Featured Articles: Original Clinical Research Report

A Positive Cocaine Urine Toxicology Test and the Effect on Intraoperative Hemodynamics Under General Anesthesia

Moon, Tiffany S. MD; Pak, Taylor J. BS; Kim, Agnes MS; Gonzales, Michael X. MD; Volnov, Yuri MD; Wright, Evan MD; Vu, Kevin Q. BA; Lu, Rachael D. BS; Sharifi, Arghavan BS; Minhajuddin, Abu PhD; Chen, Joy L. MD; Fox, Pamela E. MD; Gasanova, Irina MD, PhD; Fox, Amanda A. MD, MPH; Stewart, Jesse MD; Ogunnaike, Babatunde MD

Author Information

From the Department of Anesthesiology and Pain Management, University of Texas Southwestern Medical Center, Dallas, Texas.

Published ahead of print 16 April 2020.

Accepted for publication March 9, 2020.

Funding: Institutional and/or departmental.

The authors declare no conflicts of interest.

Clinical Trial Registration: NCT02692534.

Reprints will not be available from the authors.

Address correspondence to Tiffany S. Moon, MD, Department of Anesthesiology and Pain Management, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390. Address e-mail to [email protected].

Anesthesia & Analgesia 132(2):p 308-316, February 2021. | DOI: 10.1213/ANE.0000000000004808

Free
CME Test
Corresponding Article

Abstract

KEY POINTS

Question: Do asymptomatic cocaine-positive patients have a similar percentage of intraoperative hemodynamic events compared to cocaine-negative patients?
Findings: Appropriately selected cocaine-positive patients undergoing elective noncardiac surgery under general anesthesia have a similar percentage of intraoperative hemodynamic events compared to cocaine-negative patients.
Meaning: Patients with a positive cocaine urine toxicology test who do not have signs or symptoms of acute intoxication may have a similar intraoperative hemodynamic course as cocaine-negative patients.

See Article, p 305

Cocaine use is widespread with an estimated 18 million regular users worldwide.^¹ About 1% of elective surgical patients and 38% of major trauma patients present to the hospital with a positive cocaine urine toxicology test.^²^,^³ Cocaine has a biological half-life of 45–90 minutes and is rapidly metabolized by plasma and liver esterases.^⁴ The cocaine urine toxicology test evaluates for the presence of benzoylecgonine, an inactive metabolite that can be detected for up to a week after cocaine use.^⁵^,^⁶ Although false positives are very rare, the urine toxicology can remain positive long after cocaine has exerted its effects, creating the dilemma of a cocaine-positive surgical patient who does not exhibit any signs or symptoms of cocaine intoxication.^⁷

Historically, cocaine-positive patients would have their elective surgery postponed until their urine toxicology test was negative.^⁸^,^⁹ Same-day cancellations of surgical cases result in a delay to treatment for the patient and wastage of operating room resources for the hospital.^¹⁰ Additionally, “elective” and “nonemergent” are not synonymous. In cases where the surgery is nonemergent but is not entirely elective (eg, mastectomy), postponing treatment can contribute to disease progression (eg, cancer). This scenario has particular significance in public urban hospitals, where resources are limited and cocaine use is common.^²^,^³

The primary aim of this study was to evaluate whether asymptomatic cocaine-positive patients had a similar percentage of intraoperative hemodynamic events, defined as the 2 coprimary outcomes of (1) a mean arterial blood pressure (MAP) of <65 or >105 mm Hg and (2) a heart rate (HR) of <50 or >100 beats per minute (bpm), compared to cocaine-negative patients.^{^11–13} We hypothesized that the mean percentage of anesthesia duration outside of the specified intraoperative hemodynamic limits would be equivalent between the 2 groups. The secondary aims were to evaluate perioperative troponin values and the minimum alveolar concentration (MAC) requirements of sevoflurane. We hypothesized that the perioperative troponin values and MAC requirements between the 2 groups would be equal.

METHODS

Study Design

This study was approved by the institutional review board (IRB) of the University of Texas Southwestern Medical Center (IRB #062015-066) and registered before patient enrollment at ClinicalTrials.gov (NCT02692534; principal investigator: Tiffany Moon, MD; date of registration: January 21, 2016). Written informed consent was obtained from all patients before study enrollment. This manuscript adheres to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines.

This single-center, prospective study was performed at Parkland Health & Hospital System in Dallas, TX. For 20 consecutive months, elective surgical patients with a history of cocaine use in the previous 5 years were screened for participation in the study. Inclusion criteria were American Society of Anesthesiologists (ASA) physical status I–III, 18–70 years old, plan for general anesthesia requiring endotracheal intubation, and a normal preoperative 12-lead electrocardiogram (ECG) and vital signs. Exclusion criteria included acute cocaine intoxication as defined by the World Health Organization (eg, psychomotor agitation, delirium, psychosis, diaphoresis, tachycardia, hypertension, etc).^¹⁴ Other exclusion criteria were abnormal vital signs, abnormal ECG, or a history of coronary artery disease, arrhythmia, stroke, seizure disorder, or end-stage renal disease.^¹⁴ Patients undergoing high-risk surgeries with expected hemodynamic disturbances or significant blood loss were excluded. Patients with polysubstance abuse (except marijuana) were excluded. Pregnant or nursing women and incarcerated patients were also excluded.

Anesthesia Protocol

In the preoperative phase, a urine toxicology test (Cobas COC2; Roche Diagnostics, Indianapolis, IN) was performed to categorize patients into cocaine-positive versus cocaine-negative groups. Baseline vital signs and a 12-lead ECG were obtained. The intraoperative anesthesia protocol was standardized for all patients. All patients received premedication with midazolam 2 mg. Anesthesia was induced with propofol 1.5–2 mg·kg⁻¹ and fentanyl 1.5 mcg·kg⁻¹. Rocuronium 0.6 mg·kg⁻¹ was given to facilitate orotracheal intubation with a cuffed tube. A bispectral index (BIS) monitor (Covidien, Mansfield, MA) was applied, and the sevoflurane was titrated to a BIS target range of 40–60. A balanced anesthetic regimen of sevoflurane and fentanyl was administered throughout the procedure. After meeting standard extubation criteria, all patients were extubated in the operating room and taken to the postanesthesia care unit (PACU). Pre- and postoperative blood samples were drawn for fourth-generation cardiac troponin T (cTnT) analysis (Cobas Troponin T STAT; Roche Diagnostics), which was analyzed in real time on fresh samples.

Primary and Secondary Outcomes

There are 2 coprimary outcomes defined as the percent of anesthesia duration that a subject had (1) a MAP of <65 or >105 mm Hg and (2) a HR of <50 or >100 bpm. Both outcomes are expressed as the percentage of anesthesia duration that each subject experienced the respective hemodynamic event to account for the variable lengths of anesthesia duration among subjects. Blood pressure (BP) data were obtained using an automated BP cuff, with a sampling rate of once every 3 minutes. HR data were obtained from the intraoperative 5-lead ECG, with a sampling rate of once every 10 seconds. The total number of intraoperative hemodynamic events was divided by the total anesthesia duration, thus permitting comparison for surgeries of different lengths.

Secondary exploratory outcomes were perioperative cTnT values and the MAC requirements of sevoflurane between the 2 groups.

Statistical Analysis

Continuous data were summarized as mean ± standard deviation (SD) or median and interquartile range (IQR). Categorical variables were summarized as frequency and percentages. Statistical comparisons between cocaine-positive and cocaine-negative groups were performed using 2-sample Student t tests for continuous variables and χ² tests for categorical variables. The equivalence of the cocaine-positive and cocaine-negative groups was assessed in terms of the β coefficient from a multivariable regression model. In this model, the β coefficient measures the confounder-adjusted difference between the cocaine-positive and cocaine-negative groups in mean percentage of the case outside of the specified intraoperative hemodynamic limits. The margin of equivalence for MAP events was set at 7.5% and that for HR events was set at 5.0%. Following Mascha and Sessler,^¹⁵ 2 one-sided tests for the β coefficient of an indicator variable for cocaine-positive and cocaine-negative groups from a multivariable linear regression model were used to assess the equivalence of the adverse hemodynamic outcome measures between the groups adjusted for baseline characteristics with a standardized difference >0.2 between the 2 groups. These included age, sex, body mass index (BMI), smoking status, hypertension, renal disease, and psychiatric illnesses. Each of the 2 one-sided tests would be assessed for statistical significance at P < .025, and if both tests were statistically significant, the 2 groups would be considered to be statistically equivalent. Utilizing similar arguments, we reported 95% confidence intervals (CIs) for the β coefficients and considered the cocaine-positive and cocaine-negative groups statistically equivalent if the CI included 0 and both confidence limits were within the margin of equivalence. Normality of outcome measures defined as the percentage of anesthesia duration outside the limits as well as regression residuals were assessed using normal quantile plots. The 2 cohorts were also compared in terms of secondary outcomes of perioperative cTnT values and the MAC requirements of sevoflurane using Student t tests to assess if the mean levels of troponin T and MAC requirements in the 2 groups were equal. All analyses were performed using SAS version 9.3 (SAS Inc, Cary, NC). Type I error rate across the primary hypotheses tests was set at α = .05. Thus, using Bonferroni corrections, statistical significance for each of the 2 coprimary outcomes was assessed as P values <.025, while all secondary outcomes analyses were considered statistically significant if P < .05.

Sample Size Calculation

The study was powered to assess if the cocaine-positive and cocaine-negative groups were similar in terms of percentage of anesthesia duration outside the specified limits for MAP and HR using an equivalence test of means using 2 one-sided tests on data from a parallel-group design. A 7.5% margin of equivalence for MAP events and 5.0% for HR events was used. Therefore, the 2 groups would be considered to be equivalent if the percentages of MAP events were within 7.5% of each other and the percentages of HR events were within 5.0% of each other. Since no other preliminary data for variability of these outcomes were available in the literature, a small pilot sample of 5 cocaine-negative and 5 cocaine-positive patients was used to estimate SDs of these outcome measures. Using an SD of 13% (estimated from the pilot study) and a 5% margin of equivalence, HR events required the larger sample size of 146 per group, assuming an α of .025 and 80% power. With an SD of 14.5%, 7.5% margin of equivalence, and a sample size of 146 per group, the equivalence test for the MAP outcome has 98% power with α = .025. Assuming an estimated 5% dropout rate, the study thus required a minimum of 154 patients in each group to have 80% power. Since more patients tested positive than negative during enrollment, a total of 154 cocaine-negative and 173 cocaine-positive patients were enrolled.

RESULTS

Figure 1.:
Flow diagram of patient enrollment.

A total of 173 cocaine-positive patients and 155 cocaine-negative patients were enrolled in the study (Figure 1). One cocaine-negative patient withdrew their consent in writing after being discharged from the hospital. The other 327 patients who were enrolled completed the study procedures and were included in the data analysis.

Patient Characteristics

Table 1. - Patient Characteristics

	Cocaine Negative (n = 154)	Cocaine Positive (n = 173)	P	Standardized Differences
Age (y)	49.4 (11.3)	45.3 (11.9)	.002^a	0.35
Body mass index (kg/m²)	29.7 (6.8)	27.6 (6.6)	.004^a	0.31
Male	58.4%	73.4%	.004^a	0.32
African American	48.7%	52.6%	.481	0.08
Current smoker	54.9%	69.6%	.001^a	0.30
ASA physical status			.068
I–II	44.8%	54.9%		0.20
III	55.2%	45.1%		…
Medical history
Hypertension	55.2%	36.4%	.001^a	0.38
Cardiovascular disease	20.1%	16.1%	.355	0.10
Endocrine disease	50.7%	41.0%	.082	0.19
Gastrointestinal disease	48.7%	44.5%	.448	0.08
Renal disease	31.8%	21.4%	.033^a	0.24
Respiratory disease	27.3%	22.0%	.265	0.12
Infectious disease	26.0%	27.2%	.807	0.03
Neoplastic disease	23.4%	19.1%	.341	0.11
Nervous system disease	29.2%	30.1%	.869	0.02
Psychiatric disease	66.9%	85.0%	.0001^a	0.43
Baseline vital signs
Mean arterial pressure (mm Hg)	100	103	.064	0.21
Heart rate (bpm)	72	74	.198	0.14
Preoperative QTc interval			.090
Female (ms)	439 (28)	446 (34)		0.25
Male (ms)	431 (28)	436 (28)		0.18

Data are presented as means and standard deviations or percentages unless otherwise stated.

Abbreviations: ASA, American Society of Anesthesiologists; bpm, beats per minute; QTc, corrected QT interval.

^aP < .05, showing statistical significance.

Patient characteristics are summarized in Table 1. Patients in the cocaine-positive group were significantly younger, had a lower BMI, were more likely to be current smokers, and were more likely to be male compared to those in the cocaine-negative group. More of the cocaine-positive patients had psychiatric disease, whereas hypertension and renal disease were more prevalent among patients in the cocaine-negative group. No other significant differences in medical history were observed between the 2 groups. On the preoperative ECG, the QTc interval did not differ between the 2 groups.

Outcome Measures

The cocaine-positive group had a BP that was outside the set limits 19.4% (SD = 17.7%) of the time versus 23.1% (SD = 17.7%) of the time for the cocaine-negative group (Figure 2). The cocaine-positive group had an HR outside the set limits 9.6% (SD = 16.2%) of the time versus 8.2% (SD = 14.9%) of the time for the cocaine-negative group (Figure 2). Adjusted for demographic and other confounding variables, the percentage of intraoperative hemodynamic events for cocaine-positive and cocaine-negative patients was similar within a 7.5% margin of equivalence for MAP data (β coefficient = 2%, P = .003, CI, 2–6), and within a 5% margin of equivalence for HR data (β coefficient = 0.2%, P < .001, CI, 4–3; Table 2). For the multivariable regression models, the following factors were included as confounding variables: age, sex, BMI, smoking status, presence of comorbid hypertension, renal disease, and psychiatric illness. Statistically significant findings of equivalence between the 2 groups suggest that the cocaine-positive patients experienced an equivalent percentage of intraoperative hemodynamic events as the cocaine-negative group given the a priori–defined equivalence criterion. No gross violations of normality assumption for the outcome measures, as well as regression residuals, were observed. Additionally, the amount of antihypertensives administered to cocaine-positive and cocaine-negative patients was not statistically significant (P = .09), while cocaine-positive patients received significantly less vasopressors intraoperatively (P = .02) compared to patients in the cocaine-negative group (Table 2).

Table 2. - Intraoperative Data

Outcomes	Cocaine Negative (n = 154)	Cocaine Positive (n = 173)	β Coefficient (95% CI)	P
Percentage of hemodynamic events^a
MAP^b	23.1% (17.7%)	19.4% (17.7%)	2% (−2 to 6)	.003^c,d
HR^e	8.2% (14.9%)	9.6% (16.2%)	0.2% (−4 to 3)	<.001^c,d
Medication use
Antihypertensives	8.3%	11.0%	…	.09
Vasopressors	55.8%	42.8%	…	.02^c
New-onset cardiac arrhythmia	0	0	…
Mean MAC	0.9 (0.2)	0.9 (0.2)	…	.21
Mean fentanyl (μg·h⁻¹)^f	74 (56–106)	80 (49–115)	…	.88^g
Procedure type
ENT/OMFS	15 (9.7%)	21 (12.1%)	…	…
Neurology	14 (9.0%)	10 (5.8%)	…	…
Gynecology	20 (12.9%)	11 (6.4%)	…	…
Oncology	6 (3.9%)	8 (4.6%)	…	…
Urology	15 (9.7%)	10 (5.8%)	…	…
General	45 (29.0%)	50 (28.9%)	…	…
Orthopedic	25 (16.1%)	45 (26.0%)	…	…
Plastic surgery	15 (9.7%)	18 (10.4%)	…	…
Surgery duration (min)^f	159 (116–237)	138 (90–203)	…	.01^h
Estimated blood loss (mL)^f	100 (18–350)	75 (10–230)	…	.21

All data are presented as means and standard deviations and numbers and percentages unless otherwise stated.

Abbreviations: BP, blood pressure; CI, confidence interval; ENT, ear nose and throat; HR, heart rate; MAC, minimum alveolar concentration; MAP, mean arterial pressure; OMFS, oral and maxillofacial surgery.

^aPercentage of total anesthesia duration.

^bMAP outcome was defined as the percentage of anesthesia duration that a patient’s MAP was outside the prespecified limits of <65 or >105 mm Hg.

^cP < .025, showing statistical significance after Bonferroni corrections. The reported P value is the larger of the 2 P values resulting from the 2 one-sided tests used to assess equivalence.

^dThe P values for BP and HR are indicative of statistical significance of equivalence using the Mascha and Sessler^¹⁵ method of equivalence with a 7.5% margin for BP and a 5.0% margin for HR. The confounding variables included in the multivariable models were age, sex, body mass index, smoking status, and the presence of comorbid hypertension, renal disease, and psychiatric illness.

^eHR outcome was defined as the percentage of anesthesia duration that a patient’s HR was outside the prespecified limits of <50 or >100 beats per minute.

^fMedian, interquartile range.

^gP value reported from Wilcoxon-Mann-Whitney test for equality of the medians.

^hP < .05, showing statistical significance. The reported P value is from a 2-sample t test for equality against an alternative of inequality.

Figure 2.:
Percentages of case duration with MAP <65 or >105 mm Hg and percentages of case duration with HR <50 or >100 bpm for cocaine-positive and cocaine-negative patients. bpm indicates beats per minute; HR, heart rate; MAP, mean arterial pressure.

No significant differences were observed in the intraoperative cardiac rhythms between the 2 groups. Both groups were predominantly in normal sinus rhythm, followed by sinus tachycardia and sinus bradycardia. No other new-onset arrhythmias (eg, ventricular ectopy) were noted in either group. Mean BIS readings during anesthesia were comparable between the groups and were within the recommended range of 40–60 intraoperatively. No significant differences were observed in the mean MACs of sevoflurane for the 2 groups (Table 2). Patients in the cocaine-positive group had a shorter duration of surgery of 138 minutes (90–203 minutes) compared to patients in the cocaine-negative group of 159 minutes (116–237 minutes; P = .01). Patients in the cocaine-negative group were administered 74 μg (56–106 μg) of fentanyl each hour versus 80 μg (49–115 μg) in the cocaine-positive group (P = .88). The types of surgeries between the 2 groups did not differ significantly (Table 2).

A total of 5 patients had any detectable cTnT in the perioperative period, defined as a value ≥0.01 ng·mL⁻¹. Two patients were from the cocaine-positive group and 3 were from the cocaine-negative group (Table 3).

Table 3. - Patients With Abnormal Perioperative Troponin

Patient	Group	Preoperative cTnT (ng·mL⁻¹)	Postoperative cTnT (ng·mL⁻¹)	Δ cTnT
1	Negative	<0.01	0.05	Increase
2	Negative	0.03	0.02	Decrease
3	Negative	0.01	<0.01	Decrease
4	Positive	0.02	0.02	No change
5	Positive	0.02	<0.01	Decrease

Abbreviation: cTnT, cardiac troponin T.

Table 4. - Use of Cocaine Within 24 h of Surgery Versus More Than 24 h Before Surgery

Cocaine Use	Within 24 h of Surgery (n = 17)	More Than 24 h Before Surgery (n = 156)	P
Percentage of hemodynamic events^a
MAP^b	18.1% (16.7%)	19.1% (17.2%)	.82^c
HR^d	7.3% (9.4%)	9.2% (15.6%)	.49^c

Data are presented as means and standard deviations.

Abbreviations: HR, heart rate; MAP, mean arterial pressure.

^aPercentage of total anesthesia duration.

^bMAP outcome was defined as the percentage of anesthesia duration that a patient’s MAP was outside the prespecified limits of <65 or >105 mm Hg.

^cP value from 2-sample t tests for equality of the cocaine-positive and cocaine-negative groups in terms of respective outcomes.

^dHR outcome was defined as the percentage of anesthesia duration that a patient’s HR was outside the prespecified limits of <50 or >100 beats per minute.

To explore how the timing of last cocaine use affected hemodynamics, we divided the cocaine-positive subjects into those who reported using cocaine within 24 hours before surgery versus those who reported using cocaine more than 24 hours before surgery (based on a self-reported survey). There were 17 cocaine-positive subjects who reported using cocaine within 24 hours before surgery versus 156 who reported cocaine use more than 24 hours before surgery. We compared these 2 groups in terms of the 2 primary outcomes and observed no statistically significant differences between the groups (P = .82 for BP outcome and 0.49 for HR outcome), which are detailed in Table 4.

DISCUSSION

In this study of 327 patients with a history of cocaine use in the past 5 years, we demonstrated that asymptomatic, cocaine-positive patients with a normal preoperative ECG and vital signs have an equivalent percentage of intraoperative hemodynamic events compared to cocaine-negative patients under general anesthesia. Thus, we did not observe any hemodynamic effects that could be generated by persistent, excessive sympathetic tone or catecholamine depletion in the cocaine-positive group. Furthermore, the cocaine-positive group did not receive greater administration of vasoactive agents intraoperatively, suggesting that the similar percentages of hemodynamic events was not due to differences in vasoactive medication administration. The literature describing the effect of recent cocaine use on intraoperative hemodynamics during general anesthesia is limited, but the results of this study are consistent with previous studies.^²^,^¹⁶^,^¹⁷

QT interval prolongation is a known effect of cocaine use and can increase the risk of ventricular arrhythmias.^¹⁸^,^¹⁹ Cocaine use also increases myocardial oxygen demand and can result in acute cardiovascular dysfunction (eg, myocardial ischemia). Cocaine-positive patients who do not manifest any of the signs or symptoms of cocaine intoxication may be seen as having passed this “stress test” on the cardiovascular system. The results of this study indicate that the preoperative ECG can play an important role in the clinical decision-making of whether to proceed with surgery in cocaine-positive individuals. A normal preoperative ECG and QTc in the setting of a well-appearing patient with normal vital signs can serve as a reassuring factor when considering whether to proceed with surgery.

A secondary exploratory aim was to see if the cocaine-positive and cocaine-negative groups had differences in perioperative cTnT values. If recent cocaine use caused myocardial ischemia, one might expect to see a difference in the preoperative troponin values between groups. Alternatively, if recent cocaine use is associated with increased sympathetic tone contributing to myocardial ischemia in the perioperative period, one might expect to see a difference in the postoperative troponin values between groups. However, no significant preoperative to postoperative elevations in cTnT were noted, except for one cocaine-negative patient undergoing a cardiac electrophysiology procedure who had an increase in cTnT from <0.01 ng·mL⁻¹ preoperatively to 0.05 ng·mL⁻¹ postoperatively. On discussion with the electrophysiologists, it was determined that the cause was likely due to the intracardiac nature of the procedure, as the patient had no other signs or symptoms of ischemia. Although the study was not powered for this secondary exploratory outcome, we did not find evidence that subclinical myocardial ischemia occurred at an increased rate in cocaine-positive patients based on the cTnT analyses performed.

Patients in the cocaine-positive group received fewer vasopressor administrations than patients in the cocaine-negative group. Patients in the cocaine-negative group had a higher prevalence of hypertension, which is associated with BP volatility under general anesthesia and may have contributed to this group’s higher need for vasopressors intraoperatively.^²⁰^,^²¹ The amount of antihypertensive medications administered intraoperatively was not different between the cocaine-positive and cocaine-negative groups.

Previous animal studies have suggested that chronic cocaine use can increase inhaled anesthetic requirements.^²²^,^²³ Another secondary aim of this study was to evaluate whether cocaine-positive patients had significantly altered anesthesia requirements during a standardized anesthesia protocol with continuous BIS monitoring. Our findings show that the MAC of sevoflurane and total amounts of anesthetic and analgesic medications were comparable between groups. These findings corroborate another observational study, which reported that recent cocaine users did not have significantly altered sedative or opioid requirements compared with cocaine-negative patients.^²⁴

Limitations of our study include the single-center study design and unblinded nature of the study (ie, the clinical teams were aware of whether patients were cocaine-positive versus cocaine-negative). Thus, it is possible that the surgical teams may have been consciously or unconsciously biased toward certain actions, such as completing the surgeries more quickly in the cocaine-positive group. Given that participation in the study was voluntary, 17% of subjects who were approached to participate in the study declined, so the cocaine urine toxicology results and usage habits of these patients are unknown. It is possible that some of those patients would have been cocaine positive and differed in some manner from the cocaine-positive patients who did agree to participate in the study. Since patients with end-stage renal disease, seizures, coronary artery disease, and cardiac arrhythmias were excluded, our results may not be applicable to patients with these comorbidities. It is also possible that the study conclusions are confounded by measured or unmeasured variables that may differ between cocaine-positive and cocaine-negative patients. Additionally, we did not include high-risk surgeries with expected hemodynamic disturbances or significant blood loss, and thus these results may not be applicable in those circumstances. It is possible that our sample size was too small to capture rare events, such as new intraoperative arrhythmias or perioperative myocardial ischemia. Since the cocaine-negative cohort in this study was drawn from individuals who had a history of cocaine abuse within the past 5 years, the results may not be generalized to populations without a history of cocaine use. Additionally, subjects who were not expected to stay overnight did not receive troponin T measurements, so this secondary outcome is purely exploratory, and definitive conclusions may not be drawn from these data.

The implications of these findings are noteworthy, as they challenge the historical precedent that a patient with a positive cocaine urine toxicology test should have their elective surgery canceled.^⁹^,^²⁵ The results of this study suggest that some patients with a positive cocaine urine toxicology test who do not have signs or symptoms of acute intoxication may not exhibit a significantly altered intraoperative hemodynamic course compared to cocaine-negative patients. As such, the routine cancellation of elective cocaine-positive patients may not be warranted. This may especially be true for surgeries that are not urgent but not wholly elective either (ie, hysterectomy for a malignant mass). The most significant effects of chronic cocaine use may be accelerated atherosclerosis, development of left ventricular hypertrophy, and dilated cardiomyopathy.^{^26–28} Thus, it may be prudent to treat chronic cocaine use akin to other cardiovascular risk factors, such as coronary artery disease or diabetes mellitus.^²⁷^,^²⁹ Appropriately screened cocaine-positive patients may be able to undergo noncardiac surgery with a level of intraoperative risk that does not differ significantly from that for cocaine-negative patients. In general, appropriate screening measures include the following: (1) a detailed history and physical with assessment for signs and symptoms of acute intoxication, (2) an evaluation of concurrent comorbidities and surgical risk, and (3) a preoperative ECG to rule out QTc interval prolongation and arrhythmias. If the aforementioned criteria are favorable, then the patient may indeed be a suitable surgical candidate.

In conclusion, the presence of a positive cocaine urine toxicology test may not be an absolute contraindication to surgery if the patient has a normal ECG and vital signs and is asymptomatic with minimal comorbidities. Hospitals that serve a population in which cocaine use occurs with a high frequency may find this information valuable to advise their preoperative screening protocols and institutional practices, since very few clinical guidelines currently exist.^⁹ Additional large prospective studies evaluating the hemodynamics of cocaine-positive versus cocaine-negative patients, as well as short- and long-term outcomes, will help to corroborate and expand on the results of this study.

ACKNOWLEDGMENTS

Christopher Clark (Research Data & Analytics Specialist, Parkland Health & Hospital System, Dallas, Texas): this acknowledged contributor helped obtain study data from the electronic medical record; Anna Barden, RN, MHR, MBA, CPHRM, CHRC (Director Research Compliance, Parkland Health & Hospital System, Dallas, Texas): this acknowledged contributor helped oversee the regulatory requirements for this study; Shaina Drummond, MD (Department of Anesthesiology and Pain Management, University of Texas Southwestern, Dallas, Texas): this acknowledged contributor helped with patient enrollment; Brian Farrell, CRNA (Department of Anesthesiology and Pain Management, Parkland Health & Hospital System, Dallas, Texas): this acknowledged contributor helped with patient enrollment; Charles Koshy, MD (Department of Anesthesiology and Pain Management, University of Texas Southwestern, Dallas, Texas): this acknowledged contributor helped with patient enrollment; Emily Melikman (Department of Anesthesiology and Pain Management, University of Texas Southwestern, Dallas, Texas): this acknowledged contributor helped with regulatory oversight; Mary Christi Sunna, CRNA (Department of Anesthesiology and Pain Management, Parkland Health & Hospital System, Dallas, Texas): this acknowledged contributor helped with patient enrollment; Amy Woods, MD (Department of Anesthesiology and Pain Management, University of Texas Southwestern, Dallas, Texas): this acknowledged contributor helped with patient enrollment.

DISCLOSURES

Name: Tiffany S. Moon, MD.

Contribution: This author helped with study design, regulatory oversight, data collection, interpretation of results, manuscript drafting, statistical analysis, and revision of the manuscript.

Name: Taylor J. Pak, BS.

Contribution: This author helped with regulatory oversight, interpretation of results, manuscript drafting, and revision of the manuscript.

Name: Agnes Kim, MS.

Contribution: This author helped with study design, data collection, interpretation of results, and revision of the manuscript.

Name: Michael X. Gonzales, MD.

Contribution: This author helped with study design, data collection, interpretation of results, manuscript drafting, statistical analysis, and revision of the manuscript.

Name: Yuri Volnov, MD.

Contribution: This author helped with patient recruitment, data collection, preliminary data analysis/interpretation, and early abstract formulation and presentation.

Name: Evan Wright, MD.

Contribution: This author helped with study design, data collection, interpretation of results, and revision of the manuscript.

Name: Kevin Q. Vu, BA.

Contribution: This author helped with study design, patient recruitment, data collection, interpretation of the results, and revision of the manuscript.

Name: Rachael D. Lu, BS.

Contribution: This author helped with patient recruitment, data collection, and revision of the manuscript.

Name: Arghavan Sharifi, BS.

Contribution: This author helped with patient recruitment, data collection, data recording, and revision of the manuscript.

Name: Abu Minhajuddin, PhD.

Contribution: This author helped with study design, statistical analysis, manuscript drafting, and revision of the manuscript.

Name: Joy L. Chen, MD.