Work

Implementation of a Collaborative HIV Testing Model Between an Emergency Department and Infectious Disease Clinic

Year: 2014

Type: article

Abstract: To the Editors: Emergency departments (EDs) have been identified as a valuable site for HIV testing because they serve a broad spectrum of society and are often the only source of health care for vulnerable populations.1–4 More than 30 reports demonstrate the feasibility of HIV testing in EDs.5–7 While the Centers for Disease Control and Prevention has recommended routine, nontargeted HIV testing among populations where HIV prevalence reaches 0.1% since 2006,1 studies demonstrate that testing all ED patients is difficult.8,9 Obstacles include but are not limited to cost, lack of privacy and space for counseling, and the relevance of HIV testing in emergency settings.10,11 Recognizing these barriers, we developed and implemented an HIV testing program in the University of North Carolina (UNC) Hospital's adult ED. We describe an ED-based HIV testing program that incorporates targeted HIV screening with a conventional assay, integrated HIV consent, and collaboration with an in-hospital infectious diseases (ID) clinic that assumes responsibility for post-test counseling and linkage to care. The program was implemented at a suburban academic ED that is a level 1 trauma center. The annual census is 68,000 adult visits. Program goals were to: (1) increase HIV testing and add targeted screening criteria, (2) notify and post-test counsel all newly positive individuals, and (3) provide successful linkage to care for positives. To create a program that could be sustained without external funding, the program was designed to use existing staff and infrastructure. ID clinicians spearheaded the design of this program, but engaged the ED early in its development. A key program component was the ID clinic's commitment to assume full responsibility for notification and post-test counseling of those testing positive. The official program start date was June 1, 2008. In November 2007, the North Carolina (NC) Administrative Code was modified to eliminate separate written consent for HIV testing and pretest counseling. The UNC Health Care System subsequently modified the general consent for care to include HIV testing in January 2008. Under the revised HIV testing procedure, ED providers are required only to verbally inform patients of their desire to order an HIV test and to document assent in the patient's medical record. Blood samples are sent to UNC laboratories for testing. Testing uses a third-generation HIV-1/2 enzyme-linked immunosorbent assay with Western blot confirmation. All HIV antibody–negative specimens are automatically pooled for HIV RNA testing to detect acute HIV infection (AHI). In March 2011, UNC switched to a fourth-generation Ag/Ab combo assay with positive results tested by a multispot assay. Multispot-negative samples are tested with a real-time viral load to distinguish between AHI and false-positive results. ED providers are asked to test patients when diagnostically indicated (symptoms of AIDS/opportunistic infection or AHI) and based on HIV-related risk behaviors (sexually transmitted infection, substance abuse, or sexual risk behavior). Providers are not asked to conduct a risk assessment of patients but are encouraged to order HIV tests at their discretion. Patients who disclose a positive HIV status are not tested and are referred to the UNC ID clinic if not in care elsewhere. Patients discharged from the ED are given a referral card in English or Spanish with information about obtaining their test result at the ID clinic. The card states that patients may walk-in to the ID clinic to receive their test result on Friday mornings 1 week after their ED visit or schedule a free appointment. All HIV test results are sent electronically through a secure server to the study coordinator twice weekly. The coordinator reviews all HIV results and informs the ID clinic director of any positive results. Patients with confirmed positive results are contacted by ID clinic staff and encouraged to come to the ID clinic. HIV-positive patients receive free post-test counseling and are offered a standard-of-care baseline evaluation. If newly diagnosed patients do not present to the clinic, ID program staff notify Disease Intervention Specialists of the need for notification. Seronegative individuals who come to the clinic receive free prevention counseling. The cost of HIV testing is billed to the patient. The insurance carrier covering a majority of patients, Blue Cross/Blue Shield of NC, covers HIV testing, as do Medicare and Medicaid. A cooperative agreement with the state of NC under a federal grant (PS 07-768) has provided funds to cover post-test counseling visits for individuals testing negative or positive in the ID clinic. Outcome measures were numbers of tests ordered, confirmed positives, positives successfully notified, and positives linked to care. Successful notification was defined as occurring within 14 days of the ED visit.12 Previously known HIV-positive patients were defined as in care at UNC on the date of their HIV test by review of their UNC ID clinic history. Linkage to care was defined as attending at least 1 visit at the UNC ID clinic after the ED visit. Medical records were also reviewed to see who had at least 2 ID clinic visits within 6 months of the ED visit. HIV test data were linked with demographic and ED visit data in the electronic medical record system through medical record numbers. The unit of analysis was the patient. Organ transplant recipients and patients tested for occupational exposures were excluded from analysis since they were routinely tested before the program. From June 1, 2008 to December 31, 2011, there were 235,816 visits to the UNC ED and 3001 HIV tests (1.3%) performed. Among the 3001 tests, 461 (15.4%) were performed on patients awaiting organ transplant and 10 (0.2%) for occupational exposure, leaving 2530 tests on 2436 patients for analysis. Eighty-five patients were tested multiple times during the study period, but always tested HIV negative and were only included once in analysis. Overall, 55 of 2436 (2.3%) patients were HIV positive and 23 (0.9%) were newly positive (Table 1).TABLE 1: Demographic Characteristics of Tested and HIV-Positive Patients*Most newly positive patients were male (82.6%), and almost half were black, non-Hispanic (43.5%) (Table 1). After excluding 2 patients with AHI (viral loads 5,320,000 and 10,000,000 copies/mL), the median CD4 count and viral load among new HIV cases were 87 cells per microliter (range, 10–560) and 119,600 copies per milliliter (range, 696–4,180,892), respectively. Twenty-two (95.7%) were successfully notified of their test results within 14 days of their ED visit, and 20 (87.0%) had at least 1 visit to the UNC ID clinic within 90 days; 18 of 23 (78.3%) had 2 clinic visits within 6 months. Among previously positive patients, 69% were male, and 78% were black or Hispanic. Twenty-four of 32 previous positives (75.0%) were out of care. Fifteen of 24 (62.5%) not in care were linked to care at UNC (12 within 30 days), but only 9 (37.5%) had 2 clinic visits at UNC within 6 months. The proportion of HIV-negative participants who presented to the ID clinic for results could not be assessed due to incomplete data, but was low. In response to the change in HIV consent and increased correspondence with ED staff leading up to the program, testing notably increased in January 2008 before the program start date of June 2008. This transition period from January to June 2008 was excluded from analysis. Comparing the first 12 months (June 2008 to May 2009) of the program to a 12-month period before this increase (January to December 2007), the percent increase in HIV testing was 173% (248 tests in 2007 to 678 tests in 2008–2009). Our program facilitated HIV testing for 2436 individuals in its first 3 and a half years, representing an 882% increase in testing. Overall, the program identified 55 HIV-positive individuals, for a 2.2% HIV prevalence among those tested. Over half of those identified were previous positives due to nondisclosure. However, identification of their status was important as 75% were not in care, and 63% of those subsequently had a UNC ID clinic visit through program referral. This program is noteworthy in its use of existing resources, with external funding supporting only data analysis and post-test counseling. Using existing infrastructure allows HIV testing to be available in the ED at all times compared with rapid testing programs that rely on supplemental staff. However, utilization depends on staff integrating HIV testing into routine evaluations. Judging by the low number of tests ordered, testing reminders will need to be more frequent and pervasive. Risk-based testing also relies on providers' judgment of individual risk and misses many HIV-infected individuals who do not disclose their risks13–16 or who are not asked. This collaboratory program is novel as the ID clinic assumes full responsibility for follow-up of HIV tests, eliminating the need for time, privacy, or space to provide test results in the ED or contact patients discharged from the ED. An additional advantage of this collaboratory program is the ability of designated ID physicians to ensure prompt scheduling of the initial HIV care visit for positives. Our laboratory-based program requires only 1 patient encounter to order and draw blood for the test. One disadvantage to this approach is a longer turnaround time for results compared with rapid testing programs. HIV testing is covered by most insurance plans/companies. The charge for the HIV test billed by the ED ($111.00) is not substantially higher than charges for tests commonly ordered in the ED ($73.00 for a CBC with differential and $136.00 for a rapid strep test). Limitations in our study include lack of cost analyses to assess program cost-effectiveness and lack of data on the number of patients who declined testing, precluding evaluation of patient acceptance. Unlike many ED HIV testing programs, particularly rapid testing programs with supplemental staff, our program is sustainable due to its reliance on existing personnel and resources. Although the overall level of testing has been low, we achieved an 8-fold increase in testing and high rates of notification and linkage of newly diagnosed and previous positives into HIV care. Low CD4 counts among new diagnoses indicate a need for expanded HIV testing among ED patients. Based on our experience, eliminating a specialized HIV consent is critical in facilitating HIV testing in medical settings, particularly in fast-paced EDs. In sum, our model for expanded HIV testing is feasible, sustainable, and requires minimal additional investment. ACKNOWLEDGMENTS The authors thank Sergio Rabinovich, ED Technical Support Analyst; Orange County Disease Intervention Specialists; and UNC ID clinic staff, including Lynda Bell, Amy Heine, Clara Zelasky, Jonah Pierce, Christy Bonners, Paul Behe, and Teresa Meggett.

Source: JAIDS Journal of Acquired Immune Deficiency Syndromes

DOI: 10.1097/qai.0000000000000153