Utility of Routine Testing for Patients With Asymptomatic Severe
Blood Pressure Elevation in the Emergency Department
David J. Karras, MD Linda K. Kruus, PhD John J. Cienki, MD Marlena M. Wald, MLS, MPH Jacob W. Ufberg, MD Philip Shayne, MD David A. Wald, DO Katherine L. Heilpern, MD
From the Department of Emergency Medicine, Temple University School of Medicine, Philadelphia, PA (Karras, Kruus, Ufberg, Wald DA); the Department of Emergency Medicine, Emory University School of Medicine, Atlanta, GA (Wald MM, Shayne, Heilpern); and the Department of Emergency Medicine, Jackson Memorial Hospital–University of Miami School of Medicine, Miami, FL (Cienki).
Study objective:Recommendations for the treatment of emergency department (ED) patients with asymptomatic severely elevated blood pressure advise assessment for occult, acute hypertensive target-organ damage. This study determines the prevalence of unanticipated, clinically meaningful test abnormalities in ED patients with asymptomatic severely elevated blood pressure.
Methods:This was a prospective observational study at 3 urban academic EDs. Consecutive
patients with systolic blood pressure greater than or equal to 180 mm Hg or diastolic blood pressure greater than or equal to 110 mm Hg on 2 measurements were enrolled if they denied symptoms of hypertensive emergency. A basic metabolic panel, urinalysis, ECG, CBC count, and chest radiograph were obtained. Treating physicians were interviewed about the indication for each test and whether an abnormal result was anticipated according to clinical findings. When test results were available, physicians were asked whether abnormal findings were clinically meaningful, defined as leading to unanticipated hospitalization, medication modification, or further immediate evaluation. The primary outcome was the prevalence of unanticipated clinically meaningful test abnormalities.
Results:One hundred nine patients with asymptomatic severely elevated blood pressure were enrolled. Unanticipated abnormal test results were noted in 57 (52%) patients. Clinically meaningful unanticipated test abnormalities were found in 7 (6%) patients: basic metabolic panel in 2 (2%), CBC count in 3 (3%), urinalysis in 3 (4%), ECG in 2 (2%), and chest radiograph in 1 (1%). Five patients (5%) had abnormalities assessed as possible manifestations of acute hypertensive target-organ injury; none had abnormalities clearly related to severely elevated blood pressure.
Conclusion:Screening tests of urban ED patients with asymptomatic severely elevated blood pressure infrequently detect unanticipated hypertension-related abnormalities that alter ED management. [Ann Emerg Med. 2007;xx:xxx.]
0196-0644/$-see front matter
Copyright©2007 by the American College of Emergency Physicians. doi:10.1016/j.annemergmed.2007.03.032
INTRODUCTION Background
Severely elevated blood pressure is commonly, though arbitrarily, defined as a systolic blood pressure greater than or equal to 180 mm Hg or diastolic blood pressure greater than or equal to 110 mm Hg.1About 5% of adult emergency
department (ED) patients have at least 1 blood pressure measurement that is severely elevated.2,3Patients with severely
elevated blood pressure are considered at risk for acute hypertensive target-organ injury, which may be manifested by encephalopathy, cerebral ischemia or hemorrhage, coronary ischemia, myocardial dysfunction with pulmonary edema, aortic dissection, acute renal insufficiency, or funduscopic changes. A hypertensive emergency is defined as the presence of any of these abnormalities in a patient with severely elevated blood pressure; such individuals require careful blood pressure
reduction to limit target-organ damage.1,4-6Patients with severely elevated blood pressure values but who have no evidence of acute target-organ injury have a hypertensive urgency, which generally requires modification of the patient’s oral antihypertensive medication regimen and prompt outpatient follow-up.4,5
Importance
Emergency physicians commonly evaluate patients with severely elevated blood pressure values who manifest no symptoms of hypertensive emergency. Standard emergency medicine textbooks and reviews recommend that these patients undergo careful assessment in the ED to exclude occult hypertension-related acute target-organ injury, and diagnostic testing is advised if hypertensive emergency is a diagnostic consideration.1,4,5Testing recommendations include a basic metabolic panel, urinalysis, ECG, a CBC count, and chest radiograph. These recommendations have been challenged because no evidence-based data are available to assess their utility in patients with asymptomatic hypertension.1,5Previous investigations have documented that physicians at academic programs rarely perform a comprehensive assessment for acute hypertensive target-organ injury and that most emergency physicians do not believe that extensive diagnostic evaluation is routinely necessary.7,8
Goals of This Investigation
This study was developed to determine how frequently screening tests for acute hypertensive target-organ damage reveal
abnormalities that were not anticipated on the basis of the patient’s medical history or examination findings and to assess how often unanticipated test abnormalities were clinically meaningful, defined as resulting in changes to the patient’s ED treatment.
MATERIALS AND METHODS Study Design
This was a prospective, observational, cross-sectional study of adult ED patients presenting to 1 of 3 sites with
asymptomatic, severely elevated blood pressure. The
institutional review board of each participating site approved the study.
Setting
Demographic characteristics of the participating EDs are presented in Table 1. Each site is located in an inner-city area, serves largely economically disadvantaged and uninsured patients, and is the primary teaching hospital of a medical school. All are staffed exclusively or primarily by physicians board certified or board prepared in emergency medicine, and 2 of the 3 sites support emergency medicine residency training programs. The combined annual census of the participating sites is 295,000.
Selection of Participants
The study was conducted during 4 weeks in the summer of 2004 and subsequently extended for another 4 weeks in the summer of 2005. Research associates screened ED patients for enrollment between 8AMand midnight, the peak patient arrival
times at the participating sites. Vital signs of all patients at least 18 years of age and presenting to a study site during enrollment periods were reviewed on arrival and every 2 hours throughout the ED stay. Patients with a systolic blood pressure greater than or equal to 180 mm Hg or diastolic blood pressure greater than or equal to 110 mm Hg at any time during the ED visit were eligible for the study.
Patients were excluded from the study if they had an acute illness or injury requiring immediate intervention to prevent or Editor’s Capsule Summary
What is already known on this topic
Textbooks recommend the laboratory screening of asymptomatic patients with very high blood pressure for evidence of end-organ damage.
What question this study addressed
Whether screening such patients results in clinically meaningful changes in management.
What this study adds to our knowledge
Fifty-seven of the 109 patients had at least 1 abnormal test result. Five patients had abnormalities conceivably related to the elevated blood pressure. No patient had laboratory findings indicative of a hypertensive emergency.
How this might change clinical practice
Although screening tests are recommended, there has never been any evidence about whether patients with asymptomatic hypertension require them. These data support the concept that screening is highly unlikely to disclose results leading to changes in emergency department management.
Table 1.Characteristics of study sites.
Site Location Temple University Hospital (Philadelphia, PA) Grady Hospital (Atlanta, GA) Jackson Memorial Hospital (Miami, FL) Patients enrolled in study 32 24 53 Approximate annual adult visits 57,000 110,000 130,000 Patient ethnicity, % Black 72 80 52 Latino 14 6 34 White, non-Latino 8 10 14 Other 6 4 0 Uninsured patients, % 22 65 62
limit a life-threatening event, consist with a triage category of “emergent” or immediate initiation of medical therapy. An exception was made when the patient was classified as “emergent” or medical therapy was immediately undertaken only because the patient had asymptomatic blood pressure elevation; such patients were eligible for enrollment. Additional exclusion criteria were acute injury requiring activation of the trauma system, acute head injury, and acute spinal cord injury. Consent forms were available in English and Spanish;
non–English-speaking patients were excluded if they spoke a language other than Spanish or if the research associate was unable to communicate with the help of an interpreter. Patients under the care of a study investigator were not eligible for enrollment. No further data were collected on excluded patients.
Data Collection and Processing
An overview of the study protocol is depicted in Figure 1. Eligible patients were asked to provide written informed consent for study participation. Consenting patients underwent a brief questionnaire about demographic information and medical history. Patients were asked whether they had any symptoms in the past 24 hours suggestive of acute hypertensive target-organ damage, specifically, chest discomfort, shortness of breath, confusion or other change in mental status, numbness or weakness on one side of the body, severe headache, or nosebleed. Patients reporting any of these symptoms were excluded from further participation in the study. Patients were also excluded if they reported a history of head or spinal cord injury with persistent neurologic deficits, if they had undergone general anesthesia in the past week, or if they were pregnant. No further data were collected on excluded patients.
After the interview, patients underwent repeated blood pressure measurement by a research associate at least 30 minutes after the initial vital sign assessment. The patient was taken to a quiet area adjacent to the waiting room or to a chair in the treatment area and allowed to sit for 5 minutes. After the guidelines of the “Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and
Treatment of High Blood Pressure (JNC-7),”9the second blood pressure reading was measured using a manual blood pressure cuff, with the patient seated in a chair, feet on the floor, and arm supported at heart level. Patients with a repeated systolic blood pressure measurement less than 180 mm Hg and diastolic blood pressure less than 110 mm Hg were excluded from further data collection. Study patients whose second blood pressure reading remained severely elevated underwent tests recommended by current guidelines for detecting acute
hypertensive target-organ damage in asymptomatic patients.1,4-6 These studies were a basic metabolic profile (electrolyte, glucose, blood urea nitrogen, and creatinine levels), CBC count, ECG, urinalysis, and chest radiograph.
The patient’s treating physician was interviewed by the research associates after tests were ordered but before results were available. Associates interviewed the attending physician or
emergency medicine resident in the second, third, or fourth postgraduate year who was directly involved in the patient’s ED care. Interns (first postgraduate year) and medical students were not interviewed. Physicians were asked to choose a single reason for ordering each test from among the following options: (1) the medical history suggested that the test would be abnormal, (2) the history of present illness or physical examination suggested that the test would be abnormal, or (3) the test was obtained only because it was part of the recommended evaluation of patients with severely elevated blood pressure.
A second interview of the treating physician was conducted when study results were available. If any abnormal test result was noted, physicians were asked whether the result altered their medical decisionmaking. Response options were (1) yes, the patient was admitted specifically because of the result; (2) yes, further testing or consultation was ordered specifically because of the result; (3) yes, medications were initiated or changed specifically because of the test results; or (4) no, test results did not alter management.
Research associates were medical and other health care students trained by study investigators to enroll patients, collect study data, and measure blood pressure in accordance with guidelines contained in the JNC-7 report.9They were instructed in good research practices according to each participating site’s institutional requirements and certified by their respective institutional review boards in National Institutes of Health–mandated human subjects’ protection training. The associates had no clinical responsibilities during data collection periods. Before study initiation, all physicians scheduled to work in the ED during study periods provided written consent to be interviewed about their medical practices.
All patient treatment in the ED was performed at the discretion of the treating physicians; the protocol did not restrict or mandate additional testing, blood pressure treatment, or disposition decisions. At the conclusion of the study, patients were provided with a 1-page hypertension fact sheet, which included recommendations to follow up with a primary care physician and information about signs of hypertension-related illnesses.
Completed data collection instruments were forwarded to the coordinating site (Temple University School of Medicine) and reviewed for completeness. Site investigators were queried to provide any missing information and to resolve
inconsistencies in the data collection instruments. Data were entered into a database (SPSS, version 12.0; SPSS, Inc., Chicago, IL) by using double-data entry to verify accuracy. Outcome Measures
The primary outcome was the frequency of clinically meaningful unanticipated abnormalities in each test of acute hypertensive target-organ dysfunction among patients with severely elevated blood pressure who denied having hypertension-related symptoms. Unanticipated test
abnormalities were those in which the treating physician stated that no elements of the patient’s medical history, current
ED presentation Initial BP assessment: SBP ≥180 mm Hg or DBP ≥110 mm Hg Yes No Excluded Immediate medical care necessary
•Triage category “emergent” • Trauma system activated • Medical intervention immediately initiated
(except for isolated BP elevation)
Yes
No Excluded
Screening questionnaire: • Hypertension-associated symptoms • History of brain or spinal cord injury
• Recent general anesthesia • Pregnant
No Excluded
Yes
Repeat BP assessment: SBP ≥ 180 mm Hg or DBP ≥110 mm Hg
Written informed consent
No Excluded Yes Physician interview 1 Physician interview 2 Routine testing: BMP, CBC, ECG, UA, CXR
Final Disposition Hypertension fact sheet provided
Study complete
Figure 1.Study protocol overview.SBP,Systolic blood pressure;DBP, diastolic blood pressure;BMP,basic metabolic panel (electrolytes, blood urea nitrogen, creatinine);UA, urinalysis;CXR,chest radiograph.
symptoms, or examination suggested the presence of an
abnormality and the test was obtained only because the patient’s blood pressure was severely elevated. A clinically meaningful abnormality was defined as an abnormal result leading to hospitalization, further testing or consultation, or modification of the patient’s medication regimen. Using definitions of hypertensive emergency contained in standard emergency medicine textbooks and Joint National Committee guidelines, 2 investigators (J.W.U., D.A.W.) otherwise blinded to the data and results independently assessed the abnormalities and categorized them as being “likely,” “possible,” or “unlikely” manifestations of acute hypertensive target-organ
dysfunction.4,5,9Disagreements between the assessors were to be resolved by consensus. Secondary analyses examined the frequency of clinically meaningful unanticipated screening test abnormalities among specific patient subgroups, stratified by demographic characteristics and medical history.
There are no previous studies of the incidence of clinically meaningful laboratory abnormalities in patients with
asymptomatic severely elevated blood pressure from which we could base sample size calculations. Estimating a risk of a clinically meaningful abnormal result to be 1%, we would need to evaluate 300 patients to yield a 95% chance of detecting at least 1 event. From previous investigations by our study group, we estimated that 400 patients meeting enrollment criteria would present to the participating EDs during a 4-week enrollment period, of whom approximately 300 would present during times research associates were present.
Primary Data Analysis
Frequency counts are reported with percentages and 95% confidence intervals (CI).
RESULTS
During the cumulative enrollment periods, 409 patients with severely elevated blood pressure presented to the participating EDs and were screened for study inclusion. Figure 2 depicts the enrollment profile. The unexpectedly low enrollment rate was primarily due to patients’ frequent endorsement of symptoms suggestive of hypertensive emergency within 24 hours before ED presentation. Because enrollment was much less than anticipated during the planned 4-week enrollment period, a second 4-week period was opened to allow each study site to achieve its enrollment goal. Ultimately, 109 patients with asymptomatic severe blood pressure elevation were included in the final data analysis. Demographic characteristics of enrolled patients are shown in Table 2.
The physician interviewed was the attending physician in 82 cases (74%) and the resident physician in 29 cases (26%). Results of the initial physician interview are shown in Table 3. The most common reason for ordering each test was for routine assessment of the patients’ severely elevated blood pressure, cited by physicians as the only indication for obtaining the test in 55% to 76% of studies ordered.
Table 4 shows the results of the second physician interview, conducted when test results were available. About half of subjects had 1 or more unanticipated abnormal test results. Clinically meaningful unanticipated abnormalities were noted in 1% to 4% of tests. Cumulatively, unanticipated test abnormalities resulted in changes to the management of 7 patients (6%) with asymptomatic severely elevated blood pressure. Table 5 lists the clinically meaningful unanticipated test abnormalities in these patients, the patients’ discharge diagnosis, and assessment of whether the abnormalities
represented manifestations of acute hypertensive target-organ dysfunction. Of the patients in this group, 2 had test abnormalities considered unlikely to be related to the severely elevated blood pressure. The remaining 5 had abnormalities that were possible manifestations of acute hypertensive target-organ injury; none had abnormalities considered likely to be related to severely elevated blood pressure. There was complete agreement between the independent assessors of this outcome.
Secondary analyses were performed to determine whether any screening test could be safely eliminated in selected patients according to reported medical history. An acceptable testing model would reduce the number of tests obtained in patient subgroups without missing any clinically meaningful unanticipated abnormalities. No acceptable testing strategy could be developed with elements or any combination of elements from the patient’s demographic information and medical history.
LIMITATIONS
The study’s primary outcome was prevalence of clinically meaningful unanticipated test abnormalities. Because we provided no guidelines for decisionmaking in response to any test abnormality, this outcome was determined by the treating physicians’ assessment. It is possible that different physicians would have made different management decisions in response to the same abnormal findings. We did not provide objective criteria for determining whether the chest radiograph or ECG result was abnormal; different physicians may have interpreted these studies differently.
Because of protocol violations, not every enrolled patient underwent each of the required tests. A basic metabolic panel was not obtained in 1% of subjects, urinalysis was omitted in 9%, CBC was omitted in 4%, and ECG was omitted in 10%. After study initiation, one site declined to routinely obtain chest radiographs for study purposes; thus, only 70% of subjects received this test. We are unable to determine whether there was selective bias in test ordering by the treating physicians.
A surprisingly high number of patients met exclusion criteria for the study. Forty-six percent of patients presenting with severely elevated blood pressure were ineligible after reporting that they had experienced symptoms suggestive of acute hypertensive target-organ injury within 24 hours of ED presentation. If the definition of hypertensive emergency were rigidly applied, almost half of severely hypertensive ED patients would be diagnosed with this condition, a far greater proportion than reported in the literature. Subjects often related these symptoms only when questioned by a research associate, raising concerns of recall and elicitation biases. Nevertheless, we could not consider a patient to have asymptomatic severely elevated blood pressure if he or she reported recently having a symptom consistent with acute hypertensive target-organ injury. Even after extension of the enrollment period, the high exclusion rate resulted in a smaller sample size than anticipated.
The blood pressure criteria for enrollment were based on the definition of “stage 3” or “severely elevated blood pressure” used in the “Sixth Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.”10Although this report has been supplanted by more recent guidelines, many current hypertension
management recommendations continue to use these characteristics for defining values at which screening for hypertensive emergency is advised.1,5There is no strong scientific evidence for setting a systolic blood pressure value of 180 mm Hg or diastolic value of 110 mm Hg as “trigger values” for performing screening tests. The current JNC-7 guidelines do not use the terms “stage 3 blood pressure” or “severe
hypertension” and do not cite cutoff blood pressure values for performing screening tests.9
This study protocol was developed with then-current editions of standard emergency medicine textbooks, which advised routine, comprehensive testing of patients with severely elevated blood pressure to exclude occult, acute, hypertensive, target-organ injury.11,12Current editions of these textbooks are less prescriptive in their recommendations for the assessment of patients with asymptomatic blood pressure elevation.4,5
It is possible that the rate of unanticipated test abnormalities in the study patients reflects a high frequency of undiagnosed disease in the populations served by the sites, all of which treat a high proportion of uninsured and low-income patients. We cannot determine whether performing the same screening tests in normotensive ED patients would result in a similar rate of unexpected clinically meaningful abnormalities. The rate of unanticipated test abnormalities might have been different had
Table 2.Demographic characteristics of enrolled subjects.
Characteristic No. (%)* Female 60 (54) Age, y, mean 56.5⫾12.7 Ethnicity Black 92 (83) Latino 16 (14) White, non-Latino 2 (2) Native American 1 (1) Medical history Hypertension 95 (86) Heart disease 23 (21) Kidney disease 20 (18) Stroke 18 (16) Lung disease 7 (6) BP value on enrollment (mm Hg) Systolic, mean 199⫾17 Diastolic, mean 110⫾19 Final disposition Discharged 75 (68) Hospitalized 31 (28) Observation unit 2 (2)
Eloped or left AMA 2 (2)
AMA,Against medical advice;BP,blood pressure.
*Results are reported as frequencies, with percentages in parentheses; means are reported with SDs. Blood pressure values on enrollment were obtained man-ually at least 30 minutes after an elevated screening blood pressure was noted.
the study been performed in EDs serving patients more likely to have medical insurance and primary care providers.
DISCUSSION
Current recommendations for the evaluation of ED patients with asymptomatic severely elevated blood pressure advise careful assessment to exclude acute hypertensive target-organ injury.1,4-6Until recently, authors of standard emergency medicine textbooks advised routinely performing extensive testing to rule out hypertensive emergency, including a basic metabolic panel and urinalysis to assess renal function, CBC count to exclude microangiopathic hemolytic anemia, chest radiograph for evidence of congestive heart failure or aortic dissection, and an ECG for signs of ventricular hypertrophy or myocardial ischemia.1,6These suggestions were adapted from recommendations for assessment of patients with “stage 3 (severe)” hypertension as presented in the sixth Joint National Committee report.10Those recommendations were explicitly intended for stable outpatients and were not intended to be applied in acute care settings. An updated report from the Joint National Committee no longer contains specific testing recommendations and no longer uses blood pressure values to define a tier of patients requiring immediate intervention or assessment.9Reflecting these revisions, authors of current emergency medicine textbooks no longer dogmatically prescribe comprehensive testing of all ED patients with severely elevated blood pressure.4,5Unfortunately, these texts now contain ambiguous advice about the need for testing to exclude acute hypertensive target-organ dysfunction. As noted by one author,
any attempt to make clear recommendations is limited by the lack of evidence-based data.5
The utility of performing screening tests on ED patients with asymptomatic severely elevated blood pressure has been
questioned in the emergency medicine literature, principally because there is no evidence that these tests affect either medical decisionmaking in the ED or short-term outcomes.1The value of the ECG in assessing patients with elevated blood pressure lies primarily in detecting ventricular hypertrophy, a long-term complication of hypertension and not a finding likely to affect ED care. It can be argued that the ECG is unlikely to yield information relevant to the emergency management of patients without signs or symptoms of myocardial ischemia. The chest radiograph may reveal unrecognized cardiomegaly, pulmonary edema, or evidence of aortic aneurysm; again, the utility of screening asymptomatic ED patients for these abnormalities is questionable. The prevalence of acute microangiopathic hemolytic anemia in severely hypertensive patients is unknown, yet the value of routinely obtaining a CBC count in
asymptomatic individuals to detect hemolytic anemia is likely to be low. The basic metabolic panel has been recommended to identify patients with acute renal insufficiency as a result of severely elevated blood pressure. A previous study demonstrated that the basic metabolic panel failed to reveal acute renal dysfunction in these patients unless hematuria or nontrace proteinuria was present, suggesting that the test may be safely eliminated in many ED patients with severely elevated blood pressure.13The results of this study confirm that finding.
Table 3.Treating physicians’ reasons for ordering tests in patients with asymptomatic severely elevated blood pressure values.
Reason Test Was Ordered* BMP (nⴝ108)† CBC Count (nⴝ105) UA (nⴝ93) ECG (nⴝ94) CXR (nⴝ74)
History of specific abnormality 20 (19, 11–26) 3 (3, 0–6) 15 (16, 9–24) 10 (11, 4–17) 5 (7, 1–13) Current evaluation suggests abnormality 31 (29, 20–37) 22 (21, 13–29) 19 (20, 12–29) 21 (22, 14–31) 23 (31, 21–42) Only as part of recommended evaluation
for severe BP elevation
59 (55, 45–64) 80 (76, 68–84) 61 (66, 56–75) 67 (71, 62–80) 46 (62, 51–73)
*Physicians could select more than 1 reason for ordering a test; columns may thus total more than 100%.
†Values are frequencies, with percentages and 95% CIs in parentheses.
Table 4.Results of screening tests among patients with asymptomatic severely elevated blood pressure.
Test Result* BMP (nⴝ109)†
CBC Count (nⴝ108) UA (nⴝ85) ECG (nⴝ90) CXR (nⴝ71) Any Tests‡
(nⴝ109) Normal 66 (61, 51–70) 75 (69, 61–78) 41 (48, 38–59) 49 (54, 44–65) 57 (80, 71–90) — Abnormal 43 (39, 30–49) 33 (31, 22–39) 44 (52, 41–62) 41 (46, 35–56) 14 (20, 10–29) 87 (80, 72–87) Unanticipated abnormality 15 (14, 7–20) 19 (18, 10–25) 26 (31, 21–40) 26 (29, 20–38) 8 (11, 4–19) 57 (52, 43–62) Meaningful unanticipated abnormality 2 (2, 0–4) 3 (3, 0–6) 3 (4, 0–7) 2 (2, 0–5) 1 (1, 0–4) 7 (6, 2–11) Admission 2 (2, 0–4) 1 (1, 0–3) 1 (1, 0–4) 2 (2, 0–5) 1 (1, 0–4) 5 (5, 1–9) Additional testing 0 1 (1, 0–3) 1 (1, 0–4) 0 1 (1, 0–4) 3 (3, 0–6) Medication modified 0 0 1 (1) 0 0 1 (1, 0–3)
*Physicians could cite more than 1 action resulting from abnormal test results.
†Values are frequencies, with percentages and 95% CIs in parentheses. ‡Because of protocol violations, not all tests were obtained for each patient.
A recent multicenter study demonstrated that few ED patients with severely elevated blood pressure actually undergo the full battery of screening tests advised by most management guidelines.7In that study, only a basic metabolic panel was obtained for the majority of ED patients with severely elevated blood pressure, whereas an ECG, chest radiograph, and urinalysis test was performed in no more than half of patients. Although logical arguments and practice patterns may suggest that comprehensive screening tests are not necessary in ED patients with severely elevated blood pressure, there is no literature establishing that these tests can be safely omitted without compromising medical care. This study was undertaken to assist in the development of evidence-based
recommendations for judicious testing of ED patients with asymptomatic severely elevated blood pressure.
We found that half of ED patients with asymptomatic severely elevated blood pressure had abnormalities in 1 or more screening tests that were not anticipated by the treating physician on the basis of the patient’s medical history, symptoms, or physical findings. The majority of these abnormalities, however, did not result in changes to the patient’s medical management in the ED. We found a low prevalence of clinically meaningful unanticipated test abnormalities, defined as abnormal results leading to hospitalization, additional testing or consultation, or medical regimen changes. The prevalence of unanticipated, clinically meaningful abnormalities in each screening test ranged from 1% to 4%. In aggregate, 6% of study patients had an unanticipated abnormal screening test result that altered ED management.
Among the patients with unanticipated clinically meaningful test abnormalities, none of the abnormalities were considered likely manifestations of acute hypertensive target-organ injury. Four patients were ultimately diagnosed with chronic renal insufficiency, which may have been related to severely elevated
blood pressure. Two patients were admitted for ECG abnormalities consistent with ischemia, although neither had symptoms suggestive of cardiac ischemia at their evaluation. It is unlikely that the findings of abnormal chest radiograph and severe anemia were directly related to the patients’ elevated blood pressure; these most likely represent incidental, though fortuitous, findings.
Further analyses failed to provide the basis for developing guidelines for reducing screening tests in specific patient groups. Unanticipated, clinically meaningful abnormal results were found for each screening test, even among patients who denied a history of cardiopulmonary, cerebrovascular, or renal disease. In a small study such as this, we cannot conservatively recommend eliminating any routine screening test unless the rate of
unanticipated clinically meaningful abnormalities in a group is zero. In conclusion, we have found a high prevalence of unanticipated abnormalities in tests intended to screen for acute hypertension-related target-organ injury in adults with asymptomatic, severely elevated blood pressure in inner-city EDs. However, only a small proportion of these patients had unanticipated abnormalities that were clinically meaningful, resulting in modifications to the ED management. None of these findings were clearly manifestations of acute hypertensive target-organ injury, and some were clearly incidental findings. We are unable to recommend a strategy for safely eliminating screening tests in ED patients with asymptomatic severely elevated blood pressure.
The authors are grateful to Arthur L. Kellermann, MD, MPH, Robert M. McNamara, MD, and William B. Applegate, MD, MPH, for serving as study advisors and to Kathleen Hatala, BSN, and Carol von Hofen, BSN, for supervision of data collection and data management.
Table 5.Unanticipated clinically meaningful test abnormalities noted in study patients with asymptomatic severely elevated blood pressure. Patient Number Age, y Sex Presenting Complaint History of HTN Unanticipated Abnormalities Action
Taken Final Diagnoses
Acute Hypertensive Target-Organ Dysfunction E-019 66 Female Ran out of BP meds Yes Nonhemolytic anemia Hospitalized Hypertension, severe anemia Unlikely M-129 49 Male Referred for
elevated BP
No Abnormal CXR Hospitalized Hypertensive urgency, pleural thickening
Unlikely M-141 42 Male Referred for
elevated BP
Yes RI, hematuria, proteinuria, nonhemolytic anemia
Hospitalized Chronic RI, anemia Possible T-013 46 Male Referred for
elevated BP
Yes Ischemic ECG changes Hospitalized Hypertensive urgency, MI ruled out
Possible T-016 52 Male Referred for
elevated BP
Yes Ischemic ECG changes, RI, proteinuria
Hospitalized Hypertensive urgency, MI ruled out, chronic RI
Possible T-019 35 Male Foot pain No RI, hypokalemia, LVH,
proteinuria
Hospitalized Chronic RI Possible T-029 47 Female Ran out of BP meds Yes RI, nonhemolytic anemia,
hematuria, proteinuria
Hospitalized Chronic RI, anemia Possible
Supervising editor:Judd E. Hollander, MD
Author contributions:DJK, JJC, JWU, PS, and KLH conceived the study and designed the protocol. DJK and LKK wrote the article. LKK performed the statistical analyses and has expertise in these procedures. DAW, MMW, and JJC supervised data collection. All authors participated
substantially in the article revision. DJK takes responsibility for the paper as a whole.
Funding and support:This study was funded by an
unrestricted research grant from Pfizer, Inc. The sponsor had no role in the study design; in the collection, management, analysis, or interpretation of data; or in the preparation, review, or approval of the article.
Publication dates:Received for publication November 30, 2006. Revision received March 15, 2007. Accepted for publication March 26, 2007.
Presented at the Society for Academic Emergency Medicine annual meeting, May 2006, San Francisco, CA.
Reprints not available from the authors.
Address for correspondence:David J. Karras, MD, Department of Emergency Medicine, Temple University School of Medicine, 3401 North Broad Street, Philadelphia, PA 19140; 215-707-5032, fax 215-707-3494; E-mail david.karras@temple.edu.
REFERENCES
1. Shayne PH, Pitts SR. Severely elevated blood pressure in the emergency department.Ann Emerg Med.2003;41:513-529. 2. Karras DJ, Ufberg JW, Heilpern KL, et al. Elevated blood pressure
in emergency department patients.Acad Emerg Med.2005;12: 835-843.
3. Chiang WK, Jamshahi B. Asymptomatic hypertension in the ED. Am J Emerg Med.1998;16:701-704.
4. Gray RO. Hypertension. In: Marx JA, ed.Rosen’s Emergency Medicine: Concepts and Practice. 6th ed. St. Louis, MO: Mosby; 2006:1310-1324.
5. Wu MM, Chanmugam A. Hypertension. In: Tintinalli JE, Kelen GD, Stapczynski JS, eds.Emergency Medicine: A
Comprehensive Study Guide. 6th ed. New York, NY: McGraw-Hill; 2004:394-404.
6. Zampaglione B, Pascale C, Marchisio M, et al. Hypertensive urgencies and emergencies.Hypertension.1996;27:144-147. 7. Karras DJ, Kruus LK, Cienki JL, et al. Management of severely
elevated blood pressure in academic emergency departments: a multicenter study.Ann Emerg Med. 2006,47:230-236.
8. Rebbecchi T, Karras DJ, Heilpern KL, et al. Emergency physician evaluation of asymptomatic severe hypertension [abstract].Ann Emerg Med.1997;30:405.
9. Chobanian AV, Bakris GL, Black HR, et al. Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report.JAMA. 2003;289:2560-2572.
10. Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. The sixth report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.Arch Intern Med.1997; 157:2413-2443.
11. Gray RO, Matthews JJ. Hypertension. In: Marx JA, ed.Rosen’s Emergency Medicine: Concepts and Practice. 5th ed. St. Louis, MO: Mosby; 2002:1158 –1171.
12. Wu MM, Chanmugam A. Hypertension. In: Tintinalli JE, Kelen GD, Stapczynski JS, eds.Emergency Medicine: A
Comprehensive Study Guide. 5th ed. New York, NY: McGraw-Hill; 2000:401– 412.
13. Karras DJ, Heilpern KL, Riley LJ, et al. Urine dipstick as a screening test for serum creatinine elevation in patients with severe hypertension.Acad Emerg Med.2002;9:27-34.
Editor’s Capsule SummaryWhat is already known on this topic: Textbooks recommend the laboratory screening of
asymptomatic patients with very high blood pressure for evidence of end-organ damage.What question this study addressed:Whether screening such patients results in clinically meaningful changes in management.What this study adds to our knowledge:Fifty-seven of the 109 patients had at least 1 abnormal test result. Five patients had abnormalities
conceivably related to the elevated blood pressure. No patient had laboratory findings indicative of a hypertensive emergency. How this might change clinical practice:Although screening tests are recommended, there has never been any evidence about whether patients with asymptomatic hypertension require them. These data support the concept that screening is highly unlikely to disclose results leading to changes in emergency department management.
