Design
A prospective observational cohort study of patients diagnosed with CTS through clinical examination and history and confirmed by nerve conduction studies. Data on putative prognostic factors and outcomes of corticosteroid injection and surgery for CTS will be collected over a 2 year period.
Study setting
In Norfolk approximately 1200 patients per annum are referred with suspected carpal tunnel syndrome to clinical neurophysiology at the Norfolk and Norwich University Hospitals NHS Foundation Trust (NNUH). Nerve conduction studies (NCS) are done on both hands to confirm a clinical diagnosis, objectively grade the severity of compression and make treatment recommendations based upon it. Referrals come from primary care e.g. general practitioner, community physiotherapist, and secondary care e.g. orthopaedics, rheumatology and plastic surgery.
Inclusion criteria
Adults aged 18 years and over who attend for NCS in the Department of Neurophysiology and are confirmed to have CTS of at least mild severity (grade ≥ 1 on Padua scale [17]) based on a combination of NCS results and clinical diagnosis, in at least one hand and for which either no treatment is recommended at present or conservative or surgical treatment is recommended. Patients with idiopathic as well as secondary CTS due to diabetes or hypothyroidism will be included.
Exclusion criteria
-
Patients with unilateral CTS who have had carpal tunnel surgery in that hand in the last 12 months.
-
Pregnancy and up to and including 12 months postpartum due to the often transient nature of CTS.
-
Serious co-morbidity e.g. malignancy, severe mental illness and where the GP considers the patient to be vulnerable and for whom participation would be otherwise detrimental.
-
Other symptomatic upper limb mononeuropathies, vibration induced conditions, radiculopathy or thoracic outlet syndrome.
-
Any other sensory or motor disturbances affecting the upper limb such as stroke, multiple sclerosis, nerve injury.
-
Unable to read or write English.
Recruitment procedure
Potentially eligible patients will be identified by the consultant neurophysiologist based on the clinical history and NCS results. The patients will be informed verbally about the current study and given an information pack containing an invitation to participate, a screening questionnaire and consent form. Patients will be asked to read through these in their own time at home and a telephone number for the study coordinator is provided if patients have any further questions. Patients who wish to participate will be asked to complete and return the consent form and screening questionnaire in an enclosed stamped addressed envelope within 2 weeks. Those who do not wish to participate will be asked to still complete the screening questionnaire which does not require disclosure of any personally identifiable information. Their data will be used to compare the consenters with non-consenters and assess external validity of the study. Those patients returning a valid signed consent form will be contacted by the study coordinator via the telephone who will verify that they meet eligibility criteria and also to verify that patients have given fully informed consent. In order to maximise response rates a reminder letter will be sent by the consultant neurophysiologists to all patients who were deemed eligible and given a study invitation pack within 1 week of attendance at the department of Neurophysiology. Each patient will only be enrolled into the study once even where CTS presents bilaterally.
Care pathway
The NCS report from the Neurophysiologist will determine the severity of CTS and include treatment recommendations. The clinical management of patients will be at the discretion of the consultant clinical neurophysiologist or referring clinician informed by local guidelines for commissioning and best practice guidelines. Patients who are eligible for the study will fall into three categories: i) no further treatment recommended or required (watch and wait), ii) recommended for conservative treatment by splinting and/or corticosteroid injection; iii) recommended for surgical decompression.
-
i)
No further treatment – patients in this group are expected to have very mild symptoms and NCS confirming a mild compression. They either do not require any conservative treatment or have declined the offer of conservative treatment such as corticosteroid injection.
ii) Conservative treatment – patients in this group are expected to have mild to moderate CTS based on NCS results and will be offered steroid injection and splints.
iii) Surgical treatment – patients in this group are expected to have moderate to severe CTS based on NCS results and self-reported symptom severity. These patients are likely to meet the current thresholds for surgical referral which are: referral to secondary care if symptoms are not resolved to patient’s satisfaction after 6 months of conservative treatment from date of 1st consultation with physician or if neurological deficit is present, i.e. sensory blunting or weakness of thenar abduction (APB).
During the 2 year follow-up, patients may undergo different treatments depending on the course of the CTS. For example a patient may be recommended not to have treatment when first enrolled however 6 months later symptom severity and associated disability may have increased necessitating conservative treatment. Or a patient may be treated with corticosteroid injection initially and progress to surgery in the course of the 2 years. Patients may also seek or be offered other treatments including alternative therapies.
All types of treatments received for CTS of either hand will be recorded by patients through 6 monthly self-report. To assist patients in the recall of treatments received a 6 months diary card will be issued in which they can record date, type, dosage and out of pocket expenditures for any treatments for CTS.
Questionnaire data
Data on putative predictive factors will be collected with a study specific questionnaire through a combination of standardised measures and purposely designed questionnaires:
Symptom severity will be assessed with the 6-item short form of the Boston Carpal Tunnel Questionnaire, the CTS-6 [18].
Hand functional status will be assessed with three subscales of the Michigan Hand Questionnaire (MHQ) which capture unilateral and bilateral hand function, activities of daily living and work performance [19].
Psychological status has been identified as a predictor of outcome in several studies and will be measured using the standardised Hospital Anxiety and Depression scale (HADs) [20].
Health-related quality of life and health utility will be assessed by the EQ-5D-3 L (http://www.EuroQol.org). Results of the EQ-5D-3 L will be converted into utilities by applying UK specific preference weights to the health profiles. QALYs accrued will be calculated by integrating utility with respect to time. QALYs accrued in year two will be discounted by 3.5%.
Comorbidity will be assessed using a self-administered standardised Comorbidity Questionnaire developed by Sangha et al. [21].
Additionally, at baseline other clinical and sociodemographic factors will be collected through self-report including age, sex, ethnicity, work status and type of work, smoking status, alcohol use, household income.
Results from NCS will be requested from the neurophysiologist once patients have consented and been enrolled into the study. Median nerve sensory nerve action potential (amplitude and velocity) for digit III and distal motor latency (amplitude and velocity) will be recorded. Padua’s criteria [17, 22] for classifying CTS severity according to the neurophysiology results will be used to grade severity from 1 to 6 (mild to extremely severe).
Resource use and cost
Patients will be asked to complete a set of resource use questions asking about prescribed medications, primary, secondary and tertiary NHS and private healthcare contacts, alternative therapies and over the counter medicines purchased and time off work due to CTS since the last follow-up. The questionnaire completed at baseline will pertain to resource use in the three months prior to commencement of the study. Following completion of the baseline questionnaire and each follow-up (except final follow-up), patients will be issued with a 'resource use diary’ in the form of a card with tick-boxes and space for notes for patients to complete over the interim time period [23]. This will be used as an aide memoire when completing the resource use sections of the questionnaires.
To verify the patient-self-reported data, at the 2 year follow-up the General Practitioner of a random sample of at least 10% of patients will be contacted to confirm numbers of primary care contacts for CTS, whether the patient is still receiving treatment for CTS, details of drug and dosage where corticosteroid injection has been given, and other drugs prescribed for CTS. Records of surgery performed will also be extracted and compared with patient reported data.
Outcomes
The primary outcome used to classify treatment successes and failures will be patient-reported global assessment of outcome (GAO) at 2 years using a 5 point ordinal scale: worse = 1, unchanged = 2, slightly better = 3, much better = 4, completely cured = 5. A grade of 3 or above will be defined as treatment success. Secondary outcomes include patient-reported clinical symptoms (CTS-6), functioning (MHQ), psychological health HADs, quality of life (EQ-5D) and resource use.
Data collection methods
All patients enrolled into the study will be asked to complete a questionnaire at baseline and again at 6, 12, 18 and 24 months. Given the increasing proportion of patients who have access to the internet it is possible to collect these data online using secure encrypted web-based systems which are a cost-effective alternative to paper questionnaires. At initial enrolment patients will be asked if they have access to the internet and given the option to complete all follow-up questionnaires by mail or electronically. Data entry will be via a web-based system provided by Norwich Clinical Trials Unit Data Management team. Any individual, whether study staff or participant entering data online will gain access to the website using a unique username and password. Participants will only have access to their own data. Once a participant has confirmed the submission of data, they will not be able to change it. The database is built using MS SQL Server and the website and associated software is built using MS ASP.NET. The database and website reside on the Norwich CTU secure server at University of East Anglia. Data from patients who opt to return questionnaires via the post will be entered by a research associate. On receipt of a completed online or paper questionnaire participants will be sent a thank you card and a diary to help them record information about treatments received and out of pocket expenses incurred as a result of the carpal tunnel syndrome for the forthcoming 6 months.
Sample size
An estimated 1200 patients are seen at the Norfolk and Norwich University Hospital (NNUH) each year with suspected carpal tunnel syndrome. Of these, at least 85% (around 1020) will be confirmed as having CTS through NCS indicating a mild or worse compression. Assuming a recruitment rate of 75%, which is not unreasonable, given previous experiences, we would expect to recruit around 765 CTS subjects per annum, i.e. 1530 over the two year recruitment period. It is estimated that around 30% (460) will go to surgery immediately and 70% (1070) will be treated with corticosteroid injection. Of the latter, it is expected that, by two years follow-up, two thirds will have been referred for surgery, whereas the remaining third (about 355) will be considered a success. There is no standard agreed method for calculating sample sizes or statistical power for logistic regression models. However, following the approach by Demidenko [24] a sample size of 1070 would confer 90% statistical power, assuming a 33% success rate, to detect an increase in success with an odds ratio of 1.51 for an explanatory binary variable with an even split in classification (i.e. an equal number of subjects in each group). For an uneven split, this odds ratio would be smaller. Should recruitment be less than expected, a sample size of 535 (i.e. half that expected) would confer 90% power to detect an odds ratio of 1.79.
Analysis
The primary outcome (global assessment of outcome) will be modelled using a binary logistic regression model. Predictors will be considered on a univariate basis and those found to have a statistically significant effect on the outcome (at the 10% level) will be entered simultaneously into a multivariable model. A backwards deletion approach based upon change in likelihood will be used to remove those predictors with no significant independent effect. An ordinal logistic regression model will also be considered based upon the proportional odds assumption using the original 5 point ordinal scale as the outcome. However, as the assumption is often difficult to verify, this will be considered as a secondary analysis.
Secondary outcomes will be analysed in a similar fashion using regression models with link functions and error terms appropriate to the distribution of the outcome of interest.
Cost of illness analysis
The resource use data will be used to perform a longitudinal estimate of the cost of illness of carpal tunnel syndrome in the subject cohort over a period of two years from the perspectives of the NHS and society (defined as the sum of NHS and patient out of pocket costs and morbidity-related lost productivity). The total cost of each patient will be calculated as the sum of resource quantities multiplied by relevant unit costs extracted from standard NHS databases (e.g. NHS reference costs and Unit Costs of Health and Social Care). Unit cost sources will be from a common price year, to be determined at time of analysis. Costs incurred in year two will be discounted at 3.5% as per HM Treasury guidelines.
Regression models will be developed as described above to predict cost over two years according to a number of prognostic variables, e.g. symptom severity or functional status at baseline, age, and socioeconomic status. The analysis will make appropriate adjustments to take account of missing and censored data where necessary. Where there are systematic differences between patient self-reported data and primary care records, a secondary analysis will be conducted adjusting primary care cost for any observed bias.
Project timetable
Recruitment to the study began in July 2013 and is due to be completed by July 2015. The two-year follow-up is anticipated to be completed in July 2017.
Ethical approval
The PALMS study was approved by the East of England/Norfolk Research Ethics Committee (REC reference 13/EE/0106) on 22 April 2013.