Hospitalized cancer patients have a high risk of nosocomial COVID-19 that is associated with increased mortality, so these patients should be treated in COVID-free zones, according to researchers.
In an observational study of patients with COVID-19 and cancer, 19% of patients had COVID-19 acquired during a non-COVID-related hospital stay, and 81% had community-acquired COVID-19.
At a median follow-up of 23 days, the overall mortality rate was 28%. However, the all-cause mortality rate in patients with nosocomial COVID-19 was more than double that of patients with community-acquired COVID-19, at 47% and 23%, respectively.
"This is the first report that describes a high rate of hospital-acquired COVID-19 in patients with cancer, at a rate of 19%," Dr. Elkrief said. "This was associated with high mortality in both univariate and multivariate analyses."
The study included 250 adults and 3 children with COVID-19 and cancer who were identified between March 3 and May 23, 2020. They ranged in age from 4 to 95 years, but the median age was 73 years.
All patients had either laboratory-confirmed (95%) or presumed COVID-19 (5%) and invasive cancer. The most common cancer types were similar to those seen in the general population. Lung and breast cancer were the most common, followed by lymphoma, prostate cancer, and colorectal cancer. Most patients were on active anticancer therapy, most often chemotherapy.
Most patients (n = 236) were residents of Quebec, but 17 patients were residents of British Columbia.
"It is important to note that Quebec was one of the most heavily affected areas in North America at the time of the study," Dr. Elkrief said.
Outcomes by Group
There were 206 patients (81%) who had community-acquired COVID-19 and 47 (19%) who had nosocomial COVID-19. The two groups were similar with respect to sex, performance status, and cancer stage. A small trend toward more patients on active therapy was seen in the nosocomial group, but the difference did not reach statistical significance.
The median overall survival was 27 days in the nosocomial group and 71 days in the community-acquired group (hazard ratio, 2.2; P = .002).
A multivariate analysis showed that nosocomial infection was "strongly and independently associated with death," Dr. Elkrief said. "Other risk factors for poor prognosis included age, poor [performance] status, and advanced stage of cancer."
There were no significant differences between the hospital-acquired and community-acquired groups for other outcomes, including oxygen requirements (43% and 47%, respectively), ICU admission (13% and 11%), need for mechanical ventilation (6% and 5%), or length of stay (median, 9.5 days and 8.5 days).
The low rate of ICU admission, considering the mortality rate of 28%, "could reflect that patients with cancer are less likely to be admitted to the ICU," Dr. Elkrief noted.
Applying the Findings to Practice
The findings reinforce the importance of adherence to stringent infection control guidelines to protect vulnerable patients, such as those with cancer, Dr. Elkrief said.
In ambulatory settings, this means decreasing in-person visits through increased use of teleconsultations, and for those who need to be seen in person, screening for symptoms or use of polymerase chain reaction testing should be used when resources are available, she said.
"Similar principles apply to chemotherapy treatment units," Dr. Elkrief said. She added that staff must avoid cross-contamination between COVID and COVID-free zones, and that "dedicated personnel and equipment should be maintained and separate between these two zones.
"Adequate protective personal equipment and strict hand hygiene protocols are also of utmost importance," Dr. Elkrief said. "The threat of COVID-19 is not behind us, and so we continue to enforce these strategies to protect our patients."
Session moderator Gypsyamber D'Souza, PhD, an infectious disease epidemiologist at Johns Hopkins University in Baltimore, raised the question of whether the high nosocomial infection and death rate in this study was related to patients having more severe disease because of underlying comorbidities.
Dr. Elkrief explained that the overall mortality rate was indeed higher than the 13% reported in other studies, and it may reflect an overrepresentation of hospitalized or more severely ill patients in the cohort.
However, the investigators made every effort to include all patients with both cancer and COVID-19 by using systematic screening of inpatient and outpatients lists and registries.
Further, the multivariate analysis included both inpatients and outpatients and adjusted for known negative prognostic factors for COVID-19 outcomes. These included increasing age, poor performance status, and different comorbidities.
The finding that nosocomial infection was an independent predictor of death "pushed us to look at nosocomial infection as a new independent risk factor," Dr. Elkrief said.
Dr. Elkrief reported grant support from AstraZeneca. Dr. D'Souza did not report any disclosures.
The article was originally published on MDedge.