The association between patients’ expectations and experiences of task‐, affect‐ and therapy‐oriented communication and their anxiety in medically unexplained symptoms consultations

Abstract Background It is unknown whether patients with medically unexplained symptoms (MUS) differ from patients with medically explained symptoms (MES) regarding their expectations and experiences on task‐oriented communication (ie, communication in which the primary focus is on exchanging medical content), affect‐oriented communication (ie, communication in which the primary focus is on the emotional aspects of the interaction) and therapy‐oriented communication (ie, communication in which the primary focus is on therapeutic aspects) of the consultation and the extent to which GPs meet their expectations. Objective This study aims to explore (a) differences in patients’ expectations and experiences in consultations with MUS patients and patients with MES and (b) the influence of patients’ experiences in these consultations on their post‐visit anxiety level. Study design Prospective cohort. Setting Eleven Dutch general practices. Measurements Patients completed the QUOTE‐COMM (Quality Of communication Through the patients’ Eyes) questionnaire before and after the consultation to assess their expectations and experiences and these were related to changes in patients’ state anxiety (abbreviated State‐Trait Anxiety Inventory; STAI). Results Expectations did not differ between patients with MUS and MES. Patients presenting with either MUS or MES rated their experiences for task‐related and affect‐oriented communication of their GP higher than their expectations. GPs met patients’ expectations less often on task‐oriented communication in MUS patients compared to MES patients (70.2% vs 80.9%; P = ˂0.001). Affect‐oriented communication seems to be most important in reducing the anxiety level of MUS patients (β −0.63, 95% Cl = −1.07 to −0.19). Discussion Although the expectations of MUS patients are less often met compared to those of MES patients, GPs often communicate according to patients’ expectations. Experiencing affect‐oriented communication is associated with a stronger reduction in anxiety in patients with MUS than in those with MES. Conclusion GPs communicate according to patients’ expectations. However, GPs met patients’ expectations on task‐oriented communication less often in patients with MUS compared to patients with MES. Experiencing affect‐oriented communication had a stronger association with the post‐consultation anxiety for patients with MUS than MES.


| INTRODUC TI ON
Medically unexplained symptoms (MUS) are symptoms for which, after a thorough history taking, physical and additional investigations, no pathological cause can be found. 1 As 3%-11% of consultations in primary care concern MUS, GPs often face patients with MUS. [2][3][4] MUS represent a variety of symptoms like headache, dizziness, fatigue and abdominal discomfort. Patients with MUS often ask for extra time, emotional support and empathy when they consult a GP. 5 Furthermore, they expect to receive an explanation and a diagnosis. 6,7 Salmon et al 5 showed that expectations of patients with MUS differ from expectations of patients with medically explained symptoms (MES). For example, patients with MUS seek significantly more emotional and moral support. However, they do not seek more often for explanation, reassurance or somatic intervention than patients with MES. 5 Patients with severe MUS are often dissatisfied with the care they receive. [8][9][10] Salmon et al 11 showed that MUS patients experienced the explanations of most GPs as being at odds with their own thinking. This is line with the results of another study in which patients with chronic fatigue syndrome were dissatisfied with the quality of medical care received during their illness as they received an unacceptable psychiatric diagnosis for their symptoms. 6  with patients with MUS, as anxiety is a strong predictor for their health status and their health-care use. 14 Two widely used indicators to gain insight into individual patients' health-care needs and expectations are the significance patients adhere to specific health-care aspects (ie, importance) and the actual experience of patients with that specific health-care aspect (ie, performance). 15 Expectations are defined as the extent of importance patients attach to communication aspects, and experiences are defined as the extent to which patients receive the communication aspects from their GP. As far as we know, quantitative studies focusing on the patients' expectations and their experiences of health-care needs have not been performed before in patients with MUS. Although there is evidence that patients with MUS differ from patients with MES regarding their expectations 5 it is not known whether this also applies to patients' own experiences and whether GPs communicate in a manner that meets patients' expectations. Therefore, the first aim of this study is to explore whether patients' expectations and experiences of the consultation differ between MUS and MES patients and to test the extent to which GPs meet patients' expectations. Furthermore, a review described the influence of context effects on health outcomes and found that nonspecific therapeutic elements, such as doctor-patient communication and doctor-patient relationship may have positive effects' on patients' blood pressure, symptom distress and frequency of health-care visits. 16  and diagnostic and prognostic information may have positive effects on patients' health status, like reduction in pain and blood pressure, improvement of complaints and reduced levels of anxiety. However, it is not known whether this also applies to patients with MUS as their anxiety may be related to their potentially unmet needs and to the persistence of their symptoms. It is known that patients with MUS in general report higher rates of anxiety than patients with MES. Therefore, the GPs' communication in the clinical encounter may have a stronger impact on reducing anxiety in MUS patients compared to MES patients. On the other hand, many GPs find MUS consultations challenging and experience communication problems during these consultations. Therefore, the GPs' communication may have a stronger association with an increase in anxiety for MUS patients compared to MES patients. The second aim of this study is to explore the association of patients' experiences and meeting patients' expectations on their anxiety after the consultation. We hypothesized that patients with MUS experience GPs' communication as less adequate than patients with MES.
Moreover, we assumed that meeting patients' expectations reduce anxiety level for both patients with MUS and MES.

| ME THODS
We performed a prospective cohort study based on data from questionnaires completed by patients and GPs. We

| Study sample and procedure
We approached 36 GPs with different backgrounds regarding sex, age, years of work experiences and location of the practice, of whom 20 (56%) agreed to participate. Data were collected in primary care practices in the region of Nijmegen, the Netherlands, between April and September 2015. All patients who visited the GP clinic during pre-selected study days were invited to participate, except those who did not speak Dutch well and patients under 18 years old. The consultations with participants were video-recorded but the observation of these videos was used to examine other research questions. A researcher asked the patient before the consultation for written consent and to complete a questionnaire; after the consultation, the same questionnaire had to be filled out. The GP completed a questionnaire after each consultation, blind for the questionnaires of the patient.

| The GP questionnaire
Immediately after each consultation, the GP answered the following question: "Do you think this patient has MUS?" on a 3-point scale relating to the presentation of physical symptoms: (a) could not be explained by a recognizable disease (ie, MUS consultation), (b) could be explained by a recognizable disease (ie, MES consultation) or (c) could partly be explained by a recognizable disease (ie, partial MUS consultation). This latter group was excluded for all analyses, as we wanted to compare the two clearly defined groups of patients. This scale has face validity as it can easily be understood and applied by GPs during consultation hours and resembles clinical daily practice in which GPs have to interpret symptoms presented by patients as explained or unexplained by physical pathology. Previous studies in this field used the same scale. 7,19 The questionnaire included demographic information, ICPC 20 (International Classification of Primary Care) coding of the consultation, whether the symptom was recurrent or new, the GP's management plan and the level of GP's satisfaction with the consultation on a 5-point Likert scale.

| The patient questionnaire
The questionnaire before the consultation included demographic

| Patients' expectation and experiences score
Patients' expectations and their experiences were measured using the QUOTE-COMM. 22,23 The QUOTE-COMM has an affectoriented scale of seven items in which the primary focus is on the emotional aspects of the interaction, a task-oriented scale of 6 items in which the primary focus is on the exchange of medical content and a therapy-oriented scale of 6 items in which the primary focus is on therapeutic aspects. Before the consultation, patients assessed how important they considered various communication aspects for the next consultation on a 4-point Likert scale (expectations). After the consultation, patients rated the GPs' performance of these aspects on a 4-point Likert scale (experiences). Consequently, within communication, we distinguished task-oriented, affect-oriented and therapy-oriented communication (see Appendix S1). Cronbach's α of the QUOTE-COMM before the consultation was in our study 0.84 for the task-oriented scale, 0.87 for the affect-oriented scale and 0.79 for the therapy-oriented scale meaning good internal consistency. After the consultation, these were 0.73 and 0.81 for respectively the affect-oriented scale and therapy-oriented scale. The Cronbach's α for the therapy-oriented scale was after the consultation 0.56.

| State anxiety
The state anxiety sum score was measured before and after the consultation by the abbreviated STAI (State-Trait Anxiety Inventory) questionnaire. 24 This questionnaire has 10 items that assess anxiety levels: the score for each item ranges from 1 to 4, with higher scores indicating a greater state of anxiety (range 1-4). Cronbach's α in our study of the STAI questionnaire was 0.88 before the consultation and 0.91 afterwards indicating good internal inconsistency.

| Statistical analysis
Data were analysed using IBM Statistical Package for Social Sciences (SPSS Statistics for Windows, Version 25.0. IBM Corp., Armonk, NY, USA). Due to a skewed distribution of the item GPs' satisfaction, this item was dichotomized into somewhat/mean satisfied or (very) satisfied. Due to a skewed distribution of the item familiarity with the GP, the item familiarity with the GP was recoded into three categories: did not know the GP hardly/at all, knew the GP moderate, knew the GP (very) well. The distribution of the education level was unequal; therefore, this item was recoded into three categories: no/ primary school, secondary school or high school/university. Due to a skewed distribution for all seven items of the COOP/WONCA, all of these were recoded into three new categories: (low/mean/high score for each item of the functional health status). The expectation and experience scores were calculated for the three scales task-, affect-and therapy-oriented communication. Each scale consists of 6 or 7 communication items and the expectation and experience scores were calculated as respectively the mean that patients assessed how important they considered the communication items before the consultation and the mean that patients rated the GPs' performance of the communication items after the consultation. To calculate the extent to which GPs met patients' expectations, we dichotomized the variable "expectation" and we combined 1 (not important) and 2 (fairly important) to one single score and combined 3 (important) and 4 (extremely important) to another single score.
The variable "experience" was dichotomized as well by combining 1 (not performed) and 2 (really not performed) to one new single score and 3 (on the whole, yes) and 4 (performed) to another single score.
Next, we combined the dichotomized expectations scores with the dichotomized experiences scores with four possible outcomes: not important and not performed (congruent), important and performed (congruent), not important and performed (incongruent) important and not performed (incongruent). In the congruent category, patients did experience what they expected, and in the incongruent category, patients did not experience what they had expected.
The extent to which GPs met patients' expectations was calculated as the percentage of patients with congruent experiences divided by the total number of patients. We did this for all 19 communication items separately, which was then used to generate the mean percentage for the three main scales: task-oriented, affect-oriented and therapy-oriented communication. We calculated the expectation, the experience score and the extent to which GPs met patients' expectations for task-, affect-and therapy-oriented communication.
To explore differences in patients' expectations and experiences in consultations with patients with MUS and MES and the extent to which GPs met patients' expectations (our first aim) a t test, a Mann-Whitney U test or a chi-squared test was used, depending on the distribution of the outcome.
For the second aim (association between patients' experiences for task-, affect-or therapy-oriented communication of the consultation with their anxiety after the consultation), we used a linear regression model. We excluded questionnaires in which patients consecutively picked the same extreme side of the scale (ie, chose both the negatively formulated as well as the positively formulated answers, > 90%) to all question or questionnaires where <70% of the STAI questions were answered. We included anxiety before the consultation, complaint type (MUS vs MES), and experience score for task-, affect-or therapy-oriented communication as potential explanatory factors. We included the complaint type by experience score (for task-, affect-or therapy-oriented communication) interaction term as well and evaluated whether the interaction term was significant.
If the coefficient on the interaction term was statistically significant, there was a difference between patients with MUS and MES in how the experience score for task-, affect-or therapy-oriented communication affected their anxiety level. As we distinguished a task-, an affect-and a therapy-oriented scale, we performed three separate regression analyses. To explore the association between the extent to which GPs met patients' expectations for task-, affect-or therapyoriented communication with their anxiety after the consultation, we substituted the experience score for task-, affect-or therapyoriented communication with the extent to which GPs met patients' expectations score for task-, affect-or therapy-oriented communication in the linear regression analysis. Again, we performed three separate regression analyses (Table 3).

| RE SULTS
In total, 577 patients attended their GP during the study days, of  Table 1. Patients with MUS were younger and visited their GP more often for the same symptom. Furthermore, these patients scored significantly lower on the COOP/WONCA aspects feelings, social activities, overall health, pain and fatigue.

| Patients' expectations, patients' experiences and meeting patients' expectations
The expectation score, the experience score and the extent to which GPs met patients' expectations are shown in Table 2

| Patients' anxiety
The mean anxiety level after the consultation was 1.84 for patients with MUS and 1.72 for patients with MES. This was for both groups significantly lower than the mean anxiety level before the consulta-

| Summary of main findings
We did not find differences between expectations of patients with MUS and MES regarding their GPs' task-, affect-and therapy-

MES (n = 292) P-value
Age means (SD) 50. 5 (10), tiredness (6), neurological deficit (2), shortness of breath (2), globus sensation (2) headache (1), collapse (1) and hypersensitivity syndrome (1). The number of ICPC codes divided in chapters (n) for the MES group was as follows: musculoskeletal (48), psychological (31), respiratory (31), skin (29), digestive (27), circulatory (26), ear (14), general and unspecified (13), eye (13), endocrine (13), female genital system and breast (8), urology (7), male genital system (7), social problems (7), blood, lymphatics and spleen (4), neurological (4), pregnancy and childbirth (4), unknown (6). a Hardest physical effort during at least 2 minutes, from "very heavy" to "very light." b Extent of being bothered by emotional problems, from "not at all" to "a lot." c Extent of difficulties in doing daily activities, from "no difficulty" to "could not be done." d Extent to which social activity is limited by physical and emotional health, from "not at all" to "a lot." e Overall health, from "excellent" to "poor." f Presence of pain, from "no" to "heavy." g Presence of fatigue, from "no" to "a lot." h Bold values are statistically significant. Earlier research showed that patients with MUS sought more emotional support than patients with MES. 5 We did not measure emotional support directly but the expectations regarding affect- We found no differences in change in anxiety levels between patients with MUS and MES. It is known that patients with MUS in general are associated with higher rates of anxiety than patients with diseases with comparable symptoms. 32,33 We assume that these higher rates of anxiety concerns severe MUS patients, while the included MUS patients in our study did probably not belong to these category of severe MUS patients. The benefit of affectoriented communication on patients' anxiety as we found has been described previously. 16,18,34,35 These studies were, however, not  specifically focus on the reassurance but also on other communication elements. Therefore, we did not delete the communicational item "referral to another specialist." We excluded the partial MUS group for all analyses, as we wanted to compare two clearly distinguished groups. In case, we should merge patients with partial MUS and patients with TA B L E 3 Association between post-consultation anxiety and the explanatory factors for task-oriented, affect-oriented and therapyoriented communication

| Practical implications
Increasing

| CON CLUS ION
Many GPs think that patients with MUS differ from other patients because they want more explanation and somatic interventions.
However, as we found that patients' expectations do not differ between patients with MUS and MES, GPs should reflect on these assumptions. GPs' communication training should focus on a thorough self-reflection and should pay attention to task-and especially affectoriented communication as these are associated with reduced levels of anxiety.

ACK N OWLED G EM ENTS
We thank all the GPs and patients for their co-operation in this study.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

E TH I C S CO M M IT TE E A N D I N FO R M E D CO N S E NT
The study was carried out according to Dutch privacy legislation. The privacy regulations were approved by the Dutch Data Protection Authority. The research ethics committee of the Radboud University Nijmegen Medical Center concluded that the study could be carried out in accordance with the applicable rules in the Netherlands (file number 2015-1566). Written informed consent was obtained from all participating patients; patients were able to withdraw their consent at any time.