International studies have shown that problem alcohol or drug users have increased mortality rates in comparison to the general population. Mortality rates are further elevated among individuals with serious opiate addictions, and especially among injecting drug users (IDUs). Methadone maintenance treatment (MMT) is shown to reduce the mortality rate for this cohort of problem drug users, but in comparison to the general population MMT patients still have a higher risk of mortality.

The aim of this study1 was to explore what risk factors might contribute to increased mortality among MMT patients. The authors stated that they were specifically interested in the hypothesis that methadone dosage and problem use of non-opiate drugs might be associated with an increased risk of mortality.  They were also interested in whether there were different risk factors among cases who died from poisoning and cases who died from other causes.

The authors used a matched case-control study. They examined treatment exit records from the National Drug Treatment Centre (NDTC) in Dublin for seven years from February 2005 to February 2012. They identified all patients whose treatment outcome was recorded as death. Cause of death for these cases was determined using information from the National Drug-Related Deaths Index (NDRDI). Controls for the study were MMT patients who had not died, and matching was done by age, gender and treating team.

The authors conducted two separate statistical analyses. Uni-variable associations between risk of death and age, gender, methadone dose, non-attendance at treatment for at least one week prior to death, drug use in the month preceding death, history of imprisonment, HIV status, and medical complications were examined. For the controls, an index date, equal to the date of death, was created, and this was used to analyse the same set of death-related variables. A second analysis was conducted using binary logistic regression with alive/dead as the outcome variable.

Results

Over the seven years of the study the average number of MMT patients at the NDTC was 500, and 80 deaths were recorded over the same period. No statistical difference in methadone dosage was found between cases and controls and neither was there a significant difference in recent usage of heroin, benzodiazepines or cocaine. Cases were more likely to have a history of imprisonment (p<0.001); not attended the NDTC for at least one week prior to death (p<0.001); HIV (p=0.01); non-HIV/HCV medical problems (p<0.001); and more frequent medical/psychiatric reviews (p=0.03).

The logistic regression model reinforced these results with the significant variables found to be history of imprisonment, non-attendance at the NDTC one week prior to death, and a non HIV/HCV medical condition. HIV status was not found to be significant associated with risk of death using this model.

Poisoning deaths

The authors attempted to delineate between poisoning deaths and other types of deaths, in an attempt to determine if there were different risk factors for deaths in these cases. Poisoning deaths included all drug overdoses (intentional or accidental). Other causes of death included trauma (murder, road traffic accident, etc) and medical causes (infection, organ failure, etc). Cause of death was possible to determine in 77 out of 80 cases; no suicides were recorded, though the authors noted that this might possibly have been under-recorded owing to some coroners recording suicides as ‘accident/misadventure’.

Records showed that 33 out of the 80 patients had died of poisoning. The only significant statistical difference between this cohort and cases who did not die of poisoning was a slightly younger average age (34.2 versus 41.3, p<0.001). Comparing the poisoning cases with their controls, similar significant associations were found (HIV positive, not attending the NDTC for at least one week prior to death, history of imprisonment, non-HIV/HCV medical problems). The poisoning deaths cohort was also found to have been more likely to have a history of homelessness (p=0.03).

Discussion

The authors noted that other studies had shown that increased risk of mortality had been associated with low or very high doses of methadone. No such association was witnessed in this study. Similarly, previous studies had shown an increased risk of mortality among MMT patients with problem cocaine use, but this study showed no significant associations between ongoing problem drug use and risk of mortality.

Three significant associations were identified, however, in relation to imprisonment, medical issues and attendance at the NDTC for MMT.

• Other studies have pointed to the risk of lower tolerance on release from prison being a cause of overdose. The authors of this paper chose to focus on history of imprisonment, and showed that the cases were more likely than the controls to have a history of imprisonment. They theorised that being imprisoned could be linked to prisoners being one of the most disadvantaged groups in society, and noted that previous studies had shown an elevated risk of mortality among ex-prisoners.
• Recent studies have shown that HIV patients who were injecting drug users (IDUs) have a mortality rate 20 times higher than non-IDU patients. This study showed that there were significantly more HIV patients among the case group than the control group. The study also showed a link between non-HIV/HCV medical illness and mortality, as one would expect. Interestingly, among the cases who died of poisoning, medical issues were significantly more common in comparison to controls. The authors suggested that this might be linked to ‘increased susceptibility to the respiratory depressant effect of opioids among the medically unwell’. The authors suggested that prevention, and early detection and treatment, may help reduce mortality.
• Other studies have shown an increased risk of mortality among patients who have disengaged from MMT, and especially in the first weeks out of treatment. This study observed that a significant minority of cases had been disengaged from treatment at time of death. The authors recommended educating MMT-patients on the increased risk of overdose if they ceased treatment, and that clinics should follow up on patients, even if they have only been absent for a short period of time. A pro-active approach, the authors suggested, could help reduce the increased risk of mortality faced by those who disengage.

1.Truszkowska E, McCarron P, Konovalov P, Galander T, Lyons S, Keenan E and Smyth BP (2015) Case-control study of risks and causes of death amongst opioid dependent patients on methadone maintenance treatment. The Canadian Journal of Addiction 6 (3): 17–25. https://www.drugsandalcohol.ie/25004/