iart_a_2189245_sm1640.docx (35.33 kB)
Download file

Recurrent event analysis for time to dropout of newly-enrolled MMT participants in Guangdong, China: a retrospective study using the Prentice-Williams-Peterson model

Download (35.33 kB)
journal contribution
posted on 2023-03-17, 08:20 authored by Chaofan Xu, Chaonan Fan, Xijia Tang, Chijie Wang, Zouxiang Chen, Li Ling

Recurrent dropout often occurs among participants receiving methadone maintenance treatment (MMT), which negatively affects treatment effectiveness. However, few researchers focused on this phenomenon including those in China. This study examined systematically factors associated with dropout based on recurrent events analysis among Chinese MMT participants.

This retrospective study involved participants who firstly enrolled in MMT program between 2006 and 2017 of nine clinics in the Guangdong Province. The factors influencing recurrent dropout were identified using Prentice-Williams-Peterson model with total time (PWP-TT), then a comparison with Cox proportional hazards model was conducted.

Among a total of 1,319 participants, 1,922 treatment episodes were identified. There were 366 (27.7%) participants remained in MMT at one year follow-up. There was a progressive shortening of the treatment episode duration among participants with multiple treatment episodes. Protective factors included higher average age (≥50 years) before attending MMT, the higher last methadone dosage before each dropout (≥50ml) and higher average dosage (≥60ml) in each episode, while being divorced, arrested in the past 3 months, having contact with drug-using friends more than once per day, and being positive for the first morphine urine test were risk factors. Additionally, higher average dosage (≥60ml) had constant protective effects on subsequential treatment episodes.

The risk of dropout increased as participants experienced multiple treatment episodes in MMT. The influencing factors for recurrent dropout were variable across all treatment episodes. Greater efforts are needed to provide sufficient methadone dosage for the participants to decrease dropout rate.


This work was supported by the National Natural Science Foundation of China [Grant number: 82073664].