Interventions for fatigue in peripheral neuropathy
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
The objective is to systematically review the evidence from randomised controlled trials of interventions for fatigue in peripheral neuropathy.
Background
Description of the condition
Peripheral neuropathies (PN) collectively affect about 2.4% of the population (Martyn 1998) and may be either genetic or acquired and either acute or chronic in nature. The pathological process affects peripheral nerves resulting in neurological damage to either the axon (degeneration of the central nerve fibre), the myelin (demyelination or destruction of the nerve insulating sheath), or a combination of both. Recovery may be by remyelination, where improvements in patients' function can be rapid and almost complete in some, but not in all patients. However, regeneration of damaged axons may take many months or years and recovery, if present, may be incomplete (Tamura 2007).
Common symptoms of PN include numbness (pins and needles), altered or diminished sensation, muscle weakness and autonomic dysfunction. Patients may also experience fatigue, pain, psychological dysfunction and poor social adjustment (Lennon 1993; Pfeiffer 2001). Even when patients have improvement of neurological function, residual symptoms often persist and fatigue is frequently an ongoing problem (Merkies 1999).
The experience of fatigue (or subjective fatigue) can be described as a "an overwhelming sense of tiredness, lack of energy and feeling of exhaustion" that is "not relieved by rest" and is a common sequela of chronic conditions (Bleijenberg 2003; Karlsen 1999; Krupp 2003). Physiological fatigue or "the loss of voluntary force‐producing capacity during exercise" (Bigland‐Ritchie 1978) may contribute to subjective fatigue but subjective fatigue may be experienced in the absence of physiological factors and physiological fatigue does not necessarily result in subjective fatigue.
Physical factors such as residual muscle weakness mean that people with PN have to work harder during everyday activities than healthy individuals which will feel more effortful. Axonal degeneration prevents nerve impulses from being conducted along the peripheral nerve and demyelination may lead to conduction block that may be exacerbated by muscle activity (Cappelen‐Smith 2000) leading to fatigue (Kaji 2000). Although there is no direct problem with the muscles themselves, it has been shown that people with PN may be unable to fully voluntarily activate their muscles possibly caused by dysfunction within the central nervous system (Garssen 2007; Schillings 2003). This form of physiological fatigue is known as central fatigue and could possibly be due to reduced drive from nerve cells in the brain (cortical neurons) or the effect of alterations in the number of surviving peripheral nerve axons and increased number of muscle fibres they innervate (Martinez‐Figuerora 1977; Sanders 1996). In addition, psychological factors such as lack of motivation or low mood can also affect voluntary activation of muscles (Gandevia 2001; Kent‐Braun 1993). Fear of precipitating fatigue may also contribute and could lead to an avoidance of physical activity and physical deconditioning thus further exacerbating fatigue (Moss‐Morris 2005). Environmental factors may also precipitate fatigue since many people with PN may have to utilise compensatory strategies such as orthoses or walking aids as a result of deformities or sensory deficits due to PN. Therefore, the experience of fatigue may be clearly related to known underlying factors or more commonly may be only partly explained by these. Finally, measurement of fatigue is difficult both because the subjective experience of fatigue varies between individuals and because measures of different types of fatigue are numerous.
Description of the intervention
The efficacy of interventions for fatigue in PN and other chronic conditions is limited and unclear. Interventions aim to address physical, environmental and psychological factors contributing to fatigue and include drugs, pacing and grading of physical activity, general or specific exercise, orthotics, relaxation, counselling, cognitive behavioural strategies and others.
How the intervention might work
The presumed mode of action of drug interventions for generalised fatigue such as amantadine is unknown (Pucci 2007) and one recent RCT shows no effect in inflammatory PN (Garssen 2006). It is likely that factors contributing to subjective fatigue in individual patients with PN must be clearly understood in order to target interventions effectively. For example where fatigue is due to weakness, strengthening exercise may be beneficial, but where altered mood or well being is a factor then drug, cognitive or behavioural interventions may be more relevant. It is likely that multidimensional interventions that address individuals self‐perception and health beliefs combined with strategies to increase levels of physical activity and or participation in regular exercise may also be helpful.
Why it is important to do this review
Fatigue is a frequent and often severe problem for people with PN that affects everyday activities and quality of life. Although there have been some randomised controlled trials of drug and non‐drug interventions for fatigue in PN, no systematic review is known to exist.
Objectives
The objective is to systematically review the evidence from randomised controlled trials of interventions for fatigue in peripheral neuropathy.
Methods
Criteria for considering studies for this review
Types of studies
We will include all randomised controlled trials (RCTs) or quasi‐randomised controlled trials comparing any drug or non‐drug intervention to treat fatigue associated with peripheral neuropathy with placebo, no treatment or other drug or non‐drug interventions for fatigue. Quasi‐randomised trials are those in which randomisation was intended but may be biased.
Types of participants
We will consider all trials including participants (adults or children) with fatigue associated with a diagnosis of peripheral polyneuropathy, including sensory, motor and combined sensory and motor neuropathies. Trials including mostly cases of focal disease including local entrapment neuropathies with pain as the primary presenting feature (for example cervical radiculopathy, carpal tunnel syndrome, brachial plexus neuritis etc) and poliomyelitis will not be included. The diagnosis of peripheral neuropathy offered by the authors, provided that it stipulates the presence of clinical impairment characteristic of peripheral neuropathy, will be accepted. Diagnoses dependent on symptoms suggestive of neuropathy alone or neurophysiological abnormalities in the absence of clinical signs, will not be included.
Types of interventions
We will compare trials including any form of intervention for fatigue management including either drugs, pacing and grading of physical activity, general or specific exercise, compensatory strategies such as orthotics, relaxation, counselling, cognitive and educational strategies and others with either placebo, no intervention or an alternative form of intervention for fatigue. We will also include studies where interventions for fatigue form part of a multidimensional programme.
Types of outcome measures
Primary outcomes
Our primary outcome will be fatigue severity and symptoms as measured by any validated scale evaluated at least four weeks and less than 12 weeks after commencement of the intervention. Where studies use more than one scale the preferred ranking of commonly used scales shown below will be used to identify the scale to be used in the primary outcome analysis. Suitable instruments in rank order of preference include: Fatigue Severity Scale (FSS) (Krupp 1989), Chalder Fatigue Scale (Chalder 1993), Fatigue Impact Scale or daily‐FIS ( D‐FIS) (Fisk 1994; Fisk 2002), Visual Analogue Scale of fatigue (VAS‐F (Lee 1991), Piper Fatigue Scale (PFS) (Piper 1998).
Secondary outcomes
1. Fatigue after more than 12 weeks.
2. Activity limitations after 12 or more weeks.
3. Participation restrictions after 12 or more weeks.
4. Health related quality of life after 12 or more weeks.
5. Adverse events.
Search methods for identification of studies
Electronic searches
We will search the Cochrane Neuromuscular Disease Group Trials Register for randomised trials using the following search terms; 'peripheral neuropathy', 'fatigue', and interventions for fatigue including: 'drug therapy', 'exercise' (grading, pacing, strengthening, aerobic etc), 'physical therapy', 'physiotherapy', 'physical activity', 'behavioural therapy', 'cognitive therapy', 'cognitive behavioural therapy', 'counselling', 'relaxation', 'orthotics', 'education'. We will adapt this strategy to search the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, latest issue) MEDLINE (from January 1966 to the present), EMBASE (from January 1980 to the present), LILACS (from January 1982 to the present), CINAHL (from January 1982 to the present), AMED (from January 1985 to the present) and PEDRO. For search strategies see Appendix 1, Appendix 2, Appendix 3, Appendix 4, Appendix 5 and Appendix 6.
Searching other resources
We will review the bibliographies of the randomised trials identified, contact the authors and known experts in the field and approach pharmaceutical companies to identify additional published or unpublished data.
Data collection and analysis
Selection of studies
Two independent authors will independently examine the titles and abstracts identified by the search (CMW, RCS). We will obtain English language translations where appropriate. We will retrieve all trials that are relevant and review full text publications . We will examine retrieved publications against the selection criteria and any disagreement will be resolved by consensus and if necessary through arbitration by a third author (MPG). We will only include trials with a heterogenous sample of disorders if data from participants with PN can be isolated or if at least 75% of the participants have peripheral neuropathy.
Data extraction and management
For each publication retrieved two authors will independently extract the relevant data using standardised forms. We will gather data on :
‐eligibility criteria
‐interventions
‐details of participants
‐assignment to groups
‐outcome measures
‐time at which outcomes are measured
‐sample size
‐statistical analysis.
Assessment of risk of bias in included studies
Two authors will independently assess all included studies for risk of bias. The items will be graded according to the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2008) and presented in a risk of bias table. We will assess trials for randomisation sequence generation, allocation concealment, blinding (participants, personnel and outcome assessors), incomplete outcome data, selective outcome reporting and other sources of bias.
We will then make a judgement on each of these criteria relating to the risk of bias, such that a judgement of 'yes' indicates a low risk of bias, 'no' a high risk of bias and 'unclear' unclear or unknown risk of bias.
Measures of treatment effect
Continuous data
For studies using the same outcome measures with continuous data, data will be summarised using mean differences ‐ sometimes called 'weighted mean differences' (WMD). If the different studies use the same outcome measured in different ways then the results may need to be analysed using standardised mean differences.
Dichotomous data
For studies using similar outcomes with dichotomous data we will calculate a pooled estimate of the relative risk, with a 95% confidence interval.
Unit of analysis issues
Where there are multiple intervention groups within a trial, we shall make pairwise comparisons of similar interventions or active components versus no treatment, placebo, or another intervention. Crossover trials will be analysed using generalised inverse variance analysis (GIV) where the estimated difference in the mean treatment effects and their standard errors are available for pooling with equivalent results from parallel group studies. Where this information is not available data from the first phase only will be pooled, where possible, with parallel design studies. Trials with unique designs or outcomes will be reported without a meta‐analysis. We shall exclude non‐randomised controlled studies from the analyses but these may be commented on in the Discussion.
Dealing with missing data
If participant drop out leads to missing data, we shall conduct an intention‐to‐treat analysis. We shall contact trial authors or sponsors of studies to provide missing statistics such as standard deviations. For dichotomous outcomes, we shall regard participants with missing outcome data as treatment failures and include these in the analysis. For continuous outcomes, we shall carry forward the last recorded value for participants with missing outcome data.
Assessment of heterogeneity
We will undertake meta‐analysis when studies investigate similar interventions, use similar outcome measures and include groups of participants who are clinically homogeneous. If studies are heterogeneous we will undertake a narrative review.
If studies are similar in intervention, outcome measurement and participants, possible inconsistency across studies will be assessed using the I2 statistic (Higgins 2008). If studies are heterogeneous (Q‐statistic = 0.1 and I2 value of 25% or greater) the review authors will consider conducting subgroup analysis only. If the primary studies are considered to be heterogeneous even within subgroups, narrative reviews rather than meta‐analysis will be undertaken.
Assessment of reporting biases
We will investigate associations between effect size and study precision in terms of sample size using funnel plots. Relationships observed may be due to differences between studies involving large and small samples or systematic biases, such as publication bias.
Data synthesis
For studies with a similar type of intervention or a similar active component, we shall perform a meta‐analysis to calculate a weighted treatment effect across trials with a fixed‐effect model unless heterogeneity is demonstrated. Dichotomous outcome results will be used to obtain pooled relative risks. Where it is not possible to perform a meta‐analysis we shall provide descriptive summaries of the results from each trial.
Subgroup analysis and investigation of heterogeneity
We shall perform further subgroup analysis where adequate information is given. The groups will be (i) different severity of fatigue (mild or moderate and severe), (ii) age (49 or less and 50 years of age or older), (iii) type of neuropathy (demyelinating or axonal and immune mediated or non‐immune mediated (eg diabetic neuropathy)), (iv) gender, (v) participants with or without depression and (vi) participants with or without significant weakness. Issues of cost‐effectiveness will be considered and reported in the Discussion section.
Sensitivity analysis
We will undertake sensitivity analyses to assess the robustness of review findings by investigating the impact of study quality: effects of randomisation; inadequate concealment; blinding of outcome assessors; unequal loss to follow up; failure to employ intention‐to‐treat design.
