A state transition framework for patient-level modeling of engagement and retention in HIV care using longitudinal cohort data

Abstract

The human immunodeficiency virus (HIV) care cascade is a conceptual model used to outline the benchmarks that reflects effectiveness of HIV care in the whole HIV care continuum. The models can be used to identify barriers contributing to poor outcomes along each benchmark in the cascade such as disengagement from care or death. Recently, the HIV care cascade has been widely applied to monitor progress towards HIV prevention and care goals in an attempt to develop strategies to improve health outcomes along the care continuum. Yet, there are challenges in quantifying successes and gaps in HIV care using the cascade models that are partly due to the lack of analytic approaches. The availability of large cohort data presents an opportunity to develop a coherent statistical framework for analysis of the HIV care cascade. Motivated by data from the Academic Model Providing Access to Healthcare, which has provided HIV care to nearly 200,000 individuals in Western Kenya since 2001, we developed a state transition framework that can characterize patient-level movements through the multiple stages of the HIV care cascade. We describe how to transform large observational data into an analyzable format. We then illustrate the state transition framework via multistate modeling to quantify dynamics in retention aspects of care. The proposed modeling approach identifies the transition probabilities of moving through each stage in the care cascade. In addition, this approach allows regression-based estimation to characterize effects of (time-varying) predictors of within and between state transitions such as retention, disengagement, re-entry into care, transfer-out, and mortality. Copyright © 2017 John Wiley & Sons, Ltd.

Keywords: HIV care cascade; HIV care continuum; healthcare system evaluation; multistate models; state transition representation.

Figure 1

The HIV care cascade using AMRS reflecting retention aspect of the cascade in care of AMPATH. At each state, patients are in the following states: Engaged: engaged in care, Disen. 1: disengaged from care for one interval (i.e., disengaged for a short-term), Disen. 2: disengaged from care for two consecutive intervals (i.e., disengaged for a moderate-term), Disen. 3⁺: disengaged from care for more than two consecutive intervals (i.e., disengaged for a long-term), Xfer: transferred-out, and Death: deceased

Figure 2

Example of state ascertainment using data from four individuals in the AMPATH dataset. Grey vertical lines represent patient visits (solid if on treatment, dashed if not). The solid black vertical line is database closure date. CD4 counts are plotted as a function of time (black dot). Color-coded state membership depicted along the bottom of the graph.

Figure 3

Number of subjects in each state over time. At baseline, 92,215 unique individual records are available from AMPATH data. Six states are considered: engaged in care, disengaged from care for one interval ( Disengaged 1), 3 = disengaged for two consecutive intervals ( Disengaged 2), 4 = disengaged for more than two consecutive intervals ( Disengaged 3⁺), and 5 = transferred-out, 6=deceased. By definition, all individuals are engaged in care at baseline (day 0).

Figure 4

Temporal trends in (unadjusted) state transition probabilities (STP) calculated at every 200 days. STP are presented in the logit scale to make temporal trends and fluctuations more apparent. All patients started from engagement in care at day 0, and thus no transitions from state 2 to other states were made at day 0. In each title, S=s (s=1,2,3,4) represents prior state membership. Transition to transfer-out was not allowed from state 2,3, and 4.

Figure 5

Calendar year effects (i.e., period effects) on transition from engaged to disengaged (left panel) and transition from engaged to death (right panel). Splines were used to estimate the smoothed effect of calendar years.