The low-threshold methadone maintenance treatment (MMT) programme in Hong Kong has been in place for about 40 years. Assessment of the participation pattern of methadone users may inform future programme development to achieve effective harm reduction.

Longitudinal clinical data of methadone users who first registered for MMT in the year 2008 in Hong Kong were retrieved after ethical and institutional approval. Participation pattern of this cohort was evaluated by examining users’ frequency of attendance and then the overall retention rate. A subgroup of consistent users who remained on treatment in 2012 and/or 2013 was analysed. Comparison was made between high- and low-frequency users, and among high/moderate and low consistency users, to test their correlations with socio-demographics and social connectivity.

The cohort of methadone users registering in the year 2008 was composed of 351 persons, 77 % of whom were ethnic Chinese, with a median age of 34 and the duration of heroin dependency of 6 years. The participation pattern of methadone users was highly variable, with a 6-year retention rate of 38 %. Discontinuations or 'breaks' of >28 days had occurred in 212 (60 %) methadone users. About one third (n = 117) were high-frequency users who had attended more than twice a week for at least 90 % of their treatment periods. The dosages received by high-frequency users were generally higher. Of those continuing on treatment in the fifth and/or sixth year (n = 185), 30 (16 %), 29 (16 %) and 126 (68 %) gave a high, moderate and low level of consistency as defined by the lengths of breaks. High/moderate consistency users had a longer history of heroin use and a higher degree of connectivity with other users by social network analysis.

Despite the variability of frequency and consistency of attendance of drug users enrolling in the low-threshold MMT programme in Hong Kong, a consistent pattern could be seen in the proportional distribution of dosage and participation efforts. Whereas an adequate dosage was a potential predictor of optimal frequency of attendance, demographics and connectivity had varied between continued users with different levels of consistency.

Table 1

General characteristics of methadone treatment users first admitted to the programme in 2008 (n = 351)

	No.	%
	Median	IQR
Socio-demographics at first admission
Male	266	76 %
HKID holdera	267	76 %
Ethnic Chinese	269	77 %
Age (years)	34	28–40
History of narcotic use (years)	8	2–14
History of injection (years)	0	0–1
Methadone clinic attendance pattern
Proportion attending >2/week during treatment period
<50 %	65	19 %
50–89 %	164	47 %
≥90 %	122	35 %
Timing of attendance
Morning	60	17 %
Afternoon	100	28 %
Evening	253	72 %
No. of clinics visited	2	1–3
Longest break (months) (n = 321)	5	1–15
Methadone treatment received
Completion of detoxification (n = 20)	6	30 %
Mode dose (mg)	40	30–50
Dosage per day (mg)	6	1–21
Connectivity with clinic and among users
Connected in social networkb	168	48 %
Degreec	0	0–2
Attendance at the nearest clinic (n = 134)	94	70 %
Same residing and clinic district (n = 322)	191	59 %
Distance (m) from home to frequently visited clinic (n = 134)	609	264–1470
Shortest distance (m) from home to clinic (n = 134)	459	237–767

IQR interquartile range

aHKID holder refers to a person who is permitted to stay in Hong Kong for more than 180 days with the exclusion of visitors

bUsers who have at least one connection with any other users with F-score not less than 0.1

cThe number of connections with other users with F-score not less than 0.1

Below: Longitudinal pattern of attendance status of methadone users admitted in 2008 (n = 351). Discontinued users had discontinued methadone treatment for >28 days (‘break’), including those who did not return (defaulted). Readmitted users refer to those who returned to the MMT programme after a 28-day break. Active users of the year had attended at least once in the respective year

Below: Longitudinal pattern of weekly attendance frequency of methadone users in the cohort. The attendances were classified by the number of days per week, from once daily to once weekly

Below: Longitudinal pattern of monthly prescribed doses of methadone users in the cohort. Prescribed dosages were categorised into eight groups: ≥70, 60–69, 50–59, 40–49, 30–39, 20–29, 10–19, and <10 mg