Characteristics of Eligible Studies
We initially identified 9,363 records, and 8971 citations remained after the duplicate trials were removed; 145 RCTs were eligible after preliminary screening by title and abstract. Finally, Fifty-three RCTs [17–69] that included 12,310 patients with sepsis were included in this meta-analysis (Fig. 1). The characteristics of the included RCTs are listed in Table 1. Twenty-eight RCTs[18, 20–23, 25–29, 35, 36, 38, 40–42, 45, 46, 48, 52, 54, 57, 59–63, 67] on 8,503 patients with septic shock, 8 RCTs[37, 39, 43, 44, 50, 55, 68, 69] on 936 patients with sepsis, 4 RCTs[17, 47, 49, 65] on 390 patients with sepsis and ARDS, 9 RCTs[24, 30, 33, 34, 51, 53, 56, 58, 66] on 1,733 patients with sepsis and community-acquired pneumonia, and 4 RCTs[19, 31, 32, 64] on 748 patients with severe COVID-19 were included. Additionally, 30 RCTs[19, 23, 25, 27, 28, 30, 31, 33, 35, 36, 38–43, 46, 48, 50, 52–56, 59, 61–63, 65, 67] (7,115 patients) of which were treated with hydrocortisone, 4 RCTs[18, 20–22] (2,082 patients) with hydrocortisone plus fludrocortisone, 10 RCTs[17, 26, 32, 33, 37, 47, 49, 57, 60, 66] (1,245 patients) with methylprednisolone, 4 RCTs[34, 58, 68, 69] (364 patients) with prednisolone, 3 RCTs[29, 51, 64] (408 patients) with dexamethasone, and only one RCTs with betamethasone (85 patients) [44]and corticosteroid (21 patients)[45], respectively.
Table 1
Characteristics of the Included Studies in the Meta-analysis of Corticosteroids Vs Placebo or Standard Supportive Care in Adults With Sepsis or Septic Shock
Study | Study type | Single/Multi Center | Study period | Total Patients / Patients in Corticosteroids No. | Mean Age, y | Female/ Male of Patients No. | Type of Patients Population | Sepsis or Septic Shock Definition | Experimental Intervention | Reported Outcomes |
Annane et al (2018) | RCT | M | NA | 1241/614 | CS:66 PC:66 | 415/826 | Septic shock | Sepsis-3.0 | Hydrocortisone (50mg) intravenously every 6 hours plus fludrocortisone(50ug tablet) nasogastric tube for 7d | 28d,90d, 180,ICU discharge and hospital discharge all- cause mortality; Vasopressor‑free days; Ventilator‑free days; Organ‑failure–free days |
Venkatesh et al (2018) | RCT | M | 03/2013-04/2017 | 3658/1832 | CS:62.3 PC:62.7 | 1399/2259 | Septic shock | Sepsis-3.0 | Hydrocortisone (200mg) intravenous infusion per day for 7d | 90d and 28d mortality; ICU discharge and hospital discharge time; Ventilator‑free days |
Annane et al (2002) | RCT | M | 10/1995-02/1999 | 300/151 | CS: 62 PC: 60 | 200/100 | Septic shock | Sepsis-2 | Hydrocortisone bolus (50mg) every 6 h and fludrocortisone (50 µg) taken orally every 24 h for 7 d | 28d mortality |
Lv et al (2017) | RCT | S | 09/2015-09/2016 | 118/58 | CS: 68.8 PC: 64.8 | 70/68 | Septic shock | NA | Hydrocortisone (200 mg) daily as a continuous infusion for 6 d | 28d mortality; the reversal of shock; in-hospital mortality; the duration of ICU; hospital stay |
Klastersky et al (1971) | RCT | S | NA | 85/46 | NA | 47/38 | Severe sepsis | NA | Betamethasone 0.5 mg/kg every 12 h for 3 d | 30d mortality |
Table 1
Study | Study type | Single/Multi Center | Study period | Patients / Patients in Corticosteroids No. | Mean Age, y | Female/ Male of Patients No. | Type of Patient Population | Sepsis or Septic Shock Definition | Experimental Intervention | Reported Outcomes |
Bone et al (1987) | RCT | M | 11/1982-12/1985 | 382/191 | CS: 53.0 PC: 53.6 | 147/235 | septic shock | NA | Methylprednisolone bolus (30mg/kg) repeated every 6 h for 24 h | Shock incidence; shock reversal; overall mortality; 14d mortality |
Schumer et al (1976) | RCT | S | 1967- 1975 | 172/86 | 50 | 5/167 | Septic shock | NA | Group 1:methylprednisolone (30 mg/kg); Group 2: dexamethasone(3 mg/kg); dose was repeated once in both groups after 4 h and had to be initiated at the time of diagnosis | Mortality; age-associated mortality; severity of shock associated mortality; underlying condition; associated mortality; organ injury-associated mortality; complications |
Sprung et al (1984) | RCT | M | 8/1979-2/1982 | 59/43 | CS: 58 PC: 55 | 13/46 | Septic shock | NA | Group 1:methylprednisolone (30 mg/kg); Group 2: dexamethasone(6 mg/kg); dose was repeated once in both groups after 4 h if shock persisted | Shock reversal; hospital mortality; blood cultures; adverse events |
Vasscsg et al (1987) | RCT | M | 10/1983-04/1986 | 223/112 | CS: 60.9 PC: 60.6 | NA | Sepsis | NA | Methylprednisolone bolus (30mg/kg) repeated every 6 h for 24h | 14d mortality; adverse occurrences |
Primary Outcome
Forty trials[17, 18, 21–30, 35–39, 42–44, 47–51, 54–62, 64–69] (10,612 patients), 24 trials[21–23, 25, 27, 28, 30, 33–36, 43, 47–49, 51, 53, 55, 59, 62, 66–68] (11,613 patients), and 17 trials[21–23, 27, 28, 30, 35, 36, 38, 43, 49, 55, 56, 59, 66, 67] (7,175 patients) were included in this meta-analysis for assessing the 28-day mortality, in-hospital mortality, and ICU mortality, respectively. We used the random-effects model with RRs to assess the pooled results. Corticosteroid therapy showed a no small reduction in the 28-day mortality (RR, 0.94; 95% CI, 0.87–1.02; evidence rank, moderate; Fig. 2), with low heterogeneity among the trials (I2 = 24%). However, corticosteroids treatment resulted in a significant decreased in the in-hospital mortality (RR, 0.93; 95% CI, 0.88–0.99; evidence rank, moderate; Fig. 3) and ICU mortality (RR, 0.89; 95% CI, 0.80–0.98; evidence rank, high; Fig. 4) with low heterogeneity (I2 = 11% and I2 = 22%, respectively). The Funnel plot and Egger test showed no publication bias in the 28-day mortality (P = 0.15), in-hospital mortality (P = 0.01), and ICU mortality (P = 0.02) (Supplemental Fig. 1–3 in the Supplement). The results of sensitivity analysis showed that the models of the 28-day mortality, in-hospital mortality, and ICU mortality were credible (Supplemental Fig. 4–6 in the Supplement). Furthermore, L’Abbé plot reported that the mortality in the placebo group increased significantly than the corticosteroid group, suggesting the potential effects of corticosteroids in septic patients (Supplemental Fig. 7–9 in the Supplement).
Secondary Outcomes
Supplemental Fig. 10–23 in the Supplement present the assessment of the secondary outcomes. Corticosteroids achieved a small reduction in length of stay in hospital (MD, -1.54; 95% CI, -2.68 to -0.41; I 2 =20%; evidence rank, high), SOFA scores at day 7 (MD, -0.90; 95% CI, -1.72 to -0.08; I2 = 93%; evidence rank, low) and time to resolution of shock (MD, -1.36; 95% CI, -1.79 to -0.93; I2 = 68%; evidence rank, low) for patients with sepsis. Conversely, corticosteroids resulted in higher risk of hypernatremia (RR, 1.51; 95% CI, 1.11–2.00; I2 = 0%; evidence rank, moderate) and hyperglycemia (RR, 1.18; 95% CI, 1.09–1.28; I2 = 56%; evidence rank, high). Furthermore, corticosteroids increased the vasopressor-free days (MD, 1.93; 95% CI, 0.76–3.09; I2 = 0%; evidence rank, moderate), ventilation-free time (MD, 1.46; 95% CI, 0.27–2.65; I2 = 21%; evidence rank, moderate), shock reversal at day 7 (RR, 1.18; 95% CI, 1.08–1.29; I2 = 71%; evidence rank, moderate) and day 28 (RR, 1.07; 95% CI, 1.02–1.11; I2 = 6%; evidence rank, moderate). Additionally, corticosteroids achieve no reduction in the long-term mortality (> 60 d) (RR, 0.94; 95% CI, 0.84–1.04; I2 = 62%; evidence rank, low), length of stay in ICU (MD, -0.89; 95% CI, -1.80–0.03; I2 = 47%; evidence rank, moderate), superinfection (RR, 1.05; 95% CI, 0.92–1.19; I2 = 11%; evidence rank, moderate) and gastroduodenal bleeding (RR, 1.18; 95% CI, 1.09–1.28; I2 = 0%; evidence rank, high).
The Funnel plot and Egger test showed no publication bias in the length of stay in hospital (P = 0.53), SOFA scores at day 7 (P = 0.82), hyperglycemia (P = 0.14), the shock reversal at day 7 (P = 0.05), length of stay in ICU (P = 0.15), superinfection (P = 0.03), and gastroduodenal bleeding (P = 0.97) (Supplemental Fig. 24–30 in the Supplement). Shock reversal at day 28 showed a significant publication bias in the Funnel plot and Egger test (P = 0.046) (Supplemental Fig. 31 in the Supplement). The results of the sensitivity analysis showed that the models of the above mentioned outcomes, including length of stay in hospital, SOFA scores at day 7, hyperglycemia, shock reversal at day 7, length of stay in ICU, superinfection, gastroduodenal bleeding, and shock reversal at day 28 were credible (Supplemental Fig. 32–38 in the Supplement).
Importantly, risk of bias were reported in the forst plot of each outcome, and the evidence rank was showed in Table 2.
Table 2
The findings and evidence rank of the included studies with corticosteroids vs placebo treatment in patients with sepsis.
Pooled results | No.of Patients (No. of Studies) | Relative Effect, RR, or MD (95% CI) | Heterogeneity I2,% | Absolute effect (95%CI ) | Evidence rank |
Primary outcomes | | | | | |
28d mortality | 10,612 (40) | 0.94 (0.87, 1.02) | 24 | 17 fewer per 1000 (from37 fewer to 6 more) | Moderate1 |
In-hospital mortality | 11,613 (24) | 0.93 (0.88, 0.99) | 11 | 22 fewer per 1000 (from 3 fewer to 37 fewer) | Moderate1 |
ICU mortality | 7,199 (17) | 0.89 (0.8,0.98) | 22 | 31 fewer per 1000 (from 6 fewer to 56 fewer) | High |
Secondary outcomes | | | | | |
Long-term mortality | 6,254 (9) | 0.94 (0.84, 1.04) | 62 | 24 fewer per 1000 (from 64 fewer to 16 more) | Low2,3 |
Shock reversal at 7d | 6,651 (16) | 1.18 (1.08,1.29) | 71 | 211 more per 1000 (from 54 more to 195 more) | Moderate2 |
Shock reversal at 28d | 2,997 (12) | 1.07 (1.02,1.11) | 6 | 49 more per 1000 (from 14 fewer to 77 more) | Moderate2 |
Gastroduodenal bleeding | 5,157 (24) | 1.18 (1.09,1.28) | 0 | 3 more per 1000 (from 7 fewer to 16 more) | High |
Superinfection | 5,490 (25) | 1.05 (0.92, 1.19) | 11 | 9 more per 1000 (from14 fewer to 33 more) | Moderate2 |
Hypernatremia | 4,554 (3) | 1.51 (1.11,2) | 0 | 12 more per 1000 (from 3 more to 24 more) | Moderate2 |
Hyperglycemia | 8,787 (20) | 1.18 (1.09,1.28) | 56 | 49 more per 1000 (from 24 more to 76 more) | High |
Vasopressor-free days | 1,316 (2) | 1.93 (0.76, 3.09) | 0 | 1.93 more per 1000 (from 0.76 more to 3.09 more) | Moderate2 |
Ventilation-free days | 1,812 (4) | 1.46 (0.27, 2.65) | 21 | 1.46 more per 1000 (from 0.27 more to 2.65 more) | Moderate2 |
length of stay in hospital | 8,539 (19) | -1.54 (-2.68, -0.41) | 20 | 1.54 fewer per 1000 (from 2.68 fewer to 0.41 fewer) | High |
length of stay in ICU | 8,166 (22) | -0.89 (-1.80, 0.03) | 47 | 0.89 fewer per 1000 (from 1.8 fewer to 0.03 more) | High |
time to resolution of shock | 4,134 (5) | -1.36 (-1.79, -0.93) | 68 | 1.36 fewer per 1000 (from 1.79 fewer to 0.93 fewer) | Low2,3 |
SOFA score at day 7 | 3,076 (13) | -0.90 (-1.72, -0.08) | 93 | 0.9 fewer per 1000 (from 1.72 fewer to 0.08 fewer) | Low2,3 |
Subgroup Analysis
We performed subgroup analysis based on the sepsis subtype or type of corticosteroids used for the primary outcomes or I2 > 75% in the secondary outcomes with more than 10 trials for each outcome. The results of the subgroup analysis showed no effect on the 28-day mortality; however, the in-hospital and ICU mortality were significantly improved in the hydrocortisone plus fludrocortisone treatment and in the patients with septic shock, sepsis, and community-acquired pneumonia (Supplemental Fig. 39–44 in the Supplement). Moreover, the result of subgroup in SOFA scores at day 7 represented that the main original heterogeneity may be from the trials with smaller samples who were given hydrocortisone treatment or trials on patients with sepsis shock (Supplemental Fig. 45–46 in the Supplement).