- Email: [email protected]
Reproducibility of increased blood pressure during an emergency department or urgent care visit
CARDIOLOGY/BRIEF RESEARCH REPORT
Reproducibility of Increased Blood Pressure During an Emergency Department or Urgent Care Visit
Howard D. Backer, MD, MPH Linda Decker, RN Lynn Ackerson, PhD From the Emergency Department (Backer) and the Departments of Emergency Medicine and Internal Medicine (Decker), Kaiser Permanente Medical Center, Hayward, CA; and the Division of Research, The Permanente Medical Group, Oakland, CA (Ackerson).
Study objective: We determine the reproducibility of increased blood pressure measurements among adults in the emergency department or minor injury clinic. Methods: The study was conducted at Kaiser Permanente Medical Center in Hayward, CA, a large, group-model health maintenance organization providing capitated insurance coverage. All patients were included in the study who had no current diagnosis of hypertension but had increased blood pressure on their presentation to the ED or minor injury clinic during the 2-month study period. The staff was asked to repeat the blood pressure later during the index visit and provide these patients with written instructions to return for additional repeat measures. We compared blood pressures taken in the ED to measurements before and after the ED visit. Results: Four hundred seven patients were included in the study. Of the initial elevated blood pressures, 211 (51.8%) were stage 1 elevation, 147 (36.1%) were stage 2, and 49 (12.0%) were stage 3 by criteria of the Joint National Committee on Hypertension. Sixty-five percent of patients had repeat measures in the clinic during our 5month follow-up period, despite active outreach and reminders. Seventy percent of those who had repeat blood pressure documented had at least 1 increased blood pressure after their ED visit. The proportion of patients with at least 1 abnormal blood pressure on subsequent measurement increased with increasing stage of initial blood pressure (64.4% for stage 1, 77.1% for stage 2, 97.1% for stage 3), but was similar for patients with and without pain as a chief complaint and was similar for patients seen in the ED compared with patients seen in urgent care. Compared with blood pressures taken during the ED visit, matched blood pressures taken before or after showed no statistically significant differences. Conclusion: Increased blood pressure is common among emergency or urgent care patients without a history of current hypertension, and most of these will have mixed or consistently abnormal results on repeat measures. Patients should be referred for repeat measures after a single abnormal measure in the ED. [Ann Emerg Med. 2003;41:507-512.]
Copyright © 2003 by the American College of Emergency Physicians. 0196-0644/2003/$30.00 + 0 doi:10.1067/mem.2003.151
APRIL 2003
41:4
ANNALS OF EMERGENCY MEDICINE
5 0 7
REPRODUCIBILITY OF INCREASED BLOOD PRESSURE Backer, Decker & Ackerson
INTRODUCTION
METHODS
Hypertension is a leading risk factor for coronary heart disease, stroke, renal disease, and retinopathy.1,2 Sixty percent of people who are older than 60 years of age have hypertension, yet only half of these are aware of their hypertension and only one third are adequately treated.3 Most emergency departments record blood pressure on all patients, but increased measurements are often discounted and attributed to pain, anxiety, or random elevation, raising the question of whether the ED is an appropriate site for hypertension screening and routine referral.4 Previous studies suggest that follow-up blood pressure at a subsequent date in the clinic will exclude one third to one half of those with increased blood pressure in the ED, but as many as two thirds can benefit from further therapy or closer follow-up.4-6 Additionally, the ED may allow screening of a significant segment of the population, including many who do not make routine appointments.5,7 Despite evidence that many do require follow-up and that a simple protocol can effectively identify and refer those with increased blood pressure, recognition and follow-up of ED patients with increased blood pressure is often poor.5,8 Insurance coverage and clinic accessibility in a capitated health maintenance organization differ substantially from the inner-city populations of the aforementioned studies. In a group-model health maintenance organization like Kaiser Permanente where patients are encouraged to establish with a primary care provider and obtain preventive health care, patients with hypertension may have a higher likelihood of being previously identified and monitored. On the other hand, the ED may see many new health maintenance organization members and low-utilizing patients who have not had their blood pressure measured regularly, so the ED visit may offer an important opportunity to screen for hypertension. The aim of this study was to determine whether increased blood pressure among adults during visits to the ED or minor injury urgent care clinic is predictive of subsequent increased blood pressure measured in clinic.
This was a cohort study. Research subjects included all adults aged 21 to 80 years seen from April 1 to May 31, 2000, with an initial increased blood pressure measurement in the ED or minor injury clinic (ie, an urgent care clinic contiguous to and run by the ED) at Kaiser Permanente Medical Center in Hayward, CA. Blood pressure criteria are defined by Joint National Committee on Hypertension as: • Normal (mm Hg), systolic less than 130 and diastolic less than 85 • High normal (mm Hg), systolic between 130 or 139 or diastolic between 85 and 89 • Stage 1 (mm Hg), systolic between 140 and 159 or diastolic between 90 and 99 • Stage 2 (mm Hg), systolic between 160 and 179 or diastolic between 100 and 109 • Stage 3 (mm Hg), systolic 180 or higher or diastolic 110 or higher Exclusion criteria included chronic renal failure; diagnosis of hypertension, currently being treated or actively monitored; patients admitted to the hospital or residing in or admitted to a long-term care facility; and hypertension treatment initiated in the ED. All blood pressure measurements in the ED and minor injury clinic were taken by registered nurses, licensed visiting nurses, or clinic medical assistants using automated cuffs that are calibrated twice yearly. Except for ED patients on gurneys, all initial blood pressure measurements were performed sitting, as were all clinic measures. Appropriate cuff size was determined by standard markings on the cuffs, but arm selection varied. Followup measures in clinics outside of the ED were taken using similar practices, although abnormal readings are often confirmed by manual measurement. Subsequent references in this article to blood pressure measurement in the ED refer also to patients seen in the minor injury clinic. Reference to measures in the “clinic” refers to visits to outpatient clinics not part of the ED. The ED at Kaiser Hayward sees approximately 85,000 patients annually, divided equally between the
5 0 8
ANNALS OF EMERGENCY MEDICINE
41:4
APRIL 2003
REPRODUCIBILITY OF INCREASED BLOOD PRESSURE Backer, Decker & Ackerson
ED and minor injury clinic. The population of the Kaiser Health Plan members in this service area is 55% white, 20% Hispanic, 11% Asian, 7% black, and 7% other groups. During 2000, 94.5% of all Hayward ED patients were members of Kaiser Health Plan. Most members are insured through their work. A long-standing departmental policy requires that every patient have vital signs recorded on presentation to the ED or minor injury clinic. If the initial blood pressure taken at ED triage or on admission to the minor injury clinic was increased, nurses were instructed to repeat the blood pressure 2 more times before discharge. (One repeat was acceptable in the minor injury clinic because of the brief visits.) Discharge instructions for an increased reading consisted of written patient information and instructions to have the blood pressure rechecked at least 2 more times in the medical clinic. A study nurse retrospectively reviewed copies of every ED chart from copies that are routinely kept in ED files to capture all patients with initially increased blood pressure during their recent visit; she recorded demographic characteristics, nature of the visit, and each blood pressure obtained. If there was no documentation on the chart to indicate that their increased blood pressure was recognized and that they were given instructions during the visit, the study nurse sent the printed discharge instructions to their home, formatted as a personalized letter. During the 6-month follow-up period, which ended October 15, 2000, the study nurse searched 2 or 3 times for any follow-up visits in the outpatient chart and in a computerized patient data system that contains all emergency and outpatient visits at any Kaiser Permanente facility in Northern California, a diagnosis list, and medications. Data were captured on prior diagnosis of hypertension, measures of blood pressure before and subsequent to the index ED visit, and treatment or recommended follow-up for increased blood pressures. We did not collect racial data on our patients, and this is not available routinely from the computerized demographic database of members. For patients with no evidence of follow-up blood pressures on computer or
APRIL 2003
41:4
ANNALS OF EMERGENCY MEDICINE
chart review, active outreach by letter and then by telephone was done to emphasize the importance of repeating the blood pressure to exclude hypertension, to encourage follow-up, and to answer any questions. Data were entered into a computer and analyzed using SAS software (SAS Institute, Cary, NC). Analysis of blood pressure was performed using categoric values of “all measures normal” or “at least one measurement abnormal.” Analytic methods included McNemar’s test for correlated proportions to compare categoric paired blood pressure readings, and χ2 test to compare unpaired proportions. The Permanente Medical Group Central Research Committee and the Investigation Review Board approved the study protocol. Patients were told that they were being tracked as part of a study, but written informed consent was not required because this protocol attempted to improve adherence to established standards of care for all patients. R E S U LT S
During the study period from April 1 to May 31, 2000, 7,532 patients were seen in the ED and 6,910 patients were seen in the minor injury clinic. Initial vital signs, including blood pressure, were recorded on all patients. We identified 407 persons (2.8% of all visits) who had increased blood pressure recorded on initial measurement in the ED or minor injury clinic and met study inclusion criteria. Two hundred sixty-eight patients had clear evidence on the chart that their increased blood pressure was identified and follow-up instructions were given during the ED visit, while the study nurse sent the remaining 125 patients the same information to assure that it was received. The former group was marginally more likely to obtain follow-up (84.7% versus 76.8%) and to have higher stage of initial blood pressure. Only 12 (2.9%) patients were not health plan members at any time during the study, and 14 (3.4%) were members on the initial visit, but not at the study end date. Two hundred eighteen (53%) of the study patients were male. The mean age of all patients was 47 years and
5 0 9
REPRODUCIBILITY OF INCREASED BLOOD PRESSURE Backer, Decker & Ackerson
did not differ between patients in the ED and patients in the minor injury clinic (mean difference 1.08; 95% confidence interval [CI] –1.87 to 4.03). Two hundred sixty patients presented with an initial complaint that included pain. Only six patients gave a history of taking a medication that might have temporarily increased their blood pressure. Nine study patients had been told in the past that they had borderline increased blood pressure, 33 patients had a past diagnosis of hypertension, and 6 had been on therapy for hypertension in the past. These patients were included in the study because they had been told they did not currently have hypertension. Of the initial increased blood pressure measurements, 211 (51.8%) suggested stage 1 hypertension, 147 (36.1%) were stage 2, and 49 (12.0%) were stage 3. Age was significantly associated with higher blood pressure on initial measurement: the mean age was 45.1 years, 49.0 years, and 53.9 years for patients with stage 1, 2, and 3 elevation, respectively (P=.0003).
Two hundred-one (50%) patients had repeat blood pressure measurements during the ED or minor injury clinic visit: 63% of these had 1 repeat measurement, and 37% had 2. Two hundred sixty-six (65%) patients had repeat blood pressure measurement after their ED or minor injury clinic visit: 32% of these had 1 repeat measure, 68% had 2 or more repeat measurements (Table 1). Neither location of visit (ED or minor injury clinic), presence of pain, nor initial stage of blood pressure had an apparent effect on which patients obtained repeat blood pressure measurement. Of the 335 people who had at least 1 repeat measurement after the ED visit, 257 (76.7%) had at least 1 positive reading (Table 1). Patients were more likely to have multiple or consistently abnormal repeat blood pressure values than consistently normal measures. Subsequent measurements were more likely to be abnormal with higher stages of initial blood pressure measurement (stage 1, 69%; stage 2, 79%; stage 3,
Table 1.
Results of repeat blood pressure measurements.* A, Repeat blood pressure measurements in the ED or minor injury urgent care clinic. BP Stage on Initial ED/MIC Measurement Stage 1† Stage 2‡ Stage 3§ Column total
No. of Patients
1 Repeat BP, WNL
1 Repeat BP, Increased
211 147 49 407
26 17 0 43
41 30 12 83
2 Repeat BPs, 2 Repeat BPs, 2 Repeat BPs, Both Normal Both Increased Mixed Results 12 3 0 15
10 17 16 43
9 6 2 17
Any Repeat in ED
No Repeat BPs in ED/MIC
98 73 30 201
113 74 19 206
Any Repeat in MIC
No Repeat BPs in MIC
135 96 35 266
76 51 14 141
B, Repeat blood pressure measurements in an outpatient clinic, subsequent to the ED visit. BP Stage on Initial ED Measurement
No. of Patients
1 Repeat BP, WNL
1 Repeat BP, Increased
Stage 1† Stage 2‡ Stage 3§ Column total
211 147 49 407
19 12 1 32
19 22 12 53
2 Repeat BPs, 2 Repeat BPs, 2 Repeat BPs, Both Normal Both Increased Mixed Results 35 10 0 45
23 30 14 67
39 22 8 69
BP, Blood pressure; MIC, minor injury urgent care clinic; WNL, within normal limits (systolic <139 mm Hg and diastolic <90 mm Hg). *All values represent the number of patients with specified result. †Stage 1, 140-159 mm Hg systolic or 90-99 mm Hg diastolic. ‡Stage 2, 160-179 mm Hg systolic or 100-109 mm Hg diastolic. §Stage 3, ≥180 mm Hg systolic or ≥110 mm Hg diastolic.
5 1 0
ANNALS OF EMERGENCY MEDICINE
41:4
APRIL 2003
REPRODUCIBILITY OF INCREASED BLOOD PRESSURE Backer, Decker & Ackerson
100%; P=.0001). Patients seen in the ED without pain were as likely as patients with a painful complaint at the time of the ED visit to have abnormal blood pressures on follow-up (76.9% versus 76.7%; 95% CI –10% to 10%). We found prior blood pressure readings recorded in the outpatient chart for 332 (81.6%) of our patients, of whom 228 (68.7%) had at least 1 increased reading within the past 3 years. Patients with an increased blood pressure in their chart before the ED visit were more likely to have an increased reading after their ED visit than were those with all negative readings before the ED visit (85% versus 64%; 95% CI for difference in proportions 10% to 32%). McNemar’s test for paired proportions showed no significant difference between the proportion of patients with at least 1 abnormal repeat reading during the ED visit (73%) compared with repeat measures on the same patient obtained after the ED visit (71%; 95% CI for paired difference in proportions –7% to 11%) (Table 2). The proportion of patients with at least 1 abnormal result before the ED visit (67%) was slightly less than the proportion positive at the ED visit (72%), but this difference was not statistically significant (95% CI for paired difference in proportions –14% to 4%).
Table 2.
Comparison of paired blood pressure measurements taken during the index visit to those taken after the index visit.* Post ED† Normal Abnormal on All at Least 1 No Repeat Measurements§ Measurement Measurement
In ED‡ Normal all measurements§ Abnormal on at least 1 measurement No repeat measurement
17 21
19 75
22 47
39
95
73
*McNemar’s
test (95% CI for paired proportions –7% to 11%). † Repeat blood pressure measurements taken in clinic after the emergency or urgent care visit. ‡ Repeat blood pressure measurements taken in the ED or minor injury urgent care clinic. § Normal=systolic <139 mm Hg and diastolic <90 mm Hg.
APRIL 2003
41:4
ANNALS OF EMERGENCY MEDICINE
DISCUSSION
In our health maintenance organization population, 407 patients out of a total of 14,432 patients seen in the ED or minor injury clinic within a 2-month period had increased blood pressure without a current history of hypertension. This yield from screening (2.8%) is substantially less than the predicted 10% to 20% screening yield among inner-city, predominantly black populations.5,7 Nevertheless, it is evidence of large numbers of patients with potentially undiagnosed hypertension among insured patients in a managed care setting who have access to primary care. The exclusion of many older patients who already had a diagnosis of hypertension probably accounts for this low percentage overall of our eligible patients with increased blood pressure, as well as the relatively young mean age of 48 years.9 Of those who had any repeat blood pressures after their ED visit, nearly two thirds had at least 1 subsequent increased blood pressure and nearly one fourth had 2 or more repeat measures that were both abnormal. If the 73 people who did not have a follow-up reading all had at least 1 positive test, the proportion with at least 1 positive would be 80.9% of all 407 patients. If all 73 had only negative follow-up readings, the proportion with at least 1 positive reading would be 63.0% of all 407 patients. These results are consistent with studies in inner-city EDs that have found one half to two thirds of patients had increased blood pressures on repeat measures.4-6 Furthermore, the presence of pain in our patients did not predict that the patient will subsequently have normal blood pressures, which was also found by Chernow et al.4 Most patients had consistent results on paired measures between ED and follow-up measures or from readings before the ED visit compared with measures during or after. Multiple increased readings in the ED suggest that subsequent clinic measures will be abnormal, and the higher the initial measurement, the more likely that subsequent measurements will also be increased. This lends further support to the significance of increased blood pressure readings taken in the ED and to the need to refer patients for repeat readings after a single increased reading in the ED. Fully half of our patients had
5 1 1
REPRODUCIBILITY OF INCREASED BLOOD PRESSURE Backer, Decker & Ackerson
stage 1 elevation; however, these readings should be taken seriously. The cardiovascular risk increases linearly from levels below our usual treatment threshold.10 On the other hand, we were unable to identify a group in which we could reliably exclude hypertension by repeat measures in the ED. Only 15 (7.5%) patients had 2 normal repeat blood pressure measurements out of 201 patients who had their blood pressure rechecked during the ED visit (Table 1). Among patients with repeat normal blood pressure measurements during their ED visit, nearly equal numbers had at least 1 increased reading in the clinic as had all normal clinic measures. The primary limitation in our study was the inconsistent follow-up and number of repeat measurements. We did not limit analysis to those patients with a minimum of 2 or 3 measures in the previsit, visit, or postvisit period. Thus, a patient classified as “normal on all measures” or “abnormal on at least 1 measure” may be based on a single repeat measure. Our end point in this study was not a definitive diagnosis of hypertension based on optimal criteria, rather, it was comparing consistency among a variable number of repeat blood pressures with variation among the measures. This is similar to data that practitioners have when trying to determine a course of action. These results indicate that increased ED triage blood pressures must not be disregarded as false-positive elevations and that blood pressure screening and referral (independent of the chief complaint) should be done in the ED and urgent care clinic among this insured, managed-care population. Most of these patients will have mixed or consistently abnormal results on repeat measures and require either close follow-up or treatment. Consensus criteria suggest that persons with stage 3 elevation should have repeat blood pressure measurement within 1 week, stage 2 should be rechecked within 2 weeks, and stage 1 can be referred within 6 weeks.2
Author contributions: HDB conceived the study, designed the trial, and obtained research funding. HDB and LD supervised the conduct of the trial and data collection. LD managed the data, including quality control. LA analyzed the data. HDB drafted the manuscript, and all authors contributed substantially to its revision. HDB takes responsibility for the paper as a whole. Received for publication February 7, 2002. Revisions received July 23, 2002; September 7, 2002; September 21, 2002; and October 30, 2002. Accepted for publication November 6, 2002. Supported by the Kaiser Foundation Research Institute, Grant #149725 (Hay). Reprints not available from the authors. Address for correspondence: Howard D. Backer, MD, MPH, 2151 Berkeley Way, Room 712, Berkeley, CA 94704; 510-540-2382, fax 510-883-6015; E-mail [email protected].
REFERENCES 1. Stamler J, Stamler R, Neaton J. Blood pressure, systolic and diastolic, and cardiovascular risks: US population data. Arch Intern Med. 1993;153:598-615. 2. Joint National Committee. The sixth report of the Joint National Committee on prevention, detection, evaluation, and treatment of high blood pressure. Arch Intern Med. 1997;157:2413-2446. 3. US Public Health Service Task Force. Guide to Clinical Preventive Services. Baltimore, MD: Williams & Wilkins; 1996. 4. Chernow S, Iserson K, Criss E. Use of the emergency department for hypertension screening: a prospective study. Ann Emerg Med. 1987;16:180-182. 5. Mamon J, Green L, Levine D, et al. Using the emergency department as a screening site for high blood pressure. A method for improving hypertension detection and appropriate referral. Med Care. 1987;25:770-780. 6. Slater R, DaCruz D, Jarrett L. Detection of hypertension in accident and emergency departments. Arch Emerg Med. 1987;4:7-9. 7. Kaszuba A, Matanoski G, Gibson G. Evaluation of the emergency department as a site for hypertension screening. JACEP. 1978;7:51-55. 8. Glass R, Mirel R, Hollander G, et al. Screening for hypertension in the emergency department. JAMA. 1978;240:1973-1974. 9. Davidson RA, Hale WE, Moore MT, et al. Incidence of hypertension in an ambulatory elderly population. J Am Geriatr Soc. 1989;37:861-866. 10. Vasan R, Larson M, Leip E, et al. Impact of high-normal blood pressure on the risk of cardiovascular disease. N Engl J Med. 2001;345:1291-1297.
We thank Sue Blevins, RN, Sue Shia (analyst), Kathleen Taylor (analyst), Mala Seshagiri (health educator), Mary Lou Odom (data entry), and Beverly Wadsworth (volunteer services) for their assistance with this study.
5 1 2
ANNALS OF EMERGENCY MEDICINE
41:4
APRIL 2003
Reproducibility of Increased Blood Pressure During an Emergency Department or Urgent Care Visit
Howard D. Backer, MD, MPH Linda Decker, RN Lynn Ackerson, PhD From the Emergency Department (Backer) and the Departments of Emergency Medicine and Internal Medicine (Decker), Kaiser Permanente Medical Center, Hayward, CA; and the Division of Research, The Permanente Medical Group, Oakland, CA (Ackerson).
Study objective: We determine the reproducibility of increased blood pressure measurements among adults in the emergency department or minor injury clinic. Methods: The study was conducted at Kaiser Permanente Medical Center in Hayward, CA, a large, group-model health maintenance organization providing capitated insurance coverage. All patients were included in the study who had no current diagnosis of hypertension but had increased blood pressure on their presentation to the ED or minor injury clinic during the 2-month study period. The staff was asked to repeat the blood pressure later during the index visit and provide these patients with written instructions to return for additional repeat measures. We compared blood pressures taken in the ED to measurements before and after the ED visit. Results: Four hundred seven patients were included in the study. Of the initial elevated blood pressures, 211 (51.8%) were stage 1 elevation, 147 (36.1%) were stage 2, and 49 (12.0%) were stage 3 by criteria of the Joint National Committee on Hypertension. Sixty-five percent of patients had repeat measures in the clinic during our 5month follow-up period, despite active outreach and reminders. Seventy percent of those who had repeat blood pressure documented had at least 1 increased blood pressure after their ED visit. The proportion of patients with at least 1 abnormal blood pressure on subsequent measurement increased with increasing stage of initial blood pressure (64.4% for stage 1, 77.1% for stage 2, 97.1% for stage 3), but was similar for patients with and without pain as a chief complaint and was similar for patients seen in the ED compared with patients seen in urgent care. Compared with blood pressures taken during the ED visit, matched blood pressures taken before or after showed no statistically significant differences. Conclusion: Increased blood pressure is common among emergency or urgent care patients without a history of current hypertension, and most of these will have mixed or consistently abnormal results on repeat measures. Patients should be referred for repeat measures after a single abnormal measure in the ED. [Ann Emerg Med. 2003;41:507-512.]
Copyright © 2003 by the American College of Emergency Physicians. 0196-0644/2003/$30.00 + 0 doi:10.1067/mem.2003.151
APRIL 2003
41:4
ANNALS OF EMERGENCY MEDICINE
5 0 7
REPRODUCIBILITY OF INCREASED BLOOD PRESSURE Backer, Decker & Ackerson
INTRODUCTION
METHODS
Hypertension is a leading risk factor for coronary heart disease, stroke, renal disease, and retinopathy.1,2 Sixty percent of people who are older than 60 years of age have hypertension, yet only half of these are aware of their hypertension and only one third are adequately treated.3 Most emergency departments record blood pressure on all patients, but increased measurements are often discounted and attributed to pain, anxiety, or random elevation, raising the question of whether the ED is an appropriate site for hypertension screening and routine referral.4 Previous studies suggest that follow-up blood pressure at a subsequent date in the clinic will exclude one third to one half of those with increased blood pressure in the ED, but as many as two thirds can benefit from further therapy or closer follow-up.4-6 Additionally, the ED may allow screening of a significant segment of the population, including many who do not make routine appointments.5,7 Despite evidence that many do require follow-up and that a simple protocol can effectively identify and refer those with increased blood pressure, recognition and follow-up of ED patients with increased blood pressure is often poor.5,8 Insurance coverage and clinic accessibility in a capitated health maintenance organization differ substantially from the inner-city populations of the aforementioned studies. In a group-model health maintenance organization like Kaiser Permanente where patients are encouraged to establish with a primary care provider and obtain preventive health care, patients with hypertension may have a higher likelihood of being previously identified and monitored. On the other hand, the ED may see many new health maintenance organization members and low-utilizing patients who have not had their blood pressure measured regularly, so the ED visit may offer an important opportunity to screen for hypertension. The aim of this study was to determine whether increased blood pressure among adults during visits to the ED or minor injury urgent care clinic is predictive of subsequent increased blood pressure measured in clinic.
This was a cohort study. Research subjects included all adults aged 21 to 80 years seen from April 1 to May 31, 2000, with an initial increased blood pressure measurement in the ED or minor injury clinic (ie, an urgent care clinic contiguous to and run by the ED) at Kaiser Permanente Medical Center in Hayward, CA. Blood pressure criteria are defined by Joint National Committee on Hypertension as: • Normal (mm Hg), systolic less than 130 and diastolic less than 85 • High normal (mm Hg), systolic between 130 or 139 or diastolic between 85 and 89 • Stage 1 (mm Hg), systolic between 140 and 159 or diastolic between 90 and 99 • Stage 2 (mm Hg), systolic between 160 and 179 or diastolic between 100 and 109 • Stage 3 (mm Hg), systolic 180 or higher or diastolic 110 or higher Exclusion criteria included chronic renal failure; diagnosis of hypertension, currently being treated or actively monitored; patients admitted to the hospital or residing in or admitted to a long-term care facility; and hypertension treatment initiated in the ED. All blood pressure measurements in the ED and minor injury clinic were taken by registered nurses, licensed visiting nurses, or clinic medical assistants using automated cuffs that are calibrated twice yearly. Except for ED patients on gurneys, all initial blood pressure measurements were performed sitting, as were all clinic measures. Appropriate cuff size was determined by standard markings on the cuffs, but arm selection varied. Followup measures in clinics outside of the ED were taken using similar practices, although abnormal readings are often confirmed by manual measurement. Subsequent references in this article to blood pressure measurement in the ED refer also to patients seen in the minor injury clinic. Reference to measures in the “clinic” refers to visits to outpatient clinics not part of the ED. The ED at Kaiser Hayward sees approximately 85,000 patients annually, divided equally between the
5 0 8
ANNALS OF EMERGENCY MEDICINE
41:4
APRIL 2003
REPRODUCIBILITY OF INCREASED BLOOD PRESSURE Backer, Decker & Ackerson
ED and minor injury clinic. The population of the Kaiser Health Plan members in this service area is 55% white, 20% Hispanic, 11% Asian, 7% black, and 7% other groups. During 2000, 94.5% of all Hayward ED patients were members of Kaiser Health Plan. Most members are insured through their work. A long-standing departmental policy requires that every patient have vital signs recorded on presentation to the ED or minor injury clinic. If the initial blood pressure taken at ED triage or on admission to the minor injury clinic was increased, nurses were instructed to repeat the blood pressure 2 more times before discharge. (One repeat was acceptable in the minor injury clinic because of the brief visits.) Discharge instructions for an increased reading consisted of written patient information and instructions to have the blood pressure rechecked at least 2 more times in the medical clinic. A study nurse retrospectively reviewed copies of every ED chart from copies that are routinely kept in ED files to capture all patients with initially increased blood pressure during their recent visit; she recorded demographic characteristics, nature of the visit, and each blood pressure obtained. If there was no documentation on the chart to indicate that their increased blood pressure was recognized and that they were given instructions during the visit, the study nurse sent the printed discharge instructions to their home, formatted as a personalized letter. During the 6-month follow-up period, which ended October 15, 2000, the study nurse searched 2 or 3 times for any follow-up visits in the outpatient chart and in a computerized patient data system that contains all emergency and outpatient visits at any Kaiser Permanente facility in Northern California, a diagnosis list, and medications. Data were captured on prior diagnosis of hypertension, measures of blood pressure before and subsequent to the index ED visit, and treatment or recommended follow-up for increased blood pressures. We did not collect racial data on our patients, and this is not available routinely from the computerized demographic database of members. For patients with no evidence of follow-up blood pressures on computer or
APRIL 2003
41:4
ANNALS OF EMERGENCY MEDICINE
chart review, active outreach by letter and then by telephone was done to emphasize the importance of repeating the blood pressure to exclude hypertension, to encourage follow-up, and to answer any questions. Data were entered into a computer and analyzed using SAS software (SAS Institute, Cary, NC). Analysis of blood pressure was performed using categoric values of “all measures normal” or “at least one measurement abnormal.” Analytic methods included McNemar’s test for correlated proportions to compare categoric paired blood pressure readings, and χ2 test to compare unpaired proportions. The Permanente Medical Group Central Research Committee and the Investigation Review Board approved the study protocol. Patients were told that they were being tracked as part of a study, but written informed consent was not required because this protocol attempted to improve adherence to established standards of care for all patients. R E S U LT S
During the study period from April 1 to May 31, 2000, 7,532 patients were seen in the ED and 6,910 patients were seen in the minor injury clinic. Initial vital signs, including blood pressure, were recorded on all patients. We identified 407 persons (2.8% of all visits) who had increased blood pressure recorded on initial measurement in the ED or minor injury clinic and met study inclusion criteria. Two hundred sixty-eight patients had clear evidence on the chart that their increased blood pressure was identified and follow-up instructions were given during the ED visit, while the study nurse sent the remaining 125 patients the same information to assure that it was received. The former group was marginally more likely to obtain follow-up (84.7% versus 76.8%) and to have higher stage of initial blood pressure. Only 12 (2.9%) patients were not health plan members at any time during the study, and 14 (3.4%) were members on the initial visit, but not at the study end date. Two hundred eighteen (53%) of the study patients were male. The mean age of all patients was 47 years and
5 0 9
REPRODUCIBILITY OF INCREASED BLOOD PRESSURE Backer, Decker & Ackerson
did not differ between patients in the ED and patients in the minor injury clinic (mean difference 1.08; 95% confidence interval [CI] –1.87 to 4.03). Two hundred sixty patients presented with an initial complaint that included pain. Only six patients gave a history of taking a medication that might have temporarily increased their blood pressure. Nine study patients had been told in the past that they had borderline increased blood pressure, 33 patients had a past diagnosis of hypertension, and 6 had been on therapy for hypertension in the past. These patients were included in the study because they had been told they did not currently have hypertension. Of the initial increased blood pressure measurements, 211 (51.8%) suggested stage 1 hypertension, 147 (36.1%) were stage 2, and 49 (12.0%) were stage 3. Age was significantly associated with higher blood pressure on initial measurement: the mean age was 45.1 years, 49.0 years, and 53.9 years for patients with stage 1, 2, and 3 elevation, respectively (P=.0003).
Two hundred-one (50%) patients had repeat blood pressure measurements during the ED or minor injury clinic visit: 63% of these had 1 repeat measurement, and 37% had 2. Two hundred sixty-six (65%) patients had repeat blood pressure measurement after their ED or minor injury clinic visit: 32% of these had 1 repeat measure, 68% had 2 or more repeat measurements (Table 1). Neither location of visit (ED or minor injury clinic), presence of pain, nor initial stage of blood pressure had an apparent effect on which patients obtained repeat blood pressure measurement. Of the 335 people who had at least 1 repeat measurement after the ED visit, 257 (76.7%) had at least 1 positive reading (Table 1). Patients were more likely to have multiple or consistently abnormal repeat blood pressure values than consistently normal measures. Subsequent measurements were more likely to be abnormal with higher stages of initial blood pressure measurement (stage 1, 69%; stage 2, 79%; stage 3,
Table 1.
Results of repeat blood pressure measurements.* A, Repeat blood pressure measurements in the ED or minor injury urgent care clinic. BP Stage on Initial ED/MIC Measurement Stage 1† Stage 2‡ Stage 3§ Column total
No. of Patients
1 Repeat BP, WNL
1 Repeat BP, Increased
211 147 49 407
26 17 0 43
41 30 12 83
2 Repeat BPs, 2 Repeat BPs, 2 Repeat BPs, Both Normal Both Increased Mixed Results 12 3 0 15
10 17 16 43
9 6 2 17
Any Repeat in ED
No Repeat BPs in ED/MIC
98 73 30 201
113 74 19 206
Any Repeat in MIC
No Repeat BPs in MIC
135 96 35 266
76 51 14 141
B, Repeat blood pressure measurements in an outpatient clinic, subsequent to the ED visit. BP Stage on Initial ED Measurement
No. of Patients
1 Repeat BP, WNL
1 Repeat BP, Increased
Stage 1† Stage 2‡ Stage 3§ Column total
211 147 49 407
19 12 1 32
19 22 12 53
2 Repeat BPs, 2 Repeat BPs, 2 Repeat BPs, Both Normal Both Increased Mixed Results 35 10 0 45
23 30 14 67
39 22 8 69
BP, Blood pressure; MIC, minor injury urgent care clinic; WNL, within normal limits (systolic <139 mm Hg and diastolic <90 mm Hg). *All values represent the number of patients with specified result. †Stage 1, 140-159 mm Hg systolic or 90-99 mm Hg diastolic. ‡Stage 2, 160-179 mm Hg systolic or 100-109 mm Hg diastolic. §Stage 3, ≥180 mm Hg systolic or ≥110 mm Hg diastolic.
5 1 0
ANNALS OF EMERGENCY MEDICINE
41:4
APRIL 2003
REPRODUCIBILITY OF INCREASED BLOOD PRESSURE Backer, Decker & Ackerson
100%; P=.0001). Patients seen in the ED without pain were as likely as patients with a painful complaint at the time of the ED visit to have abnormal blood pressures on follow-up (76.9% versus 76.7%; 95% CI –10% to 10%). We found prior blood pressure readings recorded in the outpatient chart for 332 (81.6%) of our patients, of whom 228 (68.7%) had at least 1 increased reading within the past 3 years. Patients with an increased blood pressure in their chart before the ED visit were more likely to have an increased reading after their ED visit than were those with all negative readings before the ED visit (85% versus 64%; 95% CI for difference in proportions 10% to 32%). McNemar’s test for paired proportions showed no significant difference between the proportion of patients with at least 1 abnormal repeat reading during the ED visit (73%) compared with repeat measures on the same patient obtained after the ED visit (71%; 95% CI for paired difference in proportions –7% to 11%) (Table 2). The proportion of patients with at least 1 abnormal result before the ED visit (67%) was slightly less than the proportion positive at the ED visit (72%), but this difference was not statistically significant (95% CI for paired difference in proportions –14% to 4%).
Table 2.
Comparison of paired blood pressure measurements taken during the index visit to those taken after the index visit.* Post ED† Normal Abnormal on All at Least 1 No Repeat Measurements§ Measurement Measurement
In ED‡ Normal all measurements§ Abnormal on at least 1 measurement No repeat measurement
17 21
19 75
22 47
39
95
73
*McNemar’s
test (95% CI for paired proportions –7% to 11%). † Repeat blood pressure measurements taken in clinic after the emergency or urgent care visit. ‡ Repeat blood pressure measurements taken in the ED or minor injury urgent care clinic. § Normal=systolic <139 mm Hg and diastolic <90 mm Hg.
APRIL 2003
41:4
ANNALS OF EMERGENCY MEDICINE
DISCUSSION
In our health maintenance organization population, 407 patients out of a total of 14,432 patients seen in the ED or minor injury clinic within a 2-month period had increased blood pressure without a current history of hypertension. This yield from screening (2.8%) is substantially less than the predicted 10% to 20% screening yield among inner-city, predominantly black populations.5,7 Nevertheless, it is evidence of large numbers of patients with potentially undiagnosed hypertension among insured patients in a managed care setting who have access to primary care. The exclusion of many older patients who already had a diagnosis of hypertension probably accounts for this low percentage overall of our eligible patients with increased blood pressure, as well as the relatively young mean age of 48 years.9 Of those who had any repeat blood pressures after their ED visit, nearly two thirds had at least 1 subsequent increased blood pressure and nearly one fourth had 2 or more repeat measures that were both abnormal. If the 73 people who did not have a follow-up reading all had at least 1 positive test, the proportion with at least 1 positive would be 80.9% of all 407 patients. If all 73 had only negative follow-up readings, the proportion with at least 1 positive reading would be 63.0% of all 407 patients. These results are consistent with studies in inner-city EDs that have found one half to two thirds of patients had increased blood pressures on repeat measures.4-6 Furthermore, the presence of pain in our patients did not predict that the patient will subsequently have normal blood pressures, which was also found by Chernow et al.4 Most patients had consistent results on paired measures between ED and follow-up measures or from readings before the ED visit compared with measures during or after. Multiple increased readings in the ED suggest that subsequent clinic measures will be abnormal, and the higher the initial measurement, the more likely that subsequent measurements will also be increased. This lends further support to the significance of increased blood pressure readings taken in the ED and to the need to refer patients for repeat readings after a single increased reading in the ED. Fully half of our patients had
5 1 1
REPRODUCIBILITY OF INCREASED BLOOD PRESSURE Backer, Decker & Ackerson
stage 1 elevation; however, these readings should be taken seriously. The cardiovascular risk increases linearly from levels below our usual treatment threshold.10 On the other hand, we were unable to identify a group in which we could reliably exclude hypertension by repeat measures in the ED. Only 15 (7.5%) patients had 2 normal repeat blood pressure measurements out of 201 patients who had their blood pressure rechecked during the ED visit (Table 1). Among patients with repeat normal blood pressure measurements during their ED visit, nearly equal numbers had at least 1 increased reading in the clinic as had all normal clinic measures. The primary limitation in our study was the inconsistent follow-up and number of repeat measurements. We did not limit analysis to those patients with a minimum of 2 or 3 measures in the previsit, visit, or postvisit period. Thus, a patient classified as “normal on all measures” or “abnormal on at least 1 measure” may be based on a single repeat measure. Our end point in this study was not a definitive diagnosis of hypertension based on optimal criteria, rather, it was comparing consistency among a variable number of repeat blood pressures with variation among the measures. This is similar to data that practitioners have when trying to determine a course of action. These results indicate that increased ED triage blood pressures must not be disregarded as false-positive elevations and that blood pressure screening and referral (independent of the chief complaint) should be done in the ED and urgent care clinic among this insured, managed-care population. Most of these patients will have mixed or consistently abnormal results on repeat measures and require either close follow-up or treatment. Consensus criteria suggest that persons with stage 3 elevation should have repeat blood pressure measurement within 1 week, stage 2 should be rechecked within 2 weeks, and stage 1 can be referred within 6 weeks.2
Author contributions: HDB conceived the study, designed the trial, and obtained research funding. HDB and LD supervised the conduct of the trial and data collection. LD managed the data, including quality control. LA analyzed the data. HDB drafted the manuscript, and all authors contributed substantially to its revision. HDB takes responsibility for the paper as a whole. Received for publication February 7, 2002. Revisions received July 23, 2002; September 7, 2002; September 21, 2002; and October 30, 2002. Accepted for publication November 6, 2002. Supported by the Kaiser Foundation Research Institute, Grant #149725 (Hay). Reprints not available from the authors. Address for correspondence: Howard D. Backer, MD, MPH, 2151 Berkeley Way, Room 712, Berkeley, CA 94704; 510-540-2382, fax 510-883-6015; E-mail [email protected].
REFERENCES 1. Stamler J, Stamler R, Neaton J. Blood pressure, systolic and diastolic, and cardiovascular risks: US population data. Arch Intern Med. 1993;153:598-615. 2. Joint National Committee. The sixth report of the Joint National Committee on prevention, detection, evaluation, and treatment of high blood pressure. Arch Intern Med. 1997;157:2413-2446. 3. US Public Health Service Task Force. Guide to Clinical Preventive Services. Baltimore, MD: Williams & Wilkins; 1996. 4. Chernow S, Iserson K, Criss E. Use of the emergency department for hypertension screening: a prospective study. Ann Emerg Med. 1987;16:180-182. 5. Mamon J, Green L, Levine D, et al. Using the emergency department as a screening site for high blood pressure. A method for improving hypertension detection and appropriate referral. Med Care. 1987;25:770-780. 6. Slater R, DaCruz D, Jarrett L. Detection of hypertension in accident and emergency departments. Arch Emerg Med. 1987;4:7-9. 7. Kaszuba A, Matanoski G, Gibson G. Evaluation of the emergency department as a site for hypertension screening. JACEP. 1978;7:51-55. 8. Glass R, Mirel R, Hollander G, et al. Screening for hypertension in the emergency department. JAMA. 1978;240:1973-1974. 9. Davidson RA, Hale WE, Moore MT, et al. Incidence of hypertension in an ambulatory elderly population. J Am Geriatr Soc. 1989;37:861-866. 10. Vasan R, Larson M, Leip E, et al. Impact of high-normal blood pressure on the risk of cardiovascular disease. N Engl J Med. 2001;345:1291-1297.
We thank Sue Blevins, RN, Sue Shia (analyst), Kathleen Taylor (analyst), Mala Seshagiri (health educator), Mary Lou Odom (data entry), and Beverly Wadsworth (volunteer services) for their assistance with this study.
5 1 2
ANNALS OF EMERGENCY MEDICINE
41:4
APRIL 2003