Excess Deaths During COVID-19

SOURCE: Rossen LM, Branum AM, Ahmad FB, et al. Excess deaths associated with COVID-19, by age and race and ethnicity — United States, January 26-October 3, 2020. MMWR Morb Mortal Wkly Rep 2020;69:1522-1527.

Skeptics have expressed concern that mortality from COVID-19 infection is being overestimated or worse, manipulated. Compared with previous years, it is clear from national data that more people in the United States have died this year than anticipated and not just older patients, but younger age groups have been disproportionately affected.

The measurement of excess deaths, as defined by the number of persons dying from all causes in excess of that anticipated for a given place and time, is useful when questions have been raised about the attribution of death to a given cause — or, in this case, the overall impact of COVID-19 on death rates in the United States. Some of this could be due to unrecognized COVID-19 infection or an indirect effect of the pandemic on the healthcare system and the ability to receive care for other reasons.

Mortality data from the Centers for Disease Control and Prevention’s National Viral Statistics System was used to examine differences in the number of deaths as defined by age and ethnicity, compared with the same weeks from 2015-2019. The percentage of excess deaths from all causes, as well as that from all deaths excluding that attributed to COVID-19, were estimated. (Expected numbers of deaths were estimated using over-dispersed Poisson regression models, accounting for seasonal trends, and also weighted for possible incomplete reporting in more recent weeks).

Compared with death data for the same weeks from prior years, nearly 300,000 excess deaths have occurred in the United States between the weeks ending Jan. 26 through Oct. 3, 2020. Two-thirds were attributed to COVID-19 (n = 198,081); the remaining deaths were largely attributed to vascular events, respiratory disease, and dementia/Alzheimer’s.

Excess deaths reached their highest points during the weeks ending April 11 and Aug. 8, 2020. As imagined, the lowest number of excess deaths occurred in the youngest age group (< 25 years), and the highest number of excess deaths occurred in the oldest age group (75-84 years). However, the greatest percentage change in unanticipated deaths was experienced in those 25-44 years of age (26.5%). The percentage of excess deaths in 2020 compared with averages for the previous five years for other age groups was: 14.4% for ages 45-64 years, 24.1% for ages 65-74 years, 21.5% for ages 75-84 years, and 14.7% for ages > 85 years. The greatest percentage difference in excess deaths occurred in Latinos (53.6%), followed by Asians (36.6%), Blacks (34.6%), and American Indian/Native Americans (28.9%). In contrast, the percentage of excess deaths in whites was 11.9%. This finding is consistent with reported disparities in COVID-19 deaths in Latinos and other minorities.

Subclinical Influenza Infection in Healthcare Workers

SOURCE: Benet T, Amour S, Valette M, et al. Incidence of asymptomatic and symptomatic influenza among healthcare workers: A multicenter prospective cohort study. Clin Infect Dis 2020; Aug 4. doi:10.1093/cid/ciaa1109

It happens every year: Patients in the hospital for other reasons suddenly develop a fever and test positive for influenza (or respiratory syncytial virus or other viral illness). Despite all precautions, influenza vaccination, handwashing campaigns, and messaging to staff not to come to work with respiratory symptoms, healthcare workers (HCWs) are an important source of nosocomial influenza and respiratory infection. And now it is happening with COVID-19 employees with sniffles come to work, thinking they “only have a cold,” only to test positive for SARS-CoV-2.

These authors demonstrated just how common subclinical influenza really is in HCWs. A total of 278 HCWs providing active care at five French hospitals were enrolled during the 2016-2017 winter season. Participants maintained a daily diary of symptoms and were seen for physical examination the first time in October through December before the beginning of the flu season, again in January during peak flu season, and then approximately three weeks following their second visit. Nasopharyngeal swabs for influenza PCR (Virocult) and serologies by hemagglutination inhibition (IHA) for influenza A H3N2- and B Victoria lineage B/Brisbane/60/2008-specific antibodies were obtained. In the event of symptoms, an additional visit with these tests was performed. The median age of participants was 36 years and 84% were female. Vaccine coverage was 42% for 2015-2016 and 49.6% for 2016-2017. Pauci-symptomatic infection was defined as the presence of one or more signs or symptoms for more than one day, with no fever (< 37.8°F), or the absence of cough and sore throat, whereas symptomatic influenza was defined as fever 37.8°F with either cough or sore throat.

Sixty-two participants developed influenza infection during the five-month study, with a cumulative incidence of 22.3%. Impressively, 46.8% and 41.9% of these were asymptomatic and pauci-symptomatic, respectively, while only 11.3% developed more classic symptoms. Fever occurred in less than 10% of cases. At the second evaluation in January, people with confirmed influenza reported runny nose (68%), cough (64%), and headache (56%). At the third evaluation, those with confirmed influenza reported runny nose (55%), cough (45%), and headache (36%).

The cumulative incidence of influenza infection did not appear to differ between those who received the 2016-2017 influenza vaccination and those who did not (20.3% vs. 24.3%, P = 0.38), although receipt of the 2015-2016 influenza vaccination was protective (16% vs. 27%). Working in a nursing capacity increased the risk of pauci-symptomatic or symptomatic influenza. However, work in the intensive care unit (ICU) setting or the presence of three or more adults in the home was associated with an increased risk of asymptomatic infection.

Nearly nine of 10 hospital staff with confirmed influenza had subclinical infection and half of these were entirely asymptomatic. Symptoms in 42% were atypical, with at most minor sniffles, headache, or sore throat. Attempts to keep such minimally symptomatic HCWs out of the hospital has been challenging. As our emergency room medical director said the other day, physicians will still come to work with a runny nose — they are just too important and we do not have sufficient staff to keep everyone home with a cold. Although little can be done about asymptomatic infection, we must have some way to rapidly screen minimally symptomatic employees for subclinical serious infections, such as the flu or COVID-19.

Compression Garments Effective in Reducing Cellulitis

SOURCE: Webb E, Neeman T, Bowden FJ, et al. Compression therapy to prevent recurrent cellulitis of the leg. N Engl J Med 2020;363:630-639.

This single-site, nonblinded, randomized study examined the benefit of compression garments in patients with lower extremity edema at risk for cellulitis. Eligible patients had significant edema for more than three months in one or both legs, and a history of two or more episodes of cellulitis in the same leg within the previous two years. They could not be using compression garments already for more than four days per week. Patients who were end-of-life or immunosuppressed were excluded.

Participants were assigned to wear compression garments throughout the day, every day. The garments generally consisted of knee-high stockings including the foot (with or without the toes) or leg and foot wraps. Participants were followed every six months for up to three years. Control patients who developed cellulitis were crossed over to the compression therapy group. Clinical characteristics were similar between the two groups at entry to the study. A total of 84 participants were enrolled in study, including 41 in the compression group. During the study, 78% of participants in the compression group reported wearing their compression stockings or wraps five or more days per week, and 88% reported using them at least four days per week.

The trial was stopped prematurely when a large difference in outcomes between the two groups was recognized. At the time the study was stopped, 23 episodes of cellulitis had occurred, including six (15%) in the compression therapy group and 17 (40%) in the control group (P = 0.002). Three patients in the compression group (7%) and six patients in the control group (14%) required hospitalization. Three patients died (one in the compression group and two in the control group), and one in each group developed wound infection. Two patients in each group were receiving prophylactic antibacterials at the time of study entry, which were continued. The median duration of follow-up was 209 days in the compression group and 77 days in the control group — simply because patients were removed from the control group when they developed cellulitis.