Prevalence of cocaine use in ED patients with severe hypertension
The prevalence of cocaine use in the US population is reportedly rising. Recent data from the Drug Abuse Warning Network estimates that cocaine is associated with 20% of drug-related emergency department (ED) visits. In addition to patients with clinical evidence of cocaine toxicity, many patients present with complications of cocaine use that are not associated with the sympathomimetic toxidrome that characterizes severe toxicity.
The presence of cocaine may be clinically unsuspected in patients with chest pain or isolated seizures. In addition, physicians often fail to ask about cocaine use. In one study, documentation of the presence or absence of cocaine use was recorded in only 13% of patients with chest pain.Even when asked, patients are often reluctant to acknowledge the use of cocaine. In one study of ED patients who presented with chest pain, only 72% of the 17% of patients who tested positive for cocaine admitted to recent use.
β-Adrenergic receptor blocking drugs are commonly used in the ED and outpatient management of both urgent and nonurgent hypertension, ischemic coronary syndromes, and congestive cardiac failure. Serious side effects that include coronary artery vasoconstriction and paradoxical hypertension may be precipitated by use of these drugs in the presence of cocaine and. Therefore, identification of cocaine use impacts therapeutic decision making for many ED patients.
The objective of this study is to determine the prevalence of cocaine use in one ED by patients with significant hypertension who do not have clinical indicators that suggest cocaine toxicity.
We prospectively evaluated the prevalence of cocaine use in emergency department patients with elevated diastolic blood pressure greater than 120 mm Hg.
Urine was collected on adult patients with diastolic blood pressure greater than 120 mm Hg. The Institutional Review Board granted a waiver of informed consent based on complete deidentification of the data to prevent selection bias. Spontaneously voided urine specimens were analyzed for the presence of the benzoylecgonine.
Of 107 eligible patients with a diastolic blood pressure equal to or greater than 120 mm Hg, 99 were tested. Of the 99, 13 (13%) tested positive, and of which, 6 (46%) presented with a request for medication refill. Five patients (39%) had cardiovascular and/or pulmonary complaints.
Emergency department patients with hypertension who have recently used cocaine may be difficult to differentiate from those who have not. Consideration should be given to the potential for cocaine use when making treatment decisions in patients with hypertension.
2.1. Study design
This was a prospective, single-blind, observational study of cocaine use in ED patients who presented with a diastolic pressure greater than 120 mm Hg. The choice of diastolic blood pressure greater than 120 mm Hg was based on review of literature addressing severe hypertension, most commonly defined as diastolic blood pressure greater than 120 to 130 mm Hg. The study was approved by the local Institutional Review Board, who waived consent based on the complete deidentification of the data and the concern that the requirement for consent would impede the ability to conduct a meaningful study. Recorded data included age, admission diastolic pressure, chief complaint, and results of a deidentified urine screen for cocaine and cocaine metabolites. It did not include sex, ethnicity, questioning about the use of cocaine, or management or course in the ED. Study investigators were physically present in the ED 24 hours a day, with the exception of Saturday and Sunday overnights for an 8-week study period. They did not provide medical care to the patients. Patients with a diastolic blood pressure greater than 120 mm Hg were identified by continuous monitoring of the electronic ED data base (EmStat), which provides “real-time” admission vital sign data. Patients whose chief triage complaint was related to cocaine use (self-report of recent cocaine use or affirmation of recent cocaine use when queried by the triage nurse) as documented in the triage note in EmStat were excluded. In addition, one of the exclusion criteria was suspected cocaine toxicity identified by the triage nurse as a sympathomimetic toxidrome with associated agitation, aggressive behavior, or physical hyperactivity with documentation in the triage record of suspected cocaine intoxication. Patients younger than 18 years, those with a history of end-stage renal disease (ESRD), pregnancy, cognitive impairment, and patients who were unable to provide a spontaneously voided urine sample were excluded. Study investigators saved part of a spontaneously voided urine sample provided for ED testing and recorded study information on an identically numbered deidentified data sheet. Because the results of the study test were not available for use by the clinicians, toxicologic testing for clinical purposes was conducted separately.
2.2. Study setting
This study took place in an urban teaching hospital that triages more than 100,000 ED visits per year.
2.3. Data collection
Urine specimens were analyzed for the presence of the cocaine metabolite benzoylecgonine, using the ONLINE cocaine II qualitative assay. The cutoff limit for benzoylecgonine for this test is 150 ng/mL. Positive results underwent confirmatory testing with gas chromatography-mass spectrometry to ensure there were no false positives. Study characteristics (age, diastolic pressure, and chief complaint) of groups testing positive and negative for cocaine are reported.
The presence or absence of benzoylecgonine was the main outcome parameter. Characteristics of the 2 cohorts were compared using a 2-sided exact test for the one categorical variable. Continuous variables are presented as mean ± SD, and P values were calculated using a Mann-Whitney test for nonparametric variables.
One hundred seven patients with a diastolic blood pressure equal to or greater than 120 mm Hg were identified over an 8-week period in winter 2005. Eight of these patients were excluded. There were 4 who had ESRD, 1 who had an allergic reaction and could not void, 3 who did not undergo urine testing, 1 who presented for a wound check, and 2 who had headaches. No patient in our study set met the exclusion criteria for self-reported cocaine use or sympathomimetic toxidrome suggestive of cocaine intoxication.
Data sheets were missing for 3 patients. These 3 urine screens (all negative for cocaine) were included in the prevalence determination (n = 99) but excluded from comparison of the cohorts (n = 96). Of the 99 patients included in the prevalence determination, 13 (13.1%) tested positive for cocaine both on initial and confirmatory testing. There was no difference between the cohorts with regard to age, diastolic blood pressure, or chief complaint. The mean age for cocaine-positive patients was 38 ± 7 versus 44 ± 12 for those negative for cocaine (P = .09). The diastolic blood pressures for cocaine-positive patients ranged from 120 to 159 mm Hg (mean, 131 ± 12), whereas those in the cocaine-negative group ranged from 120 to 169 mm Hg (mean, 132 ± 12) (P = .3).
The chief complaints or reasons for the ED visits are detailed in Table 1. Six (46%) of the patients who tested positive for cocaine presented with a request for a medication refill. Five patients (39%) had cardiovascular and/or pulmonary complaints. The percentage of cocaine-positive patients who came for minor interventions was higher than those with negative cocaine screens; however, this difference was not significant.
Our study investigators were not present Saturday and Sunday overnights, which may have resulted in missed enrollments. These periods represent less than 10% of the total time frame in which the study was conducted. Study investigators were unable to inquire regarding drug use; thus, we cannot determine what percent of the cocaine-positive patients would have provided this information with careful history-taking by the physician. Our study population was limited to an urban county hospital, which may have a different prevalence of cocaine using patients than other practice settings. There may also be regional differences in drugs of abuse patterns that affect the ability to generalize our data. However, our findings were consistent with those described in the study of chest pain patients by Hollander et al, which included geographically diverse hospitals and urban and suburban practice settings.
Our study showed a prevalence of cocaine use of 13% in patients with hypertension who presented to the ED with common complaints not obviously related to cocaine use. Our patients included a mixture of patients who might normally be followed in an outpatient clinic setting and patients with more urgent complications of hypertension including chest pain and dyspnea. Remarkably, 46% of cocaine-positive patients presented with complaints of “high blood pressure” and/or need for medication refill, complaints that in other settings would not be seen in the ED. Thirty-eight percent had chest pain or symptoms of congestive heart failure that might have resulted in the acute ED administration of β-adrenergic receptor antagonist drugs. All others might have had β-blocking drugs prescribed for outpatient therapy. Administration of these agents in the presence of recent cocaine use has been shown to precipitate coronary vasospasm and paradoxical increases in blood pressure. Our concerns may be reasonably extrapolated to the outpatient setting, where β-blockers are commonly used to manage chronic hypertension.
Cocaine use in the United States continues to rise, and cardiovascular complications related to its use such as myocardial infarction, cardiomyopathy, and cerebral vascular accidents are well documented. Cocaine use has also been strongly linked to hypertension-related ESRD, but there is little data describing cocaine use in hypertensive patients that have not yet progressed to ESRD. To our knowledge, our study is the first that attempts to measure the prevalence of cocaine use among hypertensive patients. Considering the Drug Abuse Warning Network data that suggest that 20% of ED cases are associated with cocaine use, we expected a higher percentage in our inner city population. Because we included only patients who did not have a chief complaint or clinical evidence of cocaine toxicity, we suspect that these patients were likely chronic hypertensives who also used cocaine, although we cannot state that with certainty based on our data. Urine drug screens for cocaine can remain positive for several days after use so one cannot infer that hypertension in our study patients was caused by cocaine use. However, given the prevalence of cocaine-positive patients in our study population, we suggest that physicians who consider using a β-adrenergic blocking drug for the management of ED patients should consider the possibility of underlying cocaine use.
The data from this study show that hypertensive patients without other signs of cocaine toxicity may have cocaine metabolites in their urine consistent with recent cocaine use. Emergency department providers caring for patients with hypertensive urgencies should take a thorough medical history to screen for cocaine use and consider objective testing for cocaine metabolites when making treatment and disposition decisions that may be altered by knowledge of cocaine use. Studies of cocaine use by patients followed in primary care settings, especially in communities where cocaine-related ED visits are frequent, may alter prescribing practices or indicate a need for testing before selection of a regimen for management of hypertension. Our study underlines the need for physician awareness of cocaine use and its relevance to patient care both in emergency and primary practice settings.
The authors would like to acknowledge Jody Balko, Chief Medical Pathologist, Department of Pathology, UT Southwestern Medical School for her laboratory assistance and expertise.
 SAMHSA Administration, Results from the 2002 National Survey on Drug Use and Health: National Findings. In: SAMHSA, Editor, NHSDA Series H-22 (2002)
 Services USDoHaH, Drug Abuse Warning Network 2003: national estimates of drug related emergency department visits, DHHS, Rockville (MD) (2004).
 J.E. Hollander, D.E. Brooks and S. Valentine, Assessment of cocaine use in patients with chest pain syndromes, Arch Intern Med 158 (1998), pp. 62–68.
 J.E. Hollander, K.H. Todd and G. Green et al., Chest pain associated with cocaine: an assessment of prevalence in suburban and urban emergency departments, Ann Emerg Med 26 (1995), pp. 671–676.
 R.A. Lange, R.G. Cigarroa and E.D. Flores et al., Potentiation of cocaine induced coronary vasoconstriction by beta-adrenergic blockade, Ann Intern Med 112 (1990), pp. 897–903.
 I.C. Sand, S.L. Brody and K.D. Wrenn et al., Experience with esmolol for the treatment of cocaine associated cardiovascular complications, Am J Emerg Med 9 (1991), pp. 161–163.
 E.B. Healton, J.C. Brust and D.A. Feinfeld, Hypertensive encephalopathy and the neurologic manifestations of malignant hypertension, Neurology 32 (1982), pp. 127–132.
Melissa L. Givens MD, Rebecca Wald MD, Jason Schafer MD, Frank Wians Jr PhD and Kathleen Delaney MD
Department of Emergency Medicine, Madigan Army Medical Center, Tacoma, WA 98431, USA
University of Texas Southwestern Medical Center, Dallas, TX 75235, USA