The Utility of the Shock Index for Predicting Survival, Function and Health Status Outcomes in Major Trauma Patients: A Registry-Based Cohort Study

Round 1

Reviewer 1 Report

Congratulations on this importnat and well written piece. 

This citation: Olaussen A, Blackburn T, Mitra B, Fitzgerald M. shock index for prediction of critical bleeding post‐trauma: a systematic review. Emergency Medicine Australasia. 2014 Jun;26(3):223-8. could be included in the introduction where haemorhagic shock is discussed. 

Although you note in the limitations that removal of cases with missing HR or SBP is a problem, consider doing a sensitivity analysis on cases with a HR, but absent SBP.  

Author Response

A point-by-point response to reviewer 1

• Congratulations on this important and well written piece. 

Response: Thank you for taking time to read and comment on this manuscript. We have addressed each of the comments below.

• This citation: Olaussen A, Blackburn T, Mitra B, Fitzgerald M. shock index for prediction of critical bleeding post‐trauma: a systematic review. Emergency Medicine Australasia. 2014 Jun;26(3):223-8. could be included in the introduction where haemorhagic shock is discussed. 

Response: Thank you for finding and suggesting this source, we have included this in the introduction as source number 11 of the revised manuscript. See below:

Line 46: “A score that could help to predict haemorrhagic shock and risk of a poorer outcome is the shock index (SI) (11).”

• Although you note in the limitations that removal of cases with missing HR or SBP is a problem, consider doing a sensitivity analysis on cases with a HR, but absent SBP.  

Response: Thank you for pointing this out, it is a highly relevant aspect. The number of cases which had a valid HR recorded but a missing SBP was 49. This is very low compared to the total number of analysed cases (29,574) and the number of cases that were included when missing vital signs were substituted with the definitive hospital observations rather than the observations of the referral hospital (1,302). As the number of cases with a valid HR but missing SBP was extremely low, a sensitively analysis is not indicated as these 49 cases are unlikely to influence the findings.

Reviewer 2 Report

Dear Authors, 

I’ve read your manuscript and founded your research very interesting. 

This simple traumatic index could be very useful in clinical practice, even considering its intrinsic limits in predicting longer term outcomes including function and health status. 

In my opinion your papers shows a very high scientific level, because is based on a solid research, but also because give several simple and useful clinical messages in your final discussion.

Moreover I agree with you and totally subscribe when you write that “....Mortality is a limited measure of outcome in a mature trauma system since most patients survive their injuries and often with ongoing morbidity”. In effect I’ve appreciated your efforts to include data about residual quality of life and long term outcomes.

 Before editing i have only few requests in following item 

• Line 71: No description about Prehospital Trauma Care Organisation in Victoria State is present. I wonder if you could briefly describe, especially because also in richest countries there are several differences. This aspect could be crucial because your research lasts a ten years period, and several extra hospital clinical management algorithms are improved in this time frame. 
• Line 107: you indicate the measure of 1 m for separating the “low” falls from the “high” ones. Usually this “cut off” is higher, as 2 m (Emergency Medicine Journal. 2015;32(12):911-5), or at least the same patient’s height. Could you better explain your choice’s reason.
• Line 308: I totally agree when you admit a potential bias in your research of SI underestimation if missing vital signs in the Primary Hospital. However in fig 3 the calibration curves of the model shows a good correlation between predicted and observed mortality supporting your assumption.

Best regards

Author Response

Re: ”The Utility of the Shock Index for Predicting Survival, Function and Health Status Outcomes in Major Trauma Patients: A Registry-Based Cohort Study”

A point-by-point response to reviewer 2

• I’ve read your manuscript and founded your research very interesting. 

This simple traumatic index could be very useful in clinical practice, even considering its intrinsic limits in predicting longer term outcomes including function and health status. 

In my opinion your papers shows a very high scientific level, because is based on a solid research, but also because give several simple and useful clinical messages in your final discussion.

Moreover I agree with you and totally subscribe when you write that “....Mortality is a limited measure of outcome in a mature trauma system since most patients survive their injuries and often with ongoing morbidity”. In effect I’ve appreciated your efforts to include data about residual quality of life and long term outcomes.

Before editing I have only few requests in following item 

Response: Thank you for taking time to read and comment on this manuscript. We have addressed each of the comments below.

• Line 71: No description about Prehospital Trauma Care Organisation in Victoria State is present. I wonder if you could briefly describe, especially because also in richest countries there are several differences. This aspect could be crucial because your research lasts a ten years period, and several extra hospital clinical management algorithms are improved in this time frame.

Response: We agree with the reviewer and have added information of the Prehospital Trauma Care Organisation in the method.

Line 71: “The state has a single ambulance service, staffed by paramedics with Advanced Trauma Life Support training, which provides road and air transport of cases. There are three designated major trauma services (MTS, Level 1 trauma centre equivalent) in Victoria; two adult and one paediatric.”

• Line 107: you indicate the measure of 1 m for separating the “low” falls from the “high” ones. Usually this “cut off” is higher, as 2 m (Emergency Medicine Journal. 2015;32(12):911-5), or at least the same patient’s height. Could you better explain your choice’s reason.

Response: The cut-off for what is considered a “high” fall does differ from jurisdiction to jurisdiction. The VSTR uses the definition of the Victorian Emergency Minimum Dataset which is the State’s standardised data collection for injury prevention purposes in Victoria.

• Line 308: I totally agree when you admit a potential bias in your research of SI underestimation if missing vital signs in the Primary Hospital. However in fig 3 the calibration curves of the model shows a good correlation between predicted and observed mortality supporting your assumption.

Response: You have a point, thank you for addressing this.  

Round 2

Reviewer 2 Report

Dear Authors, 

I’ve read your revised version of the manuscript, and I’ve appreciated your answers to all my requests and comments. 

Your research is interesting, based on a very high number case, detailed and well explained. As mentioned in my previous report, your choice to insert data about residual quality of life and long term outcomes, could give a special value to your results.

Your discussion and your conclusions are clear, simple and could give useful messages for the readers. Moreover you define and underline all the intrinsic limits related to the Shock Index and to your research. 

In my opinion this second version improved its scientific value, and could be accepted for editing in present form.

Best Regards