Denver HIV Risk Calculator - Score Band and HIV Prevalence

The denver hiv risk calculator turns a short set of patient answers into a numeric score, a risk band, and an estimated prevalence of undiagnosed HIV infection, using the point values published in the 2012 Haukoos derivation and validation study.

• • Emergency department triage: An ED clinician enters age, sex, sexual practice, other risk, and optional race to flag the patient for a rapid HIV test when the score is at or above 30.
• • Outpatient targeted screening: A primary care or urgent care team uses the same inputs during a routine visit to decide whether to add HIV testing to the day's lab orders.
• • Personal risk awareness: A person curious about pre-test probability plugs in the same fields to see the score band and the matching prevalence.
• • Shortened DHRS for difficult conversations: When the sexual history is not available, the shortened DHRS skips the sexual-practices question and still gives a defensible score.

The score is a clinical decision aid, not a diagnosis. It was derived in a Denver STD clinic and validated in a Cincinnati ED, and helps clinicians decide who to test, not who is infected.

When the same visit also calls for a brief alcohol screen, the WHO's AUDIT alcohol screening is the questionnaire most often co-administered with targeted HIV testing, since alcohol and drug use are part of the same risk-factor conversation.

The calculator adds the points for the patient's age band, sex, the highest-scoring sexual practice, the highest-scoring other risk factor, and (optionally) race, and maps the total to the published Haukoos 2012 score band and prevalence of undiagnosed HIV infection.

• age_points: Seven bands: <22: 0; 22-25: +4; 26-32: +10; 33-46: +12; 47-54: +10; 55-60: +4; >60: 0.
• sex_points: Male: +21; female: 0.
• sexual_practice_points: Highest-scoring practice: sex with a male +22, receptive anal +8, vaginal -10, none 0. The shortened DHRS sets this to 0.
• other_risk_points: Highest-scoring other risk factor: injection drug use +9, past HIV testing -4, neither 0.
• race_points: Optional. Black +18, Hispanic +12, AI/AN/NH/PI/other -4, White 0, 'Prefer not to say' 0.

Switching between the full and shortened DHRS only changes whether the sexual practice line is counted. All other inputs continue to drive the score the same way.

According to Haukoos et al. 2012 (Am J Epidemiol), the Denver HIV Risk Score combines age, gender, sexual practice, other risk factors, and optional race into a total that ranges from -14 to +81, with prevalences of undiagnosed HIV infection of 0.31%, 0.41%, 0.99%, 1.59%, and 3.59% across the <20, 20-29, 30-39, 40-49, and >=50 score bands.

According to MDCalc Denver HIV Risk Score, a total of 30 or more indicates an increased risk for undiagnosed HIV infection, at which point clinicians are advised to offer HIV testing and, for negative results, prevention counseling with consideration of more frequent rescreening.

The same pretest-probability pattern shows up elsewhere in the ED, and the 4Ts score for heparin-induced thrombocytopenia sums weighted clues to place a patient in a low, intermediate, or high probability band before confirmatory testing.

The four ideas that drive the score are the additive point model, the highest-scoring rule for sexual practice and other risk, the optional-race stance, and the score-band prevalence that links the total to a published pre-test probability.

Discrimination was 0.85 (derivation) and 0.75 (validation) by area under the ROC curve, consistent with a useful but imperfect screening tool.

For a faster triage-level screen, the AUDIT-C brief alcohol screen uses just the first three WHO consumption questions and is often run before the full HIV risk conversation when a patient reports substance use.

The denver hiv risk calculator walks the same five input fields in order, runs the math in the browser, and never asks for identifying information, so it works for a personal pre-test estimate as well as a clinical triage check.

1. 1 Pick the age band: Choose one of the seven bands that match the patient's current age. The point value loads automatically.
2. 2 Pick the sex: Choose female (0 points) or male (+21 points).
3. 3 Pick the highest-scoring sexual practice: If the patient engages in more than one practice, choose the one that yields the highest score, as the original score does. The shortened DHRS option sets this row to 0 points when checked.
4. 4 Pick the highest-scoring other risk factor: Choose injection drug use, past HIV testing, or neither. Past HIV testing subtracts 4 points because a recent negative test lowers pre-test probability.
5. 5 Optionally pick the race: Choose a race or 'Prefer not to say' to skip the race line. The calculator treats 'Prefer not to say' the same as White (0 race points).
6. 6 Read the score, band, and prevalence: Read the Denver HIV Risk Score as an integer, the risk band as a verbal label, and the estimated HIV prevalence as a percentage tied to that band.

A clinician evaluating a 35-year-old Black male who has sex with males and has no injection drug use picks age band 33-46, sex male, sexual practice 'sex with a male', other risk neither, and race Black. The result panel reads 73 points, Very High, and an estimated 3.59% HIV prevalence, well above the 30-point threshold MDCalc flags for targeted testing.

When the conversation also includes substance use that may have driven the risk-factor input, Addiction Calculator turns daily use, age started, and country into a side-by-side life-lost figure on the same workflow.

The tool lays the inputs, the Haukoos 2012 point values, the score band, and the prevalence estimate out on a single page, so the math behind the result can be checked against the source.

• • Published point values: All inputs use the 2012 Haukoos coefficients, so the score is traceable to a peer-reviewed source.
• • Score band and prevalence in one place: The result panel shows the numeric score, the verbal band, and the matching prevalence of undiagnosed HIV infection side by side.
• • Full and shortened DHRS modes: A toggle lets the user run the original full score or the 2013 shortened score, so the tool works when the sexual history is not available.
• • Optional race input: The race field is optional, and a 'Prefer not to say' option returns the white-reference 0 race points, matching MDCalc's stance.
• • Awareness framing: The page flags that the result is a published average and not a clinical diagnosis, and recommends the calculator as a starting point for a candid conversation with a clinician or testing site.

A value copied from a clinical record, an intake form, an awareness leaflet, or a classroom case study can be re-run under the same fields to confirm the math.

When the patient also asks about drinking, Alcohol Units Calculator keeps the units and weekly-benchmark checks on the same risk-awareness page.

The score is one sum of points, but the inputs that drive it, the band-label rules, and the published scope of the original study all shape what the result can and cannot say.

• • The score is a screening aid, not a diagnostic test. A low score does not rule out HIV infection, and a high score does not confirm it. The CDC recommends that all patients aged 13-64 get tested at least once regardless of the score.
• • The published prevalences are population averages from a Denver STD clinic and a Cincinnati ED. They are most useful for triage, not for personal probability elsewhere, and the score is less accurate for exposures the calculator does not capture (a recent needlestick, a known exposure to a partner with HIV).

The most common reading mistake is to treat the band as a personal diagnosis. The band is a published prevalence within a derivation sample; the patient's individual probability depends on the same risk factors plus local prevalence and any recent test.

According to CDC HIV Nexus Clinical Testing Guidance for HIV, the CDC recommends that all patients aged 13-64 get tested for HIV at least once as part of routine health care, with more frequent screening for ongoing risk factors such as injection drug use or sex with people who have HIV.

HIV care also includes a routine lipid workup, and the LDL cholesterol calculator uses the same published coefficients to translate routine lab values into a comparable risk number for the same patient.