BRIEF REPORT HIV, prevalence
HIV Prevalence in a Midwestern Emergency Department Study objective: To determine the prevalence of h u m a n immunodeficiency virus (HIV) seropositivity of patients i5 years of age and older in our emergency department. Design: HIV status was determined anonymously, and the seroprevalence rate was calculated. The 95% confidence intervals also were calculated. Twenty demographic and predictor categorical variable were crosstabulated with HIV status to determine associations. Only gender and male homosexual preference were significantly associated by Fisher's exact test. Type of participants: Excess serum samples from 454 randomly selected patients 15 years of age and older who required venipuncture for their ED evaluation were included in the study, Measurements and main results: Of the 454 serum specimens, six (1.32%) were positive for HIV The 95% confidence interval was from 0.27% to 2.37%. All six positive patients were men. The only statistically significant risk factors associated with HIV seropositivity were male sex (P - .00112) and male homosexual preference (P = .0000). Conclusion: HIV seropositivity occurs in 1.32% of our ED population over the age of 15 years. The only factors that correlate with HIV seropositivity are male homosexual preference and male sex. [Sturm JT: HIV prevalence in a midwestern emergency department. Ann Emerg Med March 1991;20:276-278.]
James T Sturm, MD St Paul, Minnesota From the Departments of Emergency Medicine at Ramsey Clinic and St PaulRamsey Medical Center, St Paul, Minnesota. Received for publication April 16, 1990. Revision received August 9, 1990. Accepted for publication August 20, 1990. This study was approved by the St PaulRamsey Medical and Research Foundation. Address for reprints: James Sturm, MD, 640 Jackson Street, St Paul, Minnesota 55101.
INTRODUCTION Baker et al created a shock wave among emergency medicine health care providers in 1987 when they reported that 3% of 203 critically ill or severely injured emergency department patients were seropositive for human immunodeficiency virus (HIV) antibodyA Kelen et al continued the surveillance of HIV seropositivity in the ED and reported a 5.2% prevalence rate in 1988. z There were 27 patients who presented with known HIV infection, but 92 of the remaining 2,275 patients (4.0%) had unrecognized HIV infection, z Kelen and coworkers reported in 1989 that the HIV-seropositivity rate in their ED had risen to 6% of 2,544 consecutive patients tested. 3 In a subgroup of 126 patients in Kelen et al's population with critical illness or injury, 7.8% of patients were positive for HIV antibody, and patients between the ages of 25 and 34 years with penetrating trauma showed a seroprevalence rate of 18% for HIV antibody. 4 These statistics are alarming and point to the need for all emergency personnel to use universal precautions in dealing with patients and bodily fluids. Baker et al and Kelen et al have done the greatest amount of research regarding the incidence of HIV infection in ED patients, but their series may provide a skewed view of the overall incidence of HIV seropositivity among ED patients in other areas of the United States. To view the risk of HIV exposure to ED personnel in proper perspective, it is necessary for studies from EDs in many areas of the country to be performed, compared, and collated. To that end, Zeman and Mayhue presented the results of their study of H1V seropositivity and trauma patients admitted to their midwestern community hospital, s They studied 262 consecutive patients and found one positive for HIV (0.38%). The rates of HIV seropositivity in various subcategories of the general
20:3 March 1991
Annals of Emergency Medicine
276/87
HIV PREVALENCE Sturm
population are known. HIV infection is f o u n d in 0.043% of f i r s t - t i m e blood donors, 6 0.33% of Job Corps applicants, 6 and 0.15% of applicants for military service. 7 Rates of HIV seropositivity m a y v a r y geographically and s o c i o e c o n o m i c a l l y and may depend on the local incidence of various risk factors for HIV infection such as IV drug abuse and male homosexual preference. Our study was conducted to determine the rates of k n o w n and unknown HIV seroprevalence in a cohort of our ED patients and to determine if any risk factors predict HIV seropositivity in this group.
METHODS St Paul-Ramsey Medical Center is a trauma referral center. The Department of Emergency Medicine provides care for 180 to 240 patients a day. The ED is staffed on a 24-houra-day basis by emergency medicine staff physicians; residents from the departments of internal medicine, surgery, family practice, and obstetrics and gynecology; medical students; physician's assistants; nurses; and nursing assistants. Excess serum samples from 454 randomly selected patients 15 years of age and older who required venipuncture for other medical reasons were included in the study. The time frame of the study was January 1, 1989, through March 3, 1989. A data sheet was completed for each patient in the study by the staff physician, resident, physician's assistant, or medical student caring for the patient. Information was collected on the patient's risk factor status and clinical presentation. Risk factors were assessed by patient interview (when the patient was responsive), physical examination for needle tracks, and review of the clinical record when available (Figure). Health care providers' potential exposure to blood and other bodily fluids during invasive procedures was recorded. Each serum specimen and corresponding data sheet was labeled with the patient's name and chart number. Patients who were eligible for inclusion but had insufficient volumes of blood obtained were excluded from the study. Several patients presented one or more times to the ED during the course of the study and had blood drawn. When this occurred, only the 88/277
FIGURE. Risk factor data sheet.
Chief complaint Age
first tube was tested; therefore, there were no duplicate tests. The serum tubes were centrifuged, refrigerated, and run in batches. The data sheets were collected, and the chart was referred to for the data sheets to be as complete as possible. In some cases, the data sheet could not be completed because the patient was too critically ill, comatose, or intoxicated. Sheets were completed for 250 patients (55%). The data sheets were completed as much as possible and then given to a member of the clinical laboratory not involved with the study. This individual placed a code number on both the serum tube and the study sheet and removed the patient's name and any other identifying characteristics from both the serum tube and data sheet. At that point, the study became anonymous. This protocol was approved by the hospital institutional review board and was determined by legal counsel to be in conformance with the state's laws governing HIV testing and patient confidentiality. For the purposes of analysis, patients were classified as having a k n o w n HIV i n f e c t i o n w h e n they identified themselves as such, were identified as such by their personal physician, or had this diagnosis documented in the previous hospital record. All other patients were considered to have unknown HIV status. Patients were defined as having unrecognized HIV infection when their HIV status was unknown at ED presentation but was proved t 9 be positive during this study. All serum samples were evaluated by enzyme-linked immunoabsorbant a s s a y (ELISA, O r g a n o n - T e k n i k , Charleston, South Carolina). Each reactive specimen was run a second time by ELISA. All repeatedly reactive specimens were analyzed by the Western-Blot test (DuPont, Wilmington, Delaware). Twenty demographic and predictor categorical variables were cross-tabulated with HIV serology status (positive or negative) to determine or confirm associations or risks. Because of the extreme skewness of the sample (only six, or 1.32%, were HIV positive), these percentages broken down by HIV status can be viewed only deAnnals of Emergency Medicine
Sex Race Known HIV status Known IV drug abuse Needle tracks observed on arms Man with homosexual preference Man with bisexual preference Woman who has had sex with bisexual man Woman who has had sex with IV drug abuser Blood or blood products transfusion since 1981 rv line begun in field iV line begun in ED Trauma Nontrauma Type of third-party payor Risk factor analysis precluded Altered mental status Cardiac arrest Too critical for risk factor analysis
scriptively. Therefore, X2 values wer used only as guides to emphasiz~ these associations. Fisher's exact test was performed to clarify the significance of the apparent association between gender and serology status.
RESULTS A total of 454 random blood tube specimens were collected. Of the 454 serum specimens, six (1.32%) were positive for HIV. The 95% confidence interval was from 0.27% to 2.37%. The only statistically significant risk factors associated with HIV seropositivity were male sex (P = .00112) and male homosexual preference (P = .0000) by Fisher's exact test. Although gender and sexual preference were m a t h e m a t i c a l l y statistically significant risk factors, the skewness of the sample should modulate interpretation of the analysis. The HIV-positive patients ranged in age from 17 to 57 years. Three were known by their history or chart to be HIV positive; a 57-year-old man gave a history of being gay and hemophiliac, a 25-year-old man admitted to being gay and having AIDS-related complex, and a man of unknown age admitted to being HIV positive and gay but denied having AIDS or AIDSrelated complex. There were three men who did not identify themselves as HIV positive, and their charts did 20:3 March 1991
HIV PREVALENCE Sturm
not indicate HIV seropositivity. The risk factor data sheet was complete in 250 of the 454 patients (55%). This compares favorably w i t h Kelen et al's 29% risk factor analysis completion. 2
DISCUSSION This s t u d y is of v a l u e b e c a u s e it portrays a significantly different picture of HIV seropositivity among ED patients t h a n t h a t of the largest reported series. 1-4 Our study, like that of Z e m a n and M a y h u e 5 and Soders t r o m et al, s i l l u s t r a t e s t h a t HIV seropositivity in ED patients m a y be lower in some parts of this country than in others. T h e reason for geographical differences in HIV-seropositivity r a t e s p r o b a b l y r e l a t e s to t h e prevalence of such risk behaviors as 1V drug abuse and m a l e h o m o s e x u a l preference in different c o m m u n i t i e s . O u r 1.32% s e r o p r e v a l e n c e r a t e among ED patients stands in contrast to a 0.003% p r e v a l e n c e rate a m o n g resident M i n n e s o t a b l o o d donors. 9 The low incidence of HIV infection in M i n n e s o t a is thought to be due to the l o w p r e v a l e n c e of IV d r u g abusers. There is also the possibility that HIV seropositivity among known IV drug abusers varies geo-
20:3 March 1991
graphically. The study points out the need for universal blood precautions.10 In our department, we see b e t w e e n 180 and 240 patients a day, and an HIV-serop r e v a l e n c e rate of 1.32% i n d i c a t e s t h a t d u r i n g a 2 4 - h o u r period, one H I V - p o s i t i v e p a t i e n t w i l l be s e e n and, half the time, will not identify h i m s e l f as b e i n g H I V p o s i t i v e . Heightened awareness of the risk factors that predict HIV infection, such as m a l e h o m o s e x u a l preference and IV drug abuse, and increased awareness that an HIV-seropositive p a t i e n t m a y p r e s e n t e a c h d a y in t h e ED m i g h t encourage e m e r g e n c y personn e l to u s e u n i v e r s a l p r e c a u t i o n s more diligently. It w o u l d be useful if m o r e studies such as those of Z e m a n and M a y h u e 5 and Soderstrom et al s, and this study w e r e c a r r i e d o u t in m a j o r i n s t i t u tions, and the results were collated. This w o u l d i n d i c a t e the geographic areas of highest risk and provide inf o r m a t i o n useful to medical personnel and the general public. CONCLUSION H I V s e r o p o s i t i v i t y o c c u r r e d in 1.32% of our ED p a t i e n t s over the age of 15 years. Half of our HIV-posi-
Annals of Emergency Medicine
tive patients were not k n o w n to be positive at the t i m e of presentation.
REFERENCES 1. Baker JL, Kelen GD, Sivertson KT, et al: Unsuspected human immunodeficiency virus in critically ill emergency patients. JAMA 1987~257:2609-2611. 2. Kelen GD, Fritz S, Qaqish B, et ah Unrecognized human immunodeficiency virus infection in emergency department patients. N Engl J Med 1988;318:1645-1650. 3. Kelen GD, DiGiovana T, Bisson L, et al: Human immunodeficiency virus infection in emergency department patients. JAMA 1989;262:516-522. 4. Kelen GD, Fritz S, Qaqish B, et al: Substantial increase in human immnnodeficiency virus (HIV-1) infection in critically ill emergency patients: 1986 and 1987 compared. Ann Emerg Med 1989;18:378-382. 5. Zeman MG, Mayhue PE: Human immunodeficiency virus (HIV) seropositivity in a midwestern community trauma population {abstract). Ann Emerg Med 1988;17: 409. 6. Human immunodeficiency virus infection in the United States: A review of current knowledge. MMWR 1987~36(suppl 5-6):1-48. 7. Trends in human immunodeficiency virus infection among civilian applicants for military service United States, October 1985-March 1988. MMWR 1988;37: 677-679. 8. Soderstrom CA, Furth PA, Glasser D, et ah HIV infection rates in a trauma center treating predominantly rural b l u n t t r a u m a v i c t i m s . J Trauma 1989~29: 1526-1530. 9. McDonald KL, Jackson JB, Bowman RJ, et al: Performance characteristics of serologic tests for human immunodeficiency virus type {HIV-1) antibody among Minnesota blood donors: Public health and clinical implications. Ann Intern Med 1989;110:617 621. 10. Kelen GD: Human immunodeficiency virus and the emergency department: Risks and risk protection for health care providers. Ann Ernerg Med 1990;19:242-248.
278/89
HIV Prevalence in a Midwestern Emergency Department Study objective: To determine the prevalence of h u m a n immunodeficiency virus (HIV) seropositivity of patients i5 years of age and older in our emergency department. Design: HIV status was determined anonymously, and the seroprevalence rate was calculated. The 95% confidence intervals also were calculated. Twenty demographic and predictor categorical variable were crosstabulated with HIV status to determine associations. Only gender and male homosexual preference were significantly associated by Fisher's exact test. Type of participants: Excess serum samples from 454 randomly selected patients 15 years of age and older who required venipuncture for their ED evaluation were included in the study, Measurements and main results: Of the 454 serum specimens, six (1.32%) were positive for HIV The 95% confidence interval was from 0.27% to 2.37%. All six positive patients were men. The only statistically significant risk factors associated with HIV seropositivity were male sex (P - .00112) and male homosexual preference (P = .0000). Conclusion: HIV seropositivity occurs in 1.32% of our ED population over the age of 15 years. The only factors that correlate with HIV seropositivity are male homosexual preference and male sex. [Sturm JT: HIV prevalence in a midwestern emergency department. Ann Emerg Med March 1991;20:276-278.]
James T Sturm, MD St Paul, Minnesota From the Departments of Emergency Medicine at Ramsey Clinic and St PaulRamsey Medical Center, St Paul, Minnesota. Received for publication April 16, 1990. Revision received August 9, 1990. Accepted for publication August 20, 1990. This study was approved by the St PaulRamsey Medical and Research Foundation. Address for reprints: James Sturm, MD, 640 Jackson Street, St Paul, Minnesota 55101.
INTRODUCTION Baker et al created a shock wave among emergency medicine health care providers in 1987 when they reported that 3% of 203 critically ill or severely injured emergency department patients were seropositive for human immunodeficiency virus (HIV) antibodyA Kelen et al continued the surveillance of HIV seropositivity in the ED and reported a 5.2% prevalence rate in 1988. z There were 27 patients who presented with known HIV infection, but 92 of the remaining 2,275 patients (4.0%) had unrecognized HIV infection, z Kelen and coworkers reported in 1989 that the HIV-seropositivity rate in their ED had risen to 6% of 2,544 consecutive patients tested. 3 In a subgroup of 126 patients in Kelen et al's population with critical illness or injury, 7.8% of patients were positive for HIV antibody, and patients between the ages of 25 and 34 years with penetrating trauma showed a seroprevalence rate of 18% for HIV antibody. 4 These statistics are alarming and point to the need for all emergency personnel to use universal precautions in dealing with patients and bodily fluids. Baker et al and Kelen et al have done the greatest amount of research regarding the incidence of HIV infection in ED patients, but their series may provide a skewed view of the overall incidence of HIV seropositivity among ED patients in other areas of the United States. To view the risk of HIV exposure to ED personnel in proper perspective, it is necessary for studies from EDs in many areas of the country to be performed, compared, and collated. To that end, Zeman and Mayhue presented the results of their study of H1V seropositivity and trauma patients admitted to their midwestern community hospital, s They studied 262 consecutive patients and found one positive for HIV (0.38%). The rates of HIV seropositivity in various subcategories of the general
20:3 March 1991
Annals of Emergency Medicine
276/87
HIV PREVALENCE Sturm
population are known. HIV infection is f o u n d in 0.043% of f i r s t - t i m e blood donors, 6 0.33% of Job Corps applicants, 6 and 0.15% of applicants for military service. 7 Rates of HIV seropositivity m a y v a r y geographically and s o c i o e c o n o m i c a l l y and may depend on the local incidence of various risk factors for HIV infection such as IV drug abuse and male homosexual preference. Our study was conducted to determine the rates of k n o w n and unknown HIV seroprevalence in a cohort of our ED patients and to determine if any risk factors predict HIV seropositivity in this group.
METHODS St Paul-Ramsey Medical Center is a trauma referral center. The Department of Emergency Medicine provides care for 180 to 240 patients a day. The ED is staffed on a 24-houra-day basis by emergency medicine staff physicians; residents from the departments of internal medicine, surgery, family practice, and obstetrics and gynecology; medical students; physician's assistants; nurses; and nursing assistants. Excess serum samples from 454 randomly selected patients 15 years of age and older who required venipuncture for other medical reasons were included in the study. The time frame of the study was January 1, 1989, through March 3, 1989. A data sheet was completed for each patient in the study by the staff physician, resident, physician's assistant, or medical student caring for the patient. Information was collected on the patient's risk factor status and clinical presentation. Risk factors were assessed by patient interview (when the patient was responsive), physical examination for needle tracks, and review of the clinical record when available (Figure). Health care providers' potential exposure to blood and other bodily fluids during invasive procedures was recorded. Each serum specimen and corresponding data sheet was labeled with the patient's name and chart number. Patients who were eligible for inclusion but had insufficient volumes of blood obtained were excluded from the study. Several patients presented one or more times to the ED during the course of the study and had blood drawn. When this occurred, only the 88/277
FIGURE. Risk factor data sheet.
Chief complaint Age
first tube was tested; therefore, there were no duplicate tests. The serum tubes were centrifuged, refrigerated, and run in batches. The data sheets were collected, and the chart was referred to for the data sheets to be as complete as possible. In some cases, the data sheet could not be completed because the patient was too critically ill, comatose, or intoxicated. Sheets were completed for 250 patients (55%). The data sheets were completed as much as possible and then given to a member of the clinical laboratory not involved with the study. This individual placed a code number on both the serum tube and the study sheet and removed the patient's name and any other identifying characteristics from both the serum tube and data sheet. At that point, the study became anonymous. This protocol was approved by the hospital institutional review board and was determined by legal counsel to be in conformance with the state's laws governing HIV testing and patient confidentiality. For the purposes of analysis, patients were classified as having a k n o w n HIV i n f e c t i o n w h e n they identified themselves as such, were identified as such by their personal physician, or had this diagnosis documented in the previous hospital record. All other patients were considered to have unknown HIV status. Patients were defined as having unrecognized HIV infection when their HIV status was unknown at ED presentation but was proved t 9 be positive during this study. All serum samples were evaluated by enzyme-linked immunoabsorbant a s s a y (ELISA, O r g a n o n - T e k n i k , Charleston, South Carolina). Each reactive specimen was run a second time by ELISA. All repeatedly reactive specimens were analyzed by the Western-Blot test (DuPont, Wilmington, Delaware). Twenty demographic and predictor categorical variables were cross-tabulated with HIV serology status (positive or negative) to determine or confirm associations or risks. Because of the extreme skewness of the sample (only six, or 1.32%, were HIV positive), these percentages broken down by HIV status can be viewed only deAnnals of Emergency Medicine
Sex Race Known HIV status Known IV drug abuse Needle tracks observed on arms Man with homosexual preference Man with bisexual preference Woman who has had sex with bisexual man Woman who has had sex with IV drug abuser Blood or blood products transfusion since 1981 rv line begun in field iV line begun in ED Trauma Nontrauma Type of third-party payor Risk factor analysis precluded Altered mental status Cardiac arrest Too critical for risk factor analysis
scriptively. Therefore, X2 values wer used only as guides to emphasiz~ these associations. Fisher's exact test was performed to clarify the significance of the apparent association between gender and serology status.
RESULTS A total of 454 random blood tube specimens were collected. Of the 454 serum specimens, six (1.32%) were positive for HIV. The 95% confidence interval was from 0.27% to 2.37%. The only statistically significant risk factors associated with HIV seropositivity were male sex (P = .00112) and male homosexual preference (P = .0000) by Fisher's exact test. Although gender and sexual preference were m a t h e m a t i c a l l y statistically significant risk factors, the skewness of the sample should modulate interpretation of the analysis. The HIV-positive patients ranged in age from 17 to 57 years. Three were known by their history or chart to be HIV positive; a 57-year-old man gave a history of being gay and hemophiliac, a 25-year-old man admitted to being gay and having AIDS-related complex, and a man of unknown age admitted to being HIV positive and gay but denied having AIDS or AIDSrelated complex. There were three men who did not identify themselves as HIV positive, and their charts did 20:3 March 1991
HIV PREVALENCE Sturm
not indicate HIV seropositivity. The risk factor data sheet was complete in 250 of the 454 patients (55%). This compares favorably w i t h Kelen et al's 29% risk factor analysis completion. 2
DISCUSSION This s t u d y is of v a l u e b e c a u s e it portrays a significantly different picture of HIV seropositivity among ED patients t h a n t h a t of the largest reported series. 1-4 Our study, like that of Z e m a n and M a y h u e 5 and Soders t r o m et al, s i l l u s t r a t e s t h a t HIV seropositivity in ED patients m a y be lower in some parts of this country than in others. T h e reason for geographical differences in HIV-seropositivity r a t e s p r o b a b l y r e l a t e s to t h e prevalence of such risk behaviors as 1V drug abuse and m a l e h o m o s e x u a l preference in different c o m m u n i t i e s . O u r 1.32% s e r o p r e v a l e n c e r a t e among ED patients stands in contrast to a 0.003% p r e v a l e n c e rate a m o n g resident M i n n e s o t a b l o o d donors. 9 The low incidence of HIV infection in M i n n e s o t a is thought to be due to the l o w p r e v a l e n c e of IV d r u g abusers. There is also the possibility that HIV seropositivity among known IV drug abusers varies geo-
20:3 March 1991
graphically. The study points out the need for universal blood precautions.10 In our department, we see b e t w e e n 180 and 240 patients a day, and an HIV-serop r e v a l e n c e rate of 1.32% i n d i c a t e s t h a t d u r i n g a 2 4 - h o u r period, one H I V - p o s i t i v e p a t i e n t w i l l be s e e n and, half the time, will not identify h i m s e l f as b e i n g H I V p o s i t i v e . Heightened awareness of the risk factors that predict HIV infection, such as m a l e h o m o s e x u a l preference and IV drug abuse, and increased awareness that an HIV-seropositive p a t i e n t m a y p r e s e n t e a c h d a y in t h e ED m i g h t encourage e m e r g e n c y personn e l to u s e u n i v e r s a l p r e c a u t i o n s more diligently. It w o u l d be useful if m o r e studies such as those of Z e m a n and M a y h u e 5 and Soderstrom et al s, and this study w e r e c a r r i e d o u t in m a j o r i n s t i t u tions, and the results were collated. This w o u l d i n d i c a t e the geographic areas of highest risk and provide inf o r m a t i o n useful to medical personnel and the general public. CONCLUSION H I V s e r o p o s i t i v i t y o c c u r r e d in 1.32% of our ED p a t i e n t s over the age of 15 years. Half of our HIV-posi-
Annals of Emergency Medicine
tive patients were not k n o w n to be positive at the t i m e of presentation.
REFERENCES 1. Baker JL, Kelen GD, Sivertson KT, et al: Unsuspected human immunodeficiency virus in critically ill emergency patients. JAMA 1987~257:2609-2611. 2. Kelen GD, Fritz S, Qaqish B, et ah Unrecognized human immunodeficiency virus infection in emergency department patients. N Engl J Med 1988;318:1645-1650. 3. Kelen GD, DiGiovana T, Bisson L, et al: Human immunodeficiency virus infection in emergency department patients. JAMA 1989;262:516-522. 4. Kelen GD, Fritz S, Qaqish B, et al: Substantial increase in human immnnodeficiency virus (HIV-1) infection in critically ill emergency patients: 1986 and 1987 compared. Ann Emerg Med 1989;18:378-382. 5. Zeman MG, Mayhue PE: Human immunodeficiency virus (HIV) seropositivity in a midwestern community trauma population {abstract). Ann Emerg Med 1988;17: 409. 6. Human immunodeficiency virus infection in the United States: A review of current knowledge. MMWR 1987~36(suppl 5-6):1-48. 7. Trends in human immunodeficiency virus infection among civilian applicants for military service United States, October 1985-March 1988. MMWR 1988;37: 677-679. 8. Soderstrom CA, Furth PA, Glasser D, et ah HIV infection rates in a trauma center treating predominantly rural b l u n t t r a u m a v i c t i m s . J Trauma 1989~29: 1526-1530. 9. McDonald KL, Jackson JB, Bowman RJ, et al: Performance characteristics of serologic tests for human immunodeficiency virus type {HIV-1) antibody among Minnesota blood donors: Public health and clinical implications. Ann Intern Med 1989;110:617 621. 10. Kelen GD: Human immunodeficiency virus and the emergency department: Risks and risk protection for health care providers. Ann Ernerg Med 1990;19:242-248.
278/89
