Frequently Asked Questions

New HIV diagnoses were uncommon in absolute numbers but substantially elevated relative to the general US rate. Among 333,867 commercially insured people with bipolar disorder, the study identified 435 new HIV diagnoses between 2010 and 2022, which the authors described as an estimated rate of 130.3 per 100,000 over the study period. HIV incidence per 100 person-years increased from 0.54% in 2010 to 1.08% in 2022.

PrEP use was very low. Between 2010 and 2022, 1,337 people with bipolar disorder, or 0.40% of the cohort, were prescribed PrEP of any duration, and 909 people, or 0.27%, were prescribed PrEP for at least 3 months. The annual incidence of any PrEP prescription increased from 0.27% per 100 person-years in 2010 to 8.94% in 2022, and PrEP of at least 3 months increased from 0.00% to 6.34% over the same period.

In adjusted analyses, HIV diagnosis in people with bipolar disorder was associated with several identifiable clinical and demographic factors. Male sex had the strongest association (aOR=5.30, 95% CI 4.22-6.65, P<.001), and higher odds were also seen with a high-risk sexual behavior claim (aOR=2.66, 95% CI 1.85-3.83, P<.001), comorbid stimulant use disorder (aOR=2.40, 95% CI 1.71-3.39, P<.001), 1 psychiatric hospitalization (aOR=2.23, 95% CI 1.46-3.39, P<.001), and psychiatry-only outpatient follow-up versus primary care only (aOR=1.58, 95% CI 1.11-2.27, P=.01).

Sexually transmitted infection encounters were also associated with higher odds of HIV diagnosis compared with no STI encounters: 1 encounter (aOR=2.81), 2 encounters (aOR=4.30), 3 encounters (aOR=3.46), and 4 or more encounters (aOR=4.01), all with P<.001.

PrEP receipt was more likely in people with bipolar disorder who had markers of sexual risk and in those connected to both psychiatry and primary care. Male sex was strongly associated with any PrEP prescription (aOR=11.6, 95% CI 10.0-13.5, P<.001), as were a high-risk sexual behavior claim (aOR=16.9, 95% CI 14.7-19.4, P<.001), comorbid stimulant use disorder (aOR=1.49, 95% CI 1.19-1.87, P<.001), and any STI encounter versus none (all P<.001).

People with both psychiatrist and primary care encounters had higher odds of any PrEP prescription than those with primary care encounters only (aOR=1.38, 95% CI 1.21-1.57, P<.001). Lower odds of PrEP prescription were seen with comorbid opioid use disorder (aOR=0.67, 95% CI 0.52-0.87, P<.001), comorbid cannabis use disorder (aOR=0.69, 95% CI 0.57-0.83, P<.001), residence in the North Central or Western US versus the Northeast, and outpatient care only from providers other than psychiatrists or primary care clinicians (aOR=0.68, 95% CI 0.52-0.88, P<.001).

No. PrEP use remained low even among people with repeated STI encounters. Among people with bipolar disorder who had 4 or more STI encounters, only 3.53% (n=246) were prescribed PrEP of any duration and 2.73% (n=190) were prescribed PrEP for at least 3 months. The authors highlighted this as a missed prevention opportunity because STI diagnosis is a CDC indication for PrEP prescription.

Compared with primary care-only follow-up, psychiatry-only outpatient care was associated with higher odds of HIV diagnosis in people with bipolar disorder. People with psychiatry-only encounters had greater odds of HIV diagnosis than those with primary care-only encounters (aOR=1.58, 95% CI 1.11-2.27, P=.01).

For PrEP, the study found that people with both psychiatry and primary care encounters had higher odds of prescription than those with primary care-only encounters (aOR=1.38, 95% CI 1.21-1.57, P<.001). The discussion also states that people with bipolar disorder who only had outpatient encounters with psychiatrists were less likely to be prescribed PrEP relative to those with only outpatient primary care encounters.

This was a retrospective cohort study using the Merative MarketScan Commercial Claims Database from 2010 to 2022. The cohort included 333,867 commercially insured people aged 14 years or older with bipolar disorder, and the authors used multivariable logistic regression adjusted for clinical and demographic covariates, year of cohort entry, and follow-up duration to examine associations with HIV diagnosis and PrEP use.

Because the study used administrative claims data, it can identify associations and population-level trends but cannot directly determine individual HIV risk, PrEP eligibility, or causation. Diagnoses and exposures were based on claims definitions rather than structured clinical interviews or direct patient assessment.

The main limitations are restricted generalizability and the limits of claims data. The database included only people with commercial insurance, so the findings may not generalize to all people with bipolar disorder, especially those with more severe illness that affects employment and access to employer-based insurance. Bipolar disorder and other conditions were identified from ICD-coded claims rather than structured DSM-based assessments.

Claims data also captured only PrEP prescriptions obtained through clinical care, not PrEP obtained through self-pay or charitable programs, and they did not allow individual-level assessment of HIV risk, PrEP eligibility, or gaps between eligibility and uptake. The database did not include race or ethnicity, high-risk sexual behavior codes may be underused because of stigma, STI diagnosis codes may not perfectly reflect confirmed infection, and the study did not include claims evidence of intravenous drug use because prior work has found such indicators can be inaccurate.