The study was a non-blinded randomized trial with two arms. Both groups viewed a previously studied pre-test counseling video which contained all the elements required by New York State law to be discussed in pre-test counseling. It included information on HIV transmission, definition of AIDS and HIV infection, the nature and meaning of the HIV test, benefits of testing, reporting, partner notification and the definition of voluntary and mandatory testing. After the patients viewed this video, the research assistants administered the rapid HIV test. The control group received standard HIV post-test, in-person counseling while waiting for rapid HIV test results. The intervention group viewed an HIV prevention video in either English or Spanish. After completing an evaluation survey of the video, the intervention group received standard HIV post-test, in-person counseling. Both groups received their rapid HIV test result after their post-test counseling session. If the result was negative, further counseling was available if the patient had more questions or if he/she was at a high-risk for contracting the virus. Patients who tested positive were informed of their results by a provider and walked to the HIV clinic in the hospital by the research assistant. The research protocol was approved by the institutional review boards (IRB) at the Albert Einstein College of Medicine and the Health and Hospitals Corporation.
A convenience sample of stable patients recruited in the Jacobi Medical Center Urgent Care Area (UCA), a walk-in section of the adult ED. Jacobi Medical Center is an inner-city hospital, level I trauma and tertiary care center, located in the Bronx, New York. Emergency Medicine physicians staff the UCA Monday through Friday from 8am to 4pm. All patients seen in the UCA are ambulatory and non-acute.
Selection of Participants
Eligible patients for enrollment in this study were those 18 years of age and older visiting the UCA for any reason, between July 5, 2005 and October 5, 2005. This was a non-targeted HIV testing approach, all patients who did not fit exclusion criteria were recruited for the study. Everyone who participated received point-of-care testing. Patients were ineligible for the study if they were clinically unstable, had a previously confirmed diagnosis of HIV, tested for HIV within the past six-months or were unable to understand the consent process.
We developed a fifteen-minute rapid HIV post-test video that focused on the use of condoms as a positive behavior. The video demonstrated how to properly use male and female condoms and dental dams. Additional information in the video included interpretation of HIV test results, partner notification and domestic violence information. The education and prevention messages in the video were conveyed by one of the co-authors. The video was explicitly designed to include information that conveyed the essential elements that are required to be discussed within the post-test counseling session, as per New York State law. The video was appropriate for an eighth grade language level and was also translated into Spanish.
Data Collection and Processing
Two research assistants in the UCA used a standardized scripted approach to describe the study to potential participants. Written informed consent was obtained in either English or Spanish. Individuals who declined to participate were verbally consented to obtain reasons for their refusal. All eligible participants who agreed to participate in the study were randomized to either the control group or the intervention group. A randomization scheme was generated using a computer-generated block randomization, (available online at www.randomization.com
). Randomization assignments were placed in sealed opaque envelopes that were sequentially opened by the research assistants after the patient signed the informed consent.
After consenting, participants first responded to a demographic survey, a risk assessment survey, a self-efficacy measure and a condom intention measure. All participants then received pre-test counseling by viewing the same 6-minute pre-HIV test video. It contained all the elements required by New York State law to be discussed in pre-test counseling (as stated in the study design section). Post-test HIV counseling is required by New York State law and followed the viewing of the HIV prevention video. The control group received standard HIV post-test, in-person counseling, while waiting for their HIV test results. The intervention group viewed the HIV prevention video and then received standard HIV post-test, in-person counseling while waiting for HIV test results. The post-test video included a demonstration of how to properly use condoms, and information on interpretation of HIV test results, partner notification and domestic violence. Since the HIV-prevention video was an untested method of conveying information, standard HIV post-test counseling was also given, to provide the standard of care to patients randomized to the intervention arm of the study. (See )
Profile of randomized control trial
All participants were administered a ten-question true-false measure of retained information. Measures were administered in either English or Spanish. Those in the control group took the true-false knowledge measure after receiving post-test counseling. Those in the intervention group viewed the HIV prevention video, took the same ten-question measure of retained information, and then had a standard counseling session after the measure was completed.
The primary outcome for the study was the participant’s score on the ten-question measure. To create this measure, questions were adapted from the New York State Department of Health HIV Questions & Answers Booklet and modified to true-false responses at a fifth-grade reading level, as determined by Microsoft Word’s Flesh-Kincade grade level instrument (Microsoft Corp, Redmond, WA). The measure was circulated among the co-investigators, including HIV and Infectious Disease specialists, for consensus. The measure was initially created in English, and then translated into Spanish (see appendix A
). The primary outcome measure, the number of correct answers on the ten-question measure, formed a score that ranged from 0% to 100% correct.
Equivalence methods were used in order to evaluate whether the video was at least as good as the counselor in conveying knowledge about HIV prevention. The null hypothesis for the equivalence method is that the mean score on the true-false measure assessing retained knowledge of HIV post-test counseling in the intervention group is lower than the mean in the counseling group. The alternative hypothesis is that the mean score of the intervention is at least as high as the mean exam score of the counseling group. The groups would be considered equivalent if the mean score on the measure in the intervention group was no more than 5% lower than the mean score of those who saw the counselor. We calculated the 90% equivalence confidence interval (CI) and the corresponding one-tailed test around the difference in exam scores.
Sample size calculation, determined a priori, required 59 patients in each group for a power of 85%, equivalence limit of −5% and an alpha of 0.05. nQuery Advisor release 3.0 (Cork, Ireland, 1999) was used to calculate the sample size. Baseline characteristics of the intervention and control groups, as well as incidence of HIV positive patients, were compared using Student’s t-test for continuous variables and chi-square tests for categorical variables. All statistical analyses were performed using Stata SE 9.2 (StataCorp, College Station, TX). For all analyses, p-values are 2-tailed with an alpha of 0.05 considered for statistical significance. EquivTest 1.0 (Cork, Ireland, 1998) was used to assess equivalence.17