Detecting Unknown Cases of HIV Using CD4 Data

Tuesday, June 16, 2015: 10:30 AM
102, Hynes Convention Center
Susan L Mottice , Utah Department of Health, Salt Lake City, UT
Rachelle Boulton , Utah Department of Health, Salt Lake City, UT
Matthew S. Mietchen , Utah Department of Health, Salt Lake City, UT
Jordan Pyper , Utah Department of Health, Salt Lake City, UT
Cristie Chesler , Utah Department of Health, Salt Lake City, UT

BACKGROUND:  States have been asked to change their reportable disease rule to mandate reporting of all CD4 lab results.  The goal of this initiative is to use CD4 reports to assist with identification of previously unknown or unreported cases of HIV.  While CD4 data is exquisitely sensitive in detecting cases, the positive predictive value is very low, thus compromising the ability to easily use this important data source.  Utah has made negative HIV lab results (supplemental tests such as the MultiSpot or Western Blot, genotype data, and viral loads) reportable as per our new Communicable Disease rule.  An algorithm consisting of CD4 results coupled with negative viral load data was examined to determine its utility in identifying new cases of HIV.

METHODS:  All incoming ELR results that do not match an existing case are stored in a data warehouse.  Data from the past 18 months were queried to determine the number of individuals with an ELR-reported CD4 value and a negative HIV viral load result.  Clinicians/clinics were telephoned to determine the serostatus of patients.

RESULTS: The data warehouse contained CD4 results on 714 unique individuals, negative viral load results on 203 unique individuals, and combined CD4/negative viral load results on 42 individuals.  Initially, 10 patients who had negative viral loads (but not accompanying CD4 results) were examined.  Investigators determined that 0 out of 10 of these patients were seropositive.  The 42 individuals that had both CD4 and negative viral load testing performed were also investigated.  Nine of the patients (21%) were seronegative.  Many of the nine patients were in clinical studies that mandated these tests. Two patients (5%) were lost to followup. And 31 patients (74%) were reported as seropositive.  Seven of these patients were identified as seropositive subsequent to the receipt of the electronic labs, but 21 (50%) seropositive patients were unknown to the surveillance program.  Three patients were not matched due to computer error. 

CONCLUSIONS: The cases investigated so far indicate that this algorithm is identifying cases that are imported to Utah, have controlled HIV infection, and are in treatment and care. Fifty percent of the individuals identified via this reporting algorithm were seropositive for HIV and previously unreported to the surveillance program.  This study clearly shows that efforts should be made to combine CD4 data with other parameters, including negative viral loads, in order to identify previously unreported HIV seropositive individuals.