Approximately 21% of the 1.1 million HIV-infected persons in the United States do not know their status and therefore cannot benefit from life-saving, restoring treatment. In the VA, risk-based, routine and rapid HIV testing research projects and programs, as well as changes in policy (e.g., verbal consent replaced written consent, standard of care became routine testing, standing orders for nurse-based testing), have led to a substantial increase in the rates of HIV testing. Although positive HIV tests have increased at all VA facilities, it is not known to what degree expanded HIV testing has identified patients with previously unknown HIV infection and, more crucially, whether newly diagnosed patients are being promptly linked to appropriate medical care. Timely care linkage is associated with meaningful improvements in clinical outcomes but occurs in less than 65% of Newly Diagnosed HIV-Infected patients in the United States. Patients transferring to the VA account for more than 50% of new HIV-infected patients in the VA. Such patients tend to actively seek care and are more likely to engage in care than are newly diagnosed patients. Since there is no validated tool to specifically identify Newly Diagnosed HIV-Infected Patients in the VA, it is necessary to develop such a process.
Aim 1: Refine and validate an algorithmic decision tool to identify patients newly diagnosed with HIV infection in the VA.
Aim 2: Identify Newly Diagnosed Patients and assess geographical, facility-level and patient-level differences in the case finding rate and in terms of the diagnostic yield of both risk-based and routine HIV testing.
Aim 3: Identify Newly Diagnosed Patients with delayed linkage to care in the VA and the geographical, facility-level and patient-level predictors thereof.
The algorithm was based on preliminary work done at the VA Greater Los Angeles HCS. This pilot work utilized the timing of HIV diagnostic tests (e.g., CD4 counts, viral load), entry of ICD-9 codes for HIV infection and the ordering of antiretroviral therapy. This tool was refined and validated using data from the Houston, Atlanta and Washington D.C. VA facilities. These facilities were chosen based on their large numbers of newly diagnosed HIV infections, prior work done at these sites to identify Newly Diagnosed Patient. At each site the local investigators obtained lists of all positive HIV tests at their sites over a 6 year period beginning on August 17, 2006. Using the CAPRI remote reviewing system, the study team in Los Angeles re-reviewed the medical records of all patients with positive HIV tests to confirm the local determination of whether HIV infection was diagnosed within the VA. Discrepancies will be adjudicated through discussion with the local site investigators. This proofed list served the gold standard used to determine the performance of the algorithm.
Using the algorithm validated in Aim 1, we evaluated the rates of identification of Newly Diagnosed HIV-Infected Patients in 15 VA facilities in VISNs 1, 3, 16 and 22 during the periods of risk-based and routine HIV testing. Patient-level variances included age, marital status, race, ethnicity and HIV-related risk factors; while facility-level factors include annual patient load, baseline HIV testing rate and HIV prevalence.
Delayed linkage to care was defined by a >3 month lag from HIV diagnostic testing to evaluation by an HIV specialist.
The data needed to support the analyses was assessed through the VINCI interface to the VA Corporate data Warehouse.
During the 6 year period beginning on August 17, 2006 1,153 patients seen at the Los Angeles, Houston, Atlanta and Washington D.C. VA facilities were identified as having a positive HIV diagnostic test within the VHA. Of these, 57% were determined to have prior knowledge of their HIV status from testing at non-VHA facilities. Based on the sequence and results of available laboratory tests and ICD-9-CM entries, the refined algorithm identified new HIV diagnoses with a sensitivity of 83%, specificity of 86%, positive predictive value of 85% and negative predictive value of 90%. There were no meaningful demographic or clinical differences between newly diagnosed patients who were correctly or incorrectly classified by the algorithm.
We next applied this algorithm to patients who received HIV antibody tests at 18 VHA facilities and their satellite clinics in the Western, South Central, mid-Atlantic, and Northeastern regions of the United States from August 2006 to July 2012. We found that the rate of new diagnoses in the risk-based testing period was 0.46% while the rate of new diagnoses in the routine testing period was 0.14%. During the routine testing period, rates were highest (0.45%) among patients who were < 55 years of age, African American, and who receiving care in the Mid-Atlantic region. In contrast, the rate of new diagnoses among patients older than 75 years was 0.02% (95% confidence interval 0.0 - 0.05%) in the routine testing period.
Preliminary analyses regarding the timing of linkage to care indicate that 82% of patients with new diagnoses of HIV infection were seen by an HIV specialist within three months of the diagnosis.
As the largest provider of HIV care in the United States, the VA has a critical responsibility to assure that HIV-infected Veterans are consistently and promptly identified and referred to specialty care, while reducing variances in care. This validated method to identify cases of new diagnosis of HIV infection will enable analyses of the epidemiology of newly diagnosed HIV infection in large administrative datasets and thus help inform the development of programs to increase the efficiency of efforts to promote HIV testing and to analyze rates at which newly diagnosed patients are linked to care and otherwise proceed through the continuum of HIV care. Such data are needed to inform the development of efficient and effective programs to help the VHA build upon its successes reach or surpass national targets in providing high quality HIV care. In particular, better understanding of the magnitude of and contributors to variances in identifying HIV-infected patients and linking newly identified patients to care will facilitate the development of enhanced patient-focused programs that provide these essential services to vulnerable VA patients.
External Links for this Project
- Goetz MB, Hoang T, Kan VL, Rimland D, Rodriguez-Barradas M. Development and validation of an algorithm to identify patients newly diagnosed with HIV infection from electronic health records. AIDS Research and Human Retroviruses. 2014 Jul 1; 30(7):626-33. [view]
- Goetz MB, Hoang T, Kan VL, Rimland D, Rodriguez-Barradas MC, Asch SM. Rates and Predictors of Newly Diagnosed HIV Infection Among Veterans Receiving Routine Once-Per-Lifetime HIV Testing in the Veterans Health Administration. Journal of acquired immune deficiency syndromes (1999). 2015 Aug 15; 69(5):544-50. [view]