Pain in the older adults is a common problem, with a prevalence of 28% to 60% (Chung and Wong, 2007). This population has many aspects that need to be taken into consideration when treating pain, both acute and chronic (Ali et al, 2018). Specifically, comorbidities can restrict the use of many common pain medications (Ali et al, 2018). The changes in the pharmacokinetics and pharmaco-dynamics make it difficult to achieve a balance between the appropriate dosage for effective pain relief and the management of side effects (Chung and Wong, 2007). The effective control of pain in older adults needs more than just pharmacological management (Noroozian et al, 2018). Non-pharmacological interventions have been shown to have potential when combined with pain medication (Shropshire et al, 2019). However, previous reviews in this area included methodologically weak studies and are becoming out of date (Tang et al, 2019). The systematic review by Tang et al (2019) (Box 1) aimed to assess the effectiveness, suitability and sustainability of non-pharmacological interventions for the management of pain in community-dwelling older adults.
Box 1.Paper being reviewedTang SK, Tse MMY, Leung SF, Fotis T. The effectiveness, suitability, and sustainability of non-pharmacological methods of managing pain in community-dwelling older adults: a systematic review. BMC Public Health. 2019;19:1488. https://doi.org/10.1186/s12889-019-7831-9
Aim of commentary
This commentary seeks to critically appraise the methods used in Tang et al (2019) and expand upon their findings in the context of clinical practice.
Methods
A comprehensive multi-database search was undertaken from date of inception to March 2019. Only randomised controlled trials (RCTs) published in English from 1 January 2005 to 28 February 2019 were included. Trials that were included assessed the effectiveness of non-pharmacological interventions for non-malignant chronic pain management in older adults (over 65 years of age) in the community, but not in a care home. Screening, data extraction and assessment of bias (Jadad Scale) were carried out by three independent reviewers. The primary outcome for the review was reduction in the mean intensity of pain. Pain intensity levels were converted into a numerical rating scale ranging from 0 to 10, irrespective of which assessment tool was used. Individual study improvement was then assessed by comparing the pre-intervention mean pain intensity to the post-intervention mean pain.
Results
Out of the 662 articles, ten randomised controlled trials were identified, comparing acupressure, acupuncture, guided imagery, qigong, periosteal stimulation, and Tai Chi. A wide range of pain assessment scales were used, which included the Visual Analogue Scale, numeric rating scales, and the Western Ontario and McMaster University Osteoarthritis Index. Using the Jadad Scale, only two studies scored 5/5, with the remaining studies scoring between 2 and 3. The main methodological issue was related to the lack of double blinding. The systematic review reports that all included RCTs demonstrated a decrease in pain intensity from pre-to post-intervention. Based on the undefined conversion of multiple pain scales to a 10-point pain scale, reductions ranged from − 3.13 to − 0.65 in the intervention group from pre-to post-intervention. Six out of ten studies reported an undefined statistically significant improvement. Additionally, seven studies found an undefined statistically significant decrease in pain intensity in the follow-up assessment.
Commentary
The Amstar2 tool is used to evaluate the quality of the methods used in a systematic review by identifying if 16 key methodological processes have been undertaken (Shea, 2017). Using the Amstar2 tool for critical appraisal of systematic reviews, seven out of the 16 criteria were judged to be satisfactory for this review. This assessment was undertaken by ST and checked by JH. The main areas of concern were the methods of synthesis. This review used an unclear method of conversion for pain assessment tools to a 10-point scale and presented a range of effects. However, due to the ordinal data used in the pain assessment tools, this type of rough conversion should not be carried out. Alternatively, the more appropriate method of a meta-analysis, using standardised mean difference, could have been used (Higgins, 2021). Furthermore, when presenting the range estimation of effect, only the pre-and post-test intervention group were presented, which could be substantially misleading as there is no comparator. There were also concerns regarding the primary studies included in this systematic review, with the majority of studies not blinding participants, potentially affecting the subjective estimates of pain. Based on this and the other methodological issues in both the review and the included studies, substantial caution should be used when applying these findings to practice.
Nevertheless, there were six out of ten RCTs that found a statistically significant difference between intervention and control. These interventions were acupuncture (two RCTs), Tai Chi (two RCTs), periosteal stimulation (one RCT) and Qi-therapy (one RCT). Previous reviews have demonstrated that acupuncture and Tai Chi have potential in reducing pain in the immediate term for adults with lower back pain (Kong et al, 2016; Mu et al, 2020). However, in both reviews, the primary research was judged to be weak, and there were limited long-term benefits demonstrated.
Based on this review, there is insufficient evidence for clinical recommendations on the use of non-pharmacological interventions for pain management in community-dwelling older adults. However, pain management for older adults in the community still presents a unique challenge, due to the complexity of both polypharmacy (Dagli et al, 2014) and multimorbidity factors (Salive, 2013), which result in the need for effective non-pharmacological interventions.
Thus, further high-quality, multi-centre RCTs are still required in this area; wherever possible, both patients and assessors should be blinded to the intervention. Core outcome sets should also be used, and adverse events and attrition rates should be reported. Due to the substantial methodological weakness of this review, there is a need for this review to be updated and appropriate methods of synthesis applied. Furthermore, as it is unlikely that any accurate estimation of effect will be established (due to the wide range of intervention types and very specific population), it may be required that this review is re-scoped.
Key points
- Non-pharmacological interventions may reduce pain in community-dwelling older adults
- There is substantial variation in the intervention types, which makes specific estimations of effect difficult
- Further high-quality, multi-arm, non-pharmacological comparative blinded random controlled trials are required.
CPD reflective questions
- What are the main limitations of the primary studies included in this review?
- What future research is required in this area?
- If applying a non-pharmacological intervention to practice, what concerns would you have?