Conceived and designed the experiments: CJH JS TPG JAD MM ADC. Performed the experiments: CJH JS. Analyzed the data: CJH JS TPG JAD ADC. Contributed reagents/materials/analysis tools: JS TPG MM. Wrote the paper: CJH JS TPG JAD MM NR AS SLC ADC.
Although funding has supported the scale up of routine, opt-out HIV testing in the US, variance in implementation mechanisms and barriers in high-burden jurisdictions remains unknown.
We conducted a survey of health care organizations in Washington, DC and Houston/Harris County to determine number of HIV tests completed in 2011, policy and practices associated with HIV testing, funding mechanisms, and reported barriers to testing in each jurisdiction and to compare results between jurisdictions.
In 2012, 43 Houston and 35 DC HIV-testing organizations participated in the survey. Participants represented 85% of Department of Health-supported testers in DC and 90% of Department of Health-supported testers in Houston. The median number of tests per organization was 568 in DC and 1045 in Houston. Approximately 50% of organizations in both DC and Houston exclusively used opt-in consent and most conducted both pre- and post-test counseling with HIV testing (80% of organizations in DC, 70% in Houston). While the most frequent source of funding in DC was the Department of Health, Houston organizations primarily billed the patient or third-party payers. Barriers to testing most often reported were lack of funding, followed by patient discomfort/refusal with more barriers reported in DC.
Given unique policies, resources and programmatic contexts, DC and Houston have taken different approaches to support routine testing. Many organizations in both cities reported opt-in consent approaches and pre-test counseling, suggesting 2006 national HIV testing recommendations are not being followed consistently. Addressing the barriers to testing identified in each jurisdiction may improve expansion of testing.
HIV testing is an important step of the HIV care continuum and a critical component of prevention programs throughout the United States (US). Nationally, an estimated 18% of infected persons remain undiagnosed.
CDC funding has since supported the scale up of routine, opt-out screening nationwide,
This article reports on some of the similarities and differences in HIV testing observed in two jurisdictions highly impacted by HIV, Washington DC and Houston, TX. The jurisdictions are two of the twelve urban areas that represent 44% of the nation's AIDS cases.
| Washington, DC | Houston, TX | |
| Population (persons ≥13 yrs; Census estimates for 2010) | 528,109 | 3,249,542 |
| Number of living HIV/AIDS Cases ≥13 yrs (reported as of 2010) | 14,465 | 19,943 |
| HIV/AIDS Prevalence Rate (as of 2010) | 2.7 | 0.6 |
| Type of epidemic | Generalized | Concentrated |
| Initiation of support for routine testing (year) | 2006 | 2008 |
| HIV testing consent and counseling regulations or statutes | No | Chapter 81 of Texas Health and Safety Code |
| Mandatory HIV testing | Mandatory testing of convicted sex offenders within the criminal justice system | Mandatory testing of all offenders upon entry and release within Texas Department of Criminal Justice Texas law allows for mandatory testing in county and municipal jails |
| Texas law allows for mandatory testing in county and municipal jails | ||
| Established HIV testing reimbursement and billing laws or regulations (year) | Insurance Coverage for HIV Testing in Emergency Departments Amendment Act (2008) | No specific laws on reimbursement or billing |
| Estimated number of annual publically-funded HIV tests conducted | 129,464 (2011) | 107,458 (2011) |
| Testing campaigns/Programmatic initiatives/Research activities to scale-up HIV testing | Come Together DC Get Screened for HIV (2006) | HIP HOP for HIV Awareness (2007) |
| Ask for the Test/Offer the Test (2009) | Expanded Testing Initiative (2008) | |
| Enhanced Comprehensive HIV Prevention Planning Initiative (2010) | Routine Screening for HIV- Memorial Hermann Health System and Project RUSH at Harris Health System (2008) | |
| HIV Prevention Trials Network 065 (intervention community) (2010) | Enhanced Comprehensive HIV Prevention Planning Initiative (2010) | |
| HIV Prevention Trials Network 065 (control community) (2010) |
Jurisdiction is Houston/Harris County. Population data are for Harris County.
Since 1999, opt-out HIV testing has been required by Texas law during pregnancy at the first prenatal visit and at delivery. As of 2010, opt-out HIV testing is required during the third trimester. If no record of third trimester testing is found, opt-out testing at time of delivery is required. For all other persons (not incarcerated), general written consent for HIV testing is allowable. Informed consent may be verbal as long as test explanation and consent is documented in patient's medical record. Opt-in versus opt-out approach is not specified.
Policy in both jurisdictions has been relatively permissive of the expansion of routine HIV testing. Statutes relevant to the 2006 recommendations in DC and Texas were either neutral or consistent with all routine, opt-out recommendations.
Since 2006, the DC Department of Health (DOH) has supported targeted testing and routine, opt-out HIV testing implementation as a standard of care in health care settings.
In Houston, the CDC-funded Expanded Testing Initiative (ETI) started in 2008 when the Houston Department of Health and Human Services (HDHHS) began supporting routine, opt-out HIV testing in the EDs of two large hospitals and two community health centers. The Initiative grew to encompass four large EDs by 2009. With funding from the Texas Department of State Health Services (DSHS), routine testing was further expanded in 2010 to five additional EDs and twelve additional community health centers. Targeted testing was also funded by the HDHHS at CBOs for outreach to high risk populations and geographic areas. In 2011, the HDHHS funded over 107,000 tests in the Houston area.
Both DC and Houston health departments have provided support for the provision and funding of HIV testing in community and health care settings, including implementation of social marketing campaigns. HIV testing and promotion of HIV testing was further scaled up with the influx of HIV prevention activities funded by the CDC's Enhanced Comprehensive HIV Prevention Planning (ECHPP) Project in both DC and Houston. ECHPP was a 3-year demonstration project designed to improve program planning and implementation in support of the National HIV/AIDS Strategy.
As part of ECHPP efforts, partnerships were formed between academic institutions and public health departments to improve program planning of HIV prevention activities in both DC and Houston. These partnerships ensured that local research capacity was improved and policy and programming stakeholders were invested in the study's outcomes, both of which are enabling factors for success in operational research.
Surveys were designed to gather organizational-level data on sites conducting HIV testing in each jurisdiction in 2011. Selected survey questions were adapted from the National Association of Community Health Centers survey on HIV testing,
Sampling frames were designed to best fit the model of HIV prevention activities unique to each jurisdiction. In DC, the sampling strategy was developed through a partnership between DC DOH and researchers at the George Washington University as part of ECHPP. In 2011, DC DOH's HIV/AIDS, Hepatitis, STD, and TB Administration (HAHSTA)'s prevention model primarily supported HIV testing through distribution of HIV rapid test kits at no cost to approximately forty organizations. With stakeholder and DOH input, it was determined that these organizations conducted the vast majority of testing within the city. Therefore, organizations were identified for survey inclusion if their HIV testing services were HAHSTA-supported. These organizations included CBOs, clinics, and hospitals.
In Houston, the sampling frame and sampling tool were designed jointly by staff and researchers from HDHHS, Houston Area HIV Services Ryan White Planning Council's Office of Support, and Baylor-UTHouston Center for AIDS Research, also as part of ECHPP. The HDHHS funded two distinct HIV testing services at 10 organizations in 2011: 1) targeted HIV testing by CBOs that included prevention counseling, and 2) routine, opt-out HIV testing in EDs and county clinics. HIV testing in Houston was supported by a diverse mix of both public and private support, thus sampled organizations included both those that were and were not financially supported by the HDHHS. Utilizing a methodology described previously,
In DC, surveys were emailed to HIV testing coordinators at each organization. Confidential surveys were administered online using Research Electronic Data Capture (REDCap). If no response was received within one month of invitation, paper surveys were mailed in addition to follow up phone calls. Participants were compensated for survey completion.
In Houston, each organization was emailed a formal invitation letter, followed by phone calls if there was no response via email. Initial contact was with HIV program directors, nurse managers, and/or lab directors. Respondents were selected based on their authority to make HIV testing program decisions and/or their knowledge of HIV testing activities within the organization. When possible, surveys were completed by an in-person interview, otherwise a telephone interview was conducted. Participants were not compensated for survey completion.
In DC, data collection and storage were via REDCap. Data analysis was conducted using SAS 9.2 (SAS Institute, Cary, NC). In Houston, all data were entered and managed in Excel. Data analysis was conducted using SAS 9.3 (SAS Institute, Cary, NC). Summary data from both jurisdictions were compiled via Excel and analyzed collaboratively by researchers from both jurisdictions.
This study was approved by the Institutional Review Boards of The George Washington University, the District of Columbia Department of Health (Washington, DC) and the Baylor College of Medicine (Houston). In Houston, the Institutional Review Board approved verbal consent, with completion of the survey as documentation of consent. Consent was also documented in a secure database by the interviewer. Written consent would have been the only record linking the participant and survey; therefore, the potential harm resulting from a breach of confidentiality would have increased since no other personal identifiers were gathered. In DC, the Institutional Review Board approved the use of written consent. The survey instructions asked participants to mark “yes” or “no” to indicate their consent to the survey prior to answering any further questions. In both jurisdictions, the survey was voluntary and deemed no more than minimal risk to participants.
In DC, 41 organizations were contacted for study inclusion. Of these, 35 (85.4%) organizations participated in the survey, representing 103 facilities that conducted HIV testing. Many of the DC organizations were community-based organizations/community service organizations (40.0%) (
| Washington, DC (N = 35) | Houston, TX (N = 43) | |||
| N | % | N | % | |
| Hospital | 12 | 34.3% | 11 | 25.6% |
| Community-based or community service organization | 14 | 40.0% | 10 | 23.3% |
| Clinic or university health center | 9 | 25.7% | 22 | 51.2% |
| Yes | 29 | 82.9% | 27 | 62.8% |
| No | 6 | 17.1% | 16 | 37.2% |
| Yes | 29 | 82.9% | 20 | 46.5% |
| No | 6 | 17.1% | 23 | 53.5% |
| Continuing education | 26 | 74.3% | 20 | 46.5% |
| DOH or clinic training | 23 | 65.7% | 25 | 58.1% |
| Peer-to-peer best practices | 20 | 57.1% | 17 | 39.5% |
| Manuals, guidelines, or literature | 16 | 45.7% | 21 | 48.8% |
| Web seminars | 12 | 34.3% | 11 | 25.6% |
| Consulting site visits | 6 | 17.1% | 11 | 25.6% |
| In-person or video conferences | 5 | 14.3% | 10 | 23.3% |
| Other | 1 | 2.9% | 1 | 2.3% |
| None reported | 2 | 5.7% | 9 | 20.9% |
| Opt in approach ONLY | 18 | 51.4% | 21 | 48.8% |
| Opt out approach ONLY | 17 | 48.6% | 3 | 7.0% |
| Combination of opt in and opt out | 0 | 0.0% | 19 | 44.2% |
| Pre-test counseling only | 1 | 2.9% | 1 | 2.3% |
| Post-test counseling only | 4 | 11.4% | 7 | 16.3% |
| Both pre and post-test counseling | 28 | 80.0% | 30 | 69.8% |
| Neither pre nor post-test counseling | 1 | 2.9% | 5 | 11.6% |
| Other | 1 | 2.9% | 0 | 0.0% |
| Local Health Department Funding (DC, Houston) | 26 | 74.3% | 10 | 23.3% |
| Texas DSHS | N/A | N/A | 8 | 18.6% |
| CDC Funding | 9 | 25.7% | 3 | 7.0% |
| Other Federal Funding (Ryan White, Other HRSA, SAMHSA, VA) | 11 | 31.4% | 7 | 16.3% |
| Private Grant Funding | 7 | 20.0% | 8 | 18.6% |
| Research (HPTN 065/TLC Plus HIV testing or other research study) | 12 | 34.3% | 2 | 4.7% |
| Medicaid | 10 | 28.6% | 23 | 53.5% |
| Medicare | 4 | 11.4% | 14 | 32.6% |
| Patient/patient's insurance | 10 | 28.6% | 28 | 65.1% |
| Other | 1 | 2.9% | 14 | 32.6% |
| None reported | 12 | 34.3% | 0 | 0.0% |
| Local Health Department Funding (DC, Houston) | 24 | 68.6% | 4 | 18.2% |
| Texas DSHS | N/A | N/A | 5 | 22.7% |
| CDC Funding | 5 | 14.3% | 0 | 0.0% |
| Other Federal Funding (Ryan White, Other HRSA, SAMHSA, VA) | 6 | 17.1% | 1 | 4.5% |
| Private Grant Funding | 5 | 14.3% | 6 | 27.3% |
| Research (HPTN 065/TLC Plus HIV testing or other research study) | 5 | 14.3% | 0 | 0.0% |
| Medicaid | 9 | 25.7% | 11 | 50.0% |
| Medicare | 4 | 11.4% | 3 | 13.6% |
| Patient/patient's insurance | 7 | 20.0% | 12 | 54.5% |
| Other | 1 | 2.9% | 9 | 40.9% |
| None reported | 6 | 17.1% | 0 | 0.0% |
| Median (Range) | Total No. of Tests | Median (Range) | Total No. of Tests | |
| Rapid tests (DC n = 30; Houston n = 21) | 593 (5–10,673) | 78,765 | 1,478 (4–9,945) | 40,910 |
| Venipuncture tests (DC n = 3; Houston n = 34) | 1200 (36–3,322) | 4,558 | 585 (1.5–46,590) | 169,635 |
| Total tests (DC n = 22; Houston n = 41) | 568 (20–10,673) | 41,085 | 1,045 (1.5–47,209) | 210,565 |
Community service organizations include homeless services, substance abuse recovery centers, life skills programs, housing assistance programs, faith-based organizations, and other community-oriented service organizations.
Five organizations participating in the Houston survey only use opt-out consent during pregnancy as required by Texas law but do not use opt-out consent otherwise. Therefore, the bulk of consent used for these organizations would be opt-in. Results if these organizations are re-classified into the “opt-in approach only” category: 26 (60.5%) organizations use opt-in approach only, 3 (7.0%) organizations use opt-out approach only, and 14 (32.6%) organizations use a combination of opt-in and opt-out consent.
Respondents could check more than one response.
Other responses in DC include general revenue (n = 1). Other responses in Houston include donations (n = 12) and general revenue (n = 2).
Other responses in DC include general revenue (n = 1). Other responses in Houston include donations (n = 8) and general revenue (n = 2). One Houston organization selected both donations and general revenue for a total of n = 9 organizations selecting “other”.
One Houston organization estimated 1–2 tests were completed. The average, 1.5 tests, was recorded as the response.
In Houston, 84 organizations were contacted for study inclusion. Of these, 55 (65.5%) organizations participated in the survey. Forty-three participating organizations, representing 114 facilities in the Houston/Harris County area, conducted HIV testing. This analysis focused only on the 43 organizations conducting HIV testing of which over half (51.2%) were outpatient clinics or university health centers (
The majority of organizations reported that a specific individual coordinated HIV testing efforts for the organization. DC organizations more frequently reported a coordinator than did Houston (82.9% and 62.8%, respectively) (
The two jurisdictions reported similar approaches to provider education about HIV testing. The top three sources of provider education for DC were continuing education (74.3%), DOH or clinic training (65.7%), and peer-to-peer best practices (57.1%). Houston organizations most frequently identified DOH or clinic training (58.1%), manuals/guidelines/literature (48.8%), and continuing education (46.5%).
A similar proportion of organizations exclusively used an opt-in consent approach in DC and Houston (51.4% and 48.8% respectively). However, a substantial number of organizations in Houston used a combination of opt-in and opt-out approaches (44.2%). Further analysis revealed that 11.6% (n = 5) of organizations in Houston using a combination approach only used opt-out consent during pregnancy. While only three (7.0%) organizations in Houston reported using only the opt-out approach, 48.6% of the surveyed organizations in DC exclusively used an opt-out approach.
Despite CDC recommendations that removed the need for pre-test counseling, most organizations in both jurisdictions reported conducting both pre- and post-test counseling with HIV testing (80.0% in DC and 69.8% in Houston). Very few organizations reported conducting only post-test counseling (11.4% in DC and 16.3% in Houston).
For all HIV testing conducted by surveyed organizations, the most frequent funding sources for DC were the DOH (74.3%), research monies (34.3%), and federal funding other than that from CDC and DOH (33.3%). In contrast, Houston organizations most frequently identified the payer as the patient or the patient's insurance (65.1%), Medicaid (53.5%), Medicare (32.6%), and other sources (32.6%) such as the organization's general revenue or funds received as donations. Funding specifically for routine, opt-out HIV testing also varied between the two cities. The most frequent funding source in DC was the DOH (68.6%), while Houston organizations primarily paid for HIV testing by billing the patient or patient's insurance (54.5%) or Medicaid (50.0%).
The median total number of HIV tests performed per organization in 2011 was higher in Houston than in DC (1045 and 568 respectively). While the total number of rapid tests performed in DC was higher than in Houston (78,765 in DC vs. 40,910 in Houston), the median number of rapid tests performed per organization was lower in DC than in Houston (593 vs. 1478). Similarly, although the median number of venipuncture tests per organization was higher in DC (1200 in DC vs. 585 in Houston), the total number of venipuncture tests was higher in Houston than in DC (169,635 in Houston vs. 4,558 in DC).
On a scale of 1 (no barrier) to 5 (major barrier), the highest reported barriers to HIV testing in both DC and Houston were patient discomfort/refusal (median score = 2) and lack of funding for testing (median score = 3). When categorizing all items dichotomously (no barrier vs. any barrier), DC organizations most frequently selected “HIV not a problem for client population” (68.6% of organizations), “lack of funding” (60.0%), and “patient discomfort or refusal for testing” (59.4%) as barriers (
In comparison to Houston organizations, DC organizations more frequently reported every area as a barrier except for staff knowledge, skill, experience (41.9% Houston vs. 32.4% DC) and provider/staff resistance (20.9% vs. 20.6%) (
We examined whether barriers to testing implementation differed by organization type (CBOs, hospitals, and clinics; data not shown). Lack of funding was reported frequently across all organizational types. Clinics were particularly challenged by cost or reimbursement for testing and patient refusal, while CBOs reported staff size as a barrier. Among hospitals, the highest ranked barrier to testing in both jurisdictions was the perception that HIV was not a problem in their client population. In DC, barriers to testing were most frequently reported among hospitals in comparison to other site types. The most frequent barriers among DC hospitals included the perception that HIV is not a problem in the client population (83.3% of organizations), staff time (75.0%), staff size (75.0%), and lack of funding (66.7%). In contrast, in Houston, hospitals reported the fewest barriers, while clinics reported the most. The most frequent barriers among Houston clinics included lack of funding (68.2%), refusal to get tested (63.6%), cost/reimbursement (50.0%), and staff knowledge/skill/experience (50.0%).
Utilizing a framework of operational research for HIV prevention,
The two jurisdictions have taken different approaches to support and fund the expansion of routine testing. DC DOH's approach has involved considerable funding for the distribution of free rapid test kits to local implementation sites including clinics, hospitals and CBOs. This is in contrast to the HDHHHS' approach which has encouraged performing venipuncture tests as part of routine visits and funding large volumes of HIV testing in hospitals and clinics with rapid results of standard testing.
Funding sources for testing also varied between the two jurisdictions which could have implications for the long term sustainability and scalability of these testing programs. Houston testing organizations reported much more third party billing and reimbursement from Medicaid, Medicare, and private insurance as a primary means to support their testing programs, compared to DC sites which relied much more heavily on direct support from DC DOH. Because of this, Houston organizations may be better positioned to more quickly adapt and benefit from expanded coverage through the Affordable Care Act (ACA), especially now that routine testing is rated a recommended service for coverage by the US Public Health Services Task Force (USPHSTF). However, this strong reliance on third party reimbursement does leave these organizations more vulnerable to reimbursement challenges and inadequate reimbursement rates that may not cover the total cost of testing provision. The monitoring of HIV testing post-ACA and USPHSTF implementation will allow for further exploration of the impact of these policies on testing programs.
Overall, few barriers to testing implementation were reported by organizations in both jurisdictions however, among those reported, lack of funding and organizational capacity to conduct testing were key barriers. Capacity barriers, such as staff size and time were moderate barriers for the surveyed organizations and may have been related to lack of funding to support these efforts at the organizational level, particularly among organizations in DC. Although multiple methods of routine HIV testing have been promoted in DC, there is more reliance on rapid testing, which can be more time and labor-intensive than venipuncture testing. Time-related barriers and competing priorities have been extensively reported in literature,
This study determined current implementation practices and perceived barriers to HIV testing, a crucial step to design interventions and guide policy to improve HIV testing in cities with a high burden of HIV, such as DC and Houston. In 2010, New York State enacted a statute requiring primary care settings to offer HIV testing to all patients between the ages of 13 to 64. Since this law was enacted, one study found that only 65% of emergency departments implemented HIV testing as required by the law.
There are a number of limitations to this study. The survey was a convenience sample that could be subject to non-response bias. The self-reported responses were limited to the organizational knowledge, perceptions and/or experiences of the person completing the survey, thus the responses may not accurately reflect the practices or opinions of all providers or staff members. The primary limitation of the parts of this study that compared DC and Houston was differences in survey administration between the two cities. Non-response was potentially influenced in Houston (65.5% vs. 85.4% in DC) by the lack of compensation for survey completion. The differing routes of survey administration (in-person in Houston and online in DC) may have also led to variances in social desirability or comprehension of the survey questions. In addition, social desirability may have biased the results, especially in reporting of testing barriers. Furthermore, our ability to statistically analyze barrier data was limited as many of the barriers received low rankings. Finally, generalizability of these findings may be limited, particularly in jurisdictions that have statutes in conflict with the 2006 CDC recommendations. Results presented herein may be useful to such jurisdictions after their statutes transition.
Despite both DC and Houston having high HIV prevalence, each jurisdiction has tailored their HIV testing strategies to meet particular funding options, resources, and statutes. The range of approaches was appropriate given the distinct context of the epidemic in each jurisdiction. Reported implementation practices, such as types of tests used and the funding mechanisms most often used to pay for HIV testing, highlight this variation. Regardless of different implementation approaches, our results suggest many organizations in at least two high burden jurisdictions have not aligned with all aspects of national testing recommendations. Research has shown that neither awareness of CDC guidelines nor jurisdictional policy change alone fully scales up testing to the levels recommended in the 2006 guidelines.
The authors would like to acknowledge Christina Shaw, Divina Oweis, Jennifer Hadayia, Dr. John Chen, Dr. Monisha Arya, and Lena Williams for their contributions to the Houston survey tool design and/or data collection, and Wenze Tang, Sarah Willis, Morgan Bennett, and Kristen Camuso for their assistance with the design and analysis of the DC data. Portions of this manuscript were presented at the First National Centers for AIDS Research/AIDS Prevention Centers Enhanced Comprehensive HIV Prevention Planning Meeting, November 19, 2012, Washington DC.