S53 Figure 1Rates of mortality against cumulative number of antibiotics received per patient during inpatient spell.[Figure omitted. See PDF]ConclusionIn both COVID-19 waves, antibiotic administration correlated to increased inpatient morbidity and mortality. Given a near-linear relationship of mortality and cumulative antibiotic numbers, antimicrobial stewardship is essential, and tapering an appropriate therapy for likely responsible pathogens will yield lower mortality compared to overlapping coverage and inappropriate escalation. We strongly discourage the use of empirical antibiotics without supporting biochemical evidence of bacterial co-infection for possible future COVID-19 waves.ReferenceRussell C, et al. Lancet Microbe. 2021 Jun 2. https://doi.org/10.1016/S2666-5247(21)00090-2
Background Procalcitonin (PCT) is an established biomarker of acute bacterial infection, and elevated levels of PCT may correlate with increased severity in patients with COVID-19 infection.1 Here, we assess if initial PCT levels are a viable prognostic marker to predict significant morbidity and mortality outcomes for hospitalised patients admitted due to COVID-19 infection. Method We performed retrospective analysis of initial blood results taken from 1189 patients with RT-PCR positive COVID-19 infection presenting to our District General Hospital between 1st November 2020 and 28th February 2021. Mortality encompassed both inpatient and within 28 days post-discharge. Significant morbidity was defined as admission to the Intensive Care Unit (ICU) for organ support. PCT was measured using Brahm's chemiluminescent micro particle assay (CMIA). Elevated PCT was defined at two levels: PCT ≥0.5µg/L and PCT ≥0.2µg/L, to account for variance amongst literature.2 Regression analysis was performed to determine independent significance, accounting for comorbidity and demographics. Results We found elevated PCT levels conferred a significant two-fold increase in mortality and ICU admission. Initial PCT ≥0.5µg/L was associated with a significantly increased risk of mortality than those with PCT <0.5µg/L (46.5% vs 24.2%; OR 2.847, p=0.0001). Significantly higher mortality risk was also observed when using lower cut-off values, i.e, PCT ≥0.2µg/L vs PCT <0.2µg/L (OR 2.042, p=0.00001). A significantly higher rate of ICU admission for initial PCT ≥0.5µg/L (OR 2.041, p=0.007) or PCT ≥0.2µg/L (OR 1.918, p=0.0008) was also observed within our cohort. Conclusions Here, we report the largest single-centre study to date in analysing a UK-based population for procalcitonin in COVID-19. We observed a significant correlation between elevated initial levels of PCT and incidence of ICU admission and mortality within our cohort, thereby demonstrating promise for PCT as an effective prognostic marker. Using a higher cut-off for PCT ≥0.5µg/L increased mortality by almost 50%, but had no effect on morbidity. We suggest that a lower universal cut-off point for PCT should be used for detecting secondary bacterial infections and procalcitonin-guided antimicrobial therapy. References Hu R, et al. International Journal of Antimicrobial Agents 2020;56(2):106051. Vazzana N, et al. Acta Clin Belg. 2020 Sep 23:1–5.
Background Recent multicentre registries have shown cancer patients infected with SARS-CoV-2 have significantly higher mortality compared to patients without cancer1. Cancer-specific features associated with worse outcomes include advanced tumour stage, disease progression and lung cancer.1 Systemic anti-cancer treatments (SACT – chemotherapy, immunotherapy, targeted and hormone therapy) have been postulated to increase mortality from COVID-19 in cancer patients. Here, we aim to determine if cancer patients on SACT have a higher risk of mortality than those not on active treatment. Methods We retrospectively analysed cancer patients admitted to a Greater London District General Hospital between 1st November 2020 and 28th February 2021 with RT-PCR positive COVID-19. SACT was considered present if administered within 3 months of admission. Mortality encompassed hospitalised patients and those up to 28 days post-discharge. Association of cancer-specific demographics and mortality was assessed using logistic regression analyses adjusting for age, sex and comorbidities. Results Mortality rate was significantly higher in 122 cancer patients with COVID-19 than 1220 patients without cancer (OR 1.653, p=0.014), especially in patients with lung cancer (OR 4.664, p=0.002). 55 patients diagnosed with cancer within one year had a significantly higher mortality (OR 2.32, p=0.004). Stage 4, but not earlier stages of, cancer at diagnosis had much higher mortality (OR 2.82, p<0.001). Progression of cancer was highly predictive of mortality (OR 4.60, p=0.00002). SACT had no significant effect on mortality from COVID-19 disease when compared with cancer patients who had no active treatment. However, cancer patients that did not have SACT within 3 months were more likely to die (OR 1.80, p=0.025). Conclusion Among patients with cancer and COVID-19, mortality was high and associated with cancer-specific features. There was no evidence cancer patients on systemic anti-cancer treatments possessed higher mortality from COVID-19 disease, which correlates with findings from COVID-19 and cancer registries1. Patients that did not receive SACT within 3 months before COVID-19 and therefore more likely to have palliative treatment did demonstrate high mortality. Larger studies are needed to confirm the risk of mortality and timing of SACT before COVID-19 disease. Reference Lee AJ, et al. British Journal of Cancer 2021;124:1777–1784.
Background Bacterial infection has previously been observed in only 8% of COVID-19 patients,1 yet antibiotics are administered to 85% of inpatients. Typically, Gram negative organisms and Staphylococcus aureus are isolated as responsible pathogens.2 Here, we investigate the rates of bacterial infection during the second UK COVID-19 wave, its sources, the responsible organisms, and its impact on morbidity and mortality. Methods 1342 RT-PCR positive COVID-19 patients admitted to a Greater London District General Hospital between 1st November 2020 and 28th February 2021 were retrospectively analysed (44.6% female; mean age 68.8; mortality 26.9%). Mortality encompassed hospitalised patients and those up to 28 days post-discharge. Morbidity was assessed by length-of-stay and need for intensive care. Positive cultures due to contaminants were excluded. Independent correlation was assessed with multilinear regression analysis adjusting for demographics and comorbidities. Results 226 patients (16.8%) with COVID-19 had ≥1 bacterial infection. These patients possessed significantly higher independent mortality (35.0% vs. 25.3%, p<0.009), more frequently required intensive care support (19.5% vs 10.6%, p<0.00004), and required longer inpatient spells (19.6 days vs 9.46 days, p<0.0001). Greater number of positive culture types cumulatively increased mean length-of-stay (OR 9.08, p<0.00001); ≥2 culture type positivity (n=44) related to 31.5 days; ≥3 culture type positivity (n=12) 46.0 days. There was significantly higher mortality rate in patients receiving ≥1 antibiotic (30.2%) compared to no antibiotics (6.4%, p<0.00001). Conclusion Bacterial infection is observed far more frequently in COVID-19 patients than previously reported and adversely affects morbidity and mortality. Multiple sites of bacterial infection prolongs inpatient stay and increases mortality. Thorough culture collection should be encouraged in COVID-19 patients with biochemical evidence of bacterial infection to identify responsible pathogens and respective antimicrobial sensitivity. Given the higher mortality rates, empirical use of antibiotics in COVID-19 patients without supporting evidence of bacterial infection is strongly discouraged. References Lansbury, et al. J Infect. 2020 Aug;81(2):266–2. Russell C, et al. Lancet Microbe. 2021 Jun 2. https://doi.org/10.1016/S2666-5247(21)00090-2