Preoperative β-blockade with local or regional anesthesia for vascular surgery is associated with lower maximal heart rate intraoperatively compared to local or regional anesthesia alone. These results support the findings of a previous study that demonstrated the effectiveness of β-blockade on heart rate control during the induction period for vascular surgery patients undergoing general anesthesia (Mudumbai et al. 2012). Unlike the previous study, all patients in the present study had average maximal heart rates below 100 bpm regardless of preoperative β-blocker status. Since heart rates over 100 bpm are associated with an increase in myocardial infarction, our study results suggest a potential advantage in avoiding general anesthesia when appropriate for the surgical procedure and that preoperative β-blockade in the perioperative management of high-risk vascular surgery patients may offer additional effects in terms of heart rate control (Beattie et al. 2008).
Although this topic continues to be debated, published recommendations based on randomized controlled trials suggest that vascular surgery patients represent a unique cohort in which the advantages of perioperative β-blockade may outweigh risks (Fleisher et al. 2009; López-Sendó et al. 2004). However, patterns of intraoperative heart rate and medication-induced control during stressful periods and any association with cardiac morbidity in high-risk patients are collectively an active area of research (Fleisher and Poldermans 2008). The present study is the first to examine the effectiveness of preoperative β-blockade during the intraoperative period in a cohort of actual vascular surgery patients undergoing regional or monitored anesthesia care under situations of routine clinical practice. Although the 10-bpm decrease in maximal heart rate in favor of the β-blocker group did not directly translate into overall morbidity and mortality benefits, published studies have shown that elevated heart rates in 10-bpm increments may be associated with increased risk of myocardial ischemia, troponin-T release, and long-term mortality (Feringa et al. 2006). We did note a higher relative risk for 30-day mortality in our β-blocker population-similar to those of the POISE trial (Group et al. 2008). Since our study was neither designed nor powered to study 30-day mortality, incident major adverse cardiac event or stroke, we do not draw any definitive conclusions about these findings.
Preoperative β-blockade and intraoperative heart rate control
The 10-bpm difference during the PSI period in favor of the β-blocker group in this study is clinically relevant given recent recommendations that heart rate should be tightly controlled within a range of 60–80 bpm with β-blockers throughout the perioperative period. Although both groups’ average maximal heart rates were over 80 bpm, the β-blocker group maintained heart rates much closer to the target range during the PSI period and entire operation compared to the non-β-blocker group.
Despite the use of regional or monitored anesthesia care and avoidance of anesthetic induction and airway manipulation, the PSI period remains a vulnerable period of physiologic stress for vascular surgery patients. One source of stress and hemodynamic instability during the PSI period may be the administration of site-specific local anesthetic or regional block (Tuman et al. 1991; Christopherson et al. 1993; Christopherson et al. 1996). Since patients are awake or receiving monitored anesthesia care, they may exhibit increases in heart rate due to anxiety, positioning, needlestick stimulation, pain from the local anesthetic itself, pain from manipulation of the affected body part, or inadvertent intravascular injection of epinephrine-containing local anesthetic solutions. In addition, a significant percentage of our patients (54% in the control and 74% in the intervention group) underwent arterio-venous (AV) fistulae and grafts. There are important clinical and physiologic differences in the PSI period between AV procedures and other vascular surgery procedures (e.g., peripheral revascularization, amputation); these differences may also contribute to observed maximum heart rates. Regardless of the actual procedure they undergo, increases in heart rate for these high-risk patients may disrupt the fragile myocardial oxygen supply/demand balance, leading to myocardial ischemia or cardiac dysrythmias. The PIRAT study involving patients undergoing lower extremity vascular surgery under epidural or general anesthesia has demonstrated an association between rapid heart rate changes (> 20 bpm in 5 minutes) and intraoperative myocardial ischemia (Christopherson et al. 1996). Following surgical incision, these patients still experience physiologic perturbations related to the surgical procedure and arguably may be more hemodynamically responsive since they are not under general anesthesia. The results of the present study suggest β-blocker-associated heart rate control throughout the intraoperative period beyond the PSI interval. Future effectiveness studies are necessary to help refine existing protocols for appropriate medication selection and target heart rates.
Patterns of intraoperative medication use
Prior to surgical incision, non-β-blocked patients receive higher doses of opioids and sedatives and lower doses of β-blockers than preoperatively β-blocked patients. We speculate that, during the PSI period, anesthesiologists are likely to interpret elevated heart rate for non-β-blocked patients as pain or anxiety. After surgical incision, intraoperative doses of β-blockers are the same between groups, suggesting that anesthesiologists are more likely to interpret any increase in heart rate as not pain-related when the surgery is performed under local or regional anesthesia and feel administering β-blockers to control heart rate. These hypotheses related to medication administration patterns by providers and perioperative heart rate control must be more rigorously studied (Freundlich and Kheterpal 2011; Kheterpal 2009; Mukherjee and Eagle 2003).
Since this is a retrospective cohort study, we can only identify associations and not draw definitive conclusions regarding causality between preoperative β-blockade and heart rate control. However, we designed this study as an effectiveness study and took several steps to minimize bias (Freundlich and Kheterpal 2011; Iglehart 2009). By focusing on the PSI period, we attempted to measure the effects of preoperative β-blockade without the confounders of surgical stimuli and hemodynamic shifts from blood loss that can also affect heart rate. We attempted to control for the effects of other medications in our main regression model and sensitivity analysis. We acknowledge that anesthesiologists may have altered their perioperative management for known β-blocker patients, and we attempted to minimize this source of bias by including a consecutive series of surgical patients within a broad time frame including years before publication of the ACC-AHA recommendations (Fleisher et al. 2007; Fleisher et al. 2009). Another limitation of the present study is the use of prescription records for assignment to β-blocker or non-β-blocker groups rather than actual medication administration data (not available for outpatients); therefore, we employed intent-to-treat analysis to account for non-compliance and inadvertent cross-overs (Lachin 2000). The present study did not specifically study the inclusion of epinephrine in local anesthetic solutions, the effects of other cardiovascular medications and the potential interactions between these medications and β-blockers, so this represents another area for future investigation (Cleophas et al. 2007). Lastly, this study was performed at one tertiary care, university-affiliated VA medical center, so the results may not be applicable to every institution or practice. For example, the overrepresentation of male patients in our sample, while typical of VA hospitals, limits generalizability to females. However, the sample size and duration of study, conditions of routine clinical practice, and heterogeneity of providers support the external validity of our study within this demographic.