Patient expectations for management of chronic non‐cancer pain: A systematic review

Abstract Background Chronic pain is a major economic and social health problem. Up to 79% of chronic pain patients are unsatisfied with their pain management. Meeting patients’ expectations is likely to produce greater satisfaction with care. The challenge is to explore patients’ genuine expectations and needs. However, the term expectation encompasses several concepts and may concern different aspects of health‐care provision. Objective This review aimed to systematically collect information on types and subject of patients’ expectations for chronic pain management. Search strategy We searched for quantitative and qualitative studies. Because of the multidimensional character of the term “expectations,” the search included subject headings and free text words related to the concept of expectations. Data extraction and synthesis A framework for understanding patients’ expectations was used to map types of expectations within structure, process or outcome of health care. Main results Twenty‐three research papers met the inclusion criteria: 18 quantitative and five qualitative. This review found that assessment of patients’ expectations for treatment is mostly limited to outcome expectations (all 18 quantitative papers and four qualitative papers). Patients generally have high expectations regarding pain reduction after treatment, but expectations were higher when expressed as an ideal expectation (81‐93% relief) than as a predicted expectation (44‐64%). Discussion and conclusions For health‐care providers, for pain management and for pain research purposes, the awareness that patients express different types of expectations is important. For shared decision making in clinical practice, it is important that predicted expectations of the patient are known to the treating physician and discussed. Structure and process expectations are under‐represented in our findings. However, exploring and meeting patients’ expectations regarding structure, process and outcome aspects of pain management may increase patient satisfaction.


| INTRODUCTION
In Europe, chronic non-cancer pain of moderate to severe intensity occurs in approximately 19% of the adult population. 1 The international society for the study of pain defines chronic non-cancer pain (CNCP) as non-malignant pain lasting 3 months or more, or as pain persisting beyond the time of expected healing. CNCP often lacks a clear associated pathology; prognosis is uncertain and varies considerably between patients and therefore can be difficult to treat. 2 CNCP has a significant impact on health status, quality of life and daily activities such as paid work. 3 A large proportion of CNCP patients lack adequate pain control. 3,4 Up to 79% of the CNCP patients believe that their pain is inadequately treated, and up to 43% of the patients report not receiving pain treatment at all. 5 Given the subjective and objective burden of CNCP, the fact that a large majority of patients believe their pain is inadequately treated should alarm health-care professionals and policymakers. 3 Patients' satisfaction with CNCP management can be seen as the end result of the match between expectations and subsequent experiences. [6][7][8] From a theoretical conceptual point of view, patients' expectations are viewed by some as the major determinant for satisfaction with health care. For example, according to the expectancy disconfirmation paradigm, satisfaction arises either from positive experiences disconfirming negative expectations. Dissatisfaction arises when negative experiences disconfirm positive expectations, or when negative experiences confirm negative expectations. Disconfirmation of expectations affects perceived quality of care, and hence satisfaction. 9 Discrepancy between expectations and actual outcome portents lower satisfaction. 10 Empirical evidence for the relation between expectations and satisfaction is for instance provided by Noble et al. They found that the fulfilment of patients' satisfaction was primarily determined by patient expectations. 11 Each patient with CNCP experiences pain in a highly individualized way, and each patient has different expectations, needs and goals. Therefore, pain management should also be customized, and understanding patients' expectations is essential in shared decision making. 12,13 Meeting patients' expectations should result in more consistency between the patients' needs and health-care delivery, and subsequently in greater satisfaction with care. 14 Satisfaction with care might increase compliance, which, in turn, can improve pain management outcome. 15 The challenge, however, is to identify the patients' needs and expectations. The aim of this study was therefore to systematically explore the literature for information on patients' expectations of CNCP management. As the term "expectations" comprises a broad range of concepts which can refer to several aspects of health-care delivery, we start this review by defining and classifying expectations according to type of expectation and according to Donabedian's health-care model of structure, process and outcome of care.

| Categorizing patient expectations
Expectations are generally explained as "a strong belief that something will happen or be the case." 16 Related to anticipation, this implies that expectations are created and sustained by a cognitive process. An event, however, can be desired but not expected, 17 for example "I desire to be cured after treatment but I expect only minor pain reduction." Expectations, therefore, can also be expressed as desires, wishes and hopes. 8 In contrast to beliefs, these primarily reflect a valuation mainly based on emotions, a perception that a given event is wished for. It is therefore important to distinguish the various types of definitions of the expectations used in research papers as these are sometimes lacking, and the reader is often left to guess whether the expectations described are hopes or ideals, or anticipated outcomes. Thompson 7  , necessity (what is perceived to be needed), entitlement (that which is owed or to which one has a right) and normative standards (that which should be). 8 Kravitz 8 described a dynamic model in which patients' expectations are also defined according to content (i.e structure, process or outcome of care) 18 .

| Types of expectations
In this study, we consider "expectations as probabilities" and "predicted expectations" to reflect the same type of expectations.
Throughout the study, we will refer to this as predicted expectations.

| Content: Structure, process and outcome of care
Patients may express their expectations regarding several aspects of health-care delivery. The Donabedian's health-care model provides a standard for examining health services and evaluating quality of health care and distinguishes between structure, process and outcome of care (SPO). 18 Structure of care denotes the setting in which the care occurs, for example the characteristics of the building, accessibility, availability of therapeutic and diagnostic facilities.
Process of care reflects what is actually done in care delivery and care coordination, for example provider characteristics, timing variables. It describes how the patient moves into, through and out of the

| Framework for understanding patient expectations
The term "expectations" is sometimes undefined, imprecise or multiinterpretable; therefore, a conceptual framework is used to categorize the findings from the papers in this review ( Figure 1). Expectations are classified according to the SPO model 18 and the work of Thompson 7 and Kravitz. 8 Predicted expectations are cognitive, realistic and anticipated. Value expectations are attitudes, regulated by feelings, emotions and affections. The value expectations are divided according to Thompson 7 into ideals, necessities and normative expectations (i.e entitlements/normative standards).

| Objectives
The main objective of this systematic review was to classify patients' expectations regarding CNCP management according to the framework of understanding expectations ( Figure 1). Secondary objective of this study was exploration of the subject of patients' expectations.

| Design
This systematic review explored expectations regarding CNCP management reported in quantitative, in qualitative, as well as in mixed methods research papers. The combination of quantitative, mixed methods and qualitative research was expected to generate a more complete and deeper insight than either method alone.

| Eligibility criteria
Expectations of patients undergoing pain management continually change when experiences accumulate. 21 Furthermore, patients with acute (less than 6 weeks), subacute (6-12 weeks) and chronic (at least 3 months) pain exhibit different physiologies, courses and treatment responses. 22 Therefore, it is highly likely that expectations regarding pain therapy differ before and after pain therapies and between (sub) acute and chronic patients. For this reason, this review was restricted to papers that described expectations regarding pain therapy before or during their pain management of chronic (>3 months) non-cancerrelated pain. Pain management is defined as communication, evaluation, diagnosis and treatment, of all different types of CNCP.
Studies were considered eligible for review if they met the following inclusion criteria: (i) patients were questioned about expectations before or during CNCP management; (ii) the study population consisted of adult patients with chronic (≥3 months) non-cancer-related pain; (iii) measuring expectations was (one of) the objective(s) of the study, and the method for obtaining information on patients' expectations was described. Exclusion criteria were as follows: (i) cancer-related pain, (ii) pain duration of less than 3 months or (iii) pain duration not specified.
In case of inadequate or missing information about expectation(s) or definition of chronic pain, authors of the article were contacted for information. Studies were excluded from this review if multiple studies were identified, with overlap in study populations and findings. When this was the case, only the most appropriate (to our review objective) study was included to avoid potential duplication of data sets. text words connected to the construct expectations. [6][7][8]17 In Table 1, Chronic pain free text terms 10 ((chronic$ or intractable or refractory or persistent$ or long term or longterm or sustained or longstanding or long standing or permanent$ or unremitting or unrelenting or unceasing or constant or constantly) adj3 (pain or pains or painful$ or pained)).ti,ab,ot. (52779) 11 ((chronic$ or intractable or refractory or persistent$ or long term or longterm or sustained or longstanding or long standing or permanent$ or unremitting or unrelenting or unceasing or constant or constantly) adj3 (hurt or hurting or hurts)).ti,ab,ot. (10) 12 ((chronic$ or intractable or refractory or persistent$ or long term or longterm or sustained or longstanding or long standing or permanent$ or unremitting or unrelenting or unceasing or constant or constantly) adj3 (sore or soreness or tender$ or discomfort or ache$ or aching or agony)).ti,ab,ot. (881) 13 ((chronic$ or intractable or refractory or persistent$ or long term or longterm or sustained or longstanding or long standing or permanent$ or unremitting or unrelenting or unceasing or constant or constantly) adj3 (nociception or nociperception or algiatry)).ti,ab,ot. (230) 14 ((chronic$ or intractable or refractory or persistent$ or long term or longterm or sustained or longstanding or long standing or permanent$ or unremitting or unrelenting or unceasing or constant or constantly) adj3 (allodynia or alveolalgia or backache or causalgia or cephalalgia or cheiragra or chiragra or coxalgia or coxodynia or cystalgia or dorsalgia or dysmenorrh?ea or dyspareunia or dysuria or erythromelalgia or failed back surgery syndrome or fibromyalgia or gastralgia or headache$ or hepatalgia or intermittent claudication or ischialgia or lumbago or lumbalgia or lumbodynia or mastalgia or mastodynia or meralgia paresthetica or metatarsalgia or migraine$ or myalgia or neuralgia or odontalgia or odynophagia or orchalgia or otalgia or paroxysmal hemicrania or piriformis syndrome or piriformis muscle syndrome or pleuralgia or polymyalgia or prostatalgia or prostatodynia or psychalgia or rachialgia or radiculalgia or sciatica or SUNCT syndrome or toothache or vulvodynia)).ti,ab,ot. Patient expectation free text terms 26 ((patient$ or consumer$ or user or users or client$ or sufferer$ or person$ or people or adult$ or men or mens or man or mans or women$ or woman$) adj1 (ambition$ or aspiration$ or attitude$ or belief$ or believe$ or choice$ or concern$ or decision$ or demand$ or desire$ or drive or evaluation$ or expectation$ or experience$ or feeling$ or goal$ or hope$ or idea$ or impression$ or intention$ or judgment$ or motivation$ or motive$ or need or needs or opinion$ or perception$ or perspective$ or preference$ or reason$ or requirement$ or thought$ or value$ or view$ or wish$)).ti,ab,ot.

| Quality assessment
Quality assessment of the qualitative research papers was conducted by two independent reviewers(JG/CL) according to the Qualitative Assessment and Review Instrument (QARI). 23 The QARI software was developed by the Joanna Briggs Institute (Australia) for the evaluation and synthesis of qualitative research articles. This quality appraisal tool is a standardized 10-criteria checklist for two independent reviewers and assesses bias in relation internal validity to, for example, congruence between research methodology, philosophical perspective, methods used to collect data, analyse the data and for interpretation of the data.
Assessment of the quantitative and mixed methods research papers was performed with the Mixed Method Appraisal Tool (MMAT). 24 This appraisal tool was developed for the quality assessment in reviews that include quantitative, qualitative and mixed methods studies. With this instrument, it is possible to judge each paper in relation to its methodological domain.

| Data collection, extraction and synthesis
Extraction of findings of the qualitative papers was performed using To categorize patients' expectations, a metasynthesis of the papers is presented in a tabular summary, using the framework of Figure 1. Mixed methods studies in this systematic review were evaluated as quantitative papers because the analyses were quantitative, although the assessment often was mostly qualitatively performed.
Three authors JG/CD/PW independently categorized the types of expectations. Differences in categorization were discussed and solved in a consensus meeting.  Table 2 shows the characteristics of the included studies. In most quantitative studies (N=18), a self-constructed questionnaire 25-33 was used; six studies used a validated questionnaire. [34][35][36][37][38][39] Most studies (N=12) were conducted in the USA, seven in Europe, two in Australia and two in Asia. Research aims and management settings were diverse.

| Study characteristics
Chronic spinal pain was the most studied type of pain (11 studies).

| Quality Appraisal
The quality of the studies was appraised using the MATT 24 and QARI appraisal tools, Tables 3a,b for, respectively, quantitative (including mixed methods studies) and qualitative studies.
As our interest only related to pain management expectations, and these were collected mostly at baseline, all the quantitative and mixed methods papers were appraised as descriptive studies. On item 4.3 ("Are measurements appropriate"), for 11 of the 18 quantitative papers, the scores were zero because these studies used self-constructed questionnaires without validation. The quality of the quantitative papers was good to excellent with ten papers reaching 50-54%, seven 75% and one paper scoring 100%.
Most qualitative studies (Table 3b) scored low on item 1: "There is congruity between the stated philosophical perspective and the research methodology." Almost all studies scored zero on items 6 and 7, that is 6) "There is a statement locating the researcher culturally and theoretically" and 7) "The influence of the researcher on the research and vice versa is addressed." However, the overall quality of the accepted qualitative papers was rather high, scores ranged from 70% up to 90%; therefore, all papers were included in this review. Table 4 shows the results of the categorization by type of expectation and content (SPO) of care delivery. Two papers studied structure expectations, four process expectations and 21 outcome expectations.

| Categorization of expectations according to the framework
All quantitative papers (N=18) described outcome expectations. Onethird of the quantitative papers described both value and predicted expectations.
Qualitative studies described more frequently (N=7) value expectations. Sixty per cent of qualitative papers described both value and predicted expectations. Table 5 shows types of patients' expectations found in quantitative studies, and Table 6 shows expectations found in qualitative studies.

| Structure expectations
Only value expectations were found regarding structure of care; these value expectations were expressed as ideals or necessities.
Patients expressed the desirability of fellow patient involvement in a chronic pain management service, mostly to support the patients in their contact with the professionals and achieve validation of their pain problem 40 (Table 5). Further structure expectations were desirability of efficient flow of patients through the system (Table 5) and need for accessibility, for example parking places nearby and variable opening times (Table 6).

| Process expectations
Research addressing expectations regarding process of care was found in one quantitative 41 (Table 5) and in three qualitative studies 40,42,43 ( Table 6). All studies reported value expectations of which two also showed predicted expectations. Regarding process expectations, explanation or improved understanding of the pain problem was expressed as a necessity; validation or acknowledgement of the pain problem was expressed mostly as a normative expectation, and to get a proper diagnosis was stated as an ideal expectation. Getting a thorough consultation or referral from the GP to a specialist was once expressed as a predicted expectation and once as a normative expectation.

| Outcome expectations
Most studies, all 18  Almost all of the quantitative studies investigated predicted expectations in terms of pain management goals, like expected outcome.
Four studies focused on value expectations, for example desired, disappointing, worthwhile or outcome needed to consider the pain management a success. 27,29,33,44 Four papers studied expected pain relief before pain treatment and related this to the pain reduction acquired after treatment. All showed that patients expected a substantially larger reduction in pain from the treatment than they attained. 30,36,38,45 For instance, patients needed a mean 50.9 (scale 1-100) reduction and only attained 11.9. 36 Whenever available in the papers, the expected levels of pain reduction by type of outcome expectation are included in Table 5a. In all quantitative studies, in which the ideal pain relief and expected pain relief were assessed separately, the results showed discrepancies between desired, needed and predicted pain relief. The expected pain relief was notably less than the stated needed and desired pain relief.
The qualitative studies (Table 6) also showed great discrepancy between the desired and the expected outcome: Patients often Quantitative studies  expressed a want or a need for pain relief or pain cure but predicted substantial less pain relief or no pain reduction at all. 46,47 Within each setting of care delivery, that is primary care, CAM, surgery, rehabilitation, pain centres, most CNCP patients expected pain relief; however, some patients did not expect pain relief but expressed the desire and need for physical improvement and being able to walk with the grandkids for instance, or do daily living chores without limitations. Some patients expressed the need to learn to cope with the CNCP, or to learn tools for better control of the complaints.

| DISCUSSION
In this review, we systematically searched for quantitative and qualitative studies addressing expectations of chronic pain patients regarding CNCP management and categorized expectations according to the type of expectation and Donabedian's health-care model of structure, process and outcome.
This review found that assessment of CNCP patients' expectations for pain management is mostly limited to outcome expectations.
Furthermore, we found that patients answer differently to questions pertaining predicted expectations than to questions about ideal expectations. Patients' ideal expectations are higher than their predicted expectations; some patients hope for, or desire, a full cure, but predict to gain little or nothing from pain management. This discrepancy between ideal and predicted expectations could be due to negative experiences in the past, or it could be that patients lower their expectations as a way to avoid disappointment. 48 Another explanation, which logically follows from Thompson 7 , is that the terms "hope" and "desire" actually mean something else to patients than the term "expectation," irrespective of their previous experiences. In that case, it could well be that patients are in the process of accepting the pain and consequently suffer less pain and thus expect (predicted expectation) less gain from pain management than they would perceive as ideal (value expectation). 49,50 Empirical studies have demonstrated a positive association between acceptance and successful adaptation to chronic pain. 50 Results of the papers in our review showed that overall CNCP pa- Our results show that the expectations as expressed by the patients depended on which way the questions were asked. For instance, when asking for desired (value expectation) levels of pain after treatment, patients reported to wish for up to 98% pain relief versus when asked "what to expect the treatment to do" (predicted expectation), patients reported far more realistic pain reductions of 50%. Therefore, it is highly probable that the relationship between "value" expectations and outcome differs from the relation between "predicted" expectations and outcome. Six studies in this review demonstrated this by assessing the relation between outcome and expectations. 28,29,32,33,36,37 A significant association between expectations and outcome was found in three papers that studied predicted expectations: Higher expectations of outcome resulted in more improvement. 28,32,37 In contrast, the other three studies that assessed the association between value expectations and outcome did not find an association with outcome. 29,33,36 Therefore, it seems that not only for pain management but also for research purposes the type of expectation assessed should be clear.
We found that most quantitative papers did not use validated expectation scales. This could be due to the fact that applied research into patients' expectations is still in its infancy. Developing and validating expectation scales that comprise structure, process and outcome expectations as well as the different types of expectations would be helpful for shared decision making and could provide a useful tool for expectation management during pain therapies.
The incorporation of findings into a predefined expectation framework can be seen as a strength of this systematic review. Working with a framework to categorize types of expectations found in the papers leads to a better understanding of the broad concept and terms related to "expectations." However, the original papers did not always provide a typology of expectations, leaving this open to our interpretation. Specifically, within value expectations, distinguishing between necessities and normative expectations was particularly challenging.
The categorization was therefore performed by three authors independently (JG, PW, CD), and differences were discussed until consensus was reached.
Another strength of this systematic review is the combination of quantitative, mixed methods and qualitative studies. Qualitative findings added context or explanatory powers to the quantitative data, whereas quantitative data were useful to assess the size of the topic of interest. Furthermore, we found that qualitative findings provided more information about expectations regarding process and structure of care. However, some qualitative studies also restricted themselves to asking focused questions and explored or reported outcome expectations only. 46,47 For health-care providers, for pain management and for pain research purposes, the awareness that patients express different types of expectations is important. For health-care providers, it points at the importance of asking the right question about expectations in shared decision making and in expectation management. A validated questionnaire that incorporates all types of expectations that are assessed before the first consultation would be a useful tool to ensure manageable answers from the patient and discover genuine needs that should be incorporated into the pain treatment plan. Furthermore, this asset could also help in shared decision making to discover and discuss unrealistic expectations for treatment so as to avoid disappointment and dissatisfaction with care.
Health-care providers and policymakers should grasp the opportunity to improve on structure, process and outcome of care and thereby attain higher patient satisfaction by better meeting patients' expectations.

| Clinical implications
This systematic review showed that little information could be found about structure and process expectations of CNCP patients. We like to point out that this could be a lost opportunity to derive higher patient' satisfaction for CNCP management. It is known that structure and process components of care can influence pain patient' satisfaction. 51,52 For instance, a strong positive association was found between higher numbers of physicians and nurses and patient' satisfaction with the health-care system.
Understanding the expectations and needs of patients is essential in shared decision making. 13 Therefore, it is important to differentiate between the types of expectations. In particular, the difference between value and predicted expectations is important in clinical practice. Value expectations are ideals, and predicted expectations are the more realistic expectations. This review gives an indication that the association between high expectations and a better outcome is present when assessing predicted (i.e more realistic) expectations. In contrast, no association was found between high ideal expectations and better outcome. Patients' predicted expectations for a specific treatment can be altered by information from the professional about the evidence for potential benefits and harms of a treatment for an individual patient.
Management of expectations before and during pain management could be an important contribution to patients' satisfaction by lowering predicted expectations that are too high or heighten predicted expectations that are too low.
Differentiating between types of expectations could also be important if patients are in the process of accepting the pain better and consequently struggling less with the pain. 50 The pain management challenge should be to provide a personalized pain management programme without obstructing the patient's pain acceptance process. In shared decision making, it is likely that the process of pain acceptance is supported if predominantly predicted expectations are discussed.