Pupillary Pain Index Predicts Postoperative Pain but Not the Effect of Peripheral Regional Anaesthesia in Patients Undergoing Total Hip or Total Knee Arthroplasty: An Observational Study

Round 1

Reviewer 1 Report

I reviewed your manuscript titled, “Pupillary Pain Index Predicts Postoperative Pain, but not the Effect of Peripheral Regional Anesthesia in Patients Undergoing Total Hip or Total Knee Arthroplasty: An Observational Study”

The authors suggested that “While the effect of FIB and ACB could not be shown with PPI, postoperative pain scores after 48 h were related to target area PPI measurements before block insertion. These results suggest that preoperative PPI may be used to predict postoperative pain.”.

It was interesting to see a study on whether or not the pupillary pain index could be used to evaluate the effectiveness of regional anesthesia provided to patients undergoing orthopedic surgery.

However, there are several points that should be revised and described in more detail.

1.      I think all research should have been conducted according to the protocol registered in the clinical trial. The purpose of the study registered in the clinical trial was to investigate the effectiveness of the Nociception Level Index, ANI, and PPI in patients undergoing orthopedic surgery. However, the authors analyzed the research data on PPI among them and wrote and submitted a manuscript. In addition, the primary outcome registered in the clinical trial was to evaluate the relationship with postoperative pain, and the secondary outcome was described as cumulative postoperative opioid consumption. This is not the same information as the primary outcome of this manuscript. I have decided that this is an ethical issue.

2.      The study was registered in 2019, and the total number of study subjects was 44. Considering the number of study subjects, the study is believed to have ended in early 2020, but the report of the study results in 2023 seems to be less reliable. Therefore, I believe that the submission of raw data for this study is necessary.

3.      The authors need to describe the date of clinical trial registration and IRB approval in this manuscript.

4.      As the authors mentioned, THR and TKR can occur very differently between intraoperative and postoperative pain, and the authors provided different peripheral regional anesthesia depending on the type of surgery. I think this is a very important factor in interpreting the analysis results in the research results. Therefore, it is reasonable to analyze the effect of PPI on peripheral regional anesthesia by classifying it into THR and TKR groups.

5.      It is reasonable to present an equivalent conversion dosage of the analgesics used in the PCA provided for postoperative pain control as intravenous morphine, which should be used for analysis. In addition, the equivalent conversion formula should be presented.

6.      Information on sample size calculation for this study is ambiguous.  It is not clear which information in the pilot study you have used. In the reference, a peripheral regional anesthesia was not performed.

7.      The number of subjects registered in the clinical trial and the number of subjects registered in this study were 44, but the number of subjects presented in the sample size calculation is 29, which is inconsistent.

8.      The tables presented by the researchers are described as somewhat difficult to understand, and I recommend organizing the same characteristics. In table 1, is remifentanil used during surgery? Did you describe the usage capacity of Piritramide after surgery? In that case, the information about the group using tramadol seems to be missing.

9.      Before presenting the Prediction models of postoperative pain scores, it is necessary to show changes in PPI scores and statistically significant results before and after the provision of peripheral regional anesthesia. It is important to show this first.

10.   Table 2 shows variables with p<0.2 in relation to postoperative pain by performing regression analysis among each independent variable. In this case, you should have presented all the variables you used for the analysis. In addition, it should be described to know which variables were used to perform the final multiple regression analysis. In addition, the results of the final analysis show only the regression analysis results for 48 hours postoperative NRS. You should also see the results of the 24-hour NRS after the surgery. It is believed that the authors selectively reported the results.

 

11.   The description of the discussion is distracting and difficult to understand, so it is required to classify the content and describe it in a different paragraph. The description of the dispute does not fully support the findings of the study and does not provide sufficient evidence to support the authors' opinions.

Author Response

We thank the reviewer for the thorough work. All the points could be addressed and the manuscript and its content could be improved significantly with the help of the reviewer's constructive comments.  Please find a point-by point response below.

 

1. I think all research should have been conducted according to the protocol registered in the clinical trial. The purpose of the study registered in the clinical trial was to investigate the effectiveness of the Nociception Level Index, ANI, and PPI in patients undergoing orthopedic surgery. However, the authors analyzed the research data on PPI among them and wrote and submitted a manuscript. In addition, the primary outcome registered in the clinical trial was to evaluate the relationship with postoperative pain, and the secondary outcome was described as cumulative postoperative opioid consumption. This is not the same information as the primary outcome of this manuscript. I have decided that this is an ethical issue.

The reviewer raises an important point here. PPI was included in the NOL study at a later point (via amendment) since the same patient collective should be used. Due to the completely different nature of the methods, this was never meant to be one overall study. Therefore, this manuscript is rather a substudy with separate data analysis. Nonetheless, the same registration was used due to the reasons mentioned above. Registration is not mandatory in observational trials, it would be unfortunate if providing transparency though registration actually is a disadvantage.

Regarding primary and secondary outcome, we agree with the reviewer and apologize for the discrepancies. We modified these passages and everything should be in line with the information given in the registration.

 

2. The study was registered in 2019, and the total number of study subjects was 44. Considering the number of study subjects, the study is believed to have ended in early 2020, but the report of the study results in 2023 seems to be less reliable. Therefore, I believe that the submission of raw data for this study is necessary.

Raw data is submitted as a supplementary table.

 

3. The authors need to describe the date of clinical trial registration and IRB approval in this manuscript.

Both details have been added to the methods section of the manuscript.

 

4. As the authors mentioned, THR and TKR can occur very differently between intraoperative and postoperative pain, and the authors provided different peripheral regional anesthesia depending on the type of surgery. I think this is a very important factor in interpreting the analysis results in the research results. Therefore, it is reasonable to analyze the effect of PPI on peripheral regional anesthesia by classifying it into THR and TKR groups.

We thank the reviewer for this important note. We compared THR and TKR groups (Supplement 1). We found a difference in opioid dosage and pump types (which is also mentioned in the results section of the text). We added surgery type and PCA type to the regression model. Moreover, we recalculated our regression model for THR and TKR surgeries separately and presented the results in Supplement 1 and Figure 3. These results are also reported in the text.

 

5. It is reasonable to present an equivalent conversion dosage of the analgesics used in the PCA provided for postoperative pain control as intravenous morphine, which should be used for analysis. In addition, the equivalent conversion formula should be presented.

This has been done with the explanation in table 1. Please find the conversion formula also in the methods section.

 

6. Information on sample size calculation for this study is ambiguous. It is not clear which information in the pilot study you have used. In the reference, a peripheral regional anesthesia was not performed.

We agree with the reviewer. In the mentioned study, PPI after induction of anesthesia (with sevoflurane and propofol) before opioid given (alfentanil) and gave a value of 6, after opioid (alfentanil) – 2. We purposed that after general anesthesia induction, PPI would also be 6. We based our sample-size calculation on these measurements, changes in PPI were, naturally, estimates.

 

7. The number of subjects registered in the clinical trial and the number of subjects registered in this study were 44, but the number of subjects presented in the sample size calculation is 29, which is inconsistent.

We agree with the reviewer and are thankful for highlighting this point. From our experience we expected a large drop-out rate of 30% from the beginning due to technical problems, missing data etc. Therefore, we aimed at n=40 as mentioned in the registration. To compensate for the drop-outs already known, the number was increased to 44. We mentioned this in the methods section of the text.

 

8. 8. The tables presented by the researchers are described as somewhat difficult to understand, and I recommend organizing the same characteristics. In table 1, is remifentanil used during surgery? Did you describe the usage capacity of Piritramide after surgery? In that case, the information about the group using tramadol seems to be missing.

We thank the reviewer for the feedback. The table has been changed and the explanation about intraoperative and postoperative opioids has been modified.

 

9. Before presenting the Prediction models of postoperative pain scores, it is necessary to show changes in PPI scores and statistically significant results before and after the provision of peripheral regional anesthesia. It is important to show this first.

We thank the rewiever for highlighting this. The order of presentation has been changed.

 

10. Table 2 shows variables with p<0.2 in relation to postoperative pain by performing regression analysis among each independent variable. In this case, you should have presented all the variables you used for the analysis. In addition, it should be described to know which variables were used to perform the final multiple regression analysis. In addition, the results of the final analysis show only the regression analysis results for 48 hours postoperative NRS. You should also see the results of the 24-hour NRS after the surgery. It is believed that the authors selectively reported the results.

We apologize that the information has remained unclear. We added a simple regression analysis for each independent variable (Supplement 1, Table 2S). For multiple regression, the choice of the model was based on the best adjusted r squared value. We also added to the model surgery type and PCA type and re-calculated the model. Now all types of models, for NRS 24h and 48h are shown in figure 3.

 

11. The description of the discussion is distracting and difficult to understand, so it is required to classify the content and describe it in a different paragraph. The description of the dispute does not fully support the findings of the study and does not provide sufficient evidence to support the authors' opinions.

This has been done according to the reviewer’s suggestion. The statements in the discussion have been modified (and extended due to the numerous previous modifications) and should appear less bold. The discussion has been divided into subsections. We hope that this is in line with the reviewer’s expectations.

With the corrections mentioned above we hope that we could sufficiently address all of the points and would like to thank the reviewer again for the thorough work.

 

Reviewer 2 Report

Dear Authors,

the literature provides interesting data on the use of the pupillometry method for the objective assessment of perioperative pain. However, the objective measurement of perioperative nociception in orthopedic patients is still a challenge for scientists and physicians and has not been well described in the literature so far. In addition, only a few studies have focused on evaluating pain in orthopedic patients as a result of iliac fascia block (FIB) or adductor canal block (ACB). The manuscript may therefore be an interesting contribution to this field and could be a pilot for future research.

However, in my opinion, the quality of this article could be improved with the suggestions below.

 

INTRODUCTION:

This chapter lacks clearly formulated assumptions/hypotheses and the exact purpose(s) of the research conducted.

 

METHODS:

You should explain the reason (if it is known) for the use of FIB in selected TKR patients (referring to lines: 76, 77).

The authors in line 99 wrote that postoperative pain assessments at 24 and 48 hours were performed "at rest and in motion" using the NRS. What exactly was the form of movement during which "NRS in motion" was assessed? Was it the same test for all subjects? Or maybe a different one for patients after TKR/THR?

 

RESULTS:

Figure 1 should be converted into a flowchart diagram (current view: text placed in different places on the page).

The obtained correlations were not accurately and clearly presented (description should be corrected, lines: 189-194). Also: figure 3. - please complete the exact description of the X-axis in the correlation diagram.

 

CONCLUSIONS:

The obtained results do not entitle the authors to draw the following conclusion:

“However, PPI can be used to predict longterm postoperative pain during mobilization” (lines: 268-269). I suggest removing the term “longterm” from this sentence or even replace it with “short-term”.

 

DISCUSSION:

There are, in addition to those mentioned, some limitations of the study that should be emphasized.

It is worth taking into account in the limitations, e.g. a large difference in the number of study participants between the group of patients who received fascia iliaca block (n=27) vs adductor canal block (n=8).

There is no separate analysis of groups of patients with THR vs TKR, respectively, in terms of peripheral anesthesia and postoperative analgesics used. The studies also used a short, post-operative follow-up period, which may also be a significant limitation.

 

Therefore, the presented studies seem to be rather preliminary studies, both in terms of determining the effectiveness of peripheral anesthesia during joint replacement surgery, as well as in relation to the usefulness of using the PPI method for the assessment and prediction of perioperative pain in patients after THR/TKR. Perhaps in such a situation it is worth highlighting in the text of the manuscript the preliminary nature of the conducted research.

Author Response

We thank the reviewer for the thorough work and the constructive comments. All the reviewer’s points could be addressed and our manuscript could be improved thus. Language editing was performed by a native speaker.

Please find a point-by point response to the comments below.

 

INTRODUCTION:

This chapter lacks clearly formulated assumptions/hypotheses and the exact purpose(s) of the research conducted.

We agree and apologize, the chapter has been modified, assumptions as well as hypotheses have been added.

 

METHODS:

You should explain the reason (if it is known) for the use of FIB in selected TKR patients (referring to lines: 76, 77).

We agree that this may appear strange. Although it not part of the standard, it occurs that FIB is applied as a rescue option if the more distal ACB cannot be used due to surgical concerns. Therefore, we added “…to avoid interference with the surgical field.”

The authors in line 99 wrote that postoperative pain assessments at 24 and 48 hours were performed "at rest and in motion" using the NRS. What exactly was the form of movement during which "NRS in motion" was assessed? Was it the same test for all subjects? Or maybe a different one for patients after TKR/THR?

We thank the reviewer for this comment. This point had remained unclear. Patients were interviewed on the ward and were asked to rate their pain experiences. “Movement” was mobilization with physiotherapy which should have been equal for both groups. The passage was modified and now reads: “In the orthopedic ward, patients were interviewed about their pain experiences using the Numeric Rating Scale (NRS) at 24- and 48-hours at rest and in movement during mobilization.”

 

RESULTS:

Figure 1 should be converted into a flowchart diagram (current view: text placed in different places on the page).

Figure 1 has been modified. We hope that this is in line with the reviewer’s expectations.

The obtained correlations were not accurately and clearly presented (description should be corrected, lines: 189-194). Also: figure 3. - please complete the exact description of the X-axis in the correlation diagram

The corrections have been made. Figure 3 has been changed completely, descriptions of the X axes have been clarified.

 

CONCLUSIONS:

The obtained results do not entitle the authors to draw the following conclusion:

“However, PPI can be used to predict longterm postoperative pain during mobilization” (lines: 268-269). I suggest removing the term “longterm” from this sentence or even replace it with “short-term”.

The term “long-term” has been removed.

 

DISCUSSION:

There are, in addition to those mentioned, some limitations of the study that should be emphasized.

It is worth taking into account in the limitations, e.g. a large difference in the number of study participants between the group of patients who received fascia iliaca block (n=27) vs adductor canal block (n=8).

There is no separate analysis of groups of patients with THR vs TKR, respectively, in terms of peripheral anesthesia and postoperative analgesics used. The studies also used a short, post-operative follow-up period, which may also be a significant limitation.

Limitations have been added according to the reviewer’s suggestions. Separate analyses have also been added and can be found in supplement 1.

 

Therefore, the presented studies seem to be rather preliminary studies, both in terms of determining the effectiveness of peripheral anesthesia during joint replacement surgery, as well as in relation to the usefulness of using the PPI method for the assessment and prediction of perioperative pain in patients after THR/TKR. Perhaps in such a situation it is worth highlighting in the text of the manuscript the preliminary nature of the conducted research.

We agree with the reviewer. A passage has been added to the limitations section stating “…results are of preliminary nature and should be interpreted in this context.”

With the above mentioned corrections we hope to have sufficiently addressed all of the points and would like to thank again for the constructive comments.

 

Round 2

Reviewer 1 Report

Thank you for your effort to revise your manuscript according to my comments.

This manuscript reflected my comments well, and the author's reply was also written with sufficient persuasion.