provides a summary of the recommendations formulated by the GDG, presented by class of intervention and the associated recommendations.
Each recommendation is relevant to community-dwelling adults experiencing CPLBP, with or without spine-related leg pain. While the population definition allowed for the inclusion of comorbid spine-related leg pain, the GDG was not able to confidently interpret the effects of interventions in this subpopulation since classification systems varied across trials and some trials did not report prevalence of leg pain. Inconsistency in classification of spine-related leg pain and its reporting in trials is a recognized limitation in the literature, prompting the development of recent recommendations for terminology and the identification of neuropathic pain in people with spine-related leg pain (67). Where the included trials reported outcomes separately for older people or included older people (adults aged 60 years and over) in the mean age range of included trials, or evidence (direct or indirect) of harms are also relevant to older people, the recommendations also refer to older people. Where these criteria were not met, the recommendations refer to adults only.
Intervention class C. Psychological interventions
Psychological interventions considered for the guideline comprised five interventions: operant, respondent, cognitive, cognitive behavioural and mindfulness-based stress reduction therapies. This section provides a summary of the evidence relating to all psychological interventions combined, while subsections C.1–C.5 provide the intervention-specific evidence summaries, rationales and judgements.
Definition of the interventions
Three interventions (operant, respondent and cognitive therapies) aligned with an earlier Cochrane review of behavioural treatments for LBP (4). Each of these interventions focuses on modifying one of the three response systems which characterize emotional experiences: behaviour, physiological reactivity, and cognition, respectively. However, there is an acceptance that psychological interventions are complex and multifaceted and that treatment for chronic pain may not be appropriately bound by this classification (5). These three interventions are therefore often applied in a combined treatment approach, commonly referred to as cognitive behavioural therapy (CBT). Acceptance and commitment therapy, an extension of CBT, was not considered for the guideline. Mindfulness-based stress reduction therapy was also considered as an intervention for the guideline. This intervention aims to reduce pain through improved tolerance/acceptance of body sensations.
While other psychological interventions may be relevant to the management of CPLBP, such as those outlined in the WHO mhGAP Intervention Guide – Version 2.0 for mental, neurological and substance use (MNS) disorders in non-specialist health settings (106), the guideline considers these five interventions as stand-alone interventions for CPLBP. Intervention-specific definitions are provided in the subsequent sections C.1–C.5.
Summary of the evidence across all psychological interventions
Quantitative review
Characteristics of the evidence
The evidence for the benefits and harms of psychological interventions in the treatment of CPLBP was based on an update and extension of an earlier Cochrane review, which considered the three distinct interventions (operant, respondent and cognitive therapies) and the combined treatment approach: cognitive behavioural therapy (CBT) (4). For the current synthesis, mindfulness-based stress reduction therapy was added to this suite of interventions. The current synthesis comprised 44 trials involving a total of 5996 participants. Forty trials were carried out in 12 HICs (United States: 10 trials; Germany: 8 trials; Netherlands (Kingdom of the): 4 trials; Australia: 5 trials; United Kingdom: 3 trials; Sweden: 3 trials; Italy: 2 trials; Austria: 1 trial; Romania: 1 trial; Norway: 1 trial; Denmark: 1 trial, Israel: 1 trial), one in an UMIC (Brazil: 1 trial) and the remaining three in three LMICs (1 each in the Islamic Republic of Iran, Ghana and Pakistan). All trials included participants across an age group ranging from 20 to 65 years. No trials reported on older age groups separately. Two trials included only females while the other 42 trials included mixed male and female populations. No trial included only males. No trials reported on marginalized populations separately and eight trials described the race/ethnicity of their population.
A summary of included trials by comparator for each psychological intervention is provided in .
Number of psychological therapy trials by intervention and comparator.
Quantitative outcomes of the syntheses are reported for each specific intervention, by comparator, in the subsequent sections C.1–C.5. Where no information is provided for a specific intervention (e.g. comparator or outcome), this indicates that no evidence was available for that intervention.
Qualitative review
No qualitative evidence was identified specific to psychological interventions generally for the following EtD domains: resource implications, equity and human rights, acceptability or feasibility. Some evidence related to values and preferences for MBSR therapy specifically was identified and this is presented in section C.5.
Overall judgements and EtD considerations for psychological interventions generally
The GDG considered the EtD domains of values and preferences, resource requirements, equity and human rights, acceptability and feasibility for all psychological interventions together, since the GDG expected that judgements for these domains would not differ byfor each psychological intervention. This section therefore outlines GDG members’ judgements for those domains as they relate to psychological interventions in general. These judgements should be considered alongside the rationale for each psychological intervention.
The GDG judged that values and preferences for older people relating to psychological interventions were likely to be applicable to all adults and that there could be important or possibly important uncertainty or variability among people with CPLBP. Such variability could probably be attributed to psychological interventions not being accessible globally and the considerable variation in sociocultural attitudes and awareness about these interventions. The GDG also acknowledged that subsidy-funding from government or insurance schemes for psychological interventions vary across health systems, making accessibility and affordability difficult for some people, particularly people in low- and middle-income countries. The GDG judged that costs associated with delivery of psychological interventions were likely to be moderate to large. However, the GDG acknowledged that costs would likely vary across health and subsidy schemes, depending on whether specialist or non-specialist health practitioners delivered the intervention(s). The GDG noted that in many cases, people with CPLBP might have to attend several treatment sessions, and that this could be associated with substantial costs. The GDG also noted that costs for workforce training might also be substantial for specialist health practitioners (e.g. clinical psychologists), while costs are likely to be less significant for non-specialist health workers delivering simple behavioural interventions. Across the included trials, psychological interventions were delivered in secondary care settings, which could suggest a lack of implementation feasibility in community and primary care settings, potentially creating a travel burden for people to access care, particularly those living in rural or remote areas. The GDG noted, however, that the WHO mhGAP Intervention Guide and WHO mhGAP community toolkit offer guidance on the delivery of psychological interventions in non-specialized health settings and could provide additional implementation guidance for low-resource contexts (106). Given the potential costs to people with CPLBP and the fact that subsidies for psychological interventions are limited in most health services, health equity could be impacted, although this would vary by setting.
The GDG judged that the feasibility to deliver psychological interventions would vary. Lack of a skilled workforce in community settings and potentially high treatment costs without government or insurance subsidy could limit accessibility and affordability to people with CPLBP, particularly in low-resource settings. Different sociocultural attitudes towards these interventions, accessibility limitations in community settings and potential out-of-pocket expenses are likely to contribute to variations in the acceptability of these interventions. In particular, some GDG members identified that benefit from behavioural therapies requires more than a physiological response to the intervention; it requires an enabling socioeconomic environment, education and health literacy, which might be less available to people in low-resource settings.
◯ C.1. Operant therapy
Definition of the intervention
Operant therapy aims to replace pain-related behaviours with helpful, healthy behaviours (e.g. exercise, work). Time-contingent exercises (i.e. quotas) and encouraging people with CPLBP to increase their activity levels are its main principles. This type of therapy is aligned with behavioural activation therapy.
Recommendation
Operant therapy may be offered as part of care to adults, including older people, with CPLBP
(conditional recommendation in favour of use, very low certainty evidence).
Remarks
Depression, anxiety, stress and other mental conditions are common comorbidities in adults with CPLBP, but may not affect all people with CPLBP. Person-centred assessment with respect and dignity is required to diagnose and manage these conditions. WHO provides
mhGAP recommendations for different psychological interventions, which may be of benefit to people with CPLBP and comorbid mental disorders in non-specialist health settings.
When operant therapy is offered to people with CPLBP, it should be considered as part of a broader suite of effective treatments (i.e. not offered as a single intervention in isolation), based on a biopsychosocial assessment.
Summary of the evidence
Quantitative review
Characteristics of the evidence
Operant therapy was evaluated in 4 trials compared with no intervention (n=391). The trials were conducted in three HICs (Netherlands (Kingdom of the): 1 trial; Sweden: 1 trial; United States: 2 trials). Where reported, the age of participants ranged from 20–64 years and the proportion of females included ranged from 37% to 100%.
Outcomes
Trials did not report on harms.
Web Annex D.C1 provides the detailed GRADE evidence profile tables for the intervention, by comparator.
Qualitative review
No qualitative evidence was identified specific to operant therapy for the following EtD domains: values and preferences, resource implications, equity and human rights, acceptability or feasibility.
Rationale for judgements
For all adults and older people, the GDG judged overall net benefits for operant therapy to be largely uncertain, based on very low certainty evidence from four trials, and that no clinically important benefits were observed for most outcomes although moderate benefits for pain in the short and intermediate follow-up periods were noted. Similarly, harms were judged to be uncertain since they were not measured in the trials. However, the GDG opined that harms, if any, were likely to be trivial. The GDG judged the overall certainty of evidence to be very low, consistent with the ratings provided by the systematic review team The GDG judged that the balance of benefits to harms for operant therapy was uncertain (due to the very low certainty evidence and no evidence on harms) or probably favoured operant therapy (based on moderate effects on pain in the short and intermediate terms). The GDG noted that trials did not employ a sampling stratification based on the mental health of participants and judged that people who experience mental health comorbidities alongside CPLBP might respond differently to operant therapy compared to individuals without mental health comorbidities.
A description of GDG members’ judgements relevant to all psychological interventions for the EtD domains of values and preferences, resource implications, equity and human rights, acceptability and feasibility is presented in the introduction to this intervention class and summarized in Annex 3 (Table C10). Table C10 also provides a summary of judgements made by the GDG for benefits, harms, balance of benefits to harms, and overall certainty for operant therapy.
The GDG reached a consensus decision to make a conditional recommendation in favour of operant therapy. This decision was based on moderate benefits for pain in the short and intermediate periods, without clear evidence of harms. The GDG judged that despite concerns relating to cost, equity, acceptability and feasibility in low-resource settings as outlined above, the signal of moderate benefit to pain for operant therapy outweighed these implementation concerns. The GDG also referred to the WHO mhGAP Intervention Guide and WHO mhGAP community toolkit where guidance on the delivery of psychological interventions in non-specialized health settings and additional implementation guidance for low-resource contexts is provided (106). Six GDG members disagreed with this decision and judged instead that no recommendation would be appropriate for operant therapy, based on very low certainty evidence and the inability to confidently judge the balance of benefits to harms.
◯ C.2. Respondent therapy
Definition of the intervention
Respondent therapy aims to modify the physiological response system to pain through the reduction of muscular tension through biofeedback, progressive relaxation and applied relaxation. This type of therapy is aligned with relaxation therapy.
Recommendation
No recommendation: The balance between the benefits and harms for respondent therapy in managing CPLBP in adults, including older people, is so equivocal that a recommendation cannot be made
(no recommendation, very low certainty evidence).
Key consideration
Depression, anxiety, stress and other mental conditions are common comorbidities in adults with CPLBP but may not affect all people with CPLBP. Person-centred assessment with respect and dignity is required to diagnose and manage these conditions.
WHO provides mhGAP recommendations for different psychological interventions, which may be of benefit to people with CPLBP and comorbid mental disorders in non-specialist health settings.
Summary of the evidence
Quantitative review
Characteristics of the evidence
Respondent therapy was evaluated in ten trials (9 reports) compared with placebo (3 trials, n=144), with no intervention (6 trials; n=344) and with usual care (1 trial; n=234). Nine trials were conducted in five HICs (Australia: 1 trial; Germany: 1 trial; Netherlands (Kingdom of the): 1 trial; United Kingdom: 1 trial; United States: 5 trials) and one was conducted in an UMIC (Brazil: 1 trial). Where reported, the age of participants ranged from 20–66 years and the proportion of females included ranged from 37% to 92%.
Outcomes
In the comparison of respondent therapy with placebo, three trials were identified, although only two trials could be meta-analysed, and no clinically important differences were observed. Respondent therapy may make little to no difference to pain or function in the short term (low certainty). No trials reported on harms.
In the comparison of respondent therapy with
no intervention or where the effect of the intervention could be isolated (six trials), benefits were observed for pain, function, anxiety, coping, health-related quality of life and social participation.
- ›
Since the certainty of the evidence was very low, it was uncertain whether respondent therapy (biofeedback and relaxation methods) reduced pain in the short term (moderate effects).
- ›
Since the certainty of the evidence was very low, it was uncertain whether respondent therapy (biofeedback and relaxation methods) improved back-specific function in the short term (moderate effects).
- ›
Since the certainty of evidence was very low, it was uncertain whether respondent therapy (biofeedback methods) reduced anxiety in the short term (small effect) and made little to no difference to depression, coping or social participation at any time-point (biofeedback and relaxation methods).
Trials did not report on harms.
In the comparison of respondent therapy with usual care (1 trial), no benefits were observed. Since the certainty of evidence was very low, it was uncertain whether respondent therapy (relaxation method) made little to no difference to pain, back-specific function or health-related quality of life in the short or intermediate terms.
The trial did not report on harms.
Web Annex D.C2 provides the detailed GRADE evidence profile tables for the intervention, by comparator.
Qualitative review
No qualitative evidence was identified specific to respondent therapy for the following EtD domains: values and preferences, resource implications, equity and human rights, acceptability or feasibility.
Rationale for judgements
For all adults and older people, the GDG judged overall net benefits across outcomes to be largely uncertain, although some GDG members judged the benefits to be trivial to small. The GDG identified that while some small to moderate benefits were observed for respondent therapy compared with no intervention for pain, function and anxiety, these benefits were limited to the short-term period only and were not observed in placebo-controlled trials or in the single trial compared with usual care. The GDG was therefore uncertain about the true effects of the intervention and its longer-term benefits. Similarly, harms were judged to be uncertain since they were not measured in the trials. However, the GDG opined that harms, if any, were likely to be trivial. The GDG judged the overall certainty of evidence to be very low for respondent therapy, consistent with the systematic review team’s judgements. The GDG judged that the balance of benefits to harms for respondent therapy was largely uncertain due to the very low certainty evidence of benefit for most outcomes across comparators, and an absence of evidence of harms. The GDG noted that trials did not employ a sampling stratification based on the mental health of participants and judged that people who experience mental health comorbidities alongside CPLBP might respond differently to respondent therapy compared to individuals without mental health comorbidities.
A description of GDG members’ judgements relevant to all psychological interventions for the EtD domains of values and preferences, resource implications, equity and human rights, acceptability and feasibility is presented in the introduction to this intervention class and summarized in Annex 3 (Table C11). Table C11 also provides a summary of judgements made by the GDG for benefits, harms, balance of benefits to harms, and overall certainty for respondent therapy.
The GDG reached a consensus decision to make no recommendation for respondent therapy. This decision was primarily based on the GDG not being able to confidently judge the balance of benefits to harms. While some signals of benefit were observed, these were not consistent and the GDG expressed feasibility concerns about the delivery of the intervention in primary and community care settings, making it less comfortable about recommending in favour of the intervention. While some members of the GDG proposed that a conditional recommendation in favour of respondent therapy might be appropriate, this was not the consensus opinion.
◯ C.3. Cognitive therapy
Definition of the intervention
Cognitive therapy aims to identify and modify cognition regarding pain and disability. It is proposed that beliefs about the meaning of pain and expectations regarding control over pain can be directly modified using cognitive restructuring techniques such as imagery and attention diversion.
Recommendation
No recommendation: The balance between the benefits and harms for cognitive therapy in managing CPLBP in adults, including older people, is so equivocal that a recommendation cannot be made
(no recommendation, very low certainty evidence).
Key consideration
Depression, anxiety, stress and other mental conditions are common comorbidities in adults with CPLBP but may not affect all people with CPLBP. Person-centred assessment with respect and dignity is required to diagnose and manage these conditions. WHO provides
mhGAP recommendations for different psychological interventions, which may be of benefit to people with CPLBP and comorbid mental disorders in non-specialist health settings.
Summary of the evidence
Quantitative review
Characteristics of the evidence
Cognitive therapy was evaluated in four trials compared with no intervention (n=364). The trials were conducted in four HICs (Australia: 1 trial; Netherlands (Kingdom of the): 1 trial; Sweden: 1 trial; United States: 1 trial). Where reported, the age of participants ranged from 18–65 years and the proportion of females included ranged from 48% to 62%.
Outcomes
One trial measured harms for cognitive therapy: none was reported.
Web Annex D.C3 provides the detailed GRADE evidence profile tables for the intervention, by comparator.
Qualitative review
No qualitative evidence was identified specific to cognitive therapy for the following EtD domains: values and preferences, resource implications, equity and human rights, acceptability or feasibility.
Rationale for judgements
For all adults and older people, the GDG judged overall net benefits across outcomes to be largely uncertain, although some GDG members judged the benefits to be trivial for coping. The GDG identified that while a trivial benefit to coping might be observed for cognitive therapy compared with no intervention, the effect was not statistically significant and was identified in a single small trial and not replicated. Harms were judged to be uncertain and trivial in the single trial where harms were monitored. The GDG judged the overall certainty of evidence to be very low for cognitive therapy, consistent with the systematic review team’s judgements. The GDG judged that the balance of benefits to harms for cognitive therapy was uncertain due to the very low certainty evidence of benefit for all outcomes, and very low certainty evidence of harms. The GDG noted that trials did not employ a sampling stratification based on the mental health of participants and judged that people who experience mental health comorbidities alongside CPLBP might respond differently to cognitive therapy compared to individuals without mental health comorbidities.
A description of GDG members’ judgements relevant to all psychological interventions for the EtD domains of values and preferences, resource implications, equity and human rights, acceptability and feasibility is presented in the introduction to this intervention class and summarized in Annex 3 (Table C12). Table C12 also provides a summary of judgements made by the GDG for benefits, harms, balance of benefits to harms, and overall certainty for cognitive therapy.
The GDG reached a consensus decision to make no recommendation for cognitive therapy. This decision was primarily based on the GDG not being able to confidently judge the balance of benefits to harms – that is, largely no benefit and no harm. While a possible signal of benefit was observed for coping, this was observed in a single trial only where the effect size was unconvincing. Although harms were likely to be trivial, based on a single trial, the GDG expressed feasibility concerns about the delivery of the intervention in primary and community care settings, making it less comfortable about recommending in favour of the intervention.
◯ C.4. Cognitive behavioural therapy
Definition of the intervention
Cognitive behavioural therapy (CBT) is based on a multidimensional model of pain and focuses on reducing pain and distress by modifying physical sensation, catastrophic thinking and unhelpful behaviour(s). Treatment may include education about a multidimensional view of pain, identifying pain-eliciting and pain-aggravating situations, thoughts and behaviour, and using coping strategies and applied relaxation; that is, integrating the components of operant, respondent and cognitive therapies. Goal-setting and activity increases are encouraged as the basis of CBT to reduce feelings of helplessness and help the person gain control over their pain experience.
Recommendation
Cognitive behavioural therapy (CBT) may be offered as part of care to adults, including older people, with CPLBP
(conditional recommendation in favour of use, very low certainty evidence).
Remarks
Depression, anxiety, stress and other mental conditions are common comorbidities in adults with CPLBP, but may not affect all people with CPLBP. Person-centred assessment with respect and dignity is required to diagnose and manage these conditions. WHO provides
mhGAP recommendations for different psychological interventions, which may be of benefit to people with CPLBP and comorbid mental disorders in non-specialist health settings.
When CBT is offered to people with CPLBP, it should be considered as part of a broader suite of effective treatments (i.e. not offered as a single intervention in isolation), based on a biopsychosocial assessment.
Summary of the evidence
Quantitative review
Characteristics of the evidence
Cognitive behavioural therapy (CBT) was evaluated in 30 trials compared with no intervention (25 trials; n=3636) and with usual care (5 trials; n=1223). Twenty-eight trials were conducted in eleven HICs (Australia: 4 trials; Austria: 1 trial; Denmark: 1 trial; Germany: 7 trials; Italy: 2 trials; Netherlands (Kingdom of the): 2 trials; Norway: 1 trial; Romania: 1 trial; Sweden: 1 trial; United Kingdom: 2 trials; United States: 6 trials) and two were conducted in two LMICs (Ghana: 1 trial; Pakistan: 1 trial). Where reported, the age of participants ranged from 18–78 years and the proportion of females included ranged from 27% to 85%.
Outcomes
Four trials reported on adverse events for CBT. Two trials reported that there were no serious adverse events attributable to either treatment or control. One trial reported minor adverse effects of transient pain worsening (seven participants in the experimental group, five in the control group) and mood disorders (two participants in the experimental group, two in the control group). One trial reported that three participants in the no intervention group stopped the treatment because of increased pain in the lower back or radiating leg pain.
In the comparison of CBT with
usual care (five trials), no differences were observed.
- ›
CBT may make little to no difference to pain in the short and long term (low certainty), while in the intermediate term it was uncertain whether CBT made little to no difference to pain, since the certainty of evidence was very low.
- ›
Since the certainty of evidence was very low, it was uncertain whether CBT made little to no difference to back-specific function in the short term. CBT may make little to no difference to back-specific function in the intermediate and long terms (low certainty).
- ›
CBT may make little to no difference to depression, anxiety, self-efficacy or health-related quality of life at any time-point (low certainty).
Four trials reported on harms for CBT. Of these, one reported no harms. Across the other three trials harms were not attributed to the intervention, rates of harms either did not differ between the intervention and control groups or harms reported were non-serious and transient (e.g. temporary pain associated with muscle relaxation).
Web Annex D.C4 provides the detailed GRADE evidence profile tables for the intervention, by comparator.
Qualitative review
No qualitative evidence was identified specific to CBT for the following EtD domains: values and preferences, resource implications, equity and human rights, acceptability or feasibility.
Rationale for judgements
For all adults and older people, the GDG judged overall net benefits to be small for CBT, based on the standardized mean differences estimated in the meta-analyses, while acknowledging that the clinically worthwhile effects by outcome ranged from small to large. The GDG judged harms to be trivial, based on the observations in the trials and the experience among the GDG. The GDG also referred to indirect evidence from an aligned Cochrane review of psychological interventions, including CBT, for the management of chronic pain in adults and identified no to minimal harms with delivery of CBT compared to usual care and small benefits for pain, function and distress at the end of treatment and follow-up (6 months or more after treatment) (107). In comparison to the other psychological interventions, the GDG acknowledged the much larger body of evidence for CBT (30 trials), while noting the GRADE certainty downgrading decisions by the systematic review team because of the risk of bias and inconsistency in CBT trials. Consequently, GDG judged the overall certainty of evidence for CBT to be very low.
The GDG judged that the balance of benefits to harms for CBT probably favoured CBT, although some GDG members judged this balance to be uncertain due to very low certainty evidence for several outcomes. The GDG noted that trials did not employ a sampling stratification based on the mental health of participants and judged that people who experience mental health comorbidities alongside CPLBP might respond differently to CBT, compared to individuals without mental health comorbidities.
A description of GDG members’ judgements relevant to all psychological interventions for the EtD domains of values and preferences, resource implications, equity and human rights, acceptability and feasibility is presented in the introduction to this intervention class and summarized in Annex 3 (Table C13). Table C13 also provides a summary of judgements made by the GDG for benefits, harms, balance of benefits to harms, and overall certainty for CBT.
The GDG reached a consensus decision to make a conditional recommendation in favour for CBT. This decision was based on small to large effects on pain, function, coping and quality of life, without evidence of harms. The GDG was confident that benefits outweighed harms. The GDG judged that despite concerns relating to cost, equity, acceptability and feasibility in low-resource settings (as outlined in the section on judgements for all psychological interventions), the consistent signals of benefit for CBT outweighed these implementation concerns. The GDG also referred to the WHO mhGAP Intervention Guide and WHO mhGAP community toolkit where guidance on the delivery of psychological interventions in non-specialized health settings and additional implementation guidance for low-resource contexts is provided (106).
Two GDG members disagreed with this decision and judged that no recommendation would instead be appropriate for CBT, based on very low certainty evidence and the inability to confidently judge the balance of benefits to harms.
◯ C.5. Mindfulness-based stress reduction therapy
Definition of the intervention
Mindfulness-based stress reduction (MBSR) therapy aims to reduce stress by developing mindfulness: a non-judgemental, moment-by-moment acceptance of awareness. The intervention is free of any cultural, religious and ideological factors, but it is associated with the Buddhist origins of mindfulness.
Recommendation
No recommendation: The balance between benefits and harms for mindfulness-based stress reduction (MBSR) therapy in managing CPLBP in adults, including older people, is so equivocal that a recommendation cannot be made
(no recommendation, low certainty evidence).
Key consideration
Depression, anxiety, stress and other mental conditions are common comorbidities in adults with CPLBP but may not affect all people with CPLBP. Person-centred assessment with respect and dignity is required to diagnose and manage these conditions. WHO provides
mhGAP recommendations for different psychological interventions, which may be of benefit to people with CPLBP and comorbid mental disorders in non-specialist health settings.
Summary of the evidence
Quantitative review
Characteristics of the evidence
Mindfulness-based stress reduction (MBSR) therapy was evaluated in two trials compared with no intervention (n=155) and one trial compared with usual care (n=342). Two trials were conducted in two HICs (Israel: 1 trial; United States: 1 trial) and one trial in a LMIC (Islamic Republic of Iran: 1 trial). Where reported, the age of participants ranged from 20–70 years and the proportion of females included ranged from 66% to 100%.
Outcomes
Neither trial reported on harms.
In the comparison of MBSR therapy with
usual care (1 trial), no clinically meaningful benefits were observed.
- ›
MBSR therapy may reduce pain at all time-points, although the size of the effect was trivial (low certainty).
- ›
It was uncertain whether MBSR therapy improved back-specific function at all time-points (trivial effects), since the certainty of the evidence was very low.
- ›
MBSR therapy may improve depression in the short term only, although the size of the effect was trivial, while at other time-points MBSR therapy may make little to no difference to depression, health-related quality of life or anxiety (low certainty).
In the one trial measuring harms associated with MBSR therapy, harms were non-serious and transient (e.g. temporary increase in pain).
Web Annex D.C5 provides the detailed GRADE evidence profile tables for the intervention, by comparator.
Qualitative review
View in own window
| # | Review findings: values and preferences relevant to older people | GRADE-CERQual assessment of confidence | Explanation of confidence assessment |
|---|
| 18 | Mindfulness and meditation allowed some older people to increase their body awareness in relation to, for example, breathing, posture, cognition and pain. In some cases, this allowed for early recognition of pain. | VERY LOW | Minor concerns regarding methodological limitations, no/very minor concerns regarding coherence, serious concerns regarding adequacy, and serious concerns regarding relevance. |
| 19 | Mindfulness and meditation allowed older people to examine, assess, understand and accept their pain rather than avoid it. For some people, this lessened the significance or power of the pain experience, allowed them to gain a sense of control over their lives and increase their ability to relax and respond to stress, with improved sleep, attention, well-being and general quality of life. | LOW | Moderate concerns regarding methodological limitations, no/very minor concerns regarding coherence, minor concerns regarding adequacy, and serious concerns regarding relevance. |
| 20 | Some older people were able to use mindfulness and meditation for pain management and coping to varying degrees. Some older people experienced no relief, while others experienced some or shor9789240081789-eng-SPG-term relief, and a few were able to eliminate feelings of pain. | LOW | Serious concerns regarding methodological limitations, no/very minor concerns regarding coherence, minor concerns regarding adequacy, and serious concerns regarding relevance. |
No qualitative evidence was identified specific to MBSR therapy for the following EtD domains: resource implications, equity and human rights, acceptability or feasibility.
Rationale for judgements
For all adults and older people, the GDG judged overall net benefits across outcomes to be largely uncertain, based on a dichotomous improvement in pain in a single trial compared with no intervention, which was not replicated in the trial of MBSR compared with usual care. Furthermore, the effect size of benefits for pain, depression and function in the trial compared with usual care were below a clinically worthwhile threshold. Harms were judged to be uncertain and trivial in the single trial where harms were monitored. The GDG judged the overall certainty of evidence to be low for MBSR therapy, consistent with the systematic review team’s judgements. The GDG judged that the balance of benefits to harms for MBSR therapy was uncertain due to the low to very low certainty evidence of benefit for all outcomes, and absence of evidence for harms. The GDG noted that trials did not employ a sampling stratification based on the mental health of participants and judged that people who experience mental health comorbidities alongside CPLBP might respond differently to MBSR therapy compared to individuals without mental health comorbidities.
The GDG referred to the qualitative evidence synthesis which provided evidence on values and preferences relating to the outcomes of MBSR therapy in older people. The GDG judged that older people valued MBSR therapy as a pain management tool and as a strategy to improve their overall wellbeing (low to very low confidence). The GDG opined that these values and preferences would likely extend to all adults, yet there would be important uncertainty or variability in these values and preferences since exposure to, and acceptability of, MBSR therapy would probably vary across people with CPLBP.
GDG members’ judgements of the other EtD domains of resource implications, equity and human rights, acceptability and feasibility are described as judgements relevant to all psychological interventions in the introduction to this intervention class and summarized in Annex 3 (Table C14). Table C14 also provides a summary of judgements made by the GDG for benefits, harms, balance of benefits to harms, and overall certainty for MBSR therapy.
The GDG reached a consensus decision to make no recommendation for MBSR therapy. This decision was primarily based on the GDG not being able to confidently judge the balance of benefit to harm – that is, largely no benefit and no harm. While a possible signal of benefit was observed for pain interference, this was observed in a single trial only and measured using a dichotomous outcome, while the effect size of other benefits was below a clinically worthwhile threshold, as judged by the GDG. Although harms were likely to be trivial based on a single trial and evidence suggested that some older people valued MBSR therapy, the GDG expressed feasibility concerns about delivery of the intervention in primary and community care settings, making it less comfortable about recommending in favour of the intervention. While some members of the GDG proposed that a conditional recommendation against MBSR therapy might be appropriate, this was not the consensus opinion.
Intervention class D. Medicines
Medicines considered for evidence synthesis included nine systemic pharmacotherapies (D.1), cannabis-related pharmaceutical preparations for therapeutic use (D.2) injectable local anaesthetics delivered to extraspinal soft or connective tissues in the anatomical region of the lower back (D.3) and eight herbal medicines (D.4).
WHO’s Model List of Essential Medicines (EML) presents a list of minimum medicine needs for a basic health care system, listing the most efficacious, safe and cost-effective medicines for priority conditions. Prescribers of medicines should follow safe medication practices, as outlined by WHO in its medication without harm framework. Additional safe medication practices relevant to the agents considered in the guideline are outlined below.
Safe medication practices
Safe medication practices include assessing a person’s overall health condition including their physical and mental capacities, underlying health conditions, potential risk factors for medicines-related harms, and concurrent medicines and drug interactions as well as their values, preferences and priorities for care.
New medicines should be initiated at the lowest effective dose for the shortest duration of time, and beneficial and possible harmful effects should be closely monitored.
Age, multimorbidity and polypharmacy are the major predictors for an increased risk of experiencing a medication-related harm (
108,
109). A systematic review found that medication-related harm is most frequent in older people, with 40% of events being classed as moderate harm and 26% as clinically severe or life-threatening (
110). Clinical vigilance – which includes monitoring the effects of medicines and balancing benefits and harms – is therefore critical when prescribing medicines to older people.
Older people who visit multiple health workers or have been hospitalized recently are at greater risk of polypharmacy. Polypharmacy is commonly described as the simultaneous use of multiple (usually five or more) medicines and is associated with adverse drug reactions (
58). Because polypharmacy can contribute to losses across multiple domains of intrinsic capacity and cause adverse drug reactions (due to drug-drug and drug-disease interactions), a person-centred assessment in any older person should include a medication review. Polypharmacy can be reduced by discontinuing unnecessary, ineffective and potentially redundant medications. Discontinuing medicines suspected of causing harm is a priority.
Although WHO conditionally recommends against the routine use of opioid analgesics for the management of CPLBP (see D.1.1.), in situations where opioid analgesics are prescribed based on clinical judgement and shared decision-making it is essential to consider their safe and effective stewardship. This is because of the considerable risks of dependence, adverse drug reactions and overdose associated with opioids, and the need to manage the stigma associated with opioid use and/or withdrawal. The following practice points are provided to ensure the safety of people with CPLBP and minimize the risk of opioids inappropriately spreading into communities.
- ›
Opioid analgesics should never be used as a stand-alone treatment for adults with CPLBP. Opioids should always be prescribed in the context of the biopsychosocial model of pain care, where interventions to address the multiple contributors to a pain experience are provided.
- ›
Education about the benefits and harms of opioids as well as the safe storage of opioids in households should be provided prior to commencing therapy.
- ›
A clinical assessment in advance of opioid prescription should consider a person’s psychosocial history, patterns of earlier opioid consumption and any history of substance use in order to consider whether there is any risk of improper use or dependence, or potential for hyperalgesia. WHO provides guidance on treating people dependent on heroin or other opioids within the WHO Guidelines for the psychosocially assisted pharmacological treatment of opioid dependence (111).
- ›
The time-limited use of opioids for selected people with CPLBP should always be made at the lowest appropriate dose and shortest feasible duration, and regularly reviewed to ensure the fewest possible adverse events. Prolonged use or misuse can lead to opioid dependence4 and other adverse outcomes, including overdose and hyperalgesia.
- ›
In older people, non-pharmacological interventions and non-opioid treatments should be prioritized. If necessary, opioids might be prescribed after carefully evaluating possible benefits (especially regarding quality of life and physical and mental capacities), possible harms (e.g. worsening of respiratory conditions, dizziness, balance disorders, constipation, falls, delirium) and potential misuse. Supportive treatments, such as the concomitant use of stimulant laxatives, can reduce the burden of some adverse effects such as opioid-induced constipation.
- ›
The prescription of opioids must be undertaken by an appropriately trained and experienced prescriber after careful assessment of the benefits and risks. The prescriber should also take responsibility for regular follow-up care, monitoring and dose adjustment. Monitoring a person’s response to opioids and adjusting the dose parameters as necessary is a key component of opioid stewardship, including prescribing and discontinuation.
- ›
As part of opioid stewardship, prescribers have a responsibility to discuss and develop a tapering and discontinuation plan together with the person before initiating changes, taking into account their individual circumstances (112). An open discussion between the prescriber and the person should include perceptions of the benefits, risks, and adverse effects of continued opioid therapy, and any concerns related to tapering. A systematic review of international guidelines on opioid deprescribing in chronic non-cancer pain provides further guidance on discontinuation practices (113).
- ›
In circumstances where opioid dependence has developed, intervention support for opioid withdrawal may be indicated (111).
- ›
Risk factors for overdose of prescribed opioids include opioid dependence, higher dose, male sex, older age, multiple prescribed medicines, mental health disorder and lower socioeconomic circumstances (114). Identification and management of modifiable risk factors are important factors in preventing overdose. WHO provides guidance on management of opioid overdose, including the use of naloxone, in the WHO Guideline on community management of opioid overdose (114).
- ›
A stigma may be associated with opioid use and/or opioid withdrawal, and act as a barrier to optimal pain management and quality care. People using these medicines may require additional support to address stigma (115).
- 4
Opioid dependence is a disorder of regulation of opioid use arising from repeated or continuous use of opioids. The characteristic feature of dependence is a strong internal drive to use opioids, which manifests itself by an impaired ability to control use, increasing priority given to use over other activities and persistence of use despite harm or negative consequences. Physiological features of dependence may also be present, including increased tolerance to the effects of opioids, withdrawal symptoms following cessation or reduction in use, or repeated use of opioids or pharmacologically similar substances to prevent or alleviate withdrawal symptoms (reference: https://www.who.int/news-room/fact-sheets/detail/opioid-overdose).
New medicines should be initiated at the lowest effective dose for the shortest duration of time, and beneficial and possible harmful effects should be closely monitored.
D.1. Systemic pharmacotherapies
Section D.1 outlines the recommendations for nine systemic pharmacotherapies, described in D.1.1 to D.1.9.
Definition of the intervention
Systemic pharmacotherapies are medicines that act on the whole body or body systems that involve the entire body, such as the endocrine or/and cardiovascular systems.
Summary of the evidence
Quantitative review
Characteristics of the evidence
The evidence synthesis for benefits and harms of systemic pharmacotherapies was derived from:
- ›
recent and ongoing Agency for Healthcare Research and Quality (AHRQ) systematic reviews of opioids and non-opioids for chronic pain (;
- ›
a Cochrane review on systemic glucocorticoids for low back pain (118); and
- ›
data from two in-progress network meta-analyses of analgesic medicines for chronic low back pain that included trials with shorter duration treatment (119, 120).
The systematic review team selected evidence from these sources that met the PICO criteria for the guideline and aggregated findings into a single evidence synthesis report for the GDG.
The current synthesis of benefits and harms of systemic pharmacotherapies of short- and long-term treatment duration was derived from 63 trials with a total of 13 408 participants. Sixty-two of these trials compared a medicine to placebo and one compared a medicine to no treatment. No “usual care” comparator was included in this evidence review. The summary of included trials by medicine class and treatment duration is presented in .
Summary of trials by medicine and comparator for short and long treatment duration.
Characteristics of the included trials, by medicine class, are summarized in Annex 2 (Table B1). All systemic pharmacotherapies were administered orally, other than the four trials of skeletal muscle relaxant (botulinum toxin A) compared with placebo, which was delivered via intramuscular injection into the paravertebral muscles. There were 40 trials of long treatment duration (n=11 990), comprising six different comparisons.
There were 23 trials of short treatment duration (n=1418), comprising seven different comparisons. There were no trials evaluating paracetamol (acetaminophen) or benzodiazepines, and consequently no information about these medicines in Annex 2
Table B1. There were no trials that only examined older people.
Qualitative review
Since the qualitative evidence review did not identify medicine-specific findings (other than for opioids in the acceptability domain), this qualitative evidence summary is applicable to all systematic pharmacotherapies.
View in own window
| # | Review findings: values and preferences relevant to older people | GRADE-CERQual assessment of confidence | Explanation of confidence assessment |
|---|
| 6 | Many older people reported that medication was often the only intervention that made a difference to the severity of their pain. However, they were apprehensive of, or dissatisfied with, medication for several reasons, often viewing it as a quick fix, temporary relief or just masking the pain. Many participants were apprehensive of taking too many medications, side-effects, addiction or did not like how the medications made them feel. Some avoided taking medications all together, did not present their prescriptions for dispensing or adjusted their treatment themselves because of this. | MODERATE | Minor concerns regarding methodological limitations, no/very minor concerns regarding coherence, no/very minor concerns regarding adequacy, and moderate concerns regarding relevance. |
| # | Review findings: resource implications relevant to older people | GRADE-CERQual assessment of confidence | Explanation of confidence assessment |
|---|
| 8 | In one study conducted in rural Nigeria, older people considered medicines as a legitimate form of treatment (a cultural norm where disease was treated and “cured” with medication) and depended on medicines to be able to perform daily tasks. Other treatments were looked down on or stigmatized, such as exercise. Some older people took medication only to comply with this cultural norm. However, there was a constant struggle to be able to afford the medicines on which they depended to function normally. | LOW | No/very minor concerns regarding methodological limitations, no/very minor concerns regarding coherence, serious concerns regarding adequacy, and moderate concerns regarding relevance. |
| # | Review findings: resource implications relevant to older people | GRADE-CERQual assessment of confidence | Explanation of confidence assessment |
|---|
| 9 | Many older people expressed fear of addiction to medication, especially to opioids. This led them to not present prescriptions for dispensing, adjust the dosage or stop taking the medication often without consulting their health worker. In one case, the fear of addiction stemmed from the health care worker who refused to provide the prescription requested. | MODERATE | Minor concerns regarding methodological limitations, no/very minor concerns regarding coherence, minor concerns regarding adequacy, and moderate concerns regarding relevance. |
| 10 | In a study in rural Nigeria, some older people reported that when the locally manufactured medicines failed to relieve symptoms, they believed that they were fake or substandard. These older people thought that imported medicines were stronger and more likely to lead to a cure. | LOW | No/very minor concerns regarding methodological limitations, no/very minor concerns regarding coherence, serious concerns regarding adequacy, and moderate concerns regarding relevance. |
No qualitative evidence was identified specific to systemic pharmacotherapies for the following EtD domains: equity and human rights and feasibility.
Overall judgements and EtD considerations across systematic pharmacotherapies generally
The GDG considered the EtD domains of values and preferences, resource requirements, equity and human rights, acceptability and feasibility for all systemic pharmacotherapies together, since the GDG expected that judgements for these domains would not differ by agent. Therefore, this section outlines GDG judgements for those domains as they relate to systemic pharmacotherapies in general. These judgements should be considered alongside the rationales for each systemic pharmacotherapy agent.
The GDG judged that, for older people, values and preferences relating to systemic pharmacotherapies were likely to be applicable to all adults. The GDG judged that in both low- and middle-income and high-income countries, people with CPLBP tended to prefer symptom relief over the potential for adverse events, consistent with findings from the qualitative evidence synthesis. On this basis, there was likely to be no important uncertainty based on preferences for analgesia for symptom relief. However, the GDG noted the results of the qualitative review indicating that some older people did not value medicines as a sole intervention for their CPLBP. The GDG noted that outside a consistent preference for symptom relief, values and preferences for pharmacotherapies might vary due to differences in how people understand the benefits and harms of medicines, their experience in using medicines, the cost associated with medicines, and the relative time-and-effort burden of engaging in other non-pharmacological treatments. Depending on their personal beliefs and sociocultural context, some people might value medicines over rehabilitative interventions, since short term pain relief and lower cost could enable a more rapid return to work and participation than other non-pharmacological (e.g. behavioural) interventions, where the time to experience a therapeutic effect might take longer. The GDG judged that resource requirements would vary by setting and medicine, ranging from small costs to large costs, especially for some opioid agents. The GDG considered that the cumulative costs for medicines could amount to a large burden for people with CPLBP and their families and ultimately reduce health equity. For example, the GDG referred to the qualitative evidence synthesis findings and noted a single study from rural Nigeria, where participants expressed a constant struggle to afford the cost of medicines. In contexts where the cost burden was less significant, the GDG judged there would be less impact on health equity. The GDG judged that health workers and people with CPLBP would largely accept systemic pharmacotherapies (in general) as an analgesic intervention for CPLBP, but that this might be less true of opioid analgesics based on the risk of adverse events. This judgement was supported by data from the qualitative evidence synthesis indicating that older people were unwilling to accept the risk of addiction to medicines, especially opioid analgesics. The GDG also noted that for health workers, it might also be less acceptable to prescribe opioid analgesics for adults with CPLBP given the high risk of harms, and that they might be more willing to prescribe combination therapies rather than single agents in isolation. The GDG judged that it would be feasible to provide systemic pharmacotherapies in most settings.
The following sections (D.1.1–D.1.9) outline the recommendation, summary of quantitative evidence and rationale and evidence considerations for each systematic pharmacotherapy.
Web Annex D.D1 provides the detailed GRADE evidence profile tables for each agent, by comparator.
◯ D.1.1. Opioid analgesics
Recommendation
Opioid analgesics should not be used as part of routine care for adults, including older people, with CPLBP
(conditional recommendation against use, moderate certainty).
Remarks
There are potentially serious adverse events associated with the use of opioid analgesics, such as dependence and overdose.
Refer to information on safe medication practices (p.112) for guidance when opioids analgesics are prescribed.
Summary of the evidence
Quantitative review
Outcomes
In the comparison of opioids (treatment duration ≥ 1 month with
placebo (27 trials, 8688 participants), the certainty of evidence ranged from high to very low and benefits were observed for pain, function and health-related quality of life (physical domain).
- ›
Opioids probably offered a small reduction in pain intensity compared to placebo at one month to less than six months (moderate certainty evidence). Estimates were similar for opioid agonists, partial agonists and mixed agents. A meta-regression that evaluated the opioid dose as a continuous variable found no association between higher doses and greater effects of opioids on pain (p=0.26), with a plateauing of effects at around 50 mg morphine equivalents per day.
- ›
Opioids were probably associated with a small increased likelihood of experiencing an improvement in pain at one month to less than six months (moderate certainty evidence).
- ›
Opioids probably offered a small improvement in function compared to placebo at one month to less than six months (moderate certainty evidence). Estimates were similar when the analysis was stratified by opioid type. When measuring function as the likelihood of experiencing a > 30% reduction in disability, opioids may be associated with a trivial increased likelihood of achieving this outcome at one month to less than six months (low certainty evidence; two trials not meta-analysed).
- ›
Opioids improved health-related quality of life (physical domain) by a trivial amount at one month to less than six months (high certainty evidence). Opioids probably made little to no difference to health-related quality of life (mental domain) at one month to less than six months (moderate certainty evidence).
- ›
It was uncertain whether opioids made little to no difference in psychological well-being at one month to less than six months (very low certainty evidence; one trial with non-significant group difference).
- ›
Three trials found similar effects of opioids in older people and younger adults. Two trials reported similar effects of opioids in groups defined by sex or race.
Harms were monitored across the trials and various adverse events were identified with the potential for serious adverse events.
- ›
Opioids may be associated with a moderate increased likelihood of treatment discontinuation and large increased likelihood of nausea and pruritus (moderate certainty evidence).
- ›
Opioids are associated with a large increased likelihood of constipation, vomiting and somnolence at one month to less than six months (high certainty evidence).
- ›
Opioids may be associated with an increased risk of serious adverse events, though the estimate was imprecise and not statistically significant (low certainty).
- ›
There was probably no difference between opioids versus placebo concerning the risk of headache (moderate certainty).
- ›
Findings for harms were generally consistent in the analyses stratified by opioid type, though mixed agents were associated with higher risks of constipation, nausea, vomiting and pruritis. There was no clear dose-response relationship with harms.
In the comparison of opioids (treatment duration < 1 month) with placebo (1 trial, 25 participants), benefit was observed for pain reduction. However, it was uncertain whether opioids offer a large reduction in pain intensity compared to placebo at < 1 month, since the certainty of evidence was very low. The trial reported no adverse events in either group.
Refer to web Annex D.D1.1 for detailed GRADE evidence profile tables for opioid analgesics.
Rationale for judgements
For all adults and older people, the GDG judged overall net benefits as small to moderate. The GDG noted that there were no trials specifically in older people, but that the age ranges in the trials included older people. The majority of the GDG judged harms to be moderate, while some members considered harms to be small, based on the evidence from included RCTs. The GDG explicitly noted that RCTs were not the appropriate study design to monitor harms and referred to indirect evidence from other sources where data relating to adverse events and serious adverse events including dependence and overdose were compelling; as a consequence, some members judged harms to be large for opioids. The GDG noted indirect evidence from the recently published OPAL placebo-controlled trial from Australia (122). That trial assessed the benefits and harms of short-term use (up to 6 weeks) of oxycodone-naloxone (up to 20 mg oxycodone per day orally) in adults with at least moderately severe acute LBP and/or neck pain, identifying no difference in pain outcomes at 6 weeks between the groups and a higher risk of misuse in the opioid group at one year (122). The GDG also noted the recent Stanford-Lancet commission on responding to the opioid crisis in North America and other nations, summarizing the scale of opioid-related morbidity and mortality in the last 25 years, in particular fatal overdose and dependence outcomes (123). The GDG referred to the WHO factsheet on opioids overdose statistics (August 2021) and WHO Guideline on community management of opioid overdose (114). The GDG judged the overall certainty of evidence to be moderate. The majority of GDG judged that the balance of benefits to harms for opioid analgesics did not favour opioids based on the evidence of harms as well as other widely reported indirect evidence highlighting adverse events, such as dependence and overdose. A minority of GDG members suggested that the balance probably favoured opioids based on the systematic review evidence, particularly the moderate certainty evidence of some benefit to pain and function between one month and less than six months. When considering the totality of direct and indirect evidence available, the GDG rationalized that while some benefits might be observed, the risk of serious harm outweighed potential benefits for opioid analgesics.
A description of GDG members’ judgements relevant to all systemic pharmacotherapies for the EtD domains of values and preferences, resource implications, equity and human rights, acceptability and feasibility is presented in the introduction to this intervention class and summarized in Annex 3 (Table C15). Table C15 also provides a summary of judgements made by the GDG for benefits, harms, balance of benefits to harms, and overall certainty for opioid analgesics.
The GDG also referred to the qualitative evidence synthesis and took the view that older people were less accepting of opioid analgesics as well as having concerns about dependence (moderate confidence). From an equity perspective, the GDG took the view that many people who take opioids might encounter stigma from health workers and others in the community. The GDG highlighted the need for awareness about stigma and strategies to address stigma relating to opioid use to ensure that all people are treated with compassion, dignity and respect.
The GDG reached a consensus decision to recommend against the use of opioid analgesics. This decision was based on the GDG’s judgement that the balance of benefits to harms did not favour opioids, particularly when considering indirect evidence of adverse events associated with long-term use, including the risk of overdose. The GDG also considered data from the qualitative evidence synthesis, where moderate confidence evidence pointed to the lack of acceptability of some medicines, especially opioids, for older people, due to their potential for dependence, fatal overdose and side-effects. This decision also reflected the GDG’s judgement that in some countries opioid analgesics were not widely available and associated with high cost. One GDG member did not agree with a conditional recommendation against opioid analgesics. The GDG also noted that WHO recommends opioids for the management of cancer pain in the WHO Guidelines for the pharmacological and radiotherapeutic management of cancer pain in adults and adolescents (124).
◯ D.1.2. Non-steroidal anti-inflammatory drugs (NSAIDs)
Recommendation
NSAIDs may be offered as part of care to adults with CPLBP
(conditional recommendation in favour of use, moderate certainty).
Remarks
NSAIDs should be used/prescribed as a short-term or intermittent treatment option, irrespective of the selected class of NSAID and of the safety profile of the person with CPLBP.
When NSAIDs are offered to people with CPLBP, they should be considered as part of a broader suite of effective treatments (i.e. not offered as a single intervention in isolation), based on a biopsychosocial assessment.
When selecting an NSAID, prescribers should consider the chemical profile of the drug (COX-2 selective vs non-selective NSAIDs), its side-effect profile, and the risk profile of the person with CPLBP, including pre-existing renal impairment, cardiovascular disease, use of other drugs and prior gastrointestinal events. In general, non-selective NSAIDs are contraindicated in people at increased risk of gastrointestinal adverse events. Very careful clinical judgements of benefit and risk should be made for people with renal impairment.
A coprescribed gastroprotective agent may be considered to reduce the risk of gastrointestinal adverse events, such as ulceration or bleeding.
For older people with multiple comorbidities such as cardiovascular diseases and renal impairment and who commonly take other medicines (e.g. anticoagulant agent), NSAIDs may be contraindicated. Prescription therefore requires careful attention to the older person’s medical history, a medicines review, close follow-up and the shortest duration of treatment.
Summary of the evidence
Quantitative review
Outcomes
In the comparison of NSAIDs (treatment duration ≥ 12 weeks) with
placebo (4 trials, 1301 participants), the certainty of evidence ranged from high to low, and benefits were observed for pain and function.
- ›
NSAIDs probably offered a small reduction in pain intensity at three to six months (moderate certainty evidence). Effects were larger in trials of the COX2-selective NSAID etoricoxib than the non-selective NSAID naproxen.
- ›
NSAIDs probably offered a small increased likelihood of experiencing ≥ 30% reduction in pain at three to six months, but the estimate was imprecise and not statistically significant (moderate certainty evidence).
- ›
NSAIDs probably offered a small improvement in function at three to six months (moderate certainty evidence). Effects were larger in trials of the COX2-selective NSAID etoricoxib than the non-selective NSAID naproxen.
- ›
NSAIDs (etoricoxib only) made little to no difference to mental quality of life at three to six months (high certainty evidence) and offered trivial improvement to physical quality of life at three to six months (high certainty evidence).
Harms were monitored across the trials and were identified for nausea. Serious adverse events were not detected.
- ›
NSAIDs may be associated with an increased likelihood of nausea (low certainty evidence), but the estimate was imprecise and not statistically significant.
- ›
NSAIDs may not increase the likelihood of serious adverse events at three to six months (low certainty evidence).
- ›
NSAIDs may make little to no difference to the likelihood of discontinuation due to adverse events at three to six months (low certainty evidence).
In the comparison of NSAIDs (treatment duration < 12 weeks) with
placebo (four trials of five non-selective NSAIDs, 449 participants), the certainty of evidence ranged from high to very low and benefits were observed for pain and function.
- ›
NSAIDs may offer a small reduction in pain intensity at < 1 month (low certainty evidence).
- ›
NSAIDs offered a small reduction in pain intensity at 1 to 3 months (high certainty evidence).
- ›
NSAIDs may offer a small improvement in function at 1 to 3 months (low certainty evidence).
Harms were monitored in the trials.
- ›
NSAIDs probably made little to no difference to the likelihood of adverse events generally (moderate certainty evidence).
- ›
NSAIDs may increase the likelihood of nausea, constipation, and dizziness (low certainty evidence, small effects, risk estimates imprecise and not statistically significant).
- ›
It was uncertain whether NSAIDs increased the likelihood of discontinuation due to adverse events (very low certainty evidence, risk estimate imprecise and not statistically significant).
- ›
NSAIDs may reduce the risk of pruritis and headache (large effects, risk estimates imprecise and not statistically significant; low certainty evidence).
It was uncertain whether NSAIDs reduced the risk vomiting and pruritis (large effects), since the certainty of the evidence was very low (risk estimates imprecise and not statistically significant).
- ›
NSAIDs may make little to no difference to somnolence (low certainty evidence).
Refer to web Annex D.D1.2 for detailed GRADE evidence profile tables for NSAIDs.
Rationale for judgements
The GDG judged overall net benefits as small to moderate. The majority of the GDG judged harms to be small to moderate, with one GDG member judging harms as uncertain. The GDG referred to the indirect evidence of harms associated with non-selective (diclofenac, etodolac, flurbiprofen, ketorolac, ibuprofen, indomethacin, meloxicam, naproxen, piroxicam, sulindac) and COX-2 selective (celecoxib, rofecoxib, valdecoxib) NSAIDs in older people (mean age 80 years) with osteoarthritis or rheumatoid arthritis derived from a health claims administrative database (125). This indirect evidence highlighted that COX-2 selective users experienced an increased risk of cardiovascular events, although in a sensitivity analysis where rofecoxib and valdecoxib users were removed the risk was no longer elevated. Rates of gastrointestinal bleeding were reduced in COX-2 selective users, irrespective of the sensitivity analysis. One GDG member commented that the standard best Clinical practice for the prescription of NSAIDs was to limit the treatment duration to less than 30 days, and that judgements about benefits and harms for a treatment duration of more than 12 weeks were, therefore, not relevant to current Clinical practice. The GDG judged the overall certainty of evidence to be moderate. The GDG judged that the balance of benefits to harms for NSAIDs favoured or probably favoured NSAIDs. One GDG member suggested that judging this balance without considering the difference in benefits and harms based on NSAID selectivity was unhelpful since selectivity is usually considered when prescribing NSAIDs (refer to Remarks). The evidence review team noted, however, that there were limited trials to perform stratified analyses by NSAID selectivity and indirect evidence would need to be considered to evaluate harms more comprehensively, and by NSAID selectivity.
A description of GDG members’ judgements relevant to all systemic pharmacotherapies for the EtD domains of values and preferences, resource implications, equity and human rights, acceptability and feasibility is presented in the introduction to this intervention class and summarized in Annex 3 (Table C16). Table C16 also provides a summary of judgements made by the GDG for benefits, harms, balance of benefits to harms, and overall certainty for NSAIDs.
The GDG reached a consensus decision to make a conditional recommendation in favour of NSAIDs for adults. This decision was based on the GDG’s judgement that the balance of benefits to harms was in favour of NSAIDs for most people. Since the mean age across the included trials was less than 60 years and there was indirect evidence of increased harm for older people, a conditional recommendation in favour that included older people specifically could not be made. The GDG noted the importance of considering NSAID selectivity and careful attention and monitoring of adverse events. Members noted the limitations in the current evidence to stratify benefits and harms by COX-2 selectivity, and stated that for this reason a recommendation based on selectivity could not be made. However, the GDG elected to include information about NSAID selectivity in the Remarks to highlight this issue as an important clinical practice consideration.
◯ D.1.3. Serotonin and noradrenaline reuptake inhibitor (SNRI) antidepressants
Recommendation
SNRI antidepressants should not be used as part of routine care for adults, including older people, with CPLBP
(conditional recommendation against use, low certainty).
Remarks
The moderate increased risk of adverse events associated with SNRI antidepressants outweigh the small benefits offered.
There are potentially serious adverse events associated with SNRI antidepressants among older people, including hyponatraemia, memory impairment, gastrointestinal events and falls, without evidence of benefit.
Summary of the evidence
Quantitative review
Outcomes
In the comparison of SNRI antidepressants (treatment duration ≥ 12 weeks) with
placebo (4 trials, 1499 participants), the certainty of evidence ranged from moderate to very low and benefits were observed for pain, while benefits for function, quality of life and psychological well-being were trivial. All trials evaluated duloxetine.
- ›
SNRI antidepressants probably offered a small reduction in pain intensity at three to six months (moderate certainty evidence). Results were similar when the analysis was restricted to a duloxetine dose of 60 mg/day and when stratified by trial quality.
- ›
SNRI antidepressants probably offered a small increase in the likelihood of reduced in pain intensity (≥ 30% improvement) at three to six months (moderate certainty evidence). Results were similar when the analysis was restricted to a duloxetine dose of 60 mg/day and when stratified by trial quality.
- ›
SNRI antidepressants probably offered a trivial improvement in function at three to six months (moderate certainty evidence). Results were similar when the analysis was restricted to a duloxetine dose of 60 mg/day and when stratified by trial quality.
- ›
SNRI antidepressants may offer a trivial improvement in quality of life and psychological well-being at three to six months (low certainty evidence).
- ›
It was uncertain whether SNRI antidepressants made little to no difference to work-related outcomes (low to very low certainty evidence).
Harms were monitored across the trials and were identified for nausea, constipation, dizziness and somnolence.
- ›
SNRI antidepressants were probably associated with a large increase in the likelihood of discontinuation due to adverse events, nausea, constipation, dizziness and somnolence (moderate certainty evidence).
- ›
It was uncertain whether SNRI antidepressants were associated with a small increased likelihood of serious adverse events (very low certainty evidence).
In the comparison of SNRI antidepressants (treatment duration < 12 weeks) with placebo (5 trials, 263 participants), the certainty of evidence was very low. It was uncertain whether SNRI antidepressants made little to no difference to pain or psychological well-being at < 1 month or at 1–3 months; or to function or quality of life at 1–3 months.
Harms were monitored in the included trials, however, the certainty of evidence was very low. It was therefore uncertain whether SNRI antidepressants:
- ›
increased the likelihood of treatment discontinuation due to adverse events (large effect) and nausea (large effect);
- ›
made little to no difference to the likelihood of adverse events, serious adverse events, constipation, dizziness, somnolence, dry mouth, pruritus or headache; or
- ›
increased the likelihood of vomiting (large effect), although the risk ratio estimates exceeded 1.0.
Refer to web Annex D.D1.3 for detailed GRADE evidence profile tables for SNRI antidepressants.
Rationale for judgements
For all adults, the GDG judged overall net benefits as trivial to small, limited mostly to a small effect on pain in the short-term. The GDG judged the harms to be small to moderate based on evidence from the RCTs. In particular, the GDG noted the moderate certainty evidence of a large increased risk of discontinuation due to adverse events, nausea, constipation, dizziness and somnolence in treatment durations of three months or more, and similarly large risks for short-duration treatment, albeit with very low certainty evidence. The GDG also noted indirect evidence of publication bias towards the positive effects of anti-depressant agents when compared with the US Food and Drug Administration analysis of the evidence (126). Considering the higher certainty of evidence for outcomes of trials with longer duration treatment and lower certainty of evidence for outcomes of trials with shorter duration therapy, the GDG largely judged the overall certainty of evidence to be low (three members judged its certainty to be moderate).
The GDG was mixed in its judgement concerning the balance of benefits to harms for SNRI antidepressants. Some members judged that the balance probably favoured SNRI antidepressants based on moderate certainty of evidence of small effects on pain with longer term therapy, while most judged that the balance probably did not favour SNRI antidepressants in view of the evidence of little to no benefits offered with short-term therapy and only trivial benefits offered in terms of function, quality of life and psychosocial well-being for long-term treatment.
A description of GDG members’ judgements relevant to all systemic pharmacotherapies for the EtD domains of values and preferences, resource implications, equity and human rights, acceptability and feasibility is presented in the introduction to this intervention class and summarized in Annex 3 (Table C17). Table C17 also provides a summary of judgements made by the GDG for benefits, harms, balance of benefits to harms, and overall certainty for SNRI antidepressants.
The GDG reached a consensus decision to recommend against SNRI antidepressants (conditional recommendation). This decision was based on the GDG’s judgement that the balance of benefits to harms was not in favour of SNRI antidepressants when considering trivial to small benefits and moderate certainty evidence of a large increased risk of adverse events. The GDG also noted an absence of trial evidence in older people while acknowledging potentially serious adverse events in this group including hyponatraemia, memory impairment, gastrointestinal events and falls, without evidence of benefit. The GDG also noted that, to date, no SNRI antidepressant has been recommended in the WHO Model List of Essential Medicines or WHO mhGAP intervention guide (106).
Three GDG members did not agree with a recommendation against of SNRI antidepressants and suggested a conditional recommendation in favour would be appropriate. The GDG also acknowledged that people who experience CPLBP often concurrently experience negative impacts on mental health, for which specific management of mental health conditions might be indicated, based on a clinical assessment. WHO provides guidance through the WHO mhGAP intervention guide (106).
◯ D.1.4. Tricyclic antidepressants
Recommendation
Tricyclic antidepressants should not be used as part of routine care for adults, including older people, with CPLBP
(conditional recommendation against use, very low certainty).
Remarks
The potential harms associated with tricyclic antidepressants outwight the trivial to small benefits offered.
There are potentially serious adverse events associated with tricyclic antidepressants among older people, including postural hypotension, falls and delirium.
Summary of the evidence
Quantitative review
Outcomes
In the comparison of tricyclic antidepressants (treatment duration ≥ 12 weeks) with
placebo (3 trials, 294 participants), the certainty of evidence ranged from moderate to very low and benefits were observed for pain and function.
- ›
Tricyclic antidepressants may be associated with a trivial to small reduction in pain intensity at three to less than six months, and a small reduction in pain intensity at six months (low certainty evidence). These estimates were not meta-analysed and were not statistically significant.
- ›
Tricyclic antidepressants may be associated with a small increase in the likelihood of experiencing reduced pain by ≥ 30% or > 75% at three to less than six months (low certainty evidence), although the relative risk estimates were imprecise (derived from a single trial) and not statistically significant.
- ›
Tricyclic antidepressants may be associated with a trivial to small improvement in function at three to less than six months (low certainty evidence) and a trivial improvement in function at six months (low certainty evidence, estimate not statistically significant).
- ›
Tricyclic antidepressants may offer a trivial benefit in terms of quality of life or psychological well-being outcomes at three to less than six months or at six months (low certainty evidence, estimates not statistically significant and derived from a single trial).
- ›
Tricyclic antidepressants may be associated with a trivial decrease in work absence at three to less than six months, although the estimate was imprecise and not statistically significant (low certainty evidence). At six months, it was uncertain whether tricyclic antidepressants increased work absence: the estimate was imprecise and not statistically significant (very low certainty evidence).
Harms were monitored across the trials and were identified for dry mouth.
- ›
Tricyclic antidepressants were probably associated with an increased likelihood of dry mouth at three to less than six months (large effect; moderate certainty).
- ›
Relative risk estimates for serious adverse events, discontinuation due to adverse events, nausea, constipation and somnolence were imprecise (very low certainty evidence).
In the comparison of tricyclic antidepressants (treatment duration <12 weeks) with
placebo (6 trials, 290 participants), the certainty of evidence was very low. It was uncertain whether tricyclic antidepressants:
- ›
offered a trivial to small reduction in pain intensity at 1 to 3 months;
- ›
offered a small benefit to psychological well-being at 1 to 3 months (small effect, 1 trial, effect estimate not statistically significant); or
- ›
made little to no difference to function or quality of life at 1 to 3 months.
Harms were monitored in the included trials; however, since the certainty of evidence was very low, it was uncertain whether tricyclic antidepressants:
- ›
increased the likelihood of constipation (large effect) and dry mouth (small effect); or
- ›
made little to no difference to the likelihood of adverse events, withdrawals due to adverse events or somnolence.
Refer to web Annex D.D1.4 for detailed GRADE evidence profile tables for tricyclic antidepressants.
Rationale for judgements
For all adults, the GDG judged overall net benefits to be trivial across critical outcomes, particularly for pain where effects were not meta-analysed and not statistically significant in long-term therapy trials, while other GDG members judged benefits to be uncertain on the basis of low to very low certainty of evidence. The GDG judged the harms to be uncertain due to the very low certainty of evidence for most harms other than dry mouth. While no trials considered older people specifically, GDG members reflected (based on their own experience) on potentially serious adverse events associated with tricyclic antidepressants in older people, including postural hypotension, falls and delirium. The GDG judged the overall certainty of evidence to be very low. The GDG judged that the balance of benefits to harms for tricyclic antidepressants did not favour or probably did not favour tricyclic antidepressants based on its assessment of benefits as being trivial or uncertain, and the potential for harms based on large risks of adverse events including dry mouth (moderate certainty) and constipation (very low certainty evidence). One GDG member judged the balance to be uncertain, given the low to very low certainty evidence for benefits and harms.
A description of GDG members’ judgements relevant to all systemic pharmacotherapies for the EtD domains of values and preferences, resource implications, equity and human rights, acceptability and feasibility is presented in the introduction to this intervention class and summarized in Annex 3 (Table C18). Table C18 also provides a summary of judgements made by the GDG for benefits, harms, balance of benefits to harms, and overall certainty for tricyclic antidepressants.
The GDG reached a consensus decision to recommend against the use of tricyclic antidepressants (conditional recommendation). This decision was based on the lack of consistent evidence for a clinically worthwhile effect on critical outcomes and considering that potential harms, particularly for older people, outweighed possible benefits. One GDG member suggested no recommendation would be appropriate given the absence of sufficient evidence to judge the balance of benefits and harms.
◯ D.1.5. Anticonvulsants
Recommendation
Anticonvulsants should not be used as part of routine cate for adults, including older people, with CPLBP
(conditional recommendation against use, very low certainty).
Remarks
The large increased risk of adverse events associated with anticonvulsants substantially outweigh the small and short-term benefits offered.
Harms associated with anticonvulsants might be more pronounced in older people. Furthermore, serious breathing difficulties have been associated with the use of gabapentin in people who have respiratory risk factors, such as chronic obstructive pulmonary disease, concomitant use of opioids and/or other central nervous system depressants, and in older people.
Summary of the evidence
Quantitative review
Outcomes
In the comparison of anticonvulsants (treatment duration ≥ 12 weeks) with placebo (1 trial, 108 participants with and without spine-related leg pain, 43% had “radicular” pain), the certainty of evidence was very low and no benefits were observed. It was uncertain whether anticonvulsants (gabapentin) made little to no difference to pain intensity, the likelihood of experiencing a reduction in pain, or psychological well-being at three to six months.
Harms were monitored in the trial; since the certainty of evidence was very low however, it was uncertain whether anticonvulsants (gabapentin):
- ›
were associated with a large increased likelihood of failing memory, dry mouth, loss of balance and concentration difficulties;
- ›
were associated with a moderate increased likelihood of asthenia (weakness and fatigue);
- ›
were associated with a moderate increased likelihood of sedation and dizziness (not statistically significant);
- ›
were associated with a small increased likelihood of discontinuation due to adverse events, although the estimate was imprecise and not statistically significant; or
- ›
made little to no difference to the likelihood of serious adverse events, constipation or nausea/vomiting.
In the comparison of anticonvulsants (treatment duration < 12 weeks) with
placebo (3 trials, 206 participants), the certainty of evidence ranged from very low to moderate. All trials excluded participants with neuropathic pain, but two trials included participants with spine-related leg pain (somatic referred leg pain). Benefits were observed for pain.
- ›
Anticonvulsants probably made little to no difference to pain intensity at less than 1 month (moderate certainty evidence).
- ›
Anticonvulsants probably offered a trivial to small reduction in pain intensity at 1 to 3 months (moderate certainty evidence).
- ›
It was uncertain whether anticonvulsants offered a trivial to small improvement in function, quality of life and psychological well-being at 1 to 3 months (very low certainty evidence).
Harms were monitored in the included trials and identified for adverse events as a group.
- ›
It was uncertain whether anticonvulsants increased the likelihood of adverse events (very low certainty evidence, large effect).
- ›
Anticonvulsants may increase the likelihood of nausea, constipation, dizziness, headache and somnolence (low certainty evidence); the risk estimates were imprecise however, and not statistically significant.
- ›
It was uncertain whether anticonvulsants made little to no difference to the likelihood of withdrawals due to adverse events or pruritus (very low certainty evidence).
Refer to web Annex D.D1.5 for detailed GRADE evidence profile tables for anticonvulsants.
Rationale for judgements
For all adults, the GDG largely judged overall net benefits as trivial, while some members judged benefits to be small or uncertain due to very low certainty of evidence. Although the GDG noted moderate certainty evidence of a benefit to pain at one to three months for short-term treatment, the size of the effect was trivial to small and there was no consistent signal of benefit across other time-points for pain. The GDG judged harms to be uncertain given the very low certainty evidence for harms outcomes, while noting signs of moderate to large effects for some adverse outcomes, including failing memory, concentration difficulties, dry mouth, loss of balance and asthenia. The GDG inferred that the risk of these harms, and potentially others, would be greater in older people, based on their experience. To support their judgements on harms, the GDG also referred to indirect evidence from the US Food and Drug Authority in December 2019 regarding the risk of serious breathing difficulties associated with the use of gabapentin in people who had respiratory risk factors such as chronic obstructive pulmonary disease, concomitant use of opioids and/or other central nervous system depressants, and in older people. The GDG also referred to other important external context and indirect evidence relating to the use of anticonvulsants in people with CPLBP, highlighting that these agents were increasingly used in practice without evidence of benefit (127), and that, despite the indication, there was no current evidence of benefit to pain and disability for gabapentin or topiramate in the context of co-existing possible or probable neuropathic spine-related leg pain (moderate to high certainty) (128). The GDG judged the overall certainty of evidence to be very low. Since there was inconsistency in estimates and certainty ratings, the lowest certainty rating (very low) was applied. The GDG judged that the balance of benefits to harms for anticonvulsants did not favour anticonvulsants based on findings of little to no clinically worthwhile benefits and the potential for harm.
A description of GDG members’ judgements relevant to all systemic pharmacotherapies for the EtD domains of values and preferences, resource implications, equity and human rights, acceptability and feasibility is presented in the introduction to this intervention class and summarized in Annex 3 (Table C19). Table C19 also provides a summary of judgements made by the GDG for benefits, harms, balance of benefits to harms, and overall certainty for anticonvulsants.
The GDG reached a consensus decision to recommend against the use of anticonvulsants (conditional recommendation) on the basis of limited clinical benefit and because harms substantially outweighed possible benefit. While ten GDG members suggested that a strong recommendation against the use of anticonvulsants would be appropriate, this proportion did not meet the threshold for a consensus decision for a strong recommendation. One GDG member suggested no recommendation would be appropriate given the absence of sufficient evidence to judge the balance of benefits and harms.
◯ D.1.6. Skeletal muscle relaxants
Recommendation
Skeletal muscle relaxants should not be used as part of routine cate for adults, including older people, with CPLBP
(conditional recommendation against use, very low certainty).
Remarks
The potential harms associated with skeletal muscle relaxants outweigh the trivial to small benefits offered.
There are potentially serious adverse events such as confusion, drowsiness, falls associated with with skeletal muscle relaxants (baclofen) among older people, especially those with chronic kidney disease.
Summary of the evidence
Quantitative review
Outcomes
In the comparison of skeletal muscle relaxants (treatment duration < 12 weeks) with
placebo (4 trials, 143 participants), the certainty of evidence ranged from very low to moderate and benefits were observed for pain and function. All trials evaluated botulinum toxin A delivered via intramuscular injection into the paravertebral muscles.
- ›
It was uncertain whether skeletal muscle relaxants increased the likelihood of a reduction in pain intensity (≥ 50% difference in pre- and post-treatment scores on a 0–10 scale) at less than 1 month (very low certainty evidence, large effect)
- ›
Skeletal muscle relaxants probably increased the likelihood of a reduction in pain intensity (≥ 50% difference in pre- and post-treatment scores on a 0–10 scale) at 1 to 4 months (moderate certainty evidence, large effect).
- ›
It was uncertain whether skeletal muscle relaxants made little to no difference to pain intensity (continuous outcome) at 1 to 4 months (very low certainty evidence).
- ›
Skeletal muscle relaxants probably increased the likelihood of improved function, defined as “significant improvement” at 1 to 4 months (moderate certainty evidence, large effect).
- ›
It was uncertain whether skeletal muscle relaxants made little to no difference to function, quality of life or work-related outcomes at 1 to 4 months (very low certainty evidence).
Harms were monitored in the trials and none was identified. Skeletal muscle relaxants made little to no difference to the likelihood of adverse events (low certainty evidence).
In the comparison of skeletal muscle relaxants (treatment duration <12 weeks) with no treatment (one trial of baclofen, 42 participants), the certainty of evidence was very low, and no differences were observed. It was uncertain whether skeletal muscle relaxants (baclofen) made little to no difference to pain intensity at less than 1 month, pain intensity at 1 to 3 months, or function at 1 to 3 months. The trial did not report on harms.
Refer to web Annex D.D1.6 for detailed GRADE evidence profile tables for skeletal muscle relaxants.
Rationale for judgements
For all adults, the GDG largely judged overall net benefits to be uncertain. This judgement was based on very low certainty evidence of little to no benefit to pain and function measured on a continuous scale in only one trial (3 trials used dichotomous outcomes) and very low certainty evidence for outcomes of treatment durations of less than 12 weeks. A minority of GDG members judged benefits to be trivial or small. The GDG judged harms to be uncertain given the very low certainty evidence for harms outcomes, while noting signs of large effects for some adverse outcomes. The GDG highlighted indirect evidence of potential harms of skeletal muscle relaxants, such as confusion, drowsiness and falls in older people, especially those with chronic kidney disease and emphasized that skeletal muscle relaxants were often contraindicated for this reason in older people (129). The GDG also referred to indirect evidence of harms from a separate systematic review examining the benefits and safety of a broad range of skeletal muscle relaxants (antispastics, non-benzodiazepine antispasmodics, benzodiazepines and miscellaneous agents) for non-specific LBP of any duration (130). Considering antispastics (baclofen, dantrolene) in people with acute LBP, moderate certainty evidence of an increased risk of any adverse event was observed (RR [95% CI]: 2.0 [1.1 to 3.8]; 2 trials, 290 participants). Furthermore, participants taking antispastics for acute LBP were more likely to discontinue treatment owing to an adverse event (RR: 34.6 [2.1 to 568.0]; 1 trial, 195 participants, very low certainty evidence). Among people with chronic LBP taking miscellaneous agents (botulinum toxin, eszopiclone), moderate certainty evidence of a possible increased risk of any adverse event was observed, although not statistically significant (RR [95% CI]: 1.5 [0.4 to 5.7]; 2 trials, 95 participants). The GDG judged the overall certainty of evidence to be very low. The GDG judged the balance of benefits to harms for skeletal muscle relaxants to be uncertain, based on very low certainty evidence for benefits and harms.
The GDG noted that there might be important resource and equity considerations relevant to some skeletal muscle relaxants, particularly those that need to be injected, such as botulinum toxin A. The GDG noted that botulinum toxin A could be very expensive in some settings which would potentially increase health inequities and costs to health systems. A description of the GDG members’ judgements relevant to all systemic pharmacotherapies for the EtD domains of values and preferences, resource implications, equity and human rights, acceptability and feasibility is presented in the introduction to this intervention class and is summarisedsummarized in Annex 3 (Table C20). Table C20 also provides a summary of judgements made by the GDG for benefits, harms, balance of benefits to harms, and overall certainty for skeletal muscle relaxants.
The GDG reached a consensus decision to recommend against skeletal muscle relaxants (conditional recommendation). This decision was based on little to no benefit for pain and function outcomes when measured on a continuous scale and indirect evidence of harms, particularly for older people. The GDG also noted that skeletal muscle relaxants are typically coprescribed with other medicines, and concluded that RCTs evaluating these agents in isolation might not accurately reflect well current Clinical practice. Two GDG members did not agree with the conditional against recommendation and judged that no recommendation would be appropriate given the absence of sufficient evidence to judge the balance of benefits and harms.
◯ D.1.7. Glucocorticoids
Recommendation
Glucocorticoids should not be used as part of routine care for adults, including older people, with CPLBP
(conditional recommendation against use, very low certainty).
Remarks
The potential harms associated with long-term use of glucocorticoids outweigh the potential benefits offered, and these harms may be more serious in older people.
In circumstances where glucocorticoids are used in the short term for people with an acute flare of CPLBP, they should be prescribed at the lowest possible dose and for the shortest possible duration with a plan for tapering to discontinuation.
Summary of the evidence
Quantitative review
Outcomes
In the comparison of systemic glucocorticoids (any treatment duration) with placebo (1 trial, 100 participants with radiculopathy), the certainty of evidence was very low, and no statistically significant differences in outcomes were observed. The trial evaluated a 10-day course of dexamethasone. It was uncertain whether glucocorticoids (dexamethasone) increased the likelihood of symptom relief – “full symptom relief or greatly improved symptoms” – at 11 days (small effect, imprecise estimate, not statistically significant).
Results for harms were imprecise, although estimates suggested there may be an increased risk of worse mood (small effect, imprecise estimate, not statistically significant) and gastrointestinal symptoms (large effect, imprecise estimate, not statistically significant), with no increased risk of hyperglycaemia (defined as a blood sugar increase of at least 2.8 mmol/L or 50 mg/dL) or weight gain ≥ 1.5 kg (very low certainty evidence).
Refer to web Annex D.D1.7 for detailed GRADE evidence profile tables for glucocorticoids.
Rationale for judgements
GDG members judged the balance of benefits to harms for glucocorticoids to be largely uncertain based on the very low certainty of evidence for benefits and harms from a single trial, while some judged that the balance was not in favour of glucocorticoids based on the potential risk of gastrointestinal symptoms and in consideration of other indirect evidence of harms, particularly with long-term use. The GDG judged the overall certainty of evidence to be very low, consistent with the systematic review team’s assessment. For all adults, the GDG judged overall net benefits to be uncertain, based on very low certainty evidence. The GDG also referred to indirect evidence from an aligned Cochrane review, which identified no convincing benefit to pain for glucocorticoids in non-radicular LBP and spinal stenosis conditions, while there was equivocal evidence suggesting that short-term use of glucocorticoids might be of benefit in people with radicular symptoms with or without acute LBP (118). The GDG judged harms to be uncertain based on evidence from a single trial. In the aligned Cochrane review, harms were inconsistently reported and no clear signals of harms were identified from the trials that monitored harms (118). GDG members noted from their clinical experience that while a short course (e.g. 7–14 days) of glucocorticoids was generally well tolerated and safe (albeit with some risk of sleeplessness, increased appetite and increased blood glucose), longer duration treatment might be associated with substantial adverse events such as increased blood glucose levels leading to diabetes, bruising, weight gain, proximal weakness, decreased bone density or fracture, and infection. The GDG rationalized that the risks of harms, particularly with prolonged use of glucocorticoids outweighed the potential benefits. The GDG also noted the mean age of participants in the single trial was 47 years, suggesting uncertainty about the benefits and harms for older people based on this direct evidence.
A description of GDG members’ judgements relevant to all systemic pharmacotherapies for the EtD domains of values and preferences, resource implications, equity and human rights, acceptability and feasibility is presented in the introduction to this intervention class and summarized in Annex 3 (Table C21). Table C21 also provides a summary of judgements made by the GDG for benefits, harms, balance of benefits to harms, and overall certainty for glucocorticoids.
The GDG reached a consensus decision to recommend against glucocorticoids (conditional recommendation). This decision was based on the very low certainty evidence for benefit and the potential for harms associated with prolonged glucocorticoid use. Two GDG members did not agree with the conditional against recommendation and judged that no recommendation would be appropriate given the absence of sufficient evidence to judge the balance of benefits and harms. The GDG noted that in circumstances where glucocorticoids are used short term for people with an acute flare of CPLBP, they ought to be prescribed at the lowest possible dose and for the shortest possible duration with a plan for tapering to discontinuation.
◯ D.1.8. Paracetamol (acetaminophen)
Summary of the evidence
There were no trials identified that evaluated the benefits or harms of paracetamol (acetaminophen) in the management of CPLBP in adults.
Rationale for judgements
Given the absence of direct evidence aligned to the PICO question, the GDG ruled not to undertake an EtD process for paracetamol and no recommendation was made. However, the GDG recognized that paracetamol (acetaminophen) is used in practice and that there may be important adverse events to consider when using this medicine. As such, key considerations were formulated to guide practice. A summary of these key considerations is also provided in Box 1 in the executive summary.
◯ D.1.9. Benzodiazepines
Summary of the evidence
There were no trials identified that evaluated the benefits or harms of benzodiazepines in the management of CPLBP in adults.
Rationale for judgements
Given the absence of direct evidence aligned to the PICO question, the GDG ruled not to undertake an EtD process for benzodiazepines and no recommendation was made. However, the GDG recognized that benzodiazepines are used in practice and that there might be important adverse events to consider when using this medicine. As such, key considerations were formulated to guide practice. A summary of these key considerations is also provided in Box 1 in the executive summary.
◯ D.2. Cannabis-related pharmaceutical preparations for therapeutic use
Definition of the intervention
Cannabis-related pharmaceutical preparations for therapeutic use (referred to here as “cannabinoids”) refer to compounds that are active in cannabis. The two principal cannabinoid compounds include tetrahydrocannabinol (THC) and cannabidiol (CBD) (134).
Summary of the evidence
There were no trials identified that evaluated the benefits or harms of cannabinoids in the management of CPLBP, based on an ongoing (living) Agency for Healthcare Research and Quality (AHRQ) systematic review of cannabis for chronic pain (134, 135).
Rationale for judgements
Given the absence of direct evidence aligned to the PICO question, the GDG ruled not to undertake an EtD process for cannabinoids and no recommendation was made. However, the GDG recognized that cannabinoids are used in practice and that there might be important adverse events to consider when using this medicine. As such, key considerations were formulated to guide practice. A summary of these key considerations is also provided in Box 1 in the executive summary.
Key considerations
Cannabinoids are not likely to be an appropriate medicine for the management of CPLBP due to a current lack of direct evidence for benefit in this condition and evidence of possible adverse events, including in older people (
136). While indirect evidence from trials in populations with other pain conditions suggests benefits to pain relief, physical functioning and sleep quality, the effects were small to very small (
137).
The potential for a small to very small clinical effect associated with cannabinoids must be considered alongside evidence for harms, which may be more severe for older people. There is indirect evidence of harms in populations with other pain conditions, including harms related to non-medicinal use (
138). These harms, for example, may include short-term adverse events such as disturbances in the level of consciousness, cognition, perception, driving skills, affect or behaviour, other psychophysiological functions and responses, and adverse physical health effects, such as stroke or acute coronary syndrome. In the longer term, adverse events may include dependence, cognitive impairment, mental disorders (psychoses, depression, anxiety and suicidal behaviour), and adverse physical health effects such as cardiovascular disease, chronic obstructive pulmonary disease and respiratory and other cancers (
138).
Regarding the importance of addressing the harms of associated with cannabinoids, the WHO mhGap 2023 guideline update currently states: “Considering high prevalence and well documented health risks associated with cannabis use, wider implementation and scale up of simple advice, brief interventions and, when appropriate, offer of referral to treatment can result in significant health gains at population level” (
139).
◯ D.3. Injectable local anaesthetics
Definition of the intervention
Injectable local anaesthetics include the subcutaneous, myofascial or intramuscular delivery of anaesthetic agents (lidocaine, articaine, bupivacaine, chloroprocaine, mepivacaine, procaine, ropivacaine and tetracaine) into local soft and/or connective tissues in the region of the lower back, between the 12th rib and gluteal fold. The injectate is delivered only to the extraspinal soft tissue and not delivered to intraspinous structures, as is the case with intradiscal, epidural, intrathecal, facet joint and nerve root injections.
Recommendation
Injectable local anaesthetics should not be used as part of routine care for adults, including older people, with CPLBP
(conditional recommendation against use, very low certainty).
Remarks
The potential harms associated with local anaesthetic injections outweigh the trivial to small benefits offered in the short-term.
There are resource considerations and potentially negative equity impacts which are relevant in some contexts.
Summary of the evidence
Quantitative review
Characteristics of the evidence
The evidence for the benefits and harms of injectable local anaesthetics into extraspinal soft and/or connective tissues of the low back region for the treatment of CPLBP was based on an update and extension to an earlier Cochrane review (140). The current synthesis comprised two trials with a total of 422 participants (mean age range 42–48 years). The two trials were undertaken in one HIC (Netherlands (Kingdom of the)) and one UMIC (Brazil). Neither study reported on older age groups separately. Both trials excluded participants with spine-related radicular leg pain (referred to as “lumbosciatic pain” and “sciatica”). Neither study reported on marginalized populations nor described the race/ethnicity of their included population. Both trials used lignocaine/lidocaine injections in the treatment arm. One trial compared a single dose (0.5% with 5 mL isotonic saline) injected over the medial part of the iliac crest using a 21-gauge needle with a placebo injection of saline only. The other trial compared three paraspinous muscle lidocaine injections (3 mL 1% diffuse lidocaine infusion) injected for three consecutive weeks with a 3.7 cm 27-gauge disposable needle to two comparison groups: one group received three “placebo injections” with dry needling (no infiltration of the muscle) and the other received no treatment.
Outcomes
In the comparison of injectable local anaesthetics with placebo (2 trials), some benefit was observed for pain. It was uncertain whether local anaesthetic injections made little to no difference to pain intensity or the likelihood of a 30% pain reduction in pain in the short term, since the certainty of evidence was very low. It was uncertain whether local anaesthetic injections increased the number of participants “feeling improved” in the short term (large effect with very wide 95% CI), since the certainty of evidence was very low. Both trials monitored harms. It was uncertain whether local anaesthetics increased adverse events, since the certainty of evidence was very low and the 95% CI for the relative risk estimate exceeded 1.0. It was uncertain whether injectable local anaesthetics made little to no difference to serious adverse events, since the certainty of evidence was very low.
In the comparison of injectable local anaesthetics with no intervention (1 trial) some benefit was observed for pain. It was uncertain whether anaesthetic injections made little to no difference in pain intensity and back-specific function in the short term, since the certainty of evidence was very low. It was uncertain whether local anaesthetic injections increased the number of participants experiencing a decrease of at least 30% in the pain visual analogue scale in the short term (small effect), since the certainty of the evidence was very low. The trial monitored harms. It was uncertain whether local anaesthetics increased adverse events, since the certainty of evidence was very low and the 95% CI for the relative risk estimate exceeded 1.0.
In the comparison of injectable local anaesthetics with usual care, no trial was identified.
Web Annex D.D3 provides the detailed GRADE evidence profile tables for the intervention, by comparator. The annex also provides a summary of qualitative evidence relevant to each of the EtD domains (discussed below) and a summary of the GDG’s judgements of these domains and of the benefits and harms of the intervention.
Qualitative review
No qualitative evidence was identified specific to injectable local anaesthetics. Some evidence from the qualitative evidence synthesis related to medicines in general and GDG members’ judgements concerning EtD considerations for systemic pharmacotherapies are applicable to injectable local anaesthetics (Section D.1
page 115)
Rationale for judgements
For all adults and older adults, the GDG judged overall net benefits to be mostly uncertain or trivial, since the certainty of evidence was very low, and no outcomes data were available beyond the short-term follow-up period. Similarly, the GDG judged harms to be mostly uncertain based on very low certainty evidence, although noted from members’ clinical experience that harms might not be negligible if injections were delivered by untrained practitioners. The GDG judged the overall certainty of evidence to be very low, consistent with the systematic review team’s assessment. The GDG judged that the balance of benefits to harms for injectable local anaesthetics was uncertain and probably did not favour the intervention. The uncertainty of judgement related to very low certainty evidence and source data from a limited number of trials, while the judgement of balance not in favour of the intervention related to the lack of evidence of clinical benefit and potential for harms.
Based on their experience, GDG members judged that values and preferences for older people relating to injectable local anaesthetics were likely to be applicable to all adults. Members judged that important or possibly important uncertainty or variability could exist since people with CPLBP were likely to value short-term relief and the potential for harms differently. One GDG member noted that in situations where people understood the evidence for benefits and harms, there would be less uncertainty or variability. The GDG judged that resource requirements would be moderate to large when considering the combined costs of the medicines, storage requirements, biohazard equipment and staff training. Furthermore, given the likely requirement for repeat injections over several consultations, the direct costs to people with CPLBP could be large. In consideration of the potentially large resource requirements and need to access specialist health practitioners in some settings, the GDG judged that health equity might be negatively impacted for some groups, especially those in rural and remote locations or where health workforce shortages are significant, while for others the impact on equity could vary. The GDG judged that the acceptability of injectable local anaesthetics was likely to vary. While some people might appreciate the opportunity for potential short-term relief of symptoms, others could be less accepting based on the very limited evidence for benefit and potential for harms, based on the included trials. From the perspective of health workers, acceptability was likely to vary based on very high variations in practice between countries. Based on members’ experience, the GDG suggested that in some countries the intervention was commonly used, while not used at all in others. The GDG judged that while delivery of the intervention in many settings was feasible, at a global level feasibility would vary due to the requirements for staff training and accessibility to trained health workers, as well as infrastructure requirements relating to intervention provision and biohazard considerations.
The GDG reached a consensus decision to recommend against injectable local anaesthetics (conditional recommendation). This decision was based on very little evidence of benefit (particularly where critical outcomes were not measured on a continuous scale), potential for harms (including consideration of harms when injections were delivered by untrained health practitioners) and the potential negative consequences for health equity and resource considerations.
Some GDGs noted that in their countries injectable local anaesthetics were commonly used and easily accessible. They also noted that local anaesthetics are commonly injected into targeted trigger points and other myofascial structures and observed that the included trials might not reflect this contemporary practice for addressing specific myofascial pain syndromes. Considering these reasons and the absence of sufficient evidence of adequate certainty to balance benefits and harms, six GDG members rationalized that no recommendation would be appropriate.
Annex 3 (Table C22) provides a summary of the judgements made by the GDG for each EtD domain.
D.4. Herbal medicines
This section provides a summary of the evidence relating to all herbal medicines generally, while subsections D.4.1–D.4.8 provide the intervention-specific evidence summaries, rationales and judgements.
Definition of the intervention
WHO defines herbal medicines as herbs, herbal materials, herbal preparations and finished herbal products that contain, as active ingredients, parts of plants, or other plant materials, or combinations of both. For the purpose of the guideline, herbal medicines were restricted to plants or parts of plants used for medicinal purposes, administered orally (ingestion) or applied topically. This definition does not include plant substances, smoked individual chemicals derived from plants, or synthetic chemicals based on plant constituents.
Summary of the evidence
Quantitative review
Characteristics of the evidence
The evidence for the benefits and harms of herbal medicines for the treatment of CPLBP was based on an update and extension to an earlier Cochrane review (141). The update included all types of herbal medicine taken orally or applied to the skin (topically). Fumigation and moxibustion treatments involving heating or smoking of herbal medicines were excluded. The update included those medicines previously evaluated in the earlier Cochrane review (141), including: Cayenne pepper [Capsicum frutescens], Devil’s claw [Harpagophytum procumbens], White willow [Salix spp.], Brazilian arnica [Solidago chilensis] and all other relevant herbal medicines for CPLBP evaluated by RCTs, including combinations of multiple herbal medicines and traditional Chinese medicines. The current synthesis comprised 12 trials with a total of 1723 participants evaluating eight different individual herbal medicines or combinations of herbal medicines, as outlined in . Further details on the formulations and dose of each medicine are provided in Annex 2
Table B2.
Summary of trials of herbal medicines and combination medicines.
Nine of the trials were performed in home settings and three in hospital settings (two inpatient, one in an outpatient setting). Seven trials were carried out in HICs (Germany: 5 trials; Israel: 2 trials), three in an UMIC (Brazil: 1 trial; Thailand: 1 trial; Türkiye: 1 trial) and the remaining two in a LMIC (Islamic Republic of Iran). All trials included participants in an age group ranging from 18 to 80 years, with a mean age range of 40–60 years. One trial included older people with a mean age of 69 years in the intervention and 68 years in the control group. Ten trials included mixed male and female populations; two trials reported no details on the sex or gender of the included participants. No trials reported on marginalized populations separately and three trials described the ethnicity of their population, reported as 100% Caucasian.
Qualitative review
View in own window
| # | Review findings: values and preferences relevant to older people | GRADE-CERQual assessment of confidence | Explanation ofconfidence assessment |
|---|
| 7 | Some older people adopted alternative forms of treatment, including traditional or herbal medicines, as a part of their self-management approach when conventional treatments failed to provide relief from their chronic LBP. Some viewed this as experimenting to find a solution. Often older people did not inform their health care provider about taking this type of treatment. | LOW | No/very minor concerns regarding methodological limitations, no/very minor concerns regarding coherence, serious concerns regarding adequacy, and serious concerns regarding relevance. |
No qualitative evidence was identified specific to herbal medicines for the following EtD domains: resource implications, equity and human rights, acceptability or feasibility.
Overall judgements and EtD considerations for herbal medicines generally
The GDG considered the EtD domains of values and preferences, resource requirements, equity and human rights, acceptability and feasibility for all herbal medicines together, since the GDG expected that judgements for these domains would not differ by herbal medicine. Therefore, this section outlines GDG judgements for those domains as they relate to herbal medicines in general. These judgements should be considered alongside the rationales for each herbal medicine
The GDG judged that values and preferences for older people relating to herbal medicines were likely to be applicable to all adults. Members varied on their judgement for values and preferences with some suggesting possibly important uncertainty or variability and others suggesting no important uncertainty or variability. The GDG suggested that the variability in judgements probably reflected different practices and beliefs concerning herbal medicines across countries and that for topical Cayenne pepper in particular, people with CPLBP needed to balance likely benefits against possible harms (e.g. skin irritation). For this reason, resource requirements and impacts on equity were also judged to vary, given that costs and equity considerations are context-specific. The GDG judged that the feasibility to provide herbal medicines would vary by setting, accepting that Devil’s claw, white willow and Cayenne pepper would be accessible in most settings, while the other medicines (and in particular, combination medicines) might be less accessible. The GDG judged that the acceptability of herbal medicines would be highly variable, based on differences in the social and cultural attitudes of people and health workers to herbal medicines. Based on their experience, GDG members noted that in some countries, such as India and Brazil, acceptability varied by geography: herbal medicines were more accepted in rural than in urban centres in these countries. GDG members also noted from their experience that herbal medicines were often more acceptable when combined with other interventions, such as massage. The GDG noted that herbal medicines might not be universally acceptable to health workers, due to concerns about the quality assurance of ingredients, consistency of ingredients in preparations and absence of manufacturing quality standards for herbal medicines in general. For these reasons, assessing the risk of harms was more difficult.
The following sections (D.4.1–D.4.8) outline the recommendation, summary of quantitative evidence and rationale and evidence considerations for each herbal medicine. Web Annex D.D4 provides the detailed GRADE evidence profile tables for each herbal medicine, by comparator.
◯ D.4.1. Topical Cayenne Pepper [Capsicum frutescens]
Recommendation
Topical Cayenne pepper [Capsicum frutescens] may be offered as part of care to adults with CPLBP, including older people
conditional recommendation in favour of use, low certainty evidence).
Remarks
When topical Cayenne pepper [Capsicum frutescens] is offered, information about possible adverse skin reactions should be outlined so that people may make informed decisions about accepting this intervention.
When topical Cayenne pepper [Capsicum frutescens] is offered to people with CPLBP, it should be considered as part of a broader suite of effective treatments (i.e. not offered as a single intervention in isolation), based on a biopsychosocial assessment.
Summary of the evidence
Quantitative review
Outcomes
In the comparison of topical Cayenne pepper [Capsicum frutescens] with placebo (3 trials), benefit was observed for pain. Cayenne pepper probably reduced pain (moderate effect) in the short term (moderate certainty evidence). Cayenne pepper may increase adverse events related to skin irritation in the short term (low certainty evidence).
In the comparison of topical Cayenne pepper [Capsicum frutescens] with no intervention, or where the effect of the intervention could be isolated, no trials were identified.
In the comparison of topical Cayenne pepper [Capsicum frutescens] with usual care, no trials were identified.
Refer to web Annex D.D4.1 for detailed GRADE evidence profile tables for Cayenne pepper [Capsicum frutescens].
Rationale for judgements
For all adults and older people, the GDG judged overall net benefits as small to moderate, although a few suggested that benefits were uncertain due to limited evidence, that pain relief was estimated as a dichotomous outcome only rather than a continuous mean difference, and that outcomes were limited to short-term follow-up only. The GDG judged non-serious adverse events to be small to moderate. The GDG judged the overall certainty of evidence to be low. Most of the GDG judged that the balance of benefits to harms for topical Cayenne pepper probably favoured the intervention based on moderate certainty evidence of short-term benefit with regard to pain reduction, and considering that adverse events were likely to be non-serious, transient and expected, based on the therapeutic mechanism of action of topical Cayenne pepper. However, some GDG members considered this balance to be uncertain.
A description of GDG members’ judgements relevant to all herbal medicines for the EtD domains of values and preferences, resource implications, equity and human rights, acceptability and feasibility is presented in the introduction to this intervention class and summarized in Annex 3 (Table C23). Table C23 also provides a summary of judgements made by the GDG for benefits, harms, balance of benefits to harms, and overall certainty for topical Cayenne pepper.
The GDG reached a consensus decision to make a conditional recommendation in favour of topical Cayenne pepper based on moderate certainty evidence of a pain-reduction benefit in the short term. The GDG noted that the harms associated with Cayenne pepper were non-serious, acknowledging that unpleasant skin reactions were likely to be transient and expected as part of the therapeutic mechanism of action. The GDG was uncertain about the benefits and harms of Cayenne pepper beyond the short term.
◯ D.4.2. Devil’s claw [Harpagophytum procumbens]
Recommendation
Devil’s claw [Harpagophytum procumbens] should not be used as part of routine care for adults, including older people, with CPLBP
(conditional recommendation against use, very low certainty).
Summary of the evidence
Quantitative review
Outcomes
In the comparison of Devil’s claw [Harpagophytum procumbens] with placebo (2 trials), possible benefit was observed for pain. Devil’s claw may reduce pain (large effect) in the short term (low certainty evidence). It was uncertain whether Devil’s claw made little to no difference to back-specific function in the short term (very low certainty). It was uncertain whether Devil’s claw made little to no difference to medication use in the short term (very low certainty). It was uncertain whether Devil’s claw made little to no difference to adverse events since the certainty of the evidence was very low.
In the comparison of Devil’s claw [Harpagophytum procumbens] with no intervention, or where the effect of the intervention could be isolated, no trials were identified.
In the comparison of Devil’s claw [Harpagophytum procumbens] with usual care, no trials were identified.
Refer to web Annex D.D4.2 for detailed GRADE evidence profile tables for Devil’s claw [Harpagophytum procumbens].
Rationale for judgements
For all adults and older people, the GDG judged overall net benefits to be mostly uncertain for Devil’s claw. In particular, the GDG was uncertain about benefits since the scale used to measure pain and function was dichotomous, creating uncertainty about absolute effect sizes, and because monitoring was limited to the short term only. Two GDG members suggested that the benefits were small to moderate for Devil’s claw based on the large relative risk estimates for pain reduction. The GDG judged harms to be uncertain for Devil’s claw due to limited reporting and very low certainty evidence.
The GDG judged the overall certainty of evidence to be low to very low. Since effects for benefits were inconsistent, an overall certainty of very low was applied. Most GDG members judged the balance of benefits to harms for Devil’s claw as uncertain due to low to very low certainty evidence from two trials.
A description of GDG members’ judgements relevant to all herbal medicines for the EtD domains of values and preferences, resource implications, equity and human rights, acceptability and feasibility is presented in the introduction to this intervention class and summarized in Annex 3 (Table C24). Table C24 also provides a summary of judgements made by the GDG for benefits, harms, balance of benefits to harms, and overall certainty for Devil’s claw.
The GDG made a consensus decision to make a conditional recommendation against Devil’s claw on the basis that benefits to pain and function were limited to low to very low certainty evidence for dichotomous outcomes in the short term. In this context, the GDG was not confident that the balance of benefit to harm was in favour of Devil’s claw. The GDG also noted that no evidence was available for older adults. Three GDG members disagreed with the judgement and instead suggested that no recommendation would be appropriate given the absence of sufficient evidence to judge the balance of benefits and harms.
◯ D.4.3. White willow [Salix spp.]
Recommendation
White willow [Salix spp.] should not be used as part of routine care for adults, including older people, with CPLBP
(conditional recommendation against use, low certainty).
Summary of the evidence
Quantitative review
Outcomes
In the comparison of white willow [Salix spp.] with placebo, two trials were identified, although outcomes could only be extracted from one trial. Possible benefits were observed for pain, function and medication use. White willow may reduce pain (large effect) in the short term (low certainty evidence) and improve back-specific function (large effect) in the short term (low certainty evidence). It was uncertain whether white willow reduced medication use (large effect) in the short term (very certainty evidence). It was uncertain whether white willow was associated with fewer adverse events in the short term compared with a placebo control group (large effect), since the certainty of evidence was very low.
In the comparison of white willow [Salix spp.] with no intervention, or where the effect of the intervention could be isolated, no trials were identified.
In the comparison of white willow [Salix spp.] with usual care, no trials were identified.
Refer to web Annex D.D4.3 for detailed GRADE evidence profile tables for white willow [Salix spp.].
Rationale for judgements
For all adults and older people, the GDG judged overall net benefits to be mostly uncertain for white willow, in particular because the scale used to measure pain and function was dichotomous, creating uncertainty about absolute effect sizes for pain and function, and because monitoring was limited to the short term only. The GDG judged harms to be uncertain due to limited reporting and very low certainty evidence. The GDG judged the overall certainty of evidence to be low to very low for white willow. Since all effects were in a consistent direction for white willow, an overall certainty rating of low was applied. Most of the GDG judged the balance of benefits to harms for white willow as uncertain due to low to very low certainty evidence from two trials, with evidence available from only one of the two trials.
A description of GDG members’ judgements relevant to all herbal medicines for the EtD domains of values and preferences, resource implications, equity and human rights, acceptability and feasibility is presented in the introduction to this intervention class and summarized in Annex 3 (Table C25). Table C25 also provides a summary of judgements made by the GDG for benefits, harms, balance of benefits to harms, and overall certainty for white willow.
The GDG made a consensus decision to make a conditional recommendation against white willow on the basis that benefits to pain and function were limited to low certainty evidence for dichotomous outcomes in the short term with implausibly large effects. In this context, the GDG was not confident that the balance of benefits to harms was in favour of white willow. The GDG also noted that no evidence was available for older adults. Three GDG members disagreed with the judgement and instead suggested that no recommendation would be appropriate given the absence of sufficient evidence to judge the balance of benefits and harms.
◯ D.4.4. Topical Brazilian arnica [Solidago chilensis]
Recommendation
No recommendation: The evidence regarding the benefits and harms of topical Brazilian Arnica [Solidago chilensis] in managing CPLBP in adults is insufficient to formulate a recommendation
(no recommendation, very low certainty evidence).
Summary of the evidence
Quantitative review
Outcomes
In the comparison of topical Brazilian arnica [Solidago chilensis] with placebo (1 trial), possible benefit was observed for pain. It was uncertain whether Brazilian arnica reduced pain (large effect) in the short term since the certainty of the evidence was very low. The trial did not report on harms
In the comparison of topical Brazilian arnica [Solidago chilensis] with no intervention, or where the effect of the intervention could be isolated, no trials were identified.
In the comparison of herbal medicines with usual care, no trials were identified.
Rationale for judgements
The GDG limited its EtD discussion to three herbal medicines only, including Cayenne pepper [Capsicum frutescens], Devil’s claw [Harpagophytum procumbens] and white willow [Salix spp.], since only these three herbal medicines were evaluated in more than one trial and had higher GRADE certainty assessments. Given that the evidence for topical Brazilian arnica [Solidago chilensis] was limited to one small trial, the GDG chose not to undertake an EtD process or to make recommendations for this herbal medicine.
◯ D.4.5. Ginger [Zingiber officinale Roscoe]
Recommendation
No recommendation: The evidence regarding the benefits and harms of Ginger [Zingiber officinale Roscoe] in managing CPLBP in adults is insufficient to formulate a recommendation
(no recommendation, very low certainty evidence).
Summary of the evidence
Quantitative review
Outcomes
In the comparison of Ginger [Zingiber officinale Roscoe] with placebo, one trial was identified, although it did not report data in a format that allowed analysis or GRADE assessment. Consequently, no GRADE evidence profile was developed for this comparison.
In the comparison of Ginger [Zingiber officinale Roscoe] with no intervention, or where the effect of the intervention could be isolated, no trials were identified.
In the comparison of Ginger [Zingiber officinale Roscoe] with usual care, no trials were identified.
Rationale for judgements
The GDG limited its EtD discussion to three herbal medicines only, including Cayenne pepper [Capsicum frutescens], Devil’s claw [Harpagophytum procumbens] and white willow [Salix spp.], since only these three herbal medicines were evaluated in more than one trial and had higher GRADE certainty assessments. Given that the evidence for Ginger [Zingiber officinale Roscoe] was limited to one trial where the data were not recuperable or able to be appraised using GRADE methods, the GDG chose not to undertake an EtD process or to make recommendations for this herbal medicine.
◯ D.4.6. Topical white lily [Lilium candidum]
Recommendation
No recommendation: The evidence regarding the benefits and harms of topical white lily [Lilium candidum] in managing CPLBP in adults is insufficient to formulate a recommendation
(no recommendation, very low certainty evidence).
Summary of the evidence
Quantitative review
Outcomes
In the comparison of topical white lily [Lilium candidum] with placebo (1 trial), possible benefits were observed for pain and function. It was uncertain whether white lily reduced pain (large effect) and improved back-specific function in the short term (large effect). The trial did not report on harms.
In the comparison of topical white lily [Lilium candidum] with no intervention, or where the effect of the intervention could be isolated, no trials were identified.
In the comparison of white lily [Lilium candidum] with usual care, no trials were identified.
Rationale for judgements
The GDG limited its EtD discussion to three herbal medicines only, including Cayenne pepper [Capsicum frutescens], Devil’s claw [Harpagophytum procumbens] and white willow [Salix spp.], since only these three herbal medicines were evaluated in more than one trial and had higher GRADE certainty assessments. Given that the evidence for topical white lily [Lilium candidum] was limited to one small trial, the GDG chose not to undertake an EtD process or to make recommendations for this herbal medicine.
◯ D.4.7. Topical combination herbal compress [Zingiber cassumunar Roxb. rhizomes, Curcuma longa L. rhizomes, Cymbopogon citratus (DC.), Stapf leaves and leaf sheaths, Croton roxburghii N.P.Balakr. leaves, Tamarindus indica L. leaves, Citrus hystrix DC. peels, Blumea balsamifera (L.) DC. leaves, Vitex trifolia L. leaves and camphor]
Recommendation
No recommendation: The evidence regarding the benefits and harms of topical combination herbal compress [Zingiber cassumunar Roxb. Rhizomes, Curcuma longa L. rhizomes, Cymbopogon citratus (DC.), Stapf leaves and leaf sheaths, Croton roxburghii N.P.Balakr. leaves, Tamarindus indica L. leaves, Citrus hystrix DC. peels, Blumea balsamifera (L.) DC. leaves, Vitex trifolia L. leaves and camphor] in managing CPLBP in adults is insufficient to formulate a recommendation
(no recommendation, very low certainty evidence).
Summary of the evidence
Quantitative review
Outcomes
In the comparison of a topical combination herbal compress with placebo, no trials were identified.
In the comparison of a topical combination herbal compress with no intervention, or where the effect of the intervention could be isolated (1 trial in older people), no benefits were observed. Massage with a combination herbal compress compared with massage alone may make little to no difference to pain, back-specific function or health-related quality of life in the short or intermediate terms (low certainty). The trial did not report on harms.
In the comparison of a combination herbal compress with usual care, no trials were identified.
Rationale for judgements
The GDG limited its EtD discussion to three herbal medicines only, including Cayenne pepper [Capsicum frutescens], Devil’s claw [Harpagophytum procumbens] and white willow [Salix spp.], since only these three herbal medicines were evaluated in more than one trial and had higher GRADE certainty assessments. Given that the evidence for a topical combination herbal compress was limited to one trial that was not placebo-controlled, the GDG chose not to undertake an EtD process or to make recommendations for this herbal medicine combination.
◯ D.4.8. Topical combination transdermal diffusional patch [Oleum thymi, Oleum limonis, Oleum nigra, Oleum rosmarini, Oleum chamomilla and Oleum lauri expressum]
Recommendation
No recommendation: The evidence regarding the benefits and harms of Topical combination transdermal diffusional patch [Oleum thymi, Oleum limonis, Oleum nigra, Oleum rosmarini, Oleum chamomilla and Oleum lauri expressum] in managing CPLBP in adults is insufficient to formulate a recommendation
(no recommendation, very low certainty evidence).
Summary of the evidence
Quantitative review
Outcomes
In the comparison of a topical combination transdermal diffusional patch with placebo (1 trial), no benefits were observed. It was uncertain whether a combination transdermal diffusion patch made little to no difference to pain or function in the short term, since the certainty of the evidence was very low. It was uncertain whether a combination transdermal diffusion patch made little to no difference to adverse events in the short term, since the certainty of evidence was very low.
In the comparison of a topical combination transdermal diffusional patch with no intervention, or where the effect of the intervention could be isolated, no trials were identified.
In the comparison of a combination transdermal diffusional patch with usual care, no trials were identified.
Rationale for judgements
The GDG limited its EtD discussion to three herbal medicines only, including Cayenne pepper [Capsicum frutescens], Devil’s claw [Harpagophytum procumbens] and white willow [Salix spp.], since only these three herbal medicines were evaluated in more than one trial and had higher GRADE certainty assessments. Given that the evidence for topical combination transdermal diffusional patch was limited to one trial, the GDG chose not to undertake an EtD process or to make recommendations for this herbal medicine combination.
