perhaps this is a subset of nosocomephobia which we should name covidophobia? Regardless, people may not realize that extensive measures have been put in place to maximize patient and staff safety. Every day upon arrival to the hospital and clinic, healthcare workers such as myself are screened at the door for the presence of any symptoms of infection: “Do you feel sick? Do you have a fever? Have you had cough, shortness of breath, loose stools, muscle aches?” Everybody wears a mask inside the building at all times and certainly for each patient encounter. All patients in our emergency room are tested for the virus SARS-CoV2, which causes COVID-19. Any positive screen is isolated in a separate unit. Family members and guests are not allowed into the hospital, with rare exceptions and precautions. This degree of safeguarding actually seems higher than we see in the general community, and frankly feels safer here than in some local stores. Of course, we must also mention that in the US today, there is a third group, perhaps with a different phobia—a fear of not being told the truth—or who simply believe COVID-19 is of minimal consequence, unconnected to them. This group fully participates in society without concern or protective measures of any kind. All of these dynamics have altered the stroke landscape and as health professionals we have tried hard to mitigate such “collateral damage.”

Time is Still Brain—Even in a Pandemic

The key point of the intersection between this virus and stroke is that stroke is potentially more disabling long-term than COVID-19 and has a higher likelihood of fatal outcome (15% and 45% for ischemic and hemorrhagic stroke, respectively, vs 1%-4% for COVID-19). The gold-standard treatment for stroke–thrombolytic “clot buster” intravenous medication–that optimizes outcomes (less long-term disability) must be given by 4.5 hours of symptom onset. As mentioned above, anecdotally more people than ever are deciding to “wait-and-see” if symptoms improve before calling 911 or coming to the emergency department. What they may not realize—and the message we need to share—is that any postponed approach is suboptimal and can essentially exclude them from receiving the best treatment as they “time out” for thrombolytic therapy. This could result in more poststroke disability, longer stays in rehab, and missed time from work, driving, and other activities.

Even during a pandemic, the well-known mantra of our field, “Time is Brain,” still applies. For every minute a stroke goes untreated, a brain can lose 1.9 million neurons; multiplied by 60, that is about 120 million neurons lost per hour of delay. Hence, the “Stay at Home” advisory for an infectious pandemic does not apply to a medical emergency such as stroke. Years ago, our community invented the mnemonic FAST (face, arm, speech, time) to recall the need for urgent treatment of the most common stroke symptoms. We reiterated this in an article for the local newspaper, the Tampa Bay Times, to remind our citizenry here that we are open for business!


Fewer people are coming to the hospital with stroke, due to a panoply of reasons, ranging from nosocomephobia/covidophobia to overinterpretation of the COVID-19 “Stay at Home” orders—which should not apply to medical emergencies such as stroke. The strokes we do see in people who appear at the hospital are more severe than usual, perhaps due to COVID-19 pathogenesis itself and require additional resource utilization, as reported in our series and those from the northeastern US. It is essential that we strongly readvocate our mantra of “Time is Brain,” which supercedes “Stay at Home” in order to optimize stroke outcomes.

1. Rose, DZ, Burgin WS, Renati S. Untreated stroke as collateral damage of COVID19: “time is brain” versus “stay at home.” [published online ahead of print, 2020 May 26]. The Neurohospitalist.

2. Sweid A, Jabbour P, Tjoumakaris S. Letter to the editor: incidence of acute ischemic stroke and rate of mechanical thrombectomy during the COVID19 pandemic in a large tertiary care telemedicine network. [published online ahead of print, 2020 Jun 13]. World Neurosurg. 2020;S1878-8750(20)31311-5.

3. Rothstein A, Oldridge O, Schwennesen, Do D, Cucchiara BL. Acute cerebrovascular events in hospitalized COVID19 patients. Stroke. 2020; 51:00-00.

4. Mao L, Jin H, Wang M, et al. Neurologic manifestations of hospitalized patients with coronavirus disease 2019 in Wuhan, China. JAMA Neurol. 2020;77:1-9.

5. Yaghi S, Ishida K, Torres J, et al. SARS-CoV-2 and stroke in a New York healthcare system. Stroke. 2020;51:2002-2011.

DZR reports no relevant disclosures