If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author Claire L. Burton, MBChB, MMedSci, Arthritis Research UK Primary Care Centre, Research Institute for Primary Care and Health Sciences, Keele University, Keele, Staffordshire ST5 5BG, UK.
Arthritis Research UK Primary Care Center, Research Institute for Primary Care and Health Sciences, Keele University, Keele, Staffordshire, United Kingdom
Arthritis Research UK Primary Care Center, Research Institute for Primary Care and Health Sciences, Keele University, Keele, Staffordshire, United Kingdom
Arthritis Research UK Primary Care Center, Research Institute for Primary Care and Health Sciences, Keele University, Keele, Staffordshire, United Kingdom
Arthritis Research UK Primary Care Center, Research Institute for Primary Care and Health Sciences, Keele University, Keele, Staffordshire, United Kingdom
To summarize the available evidence regarding the course of symptoms and prognostic factors in patients with diagnosed carpal tunnel syndrome (CTS) who are treated conservatively.
Data Sources
Computerized databases, reference checking, and experts in the field were used to identify studies for inclusion in the review.
Study Selection
Multiple reviewers were used to identify studies which included adults (aged ≥18y) diagnosed with CTS in either a clinical setting or population setting. The study must have observed the course of CTS over at least a 6-week period in patients receiving no treatment or usual care that included conservative (nonsurgical) treatments. The design was of a longitudinal cohort study with either prospective or retrospective data collection. There were no language restrictions, and none of the research identified was only reported in abstract form.
Data Extraction
Methodological bias was assessed using the Quality in Prognosis Studies tool. A high risk of bias (predominantly relating to study attrition, confounding, and/or statistical analysis and reporting) was judged to be present in 8 studies. Designs showed wide variability with respect to characteristics of the included population, definition of CTS, assessment of prognostic factors, types of interventions provided, and types of outcome measures applied. This prevented pooled estimates from being produced.
Data Synthesis
A negative outcome at 3 years' follow-up of conservatively treated participants ranged from 23% to 89%. Four included studies observed the rate of surgical intervention after initial conservative management and found this to be 57% to 66%. Evidence regarding factors predicting the negative outcome of no treatment or conservative treatment was graded, taking into account the number of studies evaluating the factor, the methodological quality of these studies, and the consistency of the available evidence. There was 100% agreement in at least 3 cohorts with a medium or high risk of bias that symptom duration, a positive Phalen's test, and thenar wasting were associated with a negative outcome of conservative management; however, not all results were statistically significant, and hence the overall judgment remained inconclusive.
Conclusions
Results of this review should be treated with caution because of the heterogeneity of studies and the risks of bias identified. However, the course of CTS appears variable, and poor prognosis may be predicted by a longer symptom duration, a positive Phalen's test, and thenar wasting.
and the condition can be associated with hypothyroidism, diabetes, and rheumatoid arthritis, among others. CTS may present in late pregnancy but is usually transient.
Studies in different countries have reported varying results with respect to the incidence of CTS.
was age adjusted to the 2000 U.S. standard population to allow comparison with the results of a U.S.-based survey of the Rochester Epidemiology Project.
The estimated incidence of CTS in Sweden was reported as 324 per 100,000 in women compared with 542 per 100,000 in the United States, and in men, 166 per 100,000 in Sweden compared with 303 in the United States.
The explanation for variation between countries is unknown; however, suggested possibilities include differences in health care–seeking behavior and variation in etiologic factors including occupation, diabetes, and inflammatory joint disease.
The treatment of CTS is often categorized as either surgical or conservative (nonsurgical). Surgical treatment is generally recommended for those with severe CTS (ie, evidence of denervation of the median nerve), while conservative treatments are recommended for the initial management of those who have intermittent or mild symptoms or in whom surgery is contraindicated.
The U.S. standardized annual incidence of carpal tunnel release surgery per 100,000 persons was 166 in Sweden compared with 171 in the United States and, among men, 58 in Sweden compared with 96 in the United States.
Examples of conservative treatment include oral steroids, steroid injections, physical therapy, electrotherapy, night splinting, and workplace alterations.
In United Kingdom primary care, steroid injections and night splinting form the mainstay of conservative treatment options, as indicated by national care pathways (eg, National Institute for Health and Care Excellence Clinical Knowledge Summaries).
American Academy of Orthopaedic Surgeons. AAOS guideline on the treatment of carpal tunnel syndrome: 2011 report for the “re-issue” of the original guideline 2011. Available at: http://www.aaos.org/Research/guidelines/CTS_Treatment_REIssue.pdf. Accessed November 28, 2012.
conclude that patients with more severe and prolonged CTS may not benefit from extended conservative treatment. However, the authors were unable to recommend in which patients conservative treatments were unlikely to be effective.
American Academy of Orthopaedic Surgeons. AAOS guideline on the treatment of carpal tunnel syndrome: 2011 report for the “re-issue” of the original guideline 2011. Available at: http://www.aaos.org/Research/guidelines/CTS_Treatment_REIssue.pdf. Accessed November 28, 2012.
With respect to splinting, the authors conclude that there is limited evidence that night splinting is more effective than no treatment in the short-term. They do, however, suggest that more research is needed on the long-term effects of this intervention.
With regard to steroid injections, it was concluded that robust evidence demonstrates clinical improvement up to 1 month compared with placebo, but relief beyond this period has not yet been shown.
With ongoing clinical uncertainty regarding the most effective management strategy for CTS, there is a clear need for a greater understanding of the likely long-term course of CTS symptoms (overall prognosis) of the condition and patient factors that may be associated with outcome (prognostic factors).
Outcomes and predictors of surgical outcome have been well reported in the literature. However, few studies and no systematic reviews have been performed to summarize the evidence for prognosis and prognostic factors in conservatively managed disease—that is, that which can be delivered in a primary care environment. An estimate of average prognosis is required by public health policymakers in order for the population burden of a condition to be assessed. Understanding the future outcomes of patients with a particular condition in relation to current practice and even in the absence of clinical care (the natural history) is crucial because it allows the potential impact of interventions to be more fully assessed.
Such information is not only important when considering the potential benefits of interventions, but also in order to inform patients, clinicians, and policymakers of the potential harms, variations (such as underuse, overuse, misuse), and potential impact on health care efficiencies.
This systematic review and narrative synthesis initially focuses on summarizing the prognosis research regarding the general course of CTS. The “start point” of this review will be the point of diagnosis of CTS that is being treated conservatively or with no clinical treatment. The “endpoint” will vary depending on the primary study. This synthesis therefore seeks to describe the course of CTS being managed either with no intervention or with conservative approaches.
The second part of this systematic review aims to identify predictors of long-term outcome (prognostic factors) in CTS. A prognostic factor is “any measure that, among people with a given health condition (start point), is associated with a subsequent clinical outcome (endpoint).”
(p1) Prognostic factor research thus seeks to identify the predictive value of such factors.
Research of prognostic factors aims to identify features that could potentially contribute to the development of prognostic models or represent predictors of differential treatment response, which may further contribute to a stratified care approach to a condition. Prognostic factors may also represent modifiable targets for interventions and could hence lead to the development of new management strategies through an improved understanding of disease mechanisms.
Details of the protocol for this systematic review were registered on PROSPERO (CRD42013006608) and can be accessed at http://www.crd.york.ac.uk/PROSPERO/display_record.asp?ID=CRD42013006608#.VYk_RflVhBc. Eligible publications had to report the course of CTS symptoms (persistence/recovery or severity of pain or other symptoms) and/or the association between a potential prognostic factor and outcome, as well as meeting the following eligibility criteria: (1) The study included adults (aged ≥18y) diagnosed with CTS in either a clinical setting or population setting. Studies in pregnant women and in populations such as specific occupational groups were excluded. (2) The study observed the course of CTS over at least a 6-week period in patients receiving no treatment or usual care that included conservative (nonsurgical) treatments. Studies reporting risk factors for onset of CTS as opposed to predictors of outcome were excluded, as were studies investigating predictors of the effectiveness of a specific treatment (which would ideally require a review of randomized controlled trials and is planned for the future). (3) The design was of a longitudinal cohort study with either prospective or retrospective data collection. (4) There were no language restrictions, and none of the research identified was only reported in abstract form.
A systematic computerized search of the literature was conducted in MEDLINE, Embase, AMED, HMIC, PsycINFO, CINAHL, Cochrane, SCI-EXPANDED, and CPCI-S from their inception until December 2013. The MEDLINE search strategy can be found in supplemental appendix S1 (available online only at http://www.archives-pmr.org/). References of all included full-text articles were hand-searched, and the first 15 pages of Google Scholar results for “carpal tunnel syndrome” and “prognosis” were screened as a further check for relevant hits. Experts were contacted to identify any further studies or publications in the gray literature that had not been identified in the search. The titles were screened by 1 reviewer (C.B.) and abstracts were screened by 2 reviewers (C.B., L.C.), and full articles of potentially eligible studies were retrieved. Such articles were screened by the 2 reviewers independently for eligibility and included in the review if they met the prespecified criteria.
Quality assessment
All selected studies were assessed independently for quality by 2 reviewers (C.B. and L.C.) using the Quality in Prognosis Studies (QUIPS) tool.
The QUIPS tool assesses bias in the following 6 domains: (1) study participation; (2) study attrition; (3) prognostic factor measurement; (4) outcome measurement; (5) study confounding; and (6) statistical analysis and reporting. Judgments of low, moderate, or high risk of bias were made for each applicable domain using descriptors recommended by Hayden et al.
Summated scores for overall study quality are not generally recommended; however, assessment of the overall risk of bias is suggested to be useful when synthesizing existing evidence.
studies were judged to be of low overall risk of bias if all or most of the domains were judged as low risk, and studies in which all or most of the domains were judged as high risk were considered to be of high overall risk of bias. Studies with a moderate risk of bias were those with all or most of the domains being judged as moderate risk. Differences between reviewers were discussed, and a decision was made by agreement. Agreement between reviewers (C.B., L.C.) regarding the judgment of overall risk of bias was presented as a percentage of agreement.
Data extraction
Data were extracted by 1 reviewer (C.B.) and checked by another reviewer (L.C.). Data extraction included details of the study setting, population demographics, diagnostic criteria of CTS used, management approaches used, prognostic factors (type of factors and how measured), outcome measures (definition and instrument used), sample size, rate of attrition, and length of follow-up. With regard to clinical course, the percentage of patients with a negative outcome after conservative treatment or no treatment was recorded. All reported prognostic factors were listed and measures of association with their significance levels recorded.
Analysis
Results regarding the course of symptoms in patients with untreated and conservatively treated CTS were summarized narratively. Pooling of results was not possible because of heterogeneity with regard to study setting, case definition, follow-up periods, and measures of outcome. We summarized findings for the reported prognostic factors by taking into account the number of studies evaluating the factor, the risk of bias of these studies, and the consistency of the available evidence (as defined as significant association with the same direction). A level of evidence was defined for each factor, based on Sackett
Figure 1 presents a flow chart of study selection. A total of 15,572 citations were identified (6987 MEDLINE, 6445 Embase, 197 AMED, 19 HMIC, 92 PsycINFO, 707 CINAHL, 755 Cochrane, 370 SCI-EXPANDED and CPCI-S). After the removal of duplicates and a screen of the titles, 146 abstracts were screened and 42 full-text publications retrieved for further eligibility screening. Twenty-six articles were excluded for the following reasons: 1 foreign language duplicate was found; 3 studies reported conditions not specific to CTS (ie, wrist pain or unspecified entrapment neuropathies); 6 studies reported outcomes in a specific population; 4 studies reported the etiology of CTS only; 6 studies reported on outcomes of specific treatments; and 6 studies used a design other than that described in the selection criteria. Sixteen articles (reporting on 16 cohorts) met all eligibility criteria and were included in the review.
Table 2 summarizes the characteristics of the studies including the QUIPS score, study design and setting, study population, interventions used in the study, the primary outcome measure including the definition of a negative outcome, and the duration of follow-up. The table also presents the percentage of the cohort experiencing a negative outcome (eg, surgery) of conservative or no management.
Table 2Summary of study characteristics and results regarding the course of symptoms of prognostic cohort studies in CTS
Author, Year, Location
Risk of Bias (QUIPS Score)
Study Population
Interventions Provided to Entire Cohort
Primary Outcome Measure/Duration of Follow-Up
Measure of Negative Outcome of Conservative Management
Proportion of Patients Treated Conservatively Experiencing Negative Outcome
Setting: tertiary hand and upper limb center CTS diagnosis: clinical findings and electrophysiological abnormality 68% female Mean age: 49.3y N=25 patients (47 wrists) Dropout: 17%
Splint: all wrists Surgery: 27 (57%) wrists
No surgery vs surgery by 6mo 12wk, with an option to continue follow-up >6mo
Setting: surgical clinics CTS diagnosis: paresthesia involving at least 2 digits (thumb or index, middle or ring fingers) and symptom duration of at least 1mo 74% female Surgical cohort: >55y mean age ± SD, 68.0±9.1y; <55y compensation nonrecipient mean age ± SD 42.0±7.3y; compensation recipient mean age ± SD, 39.0±8.1y. Nonsurgical cohort >55y mean age ± SD, 64.0±7.0y; compensation nonrecipient mean age ± SD, 41.0±8.9y; compensation recipient mean age ± SD, 37.0±8.8y N=297 patients Dropout: 31%
Nonsurgical cohort: 34 patients received surgery at <3mo and were not included in analyses. By 30mo: Splint: 76 (94%) patients Injection: 36 (44%) patients Physical or occupational therapy: all
Change in status in symptom severity, functional limitations, and health status were recorded over time. Associations were measured for patients crossing between nonsurgical and surgical cohorts after >3mo. Follow-up took place at 6, 18, and 30mo
Would not be happy to live the rest of their lives with symptoms
Setting: electromyography laboratory CTS diagnosis: clinical findings, supported by electrophysiological abnormality 90% female Diabetic rehabilitation group mean age ± SD, 59.3±7.4y; diabetic untreated group mean age ± SD, 54.6±11.1y; idiopathic rehabilitation group mean age ± SD, 47.8±9.9y; idiopathic surgery group mean age ± SD, 49.2±9.8y N=42 patients (80 wrists) Dropout: 0 (assumed)
Treatment methods not controlled or standardized “Rehabilitation”: patients treated with splints, paraffin treatments, and/or oral nonsteroidal anti-inflammatories
Symptom severity score and functional status (Boston questionnaire translated into Turkish) Patients were followed up in the early follow-up period (3–5mo) and late follow-up period (6–12mo).
Percentage improvement in symptom severity scale Percentage improvement in function severity scale
Setting: electromyography laboratory CTS diagnosis: symptoms including hand paresthesia, numbness, and pain mainly at night. 95.8% female Surgical cure group mean age 46y (range, 24–70); unchanged/worse group 44y (range, 39–58y); nonsurgical cure group mean age 61y (range, 48–79y); worse group mean age 50y (range, 30–83y) N=165 patients (222 wrists) Dropout: 69%
Surgery: 147 (66%) wrists Nonsurgical (splint, local injection, medication, and others): 75 (34%) wrists
General patient satisfaction: complete relief; improved “much better”; improved “little”; unchanged; worsened Poorly recorded. Between 5 and 10y (mean, 5.9y after surgery)
Setting: electromyography laboratory CTS diagnosis: clinical history and examination, confirmed using American Association of Electrodiagnostic Medicine criteria and additional testing if this was normal 81.3% female Mean age: 53.6y N=115 Dropout: 14%
Clinician review of medical records and decision made as to category: resolved; improved; same; worse Follow-up took place at 3 and 6mo (limited data available)
An overall categorization was made at follow-up: cured; clear improvement; slight improvement; unchanged findings; further deterioration. These were then dichotomized so that groups 1 and 2 = cured, and 3, 4, and 5 = not cured. Follow-up was at least 2y and defined as when the patient had reached a “steady state.”
No evidence of cure
68% of patients
Treated Populations: Retrospective Follow-Up Study of a Population-Based Case Series
Setting: patients identified from the Marshfield Epidemiologic Study Area CTS diagnosis: ICD-9-CM code 354.0 and evidence of a clinical and/or electrophysiological abnormality in the records. 62% female Mean age: 62y N=425 Dropout: 0%
No surgery vs surgery and resolution of symptoms Median follow- up 1979–1983: 12.0y (5th and 95th percentiles: 10.0 and 14.8y, respectively). 1984–1988: 7.3y (5.0–9.8y)
Evidence of symptoms
1mo: 75% of patients 2y: 40% 8y: 22%
Treated Populations: Secondary Analysis of Katz et al,
Setting: surgical clinics CTS diagnosis: paresthesia involving at least 2 digits (thumb or index, middle or ring fingers) and symptom duration of at least 1mo 72% female Mean age ± SD: 43±11y N=253 patients Dropout: 20%
Surgery: 179 (71%) patients
Out of work at 18mo Questionnaires were completed at 6, 18, and 30mo.
Setting: electromyography laboratory CTS diagnosis: based on neurophysiological evaluation graded: negative, minimal, mild, moderate, severe, and extreme (Padua et al
Setting: electromyography laboratory CTS diagnosis: based on clinical diagnostic criteria proposed by the American Academy of Neurology and the American Association of Electrodiagnostic Medicine 82% female Mean age ± SD: 52.0±13.4y N=202 (267 wrists) with a further 62 (87 wrists) reevaluated by telephone Dropout: 34%
The course of untreated CTS was observed.
Electrophysiological changes, patient-reported changes, and clinical changes were used to describe if patients had improved, remained stationary, or worsened. 10–15mo
Neurophysiologic class
Symptoms Function Historic and Objective Scale Pain
Setting: electromyography laboratory CTS diagnosis: clinical findings, supported by electrophysiological abnormality N=12 Dropout not possible to determine
The course of untreated CTS was observed.
Clinical and electrophysiological changes were observed. 4–9y
Conduction studies
Marked improvement 25% (of which 100% had improvement in symptoms) Slight improvement 15% (of which 33% had worsening of clinical symptoms) No significant change 50% (of which 50% had worsening of clinical symptoms) Worsening 10% (of which 50% had worsening of clinical symptoms)
NOTE. Compensation recipient and nonrecipient indicates that the patient received compensation or did not receive compensation, respectively, following litigation proceedings.
Abbreviations: ICD-9-CM, International Classification of Diseases–9th Revision–Clinical Modifications; QuickDASH, shortened version of the Disabilities of the Arm, Shoulder and Hand questionnaire.
was a retrospective follow-up study of cases identified in the Marshfield Epidemiologic Study Area, a population-based cohort. All other studies were based in secondary or tertiary care, of which 6 were in surgical clinics and 8 in electromyography laboratories. No studies were based in primary care. The case definitions used to identify CTS differed: 6 studies used clinical features only, while the remaining 10 studies required accompanying electrophysiological abnormalities. The combination of clinical characteristics used and the electrophysiological criteria also varied between studies. The interventions used in the studies included wrist splinting (7 studies), nonsteroidal anti-inflammatory drugs (3 studies), other analgesia (2 studies), oral steroids (3 studies), local steroid injections (6 studies), and paraffin treatment (1 study). Three studies provided conservative management without specifying which mode exactly. In 4 studies,
the course of (clinically) untreated CTS was observed. In some studies, parts of the cohort were treated surgically. Their specific outcomes were not included in this review. A range of outcome measures were used: 3 studies used a surgical episode as a proxy for a negative outcome; 1 study used the shortened version of the Disabilities of the Arm, Shoulder and Hand questionnaire (QuickDASH) score; 5 used measures of global improvement; 2 used a change in symptom and function severity scores; 1 used the Historic and Objective Scale
; 1 used work absence; 2 observed electrophysiological changes; and 1 used absence of clinical contact as an indicator of recovery. The follow-up periods ranged from 12 weeks to 10 years.
Methodological quality
The results of the quality assessment are presented in table 3. In 4 studies that investigated the course of CTS symptoms only, the prognostic factor domain was not assessed. The percentage agreement between the authors (C.B., L.C.) with regard to judgment of the overall risk of bias was 75%, and 100% after discussion. Further adjudication was therefore not required.
Table 3Results of methodological assessment of prognostic cohort studies on CTS
Author, Year
Study Participation
Study Attrition
Prognostic Factor Measurement
Outcome Measurement
Study Confounding
Statistical Analysis and Reporting
Overall Risk of Bias
Studies Including an Analysis of Prognostic Factors
Eight studies were judged to have a moderate risk of bias and 8 to have a high risk of bias. The domains that carried a particularly high risk of bias across all studies were study attrition (12 studies), study confounding (10 studies), and statistical analysis and reporting (9 studies). Study attrition tended to be at high risk of bias because the response rates in several studies were low (see table 3), attempts to collect information on participants who dropped out were often lacking, reasons for loss to follow-up were rarely provided, and differences between those lost to follow-up and those actively followed up were not frequently compared. Study confounding was also a frequent finding largely because not all potential confounders were appropriately accounted for, and hence the observed associations of the potential prognostic factors with outcome were likely to be at least partly explained by other (unmeasured) factors. This was particularly true in studies using retrospectively collected data. Statistical analysis and reporting was commonly identified as being of high risk of bias because presentation of the data was frequently insufficient, and in some studies selective reporting of results was evident.
Course of CTS
For each included study, table 2 describes results regarding the course of CTS in conservatively treated or untreated patients by describing the proportion of patients who experience a negative outcome, the definition of which varied between studies (ie, persisting or worsening symptoms, progression to surgery, or work absence because of CTS). Table 4 further summarizes results regarding the course of CTS in terms of the percentage of patients reporting a negative outcome for different follow-up time points.
Table 4Course of CTS in conservatively treated or untreated patients
NOTE. The percentages shown are not cumulative, since it cannot be assumsed that patients reporting a change in symptoms at 6 months would not have reported something different at an earlier or later date if the study had provided them with such opportunity.
observed that of 132 patients with untreated CTS over a 2-year period, 23.5% showed a deterioration in the Historic and Objective Scale score, but most cases did not show an electrophysiological deterioration (89 remained the same, 33 recovered, and 10 deteriorated). Only 1 patient had both an electrophysiological and clinical deterioration. Padua et al
reported whether the clinical outcome was unchanged or worse in groups of patients with different electrophysiological classifications. They found the clinical outcome was worse in 50% of patients with negative electrophysiology, 27.5% with moderate studies, and 50% with extreme studies. Padua et al
further observed the electrophysiological, symptomatic, functional, Historic and Objective Scale, and pain changes in patients with CTS. They reported that 16%, 21%, 16%, 32%, and 12% of patients in each of these outcome areas worsened, while 27%, 34%, 23%, 23%, and 26% of patients improved. Resende et al
presented the change in electrophysiological measures and accompanying change in symptoms over a 4- to 9-year period and found that 25% of patients had a marked improvement in electrophysiological outcome (100% of whom had improvement in terms of symptoms); 15% showed slight improvement (of whom 33% had worsening of symptoms); 50% showed no significant change (of whom 50% had worsening in terms of symptoms); and 10% had a worsening of electrophysiological measurements (of whom 50% had a worsening of clinical symptoms).
In summary, 32% to 58% of participants receiving no treatment were reported to have a negative outcome at 12 months' follow-up in 2 studies,
both of which were of moderate risk of bias. The 2 further studies reporting at 3 and 10 years were at high risk of bias and reported a negative outcome in 23.4%
A wide variation in findings was noted according to risk of bias, with studies of a moderate risk of bias appearing to show lower percentages of patients with a negative outcome (eg, 23%–68% at 3y
used a surgical episode as a marker of negative outcome of conservative management. A range of 57% to 66% of patients were observed to receive surgery after conservative management over a period of 1 to 3 years.
In summary, the reported course of conservatively managed CTS is highly variable, but symptoms do improve over time.
Prognostic factors predicting negative outcome of CTS
Eleven of the studies presented data on the association between potential prognostic factors and a negative outcome of conservatively managed CTS. Table 5 presents potential prognostic factors observed in the studies and reported associations. Not all studies presented estimates of associations with confidence intervals. Some presented P values only; some simply reported a finding as nonsignificant. Therefore, the number of studies investigating each association, the number of studies of moderate or high risk of bias (none were of low risk), and the number showing an association (direction and significance) are summarized.
Table 5Prognostic factors and strength of association for an unfavorable outcome of CTS in patients who are conservatively treated or untreated
Prognostic Factor
Direction of Association and Significance
Risk of Bias (No. of Studies)
No. and % of Studies Demonstrating Predictive Association With a Negative Outcome (Statistically Significant)
NOTE. 0, not significant and direction not provided; +, predictive of a negative outcome; −, not predictive of a negative outcome.
Abbreviations: DASH, Disabilities of the Arm, Shoulder and Hand questionnaire; Hi-Ob, Historical and Objective Scale; SF-36, Medical Outcomes Study 36-Item Short-Form Health Survey; SF-MPQ, Short-Form McGill Pain Questionnaire.
In total, 39 potential prognostic factors were identified from the studies. All of these were found to have inconclusive levels of evidence of an association with a negative outcome. This was due to inconsistencies in study findings, nonsignificant results, low numbers of studies investigating each factor, and the moderate to high risk of bias of the studies included.
Discussion
This study is the first systematic review of the prognosis of conservatively managed CTS. A substantial amount of heterogeneity exists in terms of study setting, case definition, follow-up periods, and measures of outcome between the included studies, which prevented a meta-analysis from being conducted. A best-evidence synthesis was therefore presented.
observed the course of untreated CTS, which is helpful when considering the need for or impact of treatment. These studies suggest that a proportion (28%–62%)
of patients will recover or not deteriorate further in the absence of treatment, and hence a certain period of “watchful waiting” (not clearly defined by the available evidence) may be considered clinically when discussing treatment options with patients. When considering potential mechanisms for recovery (not including mechanisms of treatment) Padua
suggests that certain undefined CTS cases are self-limiting because of a process of neural adaption, whereby the functional relationship between the nerve and the carpal tunnel adapts over time.
Because of outcomes being measured at discrete time points by each study, it was not possible to provide a cumulative percentage of patients recovering in each period and thus provide clearer information about what is happening to patients with CTS over time. Table 4 does, however, show that a proportion of patients can be observed to have deteriorated from baseline at any point between 3 months and 10 years, suggesting that the course of CTS is likely to be highly variable. It is possible that the studies with longer follow-up periods are representative of patients who improve and relapse over time, but since none of the studies were designed to observe the longitudinal course of CTS (ie, at a week-to-week or month-to-month level), such a symptom course could not be illustrated by this review.
reported that 85% of patients initially responding to conservative treatment approaches relapsed within 1 to 4 years. The possibility of future relapse therefore puts into question the observations of all studies conducted over a shorter time frame. A further consideration is that a recurrence of symptoms after a conservative treatment that then responds to a further episode of conservative management (if deemed clinically appropriate) may not necessarily represent treatment failure. However, longitudinal data that may describe this phenomenon were not available, again emphasizing the importance of long-term studies with repeated assessment of symptoms in patients with CTS.
The observed between-study variability may be partially explained by substantial differences in study setting, study design, case definitions, interventions (the effectiveness of which cannot be compared between studies), and outcomes used, but possibly also by differences in patient or disease factors (potential prognostic factors) between studies.
Prognostic factors predicting negative outcome of conservatively managed CTS
Because of inconsistencies between study findings and the lack of studies with a low risk of bias, it was not possible to identify conclusive evidence for any of the factors reported by individual studies to predict a negative outcome of conservative management. There was, however, 100% agreement in at least 3 or more cohorts with a medium or high risk of bias that symptom duration, a positive Phalen's test, and thenar wasting were associated with a negative outcome of conservative management. However, not all results were statistically significant, and hence the overall judgment remained inconclusive.
Because of a lack of robustness in design and conduct of most of the included studies, the overall body of evidence identified was felt to be of moderate and high risk of bias. This limited whether the synthesized evidence could be considered as conclusive, and as such, evidence regarding the prognosis of untreated and conservatively treated CTS remains weak. To improve future research, key recommendations would include identifying patients with CTS at baseline using a robust case definition of the condition. Patients should be followed up for a prolonged period (>3y), preferably at a number of time points using a clinically meaningful, valid, and reliable outcome measure. This would allow a longitudinal picture of CTS to be mapped. Attempts could be made to reduce attrition or better describe the risk of attrition bias by collecting information from nonresponders and to provide a description and reason for any loss to follow-up. Ideally, all potential prognostic factors should be included and measured at baseline using valid and reliable measures.
To capture the start point of the condition and its earliest management, it would be beneficial to set such a study in primary care, where it is likely most patients present initially with their symptoms and commence treatment.
Study limitations
We searched electronic databases considered to be important and relevant to the topic. Titles were screened by 1 person because of the significant number; therefore human error may have led to some titles being missed. Studies not included in databases and not identified through reference checking, Google Scholar, and expert advice may have been overlooked, such as unpublished cohort studies. Because the review did not find strong evidence for any of the prognostic factors, it is unlikely that further unpublished material would have strongly influenced our conclusions. The review focused on studies observing the course of symptoms in patients being treated conservatively for CTS but excluded cohorts being allocated specific treatments. Predictors of differential treatment response (moderators) are best identified by randomized trials, and therefore a further systematic review of these studies is planned.
Results of studies presenting only descriptive results and P values were included in the review without any risk estimates. All evidence found could therefore be included, but there is a possibility that the lack of statistical significance was due to small sample sizes and hence represent a lack of evidence for some of the prognostic factors rather than a genuine absence of association. Future prognosis research in the area of CTS should therefore ensure that estimates of associations with outcome are adequately reported and that the study population is of adequate sample size to investigate the hypothesized associations with outcome.
The unit of analysis differed between studies; that is, some analyzed outcomes at the patient level (not necessarily taking into account the laterality of the condition), while others analyzed outcomes at the wrist level (ie, patients with bilateral symptoms may be included as 2 cases, not taking dependence of outcomes within individuals into account). Issues relating to the statistical analysis of bilateral CTS have been discussed at length for clinical trials by Page et al.
A unit-of-analysis error, which may give rise to overly narrow confidence intervals and small P values, may occur when data are analyzed on the basis of the number of wrists without adjustment for nonindependence.
Such an error may also occur in prognosis research, including the reviewed studies, and be a further source of bias. Future prognostic studies should, where possible, take into consideration this risk of bias in their design and analysis plan.
Implications for clinical practice
Patients presenting with CTS can be informed of the possibility of recovery with no treatment or conservative treatment (ie, that they will not require surgery); however, factors that help to predict their likelihood of falling into this group have not been robustly determined. Increasing symptom duration, a positive Phalen's test, and thenar atrophy are likely to be prognostic factors of poor outcome of conservatively managed CTS but need confirmation in further well-designed prognostic studies. The review did not identify electrophysiological severity as a significant predictor of a negative outcome of conservative management. This may have implications for services that ration surgery to patients with more severe results and suggests that other factors should be taken into consideration alongside laboratory investigations.
Conclusions
In this review, we found useful descriptions of both the course of untreated CTS and that of conservatively managed CTS. Although none of the studies were of low risk of bias, studies of moderate and high risk of bias showed a widely ranging course of symptoms, with 23% to 89% of participants reporting a negative outcome at 3 years' follow-up. We found no consistent evidence to support factors that predict future outcome and that help to explain the wide variability in the course of symptoms.
There is likely to be an optimum time by which conservative management should be deemed to have failed and surgical intervention considered in order to prevent long-term harm, although this point has not been clearly determined, nor is it clearly possible to predict which patients may be included in this group.
Supplemental Appendix S1 Medline Search Strategy
1.
median neuropathy/ or exp carpal tunnel syndrome/
2.
“carpal tunnel syndrome”.mp. [mp=title, abstract, original title, name of substance word, subject heading word, keyword heading word, protocol supplementary concept, rare disease supplementary concept, unique identifier]
3.
Nerve Compression Syndromes/
4.
entrapment neuropath*.ti,ab.
5.
exp Median Nerve/
6.
nerve entrapment*.ti,ab.
7.
Hand/ and Pain/
8.
Pain/ and Wrist/
9.
(carpal$ adj3 tunnel$).mp.
10.
1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9
11.
exp Prognosis/
12.
exp Disease Progression/
13.
prognos*.mp.
14.
predict*.mp.
15.
factor*.mp.
16.
risk*.mp.
17.
model*.mp.
18.
evolution.mp.
19.
history.mp.
20.
indicator*.mp.
21.
course.mp.
22.
rule*.mp.
23.
transition*.mp.
24.
determinant*.mp.
25.
pattern*.mp.
26.
subgroup*.mp.
27.
sub-group*.mp.
28.
screen*.mp.
29.
long-term.mp.
30.
progress*.mp.
31.
modif*.mp.
32.
mediat*.mp.
33.
or/11-32
34.
exp Epidemiologic Studies/
35.
cohort*.mp.
36.
follow-up.mp.
37.
follow-up.mp.
38.
(“case control” or “case controlled”).mp.
39.
retrospective*.mp.
40.
prospective*.mp.
41.
((patient* or medical) adj3 (record* or review* or histor*)).mp.
42.
longitudinal*.mp.
43.
inception.mp.
44.
observation*.mp.
45.
time series.mp.
46.
outcome*.mp.
47.
or/34-46
48.
33 and 47
49.
10 and 48
References
Alfonso C.
Jann S.
Massa R.
Torreggiani A.
Diagnosis, treatment and follow-up of the carpal tunnel syndrome: a review.
American Academy of Orthopaedic Surgeons. AAOS guideline on the treatment of carpal tunnel syndrome: 2011 report for the “re-issue” of the original guideline 2011. Available at: http://www.aaos.org/Research/guidelines/CTS_Treatment_REIssue.pdf. Accessed November 28, 2012.
Supported by the National Institute for Health Research in Practice Fellowship (NIHR-IPF-2013-07-002). The views expressed in this publication are those of the authors and not necessarily those of the National Health Service, the National Institute for Health Research, or the Department of Health.
van der Windt is a member of PROGRESS Medical Research Council Prognosis Research Strategy (PROGRESS) Partnership (G0902393/99558).
01234-4&id=gr1.jpg){kind=link}