We analyzed AHI signs and symptoms in 90 patients given a diagnosis of AHI during 2007–2014. As part of this confidential HIV testing program, routine, individual donation, HIV nucleic acid amplification testing (NAT) has been provided to all rapid antibody–negative participants since June 2007 (samples for NAT are obtained at the time of rapid antibody testing) (7,10,11). AHI was defined as having a negative or indeterminate HIV antibody test result in the presence of detectable HIV-1 RNA, corresponding to Fiebig stages I–II, with a mean estimated date of infection within the previous 10 days (95% CI 7–14 days) (12). Dates of infection were estimated for all recently infected patients using previously published criteria on the basis of serologic and virologic test results (13).

At each patient’s first visit after documentation of AHI diagnosis (median 4 days, interquartile range [IQR] 3–6 days after AHI testing), we obtained blood samples for CD4 and viral load testing and collected detailed information regarding occurrence, duration, and start and stop dates for 11 signs and symptoms associated with AHI (5,14). Participants were also asked to specify any other symptoms. In addition, patients who participated during 2007–2011 were asked if they had sought medical attention for any signs or symptoms. Typical AHI (i.e., >2 signs/symptoms) was defined according to criteria described by Braun et al. (14).

For statistical analysis, SPSS version 21 (SPSS, Inc., Chicago, IL, USA) was used. For analysis on signs or symptoms compatible with AHI, signs or symptoms that started >5 days before the estimated date of infection (i.e., before the 7–14 day 95% CI) were excluded. The University of California San Diego’s Human Research Protections Program approved the study protocol, consent process, and all study-related procedures.

All 90 participants were male and self-identified as men who have sex with men (MSM). Median age was 29 (range 18–67) years. Half (50%) of participants reported white race; 29% reported Hispanic ethnicity. Median number of male partners reported for the previous 12 months was 20 (IQR 14–31). A total of 72 (80%) patients had signs or symptoms associated with AHI that occurred within 2 weeks before undergoing NAT; of these 72 patients, 47 (52% of the study population) had ongoing signs or symptoms, while signs or symptoms had resolved by the time of testing for 25 (28% of the study population). Twelve (13%) reported signs or symptoms starting after testing, while 6 (7%) reported the absence of signs or symptoms (Table 1). A total of 66 patients (73% of the study population) reported headache, pharyngitis, or myalgia occurring during the 14 days before AHI testing.

Overall, 69 patients (77%) reported signs or symptoms that met criteria of compatibility with AHI (Table 2). Onset of signs or symptoms compatible with AHI occurred at a median of 5 days (IQR 0–8, range –4 to 15 days) after the estimated date of infection. Neither viral load nor CD4 count correlated with duration or actual number of signs or symptoms.

Data on whether a patient sought medical attention because of signs or symptoms were available for 42 (47%) of 90 patients; of these, 12 (29%) reported that they sought medical attention because of their signs or symptoms and 30 (71%) did not. Significantly higher viral loads were observed for those who sought medical attention compared with those who did not (median 6.1 [IQR 5.7–6.7] log copies/mL vs. 4.7 [IQR 3.4–5.5] log copies/mL; p<0.01).

Overall, 70 (78%) of the 90 patients fulfilled criteria for having typical AHI and 20 (22%) did not (of the latter, 14 had only 1 sign or symptom, and 6 were asymptomatic). Patients with typical AHI had significantly higher viral loads compared with patients without (p<0.01). A total of 61 (85%) of 72 patients with signs or symptoms before NAT testing fulfilled criteria for having typical AHI. In addition, 40 (85%) of 47 patients who had ongoing signs or symptoms at the time of AHI testing fulfilled criteria for having typical AHI at that time.

We characterized signs or symptoms relative to the date of AHI diagnosis among persons seeking HIV testing in a program offering universal AHI screening. Two findings are notable: 1) 52% of participants reported ongoing signs or symptoms at the time of AHI testing, and 2) 80% reported signs or symptoms occurring within 2 weeks before undergoing testing.

These findings may have major clinical implications for community-based settings that restrict AHI testing to persons with ongoing signs or symptoms. This practice may be relatively insensitive in settings where MSM undergo HIV screening frequently (11). Our results show that expansion of AHI screening to include those with signs or symptoms during the 2 weeks before the test may increase the yield of AHI diagnoses by more than half.

Although our results may allow for estimation of sensitivity of signs and symptoms for AHI in persons seeking HIV testing, specificity of signs and symptoms remains unknown (in this study, signs and symptoms were not assessed in those who tested negative, and no control group was available). Estimates on frequency of signs and symptoms in HIV-negative persons (i.e., specificity) ranged widely in previous studies. Although in one study a specificity of 65% was estimated for influenza illness–like symptoms (15), specificities ranging from 38% to 91% for recent symptoms were estimated in another study (5). Limitations of those studies include the fact that exact time frames for occurrence of signs or symptoms (e.g., ongoing at the time of testing or occurring within the last 14 days) have not been evaluated, which makes comparison of results difficult. A limitation in our study is that all cases of AHI occurred among MSM. Therefore, our results may not be applicable to populations other than MSM, although previous studies have reported that clinical features of AHI may not differ by sex and age of patients (4).

In summary, HIV diagnostic testing strategies that limit AHI testing to patients with ongoing signs or symptoms may fail to identify many persons with AHI. In contrast, HIV NAT provided for MSM who report signs or symptoms during the preceding 2 weeks (representing 80% of AHI diagnoses) may increase the yield of AHI diagnoses by more than half.