The Danish Child and Adolescent Musculoskeletal Pain (ChiBPS) Cohort: Protocol for a Prospective Nationwide Cohort Study in General Practice

This study is a prospective cohort study. Setting: Danish general practice clinics across Denmark. We have no interventions for this study. Patients will be followed from their initial pre-recruitment GP consultation and onwards. The pre-defined follow-up assessments are 3, 6 (primary), and 12 months, but the cohort will be followed continuously onwards every 1-5 years.

The aim is to recruit a sample that represents the Danish child and adolescent population with musculoskeletal pain.

The GP is suggested ways of recruitment, in order to choose which best suits the infrastructure of the individual clinic.

Recruitment by an employee:

The employee executes the recruitment process with a screen of all scheduled patients for that day. This prior to the first consultation of the day. The eligibility screening has two criteria: 1. Cause for consultation including a MSK pain complaint and 2. Age 8 to 19 years. The word MSK is written next to the names of all eligible patients to indicate and remind the employee or the consulting GP to hand out the questionnaire to the patient, before they leave the clinic.

Recruitment by the GP:

The GP screens the schedule for the day, and indicates eligibility with the word MSK written next to the patient name.

When consulting a patient with the indicated MSK, the GP is reminded to hand the patient the patient material. This can be done, prior, during or in the end of a consultation while the patient is getting dressed.

The patient hands in the completed questionnaire to an employee of the clinic, who stores it in a locked place. NP or her assistant will collect this in agreement with the clinic.

The child must first be included in the study in order for his/her parent(s) to be eligible. Of included children, data will be requested from parents during a time window of two weeks from the time of Part-Quest (participant questionnaire). The investigators are not selective as to if and how many of the parents wish to participate. Parents declining participation will not cause withdrawal of the childs' eligibility or inclusion.

Participants will be followed from their initial GP consultation initiating the inclusion and onwards. The pre-defined follow-up assessments are 3, 6 (primary), and 12 months, but the cohort will be followed continuously onwards every 1-5 years. At follow-up, a research assistant working together with and supervised by NP (Negar Pourbordbari) will contact the participants on phone, to remind them of the follow-up questionnaire they will receive by e-mail. A similar process will be repeated at all follow-up time points.

All included participants will be registered at eligibility screen, acquired written consent, and assessments, the latter collected at both baseline and follow-up time points.

With the primary outcome the investigators aim to capture pain that interferes with their everyday lives and collect the location of pain. The patient characteristics (exposures) are based on results from our systematic review, used in other studies (when including references).

The specific wording in the outcomes and throughout the questionnaire has been pilot tested among similar aged kids and adolescents.

The sample-size in this study was determined using the two following rationales: 1) a sample-size large enough to test and replicate the analyses from previous studies on. Given the prior odds (0.5, 1, 2) of follow-up MSK pain for patients, using estimates for the prognostic factors female sex, high disability index, multisite pain, and maximum HFAQ (Hannover Functional Ability Questionnaire) from our systematic review. The investigators gained an estimate of P-values according to sample size, for all factors individually (Appendix 8). Sample size of 500 patients, would result in an estimate of P-values below 0.05 for all prognostic factors and 2) investigate a range of new prognostic factors related to the sparsely investigated ethnicity and socioeconomic status. As no one has yet tested any of these potential important prognostic factors, and never in a general practice setting, the investigators decided on 500 subjects. The 500 subjects was decided based on 250 cases (50% will assume to continue to experience pain at our primary follow-up time) giving approximately 125 cases per prognostic factor (500/number of prognostic factors). The results from this analysis will be considered as explorative as no studies have previously been conducted in a general practice setting. Again, assuming 50% has pain at follow-up and 20 events for each to be tested factor are needed. The model should allow for 10-15 variables in a multivariable logistic regression analysis. This would lead to a sample size of 480. Assumed, 5% is lost to follow-up and we have a sample-size of 504 included participants. Given the high prevalence of MSK pain at follow-up, this sample-size is feasible to include.

The investigators use three questionnaires in data collection of patient and parent characteristics: participant baseline questionnaire (Part-Quest-base) (Appendix 2), participant follow-up questionnaire (Part-Quest-follow-up) (Appendix 3), and the parent questionnaire (Pare-Quest) (Appendix 4).

In relation to our previous study, all these factors have an interest as to whether they present as prognostic factors in MSK pain at follow-up.

The questionnaires will be pilot tested among target participants; Part-Quest among 8-19-year-olds with MSK pain and the Pare-Quest among parents to children with MSK pain.

Our questionnaires wil be cross-cultural adaptive for participants with poor Danish language skills, insufficient to comprehend the original Danish questionnaires. In the process of translation and validation, we will translate our questionnaires both linguistically and culturally. Our three questionnaires and our bulletin will be translated from English to Danish, which is the mother tongue of the majority of our prospective participants.

The questionnaires, will be handed to the participants and their parents in paper. If they are not able to complete the questionaires while still in the clinical setting (waiting room) and hand them to the secretary when completed, they may at a later point in time (no later than seven days) hand them in at the clinic. Seven days of response time, will be applicable also for follow-up questionnaires, before a reminder is sent out. In contrary to the baseline questionnaire, the follow-up questionnaires will be administered to the participants and parents through e-mail. If for some reason a participant do not respond within the given time limit, he/she will be reminded by e-mail or phone. This will be executed by the assistant, to ensure a satisfactory follow-up rate.

Retrieved data from both baseline and follow-up participant questionnaires and parent questionnaire will be handled according to the Danish Data Protection Agency.

Descriptive results will be stratified by sex and MSK pain type, presented with their central estimate and appropriate measure of dispersion (95% confidence interval CI). All descriptive statistics and test will be reported in accordance to the recommendations of the STROBE statement.

We will test the univariate association between each potential prognostic factor as well as combining the factors in a multivariable model to show the individual contributions.

The primary analysis will be done on those who has a first time consultation with their GP concerning their current MSK pain, but all consultations will be included.

The full statistical analysis plan will be developed concurrently and will be finalized before the last patient is enrolled. Information bias will be minimised by use of identical questionnaires for all patients included, e.g. by asking both Danes and non-Danes whether they feel Danish or not.

The adolescents included in the ChiBPS cohort will receive usual care, i.e. as usually administered at their GP. This study will not intervene in the diagnostic and therapeutic process.

Since our target audience is general practitioners, our prognostic factors must be applicable in a GP setting in relation to the terminology they are presented in as results of this study. In order to gain recognition of this, NP will create a temporary subgrouping, based on the prognostic factors from our recent systematic review. She will convey this subgrouping to a focus group consisting of 15 clinically experienced, Danish GP physician peers for feedback on the suggested terminology in subgrouping of the prognostic factors. She will ask for any concerns in comprehension including any suggestions towards an easy digestible language, in the context of general practice.