Author's response to reviews Title: Minimal clinically detectable and important changes for pain in patients with nonspecific neck pain Authors:

Objectives: To estimate the Minimal Changes in neck and referred pain intensity that are clinically detectable..... Introduction, paragraph 1: Minimal clinically detectable change is defined as the minimal change in the score measuring a symptom that the patient is able to perceive. Minimal clinically important change (MCIC) is defined as the minimal variation in the score that measures a Minimal clinically detectable change is defined as the minimal change in the score measuring symptom that is meaningful for patients. Since a change needs to be perceived before being potentially clinically important, minimal clinically detectable change is often one of the values searched for in studies aiming to assess MCIC 1-11 Therefore, MCIC is used as the generic abbreviation throughout this article. 1.1.2 The authors do not adequately describe the recruitment process and likely generalizability of the results. How many potentially-eligible patients were considered for inclusion over the recruitment time period? What were the reasons for exclusion? In accordance with the reviewer’s comment, this has been further clarified in the updated version of the manuscript, which now reads: Methods, paragraphs 5 and 6: Inclusion criteria were: seeking health care at any of the primary care centers belonging to the Ib-Salut, for neck pain (NP) lasting 14 or more days, either with or without pain referred to the arm or arms (AP), reporting a pain severity ≥ 3 points on a pain intensity numerical rating scale (PI-NRS), and undergoing a neuroreflexotherapy (NRT) intervention. Exclusion criteria were: data suggesting potential underlying diseases, (the current neck pain episode being the first one in a patient under age 20 or onset over 55 in which appropriate diagnostic procedures had not yet been performed before referral to the study, non-mechanical pain, widespread neurology (disseminated neurological findings), fever, weight loss, systemically unwell, a history of: significant trauma, systemic steroids, osteoporosis, cancer, drug abuse, HIV), widespread (>1 nerve root) or progressive motor weakness in the arms, and patients’ refusal to sign the study’s informed consent. Results paragraph 1: In this study 678 patients were recruited and 20 (2.9%) were excluded because of refusal to sign the informed consent (15 cases), fever (3 cases) and patients feeling systematically unwell (2 cases). Therefore, 658 were included, and none were lost to follow-up. 1.1.3 Background (first paragraph, first sentence) The MCIC relates to changes in scores that may measure pain, disability, impairment etc., The MCIC is not related to "variation of symptoms" as stated in the manuscript. In accordance with the reviewer’s comment, wording has been improved in the updated version of the manuscript, which now reads (Introduction, paragraph 1): Minimal clinically detectable change is defined as the minimal change in the score measuring a symptom that the patient is able to perceive. Minimal clinically important change (MCIC) is defined as the minimal variation in the score that measures a symptom that is meaningful for patients. Since a change needs to be perceived before being potentially clinically important, minimal clinically detectable change is often one of the values searched for in studies aiming to assess MCIC. Therefore, MCIC is used as the generic abbreviation throughout this article. 1.1.4 Background (second paragraph, first sentence) LBP patients do not have MCIC values. Instruments used with LBP patients have MCIC values (e.g, Oswestry, ROland Morris, numeric pain rating, etc.) Please change the wording of this sentence. In accordance with the reviewer’s comment, the wording has been changed in the updated version of the manuscript, which now reads (Introduction, paragraph 2): Different approaches can be used to determine MCIC. One is to estimate the mean change in score in patients who actually report to have improved (referred to as “mean change score”, or MCS). Another approach is to determine a cutoff point (referred to as “minimal detectable change”, or MDC) so that a patient with a change smaller or equal to the MDC has a chance of more than 95% that no real change has taken place. 1.1.5 Methods, (study population, second paragraph) The meaning of this paragraph is not clear. How does the protocol of patients with neck pain relate to individuals with subacute/chronic LBP? Do the authors intend to say that all subjects in this study received neuroflexotherapy? What does "post-marketing surveillance" have to do with this study? Please clarify this paragraph? In accordance with the reviewer’s comment, this has been further clarified in the updated version of the manuscript, which now reads: Methods, Study Population, paragraph 2 and 3: In accordance with the current treatment protocol for neck pain in routine practice in the Ib-Salut, subacute and chronic patients were derived to a specialized Unit, where patients with a pain severity ≥ 3 points on a pain intensity numerical rating scale (PI-NRS), underwent a neuroreflexotherapy (NRT) intervention. 1.1.6 This is a minimally invasive procedure that has proven to be safe, effective and cost/effective, and has been extensively described in the literature. Inclusion criteria were: seeking health care at any of the primary care centers belonging to the Ib-Salut, for neck pain (NP) lasting 14 or more days, either with or without pain referred to the arm or arms (AP), reporting a pain severity ≥ 3 points on a pain intensity numerical rating scale (PI-NRS), and undergoing a neuroreflexotherapy (NRT) intervention. Discussion, paragraphs 8 and 9: Mean duration of pain when patients entered this study was over 540 days (Table 1). During that period, they had all received many forms of treatment and many still received them during the study. Since data being analyzed in this study derive from post-marketing surveillance of neuroreflexotherapy, all of them received that specific form of treatment. This does not affect the generalizability of results from this study, since MCIC calculation relies on patients’ self-assessment of their own evolution and instruments used to assess evolution of symptoms, no matter what treatments are influencing that evolution. No data suggests that MCIC are sensitive to the treatment being received and, in fact, they seem to be consistent even across different chronic pain conditions. On the contrary, using post-marketing surveillance methods in a National Health Service to assess MCIC has a number of advantages. It makes it possible to assess MCIC values in routine practice conditions, as opposed to using data from randomized controlled trials in which Hawthorne and other unspecific effects might influence patients’ perception of global improvement and, therefore, the results. In addition, post-marketing surveillance makes it possible to recruit large representative samples and to minimize losses to follow-up, therefore giving a better general picture of what MCIC values are likely to be in “normal” clinical conditions. 1.1.6 Methods, (study population, second paragraph) Please describe the neuroflexotherapy that these patients apparently received. In accordance with the reviewer’s comment, the updated version of the manuscript reads (Methods, Study Population, paragraph 2): In accordance with the current treatment protocol for neck pain in routine practice in the Ib-Salut, subacute and chronic patients were derived to a specialized Unit, where patients with a pain severity ≥ 3 points on a pain intensity numerical rating scale (PI-NRS), underwent a neuroreflexotherapy (NRT) intervention. This is a minimally invasive procedure that has proven to be safe, effective and cost/effective, and has been extensively described in the literature. Data included in the current study derive from methods used for post-marketing surveillance of this technology in routine clinical practice. 1.1.7 Methods (external criterion) Studies examining the responsiveness of various outcome measures have frequently used 7-point or 15-point scales for patient selfassessment. This study uses a 4-point scale. It is unclear if this scale is likely to provide adequate discrimination to permit identification of patients who report meaningful improvement. Can the authors provide any support for the use of this external criterion? Has its use been described elsewhere or has this 4-point scale been validated for this purpose? In reply to the reviewer’s question: In fact, studies that used 5 or 7 seven categories collapsed them into 3 (improved, unchanged and deteriorated) for the analysis. As seen in the table below, categories used for calculating MDC, MCIC and ROC were very similar to those used in our study Studies Categories presented to the patient Categories used in the analysis MDC MCIC ROC Ostelo* 1Complete recovery 2Much improved 3Slightly improved 1Improved (1+2) 2Unchanged (3+4+5) 3Deteriorated (6+7) 3+4+5 2 1+2 vs 3+4+5 4No change 5Slightly worsened 6Much worsened 7Worse than ever Van der Roer** 1Complete recovered 2Much improved 3Slightly improved 4No change 5Slightly worsened 6Much worse 1Improved (1+2) 2Unchanged (3+4+5) 3Deteriorated (6) 3+4+5 2 1+2 vs 3+4+5 Current study 1Asymptomatic 2Improved 3No change 4Worsened 1Asymptomatic 2Improved 3No change 4Worsened 3 2 1+2 vs 3+4 *: Ostelo RW, de Vet HC, Knol DL, van den Brandt PA. 24-item Roland-Morris Disability Questionnaire was preferred out of six functional status questionnaires for post-lumbar disc surgery.J Clin Epidemiol. 2004 Mar;57(3):268-76. **: Van der Roer N, Ostelo RWJG, Bekkering G, van Tulder MW, de Wet HCW. Minical clinically important change for pain intensity, functional status and general health status in patients with nonspecific low back pain. Spine 2006;31:578-582. In accordance with the reviewer’s comment, this has been further clarified in the updated version of the manuscript, which now reads (Discussion, second to last paragraph): In this study, patients’ own classification was rated on a 4-point scale (“completely recovered”, “improved”, “unchanged” or “worsened”), while other studies have used 5 or 7-point scales to that end, in which “improvement” and “worsening” were split into further categories, such as “much improved”, “slightly improved”, “slightly worsened” or “much worse”. However, those categories must usually be collapsed at the analysis phase, so it is up to researchers to decide how to group them. On the contrary, we preferred for patients to rate their own evolution in the categories that were going to be analyzed. This may have led to patients who perceived an improvement as clinically irrelevant selecting the “unchanged” category. Taking into account the objectives of this study, we find that to be suitable. 1.1.8 Methods (Analysis) In the analysis of optimal cut-point did the authors actually select the cut-off point where sensitivity and specificity were equal? Typically the cut-off point selected is the one that maximizes the specificity and sensitivity, not necessarily the point at which these values are equal. Please clarify. This was a mistake in reporting in the previous version of the manuscript. In fact, as the referee points out, we had defined the optimal cutoff point as the one that maximizes the sum of specificity and sensitivity, and not the one in which sensitivity and specificity were equal, as was incorrectly stated in the previous version of the manuscript. In accordance with the reviewer’s comment, this has been corrected in the updated version of the manuscript, which now reads (Methods, Analysis, paragraph 4, point 3.): The optimal cutoff point was estimated by the point that maximizes the sum of specificity and sensitivity. 1.1.9. Methods (Analysis) In the text in this section the authors state that they will analyze all patients and those patients with arm pain as a separate analysis. In the results (for example Table 4) it appears that the authors examined all patients and those with neck pain only as a separate analysis. Please clarify. In accordance with the reviewer’s comment, this has been further clarified in the updated version of the manuscript, which now reads (Methods, Analysis, paragraph 6): Data from all recruited patients (both with and without AP) were included in the main analysis, in which MCIC values for neck pain were calculated. In a subgroup analysis, only patients with referred pain at baseline were included, and MCIC values for neck and referred pain were calculated. 1.1.10. Results (first paragraph) The text in this paragraph reports that 487 patients (74%) had AP. Table 1 appears to indicate that 487 patients (74%) had NP only. Please clarify this discrepancy. In accordance with the reviewer’s comment, the typo in Table 1 has been corrected, and it now reads: All included patients N=658 Patients with referred pain N=487 1.1.11. The text in this paragraph reports that Table 3 include those patients with "NP only in patients who also reported AP". This statement is unclear. It appears that Table 3 includes all subjects included in the analysis. Please clarify. In accordance with the reviewer’s comment, this has been further clarified in the updated version of the manuscript, which now reads: Results, paragraph 3: Table 3 shows the MCIC for neck pain estimated in all included patients, while Table 4 shows the MCIC values for both neck and referred pain only in patients who also reported referred pain at baseline. As seen in those tables, MCIC values range between 1.5 and 6.2 PI-NRS points, depending on the method used to estimate them, with ROC leading to the smallest values.