How do people with chronic pain explain their use, or non-use, of pain-related healthcare services? A qualitative study of patient experiences

Abstract Purpose This study aims to explore how people with chronic pain explain their use or non-use of pain-related healthcare services and their expectations of the healthcare provider, and explore how explanations and expectations vary between different levels of pain-related healthcare use. Materials and methods We conducted 20 individual semi-structured interviews with purposely sampled adults between 39 and 77 years of age with chronic pain. All interviews were audio-recorded, transcribed and analysed using a thematical template analysis approach. Results Four key drivers for pain-related healthcare use were identified: (1) the healthcare system facilitates clinical pathways, (2) appraisal of pain-related healthcare initiatives influences future use, (3) autonomy, beliefs and values determine healthcare behaviour, and (4) recommendations from others impact healthcare behaviour. Comparing explanations across different pain-related healthcare user groups (high, medium and low) showed that perceived needs, beliefs and values, and appraisal of previous healthcare experiences differed between these groups. Conclusions Beliefs, pain characteristics, recommendations, and the search for a diagnostic label, often initiate pain-related healthcare use. Healthcare is modified by two interconnected systems: (1) perceived needs, beliefs and values and (2) previous healthcare experiences. Differences related to these systems could explain some of the variance in pain-related healthcare use. Implications for Rehabilitation Different use of pain-related healthcare services for people with chronic pain could be related to differences in perceived needs, beliefs and values and appraisal of previous healthcare. It may be helpful to explore the perceived needs and beliefs of those seeking healthcare due to chronic pain. Previous healthcare experiences may impact and should be explored.


Background
Chronic pain is common and impacts behaviour, physical activity, social function, and quality of life [1,2].A common behaviour in the context of chronic pain is seeking healthcare and/or taking pain medication, although not all with chronic pain do so [3,4].About 40% of people reporting chronic pain use no, or very few, pain-related healthcare services annually, whereas just under 10% have a continuous high use of pain-related healthcare services (more than 20 annual contacts/prescriptions for pain-medication) [5].Several factors have been associated with higher use of healthcare services by people reporting chronic pain, including: older age [6-8], low socioeconomic status [6,8], being female [7-9], high pain intensity [9][10][11], high level of disability [9, 10,12,13], number of pain sites [12,14], comorbidity [6], higher body mass index (BMI), poor general health [6,7,15], negative health beliefs (health anxiety, catastrophising, fear avoidance) [14][15][16], and previous healthcare use [15,17].However, the choice of care-seeking due to chronic pain or taking pain medication is more complex than the presence or absence of certain factors and such factors do not explain why people with chronic pain have different healthcare-seeking behaviour.
Others have suggested that healthcare-seeking is the result of an interpretation process, where pain sensations are transformed into symptoms and signs of potential illness in a complex interplay between biological, psychological and cultural factors [18][19][20].Studies of first-person narratives of people reporting chronic pain have shown that some people with chronic pain experience repeated treatment failures, leaving them feeling hopeless, dislocated from the world and themselves, and trapped by the pain [21][22][23].Some seek healthcare in search of an explanation or a diagnosis to validate the pain and for some this search continues for a long time [22].In this endeavour, some come across healthcare professionals (HCPs) who appear skeptical and lack understanding of their situation, leaving the person with chronic pain feeling stigmatised and disbelieved, which gives rise to feelings of anger, frustration and despair [21].However, it is still unclear if such experiences account for the differences in the use of healthcare due to chronic pain.Therefore, to further understand the use of pain-related healthcare due to chronic pain, it would be helpful to gain insights from the lived experiences and perspectives of people with chronic pain who have different painrelated healthcare-seeking trajectories.A better understanding of healthcare experiences and the beliefs and processes underlying use or non-use of pain-related healthcare services for people with chronic pain could be a key input to inform the optimisation of chronic pain management.Therefore, the key questions underpinning this qualitative study are: (1) how do people with chronic pain explain their use, or non-use, of pain-related healthcare services and what are their expectations for the HCP they approach?and (2) how do such explanations and expectations vary between people who have different levels of pain-related healthcare service use?

Design and setting
This is a cross-sectional qualitative study seeking to describe and understand use of pain-related healthcare in people with chronic pain.This study was conducted in Denmark, and hence the primarily publicly-funded healthcare system in Denmark forms the backdrop to this study [24].Figure 1 is a graphic illustration of the organisation of current pain-related pathways in the Danish healthcare system.

Participants and recruitment
Informants in this study were drawn from a previously established Danish longitudinal cohort study.Details about recruitment, nonparticipants and characteristics for the entire cohort have been reported in detail elsewhere [25,26].In brief, the cohort is a population-based cohort established in 2008 and consists of 5068 adults (17-64 years of age in 2008) who responded to a baseline questionnaire sent to 8517 adults in 2008.The baseline questionnaire collected data on a wide range of socio-demographic, quality of life, health, work-related, lifestyle and psychosocial factors.A web-based follow-up questionnaire was distributed digitally in October 2020 to the 4865 responders of the baseline questionnaire who were still alive and registered in the Danish mandatory secure public mailbox system, in which about 93% of the cohort were registered at the time of follow up [27].The follow-up questionnaire collected data on health, psychosocial, quality of life and work-related factors.A total of 2947 individuals answered the follow-up questionnaire.To ensure that all informants had a history of long-term pain, we only considered participants who reported pain for longer than 3 months in both 2008 and 2020 (measured by a single item from the Standard Evaluation Questionnaire [28]) and who agreed to be contacted for interview in the 2020 survey.This left us with 659 potential eligible informants.

Data from questionnaires and registers
To select and describe the informants, we used variables collected by questionnaire in 2008 and 2020, and selected national registers.Sex and age were obtained from the Danish Civil Registration System [29,30].The number of pain sites in 2008 and 2020 was measured with the Standard Evaluation Questionnaire, section 1 [28].Participants were asked to rate their pain intensity (1 ¼ no pain, 7 ¼ worst imaginable pain) in seven different body locations (head and face, left upper extremity, right upper extremity, chest and stomach, neck and back, left lower extremity and right lower extremity).We considered a person to have pain in a particular region if they self-rated pain intensity in that region as �2.Health anxiety in 2008 and 2020 was measured with the Whiteley-7 Index (range 0-28) [31] and dichotomised at >5 (0-5 ¼ low risk, 6-28 ¼ high risk), as that threshold is recommended by The Danish College of General Practitioners as an indication of a clinically-relevant risk of health anxiety [32].Labour market status was measured with a single-item question.
Participants were asked to select their work status from one of the following 10 nominal categories: work, temporarily absent from work due to illness, student, health-related work placements, temporarily absent from work due to leave, trainee/apprentice, social assistance recipient, unemployed, retired, or other.
Self-rated work ability was also measured with a single-item question.Participants were asked to rate their work ability on a ten-point Likert scale, where 0 indicated unable to work and 10 indicated full work ability.Well-being was measured with the WHO-5 Well-Being Index (range 0-100), where 0 represented worst and 100 represented best imaginable well-being [33].The Danish population norm is 68.Musculoskeletal health status was measured with The Musculoskeletal Health Questionnaire (MSK-HQ).This scale consists of 14 items and is a validated scale for measuring musculoskeletal health across musculoskeletal conditions and settings [34].For scoring, all 14 items were summed, with the sum score ranging from 0 to 56.Higher scores indicated better musculoskeletal health.

Selection of informants
To ensure a variety of voices represented the range of factors commonly associated with higher prevalence of chronic pain and higher use of healthcare services, we sought to select and invite informants with different profiles in terms of gender, age, number of pain sites (2008 and 2020) and health anxiety (2008 and 2020).Potential informants were initially contacted by phone by one of the authors (SM) and invited to participate in an interview.If a person declined participation (3 cases), an informant with a similar profile was contacted.If initial participation approval was given, informants were mailed information about the study and an informed consent form.If the informant still agreed to participate, an interview date and time were scheduled.Data collection was ceased after 20 interviews as sufficient information power to give a nuanced picture of the informants' perspectives and understanding of the research question had been obtained.At the beginning of each interview, the interviewer made sure that none of the following exclusion criteria applied: evidence of severe malignant pain or any other serious pathology or severe comorbidity (e.g.cancer, severe chronic obstructive pulmonary disease or cardiovascular disease), cognitive impairment that prevented them from providing meaningful responses at interview (e.g.currently being treated for a psychiatric disorder, senile dementia, Alzheimer's disease), and being unable to speak Danish or English.After both the fifth and fourteenth interviews, a group of authors (SM, HRS, and AS) evaluated the data collection and interview profiles, and agreed on how to progress the recruitment to obtain the most information power.In this evaluation, we included additional variables [labour market status (2020) and well-being (2020)] and explored if informants who had already been interviewed also had diverse profiles related to these variables.This led to alterations in recruitment.For example, we primarily invited informants with relatively few pain sites and a low well-being index score for the final interviews, to explore if such profiles could shed light on new perspectives about pain-related healthcare-seeking behaviour.

Qualitative data collection
To accommodate government recommendations for the COVID-19 pandemic, all individual interviews were either conducted online (n ¼ 2) or over the phone (n ¼ 19).Interviews were conducted and audio-recorded by a male physiotherapist and PhD student (SM) who received training from experienced qualitative researchers (HRS & AS).The interviewer was not previously known to any of the informants, nor involved in their treatment.Interviews lasted between 30 and 90 min.To explore experiences, beliefs and thoughts underlying healthcare use, a semi-structured interview guide was developed with open-ended questions about the pain, pain-related healthcare experiences (including pain medication and alternative healthcare), drivers for healthcare-seeking and beliefs about pain (The interview guide can be found as supplementary material).The interviewer aimed to remain flexible to explore new concepts as they arose.At the beginning of the interview, informants were asked to estimate their number of pain-related healthcare contacts during the previous year (GP, physiotherapist, chiropractor, complementary and alternative medicine, hospital, and emergency room healthcare contacts) and their use of pain medication.Additionally, informants were asked to recall if this pattern of healthcare use had changed over the last 5 years.This information was evaluated by one of the authors (SM) and grouped into one of the following three categories to comply with previously identified long-term musculoskeletal healthcare trajectories [5]: low (no or very few annual musculoskeletal healthcare contacts), medium (five to 15 annual musculoskeletal healthcare contacts) and high (more than 15 annual musculoskeletal healthcare contacts).
Next, informants were asked to: � Tell their pain story.� Explain their beliefs about their pain.� Describe pain-related healthcare experiences (if any), including thoughts/beliefs/experiences underlying selecting or deselecting healthcare and/or different treatment modalities.� Suggest any advice to other people with chronic pain and HCPs.
Before data collection for the study, we conducted three pilot interviews.These were recorded, transcribed, read and discussed by two of the authors (SM and HRS) to make sure that questions were understandable and that answers reflected topics of interest, which led to small adjustments to the interview guide.

Qualitative data analysis
All audio recordings were transcribed by a research assistant with extensive experience in transcription and competence in qualitative research.Data analysis was carried out in two rounds.Round one followed the process of thematic template analysis as described by Brooks and colleagues [35] in the following stages: (1) reading and re-reading transcripts for familiarisation, (2) preliminary coding, (3) creating themes based on preliminary codes, (4) defining an initial coding template based on a subset of data, (5) applying the initial coding template to further data and modifying accordingly, and (6) finalising the coding template and applying it to the full dataset.The team of authors decided not to define any a priori coding categories in order to remain open and, via an inductive process, to conceptualise meaning based on the raw data.Four authors (SM, HRS, AS and CRB) were responsible for the analysis.This group consisted of physiotherapists with extensive clinical and teaching experience, and expertise in both qualitative and quantitative research.Data collection and analysis were an integrated process where initial analytic steps informed and qualified later sampling rounds, data collection and analysis.Three authors (AS, HRS and SM) independently read and coded the first five transcripts to develop the initial coding template (steps 1-4).Next, one author (SM) applied the initial coding template to the next nine transcripts and refined it, as new aspects of meaning were recognised in the data.To qualify this process, three authors (SM, HRS and CRB) independently read and coded four selected transcripts informed by the refined coding template.Interpretation of the coding/themes/sub-themes development and organisation were then discussed, and agreement reached before the coding template was applied to the final six transcripts.Finally, themes and subthemes were discussed and challenged by all four authors involved in the analysis, until consensus was obtained, before the final coding template was applied to the full dataset.
In round two, we explored how the identified themes differed across levels of pain-related healthcare use.This was undertaken by comparing the coding across groups of pain-related healthcare use (low, medium, high).In this process, two authors (SM and CRB) individually reread the thematised codes for each group of pain-related healthcare use and prepared summaries with condensed aspects of meaning for each group.After this, the summaries were reviewed, discussed and challenged until consensus was obtained.The main results will be presented as condensed aspects of meaning and each theme/subtheme will be accompanied with example citations.Last, the results will be contextualised to enhance the clinical interpretation and we will discuss the results based on this contextualisation.All transcripts were uploaded and analysed using NVivo 13 (QSR International, Melbourne, Australia).

Ethical aspects
This project was approved by the Danish Data Protection Agency (project number 1-16-02-141-18).Participation was based on informed consent and we only contacted informants who had a priori agreed to be contacted.All data were stored and handled according to Danish law and will be deleted by the end of the project.According to Danish law, this type of study does not require approval by a biomedical research ethics committee [36].

Characteristics of informants
Informants' profiles are presented in Table 1.Eleven women and nine men were interviewed.One informant was interviewed twice, as familiarisation with the data revealed that more clarification was required.Ten informants were working, eight were retired, while one had a "flex job" (a form of reduced capacity job) and one was not working due to other health conditions.Their ages ranged from 39 to 77 years with a mean of 58.The number of pain sites ranged from one to seven with a median of three in both 2008 and 2020.Fourteen informants had low health anxiety scores in 2008 and nine in 2020.Self-rated workability ranged from 1 to 10 with a median of 8.The Well-being Index scores ranged from 16 to 92 with a mean of 54 and Musculoskeletal Health Questionnaire score ranged from 16 to 49 with a mean of 38.Eight informants were classified as "high" healthcare users, seven as "medium" users and five as "low" users.

Explanations for use, or non-use, of pain-related healthcare services and expectations of the HCP
Regarding the outcome of the analytic process for the first question, four key themes were identified: (1) the healthcare system facilitates clinical pathways, (2) appraisal of pain-related healthcare initiatives influences future use, (3) autonomy, beliefs and values determine healthcare behaviour, and (4) recommendations from others impact healthcare behaviour, as illustrated in Figure 2.Each theme is described below, with sub-themes and supporting quotes.Each quote is marked with pseudonym and acronym that indicates the healthcare user category (_low, _med or _high).

The healthcare system facilitates clinical pathways
This theme covers explanations for the use of pain-related healthcare related to the organisation, traditions and procedures of the healthcare system.Use of pain-related healthcare and clinical pathways are often defined and driven by the healthcare system itself and the healthcare system both drives and modifies use of pain-related healthcare services.
The pathways in the healthcare system.Patients with chronic pain move in system-defined sequences back and forth between the different specialties and sectors involved, depending on the results and reports resulting from examinations and treatments.Patients need to comply with these system pathways when seeking pain-related healthcare, also in situations where other HCPs and treatments are preferred, and these system pathways exhaust them.
I17_med: … but I also felt that you were being thrown around because everything is so divided.You have to wait for x months to get to a rheumatologist whom you hope can help you.And once you have been there, you have to go back to your own GP to get a new referral to someone else who might be able to help.And that person cannot help either, so you spend a really long time getting through the system, and that's frustrating.

Referrals.
Referral is a common driver of pain-related healthcare.Often patients with chronic pain follow the clinical pathway laid out by the GP, and many have trust and faith in the evaluation and expertise of the GP.

Appraisal of pain-related healthcare initiatives influence future use
All pain-related healthcare initiatives are constantly evaluated by the person with chronic pain and the outcome of this appraisal process modifies the use of pain-related healthcare services.

Previous healthcare experiences create "treatment prototypes".
Previous healthcare experiences shape or create treatment prototypes (patterns) that influence future clinical pathways and the use of healthcare services.Treatment prototypes are based on evaluation and appraisal of previous experiences and can both facilitate ongoing/recurrent use of pain-related healthcare or reduce healthcare use.Change in symptoms or function and/or trust and relationship with the HCP are often the outcome measures against which pain-related healthcare initiatives are evaluated.
I12_High: I have been using the same chiropractor since 1992.And he knows exactly how to unlock my back.I can go there 5 -10 times, and then my back is fine again.
Treatment prototypes can also be self-management in cases where useful healthcare guidance has led to empowerment and self-efficacy or in cases where self-initiated actions have led to improvement.

I20_Low: I use my body in my daily life and I think that it really helps. It has always helped me. I remember when the pain in my back was at its worst, then I got a part-time job on the weekends and during the holidays with my cousin, who is a bricklayer and I really, really used my back quite extremely. So, when I came home from work and I sat down under the shower and whoops all the pain in my back disappeared. And it was simply because my muscles around my back just got stronger, and then that back pain stopped …
Patients with chronic pain desire to be seen, heard, and approached with interest by the HCP.Often the person with chronic pain does not expect a miracle cure for pain when approaching the healthcare system, but they hope for a thorough clinical examination, an explanation for the pain, useful advice and reassurance.If such expectations are not met, they can become frustratingly resigned to their condition or their uncertainty can be the driver of ongoing healthcare.People with chronic pain look for a HCP who approaches them with interest, with an appropriate level of empathy, from a holistic standpoint, and with sufficient time to listen, see and explain.However, experiences are common where a person with chronic pain has felt a nuisance and perceived a lack of interest from the HCP.Such experiences can lead to resignation or further searching for alternative healthcare options.If the patient gets a satisfactory examination and relevant information, then they can accept the situation and adopt active/self-management strategies, instead of invasive treatment.
I14_high: … I thought I was going to have surgery and it was just a matter of time, but I actually came out knowing that I should not have surgery.This, of course, was totally different than I expected but I was really confident that it was ok that I should not have surgery because he (the orthopaedic surgeon) explained and put it in perspective in a way which was understandable.And you were met, and you were, well, not looked down on, but you were recognised that you have an injury that was not just something … -It was not just a sprain.

Perceived improvement and alliance with the HCP influence healthcare.
To be worth the effort, pain-related healthcare has to cause some improvement and provide relevant rehabilitation opportunities.If not, the person with chronic pain may cease it and/or search for alternative healthcare options.

I5_med: … it occurred to me, well, now I have gone there for so long and it has actually not had any effect. And I just did not book a new appointment.
Furthermore, if the person with chronic pain does not experience a therapeutic alliance with the HCP, then healthcare may be ceased.
I1_high: … and then you have another therapist, and it does not work at all.Or the chemistry is just not there.And then you say no, then it's probably best we stop.

Autonomy, beliefs and values determine healthcare behaviour
Pain-related healthcare use and clinical pathways for people with chronic pain are often determined and modified by preferences, beliefs and underlying assumptions about cause and control, as well as personal values (within the constraints of the healthcare system or their financial options).

Healthcare should fit my needs, beliefs and values.
Beliefs and assumptions about healthcare often outrank the HCP's expertise and evidence.For example, healthcare initiatives that are trusted and perceived as needed will be well-regarded, despite being explicitly not recommended by their HCP or lacking in evidence.

I8_low: … and it may well be that scientifically it has not been proven that massage helps, that it is only for well-being, like my doctor said. But I can feel that it helped … I10_high: … whether it's alternative or the established medical sciencewell, I kind of don't care. I lean towards the conclusion that if it is something that can help me in my daily life and my situation, well, then I gladly accept it.
People with chronic pain often have their own rationale/thinking framework for how to control or relieve the pain and if suggested healthcare initiatives do not fit within this framework, they become skeptical or disregard it and maybe take the initiative to go somewhere else to find another HCP.Sometimes, the person with chronic pain bypasses traditional clinical pathways to get the help they want.
I12_high: … and then I got that injection, and it was simply glorious, and considering how awful it was last time … then I am glad that we called the ambulance this time.Because if I had waited and let it calm down, and then had to contact my GP the next day, and it is hopeless.The GP, they are … it is completely hopeless.And then you have to be referred for an X-ray and it takes days and weeks.No, then, it was much better that we did this.Some take control over their situation and suggest initiatives they have heard or read about.Others, in frustration or anger push the GP to act and initiate additional healthcare initiatives.
I13_high … then I said, "I have to tell you one thing", hm, "what is it then", she said (The GP)."We have a problem both of us"."Well, what was it", she said."We have to do something about this pain somehow, and now it's up to you to figure out what to do because you are a doctor"."Well", she said then she chewed on it a bit and said, "well, I think I'm sending you to the pain clinic"."Okay, then we'll try that".
A person with chronic pain can also disregard healthcare due to a lack of trust or their own belief that the pain is self-inflicted and caused by personal weakness and inactivity.Such beliefs may initiate overuse of physical activity strategies, causing a vicious cycle with increasing pain and despair.

Healthcare has to be manageable/feasible for me in my situation.
Perceived limitations and resources are important for healthcare behaviour.If the available healthcare initiative is perceived to be expensive, unhelpful, if it becomes stressful/unmanageable, or if the time demands or perceived risk of side effects exceeds the potential benefits, then that healthcare initiative may be disregarded or dismissed, despite the referral/recommendation.
I13_high: I should have been attending a 13-week course, where I had to drive down there regularly (Specialised pain rehabilitation clinic) and I simply could not manage it.And then they finished me off.[S: So, why did it end?]Well, that was probably because I had not signed up for that course (pain education and cognitive therapy).But, I had to drive almost 100 km back and forth once or twice a week; I simply could not overlook (that I would need) to sit in the car for so long.… I simply could not, I could not drive there twice a week, I just could not, it stresses me to be honest.
To be practical, healthcare due to chronic pain also has to be appealing and financially/geographically convenient and timely.Such practical aspects are sometimes beyond the influence of the person with chronic pain but still hinder relevant healthcare options.What works and fits from a practical point of view for a person with chronic pain is important, and practical aspects can both facilitate or inhibit the use of pain-related healthcare services.

Biomechanical beliefs about pain drive healthcare behaviour (cause and control).
For some, beliefs about causes of pain grounded in a biomechanical paradigm (work/physical exposures, mechanical defect, injury, age or morbidity) and imaging are considered a valid source of diagnostic information.Feelings of being fragile and/or permanently damaged often accompany such beliefs.Such a conceptual framework drives healthcare in the quest for diagnostic certainty or the endeavour to fix biomechanical damage.In some cases, such perceived biomechanical damage triggers thoughts about inevitability of future surgical procedures.
I9_med: I have used a chiropractor in the past, because I have had some joint slips because of high jump and gymnastics.Now my own GP can just pull me back in place.I've just visited him because I had acute back pain and he could just pull it into place.
I10_high: … the reason I got up there (Alternative therapy -Craniosacral Therapy), it turned out, that my skull was sitting wrongly on the outer cervical vertebra.It simply sat in the wrong place.

The pain and/or functional limitations necessitate healthcare.
Pain flairs or functional limitations influence the perceived need for healthcare and may call for action to get through the day -or through the night.In some cases, the pain can make the person with chronic pain desperate and they may see no alternative to healthcare.Often such healthcare initiatives are taken without consultation with a HCP.Pain medication or alternative medicine is frequently used in such situations but also can involve contacts with alternative HCPs or primary care HCPs.

Recommendations from others impact healthcare behaviour
A recommendation can be a powerful driver for trying different and new pain-related healthcare modalities, including alternative medicine and therapy.
Recommendations from relatives/others about healthcare.People with chronic pain often experience relatives or other caregivers eager to assist in finding ways to relieve the pain.Such recommendations and advice are often trusted and followed.
I7_high: … what mattered to me was that I had heard from others that she (alternative therapist) was really good.And that's what matters to me.

Recommendations from HCPs.
In some cases, a patient is recommended alternative treatment by HCPs.HCPs can also influence healthcare by recommending to cease or pause a course of care.Ceasing healthcare can also be based on a joint decision between the HCP and the person with chronic pain.

Differences in explanations and expectations between groups of pain-related healthcare use
Some themes were evident across all groups of pain-related healthcare use.For example, reports of: i) disappointing healthcare experiences and pain-related healthcare initiatives with no or little effect, ii) the desire to be seen, heard and approached with interest by the HCP, iii) beliefs that pain is caused by biomechanical damage and imaging is required as it is a valid source of diagnostic information, and iv) skepticism towards medication (event though high healthcare users often take it because it is considered necessary), were similar across groups.
Considerable differences in the expectations about pain/ healthcare/the HCP, and pain-related management strategies (treatment prototypes) were recognised between groups of painrelated healthcare use.The "low" healthcare user group expressed a lack of faith in the usefulness of healthcare and often prefer to self-manage their pain.They use healthcare to rule out serious conditions and pain is considered to be a natural part of life and hence not dangerous.The "medium" group have multiple pain-related healthcare experiences.They express faith in the healthcare system and healthcare management strategies are characterised by pain medication and conventional healthcare.They expect the HCP to be a mentor or collaborator who honestly and empathetically guides them and provides information -including if nothing can be done about the pain.However, such guiding is not always provided which leaves the medium healthcare user feeling their direction is unresolved.The "high" group, by contrast, expected the HCP to be a helper or a partner who provides healthcare that complies with their expectations and needs.Where the HCP does not accommodate this, attention will be directed towards other healthcare options.They seek an explanation for their pain.This group has multiple, often dissatisfying, pain-related healthcare experiences, and treatment prototypes include both conventional and alternative healthcare management strategies (including pain medication).

Main results
To better understand the use of pain-related healthcare due to chronic pain we have contextualised themes and sub-themes in the discussion and built the discussion based on the contextualised interpretation of the results.The following paragraph is our attempt to translate explanations for use, or non-use, of painrelated healthcare services into context in order to better understand the differences in healthcare behaviour.
People with chronic pain explained that beliefs about their natural history/clinical course and pain control, their pain and functional limitations, recommendations from others, and the search for a diagnostic label/reassurance, often initiate painrelated healthcare use.Once a person with chronic pain has decided to enter the healthcare system, the referrals given by the HCP, and the pathways laid out within the healthcare system, are often followed if such pathways are considered to be feasible.Healthcare due to chronic pain is highly influenced and modified by two interconnected systems.Firstly, perceived needs, beliefs and values are an important modifier of pain-related healthcare use.For example, pain medication will often be disregarded, despite being recommended, if the person with chronic pain is skeptical about it or if side effects are considered to outweigh the benefits.Similarly, alternative therapy/medicine might be selected as a pain management strategy if believed or experienced to be effective, despite recommendations against or no scientific evidence for effect.Likewise, biomechanical beliefs (e.g.beliefs that a displaced lumbar vertebra is the cause of back pain) can trigger perception of a need for certain treatments (e.g.lumbar manipulation) and examinations (e.g.scans).However, if the pain is not considered to be a threat and/or the person lacks faith in the benefit of healthcare, then "wait-and-see" or self-management strategies might be preferred.Secondly, previous healthcare experiences and the appraisal of such experiences are another powerful modifier of future healthcare use.The appraisal process can, for example, lead to treatment prototypes characterised by continuous or recurrent use of pain-related healthcare in cases where needs and expectations are met (e.g. for pain relief), and beliefs and values match the management provided.In contrast, that appraisal process can lead to self-management or wait-and-see prototypes if the healthcare provided did not match the needs, beliefs and values of the person with chronic pain.The desire to be seen, heard and approached with interest by the HCP is essential for people with chronic pain, but expectations of the HCP differ between healthcare user groups.High healthcare users seek HCPs who comply with their understanding and needs, medium healthcare users search for HCPs who honestly and empathetically guide and provide relevant information while low healthcare users lack faith in the usefulness of healthcare.When seeking healthcare, the care-seeker will assess whether the HCP meets this expectation and that assessment will influence future healthcare.These interconnected modifier systems could explain some of the different pain-related healthcare-seeking trajectories of people with chronic pain, as the explanations across levels of pain-related healthcare use show considerable differences related to those factors.

Discussion of results
People with chronic pain use healthcare services more often than people without chronic pain [37,38] but not all people with chronic pain have the same healthcare-seeking behaviour [5].In this study, we attempted to capture some explanations for seeking or deselecting pain-related healthcare and our analysis showed that certain factors initiate new pain-related healthcare contacts, while beliefs and experiences modify future healthcareseeking trajectories.Leventhal et al. introduced "The Common-Sense Model of self-regulation" (CSM) [39,40] as a theoretical framework to understand how people respond to illness.According to that framework, pain initiates a cognitive and emotional response that influences the interpretation of the pain and determines the required action.Beliefs about the cause of the pain, consequences of the pain, controllability of the pain, expected duration of the pain, together with personal experiences and external sources of information (such as that from HCPs, relatives/friends and the media), are important in that process [40].According to this evaluation, a person with chronic pain acts (which could involve seeking healthcare) and subsequently appraises the effect of this action.In this way, the CSM resonates with the findings of this study.We also found that beliefs about cause and control, functional consequences, appraisal of previous healthcare experiences and recommendations from others (external sources of information) impact pain-related healthcare use.This supports the importance of, and the potential for, HCPs addressing beliefs about the pain in order to modify how people with chronic pain respond to their condition.Our findings add to the elements of CSM by also shedding light on how healthcare system referrals and pathways facilitate healthcare use.
Studies have shown that people with chronic pain often seek alternative medicine/therapy management [41,42], which suggests that many people with chronic pain are not satisfied with conventional healthcare.This study confirms that alternative treatment options are commonly used or tried by people with chronic pain.Furthermore, dissatisfying conventional healthcare experiences were commonly reported in this sample of informants and the sources for dissatisfaction were multiple (e.g.not being engaged empathetically, not getting the help/information/examination/ medication perceived as needed).Others, who have also explored first person narratives of people with chronic pain, have also reported dissatisfying healthcare experiences, which in some cases have led to ongoing use of healthcare services [22,23].However, ongoing or high use of pain-related healthcare services is not always beneficial for patients as it can lead to overdiagnosis with unnecessary tests and treatment, in addition to challenging the provision of healthcare system resources [43,44].Lennox Thompson et al. [45] have suggested that in order to get on with life in the context of chronic pain, the person with chronic pain has to re-occupy him/herself, which involves a process of making sense of the pain, deciding to move on with life, and flexibly persisting (being committed to achieving goals in life while flexibly managing the challenges of chronic pain).On this journey, the role of the HCP is to help the person in pain to make sense of, and validate, the pain in order to empower the person and "open the door to the future" [22].The desire to be approached with an appropriate level of empathy, by a HCP intent on listening and understanding, and to experience validation of the pain experience, was put forward by almost all participants in our study.This insight has clinical implications as pain-related healthcare trajectories potentially could be modified if HCPs deliberately and intentionally approach patients in this way.The importance of such an approach and pain validation for chronic pain management has been emphasised by others [22,45].What is less often recognised is that this has not only the potential to enhance management outcomes and help a person with chronic pain embark on a healing journey towards better quality of life and function, but it could also potentially change their healthcare seeking trajectory.Our findings show that patients approached with empathy by a listening and understanding HCP are willing to accept and comply with non-invasive pain-management initiatives even though surgery was preferred initially (see: "Patients with chronic pain desire to be seen, heard, and approached with interest by the HCP").

Trustworthiness
We have aimed to establish trustworthy results worthy of attention by demonstrating reflectiveness and transparency throughout the study by using audit trails and reflexive journals [46].We have accounted for the research process and provided the readers with evidence of the decisions and choices made throughout the data collection and analysis process (see Appendix: Project audit trail).

Sampling
We purposefully selected informants from a large group of people reporting chronic pain.By that process, there was potential that we may have selected and included a group of informants with highly specific characteristics (e.g.high health literacy, which influenced their willingness to participate).However, we invited informants with very different profiles and only three declined participation, which strengthens our perception that a variety of voices are represented in this sample.
Data collection was ceased when information power for theme and sub-theme development for cross-case analysis was considered to be sufficient.For the comparison of themes and subthemes between groups of pain-related healthcare use, additional purposeful sampled informants could have potentially provided the opportunity to capture more nuances between groups.However, in this comparison we narrowed the aim by exploring already developed themes/sub-themes, and we focused on highly specific informant characteristics, where the diversity of experiences was likely to be less diverse.We therefore believe that the included informants still provide access to useful and relevant insights with high information power [47].Readers should, however, be aware that this sample may not be representative of the entire chronic pain population (even though recruited from a representative cohort).We sought to gain a rich description of experiences that explain the complexity of seeking, or deselecting, healthcare and have proposed hypotheses about what differentiates levels of healthcare use.We acknowledge that the insights from this sample may have limited generalisability to the settings in other nations and, in interpreting the results of this study, the reader should be mindful of ways the mainly government-funded healthcare system in Denmark might differ from their healthcare system.However, we still believe that our findings could be helpful and transferred to some other settings.

Data analysis
Although we aimed to create a judgement-free environment for all informants to bring forward all aspects of pain-related healthcare, some participants might have withheld undesirable or stigmatising healthcare activities (like high use of opioids or alternative therapies), even though our impression was that informants were generally candid and frank.All interviews were conducted by the same person (SM), who was relatively inexperienced in qualitative data collection at the beginning of the study.Therefore, the first phases of data collection and analysis were governed and guided by experienced qualitative researchers (HRS, AS) to ensure data quality and data saturation.In accordance with the qualitative approach, several authors were involved in the coding and analysis process (SM, AS, HRS, CRB).In the analysis phase, we decided not to approach data through a priori specified theoretical lens or define any a priori coding categories.If a more theoretical analytic approach was applied, alternative interpretations to those presented in this article might be possible.However, our audit trail is explicit for the reader to ensure a high degree of transparency.We performed no intra-or interobserver reliability test of the coding.Instead, results were reached by discussions and agreement between authors.

Conclusion
Participants in this study all reported chronic pain in both 2008 and 2020.When invited to explain drivers for their use or non-use of pain-related healthcare, we found that beliefs about pain course and control, pain and functional limitations, recommendations from trusted persons, and the search for a diagnostic label/ reassurance, often initiate pain-related healthcare use.The referrals given and the pathways laid out within the healthcare system are often followed, when such pathways are considered to be feasible.However, healthcare due to chronic pain is highly influenced and modified by two interconnected systems: (1) perceived needs, beliefs and values and (2) previous healthcare experiences and the appraisal of such experiences.It is essential for people with chronic pain to be seen, heard and approached with interest by the HCP and whether this expectation is met will influence future healthcare.Comparing explanations between participants with high, medium and low use of pain-related healthcare services showed considerable differences in perceived needs, beliefs and values and previous healthcare experiences between these groups, which could explain some of their different pain-related healthcare behaviour.This study highlights the importance for HCPs to approach a patient with an appropriate level of empathy, with the intention to listen and understand, and validate the pain experience, as such an approach provides a HCP the opportunity to modify future use of pain-related healthcare for people with chronic pain.

Figure 1 .
Figure 1.Current organisation of healthcare management of musculoskeletal care-seeking patients in Denmark (Figure with permission from Ørtenblad et al.) [24,48].

Figure 2 .
Figure 2. Explanations and modifiers for use or non-use of pain-related healthcare services.

Table 1 .
Demographics and health-related profiles of informants.