Peterson, 2018^14,15

United States

Design: SRs, RCTs, or concurrently-controlled cohort studies.

Included: 9 studies: 8 RCTs and 1 retrospective cohort study

MEDLINE (Ovid) and CINAHL were searched from 1996 to October 2016 as well as various other sources (e.g., Cochrane Database of Systematic Reviews, AHRQ, Google Scholar, hand-searching reference lists and consultation with content experts).

Total patients: 3,816 adults with chronic MSK pain (persistent for ≥ 3 months)

1 RCT (N = 851) was relevant to this review

Setting: Integrated within primary care; not to include interventions occurring entirely within intensive pain rehabilitation, specialty, or tertiary care

Intervention:

Any model with system-based mechanisms aiming to increase the uptake and organization of multimodal care (e.g., collaborative care, care management, integrated care, telecare, peer-delivered care, informal care-giving, stepped care models, and algorithms

Comparator:

Any

Outcomes:

-Effectiveness (percentages of patients obtaining reductions in pain intensity and pain-related function from BL of ≥ 30 or 50%, QoL, depression, anxiety, sleep, and opioid doses

-Unintended consequences (adverse effects on patient satisfaction, provider satisfaction, time burden, sustainability)

Follow-up:

-Duration was 12 months in the majority of studies (range: 6 to 18 months)

Cochrane, 2017¹⁶

Ireland

Design: RCTs, cluster randomized trials and quasi-RCTs.

Included: 20 studies: 19 RCTs and 1 cluster RCT

CENTRAL, MEDLINE, CINAHL, EMBASE, SCOPUS, PEDro, and OT seeker were searched between 1990 and 2016 inclusive as well as screening of existing SRs and reference lists of relevant articles.

SR was conducted in accordance with the PRISMA statement.³⁹

Total patients: 16,319 adults with MSK pain (e.g. back pain, shoulder/neck/forearm pain and knee pain) who met following criteria:

-≥ 80% of the sample were in paid employment at time of recruitment

-≤ 3 months of sickness absence from work, related to MSK pain, during the previous year

-if the sample involved participants with longer periods of sick leave, the study was included if < 20% of the sample had ≥ 3 months’ sick leave

-Trials focused on patients with inflammatory conditions (e.g., RA, AS, etc.) were excluded

-if mixed population and inflammatory conditions comprised < 10% of overall sample, the trial was included.

4 RCTs (N = 13,421) were relevant to this review

Setting: Variety of settings including hospital, community and the workplace

Intervention:

- 2 or more different components from the biopsychosocial model delivered as an integrated program by a multidisciplinary team or a single HCP; OR a physical (bio-) component and ≥ 1 psychosocial element:

-Physical/bio (participant was assessed by a HCP for causes of pain and received exercise/PT if indicated

-Psychological (education, self-management training, coping with pain and unhelpful beliefs, counseling, and CBT

-Social/occupational (workplace assessment and adaptations or barriers to work, development of communication and problem-solving skills)

Comparator:

-Usual treatment

-Wait-list

-Alternative active intervention arms

Outcomes:

-Duration of sick leave

-Time to return to work

-Pain

-Disability

-Psychological functioning

-QoL

-Fatigue

-Adverse effects

Follow-up:

-Short term (3 to 6 months)

-Long-term (≥ 12 months)

Karp, 2018¹⁷

United States

Prospective open-label RCT conducted at a single university-based late-life depression research centre

Patients (N = 227) discontinued current pharmacotherapy and entered into Phase 1 where patients received venlafaxine (≤ 150 mg/day) with supportive management. After 6 weeks, responders were required to have 2 weeks of PHQ-9 score ≤ 5 and NRS for pain improvement of ≥ 30% from BL. Non-responders (N=139) progressed to Phase 2 and were randomized to intervention and comparator treatment. Randomization was stratified by referral source (primary or other) and diagnosis of fibromyalgia (yes or no)

Inclusion criteria:

Adults (≥ 60 years of age) with diagnosis of depression and chronic LBP more days than not of at least moderate severity (i.e., score of ≥ 8 on a NRS of 0 to 20) for ≥ 3 months and failure to maintain sustained response to any physician-prescribed treatment for chronic LBP, MMSE score ≥ 80, PHQ-9 score ≥ 10

Exclusion criteria:

Medically emergent condition (e.g., vertebral fracture, infection, cancer), substance abuse, psychotic or bipolar spectrum disorders, medical instability, wheelchair-bound, or involved in legal action related to chronic LBP

Intervention:

Venlafaxine 300 mg/day or highest tolerable dose plus problem-solving therapy (i.e., stepwise approach to pain and depression management and behavioral activation) × 10 weeks (N = 68)

Comparator:

Venlafaxine 300 mg/day or highest tolerable dose plus supportive management × 10 weeks (N = 71)

Outcomes:

-Cumulative rate of response

-Back-related disability assessed with the RMDQ (MCID is improvement of 30%)

-Physical performance assessed with the SPPB (MCID is an increase of 0.5 points)

-Social problem-solving assessed with the SPSI (MCID not reported)

-Adverse events

Follow-up:

12 months

Edmond, 2018¹⁸

United States

Retrospective comparative cohort analysis based on EHR data conducted within an integrated Veterans health system comprising tertiary, inpatient, and outpatient facilities as well as community-based outpatient clinics

Veterans in a ‘pain cohort’ with moderate to severe pain (i.e. pain intensity rating of ≥ 4 on a NRS of 0 to 10) were compared with all other veterans in a ‘non-pain cohort’ (i.e., no pain or intensity ratings of 1 to 3) that received primary care treatment over a 5-year period

Inclusion criteria:

Veteran patients receiving primary care through an integrated Veterans health system who had ≥ 1 visit with a documented pain intensity rating of ≥ 4 (moderate to severe pain) over a 5-year period (July 2008 to June 2013)

Exclusion criteria:

None

Intervention:

Primary care as per SCM-PM (3-step approach as follows) (N = 31,286):

-Step 1 (primary care provider identifies and discusses patient’s pain concerns and develops a treatment plan based on self-management and primary care interventions)

-Step 2 (additional resources and collaboration such as behavioral health assessment, medication, consultation with specialists)

-Step 3 (Increased care and involvement from pain management team)

Comparator:

Primary care of ‘non-pain cohort’ (N = 157,561)

Outcomes:

-Prescription data

-Health care utilization (e.g., primary care visits, referrals, specialty pain care services)

Follow-up:

5 years

Rhon, 2018¹⁹

United States

Inclusion criteria:

Adult patients (18 to 65 years of age) seeking initial care for MSK spine or shoulder pain who received ≥ 1 visit for manual therapy.

Exclusion criteria:

NR

Intervention:

Manual therapy only (i.e., hands-on movement of joints and/or soft tissues by HCP) (N = 714)

Comparator:

Manual therapy plus opioid therapy (N = 1,162)

Outcomes:

-Healthcare utilization (e.g., medical costs and visits)

-Order and timing of care in patients who received both manual therapy and opioids

Follow-up:

1 year

Anderson, 2016²⁰

United States

Inclusion criteria:

All primary care providers who were present during the 3-year implementation of the SCM-PM (March 2011 to February 2014) and adult patients with chronic pain under the care of the providers with ≥ 1 medical visit in the previous year

Exclusion criteria:

NR

Intervention:

SCM-PM (3-step individualized approach to managing pain as described for Edmond, 2018¹⁸) (N = 25 primary care providers and N = 3,357 pre-intervention and N = 4,385 post-intervention patients)

Comparator:

NA

Outcomes:

-Provider’s knowledge and attitudes regarding pain management

-Quality of pain care

-Patient-reported pain scores

-Opioid prescribing information

-Behavioral health

-Pain referrals

Follow-up:

1 year prior to and after implementation

Comer, 2016²¹

United Kingdom

Mixed methods study with quantitative analysis and qualitative focus groups based on a community MSK service

2 separate focus groups included N = 6 patients and N = 5 PT clinicians

Inclusion criteria:

Patients who received a new MSK referral to tier 1 or tier 2 care over a 1-year period (April 2013 to March 2014)

Exclusion criteria:

Patients with referrals to allied services (e.g., domiciliary PT, podiatry service, spinal triage, community falls service) or who had not completed treatment or assessment in the MSK service

Intervention:

Community MSK service providing 3 levels of care:

-Tier 1 (standard MSK PT)

-Tier 2 (more complex assessment and PT care with an extended role or MSK physicians)

-Tier 3 (secondary care orthopedic clinics) (N = 484 patients)

Comparator:

NA

Outcomes:

-Significant predictors of requiring complex care after quantitative data analysis

-Views of patients and physiotherapy clinicians regarding case complexity

Follow-up:

1 year

Chambers, 2015²²

United Kingdom

Inclusion criteria:

Adult patients (18 to 80 years of age) with non-cardiac chest pain occurring more than once per month referred for care to clinic

Exclusion criteria:

NR

Intervention:

Stepped care, biopsychosocial management program comprising:

-Assessment only (medical therapy)

-Low intensity CBT (guided self-help)

-High intensity CBT (clinical psychologist) (N = 77)

Comparator:

NA

Outcomes:

-Change in frequency and severity of chest pain

-Negative beliefs about chest pain

-Psychosocial scores

-Healthcare resource utilization

Follow-up:

6 months

Smink, 2014²³

The Netherlands

Prospective observational cohort multicentre study

Note: Care was considered “SCS-consistent” if all modalities of the previous steps of the SCS were offered to the patient before the advanced modalities of subsequent steps. If care was inconsistent with SCS, then it was considered “SCS-inconsistent”.

Inclusion criteria:

Adult patients (≥ 18 years of age) who visited their general practitioner for a new episode of hip or knee complaints due to symptomatic hip or knee osteoarthritis

Exclusion criteria:

Joint replacement of the hip or knee, being on the wait-list for a joint replacement, language barrier, or terminal illness

Intervention:

SCS-consistent care as follows (N = 117):

-Step 1 (education, lifestyle advice, paracetamol, and glucosamine sulphate)

-Step 2 (physical therapy, dietary therapy if overweight, NSAIDs, and tramadol)

-Step 3 (multidisciplinary care, intra-articular injections, TENS)

Comparator:

SCS-inconsistent care (N = 163)

Outcomes:

-Pain and physical function assessed by the WOMAC

-Self-efficacy and active pain coping assessed with the Dutch General Self-Efficacy Scale and Pain Coping Inventory

-Healthcare resource utilization

Follow-up:

2 years

Fitzsimmons, 2014²⁴

United Kingdom

Type:

Cost-effectiveness analysis

Time Horizon:

12 months

Perspective:

UK NHS

Any intervention or comparator used to treat sciatica. The following treatments were categorized and compared in pair-wise MAs followed by MTC analysis:

-Initial treatments: Inactive control, usual care, education/advice, activity restriction, alternative or non-traditional (acupuncture), non-opioids, and opioids

-Intermediate treatments: Manipulation, traction, passive PT, active PT, biological agents

-Invasive therapies: Epidural/nerve block, disk surgery

Decision analytic model

3 main outcomes:

-Global effect (including absence of pain)

-Reduction in pain intensity (via a continuous scale)

-Improved function (based on a composite condition-specific outcome measure as continuous data using WMD and standardized MD, respectively

Clinical effect estimates derived from a SR of clinical-effectiveness and cost-effectiveness

Costs of managing patients based on expert opinion and published UK cost sources (2008–2009 prices). Drug costs were from BNF list prices. Non-traditional/alternative therapies were based on published NHS reference costs

Patients presenting with sciatica would be managed through 1 of 3 pathways: primary care, stepped approach, or immediate referral to surgery.

Base-case assumptions were that there was no reduction in utility for previous unsuccessful treatments and when individual therapies are combined in sequence, effectiveness will be as high as stand-alone treatments.

Assessment and management of pain in children and young people

Interventions include assessment and planning of care, pharmacological management, PT, psychological therapies, surgical interventions, dietary therapies, and complementary and alternative therapies.

Key questions were developed using the PICO principle. Pre-defined SIGN search strategies⁵ were used.

A search of MEDLINE, EMBASE, CINAHL, PsycINFO, and the Cochrane database of SRs was conducted for SRs (to July 14, 2015) and primary literature (to January 18, 2016).

The quality of the evidence was assessed and graded according to accredited SIGN methodology.⁵

Levels of evidence:

1++ (high quality MAs, SRs of RCTs, or RCTs with very low risk of bias

1+ (well conducted MAs, SRs of RCTs, or RCTs with low risk of bias)

1- (MAs, SRs of RCTs, or RCTs with a high risk of bias)

2++ (high quality SRs of case control or cohort studies, high quality case control or cohort studies with very low risk of confounding or bias and a high probability the relationship is causal

2+ (well conducted case control or cohort studies with a low risk of confounding or bias and a moderate probability that the relationship is causal)

2- (case control or cohort studies with a high risk of confounding or bias and a significant risk that the relationship is not causal

3 (non-analytic studies e.g., case reports, case series) 4 (expert opinion)

Assessment and management of LBP and sciatica in patients ≥ 16 years of age.

Interventions include risk assessment and stratification, imaging, self-management, exercise therapies, postural therapies, orthotics and appliances, manual therapies, acupuncture, electrotherapies, psychological interventions, pharmacological interventions, multidisciplinary biopsychosocial programs, and return to work programs.

Key questions were developed using the PICO principle and the guideline was developed in accordance with NICE guideline methodology.³¹

Systematic clinical literature searches of MEDLINE, EMBASE, the Cochrane Library, CINAHL, PsycINFO, and AMED were conducted up to December 15, 2015

Systematic economic literature searches of MEDLINE (Ovid), EMBASE (Ovid), NHS EED, HTA and HEED were conducted up to December 15, 2015.

Evidence was evaluated and presented using an adaptation of the GRADE toolbox⁹ that included risk of bias, indirectness, inconsistency, imprecision, publication bias, and other issues. Overall quality was graded as:

-High (further research is very unlikely to change confidence in the estimate of effect)

-Moderate (further research is likely to have an important impact on confidence in the estimate of effect and may change the estimate)

-Low (further research is very likely to have an important impact on the confidence in the estimate of effect and is likely to change the estimate)

-Very low (any estimate of effect is uncertain)

-Research questions and inclusion criteria included PICO components

-A comprehensive literature search was conducted. MEDLINE (Ovid) and CINAHL were searched from 1996 to October 2016 as well as numerous other sources (e.g., Cochrane Database of Systematic Reviews, AHRQ, CADTH, Google Scholar, hand-searching reference lists and consultation with content experts)

-The review was conducted in accordance with the PRISMA statement

-The review was guided by AHRQ systematic review methods but was streamlined to meet a condensed timeframe

-Study selection was performed by one reviewer and checked by another reviewer

-Data abstraction and internal validity ratings were completed by one reviewer and checked by another reviewer

-Disagreements between reviewers were resolved by consensus

-An explanation for the selection of study design was provided

-Pre-defined criteria were used to rate the internal validity of all studies: the Drug Effectiveness Review Project methods was used for RCTs and the Cochrane Risk of Bias tool was used for cohort studies

-A list of the included studies was provided (i.e., 8 RCTs and 1 retrospective cohort study were included)

-Most studies were of fair or good quality; however, 3 studies were considered of poor quality

-Characteristics of the included study were provided

-The overall risk of bias was generally considered to be low

-The strength of the evidence was graded based on the AHRQ Methods Guide for Comparative Effectiveness Reviews which incorporates 5 domains: risk of bias, consistency, directness, precision of the evidence, and reporting biases

-Strength of evidence ratings were completed by one reviewer and checked by another reviewer

-Despite strong methodology, the strength of the evidence was generally considered to be low because each intervention was supported by a single study with imprecise findings

-Conflict of interest declaration was included

-Source of funding for the systematic review was disclosed

-Although an extensive literature search was conducted, it is possible that some eligible trials were not identified (i.e., the search was limited to publications since 1996)

-Limitations among fair quality included studies were > 20% attrition and baseline differences in potential prognostic factors whereas poor quality studies had high levels of exclusions (34% to 47%) from analyses.

-Generalizability of the findings of the review may be limited due to most included studies consisting of samples from single centres and under-reporting of key patient characteristics such as pain duration, opioid use at baseline, and prevalence of common medical and mental health comorbidities

-The assessment of intervention fidelity was generally limited in most studies

-The potential confounding effects of co-interventions was largely unknown due to limited data available

-The comparator group was typically ‘usual care’ but in most studies was minimally described as regular access to primary and specialty care

-Although statistically significant differences were reported for pain and QoL outcomes, there was uncertainty regarding the clinical relevance of the reported benefits and lack of reporting on other important outcomes (e.g., depression, anxiety, sleep)

-To meet shortened timelines, the review methods were impacted by precluding a more exhaustive literature search be conducted, the use of sequential (rather than dual independent) review processes, and the scope (focus on primary care) as this limits the applicability of the findings to a broader range of specialty settings

-No meta-analyses of the data were conducted

-A list of excluded studies was not provided

-Research questions and inclusion criteria included PICO components

-A comprehensive literature search was conducted. CENTRAL, MEDLINE, CINAHL, EMBASE, SCOPUS, PEDro, and OT seeker were searched between 1990 and 2016 inclusive as well as screening of existing SRs and reference lists of relevant articles.

-The review was conducted in accordance with the PRISMA statement

-Study selection, risk of bias assessment, quality assessment, and data extraction was performed independently by two reviewers

-Disagreements between reviewers were resolved by consensus

-Risk of bias assessment was done in accordance with Cochrane guidelines: the six main domains of the risk of bias tool and the following other potential sources of bias were assessed: (1) baseline comparability of groups; (2) compliance with intervention; and (3) use of co-interventions. Each item was judged separately as being at high, low, or unclear risk of bias. 17 Studies were assigned a low quality (low risk of bias on four or less items); moderate quality (low risk of bias on 5–7 items) or high quality rating (low risk of bias on eight or more items).

-Quality of the evidence was assessed using GRADE criteria for each of the following parameters: risk of bias, inconsistency, imprecision, indirectness, and publication bias. A rating of high quality was down-graded by one level for serious concerns and by two levels for very serious concerns

-A list of the included studies was provided (i.e., 19 RCTs and 1 cluster RCT were included)

-The included studies were judged to be at low risk of bias although in some cases there was insufficient information to make a judgment

-Characteristics of the included studies were provided

-According to GRADE, the evidence was of very low to moderate quality primarily due to risk of bias and imprecise results due to small sample size; only 1 RCT was rated as being of high quality

-Meta-analysis was conducted where the homogeneity was sufficient in terms of the main components of the intervention, outcome domains, and follow-up time point

-Conflict of interest declaration was included

-Source of funding for the systematic review was disclosed

-Although an extensive literature search was conducted, it is possible that some eligible trials were not identified (i.e., the search was limited to publications since 1990)

-A number of potentially eligible studies were omitted from the review as the authors were unable to make contact with the authors of the studies to confirm that the trials met the inclusion criteria

-Lack of detail and consistency in reporting protocols and procedures and lack of agreement on defining terms such as return to work and a core set of outcomes

-Few studies provided adequate information relating to how treatment fidelity, compliance with the intervention, and use of additional healthcare resources or co-interventions were monitored.

-Limited information on the methodology used to conduct the meta-analysis was provided

-The definition of ‘biopsychosocial’ interventions was relatively broad which resulted in considerable variation in the active components included in the trials and led the authors to decide which studies could be pooled in the meta-analysis based determination of sufficient clinical homogeneity which is subjective

-The meta-analysis is also limited by combining data from a range of musculoskeletal conditions although 13 RCTs were of low back pain

-Planned subgroup analyses could not be conducted due to lack of baseline information and the small number of trials within each intervention category

-A list of excluded studies was not provided

-Objectives of the study were clearly described

-Inclusion and exclusion criteria were reported

-Baseline demographic and clinical patient characteristics were clearly presented

-The intervention, comparator, and main outcomes of the study were clearly described in the methods section

-MCID estimates for key outcomes were provided

-Study patients in the intervention and comparator groups were recruited from the same population, using the same inclusion criteria, over the same time period

-Study patients were randomized to treatment in phase 2 of the study

-Patient flow was described

-Appropriate measures of random variability were reported

-Main findings were clearly described

-Appropriate statistical tests were used to assess the study outcomes

-Actual probability values were reported

-AEs were monitored

-Certified assessors (clinicians with backgrounds in social work, psychology, education and nursing) were not blinded to treatment allocation and no attempt was made to blind those measuring the main outcomes so an important limitation is the potential for confounders

-Single centre RCT so the results may not be generalizable to other settings

-The study setting (university-based late-life depression centre) was a specialty clinic so it is uncertain if this is representative of the treatment that the majority of patients would receive

-Small sample size (N = 139)

-A placebo control was not included in the study, which is a limitation given the importance of placebo effects in trials of depression and pain interventions

-Characteristics of patients lost to follow-up were not described

-Compliance with the intervention was not reported

-Although AEs were monitored they were not reported in the study

-Objectives of the study were clearly described

-The intervention, comparator, and main outcomes of the study were clearly described in the methods section

-Study patients in the intervention and comparator groups were extracted from the same population over the same time period

-Large sample size (N = 31,286 in pain cohort)

-Baseline demographic and clinical patient characteristics were clearly presented

-Appropriate measures of random variability were reported

-Main findings were clearly described

-Appropriate statistical tests were used to assess the study outcomes

-Actual probability values were reported

-The study setting was appropriate (i.e., integrated system of care including various outpatient facilities and community-based outpatient clinics)

-Retrospective comparative cohort analysis

-Inclusion and exclusion criteria were not specified a priori

-Study patients were not randomized to treatment so an important limitation is the potential for confounders

-Study personnel were not blinded to treatment allocation

-Patient flow was not described

-Compliance with the intervention was not reported

-AEs were not reported

-The definition of the ‘pain cohort’ (i.e., veterans reporting moderate to severe pain during at least one outpatient primary care visit) did not distinguish between acute versus chronic pain, did not include veterans with pain of mild intensity, and did not provide information on pain-related diagnoses

-Objectives of the study were clearly described

-Inclusion and exclusion criteria were reported

-The intervention, comparator, and main outcomes of the study were clearly described in the methods section

-Study patients in the intervention and comparator groups were extracted from the same population over the same time period

-Large sample size (N = 1,876)

-Baseline demographic and clinical patient characteristics were clearly presented

-Adjustments were made for baseline demographic variables and comorbidites that were different between groups by including them as covariates in the generalized linear models used to compare data

-Appropriate measures of random variability were reported

-Main findings were clearly described

-Appropriate statistical tests were used to assess the study outcomes

-Actual probability values were reported

-The study setting was appropriate (i.e., healthcare claims database for all beneficiaries in the closed, single-payer US Military Health System)

-Retrospective observational cohort study

-Study patients were not randomized to treatment so an important limitation is the potential for confounders

-Study personnel were not blinded to treatment allocation

-Patient flow was not described

-Compliance with the intervention was not reported

-AEs were not reported

-Costs of care and care patterns within the military system may differ from those in civilian facilities, although the study did include the actual reimbursed cost of care for all procedures conducted in civilian hospitals by the military personnel included in the study sample—nonetheless, the results may not be generalizable to other settings

-Results rely heavily on the interpretation of the researchers and there may have been other confounding variables that could not be accounted for in the statistical models

-Utilization or timing of other interventions were not assessed in the study and could have influenced the differences seen

-Reporting on the duration of symptoms and chronicity of the problem would have assisted with interpretation of the results but these data were not available given the nature of the data

-Quality of the data in the database is limited by the quality by which it was entered in by clinicians

-Compliance with treatment was not reported

-AEs were not reported

-Objectives of the study were clearly described

-Inclusion criteria were reported (i.e., a validated identification algorithm was used to identify patients with chronic pain rather than relying on specific diagnoses)

-The intervention, comparator, and main outcomes of the study were clearly described in the methods section

-Study patients in the intervention and comparator groups were extracted from the same population over the same time period

-Large patient sample size (N = 3,357 pre-intervention and N = 4,385 post-intervention)

-Baseline demographic patient characteristics were clearly presented

-Appropriate measures of random variability were reported

-Main findings were clearly described

-Appropriate statistical tests were used to assess the study outcomes

-Actual probability values were reported

-The study setting was appropriate (i.e., 12 primary care health centres)

-Retrospective chart review

-Study patients were not randomized to treatment so an important limitation is the potential for confounders

-Selected population of patients as were cared for by 25 primary care providers within the primary health centres so the results may not be generalizable to other settings

-Study personnel were not blinded to treatment allocation

-Small sample size of primary care providers (N = 25) which was limited by provider turnover over the 3-year period which ranged from 11% to 20%

-Patient flow was not described

-Compliance with the intervention was not reported

-AEs were not reported

-Outcomes were focused on process measures and did not include other measures of patient outcomes other than pain scores

-Interventions were introduced in phases over the 3-year period which limited the ability to evaluate the impact of any one element of the intervention

-Objectives of the study were clearly described

-The authors’ definition of complex care was clearly described in the methods section

-Baseline demographic and clinical patient characteristics were clearly presented

-Appropriate measures of random variability were reported

-Main findings were clearly described

-Appropriate statistical tests were used to assess the study outcomes

-Actual probability values were reported

-Quantitative findings were largely supported by qualitative evidence from focus groups

-The study setting was appropriate (i.e., a community MSK service that receives approximately 30,000 adult patient referrals per year and delivers tier 1 and tier 2 services and acts as a triage centre for tier 3 referrals

-No prior hypothesis was stated and a large number of potential predictors were evaluated therefore results can only be considered to be exploratory

-Retrospective analysis (mixed-methods study)

-Single centre study so the findings may not be generalizable to other settings

-Study personnel were not blinded to treatment allocation so an important limitation is the potential for confounders

-No inclusion or exclusion criteria were stated, rather data was extracted from a random sample of patients who had received treatment for MSK conditions

-Main study outcomes were not clearly described in the methods section

-Study patients were not randomized to treatment

-No specific intervention or comparator was included as the aim of the study was to identify factors that are predictors of complex care that may be relevant when designing a triage tool and so were not identified a priori

-Factors such as psychosocial issues, beliefs and expectations, access to care, integration of care may be important predictors of requiring complex care but the associations were limited by the statistical power of the study and lack of data availability

-Compliance was not reported

-Potential selection bias associated with focus group participants

-AEs were not reported

-Sensitivity of the multivariable models was low

-Objectives of the study were clearly described

-Prospectively designed study

-Inclusion and exclusion criteria were reported

-Baseline demographic and clinical patient characteristics were clearly presented

-The intervention and main outcomes of the study were clearly described in the methods section

-Study patients in the cohort groups were recruited from the same population, using the same inclusion criteria, over the same time period

-Appropriate measures of random variability were reported

-Main findings were clearly described

-Appropriate statistical tests were used to assess the study outcomes

-AEs were monitored

-Pilot observational cohort study

-Study patients were not randomized to treatment so an important limitation is the potential for confounders

-Single centre study so findings may not be generalizable to other settings

-Study personnel were not blinded to treatment allocation so an important limitation is the potential for confounders

-Small sample size (N = 77)

-Selected population based on referrals to a specialty clinic so may not be representative of care available to the patient population

-Patient flow was not reported

-Length of treatment was not standardized so patients could have received treatment over different lengths of time (e.g., 3 months versus 6 months)

-Actual probabilities were not reported (i.e., only as P < 0.05 or P < 0.001)

-Compliance with treatment was not reported

-AEs were not reported

-Objectives of the study were clearly described

-Prospectively designed study

-Inclusion and exclusion criteria were reported in the methods section

-The intervention, comparator, and main outcomes of the study were clearly described in the methods section

-Study patients in the intervention and comparator groups were extracted from the same population over the same time period

-Baseline demographic and clinical patient characteristics were clearly presented

-Appropriate measures of random variability were reported

-Main findings were clearly described

-Appropriate statistical tests were used to assess the study outcomes

-Actual probability values were reported

-The study setting was appropriate (i.e., multicentre primary care general practices)

-Observational cohort study

-Study patients were not randomized to treatment so an important limitation is the potential for confounders

-Study personnel were not blinded to treatment allocation

-Statistically significant baseline imbalances were detected between the cohorts (but were adjusted for in the analyses)

-Only Dutch-speakers were enrolled in the study; therefore, cultural factors could have affected the study results and the findings may not be generalizable to other cultures

-Reasons for SCS-inconsistent care were not explored (e.g., an exercise program may not be feasible for patients with cardiovascular disease, insufficient health insurance may have precluded access to some recommended services)

-Time-frame (2 years) may have been too short to detect differences in study outcomes (e.g., change in physical function in patient with osteoarthritis)

-Patient flow was not reported

-Compliance with treatment was not reported

-AEs were not reported

-Research question and objective of the cost-effectiveness analysis were clearly stated and justified

-The perspective or viewpoint (UK NHS) was clearly stated

-The time horizon (12-months) was clearly stated

-The source of the effectiveness data was clearly stated and was based on a systematic review undertaken according to the methodology reported in the Centre for Reviews and Dissemination report and the Cochrane Handbook for Systemic Reviews of Interventions

-Clinical-effectiveness and cost-effectiveness studies were reviewed separately

-Alternatives being compared were clearly described and pair-wise meta-analyses were initially conducted followed by a MTC analysis to enable the simultaneous comparison of all treatment modalities

-A decision-analytic model was developed and clearly described and details about the model were provided

-Multiple sensitivity analyses were undertaken to address uncertainty in the modeling assumptions and inputs (e.g., best and worst case scenarios, utility values for symptoms and symptom remission, reductions in effectiveness of intermediate therapies and/or surgery, and utility achieved with symptom resolution only as a results of successive failures)

-The answer to the study question was provided and conclusions based on the data reported were clearly stated

-The majority of treatment strategies were excluded on the grounds of strict dominance (where the next regime was both more effective and less costly) and by extended dominance (where a regime has an ICER that is higher than the next more effective regime)

-The modeled time horizon was limited to 12 months with no evidence to inform the inclusion of relapse and recurrence within the model

-The perspective of the UK NHS does not allow consideration of issues relating to work and productivity and preferences of patients for symptom resolution and treatment duration

-Lack of personal social services perspective

-Currency and costs were reported in British pounds sterling (£)

-Costs associated with disk surgery were not included

-In the MTC there were a small number of relevant studies for some comparisons, statistical heterogeneity and potential inconsistency for some interventions (e.g., biological agents)

-Quantities of the resources used were not reported separately from their unit costs

-Costs were presented only in aggregated form (i.e., overall cost of treatment) and not in disaggregated form

-Costs were not discounted and no explanation was provided

-Currency (£) was not adjusted for inflation

Included 9 models of multimodal chronic pain care (mainly back pain) from 8 RCTs and 1 retrospective cohort study (N = 3,816), primarily from the USA, of fair or good quality

-Most interventions were compared to usual care (i.e., regular access to primary and specialty care)

-Risk stratification coupled with risk-matched treatment pathways using the STarT Back screening tool for back pain in 1 RCT⁹ (N = 851) demonstrated greater clinically significant improvement in pain intensity or pain-related function (≥ 30% decrease in RMDQ scores) as well as depression and QoL at 12 months than non-stratified current best practice, but did not impact anxiety or satisfaction with care

- 5 models (ESCAPE,²⁵ SEACAP,^41,42 STarT Back,⁹ SCAMP,⁴³ and SCOPE²⁶) that included algorithm-guided treatment, stepped-care analgesic optimization, or decision support based on risk stratification from prognostic screening and matched treatment pathways with pro-active ongoing treatment monitoring provided the best evidence of improvement in pain intensity and pain-related function (NNT range: 4 to 13) over 12 months and variable improvements in QoL, depression, anxiety, and sleep

-strength of evidence was generally low as each model was only supported by a single RCT with imprecise findings

Included 19 RCTs and 1 cluster RCT (N = 16,319), primarily from Europe, with interventions pooled according to main components for meta-analysis

-according to GRADE assessment, the evidence was of very low to moderate quality owing to risk of bias and imprecise results due to small sample size

-low quality evidence based on 4 RCTs and 12 months’ followup suggests that programs with a stepped care approach were more effective in promoting return to work than comparators (treatment as usual and active interventions that did not meet the biopsychosocial criterion for an intervention) (HR: 1.29 [95% CI: 1.03; 1.61]; P = 0.03)

-analyses suggested limited effectiveness in reducing sickness absences, pain reduction, or functional improvement across the intervention categories

Baseline characteristics (N=139): Mean age (SD): 70.34 (8.32); Mean duration of pain (SD): 24.02 (18.05) years; Female: 62%; White: 84%; Minor depression or dysthymia: 10.95%; Major depressive disorder: 89.05%

Depression and pain outcomes:

-Cumulative rates of response (at any time during phase) were 41.2% (VEN + PST) and 39.4% (VEN + SM)

-36.5% (95% CI: 26.1; 49.4) of VEN + PST and 37.4% (95% CI: 27.2; 49.9) of VEN + SM were responders at Week 13

-Cumulative proportion of response was not statistically significant between groups (HR: 1.07 [95% CI: 0.63; 1.80]; P = 0.81)

-No statistically significant differences in rate and time to initial response for depression and pain as independent outcomes

Functional outcomes:

-SPPB scores did not change significantly at post-randomization visits (P = 0.26)

-Change over time in SPPB scores also did not differ between groups (P = 0.88)

-Differences in decline of RMDQ scores did not differ between groups (P = 0.49)

12 month follow-up:

-PHQ-9 scores did not differ significantly between groups (P = 0.44)

-NRS for pain and RMDQ measures of pain and functional disability did not differ between groups (P = 0.74 and P = 0.21, respectively)

Change in problem-solving orientation:

-Patients receiving VEN + PST experienced greater improvement in the SPSI total score than those receiving VEN + SM (β = 5.20; SE: 2.63; 95% CI: 0.04; 0.36; z = 1.97).

-Patients who received VEN + PST did not have greater improvements on the positive problem-solving orientation subscale (β = 2.38; SE: 2.88; 95% CI: −3.26; 8.02; z = 0.83) but did have greater decreases on the negative problem-solving subscale (β = −4.87; SE: 2.29; 95% CI: −9.37; −0.39; z = −2.13)

Baseline characteristics (pain cohort) (N = 31,286): Mean age: 61.7 to 62.6 years, Female: 6.8 to 7.7%; Mean maximum pain score: 6.4 to 6.8 (on a scale of 0 to 10)

Analgesic use (pain cohort):

-Long-term opioid therapy (> 90 days) decreased from 4.2% in Year 1 to 3.3% in Year 5; P < 0.0001

-Non-opioid medication prescriptions increased from 36.7% in Year 1 to 39.8% in Year 5; P < 0.0001

-Specific medication increases (from Year 1 to Year 5) were topical analgesics (3.5% to 4.8%); NSAIDs (15.7% to 19.3%); antidepressants (5.7% to 5.9%), and anticonvulsants (10.9% to 12.8%); all P < 0.0058

-Sedatives/hypnotics decreased from 17.4% in Year 1 to 15.1% in Year 3 and increased again in Year 5 to 17.2%; P < 0.0001

Healthcare utilization (pain cohort):

-Referrals by primary care providers for any consultations increased from 43.4% in Year 1 to 51.8% in Year 5; P < 0.0001

-Increases for referrals were found for PT from 14.8% to 27.4%; P<0.0001, occupational therapy from 5.2% to 11.0%; P < 0.0001, pain medicine from 3.2% to 3.8%; P<0.0001, and neurology from 10.3% to 10.7%’ P = 0.002

-Change in referrals for mental health (from 6.0% to 5.0%) and chiropractic (from 1.4% to 3.2%) consultations were not statistically significant; P = 0.56 and P = 0.43, respectively

-Visits (≥ 1) by patients for any reason increased from 27.9% in Year 1 to 37.3% in Year 5; P < 0.0001

-Increases for visits for mental health (28.0% to 30.5%), clinical health psychology (4.7% to 7.2%), PT (16.2% to 22.9%), occupational therapy (5.6% to 10.0%), chiropractic (1.4% to 3.8%), neurology (10.0% to 13.3%), pain medicine (2.9% to 3.2%), and multimodal care (from 24.5% to 29.0%) were observed from Year 1 to Year 5; all P < 0.0019

Baseline characteristics (N = 1,876): Mean (SD) age: 36.9 (10.9) years; Female: 41.5%; Prescribed opioids: 61.9% (45.6% had opioids within first 30 days of consultation and 54.4% had opioids after 30 days); 42.8% had first MT within 30 days of consultation and 57.2% had MT after 30 days

Healthcare utilization costs:

-Mean 1-year costs in the MT-only group ($5,410 [95% CI: 5,109;5,730]) were statistically significantly lower than in the MT + opioid group ($10,498 [95% CI: 10,043; 10,973]); P < 0.05

-In patients with both MT + opioid therapy, mean 1-year costs in patients who received MT first ($10,782 [95% CI: 10,050; 11,567]) were significantly lower than in patients who received opioids first ($11,938 [95% CI: 11,272; 12,643); P = 0.030

-Patients who received MT first also had a significantly lower mean days’ supply of opioids (34.2 versus 70.9; P<0.001) and mean number of unique opioid prescriptions (3.1 versus 6.5; P < 0.001)

-Early MT (≤ 30 days from index) was associated with statistically significant lower 1-year costs for total outpatient medical visits, total outpatient medical costs, total visits for all spine or shoulder conditions, total costs for spine and shoulder care, total visits for any spine condition, total costs for all spine care, total visits for any shoulder condition, total costs for all shoulder care, and individuals with opioid prescription fills compared to delayed MT delivery (> 30 days from index); all P < 0.001

-Early opioid use (≤ 30 days from index) was associated with statistically significant lower 1-year costs for total visits for any spine condition; P = 0.048; total costs for all spine care; P < 0.001, mean unique opioid prescription fills; P<0.001, and mean days’ supply of opioids for all prescription fills; P < 0.001 compared to delayed use of opioids (> 30 days after index)

Baseline characteristics: Patients: Baseline (N = 3,357); Age (%): 40 to 49 years (31%), 50 to 59 years (26%); Female: 63%; White: 42%, Medicaid insurance: 66% and Post-Intervention (N = 4,385); Age (%) 40 to 49 years (25%); 50 to 59 years (31%); Female: 64%; Caucasian: 42%; Medicaid insurance: 64%.

Primary care providers (N = 25); Female: 56%; Caucasian: 76%; Professional degree: 68% MD/DO and 32% APRN

-Providers pain knowledge scores increased by approximately 11% from baseline; P = 0.001

-Self-rated confidence in ability to manage pain increased from 2.71 at baseline to 4.67 after implementation (on a scale of 1 to 6 where 1=strongly disagree and 6=strongly agree)

-Use of opioid treatment agreements and urine drug screens increased by 27.3% and 22.6%, respectively; P < 0.05

-Statistically significant improvements were observed in various pain care documentation data elements (e.g., documentation of pain, source or cause of pain, functional assessment, review of diagnostic tests, treatment plan, pain medication ordered, pain consult ordered, and assessment of treatment effectiveness); all P ≤ 0.025

-Referrals by the 25 primary care providers to behavioral health providers increased from 24.3% to 29.1%; P=0.009, for chiropractic consult from 0.1% to 1.1%; P=0.008, and decreased from 19.9% to 15.8% for neurologic or orthopedic surgery; P < 0.001

-There were no significant changes in opioid prescribing or chronic opioid therapy

Baseline characteristics: Patients: (N = 484); Age (%): 31 to 45 years (28.1%), 46 to 60 years (30.6%); Female: 58.1%; White: 45.7%, Working as normal: 44.6%; Incapacity/sick: 11.0%, and Retired: 23.1%

-A total of 38.0% (95% CI: 33.8; 42.4) patients required tier 2 complex care

-Final logistic regression model for predictors of requiring complex care included age group, sex, anatomic site of problem indicated by PT assessment, unclear MSK diagnosis indicated in PT assessment, and symptoms affecting sleep

-Peripheral joint problems, unclear diagnosis (atypical presentation), and symptoms affecting sleep were significant independent predictors of requiring complex care; P < 0.05

-Requirement for complex care, communication difficulties, spinal problems, level of pain (score of ≤ 7, 8, 9, or 10) were significant predictors of referral to a service other than a general practitioner; P < 0.05

Baseline characteristics: (N = 77); Mean (SD) age: 50 (10.9) years; Female: 54%; Caucasian: 47%; Chest pain present > 6 months: 79%; Atypical chest pain: 84%; Other cardiac problems: 8%, Any psychiatric disorder: 55%; Diagnosis of other physical illness (e.g., gastric, glaucoma, allergy, etc.): 70%

-Proportion of patients with daily or more often chest pain decreased from 44% at baseline to 13% at 3 months and 7% at 6 months; (P < 0.001)

-Chest pain interference reduced from 5.9 [SD 2.2] at baseline to 3.2 [SD 2.6] at 3 months and 2.6 [SD 2.1] at 6 months; P<0.001

-Depression scores measured by the PHQ-9 were 8.8 [SD 7.2] at baseline and 5.4 [SD 5.8] at 3 months and 4.4 [SD 5.0] at 6 months; P < 0.05

-Anxiety scores measured by the GAD7 decreased from 6.9 [SD6.0] at baseline to 4.6 [SD 5,0] at 3 months and 3.6 [SD:4.7] at 6 months; P < 0.05

-Impact on daily life as measured by the WSAS score decreased from 10.4 [SD 10.4] at baseline to 3.9 [SD 7.5] at 3 months and 2.5 [SD 5.7] at 6 months; P < 0.001

-Reductions in use of healthcare resources (e.g., general practitioner, emergency department, cardiologist, other hospital physician, mean number of healthcare appointments, and mean number of consultations for chest pain) were all statistically significantly reduced at 6 months compared with baseline; P < 0.05

Baseline characteristics: (N = 163 SCS-inconsistent/N = 117 SCS-consistent); Mean (SD) age: 65 (10)/62 (10) years; Female: 62%/63%; Location of OA: Hip: 49%/54% or Knee: 80%/74%

Note: Imbalances were found for number of baseline characteristics (SCS-inconsistent/SCS-consistent): Age: 65/62 (P = 0.05), Number of co-morbidities 1.2/1.0 (P = 0.04), Health insurance with additional coverage: 88%/95% (P = 0.05), Number of painful joints: 1.3/1.1 (P = 0.02)

-Over the 2-year period there were statistically significant differences between SCS-inconsistent/SCS-consistent cumulative healthcare use for the following modalities:

 -Step 1: Education (69%/100%); P < 0.01, Lifestyle advice (52%/100%); P<0.01 and Paracetamol (78%/87%); P=0.05

 -Step 2: Exercise therapy (56%/70%); P = 0.02 and Dietary Therapy (6%/15%); P = 0.02

 -Step 3: Intra-articular injections (28%/13%); P < 0.01

-There were no statistically significant differences (even after adjusting for potential confounders) between patients who received SCS-consistent or SCS-inconsistent care as follows:

-WOMAC pain score (Adjusted difference: −4.3 [95% CI: −10.3; 1.7]; P = 0.16)

-WOMAC physical function score (Adjusted difference: −1.9 [95% CI: −7.0; 3.1]; P = 0.45)

-Self-efficacy (Adjusted difference: 0.6 [95% CI: −8.3; 2.0]; P = 0.41)

-Active pain coping (Adjusted difference: 1.7 [95% CI: −1.5; 4.9); P = 0.30

Base case analysis:

-Costs for the base case were inactive control (0), usual care (£73.74), education/advice (£81), activity restriction (£70), alternative/non-traditional therapies (£70), non-opiods (£122.23), opioids (£130.26), biological agents (£1646.74), manipulation, traction, passive PT, active PT (all £349), epidural (£602.76) and disk surgery (£1433.66)

Results were expressed as incremental cost per patient with symptoms successfully resolved and incremental cost per utility gained over a 12-month period. One-way sensitivity analyses were used to address uncertainty.

-The model demonstrated that none of the strategies resulted in 100% success

-The most successful regime in the 1st treatment pathway was non-opioids with a probability of success of 0.613 (i.e., 39 patients would be unsuccessful for every 100 treated)

-The most successful strategy in the 2nd treatment pathway was non-opioids, followed by biological agents, followed by epidural/nerve block and disk surgery, with a probability of success of 0.996 (i.e., 3 patients would be unsuccessful for every 1000 treated)

-The 3rd treatment pathway of immediate surgery was not costeffective

-Compared to inactive control, the following ICERs were associated with the followed stepped approaches: treatment with non-opiods and alternative/non-traditional treatments (£999), non-opiods, alternative/non-traditional treatments, and epidural (£1992), non-opioids, alternative/non-traditional treatments, epidural, and disk surgery (£5023), and non-opioids, biological therapies, epidural and disk surgery (£388,478)

-In terms of net benefit, the stepped care approaches would be regarded as cost-effective if the ceiling ratio for an additional unit of utility gain over 12 months was < £5100 and if the ceiling ratio for each additional success was < £2500

-Sensitivity analyses identified that use of the highest cost estimates results in similar findings

-For immediate referral for surgery to be considered costeffective, the costs associated with treatment following nonopioids would have to decrease by 49% or the likelihood of success would have to increase by 10 percentage points to 0.95

A Pediatric Pain Pathway with 3 levels of intervention for the management of chronic pain in children and young people is included in the guideline:

-Level 1 (Family, Education, Healthcare with type of intervention pain education, sleep, exercise)

-Level 2 (Secondary Care: Gastroenterology, Neurology, Surgery, Pediatrics, Orthopedics, Rheumatology, Pediatric Psychology with type of intervention physiotherapy and pediatric psychology

-Level 3 (Pediatric Pain Clinic, Child and Mental Health Service, Multidisciplinary team, Rehabilitation model with type of intervention mental health, multidiscplinary rehabilitation model, pain clinic)

-Patients are fast-tracked in cases of complex regional pain syndrome, neuropathic pain, child protection concerns, life-limiting diagnosis, pain arising from medical treatment, mood disturbance, or if school attendance is affected.

Recommendations pertaining to specific treatment modalities are as follows:

“2.1 Assessment and Planning of Care

- Use of a screening tool to identify children and young people at risk of adverse outcomes due to chronic pain should be considered to aid in planning intensity and type of intervention.

- Early biopsychosocial assessment and psychological intervention should be considered, particularly where the risk of disability and distress is high.

- The potential effects (both positive and negative) of children’s interactions with family, clinicians, educators and peers on assessment and management of chronic pain should be considered. Regarding the nature of interactions with healthcare providers and clinical interventions, remote or online delivery may be considered as an alternative to face to face.

2.2 Pharmacological Management

-Pharmacological treatment should only be started after careful assessment. If being used, it should be part of a wider approach utilising supported self-management strategies within the context of a multidisciplinary approach.

-If pharmacological therapy is being used, then there should be regular review with planned reassessment of ongoing efficacy and side effects. Treatment should only be continued if benefits outweigh risks, and limited to the shortest possible duration. Review should be a minimum of once per year, to assess continued benefit in terms of pain relief and improvement in function and/or quality of life.

-Paracetamol and non-steroidal anti-inflammatory drugs (NSAIDs) should be considered in the treatment of chronic non-malignant pain in children and young people. Use should be limited to the shortest possible duration, such as during acute on chronic pain episodes.

-Topical NSAIDs should be considered for treatment of children and young people with localised, non CRPS and non-neuropathic pain.

- 5% lidocaine patches should be considered in the management of children and young people with localised neuropathic pain, particularly when aiming to improve compliance with physiotherapy regimes. They are well accepted, with a low incidence of side effects, restricted to occasional hypersensitivity reactions.

-Antiepileptic drugs should be considered as part of a multi-modal approach in the management of children and young people with neuropathic pain:

- Gabapentin should be considered as first line anti-convulsant (specialist use only). It should be used in the lowest effective dose, with ongoing monitoring for efficacy and adverse effects.

- Pregabalin should be considered as a second line anticonvulsant drug if gabapentin is not tolerated or is ineffective (specialist use only).

- Low dose amitriptyline should be considered in the treatment of children and young people with functional gastrointestinal disorders.

- Low dose amitriptyline should be considered in the treatment of children and young people with chronic daily headache, chronic widespread pain and mixed nociceptive/neuropathic back pain.

-If amitriptyline is effective but particularly sedative in an individual, nortriptyline should be considered as a less sedating alternative.

- Bisphosphonates should be considered in the management of children and young people with osteogenesis imperfecta who have bone pain.

-Intrathecal baclofen should be considered for reducing spasticity-related pain in children and young people with cerebral palsy.

-In children and young people with recurrent abdominal pain pizotifen should be considered for abdominal migraine; famotidine for dyspepsia; and peppermint oil for irritable bowel syndrome.

-Opioids and compound analgesics containing opioids are rarely indicated for chronic pain because of their adverse effect profile. Be aware of MHRA advice on codeine. Strong opioids should be used with caution and only with specialist advice or assessment.

-Use of opioids should be for as short a time as possible with regular review and monitoring of efficacy and side effects.

-The use of codeine is not recommended in children under the age of 12 (MHRA), as it can be associated with a risk of opioid toxicity and respiratory side effects. In general it should also be avoided in adolescents, particularly if they have respiratory problems and individuals known to be CYP2D6 rapid metabolisers should also avoid codeine. Caution is also needed with tramadol use due to genetic variability in metabolism, and production of active metabolites.

2.3 Physical Therapies

- Exercise should be considered as a key component of chronic pain management in children and young people.

-There should be consideration of early interventions to increase movement, physical activity and restore function.

-Exercise should be used with the aim of producing functional improvement in children and young people with CRPS. Mirror therapy should be considered.

-Exercise therapy should be considered for children and young people with Patellofemoral Pain Syndrome (PFPS) to enhance long term recovery and reduce pain.

-Relaxation and TENS are low risk interventions that should be considered for the treatment of children and young people with chronic pain.

2.4 Psychological Therapies

-Psychological interventions should be part of a multi-disciplinary approach to managing chronic pain in children and young people.

-Face-to-face psychological interventions should be delivered by suitably trained and supervised practitioners.

-Online or computerised delivery of Cognitive Behavioural Therapy (CBT) interventions should be considered if face-to-face therapy is not suitable or not available.

2.5 Surgical Interventions

-Local anaesthetic blockade or other interventions should be considered on an individual patient basis in specialist centres.

2.6 Dietary Therapies

-The use of probiotics (LGG and VSL#3) should be considered in children and young people with functional gastro-intestinal disorders.

2.7 Complementary and Alternative Therapies

-Acupuncture may be considered for managing chronic pain in children and young people, for back pain and headache. If used, efficacy should be formally assessed.

-While evidence is very limited, music therapy may be considered for children and young people with chronic migraine.”³ (pages 12–14)

Due to limitations in the evidence base, unless otherwise stated, the majority of recommendations are based on expert consensus opinion.

The level of evidence was only provided for some of the evidence considered for each recommendation. The level of evidence and quantity are summarized below for each category of recommendations:

Levels of Evidence*:

2.1 Assessment and Planning of Care:

1+, 2++, 2++, 4, 4

2.2 Pharmacological Management:

1+, 1-, 1-, 3, 3, 1-, 3, 3, 1+, 2+, 2+, 2+, 2-, 2-, 1+, 1+, 3, 1-, 3,

2.3 Physical Therapies:

1+, 1++, 1-, 2-, 4, 1+, 4

2.4 Psychological Therapies:

1+, 1+, 2++

2.5 Surgical Therapies:

1-

2.6 Dietary Therapies:

1+, 1+

2.7 Complementary and Alternative Therapies:

1-, 1-, 1+

*Definitions of levels of evidence:

1++ (high quality MAs, SRs of RCTs, or RCTs with very low risk of bias

1+ (well conducted MAs, SRs of RCTs, or RCTs with low risk of bias)

1- (MAs, SRs of RCTs, or RCTs with a high risk of bias)

2++ (high quality SRs of case control or cohort studies, high quality case control or cohort studies with very low risk of confounding or bias and a high probability the relationship is causal

2+ (well conducted case control or cohort studies with a low risk of confounding or bias and a moderate probability that the relationship is causal)

2- (case control or cohort studies with a high risk of confounding or bias and a significant risk that the relationship is not causal

3 (non-analytic studies e.g., case reports, case series)

4 (expert opinion)

An algorithm for the management of LBP and sciatica in people 16 years of age and older is included in the guideline.

“The term ‘low back pain’ is used to include any non-specific low back pain which is not due to cancer, fracture, infection or an inflammatory disease process.

1. Consider using risk stratification (for example, the STarT Back risk assessment tool) at first point of contact with a healthcare professional for each new episode of low back pain with or without sciatica to inform shared decision-making about stratified management.

2. Based on risk stratification, consider:

-simpler and less intensive support for people with low back pain with or without sciatica likely to improve quickly and have a good outcome (for example, reassurance, advice to keep active and guidance on self-management)

-more complex and intensive support for people with low back pain with or without sciatica at higher risk of a poor outcome (for example, exercise programmes with or without manual therapy or using a psychological approach).

3. Do not routinely offer imaging in a non-specialist setting for people with low back pain with or without sciatica.

4. Explain to people with low back pain with or without sciatica that if they are being referred for specialist opinion, they may not need imaging.

5. Consider imaging in specialist settings of care (for example, a musculoskeletal interface clinic or hospital) for people with low back pain with or without sciatica only if the result is likely to change management.

6. Think about alternative diagnoses when examining or reviewing people with low back pain, particularly if they develop new or changed symptoms. Exclude specific causes of low back pain, for example, cancer, infection, trauma or inflammatory disease such as spondyloarthritis. If serious underlying pathology is suspected, refer to relevant NICE guidance on:

- Metastatic spinal cord compression in adults

-Spinal injury

-Spondyloarthritis

-Suspected cancer

7. Provide people with advice and information, tailored to their needs and capabilities, to help them self-manage their low back pain with or without sciatica, at all steps of the treatment pathway. Include:

- information on the nature of low back pain and sciatica

- encouragement to continue with normal activities.

8. Consider a group exercise programme (biomechanical, aerobic, mind–body or a combination of approaches) within the NHS for people with a specific episode or flare-up of low back pain with or without sciatica. Take people’s specific needs, preferences and capabilities into account when choosing the type of exercise.

9. Do not offer belts or corsets for managing low back pain with or without sciatica.

10. Do not offer foot orthotics for managing low back pain with or without sciatica.

11. Do not offer rocker sole shoes for managing low back pain with or without sciatica.

12. Do not offer traction for managing low back pain with or without sciatica.

13. Consider manual therapy (spinal manipulation, mobilisation or soft tissue techniques such as massage) for managing low back pain with or without sciatica, but only as part of a treatment package including exercise, with or without psychological therapy.

14. Do not offer acupuncture for managing low back pain with or without sciatica.

15. Do not offer ultrasound for managing low back pain with or without sciatica.

16. Do not offer percutaneous electrical nerve simulation (PENS) for managing low back pain with or without sciatica.

17. Do not offer transcutaneous electrical nerve simulation (TENS) for managing low back pain with or without sciatica.

18. Do not offer interferential therapy for managing low back pain with or without sciatica.

19. Consider psychological therapies using a cognitive behavioural approach for managing low back pain with or without sciatica but only as part of a treatment package including exercise, with or without manual therapy (spinal manipulation, mobilisation or soft tissue techniques such as massage).

20. For recommendations on pharmacological management of sciatica, see NICE’s guideline on neuropathic pain in adults.

21. Consider oral non-steroidal anti-inflammatory drugs (NSAIDs) for managing low back pain, taking into account potential differences in gastrointestinal, liver and cardio-renal toxicity, and the person’s risk factors, including age.

22. When prescribing oral NSAIDs for low back pain, think about appropriate clinical assessment, ongoing monitoring of risk factors, and the use of gastroprotective treatment.

23. Prescribe oral NSAIDs for low back pain at the lowest effective dose for the shortest possible period of time.

24. Consider weak opioids (with or without paracetamol) for managing acute low back pain only if an NSAID is contraindicated, not tolerated or has been ineffective.

25. Do not offer paracetamol alone for managing low back pain.

26. Do not routinely offer opioids for managing acute low back pain (see recommendation 24).

27. Do not offer opioids for managing chronic low back pain.

28. Do not offer selective serotonin reuptake inhibitors, serotonin–norepinephrine reuptake inhibitors or tricyclic antidepressants for managing low back pain.

29. Do not offer anticonvulsants for managing low back pain.

30. Consider a combined physical and psychological programme, incorporating a cognitive behavioural approach (preferably in a group context that takes into account a person’s specific needs and capabilities), for people with persistent low back pain or sciatica:

- when they have significant psychosocial obstacles to recovery (for example, avoiding normal activities based on inappropriate beliefs about their condition) or

-when previous treatments have not been effective.

31. Promote and facilitate return to work or normal activities of daily living for people with low back pain with or without sciatica.

32. Do not offer spinal injections for managing low back pain.

33. Consider referral for assessment for radiofrequency denervation for people with chronic low back pain when:

-non-surgical treatment has not worked for them and

-the main source of pain is thought to come from structures supplied by the medial branch nerve and

-they have moderate or severe levels of localised back pain (rated as 5 or more on a visual analogue scale, or equivalent) at the time of referral.

34. Only perform radiofrequency denervation in people with chronic low back pain after a positive response to a diagnostic medial branch block.

35. Do not offer imaging for people with low back pain with specific facet join pain as a prerequisite for radiofrequency denervation.

36. Consider epidural injections of local anaesthetic and steroid in people with acute and severe sciatica.

37. Do not use epidural injections for neurogenic claudication in people who have central spinal canal stenosis.

38. Do not allow a person’s BMI, smoking status or psychological distress to influence the decision to refer them for a surgical opinion for sciatica.

39. Do not offer disc replacement in people with low back pain.

40. Do not offer spinal fusion for people with low back pain unless as part of a randomised controlled trial.

41. Consider spinal decompression for people with sciatica when non-surgical treatment has not improved pain or function and their radiological findings are consistent with sciatic symptoms.”² (pages 18–20)

The guideline development group considered the following factor when agreeing on the wording of the recommendations:

-The strength of the recommendation (for example the word ‘offer’ was used for strong recommendations and ‘consider’ for weaker recommendations.

-For the first and second recommendations, which are most relevant to this review, the quality of evidence for risk assessment and risk stratification was rated as low or very low quality, mainly due to risk of bias and sometimes due to imprecision.