Intraoperative hypotension has been associated with adverse perioperative outcomes including stroke,1 myocardial ischemia,2 cognitive dysfunction,3 and increased 1-year mortality.4 Therefore, defining the risk factors that predispose patients to intraoperative hypotension, including preoperative chronic antihypertensive therapy,5–8 are of clinical importance.
A number of studies suggest that patients receiving chronic angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs) experience more hypotension and require more vasopressor than patients receiving other antihypertensive medications during general anesthesia.6,7,9 Some studies suggest formulating guidelines to withhold ACEIs and ARBs preoperatively.5,8,10–12 However, few studies address the fact that patients with chronic hypertension are often taking multiple antihypertensive medications from different classes, which may interact and confound results,13 and few data are available on the effect of the number and class of antihypertensive drug on intraoperative hypotension.
We hypothesized that the magnitude of the hypotensive effect of a class of antihypertensive drug and the number of antihypertensive drugs could be quantified by the amount of intraoperative vasopressor required in patients undergoing carotid endarterectomy (CEA) under general anesthesia with a specific arterial blood pressure management. Based on these quantitative results, we could determine which classes of antihypertensive drugs, such as ACEIs, ARBs, and diuretics, increase the intraoperative vasopressor requirements compared with other classes of antihypertensive medications. We studied quantity of vasopressor as a surrogate for intraoperative hypotension, because we could not allow intraoperative hypotension in the CEA population.
This study is unique because we present data from patients having a single type of procedure, CEA, performed by a small group of surgeons, with a specific and uniform hemodynamic management strategy using arterial blood pressure goals as percentages of the patient's baseline blood pressure. Before clamping the carotid artery, the blood pressure is maintained at the preoperative baseline, during clamping approximately 20% above baseline, and after unclamping at or slightly below baseline. In addition, we include an analysis of the amount of vasopressor required as a function of not only drug class, but also the number of different chronic antihypertensive drugs taken by the patient.
This study, approved by the Columbia University IRB (New York, NY), is a post hoc analysis of 252 patients scheduled for elective CEA under general anesthesia. The patients were prospectively recruited for a study of postoperative cognitive dysfunction after CEA. For all patients, vital sign and drug administration data were available from computerized intraoperative record systems, eliminating the potential bias of handwritten data collection. Before enrollment, patients provided written consent for study participation.
All patients underwent unilateral CEA under general anesthesia. Preoperative data were collected from clinical documentation entered by anesthesiology residents, attendings, and fellows, in addition to data collected by nurses and study coordinators. The baseline arterial blood pressure was obtained by a nurse in the preoperative area on the day of surgery.
The preanesthetic evaluation sheets, completed by an attending anesthesiologist and resident by interview and physical examination on the day of surgery, were used to identify which medications patients were chronically receiving. The data were then categorized into antihypertensive class: ACEIs, ARBs, calcium channel blockers (CCBs), β-blockers (BBs), and diuretics. All patients were instructed by the neurosurgical staff to continue any antihypertensive medications they regularly take through and including the morning of surgery. All other medications were withheld. Computerized intraoperative records (CompuRecord; Philips Medical Systems, Andover, MA) were used intraoperatively in every case to record intraoperative hemodynamics, including arterial blood pressure and heart rate, as well as drug dosing.
The anesthetic technique and blood pressure management of patients were determined in advance and standardized for all study patients. Standard monitors, which included a blood pressure cuff, pulse oximeter, 5-lead electrocardiogram, and temperature probe, were used for every case. In addition, full montage electroencephalogram (EEG) and transcranial Doppler (TCD) monitoring were performed on every patient. Patients all received their IV lines in the operating room. The IV fluid goal was at least 1 L total for each case, regardless of the class of antihypertensive drug. Intraoperative blood pressure was measured using a radial intraarterial catheter that was placed before induction. The anesthetic technique involved a preinduction dose of midazolam (1–5 mg) and fentanyl (25–250 μg); induction by etomidate (0.15–0.3 mg/kg), propofol (1–2 mg/kg), or thiopental (3–5 mg/kg); and maintenance with 70% nitrous oxide in oxygen and isoflurane (0.3%–0.7%). The anesthesia was performed by 19 anesthesiologists (with one of them performing 86% of the cases), and the operations were performed by 8 surgeons (with 3 performing >50 cases each, and the remainder performing 1–5 cases).
The anesthetic management of a patient undergoing CEA was predetermined using a protocol anesthetic regimen. CEA surgery was divided into 3 intervals. The first interval of the case was the preclamp interval. This was the period of time starting at the induction of anesthesia and terminating at the clamping of the common carotid artery. During the preclamp interval, blood pressure was maintained to within 20% of the patients' mean arterial blood pressure (MAP) baseline (as determined in the holding area) before surgery. During the clamped interval, defined as starting when the clamp is placed on the common carotid artery, and ending when the clamp is removed, blood pressure was titrated up to approximately 20% above the baseline MAP. The third and final interval was the postclamp interval. During this interval, defined as after the clamp was released from the common carotid artery to the end of surgery, blood pressure was allowed to decrease to within 20% below the patients' baseline MAP. Blood pressure changes were attained using phenylephrine infusions supplemented with boluses. Phenylephrine infusions were administered using a Sigma Spectrum Infusion Pump (Baxter, Deerfield, IL) for all cases, which administers phenylephrine in micrograms/minute. To avoid a “see-saw effect” of multiple boluses, patients were started on a slow rate of phenylephrine soon after intubation, which was then titrated up or down as needed. When the infusion was altered (started or rate changed), the alteration would be entered manually by the anesthesiologist into the computerized record system. The total amount of phenylephrine administered during segments of the anesthetic was determined by the computerized record system based on the manual entry of pump rate into the computerized record system. The computerized record system collects and records data every 15 seconds.
Although a phenylephrine infusion was used from the beginning of the case, after induction of general anesthesia, phenylephrine boluses were also administered as needed at the discretion of the anesthesiologist to maintain the blood pressure at the targeted level. To further characterize the increased intraoperative vasopressor requirement, the phenylephrine requirements were converted into rates of micrograms/kilogram/minute to correct for potential differences in surgical time and patient weight. Ephedrine boluses were also administered if the patient exhibited significant bradycardia in conjunction with hypotension. Total phenylephrine and ephedrine doses were recorded on the intraoperative record system.
To address the interaction of multiple antihypertensive classes, we separated the data by the number of antihypertensive drugs chronically taken. Patients who took multiple classes of antihypertensive medications were characterized by number of classes of drugs taken, as well as which of those classes were taken. To further investigate whether there is an effect of antihypertensive class on vasopressor requirement, patients who only took a single class of antihypertensive were isolated to analyze which classes were associated with an increase in vasopressor requirement. For patients receiving only a single class of antihypertensive drug, we determined the amount of phenylephrine required by calculating the rate of phenylephrine administered during pre- and postclamp intervals of the case.
Statistical analyses were performed using Excel (Microsoft Corp., Redmond, WA) and JMP 7 software (SAS Institute, Inc., Cary, NC). All statistical tests were considered significant if the probability of a type I error was inferior to 0.05 (P value).
Univariate analyses were performed to address our hypothesis and determine whether there was a difference in the total phenylephrine requirement and the number of antihypertensive classes patients take (Fig. 1), and if there was a specific antihypertensive class that resulted in a higher phenylephrine requirement (Fig. 2). Furthermore, we analyzed whether the inclusion of diuretics, as one of the classes of a multiple antihypertensive regimen, would cause an increase in phenylephrine requirements in groups consisting of the same number of antihypertensive classes (Fig. 3).
In addition to univariate analyses, we analyzed our data with a multivariate regression model. Because the main outcome measure, total phenylephrine consumption (PHtotal), followed a log-normal distribution, it was transformed for analyses requiring linearity, and thereby, those results represent median values of phenylephrine use. We used multivariate linear regression to verify the association of antihypertensive medication class with PHtotal while adjusting for potential confounders (duration of surgery, age, body mass index, smoking status, induction drug, ephedrine dose, total fluid administered, coronary artery disease, history of congestive heart failure, and history of diabetes), which were included in the regression model using a forward, backward, and stepwise procedure, and the chosen model was identical starting with all available preoperative patient characteristics (Table 1).
To express the effect size of the independent variables in the regression model, we formulated a classification and regression tree (Fig. 4).
Two hundred fifty-two patients were included in this analysis. Of these, 68 patients (27%) were not taking a chronic antihypertensive drug. Of the patients receiving chronic antihypertensive therapy, 85 patients (35%) were taking a single antihypertensive, 68 patients (26%) were taking 2, 25 patients (9%) were taking 3, and 6 patients (3%) were taking 4 (Fig. 1). Of the patients taking a single antihypertensive, 27 were taking a BB, 23 were taking an ACEI, 22 were taking a CCB, 7 were taking a diuretic (all receiving a thiazide), and 6 were taking an ARB.
The prevalence of preoperative demographic variables of patients taking each class of antihypertensive therapy is shown in Table 2. There was a higher incidence of coronary artery disease in patients taking a BB or a CCB (P < 0.01 by Fisher right-sided exact test). Patients who were taking BBs also had a higher blood urea nitrogen and creatinine (P = 0.01 and P = 0.03, respectively, by Mann-Whitney-Wilcoxon test), and those taking diuretics had a higher blood urea nitrogen (P = 0.02 by Mann-Whitney-Wilcoxon test) than those not taking this antihypertensive class.
To maintain blood pressure at a steady level within 20% of their baseline preclamp and 20% above baseline during carotid clamping, all patients had a phenylephrine infusion started after anesthetic induction. Most of the phenylephrine was administered by continuous infusion (mean of 3888 μg, SD of 3216 μg). The remainder of the phenylephrine was administered through 40 μg/mL boluses (mean of 132 μg, SD of 190 μg). In addition to the phenylephrine infusion and boluses, one-third of patients (82) received additional ephedrine boluses. The mean total dose of ephedrine was 5.5 mg (SD 10.3 mg). There was no association between group of antihypertensive drug and either the number of ephedrine boluses administrated or the total amount of ephedrine administered during the case. No patient required other drugs such as vasopressin or epinephrine for maintaining blood pressure at the predetermined level.
Patients received a mean of 1250 mL (SD of 490 mL) of lactated Ringer solution. There was no difference between the amount of fluids received and antihypertensive class or number of classes of antihypertensive drugs taken.
The total intraoperative amount of phenylephrine required depended on the number of different classes of antihypertensive medications the patients were taking (P = 0.002 by Kruskal-Wallis test). Patients who were taking 3 classes of medications required an average of 75% more total phenylephrine than those taking 0, 1, or 2 classes. Those patients taking fewer than 3 classes of medications did not differ in the average amount of phenylephrine administered (Fig. 1).
Univariate analysis of the relationship between intraoperative pressor requirement and class of antihypertensive drug was performed using the Mann-Whitney test. Analyses showed that patients taking diuretics required significantly more total phenylephrine than those not taking diuretics (P < 0.001). There was no significant difference in a comparison of patients who were taking other classes of antihypertensives including ACEIs, ARBs, BBs, and CCBs with those who were not taking that particular class of drug (Fig. 2). This was true when the total phenylephrine dose was divided into phenylephrine administered by boluses (P = 0.0154), and the phenylephrine administered by continuous infusion (P < 0.0001). There was no other antihypertensive class in which the total phenylephrine, bolus phenylephrine, or infused phenylephrine amount was different between those taking the antihypertensive and those not taking the antihypertensive.
In addition, when comparing patients taking only 1 antihypertensive class, only patients taking diuretics required more pressor medication. Furthermore, of those taking 2 classes of antihypertensive drugs, those who chronically took diuretics required significantly more phenylephrine than those taking 2 other classes of drugs (P = 0.04) (Fig. 3). Although patients receiving diuretic therapy as part of a 3-class antihypertensive regimen required more vasopressor than those receiving other drugs, the difference was not significant (P = 0.14) (Fig. 3). Patients taking diuretics as part of an antihypertensive regimen required 125%, 50%, and 48% more total phenylephrine when taking a total of 1, 2, or 3 classes of antihypertensive drugs, respectively (Fig. 3).
When all patients were analyzed by multivariate regression model, including those taking multiple antihypertensive drugs (ACEIs, ARBs, BBs, and CCBs), patients who were taking diuretics in their antihypertensive regimen required significantly more total phenylephrine during CEA than those not taking a diuretic.
Potential confounders that were screened for included duration of surgery, age, body mass index, smoking status, induction drug, ephedrine dose, total fluid administered, coronary artery disease, history of congestive heart failure, and history of diabetes. Of all factors, only duration of surgery (P = 0.005), body mass index (P < 0.001), ephedrine dose (P = 0.015), and chronic diuretic use (P = 0.01) remained statistically significant.
To express the effect size of the independent variables in the regression model, we formulated a classification and regression tree. The duration of surgery was the covariate with the highest association with the outcome. The data were then split into 2 groups (≤135 minutes and >135 minutes). For those >135 minutes for duration of surgery, the next highly associated covariate was diuretic therapy. The data were again split into 2 groups: patients taking diuretics (diuretics >0), and patients not taking diuretics (diuretics <0). For those not taking diuretics, the most associated variable was again duration of surgery. One can similarly interpret other branches of the tree (Fig. 4).
Residual analysis showed that there were no large outliers. The residuals are shown to be normally distributed in Figure 5. Moreover, results from robust linear regression were almost identical to those from linear regression. The leverage and Cook's distance were determined for the data (Fig. 6) and none of the data points has a Cook's distance >0.5. To account for possible colinearity among the covariates, we standardized the continuous covariates. Also, estimates from ridge regression were almost identical to those from the linear regression.
To address the question of when pressor medication is required and to generalize our results, we evaluated the amount or rate of pressor use as a function of the operative interval. Because we had fewer patients receiving monotherapy, and body mass index and duration of surgery also influenced total phenylephrine requirements as found by the multivariate analysis above, we expressed the pre- and postclamp rates of phenylephrine in terms of micrograms/kilogram/minute determined by total micrograms of phenylephrine given by infusion in either interval divided by patient weight and interval time. Patients who were taking diuretic therapy required a significantly higher rate of phenylephrine to keep their blood pressure at baseline during the preclamp interval than those receiving any other antihypertensive medication (P < 0.05) (Fig. 7). This result was not found going from the preclamp interval to the clamped interval. During the clamped interval, there was no significant difference in the rate of phenylephrine needed to increase the blood pressure to approximately 20% above baseline across the different antihypertensive classes, from similar preclamp blood pressures, as a fraction of baseline (Fig. 8).
We present a post hoc analysis of vasopressor requirements of patients undergoing CEA under general anesthesia. The results demonstrate that patients taking preoperative diuretic therapy are associated with increased intraoperative vasopressor requirement during CEA when it is used as a single antihypertensive, as well as when it is used as part of a multiple antihypertensive regimen. There is no significant increase of vasopressor requirement for patients taking the other classes of antihypertensives including ACEIs, ARBs, CCBs, and BBs compared with those not taking antihypertensive medications.
Our results are applicable to patients having general anesthesia for other types of surgery. In the preclamp interval, before clamping the carotid, blood pressure is kept within 20% of baseline, a similar situation to patients having other surgery under general anesthesia.
EEG is our accepted standard to determine the adequacy of collateral blood flow to the brain. However, the EEG is insensitive at detecting focal ischemia and requires as much as a 60% decrease in normal cerebral blood flow before subtle EEG changes can be detected. Although TCD evaluates the blood flow velocity in 1 of the 4 major vessels supplying blood to the brain, to titrate blood pressure to lower values again seems risky by potentially letting focal areas become ischemic. Also, one has to remember that TCD measures cerebral blood flow velocity not cerebral blood flow. Therefore, one does not know where on the normal spectrum of blood flow the patient is when there may be a reduction of cerebral blood flow velocity. EEG has been demonstrated to reflect the adequacy of cortical cerebral blood flow to maintain cerebral activity more directly than TCD. We are aware that some experienced neuroanesthesiologists use the TCD to titrate systemic blood pressure, but we prefer to take a more conservative approach. Therefore, during clamping, our group increases blood pressure, if tolerated cardiovascularly by the patient, to 20% above their baseline. We then observe the EEG and TCD monitors and if the perfusion is inadequate, we can use these monitors to view what change an increase in blood pressure can make, often as the surgeon is preparing to place a shunt.
Diuretics are unique in altering the circadian rhythm of blood pressure. Hypertensive patients can be grouped into 2 categories based on their nocturnal (or sleeping) blood pressure: “dippers” are patients whose nocturnal blood pressure decreases by 10%, whereas “nondippers” are patients who do not experience this decrease. These different patterns are of clinical significance because of the results of a large prospective trial that showed that nondippers are at an increased risk of cardiovascular morbidity compared with dippers, whereas “extreme dippers” (those with an exaggerated ≥20% nocturnal blood pressure decrease) are at an increased risk of stroke.14,15 The period of “dipping” is characterized by a response to the decrease of sympathetic activation.16 Dipping is restored to patients taking hydrochlorothiazide diuretics probably because of increased sodium excretion during the daytime.17
In the perioperative hypertensive patient, a thiazide diuretic such as hydrochlorothiazide may enhance a decrease of blood pressure or dipping with the induction of anesthesia, similar to the dipping enhanced in patients taking thiazide diuretics during sleep. Both periods are characterized by a decrease in sympathetic stimulation. Therefore, the benefit to cardiovascular morbidity offered to the patient by maintaining circadian blood pressure may increase blood pressure decrease in the operating room upon a decrease in sympathetic stimulation.
We can expect that many patients presenting for anesthesia will be taking chronic thiazide diuretics as a first-line antihypertensive drug because thiazide diuretics have been confirmed by the Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial to be equal or superior to other antihypertensive drugs as a first-line therapy.18
There are a number of case reports of refractory hypotension in patients taking chronic ACEIs and ARBs undergoing general anesthesia. Currently, there is debate whether to withhold these medications preoperatively. To address this debate, a number of small studies have been conducted to evaluate whether patients taking chronic ACEIs or ARBs experience hypotension and require more vasopressors than patients taking other antihypertensive medications during general anesthesia.6,7,9 The major limitation of these studies is their small sample size and a large number of other antihypertensive drugs taken.
Kheterpal et al.13 performed a large retrospective study of 65,043 patients having noncardiac surgery in which they analyzed the number of vasopressor boluses (of equivalent doses of ephedrine and phenylephrine) in patients who were receiving different combinations of antihypertensives. They concluded that patients receiving chronic diuretic therapy along with ACEI/ARB therapy had greater vasopressor requirements compared with patients receiving an ACEI/ARB combined with a CCB. However, patients who were taking a CCB along with a diuretic required the same number of vasopressor boluses as the group that was taking an ACE/ARB plus diuretic. This observation is in concordance with our data, which show that when a diuretic is added to antihypertensive management, intraoperative pressor requirements increase. Similar to our findings, it suggests that ACEIs and ARBs are not responsible for increased intraoperative vasopressor requirements alone, and that diuretics have a key role in these requirements. The major limitations of the study by Kheterpal et al. are that although the study was large, the type and length of surgery varied widely and there was no set anesthetic protocol. The number of equivalent vasopressor boluses was calculated, but this information was not combined with vasopressor infusion data to yield the total amount or rate of vasopressor used. In addition, BBs were excluded from their analysis, so it is not possible to analyze exactly how many antihypertensive medications each patient was receiving.
Our study is unique because the anesthetic technique was standardized. In contrast to previous studies, this study involved a single type of surgery performed by 1 of 3 surgeons at a single institution. The entire intraoperative interval was addressed, and the data were divided into 3 intervals corrected for time, thus correcting for potential differences in surgical length. Because the patients were all undergoing the same type of surgery, they had similar preoperative factors. During the intraoperative course, the hemodynamic variables were measured precisely and the vasopressor used was standardized. In addition, all data were recorded by a computerized medical record keeper, which made drug infusion dose calculations as well as blood pressure measurement recordings more accurate.19
This study has several limitations, including its post hoc design and small sample size. In addition, it was not clear how long the patients were taking their current medication regimen and the exact amount of time before the surgery that their last dose was administered. Patient noncompliance may also interfere with the results of the study, especially because patients taking multiple medications are less likely to be taking all the medications appropriately. However, we do show the response of the patient in the “real world” clinical scenario, in which we can assume a certain degree of noncompliance. In addition, the rate of phenylephrine set on the pump was added to a computerized record system separately, and frequent changes in phenylephrine rate may not have all been recorded. However, a typical case involved 5 to 8 changes to the rate of phenylephrine.
Although the results of the study are subject to some limitations, they are in agreement with previous studies. Our study design addresses vasopressor requirements in patients managed with a standardized protocol-controlled anesthetic technique and blood pressure management strategy. The analysis of the clinical response of the hypertensive patient to different antihypertensive therapy is of clinical significance to the anesthesiologist in the prediction and prevention of intraoperative hypotension during general anesthesia. The fact that this difference occurs in the preclamp period makes this study relevant to other surgical procedures performed under general anesthesia.
Diuretics are a large confounding factor, if not an independent factor of attributing the cause of intraoperative hypotension under general anesthesia. In addition, patients taking 3 different classes of antihypertensive drugs require significantly more phenylephrine than those taking a fewer number of classes. The patients that are described in this study are of a specific category: patients with carotid disease and hypertension. Further laboratory work and prospective clinical studies must be performed to elucidate the mechanism of this interaction, as well as the other possible patient populations to which the findings may be applicable.
Name: Zirka H. Anastasian, MD.
Contribution: Study design, collection of data, data analysis, manuscript preparation.
Name: John G. Gaudet, MD.
Contribution: Data analysis, manuscript preparation.
Name: E. Sander Connolly, Jr., MD.
Contribution: Manuscript preparation, other.
Name: Srikesh Arunajadai, PhD.
Contribution: Data analysis.
Name: Eric J. Heyer, MD, PhD.
Contribution: Study design, collection of data, data analysis, manuscript preparation.
We acknowledge Drs. Alastair and Margret Wood for their help and inspiration in preparing the manuscript, as well as SaeJin Kim for her help in preparing the manuscript.
1. Limburg M, Wijdicks EF, Li H. Ischemic stroke after surgical procedures: clinical features, neuroimaging, and risk factors. Neurology 1998;50:895–901
2. Barone JE, Bull MB, Cussatti EH, Miller KD, Tucker JB. Perioperative myocardial infarction in low-risk patients undergoing noncardiac surgery is associated with intraoperative hypotension. J Intensive Care Med 2002;17:250–5
3. Yocum GT, Gaudet JG, Teverbaugh LA, Quest DO, McCormick PC, Connolly ES Jr, Heyer EJ. Neurocognitive performance in hypertensive patients after spine surgery. Anesthesiology 2009;110:254–61
4. Monk TG, Saini V, Weldon BC, Sigl JC. Anesthetic management and one-year mortality after noncardiac surgery. Anesth Analg 2005;100:4–10
5. Bertrand M, Godet G, Meersschaert K, Brun L, Salcedo E, Coriat P. Should the angiotensin II antagonists be discontinued before surgery? Anesth Analg 2001;92:26–30
6. Brabant SM, Bertrand M, Eyraud D, Darmon PL, Coriat P. The hemodynamic effects of anesthetic induction in vascular surgical patients chronically treated with angiotensin II receptor antagonists. Anesth Analg 1999;89:1388–92
7. Brabant SM, Eyraud D, Bertrand M, Coriat P. Refractory hypotension after induction of anesthesia in a patient chronically treated with angiotensin receptor antagonists. Anesth Analg 1999;89:887–8
8. Comfere T, Sprung J, Kumar MM, Draper M, Wilson DP, Williams BA, Danielson DR, Liedl L, Warner DO. Angiotensin system inhibitors in a general surgical population. Anesth Analg 2005;100:636–44
9. Rosenman DJ, McDonald FS, Ebbert JO, Erwin PJ, LaBella M, Montori VM. Clinical consequences of withholding versus administering renin-angiotensin-aldosterone system antagonists in the preoperative period. J Hosp Med 2008;3:319–25
10. Coriat P, Richer C, Douraki T, Gomez C, Hendricks K, Giudicelli JF, Viars P. Influence of chronic angiotensin-converting enzyme inhibition on anesthetic induction. Anesthesiology 1994;81:299–307
11. Pigott DW, Nagle C, Allman K, Westaby S, Evans RD. Effect of omitting regular ACE inhibitor medication before cardiac surgery on haemodynamic variables and vasoactive drug requirements. Br J Anaesth 1999;83:715–20
12. Raja SG, Fida N. Should angiotensin converting enzyme inhibitors/angiotensin II receptor antagonists be omitted before cardiac surgery to avoid postoperative vasodilation? Interact Cardiovasc Thorac Surg 2008;7:470–5
13. Kheterpal S, Khodaparast O, Shanks A, O'Reilly M, Tremper KK. Chronic angiotensin-converting enzyme inhibitor or angiotensin receptor blocker therapy combined with diuretic therapy is associated with increased episodes of hypotension in noncardiac surgery. J Cardiothorac Vasc Anesth 2008; 22:180–6
14. Verdecchia P, Porcellati C, Schillaci G, Borgioni C, Ciucci A, Battistelli M, Guerrieri M, Gatteschi C, Zampi I, Santucci A, Reboldi G. Ambulatory blood pressure: an independent predictor of prognosis in essential hypertension. Hypertension 1994;24:793–801
15. Kario K, Matsuo T, Kobayashi H, Imiya M, Matsuo M, Shimada K. Nocturnal fall of blood pressure and silent cerebrovascular damage in elderly hypertensive patients: advanced silent cerebrovascular damage in extreme dippers. Hypertension 1996;27:130–5
16. Kario K, Schwartz JE, Pickering TG. Changes of nocturnal blood pressure dipping status in hypertensives by nighttime dosing of alpha-adrenergic blocker, doxazosin: results from the HALT study. Hypertension 2000;35:787–94
17. Uzu T, Kimura G. Diuretics shift circadian rhythm of blood pressure from nondipper to dipper in essential hypertension. Circulation 1999;100:1635–8
18. ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group. The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial. Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic: the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). JAMA 2002;288:2981–97
19. Reich DL, Wood RK Jr, Mattar R, Krol M, Adams DC, Hossain S, Bodian CA. Arterial blood pressure and heart rate discrepancies between handwritten and computerized anesthesia records. Anesth Analg 2000;91:612–6
© 2011 International Anesthesia Research Society
20. Hothorn T, Hornik K, Zeileis A. Unbiased recursive partitioning. J Comput Graph Stat 2006;15:651–74