- Email: [email protected]
Resistant hypertension and preoperative silent myocardial ischaemia in surgical patients†
British Journal of Anaesthesia 1994; 73: 574-578
CLINICAL INVESTIGATIONS
Resistant hypertension and preoperative silent myocardial ischaemia in surgical patientsf K. G. ALLMAN, A. MUIR, S. J. HOWELL, A. E. HEMMING, J. W. SEAR AND P. FOEX
Summary
Key w o r d s Heart, ischaemia. Complications, hypertension.
Perioperative cardiac morbidity is the leading cause of death after anaesthesia and surgery [1]. In the UK, between 12 and 20% of patients presenting for non-cardiac surgery have preoperative evidence of cardiovascular disease and of these the commonest cardiovascular complaint is arterial hypertension [2, 3]. Despite this, the role of arterial hypertension as a determinant of perioperative morbidity is still uncertain. Early work suggested that pre-existing hypertension was associated with intraoperative cardiovascular instability and myocardial ischaemia [4-7] and recently a history of arterial hypertension has been found to be associated with a poorer postoperative outcome [8, 9]. However, a prospective multivariate analysis of more than 1000 noncardiac operations by Goldman and colleagues failed to show arterial hypertension as an independent predictor of postoperative cardiac complications [10]. The use of ambulatory ECG monitoring is now a widely accepted method of detecting myocardial ischaemia. Modern microprocessing technology enables continuous and accurate measurement of ST segment changes; in patients with coronary artery disease, these have been shown to correlate with reduced myocardial perfusion [11, 12], such episodes
Patients and methods We studied 351 patients presenting for elective noncardiac surgery. Patients gave informed consent and the study was approved by the Central Oxford Research Ethics Committee. All patients were interviewed and examined by one of four investigators (K.G.A., A.M., A.E.H., SJ.H.) and details of cardiovascular risk factors noted. Preoperative investigations, ECG and arterial pressure readings were recorded. Admission arterial pressure was defined as the first in-hospital recording taken by one of the investigators, ward nurse or house physician. All pressure readings were obtained with a manual mercury sphygmomanometer. Patients subsequently underwent ambulatory ECG monitoring using a Medilog FD-2 or 6000 FD ambulatory ECG monitoring unit (Oxford Medical Limited, Abingdon, UK) for up to 24 h before surgery (minimum 8 h). Recordings were made from silver-silver chloride skin electrodes placed to record standard lead III and CMV5. Data from the ambulatory unit were then downloaded via optical link onto an IBM compatible 286 microcomputer and analysed for ST segment changes using the Medilog Prima ECG analysis software. A significant ST segment change was defined as a horizontal or downsloping depression of greater than 1 mm or an elevation of greater than 2 mm, for a period of more than 60 s. The ECG baseline was measured at a point 60 ms before the maximum R wave and the ST segment measured 100 ms after the maximum R K. G. ALLMAN, MB, CHB, FRCA, A. MUIR, MB, BS, MSC, FANZCA, S. J. HOWELL, MA, MHCP, FRCA, A. E. HEMMING, MB, BS, BSC, FRCA, J. W. SEAR, MA, BSC, PHD, FFARCS, P. Fo£x, DM, DPHIL, FRCA,
FANZCA, Nuffield Department of Anaesthetics, John Radcliffe Hospital, Headington, Oxford OX3 9DU. Accepted for publication: May 12, 1994. Correspondence to K.G.A. fPresentcd in part at the Anaesthetic Research Society, Leicester Meeting, July 10-11, 1992 (British Journal of Anaesthesia 1992; 69: 540P).
Downloaded from http://bja.oxfordjournals.org/ at University of Birmingham on September 2, 2015
We studied 325 patients undergoing elective noncardiac surgery who had preoperative ambulatory ECG monitoring performed for a duration of 5130 h (range 8-24 h; mean 15.8h). Sixty-four subjects (20%) had one or more episodes of ST segment depression consistent with myocardial ischaemia. Of all preoperative cardiovascular variables measured, the presence of elevated arterial pressure, despite patients being maintained on long term antihypertensive therapy, was the only factor associated significantly with the presence of preoperative silent myocardial ischaemia (P < 0.002). This correlation was confirmed when arterial hypertension was defined in four separate ways. The incidence of silent ischaemia in these patients was 33-55%. We suggest that admission arterial pressure may therefore be a useful screening test to identify patients at risk of preoperative myocardial ischaemia. (Br. J. Anaesth. 1994; 73: 574-578)
being clinically silent in more than 70 % of patients [13]. The presence of silent myocardial ischaemia in patients after myocardial infarction or with known stable coronary artery disease is predictive of adverse cardiac events and sudden death [14-17]. In the present study we have investigated the relationship between incidence of preoperative myocardial ischaemia and arterial hypertension in patients undergoing general or vascular surgery.
575
Preoperative silent ischaemia Table 1 Patient characteristics and numbers of patients with silent myocardial ischaemia
Male Vascular surgery Previous myocardial infarction History of angina Peripheral vascular disease Smoking (> 20 pack years) Obesity (> 20 % ideal weight) Diabetes mellitus Age > 70 yr
STATISTICAL ANALYSIS
Admission arterial pressures for each of the four categories were compared using unpaired two-tailed Student's t tests with correction for multiple testing. The adequacy of pressure control (as defined earlier) was compared with the presence or absence of preoperative silent myocardial ischaemia using the chi-square test or contingency table analysis with a continuity correction. A similar approach was used to assess the relationship between antihypertensive therapy and the incidence of silent myocardial ischaemia. The duration of ECG monitoring was divided into 4-h epochs and compared with the presence or absence of silent myocardial ischaemia and hypertensive-treatment category, again using contingency table analysis with a continuity correction. Indices of specificity, sensitivity and positive-
No. with silent ischaemia (%) (n = 64)
P
212 (65.2) 177 (54.5) 59 (18.2) 31 (9.5) 105 (32.3) 148 (45.5) 75 (23.2) 24 (7.4) 111(34.2)
41 (19.3) 39 (22.0) 14 (23.7) 9 (29.0) 25 (23.8) 30 (20.3) 12 (16.0) 4(16.7) 25 (22.5)
0.942 0.307 0.496 0.255 0.254 0.921 0.444 0.904 0.437
negative predictive value were calculated using standard formulae. Statistical analyses were performed using Statview SE and a Macintosh LC II microcomputer. Results We studied 351 patients (table 1). Forty-one per cent were receiving antihypertensive medication. Twenty-six subjects were subsequently excluded from analysis (recorder failure 14, ECG changes precluding ST segment analysis nine, patient refusal two, operation cancelled one). The remaining 325 patients had preoperative ambulatory ECG monitoring for a total duration of 5130 h (range 8—24 h; mean 15.8 h). Sixty-four (20.0%) patients had one or more episodes of ST segment depression consistent with myocardial ischaemia. No subject complained of angina or chest pain during the study. The mean duration of ischaemia was 88.3 (range 1-1078) min with a mean ischaemic load time of 338.3 (3.6-3529) s rr 1 . The number of patients in each of the four groups, based on classifications of arterial hypertension, is shown in table 2. Of all the preoperative variables tested, only hypertensive category was associated significantly with the presence or absence of silent myocardial ischaemia (tables 1,3). Whichever definition of arterial hypertension was used, the poorly treated group had a significantly greater incidence of preoperative silent myocardial ischaemia (33—55%) compared with the other groups (P = 0.0016-0.00O1). Using a definition of arterial hypertension of three or more diastolic readings 5* 100 mm Hg, 16 of 29 (55 %) poorly treated patients had episodes of silent myocardial ischaemia compared with 19 of 103 (18%) well treated patients, nine of 32 (28%) untreated patients and 20 of 161(12%) normotensive patients (P = 0.0001). There was no association between duration of ECG monitoring and the presence or absence of silent myocardial ischaemia (P = 0.159), or with any particular hypertensive-treatment category for any of the definitions of arterial hypertension (P = 0.108-0.310). When admission pressures were compared between groups there was no significant difference between the untreated and poorly treated groups (P = 0.69), but a clinically small, statistically significant difference between the normotensive and well treated groups (P = 0.0001) (table 4).
Downloaded from http://bja.oxfordjournals.org/ at University of Birmingham on September 2, 2015
wave. Each ST segment deviation was terminated by a return to baseline of more than 60 s. All reported episodes of ST segment change were verified visually. Inclusion criteria were age greater than 40 yr, ASA grade I—III and elective admission for major noncardiac surgery. All subjects had at least three preoperative arterial pressure measurements recorded, including admission arterial pressure. Patients were then classified into one of four groups: normotensive not taking antihypertensive medication (normotensive); normotensive taking regular antihypertensive medication (well treated); hypertensive not taking regular antihypertensive medication (untreated); and hypertensive despite regular antihypertensive medication (poorly treated). Each patient was classified on the basis of four separate definitions of arterial hypertension. This was done to reduce the introduction of bias from, and to examine die effect of, different hypertensive definitions. These were as follows: (1) adm $s 150/80 = either admission systolic 5s 150 or admission diastolic 3s 80 mm Hg; (2) adm 3s 160/90 = either admission systolic 3s 160 or admission diastolic 3* 90mmHg; (3) adm 3s 170/100 = either admission systolic 3s 170 or admission diastolic 3* lOOmmHg; and (4) diast. 3 3s 100 = three or more diastolic readings 5* 100 mm Hg. After ECG analysis, the ischaemic load (duration of myocardial ischaemia/duration of recording) was calculated for each patient.
No. (%) (n = 325)
British Journal of Anaesthesia
576
Table 2 Number of patients in each of the four classification groups of arterial hypertension (n = 325) (see text for definitions of groups)
Normotensive Well treated Untreated Poorly treated
Adm > 150/80 (No. (%))
Adm > 160/90 Adm > 170/100 (No. (%)) (No. (%))
Diast. 3 ^ 100 (No. (%))
112(34) 41(13) 81 (25) 91 (28)
141 (43) 70 (22) 52(16) 62 (19)
161 (49) 103 (32) 32 (10) 29(9)
162(50) 105 (32) 31 (10) 27(8)
Table 3 Number of patients with silent myocardial ischaemia in each of the four classification groups of arterial hypertension (n = 64) (see text for definitions of groups) Adm > 160/90 (No. (%))
Adm> 170/100 (No. (%))
Diast. 3 > 100 (No. (%))
14 of 112 (13) 5 of 41 (12) 15 of 81 (19) 30 of 91 (33) P = 0.0016
17 of 141 (12) 12 of 70(17) 12 of 52 (23) 23 of 62 (37) P = 0.0005
22 of 162 (14) 21 of 105 (20) 7 of 31 (23) 14 of 27 (52) P = 0.0001
20 of 161 (12) 19 of 103 (18) 9 of 32 (28) 16 of 29(55) P = 0.0001
Table 4 Admission arterial pressure for each of the four groups classified on the basis of three or more diastolic readings Js 100 mm Hg (mean (SD)). *P < 0.0001 normotensive vs well treated group
Normotensive Well treated Untreated Poorly treated
Mean systolic pressure (mm Hg)
Mean diastolic pressure (mm Hg)
141 (21) 153(22) 181(22) 178(23)
80(11) 86 (9.2) 107(12) 106(12)
Table 5 Number of patients taking each antihypertensive treatment (n = 132). (Data analysed for each drug treatment using chi-square test with a continuity correction for small numbers)
ACE inhibitors Beta blockers Calcium antagonists Diuretics
No. (%)
No. with silent ischaemia (%)
25(19) 55 (42) 52 (39) 63 (48)
6(24) 14 (25) 16 (31) 20 (32)
0.948 0.973 0.490 0.270
For poorly treated patients classified as having three or more diastolic readings ^100 mm Hg, the sensitivity and specificity for silent myocardial ischaemia were 25 and 95%, respectively, with a positive predictive value of 55 % and a negative predictive value of 84 %. When the poorly treated and well treated groups were analysed further according to treatment, no single therapy was found to be associated with the presence of, or was protective against, silent myocardial ischaemia (table 5). Discussion The presence of silent myocardial ischaemia (SMI) has been shown to be a significant predictor of postoperative adverse myocardial events. Slogoff and Keats found silent ischaemia occurring in the anaesthetic room to be predictive of postoperative myocardial infarction for patients undergoing car-
diac surgery [7, 18]. Raby and colleagues [19-21], in a series of 176 patients, and Pasternack and coworkers [22], in a series of 200 patients undergoing vascular surgery, found preoperative SMI to be the most significant predictor of postoperative morbid cardiac events. Similar results were described for those undergoing both vascular and non-vascular surgery [23]. However, Mangano and colleagues found preoperative SMI to be weakly associated only with, and not a predictor of, postoperative cardiac events and death [24, 25]. Other studies have shown preoperative SMI to be predictive of intraoperative myocardial ischaemia [26] and postoperative SMI [27, 28]. Detection of SMI therefore has important implications in the management of patients presenting for major surgery. We have shown that patients who appear to be resistant to conventional antihypertensive therapies have an extremely high incidence (> 50%) of preoperative SMI on 24-h ambulatory ECG monitoring. The mechanism for this remains unclear. It is likely that patients with long-standing arterial hypertension who do not respond well to treatment have more severe and widespread arterial disease. Using a definition of hypertension as three diastolic readings ^ 100 mm Hg, 55% of poorly treated patients had preoperative episodes of SMI compared with 18% of well treated patients and 28% of untreated patients. It is important to note that the poorly treated group, although resistant to treatment, did not have grossly elevated arterial pressure readings. There was no significant difference between the admission pressures of the poorly treated and untreated groups. As our population consisted of patients presenting for routine major surgery and presently accepted practice recommends postponement if diastolic pressure is consistently greater than 110-115 mm Hg [5], it is not surprising that we did not find a large incidence of treatment-resistant patients. At present the importance of "white coat" hypertension on admission to hospital remains unclear. In this study, there was no significant difference in the rates of SMI among the four categories, regardless of whether hypertension
Downloaded from http://bja.oxfordjournals.org/ at University of Birmingham on September 2, 2015
Normotensive Well treated Untreated Poorly treated
Adm > 150/80 (No. (%))
Preoperative silent ischaemia
Acknowledgements This study was supported by the Wellcome Trust (K.G.A., A.E.H., S.J.H.), the trustees of the WG Ceizar Scholarship, University of Tasmania (A.M.), the British Heart Foundation and the Medical Research Fund, University of Oxford.
References 1. Mangano DT. Perioperative cardiac morbidity. Anesthesiology 1990; 72: 153-184.
2. Edwards ND, Reilly CS. Detection of perioperarive myocardial ischaemia. British Journal of Anaesthesia 1994; 72: 104-115. 3. Forrest JB, Rehder K, Cahalan MK, Goldsmith CH. Multicenter study of general anesthesia: Predictors of severe perioperative adverse outcomes. Anesthesiology 1992; 76: 3-15. 4. Prys-Roberts C, Meloche R, Foex P. Studies of anaesthesia in relation to hypertension. 1: Cardiovascular responses of treated and untreated patients. British Journal of Anaesthesia 1971; 43: 122-137. 5. Prys-Roberts C. Hypertension and anesthesia—fifty years on. Anesthesiology 1979; 50: 281-284. 6. Stone JG, Foex P, Sear JW, Johnson LL, Khambatta HJ, Triner L. Myocardial ischemia in untreated hypertensive patients: effect of a single small oral dose of a beta-adrcnergic blocking agent. Anesthesiology 1988; 68: 495-500. 7. Slogoff S, Keats AS. Does perioperarive myocardial ischemia lead to postoperative myocardial infarction? Anesthesiology 1985; 62: 107-114. 8. Browner WS, Li J, Mangano DT. In-hospital and long-term mortality in male veterans following noncardiac surgery. The Study of Perioperative Ischemia Research Group. Journal of the American Medical Association 1992; 268: 228-232. 9. Fletcher AE, Bradley IC, Broxton JS, Bulpitt CJ, Davis AJ, Dyker GS, Ganvir PL, Gostick NK, Mayhew SR, Mukerji D. Survival of hypertensive subjects identified on screening: results for sustained and unsustained diastolic hypertension. European Heart Journal 1992; 13: 1595-1601. 10. Goldman L, Caldera DL, Nussbaum SR, Southwick FS, Krogstad D, Murray B, Burke DS, O'Malley TA, Goroll A, Caplan CH, Nolan J, Carabello B, Slater EE. Multifactorial index of cardiac risk in noncardiac surgical procedures. New England Journal of Medicine 1977; 297: 845-850. 11. Wilson RF, Marcus ML, Christensen BV, Talman C, White CW. Accuracy of exercise electrocardiography in detecting physiologically significant coronary arterial lesions. Circulation 1991; 83: 412^121. 12. Boden WE, Spodick DH. Diagnostic significance of precordial ST-segment depression. American Journal of Cardiology 1989; 63: 358-361. 13. Epstein SE, Quyyumi AA, Bonow RO. Myocardial ischemia—silent or symptomatic. New England Journal of Medicine 1988; 318: 1038-1043. 14. Kennedy HL, Wiens RD. Ambulatory (Holter) electrocardiography and myocardial ischemia. American Heart Journal 1989; 117: 164-176. 15. Sheps DS, Heiss G. Sudden death and silent myocardial ischemia. American Heart Journal 1989; 117: 177-184. 16. Rocco MB, Nabel EG, Campbell S, Goldman L, Barry J, Mead K, Selwyn AP. Prognostic importance of myocardial ischemia detected by ambulatory monitoring in patients with stable coronary artery disease. Circulation 1988; 78: 877-884. 17. Tzivoni D, Gavish A, Zin D, Gottlieb S, Moriel M, Keren A, Banai S, Stern S. Prognostic significance of ischemic episodes in patients with previous myocardial infarction. American Journal of Cardiology 1988; 62: 661-664. 18. Slogoff S, Keats AS. Further observations on perioperative myocardial ischemia. Anesthesiology 1986; 65: 539—542. 19. Raby KE, Barry J, Creager MA, Cook EF, Weisberg MC, Goldman L. Detection and significance of intraoperative and postoperative myocardial ischemia in peripheral vascular surgery. Journal of the American Medical Association 1992; 268: 222-227. 20. Raby KE, Goldman L, Cook EF, Rumerman J, Barry J, Creager MA, Selwyn AP. Long-term prognosis of myocardial ischemia detected by Holtcr monitoring in peripheral vascular disease. American Journal of Cardiology 1990; 66: 1309-1313. 21. Raby KE, Goldman L, Creager MA, Cook EF, Weisberg MC, Whittemore AD, Selwyn AP. Correlation between preoperative ischemia and major cardiac events after peripheral vascular surgery. New England Journal of Medicine 1989; 321: 1296-1300. 22. Pastemack PF, Grossi EA, Baumann FG, Riles TS, Lamparello PJ, Giangola G, Primis LK, Mintzer R, Imparato AM. The value of silent myocardial ischemia monitoring in the prediction of perioperative myocardial infarction in patients undergoing peripheral vascular surgery. Journal of Vascular Surgery 1989; 10: 617-625.
Downloaded from http://bja.oxfordjournals.org/ at University of Birmingham on September 2, 2015
was defined as an admission pressure of > 170/100 mm Hg or as three diastolic readings ^ 100 mm Hg. This would suggest that in this study at least, "white coat" hypertension had little or no measurable effect. The problem of preoperative arterial hypertension is a common one. In a prospective study of more than 17000 patients presenting for routine general anaesthesia and surgery, Forrest and co-workers found a prevalence rate of 13.6% [3]. Arterial hypertension itself has been shown in several studies to be an independent predictor of perioperative SMI [7, 27, 29] and also of postoperative morbidity and mortality [8, 21]. This is the first time, however, that patients resistant to antihypertensive medication have been shown to have such a high incidence of preoperative SMI, and indeed it may be that this group is responsible for the increased morbidity associated with a history of arterial hypertension. Previous studies have found antihypertensive treatment to be protective against peroperative arterial pressure lability and associated ischaemia [4, 30]. Stone and colleagues found that preoperative treatment with beta blocking agents reduced the incidence of intraoperative myocardial ischaemia in mild to moderate untreated hypertensive patients [6]. However, beta blockers did not appear to decrease the incidence of preoperative SMI in this study. Indeed, there was no significant reduction in SMI for any particular long-term treatment. It is important to note that the precise definition of hypertension appears to be irrelevant; poorly treated patients having a significantly greater incidence of SMI, independent of the precise level at which hypertension is denned. Also interesting is that an admission arterial pressure of ^170/100mmHg appeared to be as good a predictor of SMI as three or more diastolic readings ^ 100 mm Hg. This suggests that a single admission pressure reading may be used to define a population of patients who are at risk of preoperative SMI and therefore more likely to suffer adverse postoperative myocardial events. The positive predictive value of resistant hypertension (three or more diastolic readings ^100 mm Hg) for preoperative SMI was found to be 55 %, with a negative predictive value of 84 %. With increasing concern about the most appropriate methods of preoperative assessment in patients presenting for routine major surgery, we have shown that routine preoperative arterial pressure measurement may be a simple, cheap and effective method of identifying a group of patients in whom further more invasive cardiological investigations could be warranted.
577
578 23. Fleisher LA, Rosenbaum SH, Nelson AH, Barash PG. The predictive value of preoperative silent ischemia for postoperative ischemic cardiac events in vascular and nonvascular surgery patients. American Heart Journal 1991; 122: 980-986. 24. Mangano DT, Browner WS, Hollcnberg M, Li J, Tateo IM. Long-term cardiac prognosis following noncardiac surgery. The Study of Perioperative Ischemia Research Group. Journal of the American Medical Association 1992; 268: 233-239. 25. Mangano DT, Browner WS, Hollenberg M, London MJ, Tubau JF, Tateo IM. Association of perioperative myocardial ischemia with cardiac morbidity and mortality in men undergoing noncardiac surgery. The Study of Perioperative Ischemia Research Group. New England Journal of Medicine 1990; 323: 1781-1788. 26. Knight AA, Hollenberg M, London MJ, Tubau J, Verrier E, Browner W, Mangano DT. Perioperative myocardial ischemia: importance of the preoperative ischemic pattern. Anesthesiology 1988; 68: 681-688.
British Journal of Anaesthesia 27. McHugh P, Gill NP, Wyld R, Nimmo WS, Reilly CS. Continuous ambulatory ECG monitoring in the perioperative period: relationship of preoperative status and outcome. British Journal of Anaesthesia 1991; 66: 285-291. 28. Hollenberg M, Mangano DT, Browner WS, London MJ, Tubau JF, Tateo IM. Predictors of postoperative myocardial ischemia in patients undergoing noncardiac surgery. The study of Perioperative Ischemia Research Group. Journal of the American Medical Association 1992; 268: 205-209. 29. Stone JG, Foex P, Sear JW, Johnson LL, Khambatta HJ, Triner L. Risk of myocardial ischaemia during anaesthesia in treated and untreated hypertensive patients. British Journal of Anaesthesia 1988; 61: 675-679. 30. Goldman L, Caldera DL. Risks of general anesthesia and elective operation in the hypertensive patient. Anesthesiology 1979; 50: 285-292.
Downloaded from http://bja.oxfordjournals.org/ at University of Birmingham on September 2, 2015
CLINICAL INVESTIGATIONS
Resistant hypertension and preoperative silent myocardial ischaemia in surgical patientsf K. G. ALLMAN, A. MUIR, S. J. HOWELL, A. E. HEMMING, J. W. SEAR AND P. FOEX
Summary
Key w o r d s Heart, ischaemia. Complications, hypertension.
Perioperative cardiac morbidity is the leading cause of death after anaesthesia and surgery [1]. In the UK, between 12 and 20% of patients presenting for non-cardiac surgery have preoperative evidence of cardiovascular disease and of these the commonest cardiovascular complaint is arterial hypertension [2, 3]. Despite this, the role of arterial hypertension as a determinant of perioperative morbidity is still uncertain. Early work suggested that pre-existing hypertension was associated with intraoperative cardiovascular instability and myocardial ischaemia [4-7] and recently a history of arterial hypertension has been found to be associated with a poorer postoperative outcome [8, 9]. However, a prospective multivariate analysis of more than 1000 noncardiac operations by Goldman and colleagues failed to show arterial hypertension as an independent predictor of postoperative cardiac complications [10]. The use of ambulatory ECG monitoring is now a widely accepted method of detecting myocardial ischaemia. Modern microprocessing technology enables continuous and accurate measurement of ST segment changes; in patients with coronary artery disease, these have been shown to correlate with reduced myocardial perfusion [11, 12], such episodes
Patients and methods We studied 351 patients presenting for elective noncardiac surgery. Patients gave informed consent and the study was approved by the Central Oxford Research Ethics Committee. All patients were interviewed and examined by one of four investigators (K.G.A., A.M., A.E.H., SJ.H.) and details of cardiovascular risk factors noted. Preoperative investigations, ECG and arterial pressure readings were recorded. Admission arterial pressure was defined as the first in-hospital recording taken by one of the investigators, ward nurse or house physician. All pressure readings were obtained with a manual mercury sphygmomanometer. Patients subsequently underwent ambulatory ECG monitoring using a Medilog FD-2 or 6000 FD ambulatory ECG monitoring unit (Oxford Medical Limited, Abingdon, UK) for up to 24 h before surgery (minimum 8 h). Recordings were made from silver-silver chloride skin electrodes placed to record standard lead III and CMV5. Data from the ambulatory unit were then downloaded via optical link onto an IBM compatible 286 microcomputer and analysed for ST segment changes using the Medilog Prima ECG analysis software. A significant ST segment change was defined as a horizontal or downsloping depression of greater than 1 mm or an elevation of greater than 2 mm, for a period of more than 60 s. The ECG baseline was measured at a point 60 ms before the maximum R wave and the ST segment measured 100 ms after the maximum R K. G. ALLMAN, MB, CHB, FRCA, A. MUIR, MB, BS, MSC, FANZCA, S. J. HOWELL, MA, MHCP, FRCA, A. E. HEMMING, MB, BS, BSC, FRCA, J. W. SEAR, MA, BSC, PHD, FFARCS, P. Fo£x, DM, DPHIL, FRCA,
FANZCA, Nuffield Department of Anaesthetics, John Radcliffe Hospital, Headington, Oxford OX3 9DU. Accepted for publication: May 12, 1994. Correspondence to K.G.A. fPresentcd in part at the Anaesthetic Research Society, Leicester Meeting, July 10-11, 1992 (British Journal of Anaesthesia 1992; 69: 540P).
Downloaded from http://bja.oxfordjournals.org/ at University of Birmingham on September 2, 2015
We studied 325 patients undergoing elective noncardiac surgery who had preoperative ambulatory ECG monitoring performed for a duration of 5130 h (range 8-24 h; mean 15.8h). Sixty-four subjects (20%) had one or more episodes of ST segment depression consistent with myocardial ischaemia. Of all preoperative cardiovascular variables measured, the presence of elevated arterial pressure, despite patients being maintained on long term antihypertensive therapy, was the only factor associated significantly with the presence of preoperative silent myocardial ischaemia (P < 0.002). This correlation was confirmed when arterial hypertension was defined in four separate ways. The incidence of silent ischaemia in these patients was 33-55%. We suggest that admission arterial pressure may therefore be a useful screening test to identify patients at risk of preoperative myocardial ischaemia. (Br. J. Anaesth. 1994; 73: 574-578)
being clinically silent in more than 70 % of patients [13]. The presence of silent myocardial ischaemia in patients after myocardial infarction or with known stable coronary artery disease is predictive of adverse cardiac events and sudden death [14-17]. In the present study we have investigated the relationship between incidence of preoperative myocardial ischaemia and arterial hypertension in patients undergoing general or vascular surgery.
575
Preoperative silent ischaemia Table 1 Patient characteristics and numbers of patients with silent myocardial ischaemia
Male Vascular surgery Previous myocardial infarction History of angina Peripheral vascular disease Smoking (> 20 pack years) Obesity (> 20 % ideal weight) Diabetes mellitus Age > 70 yr
STATISTICAL ANALYSIS
Admission arterial pressures for each of the four categories were compared using unpaired two-tailed Student's t tests with correction for multiple testing. The adequacy of pressure control (as defined earlier) was compared with the presence or absence of preoperative silent myocardial ischaemia using the chi-square test or contingency table analysis with a continuity correction. A similar approach was used to assess the relationship between antihypertensive therapy and the incidence of silent myocardial ischaemia. The duration of ECG monitoring was divided into 4-h epochs and compared with the presence or absence of silent myocardial ischaemia and hypertensive-treatment category, again using contingency table analysis with a continuity correction. Indices of specificity, sensitivity and positive-
No. with silent ischaemia (%) (n = 64)
P
212 (65.2) 177 (54.5) 59 (18.2) 31 (9.5) 105 (32.3) 148 (45.5) 75 (23.2) 24 (7.4) 111(34.2)
41 (19.3) 39 (22.0) 14 (23.7) 9 (29.0) 25 (23.8) 30 (20.3) 12 (16.0) 4(16.7) 25 (22.5)
0.942 0.307 0.496 0.255 0.254 0.921 0.444 0.904 0.437
negative predictive value were calculated using standard formulae. Statistical analyses were performed using Statview SE and a Macintosh LC II microcomputer. Results We studied 351 patients (table 1). Forty-one per cent were receiving antihypertensive medication. Twenty-six subjects were subsequently excluded from analysis (recorder failure 14, ECG changes precluding ST segment analysis nine, patient refusal two, operation cancelled one). The remaining 325 patients had preoperative ambulatory ECG monitoring for a total duration of 5130 h (range 8—24 h; mean 15.8 h). Sixty-four (20.0%) patients had one or more episodes of ST segment depression consistent with myocardial ischaemia. No subject complained of angina or chest pain during the study. The mean duration of ischaemia was 88.3 (range 1-1078) min with a mean ischaemic load time of 338.3 (3.6-3529) s rr 1 . The number of patients in each of the four groups, based on classifications of arterial hypertension, is shown in table 2. Of all the preoperative variables tested, only hypertensive category was associated significantly with the presence or absence of silent myocardial ischaemia (tables 1,3). Whichever definition of arterial hypertension was used, the poorly treated group had a significantly greater incidence of preoperative silent myocardial ischaemia (33—55%) compared with the other groups (P = 0.0016-0.00O1). Using a definition of arterial hypertension of three or more diastolic readings 5* 100 mm Hg, 16 of 29 (55 %) poorly treated patients had episodes of silent myocardial ischaemia compared with 19 of 103 (18%) well treated patients, nine of 32 (28%) untreated patients and 20 of 161(12%) normotensive patients (P = 0.0001). There was no association between duration of ECG monitoring and the presence or absence of silent myocardial ischaemia (P = 0.159), or with any particular hypertensive-treatment category for any of the definitions of arterial hypertension (P = 0.108-0.310). When admission pressures were compared between groups there was no significant difference between the untreated and poorly treated groups (P = 0.69), but a clinically small, statistically significant difference between the normotensive and well treated groups (P = 0.0001) (table 4).
Downloaded from http://bja.oxfordjournals.org/ at University of Birmingham on September 2, 2015
wave. Each ST segment deviation was terminated by a return to baseline of more than 60 s. All reported episodes of ST segment change were verified visually. Inclusion criteria were age greater than 40 yr, ASA grade I—III and elective admission for major noncardiac surgery. All subjects had at least three preoperative arterial pressure measurements recorded, including admission arterial pressure. Patients were then classified into one of four groups: normotensive not taking antihypertensive medication (normotensive); normotensive taking regular antihypertensive medication (well treated); hypertensive not taking regular antihypertensive medication (untreated); and hypertensive despite regular antihypertensive medication (poorly treated). Each patient was classified on the basis of four separate definitions of arterial hypertension. This was done to reduce the introduction of bias from, and to examine die effect of, different hypertensive definitions. These were as follows: (1) adm $s 150/80 = either admission systolic 5s 150 or admission diastolic 3s 80 mm Hg; (2) adm 3s 160/90 = either admission systolic 3s 160 or admission diastolic 3* 90mmHg; (3) adm 3s 170/100 = either admission systolic 3s 170 or admission diastolic 3* lOOmmHg; and (4) diast. 3 3s 100 = three or more diastolic readings 5* 100 mm Hg. After ECG analysis, the ischaemic load (duration of myocardial ischaemia/duration of recording) was calculated for each patient.
No. (%) (n = 325)
British Journal of Anaesthesia
576
Table 2 Number of patients in each of the four classification groups of arterial hypertension (n = 325) (see text for definitions of groups)
Normotensive Well treated Untreated Poorly treated
Adm > 150/80 (No. (%))
Adm > 160/90 Adm > 170/100 (No. (%)) (No. (%))
Diast. 3 ^ 100 (No. (%))
112(34) 41(13) 81 (25) 91 (28)
141 (43) 70 (22) 52(16) 62 (19)
161 (49) 103 (32) 32 (10) 29(9)
162(50) 105 (32) 31 (10) 27(8)
Table 3 Number of patients with silent myocardial ischaemia in each of the four classification groups of arterial hypertension (n = 64) (see text for definitions of groups) Adm > 160/90 (No. (%))
Adm> 170/100 (No. (%))
Diast. 3 > 100 (No. (%))
14 of 112 (13) 5 of 41 (12) 15 of 81 (19) 30 of 91 (33) P = 0.0016
17 of 141 (12) 12 of 70(17) 12 of 52 (23) 23 of 62 (37) P = 0.0005
22 of 162 (14) 21 of 105 (20) 7 of 31 (23) 14 of 27 (52) P = 0.0001
20 of 161 (12) 19 of 103 (18) 9 of 32 (28) 16 of 29(55) P = 0.0001
Table 4 Admission arterial pressure for each of the four groups classified on the basis of three or more diastolic readings Js 100 mm Hg (mean (SD)). *P < 0.0001 normotensive vs well treated group
Normotensive Well treated Untreated Poorly treated
Mean systolic pressure (mm Hg)
Mean diastolic pressure (mm Hg)
141 (21) 153(22) 181(22) 178(23)
80(11) 86 (9.2) 107(12) 106(12)
Table 5 Number of patients taking each antihypertensive treatment (n = 132). (Data analysed for each drug treatment using chi-square test with a continuity correction for small numbers)
ACE inhibitors Beta blockers Calcium antagonists Diuretics
No. (%)
No. with silent ischaemia (%)
25(19) 55 (42) 52 (39) 63 (48)
6(24) 14 (25) 16 (31) 20 (32)
0.948 0.973 0.490 0.270
For poorly treated patients classified as having three or more diastolic readings ^100 mm Hg, the sensitivity and specificity for silent myocardial ischaemia were 25 and 95%, respectively, with a positive predictive value of 55 % and a negative predictive value of 84 %. When the poorly treated and well treated groups were analysed further according to treatment, no single therapy was found to be associated with the presence of, or was protective against, silent myocardial ischaemia (table 5). Discussion The presence of silent myocardial ischaemia (SMI) has been shown to be a significant predictor of postoperative adverse myocardial events. Slogoff and Keats found silent ischaemia occurring in the anaesthetic room to be predictive of postoperative myocardial infarction for patients undergoing car-
diac surgery [7, 18]. Raby and colleagues [19-21], in a series of 176 patients, and Pasternack and coworkers [22], in a series of 200 patients undergoing vascular surgery, found preoperative SMI to be the most significant predictor of postoperative morbid cardiac events. Similar results were described for those undergoing both vascular and non-vascular surgery [23]. However, Mangano and colleagues found preoperative SMI to be weakly associated only with, and not a predictor of, postoperative cardiac events and death [24, 25]. Other studies have shown preoperative SMI to be predictive of intraoperative myocardial ischaemia [26] and postoperative SMI [27, 28]. Detection of SMI therefore has important implications in the management of patients presenting for major surgery. We have shown that patients who appear to be resistant to conventional antihypertensive therapies have an extremely high incidence (> 50%) of preoperative SMI on 24-h ambulatory ECG monitoring. The mechanism for this remains unclear. It is likely that patients with long-standing arterial hypertension who do not respond well to treatment have more severe and widespread arterial disease. Using a definition of hypertension as three diastolic readings ^ 100 mm Hg, 55% of poorly treated patients had preoperative episodes of SMI compared with 18% of well treated patients and 28% of untreated patients. It is important to note that the poorly treated group, although resistant to treatment, did not have grossly elevated arterial pressure readings. There was no significant difference between the admission pressures of the poorly treated and untreated groups. As our population consisted of patients presenting for routine major surgery and presently accepted practice recommends postponement if diastolic pressure is consistently greater than 110-115 mm Hg [5], it is not surprising that we did not find a large incidence of treatment-resistant patients. At present the importance of "white coat" hypertension on admission to hospital remains unclear. In this study, there was no significant difference in the rates of SMI among the four categories, regardless of whether hypertension
Downloaded from http://bja.oxfordjournals.org/ at University of Birmingham on September 2, 2015
Normotensive Well treated Untreated Poorly treated
Adm > 150/80 (No. (%))
Preoperative silent ischaemia
Acknowledgements This study was supported by the Wellcome Trust (K.G.A., A.E.H., S.J.H.), the trustees of the WG Ceizar Scholarship, University of Tasmania (A.M.), the British Heart Foundation and the Medical Research Fund, University of Oxford.
References 1. Mangano DT. Perioperative cardiac morbidity. Anesthesiology 1990; 72: 153-184.
2. Edwards ND, Reilly CS. Detection of perioperarive myocardial ischaemia. British Journal of Anaesthesia 1994; 72: 104-115. 3. Forrest JB, Rehder K, Cahalan MK, Goldsmith CH. Multicenter study of general anesthesia: Predictors of severe perioperative adverse outcomes. Anesthesiology 1992; 76: 3-15. 4. Prys-Roberts C, Meloche R, Foex P. Studies of anaesthesia in relation to hypertension. 1: Cardiovascular responses of treated and untreated patients. British Journal of Anaesthesia 1971; 43: 122-137. 5. Prys-Roberts C. Hypertension and anesthesia—fifty years on. Anesthesiology 1979; 50: 281-284. 6. Stone JG, Foex P, Sear JW, Johnson LL, Khambatta HJ, Triner L. Myocardial ischemia in untreated hypertensive patients: effect of a single small oral dose of a beta-adrcnergic blocking agent. Anesthesiology 1988; 68: 495-500. 7. Slogoff S, Keats AS. Does perioperarive myocardial ischemia lead to postoperative myocardial infarction? Anesthesiology 1985; 62: 107-114. 8. Browner WS, Li J, Mangano DT. In-hospital and long-term mortality in male veterans following noncardiac surgery. The Study of Perioperative Ischemia Research Group. Journal of the American Medical Association 1992; 268: 228-232. 9. Fletcher AE, Bradley IC, Broxton JS, Bulpitt CJ, Davis AJ, Dyker GS, Ganvir PL, Gostick NK, Mayhew SR, Mukerji D. Survival of hypertensive subjects identified on screening: results for sustained and unsustained diastolic hypertension. European Heart Journal 1992; 13: 1595-1601. 10. Goldman L, Caldera DL, Nussbaum SR, Southwick FS, Krogstad D, Murray B, Burke DS, O'Malley TA, Goroll A, Caplan CH, Nolan J, Carabello B, Slater EE. Multifactorial index of cardiac risk in noncardiac surgical procedures. New England Journal of Medicine 1977; 297: 845-850. 11. Wilson RF, Marcus ML, Christensen BV, Talman C, White CW. Accuracy of exercise electrocardiography in detecting physiologically significant coronary arterial lesions. Circulation 1991; 83: 412^121. 12. Boden WE, Spodick DH. Diagnostic significance of precordial ST-segment depression. American Journal of Cardiology 1989; 63: 358-361. 13. Epstein SE, Quyyumi AA, Bonow RO. Myocardial ischemia—silent or symptomatic. New England Journal of Medicine 1988; 318: 1038-1043. 14. Kennedy HL, Wiens RD. Ambulatory (Holter) electrocardiography and myocardial ischemia. American Heart Journal 1989; 117: 164-176. 15. Sheps DS, Heiss G. Sudden death and silent myocardial ischemia. American Heart Journal 1989; 117: 177-184. 16. Rocco MB, Nabel EG, Campbell S, Goldman L, Barry J, Mead K, Selwyn AP. Prognostic importance of myocardial ischemia detected by ambulatory monitoring in patients with stable coronary artery disease. Circulation 1988; 78: 877-884. 17. Tzivoni D, Gavish A, Zin D, Gottlieb S, Moriel M, Keren A, Banai S, Stern S. Prognostic significance of ischemic episodes in patients with previous myocardial infarction. American Journal of Cardiology 1988; 62: 661-664. 18. Slogoff S, Keats AS. Further observations on perioperative myocardial ischemia. Anesthesiology 1986; 65: 539—542. 19. Raby KE, Barry J, Creager MA, Cook EF, Weisberg MC, Goldman L. Detection and significance of intraoperative and postoperative myocardial ischemia in peripheral vascular surgery. Journal of the American Medical Association 1992; 268: 222-227. 20. Raby KE, Goldman L, Cook EF, Rumerman J, Barry J, Creager MA, Selwyn AP. Long-term prognosis of myocardial ischemia detected by Holtcr monitoring in peripheral vascular disease. American Journal of Cardiology 1990; 66: 1309-1313. 21. Raby KE, Goldman L, Creager MA, Cook EF, Weisberg MC, Whittemore AD, Selwyn AP. Correlation between preoperative ischemia and major cardiac events after peripheral vascular surgery. New England Journal of Medicine 1989; 321: 1296-1300. 22. Pastemack PF, Grossi EA, Baumann FG, Riles TS, Lamparello PJ, Giangola G, Primis LK, Mintzer R, Imparato AM. The value of silent myocardial ischemia monitoring in the prediction of perioperative myocardial infarction in patients undergoing peripheral vascular surgery. Journal of Vascular Surgery 1989; 10: 617-625.
Downloaded from http://bja.oxfordjournals.org/ at University of Birmingham on September 2, 2015
was defined as an admission pressure of > 170/100 mm Hg or as three diastolic readings ^ 100 mm Hg. This would suggest that in this study at least, "white coat" hypertension had little or no measurable effect. The problem of preoperative arterial hypertension is a common one. In a prospective study of more than 17000 patients presenting for routine general anaesthesia and surgery, Forrest and co-workers found a prevalence rate of 13.6% [3]. Arterial hypertension itself has been shown in several studies to be an independent predictor of perioperative SMI [7, 27, 29] and also of postoperative morbidity and mortality [8, 21]. This is the first time, however, that patients resistant to antihypertensive medication have been shown to have such a high incidence of preoperative SMI, and indeed it may be that this group is responsible for the increased morbidity associated with a history of arterial hypertension. Previous studies have found antihypertensive treatment to be protective against peroperative arterial pressure lability and associated ischaemia [4, 30]. Stone and colleagues found that preoperative treatment with beta blocking agents reduced the incidence of intraoperative myocardial ischaemia in mild to moderate untreated hypertensive patients [6]. However, beta blockers did not appear to decrease the incidence of preoperative SMI in this study. Indeed, there was no significant reduction in SMI for any particular long-term treatment. It is important to note that the precise definition of hypertension appears to be irrelevant; poorly treated patients having a significantly greater incidence of SMI, independent of the precise level at which hypertension is denned. Also interesting is that an admission arterial pressure of ^170/100mmHg appeared to be as good a predictor of SMI as three or more diastolic readings ^ 100 mm Hg. This suggests that a single admission pressure reading may be used to define a population of patients who are at risk of preoperative SMI and therefore more likely to suffer adverse postoperative myocardial events. The positive predictive value of resistant hypertension (three or more diastolic readings ^100 mm Hg) for preoperative SMI was found to be 55 %, with a negative predictive value of 84 %. With increasing concern about the most appropriate methods of preoperative assessment in patients presenting for routine major surgery, we have shown that routine preoperative arterial pressure measurement may be a simple, cheap and effective method of identifying a group of patients in whom further more invasive cardiological investigations could be warranted.
577
578 23. Fleisher LA, Rosenbaum SH, Nelson AH, Barash PG. The predictive value of preoperative silent ischemia for postoperative ischemic cardiac events in vascular and nonvascular surgery patients. American Heart Journal 1991; 122: 980-986. 24. Mangano DT, Browner WS, Hollcnberg M, Li J, Tateo IM. Long-term cardiac prognosis following noncardiac surgery. The Study of Perioperative Ischemia Research Group. Journal of the American Medical Association 1992; 268: 233-239. 25. Mangano DT, Browner WS, Hollenberg M, London MJ, Tubau JF, Tateo IM. Association of perioperative myocardial ischemia with cardiac morbidity and mortality in men undergoing noncardiac surgery. The Study of Perioperative Ischemia Research Group. New England Journal of Medicine 1990; 323: 1781-1788. 26. Knight AA, Hollenberg M, London MJ, Tubau J, Verrier E, Browner W, Mangano DT. Perioperative myocardial ischemia: importance of the preoperative ischemic pattern. Anesthesiology 1988; 68: 681-688.
British Journal of Anaesthesia 27. McHugh P, Gill NP, Wyld R, Nimmo WS, Reilly CS. Continuous ambulatory ECG monitoring in the perioperative period: relationship of preoperative status and outcome. British Journal of Anaesthesia 1991; 66: 285-291. 28. Hollenberg M, Mangano DT, Browner WS, London MJ, Tubau JF, Tateo IM. Predictors of postoperative myocardial ischemia in patients undergoing noncardiac surgery. The study of Perioperative Ischemia Research Group. Journal of the American Medical Association 1992; 268: 205-209. 29. Stone JG, Foex P, Sear JW, Johnson LL, Khambatta HJ, Triner L. Risk of myocardial ischaemia during anaesthesia in treated and untreated hypertensive patients. British Journal of Anaesthesia 1988; 61: 675-679. 30. Goldman L, Caldera DL. Risks of general anesthesia and elective operation in the hypertensive patient. Anesthesiology 1979; 50: 285-292.
Downloaded from http://bja.oxfordjournals.org/ at University of Birmingham on September 2, 2015