Summary
This systematic review has examined evidence to inform better practice, pathways and service design in the NHS to support people with chronic non-cancer pain to reduce or stop their use of opioids and address inequalities in access. A review of effectiveness was combined with reviews of safety, barriers and facilitators and inequalities.
Twenty-seven studies were identified that examined the effects of interventions to reduce (or cease) opioid use. Like other research teams who have evaluated the effectiveness of interventions to taper opioid use, we found the published evidence to be of low quality. In agreement with Sandhu 2018, our assessment of the evidence was limited by the variation in intervention investigated, reported outcomes, the low quality of studies, short follow-up and high dropout rates. Similarly, the studies included in our review reported results from behavioural interventions, acupuncture, sequential opioid dose reduction programmes, interdisciplinary programmes, a pain e-consult, pharmacological programmes, a quality improvement project and an OSI. Contact with patients varied from long sessions, over a short duration (e.g. 8 hours per day for 3 weeks) to unspecified input over a longer time period (12 months). Interventions were generally delivered in an outpatient setting but could be delivered face-to-face, over the internet or via telephone; they could be delivered to individuals, groups or, in the case of two programmes, to individuals and groups. We are therefore uncertain about which intervention approaches or their components effect peoples’ opioid use and pain but found some evidence that tapering plus support interventions may make some difference to opioid use while not increasing reported pain. Seven studies reported on AEs. None reported any serious AEs, and no patients were reported to have withdrawn from a study due to an AE. We found few studies that examined intervention acceptability but noted at least two studies that experienced significant issues related to patients leaving the study early and slow enrolment into the study, respectively.
By exploring barriers of and facilitators, we can more easily identify strategies that will support patients, carers and healthcare professionals to safely and effectively taper opioids. Sixteen studies were identified for inclusion, and eight barriers and eight facilitators were identified across multiple levels (individual, interpersonal, organisational and environmental levels) of the healthcare system. The use of theoretical frameworks to guide the analysis ensured that our research is comparable with previous research using the same frameworks and makes it easier to advance knowledge and identify gaps. At an individual level, the mapping of barriers and facilitators to the TDF59 provides practitioners with the opportunity to identify and modify patient/provider behaviours to increase the chance of a successful taper, whereas the Levesque et al.55 amended access to healthcare framework helps providers to identify when and where barriers and facilitators may arise during the opioid-tapering pathway. Tapering is a complex process. The chances of safely and effectively tapering opioids are likely to be increased where the care provided recognises this complexity and is patient-centred. Crucial to the success of opioid tapering is a willingness among both the patient and healthcare professional to taper, and ability of a patient to maintain a taper. Use of MI techniques130 and considering patient preference in the design/delivery of tapering support also appear important, as demonstrated in the EMPOWER study131 which considered patient- and provider-specific individual-level barriers in the design and delivery of an opioid taper programme. This may also help to ensure that tapering interventions and providers do not unintentionally reinforce perceptions of stigma around addiction. In addition, successful tapering concerns skilful consulting behaviours using techniques such as MI.
Little is known about opioid-tapering interventions and deferral access or outcomes according to equality groups’ characteristics and their impact on inequalities. We found that although most studies collected and reported baseline data relevant to PROGRESS-Plus categories, few considered the impact on effectiveness or in relation to inequalities in access to opioid tapering. Ultimately, any widening of inequalities in access to opioid tapering will likely impact on a widening of inequalities in the effectiveness of opioid tapering. Factors that may potentially widen inequalities in relation to access were highlighted at the individual, interpersonal and wider structural levels and may arise at any point across the entire spectrum of the tapering process, from the point of recognition of a need to taper, initiation of and engagement with a taper. Of note is the influence of ‘social capital’, which is clearly visible at every stage in a patient’s opioid-tapering pathway. While the review is comprehensive, it may not show the true impact of inequalities on the effectiveness of, and access to, opioid tapering. In particular, this review found a similar limitation to that reported in a previous review of inequalities in the uptake of, adherence to and effectiveness of behavioural weight management interventions in adults138 of low statistical power of the individual studies to detect differential effects of interventions. As reported in Birch et al.,138 this is likely because an investigation of inequalities was not the focus of the analysis. In addition, due to the small sample sizes of many of the studies assessing the effectiveness of opioid tapering, and in some studies a lack of reporting on some PROGRESS-Plus characteristics (e.g. ‘Social status’), inequalities in effectiveness may have been present but not detected.
What considerations should be made in service delivery and practice?
We found that the skills of healthcare professionals were an important theme in the barriers and facilitators review. This referred to both the skills of healthcare professionals in tapering opioids and in relation to building a positive relationship with the patient undergoing tapering. However, we do not know from the effectiveness review what training for healthcare professionals is effective. For patients, we found that a lack of knowledge and understanding about tapering was a barrier, as well as a perception that opioids play an important role in pain management. Four studies from the effectiveness review examined an intervention that involved a patient education component. One study provided detail on these components, which included education about dose-related risk, the neuropsychology of pain and how to manage pain. Training patients in skills to manage their pain without opioids was a component in eight studies. However, we are unclear what contribution, if any, education and training components make to any intervention effect on opioid use or pain.
Patients’ life circumstances play an important role in facilitating successful opioid tapering and social factors affecting patients was an important theme in the barriers and facilitators review. Seven studies from the effectiveness review, all of taper plus support interventions, were based on a model of CBT. CBT relies on a collaborative process between patients and healthcare professionals and so may address the need for trusting, non-judgemental and supportive patient-professional relationships. Among healthcare professionals, the barriers and facilitators review highlighted the need for a supportive health system environment for tapering. Five studies included intervention components that involved collaborative working or dedicated posts (such as pharmacists) to support patient management, for example, by reviewing patient records and developing recommendations. A common component of these interventions was a note or prompt being placed on patients’ health records alerting the provider to the recommendation for tapering.
Through the barriers and facilitators review, we found that negative expectations of tapering and an overarching view that tapering was a burden affected patient’s motivation for tapering. Twelve studies from the effectiveness review examined interventions that included components which targeted negative emotions. These commonly involved providing patients with access to physical, psychological or alternative therapies, including acupuncture. We also found through the barriers and facilitators review that the motivation of healthcare professionals to taper was impacted by their patient’s motivation. Few studies involved intervention components that appeared to directly target the motivation of healthcare professionals. Two studies, which were both linked to US OSI, involved feedback on prescribing practices and endorsement for opioid dose reduction from peers or senior leadership.
Patient and public involvement
A patient and public adviser was involved as a co-applicant. We included a person with lived experience in the research team, and they advised on the development of the funding application and early stages of our project. They did not have further input into the project when they left the service they had supported. Our stakeholder consultation also included people with lived experience. The stakeholders helped us contextualise our findings and consider implementation issues.
Equality, diversity and inclusion
The research looked to serve a heterogeneous population with a range of sociodemographic characteristics. There is wide variation in opioid prescribing practices in the UK,15 but more opioids are prescribed in areas of greater social deprivation. For this reason, it was important that we considered and evaluated the potential inequalities in access to interventions that aim to reduce the use of opioids. The use of the PROGRESS-Plus framework in our review enabled a comprehensive exploration of inequalities across a wide range of sociodemographic characteristics, beyond the most investigated measures of gender and age. In addition, we considered various ‘Plus’ factors (age, comorbidities and status as a long-term opioid user) known to impact on the effectiveness of, and access to, opioid-tapering interventions.
Strengths
This evidence synthesis has many strengths and we conducted our reviews in accordance with international systematic review guidance.28,29,30 The reviews of effectiveness, safety and acceptability build on previous systematic reviews21,23 and the reviews of barriers and facilitators and inequalities add novelty. To our knowledge, the review of barriers and facilitators is the first to focus specifically on opioid tapering in patients with chronic non-cancer pain and the review of inequalities the first to comprehensively review the impact of inequalities across the whole tapering pathway.
We took a sensitive approach to searching to ensure that all voices (i.e. participant, provider and carer) were heard and that all types of barriers and facilitators were captured (individual, organisational, wider contextual). We aimed to minimise publication bias by searching for both published and unpublished literature. To minimise selection bias, two reviewers screened all studies. Studies were included where barriers and facilitators were not the primary focus, for example, studies that explored broader issues around experiences of opioid tapering.
Further strengths included the consideration of a comprehensive range of effectiveness outcomes (n = 49). These outcomes were chosen based on data collected as part of previous systematic reviews on this topic,21,23 with the addition of consideration of Initiative on Methods, Measurement and Pain Assessment in Clinical Trials, recommendations for core outcome measures for chronic pain clinical trials (Dworkin 2005). This approach ensured that the review collected data relating to outcomes of importance to patients. In line with recommendations by Bach-Mortensen et al.,141 we clearly defined the concept of ‘barrier’ and ‘facilitator’ in our review of barriers and facilitators, described how factors were identified and extracted from included studies and assessed the level of confidence in our findings.
Limitations of the research
All of the studies identified for inclusion were done in USA, Europe or Australia, and none were done in the UK. This raises an important question about the generalisability of the interventions to NHS practice. Generalisability is of particular concern in relation to the six studies that were undertaken in VA facilities. While the VA is reported to often (but not always) perform better than, or similarly to, other US systems of care with regards to the safety and effectiveness of care, analyses have shown that VA users are more likely than non-VA users to be elderly, non-Hispanic black, lower income, in poorer health and live in non-metropolitan areas.
Factors including type of chronic pain, duration of pain, type and duration of opioid use and level of opioid use may influence the efficacy of a tapering intervention. These factors tended to be poorly reported in the included studies, and the included studies tended to recruit patients with multiple types of chronic pain. Further, in each study, patients were taking a range of different opioids at a range of different doses.
Currently, there is no guidance on how to assess confidence/certainty in the results of mixed-methods reviews using GRADE. By transforming quantitative data into qualitative data,30 we aimed to be inclusive in making a judgement on the confidence in the review findings applied GRADE-CERQual45 to provide some consistency, although we accept this is a somewhat modified approach. In addition, none of the five studies located in the update search identified any new themes that would change the findings of the review (i.e. data saturation had been achieved) and were therefore not incorporated into the review. However, the addition of these studies may increase confidence in some of the findings and can be incorporated in future updates.
For the review of inequalities, we did not look to analyse intersectionality of two or more PROGRESS-Plus factors. Patients may experience multiple inequalities which may add unique consequences to their health status during and following an opioid taper, and/or access to tapering support. Furthermore, we did not look at how inequalities are impacted over time, for example, whether the impact of opioid tapering on inequalities changes over time. Finally, the studies included in the review of inequalities were taken from a set of studies included in two parallel reviews on the effectiveness of, barriers and facilitators to, opioid tapering. As a result, relevant studies which did not meet the inclusion criteria for these two reviews may have been missed.