This study followed the ISO 14155:2020 guidelines for clinical trials. Patients undergoing mechanical ventilation through an artificial airway were divided into two groups: one using a test device (Lmeca.A1000, an electric medical suction device) and the other using a standard manual airway suction procedure. This study aimed to compare and assess the effectiveness and safety of the test device. Initially, 117 participants consented, of whom 116 were randomly assigned to participate in the trial. Ultimately, after excluding those who dropped out, 109 participants were included in the analysis. To our knowledge, this multicenter, prospective, randomized, open-label, pivotal study is the first attempt to evaluate the efficacy and safety of any automatic airway suction device.
For the primary efficacy assessment, the presence of grade 3 or higher airway mucosal injury was evaluated before and 3 days (72 h) after the application of the clinical trial medical device. No significant differences were observed between the two groups. The 95% two-sided confidence interval for the difference between the test group and the control group had a lower limit of -0.04 and an upper limit of 0.12. Notably, the lower limit exceeded the non-inferiority threshold (-0.2), indicating that the test group was not inferior to the control group, and thus met the criteria for non-inferiority.
For the secondary efficacy assessment, the presence of grade 3 or higher airway mucosal injury was examined before and 7 and 14 days after the application of the clinical trial medical device. However, on the 14th day, only one patient from the experimental group remained; thus, this patient was excluded from the analysis. There was no statistically significant difference in the degree of airway mucosa injury 7 days after application. The 95% two-sided confidence interval for the difference between the test group and the control group had a lower limit of -0.19 and an upper limit of 0.05. Remarkably, the lower limit exceeded the non-inferiority threshold (-0.2), indicating that the test group was not inferior to the control group and met the criteria for non-inferiority.
Furthermore, there was no statistically significant difference in the incidence of ventilator-associated pneumonia between the two groups throughout the 2 weeks following the application of the clinical trial medical device. The 95% two-sided confidence interval for the difference between the test group and the control group had a lower limit of -0.03 and an upper limit of 0.10. Notably, the lower limit exceeded the non-inferiority threshold (-0.2), indicating that the test group was not inferior to the control group.
The incidence of malfunction in the clinical trial medical device within 3 days (72 h) of application was 0.03%, and within 2 weeks of follow-up, it was 0.02%. These results indicate that the device operated effectively, with more than 90% of the patients functioning properly during the 3-day (72-hour) and 2-week follow-up periods.
Based on these results, we concluded that the degree of injury to the airway mucosa caused by the use of the clinical trial medical device was not inferior to that caused by the previous manual method. In addition, the device operated normally more than 90% of the time, thereby demonstrating its effectiveness.
In this clinical trial, adverse events occurred in eight of 56 patients (14.29%) in the test group and seven of 53 patients (13.21%) in the control group. However, it was confirmed that there was no causal relationship with the use of the airway suction device. Cardiac arrest and bradycardia were confirmed in one patient each in the test group; however, both cases occurred during the resting period when the medical device did not perform suction. The subject who developed cardiac arrest of unknown cause had spontaneous circulation restored after two cycles of CPR and recovered after concurrent treatment with therapeutic drugs. Based on this, it was concluded that there was no difference in stability compared with the previous manual method.
An analysis was conducted to determine whether the use of clinical trial medical devices would improve vital signs, assist in gas exchange, and improve blood test indicators of inflammation. Statistically significant differences between the test and control groups were confirmed in temperature, WBC count, ANC, PT (INR), and CRP levels at some time points. However, this trend was not clear upon detailed analysis.
Compared to the start of the test, the degree of decrease in body temperature was significantly greater in the experimental group than in the control group after 3 days (72 h) of application. However, the average body temperature in both groups was within the normal range, and there were no significant differences between the two groups on other days. Other vital sign indicators did not show clear trends or statistically significant differences; therefore, the use of clinical trial medical devices could not be considered to have an effect on the improvement of vital signs.
ABGA revealed that the only significant difference between the two groups was PaCO2. On the 8th day of application of the airway suction device, PaCO2 was significantly higher in the experimental group than in the control group, and the degree of increase was significantly greater than that at the start of the test. However, there were no significant differences on other days, and there was no consistent trend of higher values in the experimental group. Despite the differences between the two groups, the average PaCO2 values in each group were within the normal range. The pH and bicarbonate levels were within the normal range for both groups at all time points, and there was no significant difference in PaO2 and PaO2/FiO2 ratio between the two groups. Based on these results, there was no evidence that the clinical trial medical device helped with gas exchange.
WBC and ANC levels were initially higher in the experimental group but tended to be consistently lower than those in the control group after 2 days (48 h). However, there was only one day when both indicators were significantly lower in the experimental group, and at that point, the sample sizes for both groups were very small (< 5 patients). There was no consistent trend for CRP levels to be superior or inferior between the two groups, and there was a time when there was a statistically significant decrease in the experimental group compared with the start of the test; however, at this time, the number of patients in the two groups was ≤ 5. Procalcitonin levels tended to be consistently lower in the experimental group from day 6; however, there was no statistically significant difference between the two groups. In the case of PT (INR), the degree of decrease in PT (INR) in the control group was consistently greater during the observation period than that at the start of the test; however, there was no significant difference in the value of PT (INR) compared with the test group. Clinically, there were no differences in bleeding tendencies between the two groups. When these blood test results were combined, there was no evidence to suggest that the inflammatory response was further improved in the test group.
To assess user convenience, we surveyed the nurses in charge of the advantages and disadvantages of using medical devices in clinical trials. Using the device reduced the need for direct airway suction by the nurse, saving time by only needing to ensure that the machine's settings were appropriate and deciding whether the application interval needed adjustment. Previously, missing the right time for airway suction owing to other tasks was a problem. However, automated devices enabled the maintenance of regular suction intervals, which many nurses appreciated for reducing their workload. However, the inability to objectively judge proper suction resulted in intermittent manual suction, potentially undermining the trust in the device among medical staff. Other concerns include doubts regarding device reliability. It was noted that the machine operates based on set intervals and times regardless of the patient's reactions or abnormal signs, making it unable to address adverse reactions unless attended to by medical staff. Currently, the device lacks the capability to detect such changes and requires medical staff to recognize and intervene with alarms using patient monitors or mechanical ventilators. Future plans include establishing safety protocols to recognize excessive pressure changes in the airway or halt machine operations based on patient-monitoring readings.
Limitations
The major limitation of this study was the inconsistency in the follow-up periods of all patients. Consequently, 6 days after initiating the observation period, only half of the participants in each group remained, compared with the beginning. This discrepancy may have introduced a bias in the analysis of the long-term follow-up period after 3 days. Particularly concerning the blood test results, significant differences between the two groups occurred when the number of patients in each group was reduced to 10 or fewer, rendering a meaningful analysis impractical.
Another potential limitation is the variability in physicians performing bronchoscopies across different institutions. However, post-study endoscopic findings indicated that the degree of damage for most subjects was consistently the same when evaluated independently by researchers at each institution in a double-blind manner. In instances of discordant opinions, a double-blind evaluation method was employed to reconcile the differences. Therefore, we believe that issues stemming from different observers were minimal.
Unlike in the experimental group, in the control group, airway suction was directly performed by the attending nurse, leading to potential variations in suction strength and application time depending on the individual medical personnel conducting the procedure. The standardization of these factors proved challenging, and only the application interval could be regulated uniformly. Consequently, we were limited in our ability to achieve consistent treatment across the control group.
This study did not assess the efficacy of foreign substance removal from the airways owing to the absence of a quantitative method to evaluate the effectiveness of airway suction in removing sputum. As the quantity of sputum present in the respiratory tract varies among patients, it is impractical to determine the extent of removal solely by measuring the aspirated sputum volume. When considering the measurement of aspirate collected in the drainage container as an indirect approach, it was found that the container also contained normal saline used for cleaning the suction tube, making it impossible to accurately calculate the sputum volume. To address this limitation, the most effective approach is to observe the suction process in real-time through video monitoring. Currently, our research team is planning to develop a method for real-time observation of the airway suction process, similar to bronchoscopy.