Dr. Yonekura notes that we failed to exclude patients with potential nonischemic causes of troponin elevation, thus possibly including some patients who did not actually have myocardial injury after noncardiac surgery.1 It is likely that most patients who have troponin elevations and conditions that might falsely elevate troponin concentrations actually do have myocardial injury. Nonetheless, nonischemic causes should have been excluded from our analyses.2
Had the purpose of our study been to determine the incidence of myocardial injury, it would be critical to include only patients with scheduled (not-for-cause) troponin screening. But that was not our purpose at all. Detection and ascertainment bias are therefore irrelevant to our analysis. Instead, we asked which blood pressure components best predicted myocardial injury. Because all four components were evaluated in each patient, there was no bias.
Per request, we present a sensitivity analysis that excludes patients with medical conditions that might explain nonischemic troponin elevations (Supplemental Digital Content, table 1, http://links.lww.com/ALN/C427) and is restricted to patients who had scheduled (not for cause) troponin measurements. The analysis included 4,886 patients. The overall observed incidence of myocardial injury after noncardiac surgery was 17%.
The statistically determined thresholds and visual cutoff points did not change by clinically important amounts, although the systolic threshold decreased from 87 to 79 mmHg and pulse pressure threshold decreased from 35 to 30 mmHg (table 1, fig. 1, and Supplemental Digital Content, table 2, http://links.lww.com/ALN/C428). Univariable and multivariable relationships between blood pressure components and myocardial injury were essentially unchanged except for pulse pressure (fig. 1). Interestingly, pulse pressure demonstrated a relatively flat curve between 25 and 45 mmHg—implying a weak relationship between pulse pressure and myocardial injury. The odds of patients experiencing a composite of serious complications were significantly higher for the fourth quartile of area under curve under the threshold for systolic, mean, and pulse pressure compared to patients who never went below the threshold, but not for diastolic pressure (fig. 2).
Power was limited when analysis was restricted to patients with scheduled troponin surveillance. But among the four blood pressure components, systolic and mean pressures continue to be most predictive and were comparable in their strength of association with myocardial injury. The relationship with diastolic pressure remained poor, but pulse pressure was as well, which differs from our original analysis. We therefore conclude that systolic pressure at a threshold of 80 to 90 mmHg and mean pressure at a threshold near 65 mmHg are the blood pressure components most associated with myocardial injury in patients who have major noncardiac surgery. Similar thresholds for renal injury were identified in our original analysis.1
Drs. Sessler and Maheshwari are consultants to Edwards Lifesciences (Irvine, California). The other authors declare no competing interests.