HIV Screening in an Urban Emergency Department: Comparison of Screening Using an Opt-In Versus an Opt-Out Approach

Transcription

1 HIV SCREENING IN EMERGENCY DEPARTMENTS HIV Screening in an Urban Emergency Department: Comparison of Screening Using an Opt-In Versus an Opt-Out Approach Douglas A. E. White, MD, Alicia N. Scribner, MPH, Farnaz Vahidnia, MD, Patrick J. Dideum, BBA, Danielle M. Gordon, MS, Bradley W. Frazee, MD, Andrew C. Voetsch, PhD, James D. Heffelfinger, MD From the Department of Emergency Medicine, Alameda County Medical Center, Highland Hospital, Oakland, CA (White, Scribner, Vahidnia, Dideum, Gordon, Frazee); the Department of Epidemiology and Biostatistics, University of California at Berkeley School of Public Health, Berkeley, CA (Vahidnia); and the Division of HIV/AIDS Prevention, Centers for Disease Control and Prevention, Atlanta, GA (Voetsch, Heffelfinger). Objective: We compare outcomes of opt-in and opt-out HIV screening approaches in an urban emergency department. Methods: This was a 1-year prospective observational study comparing 2 6-month screening approaches. Eligibility for opt-in and opt-out screening was identical: aged 15 years or older, medically stable, and able to complete general consent. During the opt-in phase, triage nurses referred patients to HIV testers stationed at triage, who obtained separate opt-in written consent and performed rapid oral fluid tests. During the opt-out phase, registration staff conducted integrated opt-out consent and then referred patients to HIV testers. We assessed the proportion of potentially eligible patients who were offered screening (screening offer rate), the proportion offered screening who accepted (screening acceptance rate), the proportion who accepted screening and subsequently completed testing (test completion rate), and the proportion of potentially eligible patients who completed testing (overall screening rate) during each phase. Results: For the opt-in versus the opt-out phases, respectively, there were 23,236 potentially eligible patients versus 26,757, screening offer rate was 27.9% versus 75.8% (P.001), screening acceptance rate was 62.7% versus 30.9% (P.001), test completion rate was 99.8% versus 74.6% (P.001), and overall screening rate was 17.4% versus 17.5% (P.90). Conclusion: A significantly higher proportion of patients were offered HIV screening with an opt-out approach at registration. However, this was offset by much higher screening acceptance and test completion rates with the opt-in approach at triage. Overall screening rates with the 2 approaches were nearly identical. [Ann Emerg Med. 2011;58:S89-S95.] /$-see front matter Copyright 2011 by the American College of Emergency Physicians. doi: /j.annemergmed INTRODUCTION Background Traditional opt-in screening, in which patients are offered HIV testing and given the opportunity to accept, requires separate written informed consent for testing. The Centers for Disease Control and Prevention (CDC) revised guidelines for HIV testing in health care settings recommend opt-out screening as an important strategy to help make HIV testing a routine part of medical care. 1 With opt-out screening, patients are notified that HIV testing is routinely performed unless they decline, and the general consent for care is considered sufficient for informed consent. Because the standardized nature of testing is emphasized with opt-out, patients may be less likely to feel singled out and the stigma associated with testing may be minimized. Importance In several clinical venues, opt-out HIV screening has been shown to increase screening rates substantially compared with opt-in HIV screening. 2-8 When nurse midwives in the United Kingdom changed from an opt-in to opt-out screening approach during routine prenatal visits, screening rates increased from 35% to 88%. 6 Introduction of opt-out screening in a genitourinary clinic increased screening rates from 66% (with opt-in screening) to 80%. 7 In Denver, a program that integrated opt-out HIV screening into routine prenatal care reported screening rates of 98%. 8 The effect of changing from opt-in emergency physician directed testing to opt-out screening in an emergency department (ED) setting was recently reported. 9 Surprisingly, acceptance rates with ED registrationbased opt-out consent was only 24%. Annals of Emergency Medicine S89

2 HIV Screening in an Urban Emergency Department Goals of This Investigation The goals of this study were to compare the screening offer rate, screening acceptance rate, test completion rate, and overall screening rates for opt-in and opt-out screening approaches. METHODS Study Design This was a prospective observational study, conducted from February 2007 through January Two successive 6-month methods for offering HIV screening were carried out, beginning with opt-in and followed by opt-out. This study received institutional review board approval from the CDC and the Alameda County Medical Center. Setting The study was conducted at the Alameda County Medical Center, an urban teaching hospital and regional trauma center with an accredited emergency medicine residency program in Oakland, CA. In 2006, the annual ED census was approximately 75,000 visits; 47% of patients presenting to the ED were black, 32% were Hispanic, 21% were white, 44% were female patients, 98% were adults aged 15 years or older, and 53% did not have health insurance. Selection of Participants Patients were eligible for HIV screening if they were aged 15 years or older, medically stable, and able to consent for HIV testing (opt-in phase) or complete the general consent for care (opt-out phase). Patients requiring immediate medical evaluation, according to triage nurse evaluation, were deemed medically unstable and ineligible for HIV screening (opt-in phase). Registration staff could also classify patients as ineligible for screening if they appeared too ill (for example, patients in severe pain, actively vomiting, or in respiratory distress) (opt-out phase). Determination of a patient s inability to provide consent was at the discretion of the HIV tester (opt-in phase) or registration staff (opt-out phase) and included factors such as a language barrier, altered mental status, or acute psychiatric illness. Study Protocol Common program components. ED patients first completed a rapid nurse assessment in the waiting room area. Medically stable patients were provided a pretest HIV information brochure and were then referred to registration. After completing registration, patients were referred to the triage area for a standard triage nurse evaluation before being placed in the ED. Two full-time medical assistants were hired to be HIV testers and received training by the principal investigator (D.A.E.W.). Training consisted of a series of didactic lessons, mock patient interactions, result disclosure scenarios, hands-on practice with the rapid tests, and a written and practical competency examination. The same HIV testers worked during each phase. HIV testers were located in a testing station within the triage area weekdays from 7 AM and 10 PM. The testing station, White et al partitioned from the rest of the triage area by a sliding curtain, consisted of 1 chair, a computer terminal, and a supply cabinet with a flat surface for running tests. Chairs next to the testing station were available for patients waiting to be tested. Only 1 HIV tester was on duty at any given time. During hours when the HIV tester was not on site, screening was not performed. HIV testing was performed at the testing station with the OraQuick Advance Rapid HIV-1/2 antibody test (OraSure Technologies, Inc., Bethlehem, PA) on oral fluid specimens. HIV test results were documented in special fields in the electronic medical record (Wellsoft Corporation, Somerset, NJ). Results were interpreted as reactive, negative, or indeterminate. HIV testers placed an electronic order instructing the nurse to disclose negative results once the patient was placed in the treatment area. Nursing staff verbally disclosed negative results and provided patients with a postresult informational handout. A copy of the postresult handout was also added to the patient s electronic discharge instructions. Patients who had reactive HIV test results were immediately placed in a private room in the ED. HIV testers and emergency physicians provided patients with a verbal explanation of the reactive test result, a postresult information handout, emotional support, and information about follow-up care. Blood was drawn and sent for confirmatory testing with Western blot assays, CD4 cell counts, and HIV viral load testing. Follow-up linkage consisted of arranging a 2-week appointment during drop-in times made available by the Alameda County Medical Center HIV clinic. Patients failing to attend their scheduled appointment were contacted by telephone by the clinic s nurse administrator a minimum of 2 times to reschedule before being referred to the county s case management system. Opt-in phase, February 1, 2007, to July 31, Potentially eligible patients were referred by the triage nurse to the HIV tester. All triage nurses were instructed by the principal investigator about the procedure for referring patients to the HIV tester, through s, posted notices, and verbal reminders. In the event that the HIV tester was occupied, patients were instructed to wait in chairs adjacent to the testing station. HIV testers determined eligibility and offered HIV screening by using the following opt-in script: Would you like to have a rapid HIV test today? Patients opting in completed separate written consent, followed by immediate testing. Patients who bypassed triage and the HIV testing station (including those who arrived by ambulance, were trauma activations, or required immediate medical resuscitation) were excluded for eligibility assessment and screening. Opt-out phase, August 1, 2007, to January 31, Consent for HIV testing was integrated into the general consent form for medical care in accordance with CDC guidelines and California state law (which requires written documentation of patient refusal of testing). The general ED consent form was modified to include a statement that HIV testing may be performed during the ED visit unless the patient declined, as S90 Annals of Emergency Medicine

3 White et al HIV Screening in an Urban Emergency Department Figure. Overview of opt-in and opt-out HIV screening in an urban ED, Oakland, CA, 2007 to well as a specific signature box where patients could opt out. Registration staff determined eligibility and offered HIV screening by using the following opt-out script: HIV testing may be performed during your emergency room visit. If you do not want to be HIV tested, sign here. Patients opted out by signing in the opt-out signature box. Registration staff electronically flagged the charts of patients not opting out and referred those patients to the triage testing station. For the subset of patients who received bedside registration and did not opt out, HIV testers were instructed to go to the patient s bedside to perform testing. The principal investigator trained the registration staff in the opt-out protocol. Training consisted of a 1-hour didactic tutorial during a staff meeting, followed by direct observation of each staff member during the first week of implementation. Periodic observations with verbal reminders were performed throughout the opt-out phase. In both opt-in and opt-out phases, physician-initiated diagnostic and targeted testing occurred as an adjunct to HIV screening. Tests ordered by physicians were performed point-ofcare by nursing staff and not HIV testers. The influence of physician-initiated testing on screening acceptance rates could not be determined, and thus patients tested in this way were excluded from the analysis. Data Collection and Processing During both phases, HIV testers recorded information for every patient referred by the triage nurse (opt-in phase) or by registration staff (opt-out phase) in specific HIV testing fields that were part of the electronic medical record. Information was entered when patients arrived at the testing station and as test results became available. Information included whether patients accepted or declined screening, whether screening was completed, and results of screening. For every patient registered during the opt-out phase, the registration staff recorded screening information in special fields added to the electronic registration process. Information included whether patients were offered opt-out HIV screening and whether they accepted or declined screening. These data and data routinely collected during an ED visit, including demographic information and ED disposition, were exported to spreadsheets and patient-identifying information was removed. Western blot results, CD4 cell counts, and follow-up to the HIV clinic were obtained from review of the hospital laboratory system and examination of HIV clinic intake records, which the research coordinator entered manually into the database. Outcome Measures The primary outcome measures were screening offer rate, screening acceptance rate, and test completion rate, which together give an overall screening rate. Screening offer rate is the proportion of potentially eligible patients who were offered screening. Screening acceptance rate is the proportion of patients offered screening who accepted. Test completion rate is the proportion of patients who accepted screening and completed testing. Overall screening rate is the proportion of potentially eligible patients who completed testing. The true number of eligible patients for the opt-in phase was not known because of inability to account for patients who were not referred to the HIV tester. As a surrogate, the number of potentially eligible patients was used instead to determine screening offer rate and overall screening rate. The number of potentially eligible patients is drawn from the census during HIV screening hours. Additional outcome measures were the proportion of patients tested who received a newly positive diagnosis, their median CD4 cell counts, and admission and linkage to medical care rates. All of the definitions adhered to published Annals of Emergency Medicine S91

4 HIV Screening in an Urban Emergency Department White et al Table 1. Characteristics of patients who were offered and accepted opt-in and opt-out HIV screening in an urban ED, Oakland, CA, 2007 to 2008.* Characteristics Opt-In Phase N 6,479 Offered HIV Screening Opt-Out Phase N 20,280 Opt-In Phase N 4,061 Accepted HIV Screening Opt-Out Phase N 6,273 Age, y, mean (SD) 39 (13) 42 (14) 38 (13) 38 (13) Sex, No. (%) Male 3,461 (53) 11,236 (55) 2,103 (52) 3,319 (53) Female 3,017 (47) 9,044 (45) 1,958 (48) 2,954 (47) Race/ethnicity, No. (%) White 954 (15) 3,070 (15) 563 (14) 859 (14) Black 2,801 (43) 9,153 (45) 1,625 (40) 2,862 (46) Hispanic 1,732 (27) 5,188 (26) 1,262 (31) 1,826 (29) Asian 275 (4) 1,252 (6) 166 (4) 299 (5) Other 717 (11) 1,617 (8) 445 (11) 427 (7) Acuity score, No. (%) Low 5,477 (85) 19,754 (97) 3,426 (84) 6,151 (98) High 945 (15) 516 (3) 601 (15) 119 (2) *Total counts for each variable may differ slightly because of missing data. Percentages were calculated as a fraction of the total data set; percentages are of total offered or accepted, even when data are missing. Acuity score was measured on a 5-point triage score from 1 (highest acuity) to 5 (lowest acuity) and collapsed into 2 categories. Low acuity had a triage score greater than 2 and high acuity had a triage score of less than or equal to 2. recommendations from the National Emergency Department HIV Testing Consortium. 10 Primary Data Analysis All data were entered into an electronic database (Microsoft Office Excel 2003; Microsoft, Redmond, WA). Descriptive analyses were performed for all variables, and data are presented at the level of patient visits. Continuous data are reported as medians with interquartile ranges or means with SDs and categorical data are reported as percentages. Bivariate analyses were performed with the Wilcoxon rank sum test, 2 test, or Fischer s exact test, where appropriate. A priori sample size calculation was performed and, according to an opt-in acceptance rate of 50%, patients per group were needed to determine an absolute difference in acceptance rate of 10% between opt-in and opt-out screening, assuming P.05 and a power of 90%. All statistical analyses were performed with Stata (version 10.0; StataCorp, College Station, TX). RESULTS The Figure provides a flow chart for the 2 phases. For the opt-in versus the opt-out phase, results were as follows: there were 23,236 potentially eligible patients versus 26,757, screening offer rate was 27.9% versus 75.8% (P.001), screening acceptance rate was 62.7% versus 30.9% (P.001), test completion rate was 99.8% versus 74.6% (P.001), and overall screening rate was 17.4% versus 17.5% (P.90). There were no differences by age, sex, or race/ethnicity between patients offered or accepting screening in the opt-in and opt-out phases (Table 1). Higher-acuity patients were more likely to be offered opt-in screening than opt-out screening. During the opt-in phase, 21 patients had reactive rapid test results; 10 were confirmed positive and 11 were false positive. Ten of the 11 false-positive rapid results occurred during a 2- month period. Two of the 10 patients with confirmed positive results had received a previous diagnosis of HIV infection. During the opt-out phase, 28 patients had reactive rapid test results and all were confirmed positive. Seven of the 28 patients with confirmed positive results had received a previous diagnosis of HIV infection. Therefore, opt-in screening identified 8 patients (0.2% of tested) with newly positive results and opt-out screening identified 21 (0.4% of tested; P.04). CD4 cell counts were available for 28 of the 29 patients with newly positive results. Two (25%) of the 8 patients with newly positive results during the opt-in phase and 10 (48%) of 21 patients with newly positive results during the opt-out phase had immunologic AIDS (CD4 200 cells/ L) at diagnosis (Table 2). The majority of HIV-infected patients in each phase were discharged from the ED (opt-in 75% versus opt-out 81%) and linked to care within 90 days of their diagnosis (opt-in 75% versus opt-out 77%). All but 1 patient was notified of his reactive test result and given an appointment at the HIV clinic. This patient left against medical advice before test results were available, and attempts to contact him (including a referral to public health case management) were unsuccessful. LIMITATIONS This study has several limitations. Most significantly, the 2 phases differed in several ways, aside from how consent was approached. These confounding differences include the following: the personnel who assessed eligibility and performed consent were different between the phases; the locations where consent was performed were different; in the opt-in phase eligibility assessment, consent and testing were all performed by 1 person; in the opt-out phase, eligibility S92 Annals of Emergency Medicine

5 White et al Table 2. CD4 cell counts, disposition, and linkage to care for patients with newly positive results identified by HIV screening in an urban ED by study phase, Oakland, CA, 2007 to 2008.* Characteristics Patients With Newly Positive Diagnoses, Opt-In N 8 Patients With Newly Positive Diagnoses, Opt-Out N 21 CD4 count cells/ L, mean (SD) 415 (237) 307 (274) CD4 count, cells/ L, No. (%) (25) 10 (48) (75) 10 (48) Disposition, No. (%) Admitted 2 (25) 4 (19) Discharged 6 (75) 17 (81) Linkage to care, No. (%) Successful 6 (75) 14 (67) Delayed 0 2 (10) Unsuccessful 2 (25) 5 (24) Time to follow-up in HIV clinic, median days (IQR) 7 (7-14) 13 (6-16) IQR, Interquartile range. *Total counts for each variable may differ slightly because of missing data. Percentages were calculated as a fraction of the total data set. CD4 count was missing for 1 patient during the opt-out phase. Linkage to care was defined as receipt of any HIV-related care in a nonemergency setting after diagnosis. Successful linkage refers to linkage fewer than 30 days after the date that testing was conducted in the ED (includes individuals who were hospitalized on the day of their diagnosis); delayed linkage refers to linkage 30 to 90 days after testing was conducted in the ED; unsuccessful linkage refers to lack of linkage within 90 days of ED diagnosis. Number of days to follow-up in HIV clinic was calculated only for patients who were discharged from the ED after the initial encounter. assessment and consent were separate from testing. How these various factors might have influenced the outcomes, as opposed to the effect of opt-in versus opt-out consent, could not be quantified. Second, in the opt-in phase, the exact number of eligible patients was not known because the triage nurse referred only a small proportion of the ED census to the HIV tester. Therefore, we chose to use the number of potentially eligible patients (ED census during hours testing staff on site) in each phase as the denominator when comparing the offer and screening rates. Although this surrogate marker overestimates the true number of eligible patients and thereby reduces the reported offer and screening rates, it does so equally between the 2 phases. Our method of implementing opt-out HIV screening represents only 1 of several methods. Because California state law requires documentation of patient refusal of HIV screening, we added an opt-out signature box to the general consent form, essentially creating an abbreviated written consent form for patients to sign if they declined testing. This requirement for documentation did not permit truly seamless integration of optout screening into care and may not be fully consistent with the intent of CDC recommendations. 1 Last, the initial training sessions for staff may have been insufficient, leading to noncompliance with the protocol. We HIV Screening in an Urban Emergency Department could not monitor how closely all staff complied with the study protocol. DISCUSSION In this report, we compared 2 models of integrating HIV screening into the routine care of ED patients: opt-in HIV screening offered by exogenous staff at triage using separate signed consent, and opt-out HIV screening integrated into the general ED consent registration process. Although we found the screening offer rate to be significantly higher with the opt-out approach, this was counterbalanced by a lower screening acceptance rate and a lower test completion rate, resulting in similar overall screening rates. Incorporating opt-out consent for HIV testing into the general consent increased the screening offer rate nearly 3-fold compared with that in the opt-in phase. Although not formally studied, the reason for this difference may be readily explained. Opt-out consent was linked to registration, a mandatory process for all patients, including those arriving by ambulance, and performed by multiple registration staff at all times. As a result, the entire ED census was assessed for eligibility and all patients deemed eligible were offered HIV screening. During the opt-in phase, by contrast, there was only 1 HIV tester on site to assess eligibility and offer screening. Also, with the opt-in design, many patients arriving by ambulance bypassed the triage area, thereby missing the opportunity to be referred for screening. Many patients were also simply not referred by the triage nurse to the HIV tester, despite the study protocol. We found that the screening acceptance rate during the optout phase was only half that during the opt-in phase. Another ED study comparing outcomes of emergency physician directed HIV testing with opt-out HIV screening reported acceptance rates with registration-based opt-out consent to only be 24%. 9 These findings are dramatically different from those of most other published studies. 4,6,7,12 An ED survey study reported that a similarly high proportion (approximately 80%) of patients would accept HIV screening, regardless of whether opt-in or opt-out approaches were used. 12 This study, however, reflects patients perspectives and no actual testing was conducted. Studies of opt-out HIV screening in non-ed settings have generally found an increase in acceptance rates when switching from opt-in to opt-out screening. 4,6,7 These studies, however, were carried out in prenatal and genitourinary clinics. Patient expectations at a prenatal visit, in which laboratory panel tests are routinely performed, and a genitourinary clinic, in which testing for sexually transmitted infections is expected, may better lend themselves to an opt-out screening approach than an urban ED. The low acceptance rate in the opt-out phase could be due to confounding features of the study design, rather than intrinsic limitations of the opt-out approach itself. It has been demonstrated that acceptance rates are greatest when HIV tests are offered by physicians, suggesting that acceptance rates may depend on who is offering screening. 2,13,14 In our study, optout screening was offered by registration staff and opt-in Annals of Emergency Medicine S93

6 HIV Screening in an Urban Emergency Department screening was offered by HIV testers. Patients may find it easier to decline screening offered by nonclinical registration staff than that offered by dedicated HIV testers. Also, the requirement of an opt-out signature box into the general consent form may have highlighted that HIV testing was different from other tests and increased the likelihood of refusal. Last, some members of the registration staff were observed asking patients to sign in the opt-out signature box without saying the opt-out script. Future studies should systematically monitor for protocol adherence, incorporate ongoing training, and evaluate whether variations in staffing and consent are barriers to acceptance. It has also been shown that opt-in consent is easier for patients to understand than opt-out consent, 12 which raises the possibility that some patients may have misunderstood the optout process, incorrectly signing the opt-out signature box on the general consent form in the belief that they were agreeing to testing. Patient inability to understand the process may be an intrinsic limitation of opt-out screening and requires additional study. Patient comprehension may be improved through optout specific pretest information (verbal or written) and additional staff education and training. The nearly identical overall screening rates in the 2 phases may simply reflect a ceiling on the capacity of a single tester to complete rapid HIV testing rather than the equivalence of opt-in and opt-out screening. During the opt-in phase, the number of patients who accepted testing was roughly the same as the number who were tested because the same person performed these processes sequentially. During the opt-out phase, however, the number of patients who accepted screening at registration could freely exceed the HIV tester s capacity to complete testing. In the opt-out phase, we observed that during busy shifts, HIV testers simply could not keep up with numbers of patients accepting screening. Future opt-out testing programs might be improved with laboratory-based rapid HIV testing that would provide additional testing capacity, continuous availability, and more seamless integration with medical care. Although patients receiving testing appeared similar between phases with respect to basic characteristics, the proportion with newly positive infection during the opt-out phase was 2.5 times higher during the opt-in phase. Reasons for this difference are unclear, and it may have been due to chance. The significance of any differences between study phases with regard to the CD4 counts is also difficult to assess because of the small number of patients with newly positive results. Future research comparing the positive yield and stage of illness at diagnosis between opt-in and opt-out screening approaches may help to clarify our results. All the false-positive rapid test results occurred during the 6-month opt-in phase, and 10 of 11 false-positive results occurred during a 2-month period. After an investigation, the reason for this cluster of false-positive rapid test results remains unknown. Similar clusters of unexplained false-positive rapid test results with oral fluid specimens have been previously reported by several HIV testing programs Exporting any HIV screening protocol from one institution to another is bound to be problematic. Modification of the screening program should be guided by site-specific variables, such as the triage process, physical layout, staffing roles, and registration procedures. Further, HIV screening programs all require ongoing financial and administrative assistance, which might be provided through in-kind medical center support, collaboration with public health departments, research grants, and insurance reimbursements. In several states, amendments to legislation would need to be enacted before implementation of opt-out consent methods for HIV screening. 18,19 In conclusion, integrating opt-out HIV screening into general consent at registration resulted in a higher screening offer rate than opt-in screening by HIV testers. The screening acceptance rate, however, was unexpectedly low with our optout protocol. Potentially confounding features of the study design should be controlled for in future research. The authors acknowledge Harrison J. Alter, MD, MS, for his assistance with database management and statistical expertise. Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article as per ICMJE conflict of interest guidelines (see This study was funded by cooperative agreements from the Centers for Disease Control and Prevention to Dr. White (U18 PS000321). The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention. Publication of this article was supported by Centers for Disease Control and Prevention, Atlanta, GA. White et al Address for correspondence: Douglas A. E. White, MD, Alameda County Medical Center-Highland Hospital, 1411 East 31st St, Oakland, CA 94602; , fax ; daewhite@gmail.com. REFERENCES 1. Branson BM, Handsfield HH, Lampe MA, et al. Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings. MMWR Recomm Rep. 2006;55: Simpson WM, Johnstone FD, Boyd FM, et al. Uptake and acceptability of antenatal HIV testing: randomised controlled trial of different methods of offering the test. BMJ. 1998;316: Stringer EM, Stringer JS, Cliver SP, et al. Evaluation of a new testing policy for human immunodeficiency virus to improve screening rates. Obstet Gynecol. 2001;98: Stanley B, Fraser J, Cox NH. Uptake of HIV screening in genitourinary medicine after change to opt-out consent. BMJ. 2003;326: Creek TL, Ntumy R, Seipone K, et al. Successful introduction of routine opt-out HIV testing in antenatal care in Botswana. J Acquir Immune Defic Syndr. 2007;45: S94 Annals of Emergency Medicine

7 White et al HIV Screening in an Urban Emergency Department 6. Simpson WM, Johnstone FD, Goldberg DJ, et al. Antenatal HIV testing: assessment of a routine voluntary approach. BMJ. 1999; 318: Shagufta Z, Mahto M. Uptake of HIV testing in a genitourinary medicine clinic following opt-out screening method and uptake rate by individual clinicians. Int J STD AIDS. 2007;18: Breese P, Burman W, Shlay J, et al. The effectiveness of a verbal opt-out system for human immunodeficiency virus screening during pregnancy. Obstet Gynecol. 2004;104: Haukoos JS, Hopkins E, Conroy AA, et al, for the Denver Emergency Department HIV Opt-Out Study Group. Routine opt-out rapid HIV screening and detection of HIV infection in emergency department patients. JAMA. 2010;304: Lyons MS, Lindsell CJ, Haukoos JS, et al. Nomenclature and definitions for emergency department human immunodeficiency virus (HIV) testing: report from the 2007 conference of the National Emergency Department HIV Testing Consortium. Acad Emerg Med. 2009;16: White DA, Scribner AN, Schulden JD, et al. Results of a rapid HIV screening and diagnostic testing program in an urban emergency department. Ann Emerg Med. 2009;54: Haukoos JS, Hopkins E, Byyny RL. Patient acceptance of rapid HIV testing practices in an urban emergency department: assessment of the 2006 CDC recommendations for HIV screening in health care settings. Ann Emerg Med. 2008;51: Haukoos JS, Hopkins E, Eliopoulos VT, et al. Development and implementation of a model to improve identification of patients infected with HIV using diagnostic rapid testing in the emergency department. Acad Emerg Med. 2007;14: Lyss SB, Branson BM, Kroc KA, et al. Detecting unsuspected HIV infection with a rapid whole-blood HIV test in an urban emergency department. J Acquir Immune Defic Syndr. 2007;44: Centers for Disease Control and Prevention. False-positive oral fluid rapid HIV tests New York City, MMWR Morb Mortal Wkly Rep. 2008;57: Jafa K, Patel P, Mackellar DA, et al. Investigation of false positive results with an oral fluid rapid HIV-1/2 antibody test. PLoS One. 2007;2:e Delaney KP, Branson BM, Uniyal A, et al. Performance of an oral fluid rapid HIV-1/2 test: experience from four CDC studies. AIDS. 2006;20: Bartlett JG, Branson BM, Fenton K, et al. Opt-out testing for human immunodeficiency virus in the United States: progress and challenges. JAMA. 2008;300: Janssen RS. Implementing HIV screening. Clin Infect Dis. 2007; 45(suppl 4):S Annals of Emergency Medicine S95