91������

91������ does not support MIPS measure ID# 475 (NQF ID# 3067): “HIV Infection Screening” because of uncertain validity. To the extent the intent of this measure is to standardize HIV screening, thereby increasing early diagnosis and reducing the stigma of testing, including some measure of “ever tested” seems like a reasonable first step. However, we note several implementation and methodological flaws that reduce the measure’s ability to lead to measurable and meaningful improvements in clinical outcomes. First, while evidence suggests the benefit of screening for HIV in all adults on clinical outcomes is high, the patient’s consent to testing is often beyond the clinician’s control. Second, poor interoperability across EHRs poses a significant burden on clinicians who report this measure. Additionally, clinicians may encounter confidentiality barriers to retrieving patient sensitive information around test results. If clinicians are unable to retrieve previous results, they may feel inclined to order additional tests. Second, the specifications should include exclusion criteria for patient refusal, patients who are diagnosed with limited life expectancy, and patients who are already infected with HIV. Finally, developers not cite any evidence to form the basis of the annual screening frequency described in the denominator specifications. Data are far better for the frequent screening of high-risk patient. One-time screening is an odd idea for an infectious disease— patients are either at risk, in which case they should be screened, or not at risk with limited benefit of screening. Additionally, one-time screening in low-risk patients has mixed data on effectiveness and is highly dependent on the assumptions about the underlying prevalence. For example, two major papers on the topic conclude that the cost-effectiveness is >$100,000 per quality-adjusted life-year per (QALY) and >$15,000 per QALY.