Characteristics of outpatients with functional somatic syndromes at a university hospital's general medicine clinic

Abstract Background The term medically unexplained symptoms (MUS) is unhelpful for both patients and physicians, and more acceptable illness categories are needed as substitutes for MUS. While some potential substitutes are characterized by excessive psychological burden related to somatic symptoms, “functional somatic syndromes” (FSS) is a category that focuses on physical dysfunction and emphasizes similarities among individual syndromes. Examples of FSS include irritable bowel syndrome, functional dyspepsia, and fibromyalgia syndrome. This study aimed to distinguish FSS from MUS and compare the somatic and psychobehavioral characteristics of FSS with those of other diseases. Methods This study included 1975 first‐visit outpatients at a Japanese university hospital's general medicine clinic. According to their first‐listed diagnosis, they were classified as having FSS, acute infection, organic disease (OD), psychiatric disorder, and unknown condition (UC). The somatic symptom burden and health‐related quality of life (HRQoL) were assessed using the Somatic Symptom Scale‐8 and EuroQol‐5 Dimension, respectively; the involvement of psychobehavioral factors affecting somatic symptoms was also evaluated. Results Overall, 33% of patients were included in the FSS category, and 93% of the supposed MUS (FSS and UC) were diagnosed with FSS. Compared with OD, FSS showed more severe somatic symptom burden, similar reduced HRQoL, and higher involvement of psychobehavioral factors. Conclusion It can be useful to improve FSS diagnostic skills for the reduction of MUS misdiagnosis. Psychobehavioral factors might be less associated with MUS (in the narrow sense of the term) than FSS.


| INTRODUC TI ON
There are patients in whom histological and organic disorders cannot be identified based on somatic symptoms in both primary and secondary care settings. Their conditions are recorded as medically unexplained symptoms (MUS), 1 and medical treatment of such patients places a considerable burden on physicians. 2 As the term "MUS" can cause an unhelpful explanatory gap between patients and physicians, 3 more acceptable illness categories are needed. For example, bodily distress disorder 4 and somatic symptom disorder (SSD) 5 are illness categories characterized by excessive psychological burden relating to somatic symptoms and are candidates for MUS substitutes.
It is difficult to identify whether patients' functional physical symptoms are rooted in somatic diseases or psychiatric disorders, 6 and patients exhibiting such symptoms are often misdiagnosed with MUS. FSS refers to several related syndromes that are characterized based on symptoms, suffering, and disability rather than consistently demonstrable tissue abnormalities. 7 It is a category focused on physical dysfunction, and it emphasizes that similarities among the individual syndromes outweigh the differences. These syndromes include irritable bowel syndrome (IBS), functional dyspepsia, fibromyalgia syndrome (FMS), and chronic fatigue syndrome. Moreover, Wessely et al. have suggested many other syndromes and symptoms, such as premenstrual syndrome and tension headache, for inclusion as FSS. 8 In our previous studies, we reported lower health-related quality of life (HRQoL) associated with autonomic dysfunction 9 and improvement in somatic symptoms and mood disturbance through autogenic training 10 in patients with FSS. As it is usually less problematic for physicians to assess patients' physical dysfunction than their psychological conditions, the classification of FSS rather than MUS is expected to minimize misdiagnosis. However, the prevalence of FSS in MUS is uncertain.
Most patients with FSS receive unnecessary tests and unfounded reassurance that the disease is not serious and are likely to repeatedly consult other physicians. 11,12 As a result, the cost of treating such patients becomes a major societal economic issue. 13 Given that illness behavior depends on the healthcare system and culture of a country, 1 it is necessary to investigate the behavior in each healthcare system. In systems wherein patients must first visit family physicians for referrals, MUS accounts for 5%-30% of primary care patient diagnoses 2 and 37%-68% of specialty care patient diagnoses. 14,15 On the contrary, Japan has a "free access" system wherein patients can utilize whatever kinds of outpatient services they like without referrals. As patients who seek both primary and specialty care can visit university hospitals, epidemiological investigation at a general medicine clinic of a university hospital is suitable to grasp the actual incidence of MUS in Japan. Therefore, this study aimed to identify patients with FSS from patients with MUS and compare their somatic and psychobehavioral characteristics with those of other diseases diagnosed at a university hospital's general medicine clinic.

| Design
This was a cross-sectional survey of outpatients with FSS in the secondary healthcare area in Japan. Before examination, the outpatients who presented at a university hospital's general medicine clinic for the first time were required to complete two questionnaires related to somatic symptom burden and HRQoL. After the examination was complete, the attending physician decided on a diagnosis and the involvement of psychobehavioral factors contributing to the condition.

| Participants
The participants in this study were outpatients who presented for the first time at a general medicine clinic of a university hospital in an urban city in western Japan between January 1 and December 31, 2016. Those who were examined by residents and did not complete the questionnaire before examination were excluded from the study.
The study protocol was approved by the Institutional Review Board of the university hospital. However, informed consent was not obtained because this study neither involved patients' interventions nor used samples obtained from patients. An opt-out method was used so that patients could refuse to participate in the study.

| Somatic symptom burden
The Japanese version of the Somatic Symptom Scale-8 (SSS-8) is a self-administered questionnaire that assesses somatic symptom burden; 16,17 it was originally developed as a shorter version of the Patient Health Questionnaire-15, 18 and comprised eight items pertaining to the following categories: (1) stomach or bowel problems; (2) back pain; (3) pain in the arms, legs, or joints; (4) headaches; (5) chest pain or shortness of breath; (6) dizziness; (7) feeling tired or having low energy; and (8) having trouble sleeping. Each item is scored 0 (not at all) to 4 (very much).

| HRQoL
The Japanese version of the EuroQol-5 Dimension (EQ-5D) has been approved by the EuroQol Group as a concise generic health status instrument that can be used to measure, compare, and value health status across diseases; [19][20][21] it assesses self-reported health status across five dimensions (mobility, self-care, usual activities, pain/ discomfort, and anxiety/depression) and three levels of severity (no problems, moderate problems, and severe problems). The responses for the five dimensions were combined into a five-digit number and were converted to a HRQoL score using Japanese value sets.

| Assessment
As of 2016, healthcare insurance at Japanese medical institutions has been organized using diagnostic conditions based on the 10th revision of the International Classification of Diseases (ICD- 10) codes. Therefore, the attending physician decided upon the appropriate diagnostic condition code.
In addition, the condition was classified according to the involve-

| Classification
For each patient, the first-listed diagnosis at initial examination was therefore, there could be potential discrepancies depending on the individual performing the classification. To address this, classification was performed independently by three coding physicians; two of them were employees at the general medicine clinic, and one was a psychosomatic medicine specialist. Whenever the classification by all the three physicians was not consistent, the final classification was made after discussion among them.

| Statistical analysis
Data are presented as mean ± standard deviation. Differences between diagnostic categories in terms of age and somatic symptom burden using the SSS-8 were assessed using unpaired one-way analysis of variance. Differences in the HRQoL status between diagnostic categories were assessed using the nonparametric Kruskal-Wallis test, with EQ-5D utility values. Differences in the involvement of psychobehavioral factors among the five diagnostic categories were assessed using the chi-squared test. For all the differences identified, multiple comparisons were made using Bonferroni's method. All statistical analyses were performed using SPSS version 26.0 (SPSS Inc.). For all statistical tests, the α level was set at 0.05.

| Participants
A total of 2366 patients presented at a general medicine clinic of a university hospital, but 335 patients who were examined by residents and 56 patients who did not complete the questionnaire before examination were excluded from the study. As a result, 1975 patients were included in the final analysis. The patient population comprised 1239 women (63%) and 736 men. The mean (± standard deviation) age of the patients was 52.6 ± 19.8 years. Overall, 1287 (65%) patients had to pay a supplementary fee of ¥5000 ($47) because they wished to consult a general physician at the university hospital without a referral letter, while 688 (35%) patients were referred to our clinic either from within the hospital or from primary or secondary healthcare centers owing to the nature of their symptoms (unknown cause).

| Classification
The proportion (number) of patients classified in each diagnostic category was as follows: FSS, 33% (653); AI, 15% (306); OD, 45% (881); PSYD, 5% (86); and UC, 2.5% (49). Age differences were noted between these categories [F (4, 1970) = 18.160; p < 0.001], in that patients in the AI category were younger than those in the other categories, and patients in the FSS category were younger than those in the OD category. No gender differences were noted between the categories (Table 1). AI was mostly reported in the respiratory, urinary, or gastrointestinal tracts.

| DISCUSS ION
This study investigated the characteristics of patients with FSS at a general medicine clinic of a university hospital. FSS occupied onethird of our study population and was associated with severe somatic symptom burden, moderate HRQoL, and high involvement of psychobehavioral factors.
To the best of our knowledge, this is the first large-scale study to elucidate the proportion and characteristics of patients with FSS at a general medicine clinic of a university hospital. This diagnostic category has not been widespread until recently, and there are no previous studies that have classified MUS by functional pathology or problem. Therefore, FSS and UC in our data were considered to correspond to MUS in conventional studies. For our study, the combined prevalence of FSS and UC was 36% in general medicine clinic, and in other studies, MUS accounted for 5%-30% of primary care diagnoses 2 and 37%-68% of specialty care diagnoses. 14,15 In secondary care settings, each specialty of the clinics is usually related to the characteristic of the patient with MUS. Therefore, the prevalence of FSS in MUS varies according to the settings. As our study population consisted of both primary and secondary healthcare patients, it is suitable that our result was between those reported in the two studies mentioned. Our data revealed that the prevalence of FSS among MUS in general medicine clinic was 93%, indicating the potential usefulness of FSS diagnostic skills in reducing the misdiagnosis of MUS.
Functional somatic syndromes shows high somatic symptom burden rather than demonstrable tissue abnormalities. We still do not have a simple method to diagnose FSS. While the SSS-8 is a questionnaire to assess somatic symptom burden, our results showed that it was useless as a diagnosis assistance tool to distinguish FSS TA B L E 1 Characteristics of diagnostic categories   The HRQoL status in FSS was approximately the same as that in OD. HRQoL status is reported to be reduced by the same extent for IBS and other chronic organic diseases, 23 and to decrease more in psychiatric disorders than in physical disorders. 24 Our results support these findings. In addition, the HRQoL in AI may not have been reduced because of the short morbidity period.
Functional somatic syndromes was associated with high involvement of psychobehavioral factors and differed quite markedly from UC in this respect. In our study, psychobehavioral factors were assessed by a physician only during the first visit, and this might not be sufficient to differentiate between SSD and "psychological factors affecting other medical conditions", as defined in the DSM-5. However, the distinction between FSS and SSD is conceptually difficult when somatic symptoms are affected by psychobehavioral factors. 6 In our study, we did not distinguish these conditions because that was not our aim. Numerous studies have established the involvement of psychobehavioral factors in FSS, [25][26][27] and our results indicated that the psychosomatic component is one important

TA B L E 3
Multiple comparison of EQ-5D utility values mechanism of FSS. Thus, in a narrow sense, MUS (corresponding to UC in our data) might be less associated with the involvement of psychobehavioral factors than was supposed.

| LI M ITATI O N S
First, this study was performed at a single medical institution, and therefore, it might have been subject to selection bias. It is desirable to conduct multicenter studies including general medicine clinics of other university hospitals and community hospitals or primary care offices to validate our findings. Second, the term MUS was not used in the classification, and FSS and UC categories were used instead.
However, the definitions of these terms were somewhat vague, and the decision as to whether or not a condition was a functional pathology was partly subjective on the part of the examining physician. We tried to resolve this issue by including three physicians; however, it is uncertain whether this approach was sufficient to eliminate this problem. Autonomic dysfunction, which is considered as one of the mechanisms of FSS, 9 is expected to be an objective index to distinguish FSS from MUS in the future studies. Third, only the first-listed diagnosis for each patient was used for classification, and no other comorbid diseases were investigated. It is necessary to evaluate the classification of patients according to the final diagnoses or comorbid diseases.

| CON CLUS IONS
We investigated the characteristics of patients with FSS at a gen-

ACK N OWLED G M ENT
We would like to thank Editage (www.edita ge.com) for Englishlanguage editing.

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