Talk to the Veterans Crisis Line now
U.S. flag
An official website of the United States government

VA Health Systems Research

Go to the VA ORD website
Go to the QUERI website

RRP 07-283 – HSR Study

 
RRP 07-283
Evaluating the Implementation of Rapid Testing for HIV infection in a VA Emergency Department
Matthew B Goetz, MD
VA Greater Los Angeles Healthcare System, West Los Angeles, CA
West Los Angeles, CA
Funding Period: March 2008 - September 2008
BACKGROUND/RATIONALE:
HIV testing is justified in patient populations such as in the VA, where the expected sero-prevalence of infection is >0.05%. However, throughout the US and within the VA, many at-risk patients are not tested. Not only are up to 25% of all HIV-infected patients unaware of their positive serostatus, but many opportunities for testing are missed (i.e., infected persons have encounters with medical providers but testing is not offered). Consequently, many patients with HIV infection are diagnosed only when they develop avoidable life-threatening complications. Emergency Department (ED) patients are particularly vulnerable. While >40% of the 23,000 yearly VISN 22 ED patients are at-risk, only 18% of such patients are tested for HIV within one year of an ED visit (VISN 22 data warehouse). At the VA Greater Los Angeles Healthcare System (GLA), 23% of untested at-risk ED patients have no primary care (PC) visit within one year of their ED visit. Finally, compared with other at-risk GLA patients, at-risk ED patients who lack PC visits are younger (mean age 53 years) and more like to be substance users (45%) or homeless (65%).

QUERI-HIV/Hepatitis has previously characterized the barriers to HIV testing, identified solutions, evaluated facilitators, impediments and related costs, and assessed the effectiveness of several quality improvement interventions that target HIV testing performance in IIR 04-023-1 (Improving HIV Screening by Nurse Rapid Testing/Streamlined Counseling; PI: S Asch) and SDP 06-001 (Implementation and Evaluation of a VISN-based program to improve HIV Screening and Testing; PI: M Goetz). Prominent barriers to HIV testing include
- Lack of knowledge of HIV risk factors
- Lack of physician time required to fulfill counseling processes
- Face-to-face post-test counseling is difficult to arrange and patients often miss follow-up appointments

Interventions developed and tested by our group to overcome them are
- Implementation of the electronic HIV Testing Clinical Reminder
- Use of streamlined HIV counseling and Nurse-based pre-test counseling
- Use of HIV rapid testing and academic detailing, social marketing

The effectiveness of our multi-modal interventions has been demonstrated in our ongoing SDP to improve HIV testing rates that utilizes an HIV Testing Clinical Reminder, academic detailing and social marketing (SDP 06-001) and our IIR that has evaluated the effectiveness of nurse-based streamlined HIV counseling and HIV rapid testing (IIR 04-023-1). The one year results from SDP 06-001 demonstrates an increase of HIV testing rates among previously untested at-risk individuals from 8.7% to 20.5%. Similarly results from IIR 04-023-1 indicate that nurse-based streamlined counseling, which covers all the required elements of pre-test counseling and reduces the time of pre-test counseling to 2-3 instead of 10-15 minutes, combined with HIV rapid testing (a 20 minute CLIA-exempt, saliva assay) can double the rate of HIV testing and increase the rate at which patients receive their results from 41% to 90% of persons tested.

Our previous work has focused on increasing HIV testing rates in primary care, mental health, urgent care and substance use clinics. In preparation for a broader roll-out that will build on our initial work, we now propose to use the RRP mechanism to determine the feasibility and effectiveness of nurse-based HIV streamlined counseling and rapid testing for VA ED patients with known risk factors for HIV infection. This proposal is a component of the QUERI-HIV/Hepatitis pipeline of projects and is an essential Phase 1 step towards a large-scale rollout of HIV streamlined counseling and rapid testing. This project has support from the Chief, GLA Laboratory Medicine Service and the GLA ED Medical Director and Nurse Manager.

OBJECTIVE(S):
We will perform a quasi-experimental evaluation of the effectiveness of nurse-based HIV streamlined counseling and HIV rapid testing in the VA GLA ED for patients with known HIV risk factors. Our specific goals are to assess:
The effectiveness of the intervention in increasing HIV testing among at-risk ED department users
The distribution of HIV risk factors among users of the GLA ED and the patterns of use of other VA services among persons who use the ED and have HIV risk factors
The barriers and costs related to implementation of nurse-based HIV streamlined counseling and rapid testing in the GLA ED.
The seroprevalence of undiagnosed HIV infection among tested at-risk ED department users

METHODS:
Interventions:
- Multimodal promotion of HIV testing: Nurses and other providers will be alerted to the presence of HIV risk factors by information provided by the HIV Clinical Reminder (5). This Reminder, which has already been implemented at GLA, prompts healthcare workers to offer HIV testing to persons with the risk factors for HIV exposure that can be extracted from VistA/CPRS. In addition, we will promote the use of streamlined HIV counseling and of rapid HIV testing for ED patients with HIV risk factors that do or do not trigger the HIV Testing Clinical Reminder through educational programs, academic detailing, social marketing and audit feedback techniques. These methodologies are currently in use in SDP 06-001.
- Nurse-based HIV streamlined counseling and rapid testing: We will implement these processes in the ED using the procedures developed under the aegis of IIR 04-023-1.

Evaluation Methods:
- Outcome measure: Our primary outcome measure is the fraction of previously untested, patients with identified risk factors for HIV infection who are tested for HIV infection. We will compare the rate of HIV testing in at-risk patients in the intervention period (May 2007 - October 2007) with that in the prior 6 months. The primary comparison will be the change in the HIV testing rate among GLA ED patients; a secondary measure will be the change in testing rates in at-risk ED patients at the three other VISN 22 facilities with EDs. Data relevant to the rates of HIV testing, distribution of HIV risk factors, seroprevalence of HIV infection and use of VA services among at-risk ED users will be obtained from information in the VISN 22 data warehouse (to which we already have access). With about 3,000 untested, at-risk patients seen per year in the GLA ED we will have >80% power to detect an increase the semi-annual testing rate from 18% to 23%.

- Formative evaluation: Using previous established procedures from IIR 04-023-1 and VA HSR&D SDP 06-001, we will conduct a formative evaluation to assess the barriers to implementation of HIV rapid testing in the GLA ED. The overall aim will be to better understand the influences that impact the success of the intervention by identifying contextually relevant factors (i.e., facilitators and barriers) (7). Semi-structured interviews with key informants from the nurse and physician clinical leadership will provide qualitative data regarding the barriers to offering HIV testing in the ED, the utility of the HIV rapid test and ED patient acceptance of HIV testing.

- Cost Evaluation: We will ask the ED RNs to estimate the time required for streamlined counseling (pre and post-test) and performance of HIV rapid tests. We will then estimate costs of counseling based on the salary of VA licensed nursing personnel. We will then compare these results from those derived from the ongoing cost-effectiveness analyses in IIR 04-023-1

FINDINGS/RESULTS:
Over a 15 week period, 296 patients were deemed eligible for HIV testing based on nursing staffing profiles (i.e., adequate time and training to perform rapid testing) in the VA Greater Los Angeles Healthcare System ED. Of these 296 patients 41% (121 patients) consented to undergoing testing. With no positive test results the estimated upper 5% bound of positive HIV tests is estimated to be 3% in this patient population. This range remains far above the 0.1% prevalence of undiagnosed HIV infection at which routine offers of HIV testing are cost-effective. After the protocol for identifying eligibility evolved to incorporate ED nursing concerns regarding staffing limitations, from a random sampling model to one focused on testing during non-peak hours, the weekly number of tests increased.

Provider surveys indicated 18% of providers favored nontargeted HIV screening, 27% favored the model of testing at non-peak hours (i.e., during period of adequate nurse staffing), 32% supported diagnostic testing, 18% favored no testing or "other". Identified barriers to performance of HIV rapid testing included time constraints (8/22, 36%), ineligibility of patient (illness/capacity to consent/age>65). Other reasons included lack of comfort with sensitive topic, lack of knowledge of test availability and provider's personal assessment of patient's lack of risk. Primary operational barriers included time required for provider to complete separate written informed consent and time required for nurse to document HIV testing in the electronic medical record (both cited by 5/22 staff members (23%). 18% (4/22) could not identify any barriers to HIV RT implementation.

IMPACT:
HIV rapid testing is now available as a routine clinical procedure in the VA Greater Los Angeles Healthcare System Emergency Department. The insights gained as a consequence of the study provide the basis for readdressing the performance of HIV rapid testing in VA Emergency Departments after the implementation of new regulations that remove the requirement for written informed consent which are anticipated to occur in mid August 2009.

The budget impact analysis indicates that the implementation of the Rapid Testing program was never significantly more costly than Usual Care. While Rapid Testing had substantial screening expenses, it had lower inpatient costs, resulting in similar overall costs between the two programs. Assuming a 1% prevalence of disease, the cost of Rapid Testing was $1,418,088 versus $1,320,338 for Usual Care (p = 0.5854). These results support implementation of routine rapid HIV screening in integrated systems such as the VA.


External Links for this Project

Dimensions for VA

Dimensions for VA is a web-based tool available to VA staff that enables detailed searches of published research and research projects.

Learn more about Dimensions for VA.

VA staff not currently on the VA network can access Dimensions by registering for an account using their VA email address.
    Search Dimensions for this project

PUBLICATIONS:

Journal Articles

  1. Gidwani R, Goetz MB, Kominski G, Asch S, Mattocks K, Samet JH, Justice A, Gandhi N, Needleman J. A budget impact analysis of rapid human immunodeficiency virus screening in Veterans Administration emergency departments. The Journal of emergency medicine. 2012 Jun 1; 42(6):719-26. [view]
  2. Chen JC, Goetz MB, Feld JE, Taylor A, Anaya H, Burgess J, Flores Rde M, Gidwani RA, Knapp H, Ocampo EH, Asch SM. A provider participatory implementation model for HIV testing in an ED. The American journal of emergency medicine. 2011 May 1; 29(4):418-26. [view]
  3. Saifu HN, Asch SM, Goetz MB, Smith JP, Graber CJ, Schaberg D, Sun BC. Evaluation of human immunodeficiency virus and hepatitis C telemedicine clinics. The American journal of managed care. 2012 Apr 1; 18(4):207-12. [view]
  4. Knapp H, Hagedorn H, Anaya HD. HIV rapid testing in a Veterans Affairs hospital ED setting: a 5-year sustainability evaluation. The American journal of emergency medicine. 2014 Aug 1; 32(8):878-83. [view]


DRA: none
DRE: Diagnosis, Prevention
Keywords: Emergency care, HIV/AIDS
MeSH Terms: none

Questions about the HSR website? Email the Web Team

Any health information on this website is strictly for informational purposes and is not intended as medical advice. It should not be used to diagnose or treat any condition.