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

Abstract

Background

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.

Methods

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.

Results

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.

Conclusions

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.

Keywords:

• Methadone maintenance treatment
• dropout
• recurrent event analysis
• China

Acknowledgements

The authors would like to thank the staff and participants of all MMT clinics for their cooperation.

Author contributions

Conceptualization, CX, CW and LL; data curation, CX; formal analysis, CX; funding acquisition, LL; methodology, CX, CW and ZC; project administration, LL; resources, LL; supervision, LL; writing—original draft, CX, CF and XT; writing—review and editing, CX, CF, XT and LL.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

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