Abstract
BACKGROUND: Anal high-grade squamous intraepithelial lesion (aHSIL) is the immediate precursor of anal cancer. Anal cytology is a recommended screening test to identify aHSIL among people with human immunodeficiency virus (HIV; PWH). Heterogeneity of risk for invasive anal cancer among PWH suggests the value of a shared decision-making framework regarding screening. METHODS: Using a longitudinal HIV cohort with a comprehensive anal cancer screening program, we estimated the adjusted probabilities of having aHSIL on the first anal cytology. We used logistic regression models with inverse probability weighting to account for differential screening in the cohort and to construct a predicted probability nomogram for aHSIL. Sensitivity analysis was performed to estimate aHSIL prevalence corrected for misclassification bias. RESULTS: Of 8139 PWH under care between 2007 and 2020, 4105 (49.8%) underwent at least 1 anal cytology test. First-time cytology aHSIL was present in 502 (12.2%) PWH. The adjusted probability of having aHSIL varied from 5% to 18% depending on patient characteristics. Prespecified factors in the aHSIL prediction model included nadir CD4 cell count, ethnicity, race, age, sex, gender identity, and HIV risk factors. The ability of the model to discriminate cytological aHSIL was modest, with an area under the curve of 0.63 (95% confidence interval, .60-.65). CONCLUSIONS: PWH are at increased risk for aHSIL and invasive anal cancer. Risk, however, varies by patient characteristics. Individual risk factor profiles predictive of aHSIL can be modeled and operationalized as nomograms to facilitate shared decision-making conversations concerning anal cancer screening.