Moein Tavakkolizadeh, William Harvey Hospital, Kent, UK
11 May 2006
With many thanks to authors of the above article for highlighting the clinical implication of different methods of monitoring arterial blood pressure, we had concerns about waiving informed consent on the basis of “no additional procedures apart from usual intensive care practice needed” as inserting both radial and femoral artery cannulas is not part of standard intensive care practice. Since arterial cannulation is not without risks, complication rate is higher when the patient is cannulated at two sites. Exposing the patient to higher risk does require informed consent.
In that study data collection was based on repeated measurements of arterial blood pressure of each individual, so the gathered data are subject to paired sample t-test not an unpaired one. Using unpaired sample t test is less sensitive in detecting significant differences and makes the analysis less powerful thus questioning the conclusion of above study.
• Armitage P, Berry G. Statistical Methods in Medical Research. 3rd ed. Oxford: Blackwell Scientific Publications, 1994:207-14
• WORLD MEDICAL ASSOCIATION DECLARATION OF HELSINKI
Ethical Principles for Medical Research Involving Human Subjects
Adopted by the 18th WMA General Assembly, Helsinki, Finland, June 1964
• Steven G. Wallach, Cannulation injury of the radial artery: diagnosis and treatment algorithm American Journal of Critical Care, July 2004
• Bedford RF, Wollman H. Complications of percutaneous radial artery cannulation: an objective prospective study in man. Anesthesiology 1973;38(3):228-236
• Frezza EE, Mezghebe H. Indications and complications of arterial catheter use in surgical or medical intensive care units: analysis of 4932 patients. Am Surg 1998;64(2):127-131
Moein. Tavakkolizadeh, MD
Hooman. Parsi, MD
Anaesthetic Department, William Harvey Hospital,Kent TN24 0LZ, UK
Competing interests
None declared
Ethical and statistical concern
MARIANO ALEJANDRO MIGNINI, Clinica Bazterrica, Bs As ARGENTINA
19 June 2006
We thank Drs. Tavakkolizadeh and Parsi for their comments on our article [1]. About their ethical concerns we want to amplify the information. The orders to change arterial catheters were given attending physicians, not by the investigators. After the change, both catheters were left in place for a short period of time (never greater than five minutes). During this period, which is usually required to test the adequate functioning of the new catheter - a standard practice in our ICU, is when patients were studied. Thus, no additional procedures apart from usual intensive care practices were performed. For this reason, Hospital Ethics Committee waived the need for informed consent.
With regard to statistical concerns, we respectfully believe that there is a misinterpretation. We stated in “Statistical analysis” section that “Bias and precision between groups were compared using unpaired t tests.” Indeed, we compared the bias ± precision of two independent groups: those receiving high doses of vasopressors and those receiving low doses of vasopressors. We did not compare “repeated measurements of arterial blood pressure of each individual”, in which case paired t test should have been used, as Drs. Tavakkolizadeh and Parsi wrote. Unpaired t test is the test indicated for the purpose of our study.
Finally, we disagree with Drs. Tavakkolizadeh and Parsi’s statement that “Using unpaired sample t test is less sensitive in detecting significant differences and makes the analysis less powerful thus questioning the conclusion of above study.” Apart from being unpaired t test the right choice for this comparison, 3 ± 4 (high-dose group) vs. 3 ± 4 mm Hg (low-dose group) will display non-significative p-value, regardless the used test. Our conclusion that radial and femoral mean arterial blood pressures are interchangeable, independently whether vasopressors are used or not, is based on the finding of equal bias and precision in both groups. Therefore, we are convinced that our conclusion is based on robust results and analysis.
1.-Peripheral arterial blood pressure monitoring adequately tracks central arterial blood pressure in critically ill patients: an observational study. Mariano Alejandro Mignini, Enrique Piacentini and Arnaldo Dubin. http://ccforum.com/content/10/2/R43
Critical Care
Ethical and statistical concern
11 May 2006
With many thanks to authors of the above article for highlighting the clinical implication of different methods of monitoring arterial blood pressure, we had concerns about waiving informed consent on the basis of “no additional procedures apart from usual intensive care practice needed” as inserting both radial and femoral artery cannulas is not part of standard intensive care practice. Since arterial cannulation is not without risks, complication rate is higher when the patient is cannulated at two sites. Exposing the patient to higher risk does require informed consent.
In that study data collection was based on repeated measurements of arterial blood pressure of each individual, so the gathered data are subject to paired sample t-test not an unpaired one. Using unpaired sample t test is less sensitive in detecting significant differences and makes the analysis less powerful thus questioning the conclusion of above study.
• Armitage P, Berry G. Statistical Methods in Medical Research. 3rd ed. Oxford: Blackwell Scientific Publications, 1994:207-14
• WORLD MEDICAL ASSOCIATION DECLARATION OF HELSINKI
Ethical Principles for Medical Research Involving Human Subjects
Adopted by the 18th WMA General Assembly, Helsinki, Finland, June 1964
• Steven G. Wallach, Cannulation injury of the radial artery: diagnosis and treatment algorithm American Journal of Critical Care, July 2004
• Bedford RF, Wollman H. Complications of percutaneous radial artery cannulation: an objective prospective study in man. Anesthesiology 1973;38(3):228-236
• Frezza EE, Mezghebe H. Indications and complications of arterial catheter use in surgical or medical intensive care units: analysis of 4932 patients. Am Surg 1998;64(2):127-131
Moein. Tavakkolizadeh, MD
Hooman. Parsi, MD
Anaesthetic Department, William Harvey Hospital,Kent TN24 0LZ, UK
Competing interests
None declared
Ethical and statistical concern
19 June 2006
We thank Drs. Tavakkolizadeh and Parsi for their comments on our article [1]. About their ethical concerns we want to amplify the information. The orders to change arterial catheters were given attending physicians, not by the investigators. After the change, both catheters were left in place for a short period of time (never greater than five minutes). During this period, which is usually required to test the adequate functioning of the new catheter - a standard practice in our ICU, is when patients were studied. Thus, no additional procedures apart from usual intensive care practices were performed. For this reason, Hospital Ethics Committee waived the need for informed consent.
With regard to statistical concerns, we respectfully believe that there is a misinterpretation. We stated in “Statistical analysis” section that “Bias and precision between groups were compared using unpaired t tests.” Indeed, we compared the bias ± precision of two independent groups: those receiving high doses of vasopressors and those receiving low doses of vasopressors. We did not compare “repeated measurements of arterial blood pressure of each individual”, in which case paired t test should have been used, as Drs. Tavakkolizadeh and Parsi wrote. Unpaired t test is the test indicated for the purpose of our study.
Finally, we disagree with Drs. Tavakkolizadeh and Parsi’s statement that “Using unpaired sample t test is less sensitive in detecting significant differences and makes the analysis less powerful thus questioning the conclusion of above study.” Apart from being unpaired t test the right choice for this comparison, 3 ± 4 (high-dose group) vs. 3 ± 4 mm Hg (low-dose group) will display non-significative p-value, regardless the used test. Our conclusion that radial and femoral mean arterial blood pressures are interchangeable, independently whether vasopressors are used or not, is based on the finding of equal bias and precision in both groups. Therefore, we are convinced that our conclusion is based on robust results and analysis.
1.-Peripheral arterial blood pressure monitoring adequately tracks central arterial blood pressure in critically ill patients: an observational study. Mariano Alejandro Mignini, Enrique Piacentini and Arnaldo Dubin. http://ccforum.com/content/10/2/R43
Competing interests
None declared