Today’s frontline clinicians often cannot wait for randomized controlled trials (RCTs) to make lifesaving decisions. They rely largely on the experiences of their colleagues. This article distills some of the challenges and lessons learned in key areas of COVID-19 management: convalescent plasma therapy and strategies for diabetes/hyperglycemia.
Convalescent Plasma
Craig M. Coopersmith, MD, FACS, MCCM
COVID-19 Pearls
- Currently, we are not clear whether or not convalescent plasma is effective, and it is not standard of care. I strongly urge healthcare professionals to enroll patients in clinical trials of convalescent plasma so we can truly understand whether or not it is effective, and if so, in which patient population.
- Based on a large sample size, convalescent plasma appears to have no higher risk than any other plasma transfusion. If convalescent plasma is used, there are theoretical reasons to start it early in the hospitalization.
COVID-19 Rapid Resource Center:
Convalescent plasma is available to hospitalized patients with COVID-19. However, to date there are no RCTs demonstrating its efficacy, which has made the Food and Drug Administration’s (FDA) Emergency Use Authorization (EUA) somewhat controversial.
In his role as a Society of Critical Care Medicine (SCCM) representative to the National Institutes of Health (NIH) COVID-19 Treatment Guidelines Panel, Dr. Coopersmith has a front-row seat to the most recent developments in the use of convalescent plasma. Dr. Coopersmith is the director of the Emory Critical Care Center and a professor in the university’s Department of Surgery.
So far, in the absence of RCTs, the data used to support the EUA is based upon an analysis of a large patient cohort that received convalescent plasma under an Expanded Access Protocol from the Mayo Clinic. When comparing patients with high-titer versus low-titer convalescent plasma, there was no benefit in the entire group. However, multiple subset analyses demonstrated that there may be a benefit in nonintubated patients receiving high-titer convalescent plasma within the first three days of hospitalization. Further studies are required to determine whether this represents a true benefit, Dr. Coopersmith said.
“Conceptually, convalescent plasma in COVID-19 is similar to other infectious diseases. It has been shown to be effective in many—but not all—infections it has been tried in,” he added. “In general, because of the theoretical benefits of treating earlier with convalescent plasma, we have tried to use it earlier in a patient’s hospitalization.
“We urge clinicians to enroll patients in trials so we can get a clear understanding of whether convalescent plasma is beneficial or not,” he said.
“Since we are doing so many things simultaneously, it is impossible on an individual basis to determine if convalescent plasma is effective or not,” Dr. Coopersmith noted. “That is why we need larger studies. The FDA relied on data from thousands of patients to issue the EUA—and despite that, we still don’t know if it works and in whom.”
Hyperglycemia and Diabetes
Amado X. Freire, MD, MPH, FCCM
COVID-19 Pearls
- Be alert: Hyperglycemia, with the widespread use of steroids, is more common among COVID-19 patients than among most other ICU patients, so you need to monitor for it and be ready to control it. Expect about half of your COVID-19 patients to have hyperglycemia at some point—it could be more.
- Have a low threshold to identify patients with hyperglycemia and a compulsiveness to control it appropriately.
Unmanaged hyperglycemia is recognized not only as a marker of more severe disease, but also as an indicator of poorer outcomes in critically ill patients. Dr. Freire is a professor of medicine, pulmonary critical care, and sleep medicine at the University of Tennessee Health Science Center. He discusses his experience treating COVID-19-positive patients with hyperglycemia.
“In the early days of the epidemic, we were mostly seeing older patients and those with more severe comorbidities in our ICUs,” he noted. “In these patients, the risk of mortality was greater than 50%. As time went by and we learned more about the virus, people were treated earlier in the community by their primary care doctors. At the same time, we became able to identify the consequences of the virus earlier,” he said. “Now we tend to see patients who are in the earlier stages of infection, as well as patients who are younger. These patients tend to have less glycemic stress. Also, patients who are baseline diabetic, we have learned, tolerate a little more hyperglycemia. That can make a difference in how we treat these patients.”
The other thing that has changed in the past couple of months is that now it is accepted and appropriate to use steroids for the acute respiratory distress syndrome (ARDS) component of COVID-19,” he added. “But when we use steroids, we also see more hyperglycemia. That means we need to use hourly checks and administer supplemental insulin to keep glucose levels under control.”
Dr. Freire uses IV insulin when necessary, ensuring that patients are monitored every hour. The unit also uses subcutaneous insulin when appropriate, he said.
“We use SCCM’s
Guidelines for the Use of an Insulin Infusion for the Management of Hyperglycemia in Critically Ill Patients in our ICU, which means targeting a glucose level of 150 mg/dL. If a patient has diabetes, that patient may be a little more tolerant of higher glycemia levels. In these patients, don’t try to push the glucose level too low because you may be more likely to have problems. If the patient is not known to have diabetes—that is, if they do not have diagnosed diabetes when they come to the ICU—then you should try harder to maintain a 150-180 mg/dL window. Some clinicians use tighter control with higher risk of hypoglycemia, but in our ICU, we stick to that 150-180 mg/dL range.”
“It’s true that steroids compound the problem of hyperglycemia in COVID-19 patients, but we must give steroids to many of these patients—there’s no getting around that,” he said. “We just need to be mindful of the possible effect on glycemia and monitor and control that effect.
“Initially, I thought perhaps remdesivir was contributing to hyperglycemia in these patients,” he added. “But now I think when we see more hyperglycemia in patients treated with remdesivir, it’s because these patients are sicker to begin with and not because of the drug itself.”
Many sicker patients have another concern to contend with—ICU related weakness. “Mortality rates are much lower now than they were in the beginning of the epidemic. In general, if we can keep a patient alive for two weeks, then they have an excellent chance of survival. But afterward, they may be left with this extreme weakness that causes difficulties with liberation, ambulation, and performing tasks of everyday living,” he noted. “I don’t know how that will play out for these patients in the long term, but it is something to think about when treating and discharging patients who have received paralytics and/or steroids.”
Dr. Freire has not seen much difference in outcomes between patients with well-controlled diabetes and those with poorly controlled diabetes. “The difference lies with controlling hyperglycemia,” he noted. “Hyperglycemia occurs in response to the stress of the disease, so if you come in with hyperglycemia, that means you’re sicker to begin with. What’s even more significant is a patient who comes in with normal A1C and then develops hyperglycemia acutely. That’s an indicator of the severity in the patient's disease.
James S. Krinsley, MD, FCCM
COVID-19 Pearls
-Frequent monitoring is mandatory to avoid hyper- and hypoglycemia. Blood glucose levels may change dramatically throughout the day. If you monitor only four times a day, you will miss episodes of hyper- and hypoglycemia and you will not be able to control glucose variability.
-Diabetes and hyperglycemia are both strongly associated with risk of death in patients with COVID-19. Right now, the mechanisms for this are just speculative. There is a huge target here for an RCT on the effect of better glucose control on diabetes and hyperglycemia outcomes.
As director of critical care at Stamford Hospital, Dr. Krinsley has significant experience treating COVID-19 patients with diabetes and hyperglycemia. Like many hospitals on the East Coast, Stamford saw the greatest influx of COVID-19 patients from mid-March through May.
“We went from having a typical ICU ventilator census of five to seven patients to having 43 patients on ventilators at one point. Even finding the right people to take care of them was a challenge,” he said. “We’re very proud of our outcome statistics. For the most critically ill group of patients, among whom 85% were mechanically ventilated, our overall mortality was 35%.”
During those early days of the pandemic, Dr. Krinsley and his team treated many patients with both COVID-19 and diabetes. Data collection included a series of 109 ICU COVID-19 patients for whom at least four blood glucose values were available. The series has provided valuable insights regarding the link between diabetes, glucose levels, and outcomes in this patient population.
“In this series, there is a very strong relationship between glucose levels in the ICU and mortality rates,” he noted. “What’s different in this analysis is that we were able to do a multivariate analysis that included severity of illness, not just diabetes status. We also looked at age and whether or not patients were mechanically ventilated, which in itself is an important marker of disease severity.
“In these patients, for every 10 mg/dL increase in mean blood glucose during their ICU stay, there was a 20% increased odds of death,” he noted. “This was an independent association. That’s a very strong relationship between glucose level and increased odds of death.” Dr. Krinsley also noted that, in his experiences, diabetes was overrepresented among COVID-19 patients, especially those with severe disease.
“Typically, in our mixed ICU, 15%-20% of our patients have diabetes. In our cohort of 109 COVID-19 patients, 44% had diabetes. These are the sickest patients with COVID-19,” he said. “At the height of the surge, with multiple ICUs, 85% of our COVID-19 patients were intubated because they could not be managed with high-flow nasal cannula or other modalities. This is the group in which diabetes is overrepresented.” Increasing glucose levels was associated with increased risk of mortality in patients, regardless of their pre-admission glycemia, he said.
“In these patients, diabetes was not just a marker of increased risk of serious illness but was also associated with an increased risk of death,” he said. “For COVID-19-postive patients, the risk associated with hyperglycemia is across the board and independent of A1C levels.”
Before the pandemic, Dr. Krinsley said his ICU implemented two targets for ICU patients: 80-140 mg/dL and 110-160mg/dL. Patients are monitored at least every three hours, with IV insulin patients monitored every hour. But with the advent of the pandemic, the standard of care had to be adjusted to handle not only the increased patient load, but also concerns about clinician exposure.
“During the surge, we went from a typical ICU census of 10 to 12 patients to having up to 45 patients across three ICUs,” he said. “In addition to our regular staff, we had nurses working with us who were trained and supervised to perform ICU care, but who were not ICU nurses. Personal protective equipment was a challenging issue. We never ran out, but we had to be progressively more careful about its use.”
“With such frequent monitoring, we also had to be mindful of staff exposure to these very sick patients, and we wanted to try to limit exposure as much as we could. So we bit the bullet and changed our glucose target during the surge,” he said. “Everyone was treated with a target of 120-180 mg/dL, which was a looser target overall.”
The use of IV insulin was also reduced. “IV insulin requires hourly monitoring, and that means a lot of interactions with these very sick patients,” he said. “Instead, we started to use more subcutaneous long-lasting insulin, replacing part of the daily insulin requirement with long acting insulin administered on a Q12 schedule to increase dosing flexibility. I wanted to simplify the regimen for the nurses, especially for those who were being repurposed during the surge.”
SCCM is committed to providing its members with up-to-the-moment information regarding the COVID-19 pandemic as it evolves. For more information, visit the SCCM COVID-19 Rapid Resource Center.