To the Editor:
We read with great interest the article by Ahuja et al.1 on the association between various intraoperative blood pressure components and postoperative morbidity. They identified an association among the arterial systolic, mean, and pulse pressure hypotension with myocardial and renal injury. Although the main finding—a lower blood pressure can be associated with postoperative myocardial injury after noncardiac surgery and acute kidney injury—offers clinically valuable information, we believe that some inherent bias in the study design should be discussed and clarified.
First, compared with previously published studies, this study had an important difference in the definition of myocardial injury after noncardiac surgery.2 The outcome definition of myocardial injury in this study (i.e., elevation of troponin or creatinine kinase-myocardial bound during the first 7 postoperative days)1 was different from that approved by the consensus diagnostic criteria in 2014, which defined myocardial injury after noncardiac surgery as “elevated post-operative troponin measurement judged as resulting from myocardial ischemia during or within 30 days after non-cardiac surgery.”3,4 In addition, this study did not exclude nonischemic etiologies (sepsis, arrythmias, pulmonary embolism, etc.). A previous study showed that elevation of troponin levels in 11 to 14% cases after noncardiac surgery was due to nonischemic etiologies.3 Therefore, without an adequate outcome assessor, the results of Ahuja et al.1 tend to overestimate the actual incidence of myocardial injury by including nonischemic etiologies. Although this exclusion was not possible because of the retrospective study design using electronic medical records, the authors should discuss this aspect in the study limitations.
Second, residual confounding might have been present because compared with routine postoperative biomarker screening, the postoperative measurements of cardiac biomarkers were likely to be influenced by clinical indication (“confounding by indication”).5 As stated by the authors, the postoperative troponin concentration was measured in only 25% of the samples, and the authors assumed that myocardial injury was absent in patients without troponin surveillance.1 This assumption can induce serious outcome detection bias. In clinical practice, postoperative cardiac biomarkers (troponin or creatinine kinase–myocardial bound) are not measured routinely. Results of previous studies indicate that majority of the patients with myocardial injury after noncardiac surgery (approximately 90%) do not have clinical cardiac symptoms and are not surveilled for troponin measurement.6,7 Therefore, measurement of postoperative cardiac biomarkers was mostly restricted to only high-risk patients or those with clinical signs of myocardial ischemia.8 In addition, differential surveillance for outcome assessment in patients who experienced or did not experience intraoperative hypotension can induce surveillance bias.9,10 If possible, Ahuja et al. should perform sensitivity analyses in patients who undergo troponin surveillance (5,699 patients) to test the robustness of their results. This information will be valuable to the readers of Anesthesiology.
The author declares no competing interests.