Association of psychological distress, quality of life and costs with carpal tunnel syndrome severity: a cross-sectional analysis of the PALMS cohort

Objectives The Prediciting factors for response to treatment in carpal tunnel syndrome (PALMS) study is designed to identify prognostic factors for outcome from corticosteroid injection and surgical decompression for carpal tunnel syndrome (CTS) and predictors of cost over 2 years. The aim of this paper is to explore the cross-sectional association of baseline patient-reported and clinical severity with anxiety, depression, health-related quality of life and costs of CTS in patients referred to secondary care. Methods Prospective, multicentre cohort study initiated in 2013. We collected baseline data on patient-reported symptom severity (CTS-6), psychological status (Hospital Anxiety and Depression Scale), hand function (Michigan Hand Questionnaire) comorbidities, EQ-5D-3L (3-level version of EuroQol-5 dimension) and sociodemographic variables. Nerve conduction tests classified patients into five severity grades (mild to very severe). Data were analysed using a general linear model. Results 753 patients with CTS provided complete baseline data. Multivariable linear regression adjusting for age, sex, ethnicity, duration of CTS, smoking status, alcohol consumption, employment status, body mass index and comorbidities showed a highly statistically significant relationship between CTS-6 and anxiety, depression and the EQ-5D (p<0.0001 in each case). Likewise, a significant relationship was observed between electrodiagnostic severity and anxiety (p=0.027) but not with depression (p=0.986) or the EQ-5D (p=0.257). National Health Service (NHS) and societal costs in the 3 months prior to enrolment were significantly associated with self-reported severity (p<0.0001) but not with electrodiagnostic severity. Conclusions Patient-reported symptom severity in CTS is significantly and positively associated with anxiety, depression, health-related quality of life, and NHS and societal costs even when adjusting for age, gender, body mass index, comorbidities, smoking, drinking and occupational status. In contrast, there is little or no evidence of any relationship with objectively derived CTS severity. Future research is needed to understand the impact of approaches and treatments that address psychosocial stressors as well as biomedical factors on relief of symptoms from carpal tunnel syndrome.

4. Page 4, lines 36-37: What type of -severity‖ is the debate about? The illness or the disease? Symptoms or pathophysiology? I think there is little debate about the indications for CTR in terms of pathophysiology. The only debate there might be that advanced CTS has permanent nerve damage and surgery can be disappointing.
5. Page 4, lines 41-47: It's not clear from the Introduction why you looked at psychological factors? I would argue that patients and surgeons often consider surgery based on symptoms when they should base surgery on pathophysiology. Your confirmatory study that symptoms are related to stress, distress, and less effective coping strategies would further establish that symptom severity is not a good basis for considering surgery in patients with CTS. Given the nature of CTS as a disease where surgery might be considered -necessary‖ because preserving sensation in the median nerve innervated digits is so important, we can apply the finding that symptom intensity is not a good guide to the role of surgery to other hand conditions-most of which are highly discretionary and preference-sensitive. 6. Page 5, line 3: I recommend you separate symptoms from pathophysiology. Describe the diagnosis of idiopathic median neuropathy at the carpal tunnel using electrodiagnostic testing (objective measurable pathophysiology). Distinguish that from symptoms. Presumably everyone in this cohort had testing for symptoms. So for this study you can disregard symptoms as a part of the diagnosis and just focus on electrodiagnostic objective verification of disease in a cohort of patients with symptoms. 7. Page 5, line 8: CTS is never -due‖ to hypothyroidism or diabetes. It's debatable whether it's even related. Severe hypothyroid is a thing of the past. Omit this. 8. Results: People with more advanced disease were less anxious and more adaptive.
9. Lines 38-There are plenty of studies of symptoms intensity and limitations in people with CTS. Many of them look at electrodiagnostic severity. Here's two: 23027833, 23890497. Keep looking! 10. I completely disagree with your conclusions. Your studybuilding on lots and lots of evidence on CTS, hand surgery, and medicine in general-shows that there is a substantial divide between pathophysiology (disease) and symptoms/limitations (illness). In the biopsychosocial paradigm, disease is treated appropriately (e.g. penicillin for strep throat, surgery for moderate CTS on EMG) and symptoms and limitations are managed by addressing stress, distress, and less effective coping strategies. It would be a major step backwards to say that we should treat pathophysiology based on symptoms. That represents a misdiagnosis and will continue the shameful undertreatment of the psychosocial aspects of illness.

GENERAL COMMENTS
This paper covers an interesting topic but it needs further revision and several sections should be completely rewritten. The introduction is poor. Authors should discuss current knowledge on anxiety and depression and the other variables in CTS and chronic pain. The current introduction does not help the readers for the current study. Methods: If more data from the same sample size has been published it should be included in the methods. This section is clearly poorly described. All outcomes should be described and justify their use in CTS. More description of the EMG analysis and classification of the patients is clearly needed. I presume that authors first conducted correlation analysis since this is needed for doing after linear regression models. The statistical analysis section should be clarified and expanded. Results: If almost 80% of the patients had normal HADS scores, depression and anxiety could not be determined, so the results would be not valid. Authors should include r and adjust R values of the linear regression models, and not just the P values.
We cannot say that severity is associated to anxiety and depression where there were not these psychological aspects. Please clarify in the discussion. Authors should include direct association coefficients before linear regression. There is no data on how costs were analyzed. It is important to include a brief comment in the methods, and not just refer to the published protocol. Discussion: More discussion on depression and anxiety is clearly needed. There are some studies including sample sizes of 200 patients (Clin J Pain 2014, Pain Med 2015 investigating the role of depression in function. These should be also included. Authors should include some hypothesis why depression can be related to symptom's severity.

GENERAL COMMENTS
This is a large study with new interesting information that is useful to clinicians and researchers in the field of carpal tunnel syndrome. The manuscript is well-written and generally clear.

Introduction:
The statement that surgical decompression rates after initial corticosteroid injection -vary by the country and referral criteria‖ is surprising because they should depend on efficacy of treatment in resolving patients' symptoms rather than in which country the patients live. Probably these variations reflect the study design and other study-related factors rather than country.

Methods:
Was the study registered? Were the primary and secondary analyses for this study prespecified in a study protocol before patients were recruited. What was the rational of recruiting patients from 4 neurophysiological departments but only 1 hand surgery center? Usually surgeons in clinical practice refer patients when they need confirmation (ie clinical diagnosis not adequately clear to proceed to surgery) while patients wíth typical history are often treated without nerve conduction tests. If there are data available it would be helpful to report the proportion of patients referred to NCS by primary care physician and by surgeons. It would be helpful to describe the Bland criteria because not all readers are familiar with them and they are important in the analyses and conclusions. Usually the EQ-5D value that is derived from weights is called the EQ-5D -index‖ not score.
Considering that false positive NCS may be more common than believed, is it possible that the paradoxical association between lowest NCS severity (near normal) and high anxiety may be incorrect CTS diagnosis? Why do NHS and societal costs increase with self-reported CTS severity? Are younger patients with clinical diagnosis of CTS less likely to have abnormal NCS than older patients? This is important because the severity of self-reported symptoms are likely to be similar or even as sometimes suggested that old persons with CTS may even have less symptoms despite severe median nerve dysfunction. Is it known how large proportion of patients that are not referred to NCS but treated entirely based on clinical diagnosis? are there any data about this in the study region? Table 1 The mean age of the patients (60 years) is somewhat higher than the average CTS population and only about 50% working, do we have any data about the group that were eligible but did not participate in the study? Table 2 The authors show data about activity limitations (MHQ) but these are not further addressed. It would be interesting to see for example the relationship between activity limitations and CTS-6 and NCS. Anxiety and depression categories show that only a small proportion have at least moderate (18% for anxiety and 8% for depression). Table 4 It would be helpful to add number of patients for each of the CTS-6 score and NCS grade categories. It would also be interesting to add the mean CTS-6 score for each NCS grade. 95% confidence interval is usually easier to interpret than SE. Any correlation between age and NCS grade? Table 5 Is it correctly interpreted that patients with CTS-6 score of 1 (ie have no CTS symptoms according to the CTS-6 scale) have the highest costs? Any explanations for this? It may indicate that these patients have some other diagnosis than CTS. In both Table 4 and 5 the CTS-6 score categories (for example 1-2 and then 2-3 etc) should be more accurately specified without overlapping. It is unclear how to interpret the data in Table 5 regarding -Age & CTS-6‖ and -Age & NCS grade‖. The reference list does not follow a uniform style, needs to be checked. Figure 1 Were 1918 patients eligible? 140 declined and 820 accepted, who were the remaining patients?

Discussion
There are longitudinal CTS studies that have measured mental health aspects, for example using the SF-36, before and after surgery, the authors could discuss the findings from these studies.

VERSION 1 -AUTHOR RESPONSE
Reviewer: 1 Reviewer Name: David Ring Institution and Country: Dell Medical School --The University of Texas at Austin, USA Competing Interests: None.
Comment 1. Page 4, lines 15-16: You must distinguish between symptoms/limitations (illness) and disease (median neuropathy at the carpal tunnel). Symptoms might be relieved or wax and wane, but the disease marches on and leads to atrophy, weakness and static numbness over several decades.
authors response: Thank you we agree and have changed the sentence to : Whilst in some people symptoms of CTS resolve spontaneously or respond to non-operative treatment by corticosteroid injection, many undergo surgical decompression to prevent irreversible nerve damage. authors response: Thank you this indeed an important distinction and it is the lack of clarity in the existing guidelines about what is meant by severity which compounds the problem. Some define it using neurophysiology, others rely on patient-reported severity. We have therefore placed ‗moderate to severe' in quote marks as we are citing from these documents and added a sentence highlighting this inconsistency in the terms used.
Comment 5. Page 4, lines 41-47: It's not clear from the Introduction why you looked at psychological factors? I would argue that patients and surgeons often consider surgery based on symptoms when they should base surgery on pathophysiology. Your confirmatory study that symptoms are related to stress, distress, and less effective coping strategies would further establish that symptom severity is not a good basis for considering surgery in patients with CTS. Given the nature of CTS as a disease where surgery might be considered -necessary‖ because preserving sensation in the median nerve innervated digits is so important, we can apply the finding that symptom intensity is not a good guide to the role of surgery to other hand conditions-most of which are highly discretionary and preference-sensitive.
authors response: This is also commented on by reviewer 2 and we have now added further justification in the introduction of why we have examined psychological factors and HRQoL. We also agree that symptoms are not a good indicator for how to treat CTS but would maintain than they are a measure of disease impact and therefore can inform ‗whether' to treat someone. Comment 6. Page 5, line 3: I recommend you separate symptoms from pathophysiology. Describe the diagnosis of idiopathic median neuropathy at the carpal tunnel using electrodiagnostic testing (objective measurable pathophysiology). Distinguish that from symptoms. Presumably everyone in this cohort had testing for symptoms. So for this study you can disregard symptoms as a part of the diagnosis and just focus on electrodiagnostic objective verification of disease in a cohort of patients with symptoms.
authors response: All patients were examined for signs and symptoms and objective nerve conduction studies. It is common practice that neurophysiologists offer an interpretation of the NCS results alongside signs and symptoms in their written reports to the referrer. We have reworded this sentence for clarity.
Comment 7. Page 5, line 8: CTS is never -due‖ to hypothyroidism or diabetes. It's debatable whether it's even related. Severe hypothyroid is a thing of the past. Omit this.
authors response: These were the prespecified inclusion criteria in the published protocol and we feel it is important to report these transparently. However we have changed the sentence so that the attribution of CTS to these conditions is removed.
Comment 8. Results: People with more advanced disease were less anxious and more adaptive.
authors response: Yes, we agree that they were less anxious however we cannot say that they were more adaptive as we did not measure that.
Comment 9. Lines 38-There are plenty of studies of symptoms intensity and limitations in people with CTS. Many of them look at electrodiagnostic severity. Here's two: 23027833, 23890497. Keep looking! authors response: We assume the reviewer is referring to the discussion section line 38 and that the claimed ‗primacy' is being questioned here. We have reworded this section now and included further studies as suggested and discussed our findings in relation to these.
Comment 10. I completely disagree with your conclusions. Your study-building on lots and lots of evidence on CTS, hand surgery, and medicine in general-shows that there is a substantial divide between pathophysiology (disease) and symptoms/limitations (illness). In the biopsychosocial paradigm, disease is treated appropriately (e.g. penicillin for strep throat, surgery for moderate CTS on EMG) and symptoms and limitations are managed by addressing stress, distress, and less effective coping strategies. It would be a major step backwards to say that we should treat pathophysiology based on symptoms. That represents a misdiagnosis and will continue the shameful undertreatment of the psychosocial aspects of illness.
authors response: We agree with the reviewer's stance regarding the approach to medical treatment of CTS and are unclear how our conclusions contradict this. We don't believe that we have said that pathophysiology should be treated based on symptoms. We have reworded the conclusion to clarify this distinction between whether someone needs treatment for CTS and what treatment should be given and the likely prognosis. The former should be informed by symptoms and impact on function, whereas the latter should be informed by pathophysiology e.g. if grade 1 then non-operative treatment may be tried.

Reviewer: 2
Reviewer Name: Cesar Fernandez-de-las-Peñas Institution and Country: Universidad Rey Juan Carlos, Spain Competing Interests: None declared Comment: This paper covers an interesting topic but it needs further revision and several sections should be completely rewritten. The introduction is poor. Authors should discuss current knowledge on anxiety and depression and the other variables in CTS and chronic pain. The current introduction does not help the readers for the current study.
authors response: We have added a section including previous research on CTS and psychological aspects as well as clarified the rationale for our analysis. We also want to point out that the focus of our study was not chronic pain and whilst pain features in the measurement of symptoms in the CTS-6 we cannot say that it is ‗chronic'.

Methods
Comment: If more data from the same sample size has been published it should be included in the methods. This section is clearly poorly described. All outcomes should be described and justify their use in CTS. More description of the EMG analysis and classification of the patients is clearly needed. I presume that authors first conducted correlation analysis since this is needed for doing after linear regression models. The statistical analysis section should be clarified and expanded.
authors response: No other data from this sample has been published. The study is still in follow-up and analysis of that data is still to be completed. The justification for the included outcome measures is fully described in the previously published protocol which has been referenced. We did not undertake EMG analysis but nerve conduction studies (NCS), however we have given further details on the methods for classifying objective severity by NCS in a supplementary table (also suggested by reviewer 3). We did examine bi-variate correlations between CTS measures and our selected outcomes. These ranged in magnitude from 0.03 to 0.56. However, the statistical analysis was ‗hypothesis driven', i.e. we wished to examine the strength of relationship between CTS measures and outcomes, rather than ‗exploratory', i.e. assessing which measures were related to outcomes. As such, there was no ‗model construction' or process of selection of explanatory variables; we included CTS-6 and NCS measures as predictors and then further variables as potential confounders. We do not feel that the bi-variate correlations contribute substantially to the paper and have not included them.

Results
Comment: If almost 80% of the patients had normal HADS scores, depression and anxiety could not be determined, so the results would be not valid. Authors should include r and adjust R values of the linear regression models, and not just the P values. We cannot say that severity is associated to anxiety and depression where there were not these psychological aspects. Please clarify in the discussion. Authors should include direct association coefficients before linear regression.

Introduction
Comment: The statement that surgical decompression rates after initial corticosteroid injection -vary by the country and referral criteria‖ is surprising because they should depend on efficacy of treatment in resolving patients' symptoms rather than in which country the patients live. Probably these variations reflect the study design and other study-related factors rather than country.
authors response: Yes we agree that these should be the same however there are differences in access to surgical decompression in the UK National Health Service depending on the specific local clinical commissioning criteria. We have reworded this sentence as suggested.

Methods
Comment: Was the study registered? Were the primary and secondary analyses for this study prespecified in a study protocol before patients were recruited.
authors response: The PALMS cohort study was not registered but the protocol has been published (referenced in text). The primary analysis, which is to develop multivariable models for predictors of outcome and predictors of cost are pre-specified. The focus of this paper is the cross-sectional analysis which is a secondary analysis of the baseline data for the cohort.
Comment: What was the rational of recruiting patients from 4 neurophysiological departments but only 1 hand surgery center? Usually surgeons in clinical practice refer patients when they need confirmation (ie clinical diagnosis not adequately clear to proceed to surgery) while patients wíth typical history are often treated without nerve conduction tests. If there are data available it would be helpful to report the proportion of patients referred to NCS by primary care physician and by surgeons.
authors response: We chose participating centres where neurophysiology testing is undertaken in all patients referred from primary care with a suspected diagnosis of CTS as this was considered the best time point at which to enrol patients into this prospective study. We have reworded that sentence to clarify that these were secondary care sites and that patients were referred from primary care.
Comment: Who diagnosed the patients with CTS and who invited them to participate? Who graded the NCS results?
authors response: Either a neurophysiologist or hand surgeon made the diagnosis based on the combination of signs and symptoms and NCS reports. The grading of all NCS reports according the Bland criteria was done by the first author (CJH).
Comment: Do these 4 neurophysiological units use identical testing methods and identical reference values for what is abnormal?
authors response: The method and equipment for orthodromic testing is the same in all units and standard operating procedures were followed for identifying eligible patients who were at least grade 1 based on the Bland's criteria (now given in a supplementary table).
Comment: In patients with bilateral CTS, how was the worst hand determined, by the patient or by CTS-6 score, or nerve conduction test results?
authors response: Worst hand was determined by the patient based on CTS-6 score and we have added this in the text.
Comment: A description of the HADS scoring would be helpful.
authors response: We have added this to text under methods/data collection.
Comment: Was -time of work due to CTS‖ self-reported, like number of days? Needs some clarification.
authors response: Yes it is patient reported and in dayswe have amended this.
Comment: It would be helpful to describe the Bland criteria because not all readers are familiar with them and they are important in the analyses and conclusions.
authors response: We have provided this as a supplementary file/table with the criteria (also suggested by reviewer 2).
Comment: Usually the EQ-5D value that is derived from weights is called the EQ-5D -index‖ not score.
authors response: Thank you for pointing this out, we have amended as suggested.
Comment: Considering that false positive NCS may be more common than believed, is it possible that the paradoxical association between lowest NCS severity (near normal) and high anxiety may be incorrect CTS diagnosis?
authors response: Thank you, yes that is also a possible explanation and we have included this in the discussion.
Comment: Why do NHS and societal costs increase with self-reported CTS severity?
authors response: We have added a sentence to elaborate on this within the results and believe that in the discussion we did give a possible explanation as follows: ‗This may also explain the increased costs, both from an NHS and personal perspective, driven by increased treatment-seeking behaviour in those who perceive their symptoms as worse.' Comment: Are younger patients with clinical diagnosis of CTS less likely to have abnormal NCS than older patients? This is important because the severity of self-reported symptoms are likely to be similar or even as sometimes suggested that old persons with CTS may even have less symptoms despite severe median nerve dysfunction.
authors response: We did consider whether the more severe NCS grades were from older patients who, as you say, sometimes report less severe symptoms, however we found that the distribution of grades 5 and 6 were across the age span. Also the multivariable model accounts for several variables including age and so any association between NCS and psychological status or QoL is independent of age.
Comment: Is it known how large proportion of patients that are not referred to NCS but treated entirely based on clinical diagnosis? are there any data about this in the study region?
authors response: Indeed, it is the case that many patient undergo non-operative and operative treatment without ever having had any NCS. We do not have data on this for our cohort as we deliberately selected centres, where patients could be recruited immediately after having CTS confirmed with NCS.
Comment: Table 1 The mean age of the patients (60 years) is somewhat higher than the average CTS population and only about 50% working, do we have any data about the group that were eligible but did not participate in the study?
authors response: Yes we concur this is not typical but because this was a research study as opposed to an analysis of routinely collected clinical data we depended on patients who gave consent to be enrolled in this study. This has biased the sample more towards those retired and older. We do not have any information on those who did not respond as we do not have ethical approval to access their clinical records. We have added a comment in the discussion acknowledging this limitation.
Comment: Table 2 The authors show data about activity limitations (MHQ) but these are not further addressed. It would be interesting to see for example the relationship between activity limitations and CTS-6 and NCS.
authors response: Yes we agree and have now extended the statistical modelling to include the MHQ as outcomes. These results are now shown in table 4(b), with the original table 4 now being relabelled table 4(a).
Anxiety and depression categories show that only a small proportion have at least moderate (18% for anxiety and 8% for depression). authors response: Yes, this is correct.
Comment: Table 4 It would be helpful to add number of patients for each of the CTS-6 score and NCS grade categories.
authors response: These numbers have now been added to tables 4 and 5. We would like to note that when doing this for tables 4 and 5 we found some minor discrepancies and the analysis of costs had to be re-run. As a result the values in the text and table 5 changed slightly however do not affect the overall results or interpretation.
Comment: It would also be interesting to add the mean CTS-6 score for each NCS grade.
authors response: The mean (and standard deviation) CTS-6 score by NCS grade is as follows: 3. I disagree with your conclusions. NCS should be used to assess disease severity and disease severity should be used to determine the indication for surgery. Greater symptoms and limitations than expected for a given disease severity should prompt a screen for psychological distress, greater social stressors, and less effective coping strategies. Treatment can then be tailored to address these aspects of the illness. Surgery for normal or mild NCS would be likely to result in a missed opportunity to treat depression or catastrophic thinking. Avoiding surgery because symptoms aren't severe enough could lead to numbness, weakness, and atrophy that could have been avoided. At a minimum, this possibility should be acknowledged in your paper, even it does not fit your bias.

GENERAL COMMENTS
Authors have edited all comments properly

GENERAL COMMENTS
The questions have been addressed very well. Two minor issues: 1. Because the CTS-6 scale ranges from 1 to 5, the mean CTS-6 score can never be less than 1.0 and therefore the score category -0 -1.00‖ should be corrected to -1.00‖ in all tables. It may be of interest when discussing the results for this category to highlight that a mean score of 1 means all patients in that category had no symptoms of CTS as measured by the CTS-6 (numbness, tingling and pain experienced during the past 2 weeks). 2. In the Discussion about health utility measures, a statement mentions the exact value from a previous study and refers to it as the SF-36 but it is more accurately the SF-6D index value (even though the SF-6D is derived from the SF-36).

VERSION 2 -AUTHOR RESPONSE
Reviewer: 1 Reviewer Name: David Ring Institution and Country: Dell Medical School --The University of Texas at Austin, USA Competing Interests: None.
Comment 1. Intro, first paragraph: This background information is not particularly necessary or helpful. Some of it introduces debatable concepts such as the idea that idiopathic median neuropathy at the carpal tunnel can resolve spontaneously or be cured with a steroid injection. This represents confusion of the illness (symptoms and limitation) with the disease (the pathophysiology of the median nerve). As your data clearly shows symptoms correlate with psychosocial factors, not pathophysiology. That means that people with pathophysiology may not experience symptoms. With a structural disease like carpal tunnel syndrome, symptoms may resolve or become less bothersome, but the disease most likely persists. You may do people a disservice by suggesting that median neuropathy can resolve on it's own. Such a person might adapt and then end up with advanced neuropathy that is permanent. I realize there is room for debate here. So my advice is to just delete this background information. You can do without it.
Authors response: We have removed any reference to CTS resolving by its own or in response to steroid injections, however we feel that it would not be appropriate to remove this whole first paragraph as it sets an important context for the next paragraph and the study.
Comment 2. Introduction, paragraph 2. Moderate to severe refers to electrodiagnostic testing (pathophysiology). Since symptoms don't correlate with disease severity, greater symptoms indicate opportunities for relief of distress and work on more effective coping strategies. Greater symptoms should not affect the indication for disease-modifying surgery.
Authors response: the words ‗moderate to severe' have been used here in the context of how the clinical commissioning group guidelines use them and have been placed in quotation marks. We are unclear what else is being requested here.
Comment 3. I disagree with your conclusions. NCS should be used to assess disease severity and disease severity should be used to determine the indication for surgery. Greater symptoms and limitations than expected for a given disease severity should prompt a screen for psychological distress, greater social stressors, and less effective coping strategies. Treatment can then be tailored to address these aspects of the illness. Surgery for normal or mild NCS would be likely to result in a missed opportunity to treat depression or catastrophic thinking. Avoiding surgery because symptoms aren't severe enough could lead to numbness, weakness, and atrophy that could have been avoided. At a minimum, this possibility should be acknowledged in your paper, even it does not fit your bias.