Prognosis of frail hospitalized COVID-19 patient: Better than expected?

Severe acute respiratory syndrome coronavirus 2 (SARSCoV-2) is a single-stranded RNA virus that is responsible for the novel coronavirus disease (COVID-19), a lower respiratory tract infection that can lead to severe and even fatal disease. Transmission is thought to be primarily via respiratory droplets, although other mechanisms are still being investigated [1]. COVID-19 has affected close to 6.2 million people worldwide [2], with a variety of presenting symptoms that include fever, cough, dyspnea, myalgia, headache, sore throat, diarrhea and vomiting, along with sudden gustatory and olfactory dysfunctions [3].

Current treatments are largely supportive, including oxygen therapy, non-invasive and invasive mechanical ventilation, fl uids, antivirals such as lopinavir-ritonavir and remdesivir, or antimalarial as chloroquine, and broad-spectrum antibiotics when there is evidence of secondary bacterial infection [4].
As of April 22, 2020 in Switzerland, 28,268 individuals have tested positive for SARS-CoV-2 (with a cumulative incidence of 329 per 100,000 inhabitants), among whom 1,207 have died [5]. In Valais, one of the 26 Swiss states and home to 345,000 inhabitants, 1,729 cases have been detected (cumulative incidence of 517 per 100,000 inhabitants), including 114 deaths (54 hospitalized, 60 non-hospitalized) [6].
At the beginning of the confi nement period that the Swiss Federal Council issued on March 16, 2020, we opened a COVID-19 unit in the Valais Hospital (900 beds), within the geriatric department (210 beds), where we hospitalized elderly patients who did not desire intensive resuscitation (intubation and reanimation). Geriatric patients who desired intensive resuscitation were instead admitted to the internal medicine or intensive care wards.
Patients' decisions not to pursue resuscitation were made of their own free will, based on advance directives when available, as well as formal discussions at the time of admission with the patient, his family, and/or his therapeutic representative(s).
As per our hospital's standards, in the setting of the ongoing pandemic, no prior explicit consent for participation in COVID-19-related studies was required, as long as patients' identifying characteristics were not used.
We followed our hospital's guidelines for diagnosis and treatment, as defi ned collaboratively by our infectious disease experts and intensive care physicians, and adapted specifi cally to this geriatric do-not-resuscitate population by the Chairman of the Geriatric Department. Diagnosis was confi rmed by PCR on nasopharyngeal swabs. When confronted with negative results but high clinical suspicion, up to two addition swabs and a thoracic CT scan were performed before infection was defi nitively ruled out. Admission tests included complete blood count with differential, chemistry panel with liver function tests, CRP and procalcitonin. A chest x-ray was also obtained at time of admission. CT scans were obtained when (1) the chest x-ray was normal but laboratory values pointed towards secondary bacterial infection, such as procalcitonin > 0.25 μg/L, or (2) signifi cant deterioration in respiratory status, as long as the patient's clinical status was stable enough to allow transfer to the radiology unit. If a purely palliative approach was in place, additional imaging was often forgone. The recommended treatment protocol was modifi ed several times during the onemonth period, based on available studies at the time. Antibiotics were administered primarily when procalcitonin was elevated or when imaging suggested a bacterial superinfection (Meropenem 1g every 8 hours was recommended until April 1, at which point the protocol changed to Co-amoxicillin 1.2g every 8 hours for 10 days, with Azithromycin 500mg once daily for 5 days, used at individual physicians' discretion). We did not categorically provide antimalarial or antivirals, but the geriatricians were encouraged decide whether the potential benefi ts of chloroquine, remdesivir and lopinavir-ritonavir outweighed risks for their individual patients. In practice, we rarely initiated antiviral treatment in our patients, but at times continued antiviral treatments had been started in another service.
This study represents a novel perspective, offering insight on the impact of COVID-19 illness on this unique cohort of hospitalized patients over 65 years old who did not desire resuscitation.

Clinical characteristics
In this retrospective case series, we present relevant Among the 65 discharged patients, average age was 84.5 years, and 62% were male. At time of admission, 78% presented with respiratory symptoms (cough, dyspnea, pharyngitis), 71% had a temperature greater than 38°C, and 89% had severe asthenia and/or myalgia. In addition, 60% presented with bilateral pulmonary infi ltrates on chest x-ray. oxygen saturation on room air (average O2 saturation of 78.4% for those who died, vs 88.6% for those who survived to discharge). Additionally, we found that they had more intense oxygen demands (72% of those who died required 6L of 50% FiO2, vs 5% of survivors) and required supplemental oxygen for a longer time period (9 vs 3.3 days).

Comment
This case series suggests that, among a vulnerable geriatric patient cohort, the cumulative presence of inherent risk factors did not seem a reliable predictor of prognosis. It remains unclear which risk factors and clinical characteristics dictate that a patient will be more likely to develop acute respiratory decompensation, thereby predicting higher mortality. The theory that thromboembolic events explain this decline was not corroborated within our population, even though our patients were admittedly less likely to receive repeat CT-scans than those in more intensive care settings. Additionally, our data was limited by the fact that we were frequently providing palliative care, and thereby choose to forgo exhaustive investigations unless the results were expected to signifi cantly alter treatment. Due to the observational nature of this small study, we were unable to perform a statistic analysis that would allow us to draw conclusions on which risk factors and therapies had the greatest infl uence on outcome. Future studies will allow further clarifi cation of the associations that we observed on our unit. We hope that future studies will also help better identify the elements that determine which patients experience high oxygen demands, thereby permitting earlier interventions and improved outcomes. Additionally, further explorations of methods for providing geriatric palliative care within the confi nes of a pandemic will improve the emotional and psychologic support we can offer our patients, their families and our medical teams.