Adverse Events during Prone Positioning of Patients with COVID-19 during a Surge in Hospitalizations—Results of an Observational Study
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors
Thank you for the opportunity to review this manuscript.
As a general remark, it seems, the authors reported extraordinarily high incidences of catheter dislocations and pressure ulcers in patients with ARDS treated with prone position in their own center. Although still speculative, the causes for the high number of PP-associated events are specific to the authors ARDS center and their specific pandemic-associated setting including newly recruited staff and lack of appropriate training. The results are not comparable to other published data or guidelines. It is unclear if this observational single center data collection will lead to improvements in patient care in the ICU of the authors, however, the results of this study appear not generally applicable.
LL 55 ff “The 56-bed COVID ICU was located in converted hospital departments (ex cardiac surgery, neurosurgery and psychiatry departments) for the purpose of treating a large number of patients who required mechanical ventilation.”
Please provide more information about the structure of the ICU and the setting of the ARDS treatment center
LL 70 “around 24h regardless of the time of the initiation of PP “
Why 24h? 24h are not evidence based standard and more flexibility gives the opportunity to have experienced staff present in case there is a shortage e.g. at nighttime.
LL 73 ff. Was standardized equipment used? Or just pillows?
LL 79 ff “During the PP sessions patients were deeply sedated and neuromuscular blocking agents were used as a lung protective ventilation strategy.”
Is this according to current or at that time actual official international guidelines and recommendations? – please discuss.
How was ventilation performed? Is deep sedation and the use of neuromuscular blocking agents considered a “lung protective ventilation strategy”? How were ventilator settings adapted?
LL 91 ff “Removal of central venous catheters, arterial catheters, drainage tubes, endotracheal tubes or other tubes was defined as any unplanned removal during PP session or within 24 hours after subsequent supination.“
Line exchange for central lines can usually only be done in supine position. In case line exchange was necessary for e.g. infectious reasons this would also count as adverse event of cvc-removal due to PP?
LL 99 ff: “after Kolmogorov-Smirnov test of normality”
Why was normality tested when only non-parametric testing for group comparison was performed? With a sample size of <100 and in these heterogenic patient population, normal distribution of data will be unlikely. Therefore, general use of non-parametric testing would be recommended.
Table 1 – Please specify which patients were included? All infected with COVID-19, all with mechanical ventilation or all patients with PP?
Results: The total number of PP should be reported to ease evaluation of the frequency of adverse events.
Table 2 – Reintubation in 19 patients was only in patients that had a tracheostomy?
Table 3 – Relative frequencies of total pressure ulcers sum up to 100%. Most patients had multiple pressure ulcers. Did they have different pressure ulcers of different stages? What location combinations were mostly detected?
Table 4 is hard to understand. What happened to patients with only one session of PP? Did all patients reported receive at least 5 sessions of PP? Did all survive at least 5 PP? Not considering these drop-outs could introduce a bias.
Given the next paragraph, average number of PP ranks between 2 without and 3 with adverse events.
Table 5 – see comments for Table 4. How are drop-outs considered? Were adverse events in the same patients counted separately?
Table 5 – number of patients per session – did 24 patients only receive one PP? and only 3 patients 5 PPs? What reasons were documented to proceed or to stop with PP?
LL 142 ff.: “However, the duration of individual sessions was observed only for the second PP session (the duration of the first session was 24h as per departmental strategy).“
So how long were the second and the other sessions? Was there a general protocol or standardization for PP apart from the first PP?
LL 146 “Presence of a third and fourth PP session did not correlate with a higher EPUAP stage pressure ulcer, most likely because of lack of power”
Regarding the number of investigated patients?
Figure 1 – what are the three similarly labeled Figure panels demonstrating?
193: ARDS
198: These numbers mainly demonstrate an inappropriate clinical management. Moreover, reasons for these rates of complications remain speculative. Setting, equipment, staff training, power, and many more all possible factors except the self-removals by the patients. Is there an opportunity for at least a historic control that could support the speculation?
LL 225 ff: “These differences can be partially explained by methodology – we defined unplanned removal as any removal within 24 after supination following PP.”
Only minimal contribution I sense. All patients were deeply sedated and had received neuromuscular blocking agents. Kinking and occlusion are also PP-associated and management-associated complications.
LL 231 ff: Speculative but highly likely part of the problem. However, it mainly demonstrates inappropriate standard of care and most importantly inappropriate and insufficient education and training of an evidence-based ARDS rescue treatment with “PP is a low-cost measure that can be performed without any (or minimal) additional medical devices”
Conclusion - Although still speculative, the causes for the high number of PP-associated events appears specific to the authors ARDS center and their specific pandemic-associated setting including newly recruited staff and potentially inappropriate training. If the assumed causes are valid, recommendation for appropriate education and training of PP could be suggested. Have the results triggered certain measures to improve quality of PP? Whether treatment of patients with ARDS apart from treatment with ECMO need to be done in specialized centers only can be discussed, although PP can be safely performed without special equipment and devices in many ICU settings.
Author Response
Answers to Reviewer 1:
We would like to thank Reviewer 1 for their discerning comments. The answers to the comments are as follows:
More information was provided about the structure and setting of the COVID ICU in section 2.1 Study Design and Setting (”The COVID ICU functioned as a separate hospital department, but it was organized into 3 units with staff allocated to a distinctive unit and inter-unit staff changeovers were kept to minimum. Approximately 1/3 of staff were in the COVID ICU permanently during the pandemic waves and others rotated into and out of the COVID ICU usually after 3-6 months. Distinct units of COVID ICU were provisioned with similar or identical equipment to simplify equipment supply.”).
24 hour prone positioning sessions were adopted as a means to counteract staff shortage. The thinking was that the changes of body position from supination to pronation and vice versa were the most demanding in terms of the likelihood for adverse events and in terms of staff required to perform the procedure. This strategy has been used in other centres and this was mentioned in the discussion section (fifth paragraph in the discussion section).
If available specialized pillows and other equipment were used aimed at reducing the risk for pressure sores, and in case it were not available, normal hospital pillows were used. This was not noted in medical documents. Mepilex polyurethane wound dressings were used in all patients. This was mentioned in the methods section (section 2.2 Study Patients and Interventions, third paragraph).
Prone positioning was used as a survival improvement strategy, along with neuromuscular blockade and lung-protective ventilator settings. Target ventilator parameters were added to the section 2.2 Study Patients and Interventions, the end of second paragraph (“... with target tidal volumes < 8 ml/kg, plateau pressure < 30 cm H2O and driving pressure < 12 cm H2O.”). The decision to use prone positioning or not was left to the treating physician, but it was perceived as strategy to be used in case routine mechanical ventilation did not suffice. Similarly, ample evidence exists that neuromuscular blockade improves survival in patients with ARDS. An additional reason to use neuromuscular blockade was as a safety measure to prevent patients from removing catheters, tubes or drains from self-removal. Because neuromuscular blockade is at least uncomfortable patients were deeply sedated before initiation of neuromuscular blockade. This was added to the discussion section and appropriate references (references 30 and 31) were added (last sentences in the fourth paragraph in the discussion section: “Other factors, such use of neuromuscular blocking agents and subsequent need for deep sedation might have also contributed to pressure ulcers in our setting. The decision to use PP was as per the treating physician, but it was perceived as strategy to be used in higher-severity of illness patients. The rationale for neuromuscular blockade was twofold – as a lung protective strategy along with lung-protective ventilation parameters [30] and as a safety strategy to prevent patient self-removal of catheters, tubes and drains [31].”).
Because of the retrospective nature of the study, we defined any removal during PP or within 24 hours after supination as unplanned, unless explicitly otherwise stated (e.g. removal of pleural drain after evacuation of pleural effusion). If CVC line change occurred during the time period above and it were not explicitly described as caused by another reason, it would have been ascribed to PP.
Kolmogorov-Smirnov test was chosen as the most appropriate test for our population. We sincerely hope the test suffices to analyze our population.
Patients who were infected with COVID-19, mechanically ventilated and pronated were included to our study, and these patients were descibed in Table 1. Thi swas clarified in the table title.
A total of 228 PP sessions were performed in our patient cohort. This was added to section 3.1 Basic demographic Data.
Reintubation in Table 2 was corrected to “reintubation of repositioning of tracheostomy tube”.
Frequencies sum up to 100% of patients. Most patients had multiple pressure ulcers and the most common combination were facial and thoracic pressure ulcers. No changes were made to Table 3.
Table 4 includes all patients who survived to that point – not all patients reported in Table 4 survived to the fifth session of PP. Because only a very small number of patients survived to the fifth session (only 3) any other analysis would have been impossible.
All patients were included in Table 5. Adverse events occurring in that specific PP session were counted.
24 patients received 1 PP session, 39 patisnts 2 PP sessions, etc. The decision to continue/stop PP was as per the treating physician. The reason for stopping was usually patients improved and PP was no longer required.
Generally, the aim was to use 24h PP sessions, but subsequent PP sessions varied in duration (e.g. supination required for lung CT scan on day 3).
Figure 1 was labeled more accurately.
The aim of the study was to present our real-life data from a specific time period. Comparison with a historic control would require a separate ethic committee application and is not possible within the given time constraints.
We agree that methodological definition of removal as within 24 h after supination probably had little. We think that this was expressed in the text.
We agree with Reviewer 1 that appropriate education and staffing are required to ensure safe conduct of any medical procedure, PP included, and the aim of our manuscript is at least in part to perform a quality review of our real-life clinical setting. We tried to determine more objective parameters and parameters that are more easily quantified compared to adequacy of education and staffing levels (i.e., duration of PP, weight, etc) to outline those parameters that are associated with adverse events, and we hope that this is at least in some part generalizable.
Reviewer 2 Report
Comments and Suggestions for Authors
Dear Authors,
Your manuscript presents important findings on the adverse events associated with prone positioning in mechanically ventilated COVID-19 patients. Below are some specific comments and suggestions to help improve the quality and clarity of your paper:
- Introduction: Consider expanding the introduction to provide a more detailed background on prone positioning in mechanically ventilated COVID-19 patients and its known adverse events. Additional relevant references would enhance the context.
- Methods: The methodology section needs more detailed descriptions of the data extraction process and the criteria used for defining adverse events. Clarify specific procedures or protocols followed during the study.
- Results: Summarize key findings concisely and consider using more visual aids, such as graphs or charts, to improve clarity and reader understanding.
- Sample Size Calculation: Although you performed a machine learning simulation with a random forests algorithm, it is necessary to formally calculate a post hoc sample size. This calculation should establish not only the type I error but also the type II error, ensuring the robustness and statistical validity of your findings.
By addressing these recommendations, your manuscript will significantly improve in clarity, comprehensiveness, and overall quality.
Sincerely,
Comments on the Quality of English Language
Moderate editing is required to improve readability and correct grammatical errors. Consider having the manuscript reviewed by a professional editor or a native English speaker.
Author Response
Answers to Reviewer 2:
We would like to thank Reviewer 2 for their perceptive comments. The answers to the comments are as follows:
The introduction section was expanded to explain the basic mechanisms of action of prone positioning. A new section was added to the introduction section (Shifting the body position from a supine to prone position helps reduce pleural pressure gradients from dependent to nondependent lung regions resulting in increased functional residual capacity, improved matching of ventilation to perfusion, and, in some cases, improved lung drainage. Evidence supports the efficacy of PP in improving outcomes for ARDS patients both in pre-pandemic settings and during the COVID-19 pandemic [1,2].)
A more detailed description of data extraction and criteria using adverse events was added to section 2.2. (”Temperature and therapeutic charts were paper-based, and laboratory, radiology and microbiology reports were electronic. Removals of catheters and tubes were noted on the temperature and therapeutic charts on the day of the removal, and the reason for the removal was noted in the nursing and physicians’ notes. Data extraction was performed retrospectively from September 2022 to January 2023 by individual exploration of source data by researcher NR. Removals of catheters and tubes were defined as unplanned if a removal was noted during a PP session or within 24h after the patients was supinated, unless explicitly noted in medical documents that the removal was pre-planned (e.g., removal of a pleural drainage tube after evacuation of pleural effusion). Other adverse events were noted to the discretion of the researcher NR (e.g., brachial plexus injuries, corneal injuries etc.) and were not explicitly predefined.”).
Key results were summarized at the beginning of the results section ina a new paragraph (“In 100 mechanically ventilated adult patients with COVID-19 who required PP we observed in all 118 removals of catheters, tubes and drains, and a development of 184 pressure ulcers during a total of 228 PP sessions. Naso-gastric tube removals were most common, but removals of vascular, endotracheal and tracheal tubes were also observed. Facial and thoracic pressure ulcers were most common. Increasing weight, greater cumulative duration of PP, higher age and height, female gender and increasing number of PP sessions were associated with the development of any adverse event during PP.”).
We performed a post-hoc sample size calculation regarding the removal of catheters, tubes and drains (added to section 3.2. “A post-hoc sample size calculation was performed yielding a sample size of 78 patients with 95% confidence level and 5% margin of error.”) and regarding the development of pressure ulcers (added to section 3.3. A post-hoc sample size calculation revealed a sample size of 73 patients with 95% confidence level and 5% margin of error.)
The manuscript was reviewed by a native English speaker to improve the quality of the language.
Reviewer 3 Report
Comments and Suggestions for Authors
Hello,
GENERAL COMMENTS ABOUT THE STUDY:
- What is the aim of your study? It is defined differently in the abstract and introduction:
In the abstract (lines 14-15 "...determine the prevalence of adverse events in mechanically ventilated adults with Covid-19") and at the end of the introduction (lines 46-47"...aimed to explore the incidence of adverse events in relation with the duration of PP in mechanically ventilated patients with COVID-19").
The objective of a study is unique and should be defined in the same way throughout the document. Therefore, they should reformulate the objective according to what they want to achieve with the study.
- During the discussion chapter, they sometimes attribute the same meaning to the epidemiological variables incidence and prevalence, which should not be the case. Therefore, as mentioned above, they should reformulate the discussion bearing in mind the objective of the study, which seems to me to be" ...determine the prevalence of adverse events in mechanically ventilated adults with Covid-19"), taking into account the type of study carried out.
- The conclusion of the study adds very little to what was said in the discussion, so I would propose that you add some corrective measures or suggestions for improvement that, in your opinion, could improve the professional practice of doctors and nurses in ICUs, in order to reduce the prevalence of adverse events in mechanically ventilated patients submitted to prone positioning.
SPECIFIC OBSERVATIONS ON THE STUDY:
Lines 276-279: "...in the swimmer's position, the use of cushions and the frequency of body position changes may have led to a higher prevalence of pressure ulcers in your study. The position of the head, arms and legs was changed once per shift (8-8am), whereas in other studies the position of the head, arms and legs was generally changed every two hours."
Since the variables swimmer's position, use of a pillow and frequency of changing body position influence the prevalence of ulcers, it was important for us to know why, in your study, the patient's position was only changed once per shift, unlike other studies.
GOOD WORK!
Author Response
Answers to Reviewer 3:
We would kindly like to thank Reviewer 3 for their insightful comments and words of encouragement!
Answers to general comments:
We have unified the aim of the study in abstract and introduction sections. The same terminology was used as in the abstract (”To determine the prevalence of adverse events in mechanically ventilated adults with Covid-19 who have undergone prone positioning”).
In the discussion section the term prevalence was used when necessary.
One sentence was added to the conclusion section in order to highlight adverse events associated with prone positioning (”PP should be used with adverse events in mind and appropriate ICU staff numbers and education are needed to ensure appropriate quality and safety of PP”).
Answers to specific observations:
The patients' position was changed once per shift because of the lack of healthcare workers during the pandemic waves. This was explained in section 2.2. Study Patients and Interventions, penultimate sentence in the second paragraph.
Round 2
Reviewer 2 Report
Comments and Suggestions for Authors
NA
Author Response
I would kindly like to thank Reviewer 2 for their report.