Predicting short-term interruptions of antiretroviral therapy from summary adherence data: Development and test of a probability model

Abstract

Antiretroviral therapy (ART) for HIV is vulnerable to unplanned treatment interruptions–consecutively missed doses over a series of days–which can result in virologic rebound. Yet clinicians lack a simple, valid method for estimating the risk of interruptions. If the likelihood of ART interruption could be derived from a convenient-to-gather summary measure of medication adherence, it might be a valuable tool for both clinical decision-making and research. We constructed an a priori probability model of ART interruption based on average adherence and tested its predictions using data collected on 185 HIV-infected, treatment-naïve individuals over the first 90 days of ART in a prospective cohort study in Mbarara, Uganda. The outcome of interest was the presence or absence of a treatment gap, defined as >72 hours without a dose. Using the pre-determined value of 0.50 probability as the cut point for predicting an interruption, the classification accuracy of the model was 73% (95% CI = 66%– 79%), the specificity was 87% (95% CI = 79%– 93%), and the sensitivity was 59% (95% CI = 48%– 69%). Overall model performance was satisfactory, with an area under the receiver operator characteristic curve (AUROC) of 0.85 (95% CI = 0.80–0.91) and Brier score of 0.20. The study serves as proof-of-concept that the probability model can accurately differentiate patients on the continuum of risk for short-term ART interruptions using a summary measure of adherence. The model may also aid in the design of targeted interventions.

Measuring patient engagement with HIV care in sub‐Saharan Africa: a scoping study

“…Flag patients early for intervention [ 57 , 120 , 121 , 122 , 123 , 124 , 125 , 126 , 127 , 128 , 129 ]…”

“…While VL is considered the gold standard for monitoring treatment response [ 50 ], the mixed evidence means no measure perfectly reflects successful engagement or its dimensions. Measures are chosen to prioritize sensitivity or specificity in predicting outcomes, resulting in a trade‐off between missing people in need of support or over‐intervening and wasting resources [ 125 ].…”

“…• Routine evaluation of engagement in patients in facility care [6,61,62,77,78] • Rationalize resources, such as expensive genotype [33,[79][80][81][82] or viral load [50,[82][83][84][85][86][87] testing and resources to trace those lost to follow-up [88] • Understanding the behaviour of people on ART [15,21,25,63,65, • Routine evaluation of engagement in patients in differentiated service delivery models [49,119] • Flag patients early for intervention [57,[120][121][122][123][124][125][126][127][128][129] • a At facility and population levels for quality improvement and strategic decision-making.…”

<p>Patterns of non-adherence to oral antiretroviral medication: frequencies of consecutively missed doses</p>

“…We found that observed frequencies of runs of ≥3 and ≥4 consecutively missed doses were significantly greater than expected from random chance. In a similar approach, Harris et al 20 compared the observed and calculated distributions of ≥3 consecutively missed doses of an antiretroviral drug regimen. ( Figure 1 in the study by Harris et al, 20 comparing the observed and theoretical distributions of ≥3 consecutively missed doses in 185 patients over a period of 90 days, is the formal equivalent of Figure 2 , middle panel in the present work.)…”

“…In a similar approach, Harris et al 20 compared the observed and calculated distributions of ≥3 consecutively missed doses of an antiretroviral drug regimen. ( Figure 1 in the study by Harris et al, 20 comparing the observed and theoretical distributions of ≥3 consecutively missed doses in 185 patients over a period of 90 days, is the formal equivalent of Figure 2 , middle panel in the present work.) In the study by Harris et al, the observed frequencies of ≥3 consecutively missed doses were statistically significantly higher than the calculated random probabilities at levels of average adherence of 0.85–0.97, corresponding to between 1 and 4–5 missed doses in a 30-day period, whereas there was no statistically significant difference between the observed and calculated frequencies at levels of average adherence of 0.40 to <0.85, corresponding to 4–5 to 18 missed doses in a 30-day period.…”