Factors associated with patients who prefer HIV self-testing over health professional testing in an emergency department-based rapid HIV screening program

Abstract

Kiosk-facilitated HIV self-testing has been shown to be accurate and well accepted by emergency department (ED) patients. We investigated factors associated with patients who preferred self-testing over testing performed by health professionals in an ED-based HIV screening program. This opt-in program evaluation studied 332 patients in an inner-city academic ED from February 2012 to April 2012, when a kiosk-based HIV self-testing program was standard of care. The first kiosk in the 2-stage system registered patients and assessed their interest in screening, while the second kiosk gathered demographic and risk factor information and also provided self-testing instructions. Patients who declined to self-test were offered testing by staff. Broad eligibility included patients aged 18–64 years who were not critically ill, English-speaking, able to provide informed consent, and registered during HIV program operational hours. Data were analyzed using descriptive statistical analysis and Chi squared tests; 160 (48.2%) of 332 patients consenting to testing chose to use a kiosk to guide them performing self-testing. Patients aged 25–29 years and those whose primary ED diagnosis was not infectious disease-related were more likely to prefer HIV self-testing (OR = 2.19, 95% CI: 1.17–4.10; OR = 1.79, 95% CI: 1.03–3.12). HIV self-testing in the ED could serve as a complementary testing approach to the conventional modality.

Keywords

Emergency department HIV screening self-testing

Introduction

The Centers for Disease Control and Prevention (CDC) has suggested that emergency departments (EDs) are critical venues for HIV screening.¹ Since 2006, U.S. EDs have successfully identified thousands of previously undiagnosed cases of HIV.² EDs can play a pivotal role towards improving the HIV Care Continuum in the United States.^3–5

However, recent data demonstrated that HIV testing is only performed during 0.2% of U.S. ED visits for patients aged 13–64 years.⁶ Reasons for this gap are unclear but may be due to staff perception of opportunity, and work flow issues.^7–12 Additionally, challenges exist in engaging patients in acute care for HIV testing because of their chief complaint issue or their concern about confidentiality.

Researchers have developed innovative strategies to adapt HIV testing into routine practice such as ED-based kiosk-facilitated HIV testing.¹³ Our group developed a prototype computerized kiosk system as a means of offering HIV testing and other public health information to ED patients,¹⁴ after a pilot study using tablet-based kiosks for facilitating HIV testing in our ED.¹⁵ Following the success of our prototype system, we implemented a kiosk-facilitated HIV screening program to engage patients at ED registration. Our results demonstrated that a kiosk-driven program offering HIV testing is not only feasible but also engages more high-risk patients for testing.¹⁶

HIV self-testing in ED settings is another novel approach we have successfully explored to increase patient engagement without increasing burden on ED clinical staff.¹⁵ We also studied the feasibility of using a tablet kiosk to guide ED patients to perform HIV self-testing using a rapid oral fluid HIV test. This pilot study found that tablet-facilitated HIV self-testing was highly accurate and well-accepted by patients.¹⁵ Combining the successful outcomes from both the kiosk-facilitated HIV screening program and the tablet-based self-testing feasibility study, we implemented a pilot rapid oral fluid HIV self-testing program as an option for patients engaged in the kiosk-facilitated HIV screening program.

Our aim was to determine which factors were associated with patients who were more likely to accept kiosk-facilitated self-testing by exploring socio-demographic information, computer/kiosk technology experience, behavior patterns, and ED visit characteristics compared to those who chose conventional HIV testing by trained clinical staff.

Methods

Our program was implemented in an inner-city academic ED with approximately 65,000 annual patient visits in 2012 and an HIV seroprevalence rate of approximately 8%.⁵ The study took place on 20 weekdays between 20 March 2012 and 27 April 2012 when kiosk-based self-testing was standard of care. Eligibility included English-speaking patients aged 18–64 years able to provide informed consent, not critically ill, and registered as an ED patient during HIV program operational hours. Patients were directed by HIV testing staff to a touchscreen computerized kiosk near registration/triage area to respond to questions regarding general medical and public health information questions as well as their interest in HIV testing after the routine triage process. Then, eligible patients were directed to the backend kiosk where the HIV testing staff provided verbal opt-in consent for HIV testing. Consented patients then responded to a series of questions on the backend kiosk, including the option for either performing an HIV test for themselves or having trained testing staff to test them. The quality improvement of the HIV screening program was approved by the Johns Hopkins University Institutional Review Board.

The first kiosk in the 2-stage system registered patients and assessed their interest in screening, while the second kiosk served to gather demographic and risk factor information and provide self-testing instructions. This ‘‘Risk Assessment Module’’ included up to 26 screens, depending on user responses. Screens included: login; confirmation of desire to test; the option of self-testing; a series of screens regarding procedural instructions if self-testing was selected; sociodemographic data collection (including computer and kiosk experience); HIV risk assessment questions; and a survey on the patient’s preferences and ease of use regarding the kiosk program. The instructions guided patients in collecting and testing the oral fluid specimen (Figure 1). HIV test results were read by program staff and patients were asked to return to the booth 20 minutes later to receive their results. HIV testing staff were on hand to observe patients performing self-testing and to provide assistance as needed. Patients who declined self-testing were tested by trained testing staff.

Figure 1.

Selected computerized touch-screen kiosk screens for HIV self-testing instructions.

Demographic and visit data were obtained from ED administrative data while HIV testing/self-testing, computer/kiosk experience, risk behavior assessment were from the testing program. Data regarding ED diagnoses were coded according to the International Classification of Disease, Ninth Revision (ICD-9). We classified the primary ED diagnosis of each visit as infectious or non-infectious based on a previously described protocol (see Table 1 in the Supplement).¹⁷ Data were first analyzed using descriptive analysis. Multiple imputation technique for missing values was performed for the questions when >20% of patients did not respond on the kiosk to minimize bias. Chi square tests were performed for bivariate analysis and logistic regression was performed using SAS V.9.4 (SAS Institute, Cary, North Carolina, USA) followed by multiple imputation. Variables with a p-value < 0.2 in bivariate analysis and other variables a priori considered as potential important confounders were entered into a full multivariate logistic regression model for stepwise model selection approach for a final multivariate regression model.

Table 1.

Characteristics of 332 patients who were offered an HIV test in an emergency department-based HIV screening program by HIV self-testing status.

Characteristics	Categories	Number	Number (%) patients who chose to perform HIV self-testing
Overall		332	160 (48.2)
Age (years)	18–24	106	50 (47.8)
	25–29	60	37 (61.7)
	30–39	52	22 (42.3)
	40–49	62	26 (41.9)
	50–64	52	25 (48.1)
Gender	Female	184	90 (48.9)
	Male	147	70 (47.6)
	Transgender	1	0 (0.0)
Race	African American	231	107 (46.3)
	White	40	22 (55.0)
	Other	61	31 (50.8)
Ethnicity	Hispanic	6	4 (66.7)
	Non-Hispanic	326	156 (47.8)
Highest education level	High school diploma or less	155	72 (46.5)
	Some college or more	110	64 (58.2)
	Missing	67	24 (35.8)
High risk behavior^a	Yes	228	119 (52.2)
	No	37	17 (46.0)
	Missing	67	24 (35.8)
Computer use	<30 min	114	60 (52.6)
(per day)	≥30 min	152	77 (50.7)
	Missing	66	23 (34.8)
Computer use	<3 days	115	55 (47.8)
(per week)	≥3 days	151	82 (54.3)
	Missing	66	23 (34.8)
Previous kiosk experience	Yes	149	73 (49.0)
Previous kiosk experience	No	120	64 (53.3)
	Missing	63	23 (36.5)
Triage acuity level	3–‘Urgent’	229	112 (48.9)
	4–‘Less Urgent’	97	44 (45.4)
	5–‘Non-Urgent’	1	0 (0.0)
	Missing	5	4 (80.0)
Chief complaint	Abdominal pain	27	14 (51.9)
	Headache	14	7 (50.0)
	Back pain	12	7 (58.3)
	Sore throat	11	5 (45.5)
	Other	268	160 (48.2)
Primary ED diagnosis	Infectious diseases	67	25 (37.3)
	Other	265	135 (50.9)

High risk behaviors were defined as: men who have sex with men (MSM), injection drug use, having sex with both genders, having >4 sexual partners in the past year, having sex with >1 person in the past three months, having a new sexual partner in the past three months, ever having had receptive anal sex, having anal sex in the past three months, having a new anal sex partner in the past three months, having >1 anal sex partner in the past three months, having anonymous sex, having sex with an HIV-positive person, having sex with an injection drug user, having sex with MSM, having sex with someone without knowing their HIV status, having sex with anyone who trades sex for drugs or money, having sex against your will, not always using condoms when having sex, or having a partner who has a sexually transmitted infection.

Results

During the study period, 332 ED patients consented for HIV testing. The majority were female (55%), African American (70%), and 18–39 years (66%) (Table 1). The majority had a triage acuity level of 3, at which patients required urgent attention from ED providers (69%). The leading chief complaint was abdominal pain (8%), followed by headache (4%), back pain (4%), and sore throat (3%); 20% of the patients had an infectious disease-related primary ED diagnosis.

Among 332 patients, 160 (48.2%) patients opted for self-testing (Figure 2). No patients performing HIV self-tests were reactive for HIV; one patient tested by clinical staff had a reactive test. In the bivariate analysis, we found that patients who were 25–29 years or who did not have an infectious disease-related primary ED diagnosis were more likely to test themselves than their counterparts (p = 0.017 and p = 0.048, respectively). Patients with higher education levels (some college or higher) were marginally more likely to test themselves than those with high school education or less (p = 0.092). Gender, race, high risk behavior for HIV infection, computer or kiosk technology experience, chief complaint, and triage acuity were not associated with HIV self-testing. In the multivariate analysis, age of 25–29 years and absence of an infectious disease-related primary ED diagnosis were associated with HIV self-testing (OR = 2.19, 95% CI: 1.17, 4.10; OR = 1.79, 95% CI: 1.03, 3.12, respectively) (Table 2).

Figure 2.

A flow diagram of outcomes of 332 emergency department patients who opted for HIV self-testing and who opted for HIV staff-based testing.

Table 2.

Multivariate analysis of factors associated with emergency department patients who opted to choose self-testing for HIV.

		Full model	Final model
Characteristics	Categories^a	OR (95% CI)	OR (95% CI)
Age (years)	25–29	1.85 (0.94, 3.64)	2.19 (1.17, 4.10)
	18–24 or ≥30	1.00	1.00
Gender	Female	1.02 (0.64, 1.60)	–
	Male or transgender	1.00
Race	African American	1.13 (0.61, 2.09)	–
	White	0.97 (0.41, 2.25)	–
	Other race	1.00
Highest education	High school or less	1.49 (0.74, 2.98)	–
level	Some college or higher	1.00
High risk behavior^b	Yes	0.68 (0.34, 1,37)	–
	No	1.00
Computer use	<30 min	0.70 (0.39, 1.25)	–
(per day)	≥30 min	1.00
Computer use	<3 days	1.29 (0.63, 2.62)	–
(per week)	≥3 days	1.00
Previous kiosk	Yes	0.76 (0.46, 1.24)	–
experience	No	1.00
Triage acuity level	‘Urgent’	0.91 (0.55, 1.51)	–
	‘Less Urgent’ or ‘Non-Urgent’	1.00
Primary ED diagnosis	Non-infectious diseases	1.71 (0.96, 3.03)	1.79 (1.03, 3.12)
	Infectious diseases	1.00	1.00

High risk behaviors are defined as: men who have sex with men (MSM), injection drug use, having sex with both genders, having >4 sexual partners in the past year, having sex with >1 person in the past three months, having a new sexual partner in the past three months, ever having had receptive anal sex, having anal sex in the past three months, having a new anal sex partner in the past three months, having >1 anal sex partner in the past three months, having anonymous sex, having sex with an HIV-positive person, having sex with an injection drug user, having sex with MSM, having sex with someone without knowing their HIV status, having sex with anyone who trades sex for drugs or money, having sex against your will, not always using condoms when having sex, or having a partner who has a sexually transmitted infection.

Multiple imputation technique for missing values.

Discussion

The results of this pilot HIV kiosk-self-testing program demonstrated that ED patients were accepting HIV self-testing in the ED, as approximately half of the interested eligible patients chose to test themselves. The magnitude of preference and/or acceptability for self-testing was similar to the participation rate (49.5%) for our earlier ED HIV self-testing research study¹⁵ and was slightly lower than a hypothetical acceptance rate (56.2%) in a phone survey study in New York City.¹⁸ HIV self-testing in healthcare settings is an alternative, innovative approach to empower and engage patients in testing,^19,20 and could minimize burden on clinical providers and eliminate the need for additional testing staff. Coupling self-testing with an electronic vending machine for dispensing the self-testing kit,²¹ may provide a cost-effective HIV testing program model with long-term sustainability in busy acute care settings.

Our study may be the first to investigate factors associated with preference of HIV self-testing in the clinical setting. We identified age of 25–29 years and absence of primary ED diagnosis of infectious diseases as two independent correlates with the preference of HIV self-testing in EDs. Younger age groups (18–24 years and 25–44 years vs. 45–64 years) have been identified as key factors associated with perceived acceptability of HIV home testing, if the test were available based on a telephone survey study conducted in New York City prior to the Food and Drug Administration’s approval of the use of home self-testing HIV kits.¹⁸ It appears that HIV self-testing in EDs could engage more young adults for HIV testing, an important demographic group that is at higher risk for HIV infection. Interestingly, an acute infectious disease-related ailment was correlated with decreased willingness to self-test, even though patients expressed interest in HIV testing. One possible explanation is that this subset did not feel well enough (due to symptoms of an acute infection) to follow the HIV self-testing kiosk instructions and chose conventional healthcare professional testing instead.

Some study aspects limit interpretation of our findings. First, a long-standing oral-fluid rapid HIV screening program was launched in 2005 and several research projects on HIV self-testing have been conducted in our ED. Our patients might be more familiar with the technology and procedures of the HIV test than those in other EDs without similar programs. Our acceptance rate of HIV self-testing might be higher and the factors associated with self-testing might be different from others. Additionally, approximately one-third of participants opted not to respond to socio-demographic, computer/kiosk experience, and risk behavior questions on the back-end kiosk. The data analysis on these variables might not truly reflect their influence on the acceptance of self-testing for HIV in the ED, even though an advanced multiple imputation technique for missing values was employed to mitigate the impact of missing data. Finally, we were not able to detect acute HIV infection using this third-generation oral-fluid point-of-care test.

In conclusion, approximately 50% of ED patients who accepted HIV testing chose to perform self-testing by following the instructions provided by a kiosk. Preference of self-testing was positively correlated with young age but negatively correlated with an ED primary diagnosis of infectious disease. With 50% of patients choosing conventional healthcare professional testing over self-testing in the ED, our findings indicate that ED patient self-testing for HIV is likely a complementary testing approach to the current, traditional format. Future research should be directed towards other novel strategies (e.g. electronic vending machine with a result reader) to couple with self-testing to increase patient engagement in HIV testing in acute care settings.

Footnotes

Acknowledgement

Part of this work was presented at the annual meeting of the Society for Academic Emergency Medicine, 14–18 May 2013, Atlanta, GA.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: HIV screening program for this study was supported by Gilead Sciences, Inc.’s HIV FOCUS program and the Baltimore City Health Department. Hsieh, Rothman, and Gaydos are also supported by NIH U54EB007958. Dr Hsieh is also supported in part by a National Institutes of Health award, K01AI100681 from National Institute of Allergy and Infectious Diseases.

References

Branson

Handsfield

Lampe

et al.

Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings. MMWR Recomm Rep 2006; 55: 1–17.

Centers for Disease Control and Prevention. Results of the expanded HIV testing initiative – 25 jurisdictions, United States, 2007-2010. MMWR Morb Mortal Wkly Rep 2011; 60: 805–810.

Gardner

McLees

Steiner

et al.

The spectrum of engagement in HIV care and its relevance to test-and-treat strategies for prevention of HIV infection. Clin Infect Dis 2011; 52: 793–800.

Hsieh

Kelen

Laeyendecker

et al.

HIV care continuum for HIV-infected emergency department patients in an inner-city academic emergency department. Ann Emerg Med 2015; 66: 69–78.

Kelen

Hsieh

Y-H

Rothman

et al.

Improvements in the continuum of HIV care in an inner-city emergency department. AIDS 2016; 30: 113–120.

Hoover

Tao

Heffelfinger

. Monitoring HIV testing at visits to emergency departments in the United States: Very-low rate of HIV testing. J Acquir Immune Defic Syndr 2013; 62: 90–94.

Akhter

Gorelick

Beckmann

. Rapid human immunodeficiency virus testing in the pediatric emergency department: A national survey of attitudes among pediatric emergency practitioners. Pediatr Emerg Care 2012; 28: 1257–1262.

Arbelaez

Wright

Losina

et al.

Emergency provider attitudes and barriers to universal HIV testing in the emergency department. J Emerg Med 2012; 42: 7–14.

Brown

Shesser

Simon

. Establishing an ED HIV screening program: Lessons from the front lines. Acad Emerg Med 2007; 14: 658–661.

10.

Burke

Sepkowitz

Bernstein

et al.

Why don't physicians test for HIV? A review of the US literature. AIDS 2007; 21: 1617–1624.

11.

Hsieh

Y-H

Jung

Shahan

et al.

Emergency medicine resident attitudes and perceptions of HIV testing before and after a focused training program and testing implementation. Acad Emerg Med 2009; 16: 1165–1173.

12.

Signer

Peterson

Hsieh

et al.

Scaling up HIV testing in an academic emergency department: An integrated testing model with rapid fourth-generation and point-of-care testing. Pub Health Rep 2016; 131: 82–89.

13.

Haukoos

Hopkins

Bender

et al.

Use of kiosks and patient understanding of opt-out and opt-in consent for routine rapid human immunodeficiency virus screening in the emergency department. Acad Emerg Med 2012; 19: 287–293.

14.

Rothman

Gauvey-Kern

Woodfield

et al.

Streamlining HIV testing in the emergency department – Leveraging kiosks to provide true universal screening: A usability study. Telemed J E Health 2014; 20: 122–127.

15.

Gaydos

Solis

Hsieh

et al.

Use of tablet-based kiosks in the emergency department to guide patient HIV self-testing with a point-of-care oral fluid test. Int J STD AIDS 2013; 21: 716–721.

16.

Hsieh

Y-H

Gauvey-Kern

Peterson

et al.

An emergency department registration kiosk can increase HIV screening in high risk patients. J Telemed Telecare 2014; 20: 454–459.

17.

Armstrong

Conn

Pinner

. Trends in infectious disease mortality in the United States during the 20th century. JAMA 1999; 281: 61–66.

18.

Myers

Bodach

Cutler

et al.

Acceptability of home self-tests for HIV in New York City, 2006. Am J Public Health 2014; 104: e46–e48.

19.

Pant Pai

Sharma

Shivkumar

et al.

Supervised and unsupervised self-testing for HIV in high- and low-risk populations: A systematic review. PLoS Med 2013; 10: e1001414–e1001414.

20.

Youngs

Hooper

. Ethical implications of HIV self-testing. J Med Ethics 2015; 41: 809–813.

21.

Young

Klausner

Fynn

et al.

Electronic vending machines for dispensing rapid HIV self-testing kits: A case study. AIDS Care 2014; 26: 267–269.