Study design and setting
This was a prospective, multicenter, randomized study conducted at seven participating hospitals in Beijing, China. Patients with sepsis will be enrolled between December 2023 and February 2025. The investigators at each participating center will screen potential patients with sepsis to determine their eligibility for the study. Prior to commencing the study, the study underwent review and received approval from the institutional review board of Beijing Chaoyang Hospital, which is affiliated with Capital Medical University (Approval No.2022-620-1), as well as the ethics committee of each participating center. Informed consent will be obtained from all participants or their next of kin before their inclusion in the study.
The study was registered on the Clinical Trials website. gov (registration no. NCT05634057).
Inclusion and exclusion criteria
The inclusion criteria for this study were as follows:
1) Patients must be over 18 years old.
2) Patients must have a diagnosis of sepsis, as defined by The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) [1].
3) Patients must be willing to provide informed consent by signing the consent form.
The exclusion criteria for this study were as follows:
1) Patients who were expected to die within 24 hours after enrollment.
2) Contraindications to low-molecular-weight heparin and scopolamine butylbromide drugs.
3) Thrombotic disease requiring treatment with low-molecular-weight heparin.
4) Patients with terminal-stage malignancies, severe immunodeficiency, immunosuppression, severe liver or kidney dysfunction (defined as liver or kidney SOFA score ≥ 3 points), etc.
5) Pregnant or lactating women.
6) Patients participating in other clinical trials.
The methods of recruiting subjects involved utilizing hospital records, briefing family members and/or patients about the study by doctors and researchers, and allowing family members and/or patients to make a decision regarding participation in the trial.
Diagnostic criteria for sepsis (including septic shock) [1]
1. Patient with suspected infection.
2. qSOFA ≥ 2.
3. Assessment of evidence of organ dysfunction
4. SOFA ≥ 2.
5. Despite adequate fluid resuscitation, vasopressors are required to maintain a MAP ≥ 65 cm Hg and a serum lactate level > 2 mmol/L.
Intervention
In the control group, the initial treatment strategy followed the recommendations outlined in the Surviving Sepsis Campaign (SSC) guidelines [7] within the first 24 hours after the diagnosis of sepsis. This strategy included two variables:
1. The following resuscitation measures were implemented within the first 6 hours of septic shock:
-
Measurement of lactate levels
-
Collection of cultures prior to initiating antibiotic treatment
-
Administration of broad-spectrum antibiotics
-
Administration of fluids and vasopressors to achieve a mean arterial pressure (MAP) > 65 mmHg
2. All treatment measures were implemented within 24 hours, including the following:
In the intervention group, a combination therapy approach was employed, which included Ani HBr injection, low-molecular-weight heparin, and traditional treatment as per the SSC guidelines. The intervention protocol was as follows:
Ani HBr Injection:
Low-molecular-weight heparin:
Study endpoint
The primary endpoint of this study was 28-day mortality, which was assessed by determining the mortality rate of patients over 28 days. The patients will be followed up for 28 days. In the event that patients are discharged earlier than 28 days, their survival time will be recorded as of day 28 after admission or diagnosis with septic shock.
The secondary endpoints of the study included the following:
- Lactate clearance rate: The rate of lactate clearance will be evaluated at specific time points, namely, 6 hours, 24 hours, and 72 hours after admission. This will involve measuring the reduction in lactate levels, which serve as an indicator of the patient's response to treatment.
- Length of stay in the ICU and hospital: The duration of ICU and hospital stay will be recorded, providing insights into the patient's healthcare resource utilization and recovery trajectory.
- Organ failure-free days: Organ failure will be evaluated using the Sequential Organ Failure Assessment (SOFA) score. The number of organ failure-free days was calculated, indicating the duration during which the patient did not experience organ failure.
- Number of days with vasopressor use within 28 days.
- Septic shock conversion rate.
- The 28-day mortality rate of septic shock patients.
- The 72-hour sublingual microcirculation.
These secondary study endpoints aim to provide additional information on the effectiveness of the intervention and its impact on various clinical outcomes related to sepsis.
Adverse events
Patient safety will be a paramount consideration throughout the study, and measures will be in place to monitor and address any potential adverse events. At each study visit, the safety of the patients will be carefully assessed and monitored.
Adverse events [19] that are closely monitored include but are not limited to:
1. New-onset psychosis: Any occurrence of new psychiatric symptoms or changes in mental status will be thoroughly evaluated and reported.
2. Urinary retention: Instances of urinary retention, where the patient experiences difficulty or inability to empty the bladder, will be monitored and documented.
3. Significant hypotension: Patients with severe low blood pressure will be closely monitored and managed appropriately.
4. Tachycardia: Instances of rapid heart rate exceeding the normal range will be monitored and documented.
In the event of any adverse event, patients will be promptly informed and instructed to discontinue their participation in the study. Patient safety is of utmost importance, and the study protocol includes provisions to ensure that the well-being and safety of the patients are prioritized. This includes closely monitoring adverse events and taking appropriate actions to mitigate any risks or harm to the patients involved in the study.
Data collection
In this study, data collection will be conducted by research coordinators using paper data forms. The case report form (CRF) will be written in Chinese to facilitate effective communication among investigators. To ensure accuracy, the correctness of the original CRFs will be carefully verified.
Subsequently, a database will be established using an electronic data capture (EDC) system. The data will be entered into the database by two trained data entry clerks independently. Following the initial data entry, a process of double entry verification will be conducted to compare the entries and identify any inconsistencies or errors.
In case of any inconsistencies or discrepancies found within the database, a thorough item-by-item review of the original record table will be performed for proofreading and correction. Once all the necessary corrections and verifications are completed, the database will be locked to prevent further changes.
This data collection and entry process ensures the integrity and accuracy of the collected data, reducing the chances of errors or discrepancies during the study analysis phase.
Group sequential analysis and sample size estimation
This study follows a superiority trial design and employs the group sequential method to estimate the required sample size. Considering the cost associated with Ani HBr and the potential effectiveness of interim analyses, the trial has provisions to stop early under certain circumstances. The reasons for early termination of the clinical trial may include the following:
1. Serious toxicity or adverse events: If there are significant safety concerns or unexpected severe adverse events associated with the treatment, the trial may be stopped early to ensure patient safety.
2. Established benefit: If interim analyses indicate a clear and statistically significant benefit of the treatment being investigated, it may be considered unethical to continue the trial, and it may be stopped early to provide the treatment to a larger population.
3. Design or logistical difficulties that are too serious to fix: In situations where there are significant design or logistical challenges that cannot be resolved without compromising the integrity or validity of the trial, the study may be stopped early.
To facilitate the design of group sequential trials, the R package "gsDesign" will be utilized. This package allows for the implementation of group sequential analyses in which promising treatments can be selected at various interim analyses. The group sequential method enables efficient monitoring of treatment effects and provides the flexibility to potentially stop the trial early if specific criteria are met.
By incorporating the group sequential method and the ability to perform interim analyses, this trial aims to optimize the use of resources, reduce costs, and efficiently evaluate the effectiveness and safety of the treatment under investigation.
Parameters:
Based on the information provided, it appears that a group sequential design will be implemented with lower bound spending under the null hypothesis, three-stage analyses, 80% power, and a 2.5% (1-sided) type I error rate. The following details were specified:
1. The mortality rate in the control group is assumed to be 29%, and the new intervention is expected to reduce the mortality rate by 9%.
2. Fixed sample size: The sample size for the fixed design with no interim was 782, which was calculated by PASS21.
3. Timing of interim analyses: Interim analyses will be conducted at three equally spaced accrual sample sizes: 266, 531, and 796 (Table 1).
4. Spending functions: The Hwang-Shih-DeCani [20] spending function will be used, with lower bound spending with gamma=-2 and alpha spending with gamma=-4.
The expected behavior of the trial and the sample size calculations are summarized as follows:
- Maximum total sample size: The maximum total sample size for the trial was 796.
- Alpha spending: Table 2 provides the alpha spending function, which indicates the cumulative type I error rate at each interim analysis.
- Beta spending: Table 3 presents the beta spending function, which indicates the cumulative type II error rate at each interim analysis.
- Boundary crossing probabilities: Figure 2 illustrates the upper boundary crossing probabilities. If any interim analysis crosses the upper boundary, the trial will be stopped. However, the trial will continue if any analysis crosses the lower boundary.
- Spending function plot: Figure 3 displays the spending function plot for alpha and beta, derived using the Hwang-Shih-DeCani spending function.
- Expected sample size: Figure 4 shows the expected sample sizes for different underlying risk factors. At the assumed risk difference of 0.1, the expected sample size is estimated to be 661, which is smaller than the fixed sample size of 782.
By implementing the group sequential design with the specified parameters, the trial aims to achieve more efficient sample size utilization while maintaining sufficient statistical power to evaluate the effectiveness of the new intervention.
Table 1
Upper and lower bounds expressed at different scales
|
|
|
---------Lower bounds---------
|
--------- Upper bounds---------
|
Analysis*
|
n.I**
|
N***
|
Z****
|
Nominal p*****
|
Spend+
|
Z
|
Nominal p
|
Spend++
|
1
|
0.3393
|
266
|
-1.06
|
0.1446
|
0.1446
|
3.01
|
0.0013
|
0.0013
|
2
|
0.6786
|
531
|
-0.23
|
0.4082
|
0.2817
|
2.55
|
0.0054
|
0.0049
|
3
|
1.0179
|
796
|
2.00
|
0.9772
|
0.5487
|
2.00
|
0.0228
|
0.0188
|
Total
|
|
|
|
|
0.9750
|
|
|
0.0250
|
*3 equally spaced analyses.
**Average sample number/expected sample size is computed using n.I. at each analysis times the probability of crossing a boundary at that analysis. If no boundary is crossed at any analysis, this is considered stopping at the final analysis.
***Expected sample size for each analysis.
**** Test statistics of each analysis.
*****P value of each analysis.
++β-spending of each analysis. ++αspending of each analysis.
Table 2
Upper boundary (power or Type I Error)
|
Analysis
|
|
|
|
|
δ
|
1
|
2
|
3
|
Total
|
E{N}
|
0.00
|
0.0013
|
0.0049
|
0.0187
|
0.025
|
642.5
|
0.10
|
0.0840
|
0.3284
|
0.3876
|
0.8000
|
661.2
|
δ represents the risk difference. Upper boundary crossing probabilities and assume that any cross stops the trial.
Table 3
Lower boundary (futility or Type II Error)
|
Analysis
|
|
|
|
δ
|
1
|
2
|
3
|
Total
|
0.00
|
0.1446
|
0.2817
|
0.5487
|
0.975
|
0.10
|
0.0036
|
0.0045
|
0.1919
|
0.200
|
δ represents the risk difference. We set lower bound spending under the null hypothesis with a nonbinding lower bound and assume that any cross stops the trial.
Randomization
In this study, blocked randomization was employed to allocate subjects to either the combination therapy group or the control group. The randomization was performed at a 1:1 ratio using blocks of sizes 2,4,6,8 to 796 subjects.
To ensure random assignment concealment, a web-based central random system was utilized. The person in charge of each center accessed the network system and entered the basic information of the subjects who met the inclusion and exclusion criteria. The central random system automatically assigned a random number to each subject. Once a random number was assigned, it could not be assigned to another subject. It is important to note that subjects who failed to complete the entire study could not be included.
Random sequence allocation and allocation hiding were implemented through the use of a central stochastic system. This ensured that the allocation process was unbiased and concealed from the investigators involved in the study.
This study is an open-label trial, meaning that both the researchers and participants are aware of the treatment assignment. The protocol details can be found in Fig. 1, which provides a summary of the study design and procedures.
Statistical analysis
The primary and secondary efficacy analyses in this study were conducted using the intention-to-treat (ITT) analysis set. The analysis will involve baseline variables, which include both measurement data and count data.
The measurement data are presented as the mean (standard deviation, SD) or number (percent), while the count data are expressed as the median and interquartile range.
To compare the means of continuous random variables between the two groups, t tests will be utilized. The proportions test will be employed for dichotomous response variables, and the log-rank test will be used for censored survival data. For the binary variable of mortality, a chi-square test will be used for the comparison.
A mixed linear model will be utilized to assess the effect of treatment on outcomes and to perform repeated-measures analysis of variance. The statistical tests will be two-sided, and a p value less than 0.05 will be considered to indicate statistical significance.
All the statistical analyses will be conducted using R software, specifically version 4.3.1. R is a widely used statistical programming language that provides a range of statistical functions and packages for data analysis.
By employing appropriate statistical methods and software, this study aimed to analyze the data accurately and determine the significance of the treatment effect on the evaluated outcomes.