Critical Care for Severe COVID-19: Insights from Sichuan Province, China
The COVID-19 pandemic has had a profound global impact, with approximately 19% of cases progressing to severe or critical disease, a high-risk group for death. The case-fatality rates among severe COVID-19 patients vary widely across regions, from zero to 61.5%, and the reasons behind this remain unclear. In this study, we explore the critical care management of severe COVID-19 in Sichuan province, China, where the 28-day case-fatality rate was notably low (0.6% in all patients and 3.7% in severe cases).
Study Design and Patient Selection
The multicentre cohort study (SUNRISE, ChiCTR2000029758) involved 21 hospitals designated for severe COVID-19 patients in Sichuan. Data were collected prospectively (for patients still in the hospital after enrolment) and retrospectively (for others) between January 16 and March 15. Severe cases were defined based on five criteria from the National Health Commission of China, including dyspnoea, SpO2 <93% without oxygen therapy, P/F ratio 50% within 24-48 hours, and respiratory failure, septic shock, or multiple organ dysfunction. Patients were followed from day 1 (D1) until discharge, death, or the end of the study. Clinical outcomes at D28 were classified as rapid recovery (RR), prolonged recovery (PR), or no recovery (NR).
Patient Characteristics
Of 539 patients, 81 were identified as severe. The median age was 50 (39-65) years, 37.0% were female, and 50.6% had a body mass index ≥24 kg/m². Elderly patients (age ≥65 years) accounted for 28.4%, and 53.1% had chronic comorbidities. The most common severe diagnostic criteria were P/F ratio <300 mmHg (87.7%), SpO2 <93% without oxygen therapy (66.7%), and dyspnoea (27.2%).
Clinical Outcomes
By D28, 65.4% (53 patients) were discharged (RR), 22.2% (18 patients) were PR (13 on COT and 5 awaiting negative PCR results), and 12.3% (10 patients) were NR (3 deaths and 7 requiring advanced respiratory support).
ICU and Respiratory Support
Among the 18 designated hospitals, only two had standard ICUs, while 16 had provisional ICUs. All 81 severe cases were centralized to these ICUs, with 63.0% treated in provisional ICUs. Most patients (86.4%) were transferred from other hospitals, and 95.1% were admitted to ICUs by D1.
Respiratory support was crucial. On D1, 93.8% of patients received respiratory support (COT: 67.9%, NIV: 16.0%, HFNC: 9.9%). During the study, 97.5% used COT, 38.3% HFNC, 27.2% NIV, 12.3% IV, and 1.2% ECMO. COT was the most commonly used (62.7% of person-days), and patients using only COT had a high discharge rate (79.4%). For those needing escalation from COT, 48.0% were discharged by D28.
Implications for Diagnosis and Treatment
Early Identification
Using P/F ratio as a diagnostic criterion was more sensitive (87.7% of severe cases identified). Relying solely on dyspnoea or SpO2 would miss many early severe cases. Early identification led to timely respiratory support (93.8% on D1 with a median P/F ratio of 204 mmHg), contrasting with lower P/F ratios reported in previous studies. This is especially important for elderly or comorbid patients with atypical symptoms and faster disease progression.
Respiratory Support and Ventilator Shortage
The low use of IV (12.3%) and ECMO (1.2%) in Sichuan, compared to higher rates in other studies (e.g., 38.9%-71% for IV in Wuhan and US studies), may be due to timely intervention. Hypoxemia can contribute to multiple organ injury if not treated promptly. This finding is significant for regions facing ventilator shortages, suggesting that early and appropriate respiratory support (like COT and HFNC) can be effective.
Conclusion
This study highlights the importance of early and accurate diagnosis (using sensitive criteria like P/F ratio) and timely, appropriate respiratory support in managing severe COVID-19. The low case-fatality rate in Sichuan may be attributed to these strategies, providing valuable insights for global efforts in treating severe COVID-19, especially in resource-constrained settings.
doi: 10.1097/CM9.0000000000001187
Was this helpful?
0 / 0