Summary of results in context of previous research
In this study, we examined trends in the prescribing of opioids, benzodiazepines, Z-drugs, and gabapentinoids in Irish Prisons between 2012 and 2020. We identified an overall significant reduction over time in prescribing of opioids (since 2015), benzodiazepines and Z-drugs, and an increase in gabapentinoid prescribing. The prison population is younger and has disproportionately more men compared to available evidence from the national pharmacy claims database in Ireland, which renders comparisons difficult. Nevertheless, observations of long-term prescribing trends in the community identified an increase in prescription opioids from 2010 to 2019 [40] and Z-drugs between 2005 and 2015 [41], which are not reflected in our prison population. Observations of decreasing benzodiazepines between 2005 and 2015 [41], and increasing pregabalin prescribing between 2013 and 2016 [42] in the community are consistent with trends observed in our prison population.
It is also worth noting that, compared to the community [40, 41], a limited number of benzodiazepines and opioids were consistently prescribed in prison. Chlordiazepoxide and diazepam were the benzodiazepines of choice in prisons, whereas diazepam, alprazolam and temazepam were identified as the most commonly dispensed benzodiazepines in a previous study using pharmacy dispensing records in the community [41]. The high rate of prescribing for chlordiazepoxide may be due to the high prevalence of alcohol use disorders among prisoners [16, 17]. Chlordiazepoxide is recommended for uncomplicated alcohol withdrawal as it has a low dependence-forming potential [22, 43], whereas diazepam is recommended if there is a history of concurrent benzodiazepine dependence [22]. Opioid prescriptions in prisons were almost entirely limited to tramadol, similar to findings from a study in Swiss prisons [44]. A repeated cross-sectional analysis of the national community pharmacy claims database in Ireland between 2010 and 2019 also identified tramadol as the most frequently prescribed product, however oxycodone and tapentadol prescribing were increasing over the study period [40].
The Irish prison population was largely composed of men (96%), similar to other EU countries, where women generally represent only 3 to 8% of the prison population [45]. Although women represented less than 5% of the Irish prison population, they had a high burden of OUD. While a recent meta-analysis of 24 prison studies also identified a higher pooled prevalence estimate of OUD among women (51%) compared to men (30%), it would appear that OUD among women in Irish prisons (60%) is high compared to international studies, but low for men (16%) in Irish prisons [16]. This is also in contrast to recent estimates in the Irish community where the prevalence of problematic opioid use among men (10 per 1,000) is over twice that of women (4 per 1,000) [46].
Prescribing rates of benzodiazepines and Z-drugs in prison were almost 12 and 14 times higher, respectively, in women compared to men. These estimates are comparable to recent UK results which reported an 8-fold increase in prescriptions for hypnotics and anxiolytics to women relative to male prisoners [18]. While women in Irish prisons were also more likely to be prescribed gabapentinoids compared to men, the difference was of a lesser magnitude.
A history of OUD affected men and women prescribing rates differently. In men, a history of OUD was associated with significantly higher rates of prescribing for opioids, benzodiazepines, z-drugs, and gabapentinoids. By contrast, women with a history of OUD were less likely than other women to be prescribed gabapentinoids, and no difference was observed for benzodiazepines and Z-drugs. Co-prescribing of opioids, benzodiazepines, Z-drugs or gabapentinoids with OAT drugs remained uncommon throughout the study period, with the exception of benzodiazepines in women. Co-prescribing benzodiazepines, Z-drugs or gabapentinoids with OAT is identified as a risk factor for drug related mortality [30]. A prior study in specialist addiction clinic settings in Ireland identified up to 65% of OAT clients with a history of co-prescribing of OAT and benzodiazepines between 2010 and 2015 [24]. In this regard, co-prescribing rates appear conservative in prison settings, reducing the risk of drug poisoning mortality in prisoners on OAT. It is plausible that a higher proportion of people receive short-term detoxification in prisons [47] compared to community and specialist addiction clinic services, which can explain, in part, the lower co-prescribing rates found here.
In 2020, we observed a marked decrease in prescribing of benzodiazepines and gabapentinoids, contrasting with previous years (Fig. 1a). COVID-19 public health measures introduced in the first quarter of 2020 resulted in a reduction of court activity and delays in the justice system, with a decrease in the number of committals, particularly for short sentences (< 3 months) [31]. In addition, contingency measures were introduced to mitigate the effects of COVID-19 in people who use drugs and ensure continuity of treatment for people on OAT. These included accelerated access to OAT for people not already in treatment, additional emergency accommodation in COVID-19 facilities to allow for self-isolation/ social distancing among homeless people [48]. This may have reduced the number of committals of people with a history of OUD, particularly the more complex cases, and in turn reduced the observed prescribing rates.
Clinical implications
In several respects, prescribing practices in Irish prisons appear to adhere to the UK Royal College of General Practitioners’ guidelines for safe prescribing in prisons [22]. Firstly, methadone was the OAT drug of choice, and, with the exception of benzodiazepines in women, the levels of co-prescribing of OAT with other opioids, benzodiazepines, Z-drugs or gabapentinoids remained very low. Secondly, the reduction in prescribing for Z-drugs, benzodiazepines and opioids may reflect an increased awareness of the potential for misuse of these drugs in prison settings. In addition, the sharp decline in opioid prescriptions from 2015 may be in response to emerging evidence from the US opioid crisis which identified an increase in opioid overdose deaths arising from poor prescribing practices, with synthetic opioids such as tramadol increasingly implicated in drug poisonings in the US from 2015 [49]. It also coincides with tramadol being classified as a Schedule 3 controlled substance in June 2014 in the U.K, after concerns about safety and potential risk of misuse were raised [50]. In contrast, gabapentinoid prescribing increased during the observation period despite recommendations to avoid in prison [22]. It does, however, appear to be reducing in 2020, following recent advice regarding appropriate prescribing of pregabalin issued by the Health Service Executive in Ireland in 2019 to all general practitioners [42]. Nevertheless, given the risk of dependence and the potential for diversion and medicinal misuse of pregabalin, prescribing trends should be monitored, to determine whether the downturn observed in 2020 continues, or rebounds to pre-pandemic levels. The sharp reduction observed in opioid prescribing since 2015 and increase in gabapentinoids should also prompt an examination of pain management practices. Acknowledging it is challenging for prescribers to balance the risk of misuse of strong analgesic medications in prison settings [44], prisoners with untreated chronic pain may seek illicitly sourced analgesic drugs, increasing the risk of adverse effects, including dependence and mortality.
Findings from this study add to existing evidence on prescribing practices in prisons but also provide new evidence in relation to how prescribing practices vary by gender and history of OUD. Women were more often prescribed benzodiazepines, Z-drugs and gabapentinoids than men, and men with a history of OUD were more often prescribed benzodiazepines, Z-drugs and gabapentinoids compared to other men. While this may reflect the more complex needs of women and people with a history of OUD, including increased levels of pain [51] and mental health issues [17, 52], further work is needed to understand if other factors are driving these differences. As previously noted, additional targeted social and psychological support could be beneficial, to reduce the reliance on benzodiazepines and Z-drugs prescribing, when clinically appropriate [18]. In line with existing evidence, we found a high prevalence of OUD in Irish prisons. Prison represents an opportunity to engage people in addressing health issues including OUD, in this otherwise difficult to reach population [53]. Easy access to services and improved adherence to OAT within the prison environment can be viewed as favourable conditions to initiate maintenance treatment, if OAT can be secured seamlessly in the community after release [54, 55].
Due to the limited number of prisons and prescribers involved, a change of prescribers and/or prescribing practices in a single prison can greatly affect national estimates. While this can be seen as a challenge to standardisation and continuity of care, it can also provide opportunities for rapid change and implementation of recommended prescribing practices.
Strengths and limitations
This study is not without limitations. Firstly, we used the number of people in prison on the last day of the month as the denominator to estimate monthly prescribing rates, with the number of prisoners prescribed at any time during the month as the numerator. This results in an overestimate of monthly prescribing rates. However, the overestimation is expected to be consistent throughout the study period, therefore providing an accurate evaluation of trends. In addition, people with a history of OUD who did not receive a prescription for OAT during the study period will be misclassified as not having a history of OUD. This could affect the external validity of the study, however, as all people with OUD in prison should be offered treatment (detoxification or maintenance), misclassification is expected to remain low. Secondly, due to insufficient numbers in certain subgroups, we limited adjustment factors to gender and history of OUD. Residual confounding cannot be excluded from the trend analysis. Thirdly, prescription records did not contain information on diagnosis or indication for prescribing, or dosage or duration of treatment, therefore limiting our ability to assess appropriateness of prescribing against existing guidelines. Fourthly, we selected a minimum of 7-days overlap per month, as a meaningful indication of co-prescribing among people in receipt of OAT in prison. While aiming to exclude once-off prescriptions for acute situations, our cut-off misclassifies co-prescriptions of 1–6 days per month as none. Thus, co-prescribing rates are underestimated. Fifthly, while all drugs analysed in this study would have been taken under supervision in the prison, we do not know to what extent doses were concealed within the mouth and later removed and diverted to other prisoners, either voluntarily or under duress. Sixthly, benzodiazepine prescribing rates did not include Prazepam (N05BA11), as it was not retrieved from the prescription records. However, considering the relatively low prescribing rates in the community [41], and the predominance of diazepam and chlordiazepoxide in prison, we expect this had little impact on the overall prescribing trends for benzodiazepines in this study. Finally, our analysis did not include sedating anti-depressants (e.g. mirtazapine) or antipsychotics (e.g. quetiapine), which are identified as high risk of misuse, diversion and dependence in prison by the UK Royal College of General Practitioners Safer Prescribing in Prisons (2019) guidelines [22]. Future work is needed to examine prescribing trends for these drugs in Irish prisons.
Notwithstanding these limitations, this is the first study to examine prescribing trends for opioids, benzodiazepines, Z-drugs and gabapentinoids in Irish prisons in recent years. This is important, as up-to-date, robust evidence is necessary to inform practices and policies. In addition, while other studies typically used census data [18, 56], this work reports on electronic prescribing record data from all Irish prisons from 2012 to 2020, providing national estimates of prescribing rates and long-term trends. All prescribing in prison being electronically recorded, missing data are unlikely. Moreover, given the great imbalance in gender ratio observed in prison, it is critical to run gender sensitive analyses, as women specific results would otherwise remain invisible. Finally, history of OUD was taken into account in analyses and appears as a highly relevant factor for prescribing patterns in prison.
This post was originally published on this site be sure to check out more of their content.
