- Email: [email protected]
Propofol Use in the Elderly Population: Prevalence of Overdose and Association With 30-Day Mortality
Clinical Therapeutics/Volume 37, Number 12, 2015
Original Research
Propofol Use in the Elderly Population: Prevalence of Overdose and Association With 30-Day Mortality Adam T. Phillips, MD1; Stacie Deiner, MD2,3,4; Hung Mo Lin, ScD5; Evie Andreopoulos, MA5; Jeffrey Silverstein, MD2,4,5; and Matthew A. Levin, MD2 1
Department of Internal Medicine, School of Medicine, Yale University, New Haven, Connecticut; Departments of 2Anesthesiology; 3Neurosurgery; 4Geriatrics and Palliative Care; and 5Health Evidence and Policy, Icahn School of Medicine at Mount Sinai, New York, New York
ABSTRACT Purpose: Geriatric patients are more sensitive to the anesthetic effects of propofol and its adverse effects, such as hypotension, than is the general population; thus, a reduced dose (1–1.5 mg/kg) is recommended for the induction of anesthesia. The extent to which clinicians follow established dosing guidelines has not been well described. Therefore, we investigated the prevalence of propofol overdose in the elderly population to determine whether propofol overdose occurs and is associated with increased hypotension and 30-day mortality. Methods: In this retrospective study in patients who received propofol for the induction of general anesthesia, data on demographic characteristics, preoperative medications, intraoperative management, and 30-day mortality were collected. The dose of propofol used for the induction of anesthesia and the median blood pressure in the pre- and immediate postinduction periods were determined. Hypotension was defined as either: (1) a decrease in mean arterial pressure (MAP) of 440% concurrent with a MAP of o70 mm Hg; or (2) a MAP of o60 mm Hg. Findings: A total of 17,540 patients were included in the analysis; 4033 (23.0%) were aged 465 years. The median (interquartile range) propofol dose in the group aged 465 years was 1.8 (1.4–2.2) mg/kg, above the recommended dose, in comparison to 2.2 (1.9–2.5) mg/kg in younger patients. On multivariate
Accepted for publication October 6, 2015. http://dx.doi.org/10.1016/j.clinthera.2015.10.005 0149-2918/$ - see front matter & 2015 Elsevier HS Journals, Inc. All rights reserved.
2676
analysis, increased propofol dose was associated with increased postinduction hypotension, especially in patients over 70 years of age, but not 30-day mortality. Implications: Older patients received greater-thanrecommended doses of propofol for induction, which may have led to significant dose-dependent hypotension. Despite this finding, the dose of propofol for induction was not independently associated with a greater 30-day mortality rate. More education regarding geriatric concerns is needed for encouraging anesthesiologists to tailor the plan for anesthesia in geriatric patients. However, overall postsurgical mortality is a function of preoperative risk and type surgical procedure. (Clin Ther. 2015;37:2676–2685) & 2015 Elsevier HS Journals, Inc. All rights reserved. Key words: 30-day mortality, elderly, propofol overdose, hypotension.
INTRODUCTION Elderly patients constitute a large proportion of the surgical population in the United States, and anesthesiologists are often faced with the need for tailoring the anesthesia technique to account for geriatric physiology.1 The anesthesiologist may change the selection of medications or modify dosing in an attempt to optimize postoperative outcomes. However, geriatric-specific guidelines on anesthesia do not exist, and an anesthesiologist Scan the QR Code with your phone to obtain FREE ACCESS to the articles featured in the Clinical Therapeutics topical updates or text GS2C65 to 64842. To scan QR Codes your phone must have a QR Code reader installed.
Volume 37 Number 12
A.T. Phillips et al. may modify the technique based on a nonscientific impression of what is best for the patient. The use of propofol in the elderly population is one specific area of equipoise. Propofol is the medication most commonly used for inducing anesthesia in surgical patients in the United States. However, a common adverse effect of propofol-based induction is dose-dependent hypotension. Propofol-related hypotension increases with dose and results from a decrease in systemic vascular resistance and, to a lesser extent, myocardial depression.2–7 Elderly patients in particular have an increased sensitivity to this effect, leading to the manufacturersʼ recommendation that the dose of propofol used for induction be decreased in elderly patients from a recommended dose of 2 to 2.5 mg/kg, to 1 to 1.5 mg/kg.8–14 Reich et al15 reported that postinduction hypotension was associated with increased mortality despite controlling for the effects of age, preexisting hypotension, and American Society of Anesthesiologists (ASA)–defined physical status. Those investigators recommended that propofol induction be avoided in the elderly population, especially in patients with baseline low blood pressure. Nonetheless, many anesthesiologists continue to use propofol in the elderly population instead of induction drugs that cause less hypotension, such as etomidate. There is a distinct impression that propofol is well tolerated; however, it is unclear whether anesthesiologists employ reduced doses and whether reduced propofol doses can be used for avoiding hypotension. To address these areas of equipoise, we performed a retrospective review of data from a large clinical cohort of patients aged 19 to 89 years who received propofol for the induction of general anesthesia. Our hypothesis was that elderly patients often receive propofol in excess of 1.5 mg/ kg (the recommended dose for age 465 years) and that this relative overdose is associated with increased postinduction hypotension and increased 30-day mortality.
PATIENTS AND METHODS The study protocol was approved by the Program for the Protection of Human Subjects, Icahn School of Medicine at Mount Sinai (New York, New York), and the requirement for written informed consent was waived pursuant to 45 CFR 46.101(i). Data from all inpatients and day-of-admission surgical patients aged 19 to 89 years who had received a general anesthetic with induction using intravenous propofol between January
December 2015
2006 and December 2009 were included, with the exception of those undergoing cardiac or obstetric surgery (eg, cesarean section). In patients who underwent multiple surgeries in the time period studied, only data from the most recent surgery were included in the analysis. Primary outcomes were hypotension in the 10 minutes after induction and 30-day mortality. Perioperative demographic, medication, and physiologic data were extracted from the anesthesia information management system (AIMS) (CompuRecord; Phillips Medical Systems, Andover, Massachusetts). Classification of preoperative antihypertensive medication was based on the major classes of antihypertensive medications recommended by the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure for the treatment of hypertension.16 Additional data on demographic characteristics and mortality were obtained from the institutional data warehouse. We also utilized the Social Security Administration Death Master File (National Technical Information Service, Alexandria, Virginia) as an additional source of data on mortality. As a measure of coexisting illness, we used the both the ASA physical status score and the All Patient Refined–Diagnosis Related Group risk for mortality (APR-DRG ROM) scale score (extreme, major, moderate, and minor). ASA scores range from 1 (healthy with no comorbidities) to 5 (not expected to survive without the operation).17 APR-DRG is a classification scheme developed by 3M (Maplewood, Minnesota) based on nationwide samples and used by Medicare to estimate resource consumption by patients.18 ROM score is calculated based on primary and secondary discharge diagnoses and thus captures all comorbidities present during an admission. Although ASA and ROM are correlated in many conditions, they sometimes differ because the ASA score is a qualitative score assigned by the anesthesiologist, whereas ROM is determined based on administrative coding data. APR-DRG ROM has been reported to correlate well with mortality among patients in the intensive care unit and to be a suitable tool for risk adjustment of 30-day mortality in patients with acute myocardial infarction.19–21
Determination of Dose of Propofol Used for Induction and Co-Induction Agents At the Icahn School of Medicine, propofol is typically injected manually, usually as a single bolus,
2677
Clinical Therapeutics for the purpose of induction. Occasionally a provider may titrate the induction using divided doses. Induction dose was therefore measured as the total dose of propofol given in the period beginning 5 minutes before and ending 2 minutes after the induction timestamp as recorded in the AIMS. Total dose was then divided by the patientʼs weight as recorded in the AIMS to give an induction dose by weight (in mg/kg).
(in liters), transfusion of packed red blood cells (dichotomous variable), and transfusion of platelets or fresh frozen plasma (dichotomous variable). To determine whether the effects of propofol dose on both outcomes differed depending on the preoperative risk profile, we tested the 2-way interactions between propofol and age and ASA physical status, respectively.
Determination of Baseline and Postinduction Blood Pressure
RESULTS
Data on blood pressure were electronically captured by the AIMS as often as every 15 seconds for an intra-arterial line and every 1 to 5 minutes for oscillometric noninvasive blood pressure cuff. We report baseline (10 minutes preinduction) and 10 minutes postinduction as the median mean arterial pressure (MAP) of each period. Medians were used for filtering out transient changes and artefact.22 Postinduction hypotension was defined, as per Reich et al,15 as either: (1) a decrease in MAP of 440% from baseline and a MAP of o70 mm Hg; or (2) a MAP of o60 mm Hg.
Statistical Analysis Patientsʼ characteristics are described as numbers (%), medians (interquartile ranges) or means (SD). For comparisons between groups, χ2 tests were used for categorical variables, and 2-sample t tests were used for continuous variables, as appropriate. Forward stepwise logistic regression analyses (with significance levels for both the stay and entry criteria set at 0.05) were performed for the identification of significant predictors of postinduction hypotension and 30-day mortality, respectively. For the outcome of postinduction hypotension, the covariates considered in the initial models included the dose of propofol used for induction, age, sex, weight, ASA physical status, APRDRG ROM classification, emergency status, baseline MAP, use of antihypertensive medication, class of antihypertensive used, coadministration of fentanyl or midazolam during induction, and type of surgical procedure. A dichotomous variable was used for indicating whether the attending anesthesiologist was working alone. In addition, we considered the case volume of the attending anesthesiologist during the study period. For the 30-day mortality model, the following additional intraoperative covariates were added: duration of anesthesia (in minutes), total amount of crystalloid given
2678
A total of 17,810 patients met the inclusion criteria, of whom 17,540 patients (98.5%) who received propofol had complete covariate data available. The demographic characteristics of the cohort are shown in Table I. Approximately half the population was female, and 4033 (23.0%) were aged 465 years. Most patients were either ASA status 1–2 (58.2%) or 3 (37.4%). The type of surgery most commonly reported was general surgery, followed by orthopedic and gynecologic surgery. Overall, the surgeries were relatively long; the median (interquartile range) duration of anesthesia was 192 (139–270) minutes. The median propofol dose in the overall cohort was 2.1 (1.8–2.5) mg/kg, with 2.2 (1.9–2.5) mg/kg in the group aged r65 years versus 1.8 (1.4–2.2) mg/kg in the group aged 465 years. Figure 1 demonstrates that the median dose of propofol used for induction decreased with age; however, even at 80 years of age, the median dose was still greater than the recommended dose of 1 to 1.5 mg/kg. The dose administered was also highly variable at all ages.
Results of Univariate Analysis Table I also shows the results of the univariate analysis of postinduction hypotension and 30-day mortality. Postinduction hypotension occurred in 5.96% of all patients. The hypotensive patients tended to be older (median, 60 vs 53 years; P o 0.0001), have a higher ASA physical status, and were more likely to have been taking a β-blocker before surgery. The 30day mortality rate was 0.89%. Patients who died were older, male, had greater preoperative APR-DRG ROM scores and ASA scores, were more likely to have been on preoperative β-blockers, and were more likely to have undergone a neurosurgical procedure.
Results of Multivariate Analysis In the multivariate model, APR-DRG ROM score, emergency surgery, otolaryngologic procedure, dose of
Volume 37 Number 12
December 2015
Table I. Patients’ characteristics, overall and by study outcome (univariate analysis). Hypotension Within First 10 Min Parameter Age Median (IQR), y 465 y, no. (%) Female, no. (%) Weight, median (IQR), kg ASA physical status,* no. (%) 1 or 2 3 4 or 5 APR-DRG risk for mortality, no. (%) Minor Moderate Major Extreme Emergency case, no. (%) Antihypertensive medication, no. (%) β-Blocker Other None Baseline MAP, median (IQR), mm Hg
2679
Group, no. (%) 0–1 h 41–2 h
Yes (n ¼ 1047)
No (n ¼ 16,493)
53 (41–64) 4033 (23.0) 10,005 (57.0) 75.0 (63.0–88.1)
60 (49–69) 360 (34.4) 650 (62.1) 70.0 (59.1–84.1)
53 (49–64) 3673 (22.3) 9355 (56.7) 75.0 (59.1–88.6)
10,213 (58.2) 6562 (37.4) 765 (4.4)
488 (46.6) 476 (45.5) 83 (7.9)
9725 (59.0) 6086 (36.9) 682 (4.1)
12,914 (73.6) 3072 (17.5) 1111 (6.3) 443 (2.5) 1869 (10.7)
658 (62.8) 220 (21.0) 100 (9.6) 69 (6.6) 144 (13.8)
12,256 (74.3) 2852 (17.3) 1011 (6.1) 374 (2.3) 1725 (10.5)
2372 (13.5) 2828 (16.1) 12,340 (70.4) 103 (93–114)
178 (17.0) 194 (18.5) 675 (64.5) 102.9 (87.4–114)
2194 (13.3) 2634 (16.0) 11,665 (70.7) 103.0 (87.4–114)
192 (139–270)
187.0 (132–265)
192.0 (132–270)
245 (1.4) 2799 (16.0)
16 (1.5) 192 (18.3)
229 (1.4) 2607 (15.8)
P o0.001 o0.001 o0.001 o0.001
Yes (n ¼ 156)
No (Survivors) (n ¼ 17,384)
63 (53–74) 69 (44.2) 66 (42.3) 72.7 (62.6–88.1)
53 (53–64) 3964 (22.8) 9939 (57.2) 75.0 (62.6–88.1)
11 (7.1) 67 (42.9) 78 (50.0)
10,202 (58.7) 6495 (37.4) 687 (4.0)
5 (3.2) 4 (2.6) 36 (23.1) 111 (71.2) 37 (23.7)
12,909 (74.3) 3068 (17.6) 1075 (6.2) 332 (1.9) 1832 (10.5)
41 (26.3) 25 (16.0) 90 (57.7) 101.0 (88.7–113)
2331 (13.4) 2803 (16.1) 12,250 (70.5) 103.0 (88.7–114)
180.5 (103–296)
192.0 (103–270)
4 (2.6) 41 (26.3)
241 (1.4) 2758 (15.9)
o0.001
o0.001
0.12
o0.001 o0.001 o0.001 0.42 o0.001
o0.001
o0.001 o0.001
P
o0.001
o0.001 o0.001
0.12
0.196 o0.0001
0.07
(continued)
A.T. Phillips et al.
Duration of anesthesia Median (IQR), min
All Patients (N ¼ 17,540)
30-D Mortality
Hypotension Within First 10 Min Parameter 42–4 h 44–6 h 46 h Surgery type, no. (%) General Otolaryngologic Gynecologic Urologic Neurologic Spinal Vascular Transplantation Endoscopic Other Crystalloid given, no. (%) 0–1 L 41–3 L 43–5 L 45 L Not documented Propofol dose, median (IQR), mg/kg Volume 37 Number 12
Transfusion, no. (%) RBCs Fresh frozen plasma Platelets
All Patients (N ¼ 17,540)
Yes (n ¼ 1047)
No (n ¼ 16,493)
8803 (50.2) 3885 (22.1) 1808 (10.3)
519 (49.6) 203 (19.4) 117 (11.2)
8284 (50.2) 3682 (22.3) 1691 (10.3)
5998 (34.2) 2176 (12.4) 1973 (11.2) 1432 (8.2) 1318 (7.5) 1226 (7.0) 460 (2.6) 425 (2.4) 119 (0.7) 2413 (13.8)
326 (31.1) 219 (20.9) 73 (7.0) 55 (5.3) 82 (7.8) 77 (7.4) 28 (2.7) 21 (2.0) 10 (1.0) 156 (14.9)
5672 (34.4) 1957 (11.9) 1900 (11.5) 1377 (8.4) 1236 (7.5) 1149 (7.0) 432 (2.6) 404 (2.5) 109 (0.7) 2257 (13.7)
P
30-D Mortality Yes (n ¼ 156)
No (Survivors) (n ¼ 17,384)
56 (35.9) 29 (18.6) 26 (16.7)
8747 (50.3) 3856 (22.2) 1782 (10.3)
43 (27.6) 24 (15.4) 8 (5.1) 8 (5.1) 23 (14.7) 8 (5.1) 9 (5.8) 6 (3.9) 11 (7.1) 16 (10.3)
5955 (34.3) 2152 (12.4) 1965 (11.3) 1424 (8.2) 1295 (7.5) 1218 (7.0) 451 (2.6) 419 (2.4) 108 (0.6) 2397 (13.8)
o0.001
o0.001
o0.001
0.97 5504 (31.4) 8248 (47.0) 2402 (13.7) 685 (3.9) 701 (4.0) 2.1 (1.8–2.5)
324 (30.9) 496 (47.4) 148 (14.1) 38 (3.6) 41 (3.9) 2.08 (1.7–2.5)
5180 (31.4) 7752 (47.0) 2254 (13.7) 647 (3.9) 660 (4.0) 2.11 (1.7–2.5)
1087 (6.2) 187 (1.1) 156 (0.9)
99 (9.5) 14 (1.3) 7 (0.7)
988 (6.0) 173 (1.1) 149 (0.9)
0.10
o0.001 0.38 0.43
P
57 (36.5) 52 (33.3) 21 (13.5) 11 (7.1) 15 (9.6) 1.59 (1.1–2.1)
5447 (31.3) 8196 (47.1) 2381 (13.7) 674 (3.9) 686 (4.0) 2.11 (1.1–2.5)
46 (29.5) 20 (12.8) 18 (11.5)
1041 (6.0) 167 (1.0) 138 (0.8)
o0.001
o0.001 o0.001 o0.001
APR-DRG ¼ All Patient Refined–Diagnosis Related Group; ASA ¼ American Society of Anesthesiologists physical status scale; IQR ¼ interquartile range; MAP ¼ mean arterial pressure; RBC, red blood cell. * Score range: 1 (healthy with no comorbidities) to 5 (not expected to survive without the operation).
Clinical Therapeutics
2680
Table I. (continued).
A.T. Phillips et al.
Propofol Dose (mg/kg)
6
4
2
0 20
40
60
80
Age (yr)
Figure 1. Scatterplot showing the relationship between age and propofol-induction dose (mg/kg). Dose of propofol decreases as age increases, but not as much as published guidelines would recommend. The dark line is a smoothed curve generated using the penalized B-spline method (PROC SGCATTER, SAS; SAS Institute Inc, Cary, North Carolina).
propofol used for induction, and age were all strong predictors of postinduction hypotension (Table II). The interaction between propofol and age was significant. Figure 2 demonstrates that for each decade of increasing age 460 years, a given dose of propofol was associated with a progressively greater risk for hypotension. After adjustment for covariates, propofol dose was not significantly associated with 30-day mortality (Table II).
DISCUSSION In this large, retrospective study, we found that elderly patients commonly receive greater-than-recommended doses of propofol for the induction of general anesthesia. To our knowledge, this result has not been previously reported. Greater doses were associated with an increased prevalence of significant hypotension in the postinduction period, consistent with earlier reports.9 An increased dose of propofol used for induction, however, was not associated with increased 30-day mortality. Instead, mortality was more strongly associated with the inherent risk of the procedure and patient characteristics.
December 2015
Our results conflict with those from Reich et al,15 who reported that hypotension was not dose dependent. However, the findings from both studies agree that death is more common in patients with postinduction hypotension. Our findings suggest that large doses of propofol cause hypotension but also that, overall, there is not a “failure to rescue.” Other important differences between the 2 studies are that Reich et al did not examine or interpret the interaction of age and propofol dosing and that they treated age as a dichotomous variable, in which patients aged 450 years were classified as “old.” We chose to look at age in a more nuanced fashion and report that as age increased by decade, the dose-dependent effect of propofol-induced hypotension became increasingly dramatic (Figure 2). This finding suggests that even the recommended dosing of 1-1.5 mg/kg should be further reduced for patients over the age of 70-80. In the multivariate model adjusted for preoperative risk and type of surgery, the initial propofol dose was not associated with 30-day mortality. Whether persistent overdose of anesthetic agent occurs with propofol infusion, whether there is a failure to rescue with “pressors,” and whether doses of inhalational anesthetic agents are not being adjusted for age were not addressed in this study. However, Sessler et al23 suggested that the absolute dose of anesthetic is less important to postsurgical outcomes than is the larger picture of whether the patient has a susceptible phenotype of both sensitivity to anesthetic agents and persistent low blood pressure throughout the case. Regarding the contribution of intraoperative course to postoperative mortality, we did examine the role of transfusion as a surrogate marker for intraoperative misadventure. In our study, the need for red blood cell transfusion was not associated with hypotension, but the need for platelet transfusion was. This finding may have been due to the fact that platelets are generally given in the setting of large volume resuscitation. Whether patients who required platelet transfusion also had intraoperative hypotension after the first 10 minutes could not be determined from this dataset. We did not include the duration of hypotension in the 10-minute window after induction in the multivariate model of 30-day mortality. The duration of hypotension may not add information because it was likely collinear with other variables in the model, including propofol dose and APR-DRG
2681
Clinical Therapeutics
Table II. Final multivariate logistic models for hypotension and 30-day mortality. Hypotension Parameter Propofol (per mg/kg)* Female vs male Weight (per 5 kg) APR-DRG risk for mortality Moderate Major Extreme ASA physical status‡ 1 or 2 (Ref) 3 4þ Baseline MAP (per 5 mm Hg) Emergency case Surgical indication General (Ref) Otolaryngologic Spinal Neurologic Endoscopic Vascular Urologic Transplantation Gynecologic Other Need platelet transfusion
OR
95% CI
30-D Mortality P
OR
95% CI 0.690–1.182
0.457
†
P
1.25 0.93
1.08–1.45 0.89–0.97
0.003 o0.001
0.90 NS‡ NS
1.10 1.28 2.31
0.92–1.30 1.00–1.64 1.70–3.16
0.304 0.051 o0.001
2.55 51.0 440
0.66–9.76 18.6–139 164–4999
0.173 o0.001 o0.001
1.24 1.42 0.910 1.46
1.07–1.45 1.06–1.92 0.891–0.929 1.20–1.78
0.005 0.020 o0.001 o0.001
1.69 4.64 NS NS
0.83–3.44 2.18–9.87
0.152 o0.001
2.00 1.30 1.11 1.04 0.81 0.80 0.76 0.70 1.16 NA
1.66–2.45 1.00–1.69 0.86–1.44 0.53–2.05 0.54–1.22 0.59–1.08 0.48–1.22 0.53–0.91 0.95–1.42
o0.001 0.048 0.407 0.912 0.317 0.147 0.259 0.008 0.148
0.95 1.32 1.02 2.12 0.45 1.16 0.23 1.15 0.60 3.03
0.52–1.72 0.54–3.24 0.56–1.86 0.92–4.90 0.20–1.03 0.49–2.74 0.09–0.60 0.49–2.69 0.32–1.13 1.55–5.92
0.853 0.545 0.950 0.078 0.057 0.741 0.002 0.749 0.112 0.001
APR-DRG ¼ All Patient Refined–Diagnosis Related Group; ASA ¼ American Society of Anesthesiologists physical status; MAP ¼ mean arterial pressure; NA ¼ not applicable because the variable was not considered in this initial model selection; NS ¼ not significant after forward stepwise selection; OR ¼ odds ratio. * Propofol was forced into the model. † The age and propofol interaction is significant (P ¼ 0.016). The joint effects of propofol and age on hypotension in the first 10 minutes after induction are illustrated in Figure 2. ‡ Score range: 1 (healthy with no comorbidities) to 5 (not expected to survive without the operation).
classification. Overall, the findings from our study support that a preoperative risk for mortality and comorbidity are much stronger predictors than is either the use of propofol for induction or transient postinduction hypotension. There were several limitations of this study. This study was performed at a single institution, and thus the information regarding anesthetic practice may not
2682
be generalizable. It is possible that practitioners at other centers are more diligent about reducing the dose of propofol in elderly patients. The anesthesiologists at our institution appeared to have used criteria other than age for determining propofol dose. In a subanalysis of data from the nonsurvivors, the dose of propofol used for induction was lesser, suggesting that some combination of chart review and “eyeball test”
Volume 37 Number 12
A.T. Phillips et al.
60yrs
70yrs
80yrs
2.50
2.25
Odds Ratio
2.00
1.75
1.50
1.25
1.00 2.0
2.5
3.0
3.5 2.0
2.5
3.0
3.5 2.0
2.5
3.0
3.5
Propofol Dose (mg/kg)
Figure 2. Odds ratios (95% CI) of postinduction hypotension per increase in dose of propofol used for induction (mg/kg), stratified by decade over 60 years of age. The steeper slopes seen in older age groups suggest that these groups are more susceptible than are younger groups to hypotension with greater doses of propofol.
influenced decision making. With the retrospective cohort study design, it is possible that there were confounding factors that were not controlled between nonsurvivors and survivors. The influences of type of procedure, sex, and weight were adjusted for in the multivariate analysis. Although some of the influence of these factors remains, the risk for death with
December 2015
increasing preoperative morbidity and risk were orders of magnitude greater. Finally, the present analysis included hemodynamic data from only the first 10 minutes of each surgical case. This method was based on the study of Reich et al,15 which reported an association between postinduction hypotension and mortality.
2683
Clinical Therapeutics
CONCLUSIONS The data from the present study suggest that older patients commonly receive propofol for induction at doses in excess of those in current recommendations. These relative overdoses are associated with hypotension that is dose dependent and is most dramatic in the oldest patients. However, propofol dose alone is not associated with a greater 30-day mortality rate, and propofol can be tolerated in the elderly population, even in the oldest subgroups. In the elderly population, cardiac physiology is characterized by increased afterload due to arterial stiffening, elevated systolic blood pressure, and left ventricular hypertrophy and calcification of the valves. Baroreceptor function is depressed. To minimize postinduction hypotension, slow induction may be combined with comprehensive clinical assessment and monitoring using processed electroencephalography when available. More education regarding concerns in the geriatric surgical population is needed for encouraging anesthesiologists to tailor the plan for anesthesia in geriatric patients; however, overall postsurgical mortality is a function of preoperative risk and type of surgical procedure.
ACKNOWLEDGMENTS This research was sponsored by the Icahn School of Medicine at Mount Sinai. There were no external sources of funding.
CONFLICTS OF INTEREST The authors have indicated that they have no conflicts of interest with regard to the content of this article.
REFERENCES 1. FastStats—Inpatient Surgery. Centers for Disease Control and Prevention. 2013. http://www.cdc.gov/nchs/fastats/ inpatient-surgery.htm. Accessed 2015. 2. Sebel P, Lowdon J. Propofol: a new intravenous anesthetic. Anesthesiology. 1989;71:260–277. 3. Hug CC, McLeskey CH, Nahrwold ML, et al. Hemodynamic effects of propofol: data from over 25,000 patients. Anesth Analg. 1993;77(Suppl):S21–S29. 4. Grounds RM, Morgan M, Lumley J. Some studies on the properties of the intravenous anaesthetic, propofol (“Diprivan”)—a review. Postgrad Med J. 1985;Suppl 3:90–95. 5. Kaplan JA, Guffin AV, Mikula S, et al. Comparative hemodynamic effects of propofol and thiamylal sodium during anesthetic induction for myocardial revascularization. J Cardiothorac Anesth. 1988;2:297–302.
2684
6. Butterworth J, Mackey DC, Wasnick J. Morgan and Mikhailʼs Clinical Anesthesiology, 5 ed. New York: McGraw Hill Professional; 2013. 7. McCollum JS, Dundee JW. Comparison of induction characteristics of four intravenous anaesthetic agents. Anaesthesia. 1986;41:995–1000. 8. Claeys MA, Gepts E, Camu F. Haemodynamic changes during anaesthesia induced and maintained with propofol. Br J Anaesth. 1988;60:3–9. 9. Dundee JW, Robinson FP, McCollum JS, Patterson CC. Sensitivity to propofol in the elderly. Anaesthesia. 1986;41:482–485. 10. Shafer SL. The pharmacology of anesthetic drugs in elderly patients. Anesthesiol Clin North Am. 2000;18: 1–29–v. 11. Maneglia R, Cousin MT. A comparison between propofol and ketamine for anaesthesia in the elderly. Haemodynamic effects during induction and maintenance. Anaesthesia. 1988;Suppl:109–111. 12. McEvoy MD, Reves JG. Intravenous hypnotic anesthetics. Geriatric Anesthesiology. New York: Springer New York; 2008: 229–245. 13. Propofol [package insert]. Hospira, Inc. 2014. 14. Steib A, Freys G, Beller JP, et al. Propofol in elderly high risk patients. A comparison of haemodynamic effects with thiopentone during induction of anaesthesia. Anaesthesia. 1988;Suppl:111–114. 15. Reich DL, Hossain S, Krol M, et al. Predictors of hypotension after induction of general anesthesia. Anesth Analg. 2005;101:622–628. 16. Chobanian AV, Bakris GL, Black HR, et al. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. JAMA. 2003;289:2560–2572. 17. ASA Physical Status Classification System. American Society of Anesthesiologists. 2014. http://www.asahq.org/resources/ clinical-information/asa-physical-status-classification-sys tem. Accessed 2015. 18. Averill RF, Goldfield NI, Muldoon J, et al. A closer look at all-patient refined DRGs. J Ahima. 2002;73:46–50. 19. Romano PS, Chan BK. Risk-adjusting acute myocardial infarction mortality: are APR-DRGs the right tool? Health Serv Res. 2000;34:1469–1489. 20. Goldfield N, Averill R. On “risk-adjusting acute myocardial infarction mortality: are APR-DRGs the right tool?” Health Serv Res. 2000;34:1491–1495. 21. Baram D, Daroowalla F, Garcia R, et al, Use of the All Patient Refined-Diagnosis Related Group. (APR-DRG) risk of mortality score as a severity adjustor in the medical ICU. Clin Med Circ Respirat Pulm Med. 2008;2:19–25. 22. Reich DL, Wood RK, Mattar R, et al. Arterial blood pressure and heart rate discrepancies between handwritten and computerized anesthesia records. Anesth Analg. 2000; 91:612–616.
Volume 37 Number 12
A.T. Phillips et al. 23. Sessler DI, Sigl JC, Kelley SD, et al. Hospital stay and mortality are increased in patients having a “triple low” of low blood pressure, low bispectral index, and low minimum alveolar concentration of volatile anesthesia. Anesthesiology. 2012;116:1195–1203.
Address correspondence to: Stacie Deiner, MD, Department of Anesthesiology, Icahn School of Medicine at Mount Sinai, 1 Gustave L. Levy Place, Box #1010, New York, NY 10029-6574. E-mail: [email protected]
December 2015
2685
Original Research
Propofol Use in the Elderly Population: Prevalence of Overdose and Association With 30-Day Mortality Adam T. Phillips, MD1; Stacie Deiner, MD2,3,4; Hung Mo Lin, ScD5; Evie Andreopoulos, MA5; Jeffrey Silverstein, MD2,4,5; and Matthew A. Levin, MD2 1
Department of Internal Medicine, School of Medicine, Yale University, New Haven, Connecticut; Departments of 2Anesthesiology; 3Neurosurgery; 4Geriatrics and Palliative Care; and 5Health Evidence and Policy, Icahn School of Medicine at Mount Sinai, New York, New York
ABSTRACT Purpose: Geriatric patients are more sensitive to the anesthetic effects of propofol and its adverse effects, such as hypotension, than is the general population; thus, a reduced dose (1–1.5 mg/kg) is recommended for the induction of anesthesia. The extent to which clinicians follow established dosing guidelines has not been well described. Therefore, we investigated the prevalence of propofol overdose in the elderly population to determine whether propofol overdose occurs and is associated with increased hypotension and 30-day mortality. Methods: In this retrospective study in patients who received propofol for the induction of general anesthesia, data on demographic characteristics, preoperative medications, intraoperative management, and 30-day mortality were collected. The dose of propofol used for the induction of anesthesia and the median blood pressure in the pre- and immediate postinduction periods were determined. Hypotension was defined as either: (1) a decrease in mean arterial pressure (MAP) of 440% concurrent with a MAP of o70 mm Hg; or (2) a MAP of o60 mm Hg. Findings: A total of 17,540 patients were included in the analysis; 4033 (23.0%) were aged 465 years. The median (interquartile range) propofol dose in the group aged 465 years was 1.8 (1.4–2.2) mg/kg, above the recommended dose, in comparison to 2.2 (1.9–2.5) mg/kg in younger patients. On multivariate
Accepted for publication October 6, 2015. http://dx.doi.org/10.1016/j.clinthera.2015.10.005 0149-2918/$ - see front matter & 2015 Elsevier HS Journals, Inc. All rights reserved.
2676
analysis, increased propofol dose was associated with increased postinduction hypotension, especially in patients over 70 years of age, but not 30-day mortality. Implications: Older patients received greater-thanrecommended doses of propofol for induction, which may have led to significant dose-dependent hypotension. Despite this finding, the dose of propofol for induction was not independently associated with a greater 30-day mortality rate. More education regarding geriatric concerns is needed for encouraging anesthesiologists to tailor the plan for anesthesia in geriatric patients. However, overall postsurgical mortality is a function of preoperative risk and type surgical procedure. (Clin Ther. 2015;37:2676–2685) & 2015 Elsevier HS Journals, Inc. All rights reserved. Key words: 30-day mortality, elderly, propofol overdose, hypotension.
INTRODUCTION Elderly patients constitute a large proportion of the surgical population in the United States, and anesthesiologists are often faced with the need for tailoring the anesthesia technique to account for geriatric physiology.1 The anesthesiologist may change the selection of medications or modify dosing in an attempt to optimize postoperative outcomes. However, geriatric-specific guidelines on anesthesia do not exist, and an anesthesiologist Scan the QR Code with your phone to obtain FREE ACCESS to the articles featured in the Clinical Therapeutics topical updates or text GS2C65 to 64842. To scan QR Codes your phone must have a QR Code reader installed.
Volume 37 Number 12
A.T. Phillips et al. may modify the technique based on a nonscientific impression of what is best for the patient. The use of propofol in the elderly population is one specific area of equipoise. Propofol is the medication most commonly used for inducing anesthesia in surgical patients in the United States. However, a common adverse effect of propofol-based induction is dose-dependent hypotension. Propofol-related hypotension increases with dose and results from a decrease in systemic vascular resistance and, to a lesser extent, myocardial depression.2–7 Elderly patients in particular have an increased sensitivity to this effect, leading to the manufacturersʼ recommendation that the dose of propofol used for induction be decreased in elderly patients from a recommended dose of 2 to 2.5 mg/kg, to 1 to 1.5 mg/kg.8–14 Reich et al15 reported that postinduction hypotension was associated with increased mortality despite controlling for the effects of age, preexisting hypotension, and American Society of Anesthesiologists (ASA)–defined physical status. Those investigators recommended that propofol induction be avoided in the elderly population, especially in patients with baseline low blood pressure. Nonetheless, many anesthesiologists continue to use propofol in the elderly population instead of induction drugs that cause less hypotension, such as etomidate. There is a distinct impression that propofol is well tolerated; however, it is unclear whether anesthesiologists employ reduced doses and whether reduced propofol doses can be used for avoiding hypotension. To address these areas of equipoise, we performed a retrospective review of data from a large clinical cohort of patients aged 19 to 89 years who received propofol for the induction of general anesthesia. Our hypothesis was that elderly patients often receive propofol in excess of 1.5 mg/ kg (the recommended dose for age 465 years) and that this relative overdose is associated with increased postinduction hypotension and increased 30-day mortality.
PATIENTS AND METHODS The study protocol was approved by the Program for the Protection of Human Subjects, Icahn School of Medicine at Mount Sinai (New York, New York), and the requirement for written informed consent was waived pursuant to 45 CFR 46.101(i). Data from all inpatients and day-of-admission surgical patients aged 19 to 89 years who had received a general anesthetic with induction using intravenous propofol between January
December 2015
2006 and December 2009 were included, with the exception of those undergoing cardiac or obstetric surgery (eg, cesarean section). In patients who underwent multiple surgeries in the time period studied, only data from the most recent surgery were included in the analysis. Primary outcomes were hypotension in the 10 minutes after induction and 30-day mortality. Perioperative demographic, medication, and physiologic data were extracted from the anesthesia information management system (AIMS) (CompuRecord; Phillips Medical Systems, Andover, Massachusetts). Classification of preoperative antihypertensive medication was based on the major classes of antihypertensive medications recommended by the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure for the treatment of hypertension.16 Additional data on demographic characteristics and mortality were obtained from the institutional data warehouse. We also utilized the Social Security Administration Death Master File (National Technical Information Service, Alexandria, Virginia) as an additional source of data on mortality. As a measure of coexisting illness, we used the both the ASA physical status score and the All Patient Refined–Diagnosis Related Group risk for mortality (APR-DRG ROM) scale score (extreme, major, moderate, and minor). ASA scores range from 1 (healthy with no comorbidities) to 5 (not expected to survive without the operation).17 APR-DRG is a classification scheme developed by 3M (Maplewood, Minnesota) based on nationwide samples and used by Medicare to estimate resource consumption by patients.18 ROM score is calculated based on primary and secondary discharge diagnoses and thus captures all comorbidities present during an admission. Although ASA and ROM are correlated in many conditions, they sometimes differ because the ASA score is a qualitative score assigned by the anesthesiologist, whereas ROM is determined based on administrative coding data. APR-DRG ROM has been reported to correlate well with mortality among patients in the intensive care unit and to be a suitable tool for risk adjustment of 30-day mortality in patients with acute myocardial infarction.19–21
Determination of Dose of Propofol Used for Induction and Co-Induction Agents At the Icahn School of Medicine, propofol is typically injected manually, usually as a single bolus,
2677
Clinical Therapeutics for the purpose of induction. Occasionally a provider may titrate the induction using divided doses. Induction dose was therefore measured as the total dose of propofol given in the period beginning 5 minutes before and ending 2 minutes after the induction timestamp as recorded in the AIMS. Total dose was then divided by the patientʼs weight as recorded in the AIMS to give an induction dose by weight (in mg/kg).
(in liters), transfusion of packed red blood cells (dichotomous variable), and transfusion of platelets or fresh frozen plasma (dichotomous variable). To determine whether the effects of propofol dose on both outcomes differed depending on the preoperative risk profile, we tested the 2-way interactions between propofol and age and ASA physical status, respectively.
Determination of Baseline and Postinduction Blood Pressure
RESULTS
Data on blood pressure were electronically captured by the AIMS as often as every 15 seconds for an intra-arterial line and every 1 to 5 minutes for oscillometric noninvasive blood pressure cuff. We report baseline (10 minutes preinduction) and 10 minutes postinduction as the median mean arterial pressure (MAP) of each period. Medians were used for filtering out transient changes and artefact.22 Postinduction hypotension was defined, as per Reich et al,15 as either: (1) a decrease in MAP of 440% from baseline and a MAP of o70 mm Hg; or (2) a MAP of o60 mm Hg.
Statistical Analysis Patientsʼ characteristics are described as numbers (%), medians (interquartile ranges) or means (SD). For comparisons between groups, χ2 tests were used for categorical variables, and 2-sample t tests were used for continuous variables, as appropriate. Forward stepwise logistic regression analyses (with significance levels for both the stay and entry criteria set at 0.05) were performed for the identification of significant predictors of postinduction hypotension and 30-day mortality, respectively. For the outcome of postinduction hypotension, the covariates considered in the initial models included the dose of propofol used for induction, age, sex, weight, ASA physical status, APRDRG ROM classification, emergency status, baseline MAP, use of antihypertensive medication, class of antihypertensive used, coadministration of fentanyl or midazolam during induction, and type of surgical procedure. A dichotomous variable was used for indicating whether the attending anesthesiologist was working alone. In addition, we considered the case volume of the attending anesthesiologist during the study period. For the 30-day mortality model, the following additional intraoperative covariates were added: duration of anesthesia (in minutes), total amount of crystalloid given
2678
A total of 17,810 patients met the inclusion criteria, of whom 17,540 patients (98.5%) who received propofol had complete covariate data available. The demographic characteristics of the cohort are shown in Table I. Approximately half the population was female, and 4033 (23.0%) were aged 465 years. Most patients were either ASA status 1–2 (58.2%) or 3 (37.4%). The type of surgery most commonly reported was general surgery, followed by orthopedic and gynecologic surgery. Overall, the surgeries were relatively long; the median (interquartile range) duration of anesthesia was 192 (139–270) minutes. The median propofol dose in the overall cohort was 2.1 (1.8–2.5) mg/kg, with 2.2 (1.9–2.5) mg/kg in the group aged r65 years versus 1.8 (1.4–2.2) mg/kg in the group aged 465 years. Figure 1 demonstrates that the median dose of propofol used for induction decreased with age; however, even at 80 years of age, the median dose was still greater than the recommended dose of 1 to 1.5 mg/kg. The dose administered was also highly variable at all ages.
Results of Univariate Analysis Table I also shows the results of the univariate analysis of postinduction hypotension and 30-day mortality. Postinduction hypotension occurred in 5.96% of all patients. The hypotensive patients tended to be older (median, 60 vs 53 years; P o 0.0001), have a higher ASA physical status, and were more likely to have been taking a β-blocker before surgery. The 30day mortality rate was 0.89%. Patients who died were older, male, had greater preoperative APR-DRG ROM scores and ASA scores, were more likely to have been on preoperative β-blockers, and were more likely to have undergone a neurosurgical procedure.
Results of Multivariate Analysis In the multivariate model, APR-DRG ROM score, emergency surgery, otolaryngologic procedure, dose of
Volume 37 Number 12
December 2015
Table I. Patients’ characteristics, overall and by study outcome (univariate analysis). Hypotension Within First 10 Min Parameter Age Median (IQR), y 465 y, no. (%) Female, no. (%) Weight, median (IQR), kg ASA physical status,* no. (%) 1 or 2 3 4 or 5 APR-DRG risk for mortality, no. (%) Minor Moderate Major Extreme Emergency case, no. (%) Antihypertensive medication, no. (%) β-Blocker Other None Baseline MAP, median (IQR), mm Hg
2679
Group, no. (%) 0–1 h 41–2 h
Yes (n ¼ 1047)
No (n ¼ 16,493)
53 (41–64) 4033 (23.0) 10,005 (57.0) 75.0 (63.0–88.1)
60 (49–69) 360 (34.4) 650 (62.1) 70.0 (59.1–84.1)
53 (49–64) 3673 (22.3) 9355 (56.7) 75.0 (59.1–88.6)
10,213 (58.2) 6562 (37.4) 765 (4.4)
488 (46.6) 476 (45.5) 83 (7.9)
9725 (59.0) 6086 (36.9) 682 (4.1)
12,914 (73.6) 3072 (17.5) 1111 (6.3) 443 (2.5) 1869 (10.7)
658 (62.8) 220 (21.0) 100 (9.6) 69 (6.6) 144 (13.8)
12,256 (74.3) 2852 (17.3) 1011 (6.1) 374 (2.3) 1725 (10.5)
2372 (13.5) 2828 (16.1) 12,340 (70.4) 103 (93–114)
178 (17.0) 194 (18.5) 675 (64.5) 102.9 (87.4–114)
2194 (13.3) 2634 (16.0) 11,665 (70.7) 103.0 (87.4–114)
192 (139–270)
187.0 (132–265)
192.0 (132–270)
245 (1.4) 2799 (16.0)
16 (1.5) 192 (18.3)
229 (1.4) 2607 (15.8)
P o0.001 o0.001 o0.001 o0.001
Yes (n ¼ 156)
No (Survivors) (n ¼ 17,384)
63 (53–74) 69 (44.2) 66 (42.3) 72.7 (62.6–88.1)
53 (53–64) 3964 (22.8) 9939 (57.2) 75.0 (62.6–88.1)
11 (7.1) 67 (42.9) 78 (50.0)
10,202 (58.7) 6495 (37.4) 687 (4.0)
5 (3.2) 4 (2.6) 36 (23.1) 111 (71.2) 37 (23.7)
12,909 (74.3) 3068 (17.6) 1075 (6.2) 332 (1.9) 1832 (10.5)
41 (26.3) 25 (16.0) 90 (57.7) 101.0 (88.7–113)
2331 (13.4) 2803 (16.1) 12,250 (70.5) 103.0 (88.7–114)
180.5 (103–296)
192.0 (103–270)
4 (2.6) 41 (26.3)
241 (1.4) 2758 (15.9)
o0.001
o0.001
0.12
o0.001 o0.001 o0.001 0.42 o0.001
o0.001
o0.001 o0.001
P
o0.001
o0.001 o0.001
0.12
0.196 o0.0001
0.07
(continued)
A.T. Phillips et al.
Duration of anesthesia Median (IQR), min
All Patients (N ¼ 17,540)
30-D Mortality
Hypotension Within First 10 Min Parameter 42–4 h 44–6 h 46 h Surgery type, no. (%) General Otolaryngologic Gynecologic Urologic Neurologic Spinal Vascular Transplantation Endoscopic Other Crystalloid given, no. (%) 0–1 L 41–3 L 43–5 L 45 L Not documented Propofol dose, median (IQR), mg/kg Volume 37 Number 12
Transfusion, no. (%) RBCs Fresh frozen plasma Platelets
All Patients (N ¼ 17,540)
Yes (n ¼ 1047)
No (n ¼ 16,493)
8803 (50.2) 3885 (22.1) 1808 (10.3)
519 (49.6) 203 (19.4) 117 (11.2)
8284 (50.2) 3682 (22.3) 1691 (10.3)
5998 (34.2) 2176 (12.4) 1973 (11.2) 1432 (8.2) 1318 (7.5) 1226 (7.0) 460 (2.6) 425 (2.4) 119 (0.7) 2413 (13.8)
326 (31.1) 219 (20.9) 73 (7.0) 55 (5.3) 82 (7.8) 77 (7.4) 28 (2.7) 21 (2.0) 10 (1.0) 156 (14.9)
5672 (34.4) 1957 (11.9) 1900 (11.5) 1377 (8.4) 1236 (7.5) 1149 (7.0) 432 (2.6) 404 (2.5) 109 (0.7) 2257 (13.7)
P
30-D Mortality Yes (n ¼ 156)
No (Survivors) (n ¼ 17,384)
56 (35.9) 29 (18.6) 26 (16.7)
8747 (50.3) 3856 (22.2) 1782 (10.3)
43 (27.6) 24 (15.4) 8 (5.1) 8 (5.1) 23 (14.7) 8 (5.1) 9 (5.8) 6 (3.9) 11 (7.1) 16 (10.3)
5955 (34.3) 2152 (12.4) 1965 (11.3) 1424 (8.2) 1295 (7.5) 1218 (7.0) 451 (2.6) 419 (2.4) 108 (0.6) 2397 (13.8)
o0.001
o0.001
o0.001
0.97 5504 (31.4) 8248 (47.0) 2402 (13.7) 685 (3.9) 701 (4.0) 2.1 (1.8–2.5)
324 (30.9) 496 (47.4) 148 (14.1) 38 (3.6) 41 (3.9) 2.08 (1.7–2.5)
5180 (31.4) 7752 (47.0) 2254 (13.7) 647 (3.9) 660 (4.0) 2.11 (1.7–2.5)
1087 (6.2) 187 (1.1) 156 (0.9)
99 (9.5) 14 (1.3) 7 (0.7)
988 (6.0) 173 (1.1) 149 (0.9)
0.10
o0.001 0.38 0.43
P
57 (36.5) 52 (33.3) 21 (13.5) 11 (7.1) 15 (9.6) 1.59 (1.1–2.1)
5447 (31.3) 8196 (47.1) 2381 (13.7) 674 (3.9) 686 (4.0) 2.11 (1.1–2.5)
46 (29.5) 20 (12.8) 18 (11.5)
1041 (6.0) 167 (1.0) 138 (0.8)
o0.001
o0.001 o0.001 o0.001
APR-DRG ¼ All Patient Refined–Diagnosis Related Group; ASA ¼ American Society of Anesthesiologists physical status scale; IQR ¼ interquartile range; MAP ¼ mean arterial pressure; RBC, red blood cell. * Score range: 1 (healthy with no comorbidities) to 5 (not expected to survive without the operation).
Clinical Therapeutics
2680
Table I. (continued).
A.T. Phillips et al.
Propofol Dose (mg/kg)
6
4
2
0 20
40
60
80
Age (yr)
Figure 1. Scatterplot showing the relationship between age and propofol-induction dose (mg/kg). Dose of propofol decreases as age increases, but not as much as published guidelines would recommend. The dark line is a smoothed curve generated using the penalized B-spline method (PROC SGCATTER, SAS; SAS Institute Inc, Cary, North Carolina).
propofol used for induction, and age were all strong predictors of postinduction hypotension (Table II). The interaction between propofol and age was significant. Figure 2 demonstrates that for each decade of increasing age 460 years, a given dose of propofol was associated with a progressively greater risk for hypotension. After adjustment for covariates, propofol dose was not significantly associated with 30-day mortality (Table II).
DISCUSSION In this large, retrospective study, we found that elderly patients commonly receive greater-than-recommended doses of propofol for the induction of general anesthesia. To our knowledge, this result has not been previously reported. Greater doses were associated with an increased prevalence of significant hypotension in the postinduction period, consistent with earlier reports.9 An increased dose of propofol used for induction, however, was not associated with increased 30-day mortality. Instead, mortality was more strongly associated with the inherent risk of the procedure and patient characteristics.
December 2015
Our results conflict with those from Reich et al,15 who reported that hypotension was not dose dependent. However, the findings from both studies agree that death is more common in patients with postinduction hypotension. Our findings suggest that large doses of propofol cause hypotension but also that, overall, there is not a “failure to rescue.” Other important differences between the 2 studies are that Reich et al did not examine or interpret the interaction of age and propofol dosing and that they treated age as a dichotomous variable, in which patients aged 450 years were classified as “old.” We chose to look at age in a more nuanced fashion and report that as age increased by decade, the dose-dependent effect of propofol-induced hypotension became increasingly dramatic (Figure 2). This finding suggests that even the recommended dosing of 1-1.5 mg/kg should be further reduced for patients over the age of 70-80. In the multivariate model adjusted for preoperative risk and type of surgery, the initial propofol dose was not associated with 30-day mortality. Whether persistent overdose of anesthetic agent occurs with propofol infusion, whether there is a failure to rescue with “pressors,” and whether doses of inhalational anesthetic agents are not being adjusted for age were not addressed in this study. However, Sessler et al23 suggested that the absolute dose of anesthetic is less important to postsurgical outcomes than is the larger picture of whether the patient has a susceptible phenotype of both sensitivity to anesthetic agents and persistent low blood pressure throughout the case. Regarding the contribution of intraoperative course to postoperative mortality, we did examine the role of transfusion as a surrogate marker for intraoperative misadventure. In our study, the need for red blood cell transfusion was not associated with hypotension, but the need for platelet transfusion was. This finding may have been due to the fact that platelets are generally given in the setting of large volume resuscitation. Whether patients who required platelet transfusion also had intraoperative hypotension after the first 10 minutes could not be determined from this dataset. We did not include the duration of hypotension in the 10-minute window after induction in the multivariate model of 30-day mortality. The duration of hypotension may not add information because it was likely collinear with other variables in the model, including propofol dose and APR-DRG
2681
Clinical Therapeutics
Table II. Final multivariate logistic models for hypotension and 30-day mortality. Hypotension Parameter Propofol (per mg/kg)* Female vs male Weight (per 5 kg) APR-DRG risk for mortality Moderate Major Extreme ASA physical status‡ 1 or 2 (Ref) 3 4þ Baseline MAP (per 5 mm Hg) Emergency case Surgical indication General (Ref) Otolaryngologic Spinal Neurologic Endoscopic Vascular Urologic Transplantation Gynecologic Other Need platelet transfusion
OR
95% CI
30-D Mortality P
OR
95% CI 0.690–1.182
0.457
†
P
1.25 0.93
1.08–1.45 0.89–0.97
0.003 o0.001
0.90 NS‡ NS
1.10 1.28 2.31
0.92–1.30 1.00–1.64 1.70–3.16
0.304 0.051 o0.001
2.55 51.0 440
0.66–9.76 18.6–139 164–4999
0.173 o0.001 o0.001
1.24 1.42 0.910 1.46
1.07–1.45 1.06–1.92 0.891–0.929 1.20–1.78
0.005 0.020 o0.001 o0.001
1.69 4.64 NS NS
0.83–3.44 2.18–9.87
0.152 o0.001
2.00 1.30 1.11 1.04 0.81 0.80 0.76 0.70 1.16 NA
1.66–2.45 1.00–1.69 0.86–1.44 0.53–2.05 0.54–1.22 0.59–1.08 0.48–1.22 0.53–0.91 0.95–1.42
o0.001 0.048 0.407 0.912 0.317 0.147 0.259 0.008 0.148
0.95 1.32 1.02 2.12 0.45 1.16 0.23 1.15 0.60 3.03
0.52–1.72 0.54–3.24 0.56–1.86 0.92–4.90 0.20–1.03 0.49–2.74 0.09–0.60 0.49–2.69 0.32–1.13 1.55–5.92
0.853 0.545 0.950 0.078 0.057 0.741 0.002 0.749 0.112 0.001
APR-DRG ¼ All Patient Refined–Diagnosis Related Group; ASA ¼ American Society of Anesthesiologists physical status; MAP ¼ mean arterial pressure; NA ¼ not applicable because the variable was not considered in this initial model selection; NS ¼ not significant after forward stepwise selection; OR ¼ odds ratio. * Propofol was forced into the model. † The age and propofol interaction is significant (P ¼ 0.016). The joint effects of propofol and age on hypotension in the first 10 minutes after induction are illustrated in Figure 2. ‡ Score range: 1 (healthy with no comorbidities) to 5 (not expected to survive without the operation).
classification. Overall, the findings from our study support that a preoperative risk for mortality and comorbidity are much stronger predictors than is either the use of propofol for induction or transient postinduction hypotension. There were several limitations of this study. This study was performed at a single institution, and thus the information regarding anesthetic practice may not
2682
be generalizable. It is possible that practitioners at other centers are more diligent about reducing the dose of propofol in elderly patients. The anesthesiologists at our institution appeared to have used criteria other than age for determining propofol dose. In a subanalysis of data from the nonsurvivors, the dose of propofol used for induction was lesser, suggesting that some combination of chart review and “eyeball test”
Volume 37 Number 12
A.T. Phillips et al.
60yrs
70yrs
80yrs
2.50
2.25
Odds Ratio
2.00
1.75
1.50
1.25
1.00 2.0
2.5
3.0
3.5 2.0
2.5
3.0
3.5 2.0
2.5
3.0
3.5
Propofol Dose (mg/kg)
Figure 2. Odds ratios (95% CI) of postinduction hypotension per increase in dose of propofol used for induction (mg/kg), stratified by decade over 60 years of age. The steeper slopes seen in older age groups suggest that these groups are more susceptible than are younger groups to hypotension with greater doses of propofol.
influenced decision making. With the retrospective cohort study design, it is possible that there were confounding factors that were not controlled between nonsurvivors and survivors. The influences of type of procedure, sex, and weight were adjusted for in the multivariate analysis. Although some of the influence of these factors remains, the risk for death with
December 2015
increasing preoperative morbidity and risk were orders of magnitude greater. Finally, the present analysis included hemodynamic data from only the first 10 minutes of each surgical case. This method was based on the study of Reich et al,15 which reported an association between postinduction hypotension and mortality.
2683
Clinical Therapeutics
CONCLUSIONS The data from the present study suggest that older patients commonly receive propofol for induction at doses in excess of those in current recommendations. These relative overdoses are associated with hypotension that is dose dependent and is most dramatic in the oldest patients. However, propofol dose alone is not associated with a greater 30-day mortality rate, and propofol can be tolerated in the elderly population, even in the oldest subgroups. In the elderly population, cardiac physiology is characterized by increased afterload due to arterial stiffening, elevated systolic blood pressure, and left ventricular hypertrophy and calcification of the valves. Baroreceptor function is depressed. To minimize postinduction hypotension, slow induction may be combined with comprehensive clinical assessment and monitoring using processed electroencephalography when available. More education regarding concerns in the geriatric surgical population is needed for encouraging anesthesiologists to tailor the plan for anesthesia in geriatric patients; however, overall postsurgical mortality is a function of preoperative risk and type of surgical procedure.
ACKNOWLEDGMENTS This research was sponsored by the Icahn School of Medicine at Mount Sinai. There were no external sources of funding.
CONFLICTS OF INTEREST The authors have indicated that they have no conflicts of interest with regard to the content of this article.
REFERENCES 1. FastStats—Inpatient Surgery. Centers for Disease Control and Prevention. 2013. http://www.cdc.gov/nchs/fastats/ inpatient-surgery.htm. Accessed 2015. 2. Sebel P, Lowdon J. Propofol: a new intravenous anesthetic. Anesthesiology. 1989;71:260–277. 3. Hug CC, McLeskey CH, Nahrwold ML, et al. Hemodynamic effects of propofol: data from over 25,000 patients. Anesth Analg. 1993;77(Suppl):S21–S29. 4. Grounds RM, Morgan M, Lumley J. Some studies on the properties of the intravenous anaesthetic, propofol (“Diprivan”)—a review. Postgrad Med J. 1985;Suppl 3:90–95. 5. Kaplan JA, Guffin AV, Mikula S, et al. Comparative hemodynamic effects of propofol and thiamylal sodium during anesthetic induction for myocardial revascularization. J Cardiothorac Anesth. 1988;2:297–302.
2684
6. Butterworth J, Mackey DC, Wasnick J. Morgan and Mikhailʼs Clinical Anesthesiology, 5 ed. New York: McGraw Hill Professional; 2013. 7. McCollum JS, Dundee JW. Comparison of induction characteristics of four intravenous anaesthetic agents. Anaesthesia. 1986;41:995–1000. 8. Claeys MA, Gepts E, Camu F. Haemodynamic changes during anaesthesia induced and maintained with propofol. Br J Anaesth. 1988;60:3–9. 9. Dundee JW, Robinson FP, McCollum JS, Patterson CC. Sensitivity to propofol in the elderly. Anaesthesia. 1986;41:482–485. 10. Shafer SL. The pharmacology of anesthetic drugs in elderly patients. Anesthesiol Clin North Am. 2000;18: 1–29–v. 11. Maneglia R, Cousin MT. A comparison between propofol and ketamine for anaesthesia in the elderly. Haemodynamic effects during induction and maintenance. Anaesthesia. 1988;Suppl:109–111. 12. McEvoy MD, Reves JG. Intravenous hypnotic anesthetics. Geriatric Anesthesiology. New York: Springer New York; 2008: 229–245. 13. Propofol [package insert]. Hospira, Inc. 2014. 14. Steib A, Freys G, Beller JP, et al. Propofol in elderly high risk patients. A comparison of haemodynamic effects with thiopentone during induction of anaesthesia. Anaesthesia. 1988;Suppl:111–114. 15. Reich DL, Hossain S, Krol M, et al. Predictors of hypotension after induction of general anesthesia. Anesth Analg. 2005;101:622–628. 16. Chobanian AV, Bakris GL, Black HR, et al. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. JAMA. 2003;289:2560–2572. 17. ASA Physical Status Classification System. American Society of Anesthesiologists. 2014. http://www.asahq.org/resources/ clinical-information/asa-physical-status-classification-sys tem. Accessed 2015. 18. Averill RF, Goldfield NI, Muldoon J, et al. A closer look at all-patient refined DRGs. J Ahima. 2002;73:46–50. 19. Romano PS, Chan BK. Risk-adjusting acute myocardial infarction mortality: are APR-DRGs the right tool? Health Serv Res. 2000;34:1469–1489. 20. Goldfield N, Averill R. On “risk-adjusting acute myocardial infarction mortality: are APR-DRGs the right tool?” Health Serv Res. 2000;34:1491–1495. 21. Baram D, Daroowalla F, Garcia R, et al, Use of the All Patient Refined-Diagnosis Related Group. (APR-DRG) risk of mortality score as a severity adjustor in the medical ICU. Clin Med Circ Respirat Pulm Med. 2008;2:19–25. 22. Reich DL, Wood RK, Mattar R, et al. Arterial blood pressure and heart rate discrepancies between handwritten and computerized anesthesia records. Anesth Analg. 2000; 91:612–616.
Volume 37 Number 12
A.T. Phillips et al. 23. Sessler DI, Sigl JC, Kelley SD, et al. Hospital stay and mortality are increased in patients having a “triple low” of low blood pressure, low bispectral index, and low minimum alveolar concentration of volatile anesthesia. Anesthesiology. 2012;116:1195–1203.
Address correspondence to: Stacie Deiner, MD, Department of Anesthesiology, Icahn School of Medicine at Mount Sinai, 1 Gustave L. Levy Place, Box #1010, New York, NY 10029-6574. E-mail: [email protected]
December 2015
2685