Do Only 21% of HIV-Positive Medicaid Enrollees Link to Treatment? Challenges in Interpreting Medicaid Claims Data

To the Editors In a recent article, Johnston et al.1 found that only 21% of Medicaid enrollees with a new HIV diagnosis code were linked to appropriate care within a year after the HIV test. This finding contrasts with Centers for Disease Control and Prevention estimates based on surveillance data, which indicate that 75% of people with HIV-positive test results link to care within a year. This analysis was based on Medicaid claims, which lack information on laboratory test results; consequently, Johnston et al. inferred a positive test result if the claims contained an HIV diagnosis code on the same day as the test or at a later date. This method likely resulted in including many non–HIV-positive adults in the sample. Comparable with Johnston et al., we selected from Medicaid claims files all individuals with an HIV diagnosis code, initially identifying 14,402 individuals in California with Medicaid, but not also Medicare coverage. Of those, 14% had HIV diagnoses coded only on the same day they were screened for HIV, and there was no evidence of a subsequent diagnosis from a confirmatory test. Of this same-day group, 82% were female and 49% were entitled not to full benefits, but only to services for pregnancy, family planning, breast cancer treatment, or other limited services. One quarter of this group received services for pregnancy. We concluded that many of these enrollees were receiving HIV screening tests in prenatal care or at delivery and did not have HIV disease. Only 1% of those whose diagnoses were recorded only on screening days, with no confirmatory test; had claims for viral loads or CD4 tests; or had claims for antiretroviral medications. Conversely, of those with diagnoses recorded on days other than screening days, 63% had viral load or CD4 tests and 70% had claims for antiretroviral medication. Many of the “HIV diagnoses” in 2007 California Medicaid claims seemed to be “rule-out HIV” diagnoses. Thus, the strategy of Johnston et al. regarding using a same-day HIV diagnosis to identify a new HIV case may have inadvertently included many HIV testers who did not turn out to be HIV infected. Including many Medicaid recipients without confirmed HIV diagnoses may be responsible for the unexpected findings of Johnston et al. that 70% of the HIV-positive Medicaid enrollees were female and that 21% received their HIV test in an inpatient setting.1 If women represent 70% of Medicaid enrollees with HIV, it implies a greater infection rate among women than men because women account for only 59% of the adult Medicaid population. That the infection rate is 3.7 times higher among men than among women casts doubt on the selection criteria used by Johnston et al.2 Including individuals who do not actually have HIV disease when calculating the percentage of HIV-positive individuals who receive appropriate medical follow-up understates the true level of linkage to care and also biases other measures of interest, such as Medicaid expenditures for HIV. The difficulties in identifying new HIV cases in claims data raise serious questions about the conclusion that Medicaid enrollees with HIV are not being linked to effective treatment in a timely way. Arleen A. Leibowitz, PhD Katherine Desmond, MS Department of Public Policy UCLA Luskin School of Public Affairs Los Angeles, CA [email protected]

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.