Studies in which standardized methods have been used. Somatization in Primary Care

Transcription

1 Prevalence, Health Care Utilization, and General Practitioner Recognition PER FINK, M.D., PH.D., DR. MED. SC., LISBETH SøRENSEN, M.D., MSC. ECON. MARIANNE ENGBERG, M.D., PH.D., MARTIN HOLM, M.D. PAUL MUNK-JøRGENSEN, M.D., DR. MED. SC. To study the prevalence of somatoform disorders (SDs) in primary care, a questionnaire including the modified 25-item version of the Symptom Checklist-90 was administered to 191 patients consecutively consulting their family physician. A stratified sample of the patients was interviewed with the Schedules for Clinical Assessment in Neuropsychiatry. The study showed that 22.3% (confidence interval [CI]: 95%: ) of the patients fulfilled the diagnostic criteria for an International Classification of Diseases-10th Revision (ICD-10) SD, excluding SD, unspecified, and 57.5% (CI: 95%: ) for DSM-IV SD. But 30.3% met the criteria (CI: 95%: ) when the DSM-IV Not Otherwise Specified (NOS) diagnostic group is excluded. The most frequent ICD-10 diagnosis was autonomous dysfunction, for which 14.1% of the patients fulfilled the criteria, whereas the prevalence of the other somatoform diagnosis was between 3.0% and 8.1%. The most frequent DSM-IV diagnoses were SD NOS and undifferentiated SD, which 29.93% and 27.3% of the interviewed patients, respectively, received, whereas the prevalence of the other diagnoses was between 1.0% and 8.1%. A high comorbidity between SDs and other mental disorders was found. The general practitioners identified between 50% and 71% of the patients with an ICD-10 SD and between 36% and 48%, according to DSM-IV criteria. Patients with SDs used more nonpsychiatric health care facilities than other patients (P 0.01). (Psychosomatics 1999; 40: ) Studies in which standardized methods have been used show that from 14% to 36% of patients in primary care settings in western countries have mental disorders. 1,2 Somatoform disorders in primary care are less intensely studied compared with other mental disorders, and only a few Received April 22, 1998; revised November 4, 1998; accepted December 4, From the Department of Psychiatric Demography, Institute for Basic Psychiatric Research, Psychiatric Hospital in Aarhus, Aarhus University Hospital, DK 8240 Risskov, Denmark. Address correspondence and reprint requests to Dr. Fink at the same address; perfink@aaa.dk. Copyright 1999 The Academy of Psychosomatic Medicine. studies have used standardized criteria for somatoform disorders. Bridges and Goldberg 3 examined 500 patients consulting their general practitioners (GPs) in Manchester because of new illness episodes. By using the Psychiatric Assessment Schedule interview, the researchers found that one-third of the patients fulfilled the criteria for a psychiatric disorder, according to DSM-III. More than half of the patients with a psychiatric diagnosis also fulfilled some ad hoc-defined research criteria for somatization, and one-fifth of all consultations were due to somatization. In Montreal, Kirmayer and Robbins 4,5 examined 685 consecutive pa- 330 Psychosomatics 40:4, July-August 1999

2 Fink et al. tients who consulted a general family medical clinic. By using the Diagnostic Interview Schedule (DIS), the Somatic Symptom Index, and a modified version of the Whiteley Index (WI), the researchers found that a total of 26.3% of the patients fulfilled the criteria for one or more types of somatization. In a worldwide multicenter study, Üstün et al. 1 found that 5.4% of the patients fulfilled the International Classification of Diseases-10th Revision (ICD-10) diagnostic criteria for neurasthenia, 2.7% for somatization disorder, and 0.8% for hypochondriasis, according to the Composite International Diagnostic Interview, Primary Health Care version. By conducting a standardized psychiatric interview, Lobo et al. 6 found that 9.4% of the patients in a study on primary care in Spain fulfilled the Bridges and Goldberg criteria for somatization. All referred studies 1,3 6 showed a high comorbidity of somatization illness with other mental disorders. Mental disorders are only diagnosed in 33% 60% of the cases by the GPs, 4,7 but whether somatization is recognized by GPs has only been sparsely studied. Usually, the diagnosis of a somatoform disorder in nonpsychiatric settings is a diagnosis of exclusion that is made when organic pathology beyond all doubt has been ruled out. More problematic, this strategy delays the diagnosis, and thereby the possibility of a sufficient treatment, making the detection process very costly, which results in a heavy use of health care resources By use of a standardized methodology, the aims of this study were to find the prevalence of somatoform disorders in a primary care sample of consecutive patients, the comorbidity with other mental disorders, to investigate GPs ability to detect somatization, and to study the impact of somatization on health care utilization. METHODS In our study, 223 consecutive patients, residents in Aarhus, Denmark (age 18 65), consulting their GP, were asked to participate in the study. Of these, 191 accepted (134 [70.2 %]: females, 57 [29.8%]: males). The median age of the sample was 34.4 (Q1 Q3: ). Q1 stands for 25th percentile; Q3 stands for 75th percentile. The 191 participating patients were presented with a self-rating questionnaire including a modified 25-item version of the Hopkins Symptom Checklist (SCL-90). 11 This SCL-25 includes most of the items on the anxiety and the depression subscales of the full SCL-90 version and 6 out of 12 items in the full SCL-90 somatization subscale. The patients rated each item on a 4-point distress scale ranging from 1 not at all to 4 very much. Patients with a mean score of 1.55 or higher on the SCL-25 (n 44), and a corresponding number randomly selected among patients scoring below 1.55 (n 55), were chosen for a SCAN interview (Schedules for Clinical Assessment in Neuropsychiatry). 12 This interview is the latest version of the Present State Examination (i.e., PSE-10) and one of the standardized psychiatric interviews recommended by the World Health Organization. 2,7 The two interviewers (PF and ME) are psychiatrists certified in the use of this interview. Of the interviewed patients, 74.7% were females and 25.3% males, and the subsample s median age was 37.2 (Q1 Q ). The 99 interviewed patients were, moreover, given a 14-item version of the WI for hypochondriacal traits. 13,14 The psychometric performance of this test has been reported in an accompanying paper. 15 The population prevalence was calculated according to the method described by Tarnopolsky et al., 16 that is, prevalence (N [high score] positive predictive value)/ (N [low score] low score predictive value)/n. The somatization subscale of the SCL-25 was used to divide the patients into low scores (0 2) and high scores ( 2). The SCAN interviews were processed with the CATEGO computer program, 17 which creates an Index-of- Definition (ID). The ID consists of eight levels of severity, ranging from no disturbances to maximum severity. As in other studies using the SCAN/PSE instrument, patients with an ID of 5 or higher were defined as psychiatric cases, as they fulfill the diagnostic criteria for an ICD-10 diagnosis excluding the group of somatoform disorders. The CATEGO computer program has not yet been sufficiently refined on the physical health section of the SCAN, which covers the somatoform disorder diagnoses. Therefore, it was necessary to process data from the physical health section of the SCAN separately by algorithms based on ICD and DSM-IV research criteria, 19 respectively. GP Recognition After seeing the patient, the GPs were asked to complete a questionnaire on the consultation. Among other questions, they were asked to assign the reason for consultation into one of six predefined categories. The responses were in this study dichotomized, in the way that the two categories ill-defined conditions and mental illness with physical symptoms were classified as somatiz- Psychosomatics 40:4, July-August

3 ers, according to the GP, and all other responses as GP nonsomatizers. Second, the GPs were asked whether they judged the patient to be more preoccupied with physical symptoms and bodily sensations than could be expected from their medical conditions, rating on a 4-point scale from no to in the highest degree. The responses were dichotomized with a cut-point between no and all other responses. Statistics The data were processed with the Statistical Package for the Social Sciences Windows release Mann- Whitney U-test, chi-square test, and Fisher s exact test were used. Kappa statistics were used to express agreement between GP rating and diagnoses based on psychiatric interviews. RESULTS Of the interviewed patients, 60.6% (56% of males and 62.5% of females) presented at least one medically unexplained symptom, according to the SCAN interview, and 24.2% fulfilled diagnostic research criteria for at least one ICD-10 somatoform disorder, excluding somatoform disorder, unspecified, 26.3% if dissociative diagnoses were included (Table 1). Also, 59.9% met diagnostic research criteria for at least one DSM-IV somatoform disorder, 30.3 %, excluding the not otherwise specified (NOS) diagnostic group, and 13.1 % excluding both the NOS and undifferentiated somatoform diagnostic groups. There were no statistical significant differences as to age or gender (P 0.10). When applying these results on the total sample of 191, a prevalence of somatoform disorders, excluding somatoform disorder (unspecified), as to ICD-10 criteria can estimated to be 22.3% (CI: 95%: ) and 57.5% (CI: 95%: ) as to DSM-IV criteria, and 30.3% (CI:95%: ) when the NOS group is excluded. Table 1 also displays the distribution of diagnosis. There were no statistically significant differences as to age or gender in any category. The four patients with a diagnosis of neurasthenia also fulfilled criteria for an ICD-10 somatoform disorder. The most frequent ICD-10 diagnosis was autonomous dysfunction, for which 14.1% of the patients fulfilled the criteria, whereas the prevalence of the other diagnoses was between 3.0% and 8.1%. There was a complete concordance between ICD-10 and DSM-IV pain disorder and hypochondriasis. However, there was a marked difference in SD prevalence, especially when DSM-IV and ICD criteria were used for the undifferentiated SD diagnosis: 27.3% of the patients fulfilled the DSM-IV criteria, whereas 7.1% met the ICD-10 criteria. The most frequent DSM-IV diagnosis was somatoform disorder NOS, which 29.3% of the interviewed patients received. One-quarter of the patients with either a DSM-IV undifferentiated somatoform disorder or NOS diagnosis fulfilled the diagnostic criteria for the ICD-10 autonomous dysfunction diagnosis (29.6% and 20.7%, respectively). Twelve patients fulfilled the diagnostic criteria for only one of the ICD-10 diagnoses listed in Table 1, nine patients received two, and five received three or more different diagnoses. Thirty-one patients fulfilled the diagnostic criteria for one DSM-IV diagnosis, eight for two, and three for three or five of the diagnoses listed in Table 1. Table 1 also shows that there was a high comorbidity of other mental disorders. Sociodemographic Characteristics The ICD-10 somatizing patients belonged to low social class and were less educated; more were unemployed or on a pension; and more were living on sick pay, social security benefits, or pension (Table 2). There were no statistically significant differences as to living alone or with a partner or having children at home or not. The picture was the same if the DSM-IV criteria excluding the NOS diagnostic group was used. However, if including the NOS group, there was no statistically significant difference as to education and social group between the nonsomatizers and the somatizers. Utilization of Health Care Resources Table 3 shows the utilization of nonpsychiatric health care, according to the patients during the year before the index contact. The patients with an ICD-10 somatoform disorder at median had 10 (Q1 Q3: 6 21) contacts with GPs and other physicians or general hospital admissions for physical ailments during the previous year, compared with 5 (Q1 Q3: 3 10) among patients without the ICD-10 somatoform disorder (P 0.004). The differences were less pronounced if the DSM-IV criteria were used and did not reach statistical significance at a 5% level in any of the variables listed in Table 3. No patients had been admitted to a psychiatric department during the prior year. Two patients with a somatoform disorder (either ICD-10 or DSM-IV criteria) had 332 Psychosomatics 40:4, July-August 1999

4 Fink et al. TABLE 1. Diagnostic distribution SCAN Interviewed, % Total Sample (N 199) Calculated Psychiatric All Males Females Prevalence a Comorbidity, ICD-10 Research criteria (N 99) (n 24) (n 75) % CI 95% % Somatization disorder (SD) (F45.0) Undifferentated SD b (F45.1) Pain disorder (F45.4) Hypochondriasis (F45.2) Autonomous dysfunction (F45.3) Neurasthenia (F48.0) Dissociative movement and sensibility (F44.4 6) Dissociative other (F44.1 3) Above combined SD combined (F45.X) Excluding autonomous dysfunction DSM-IV research criteria SD (300.81) Undifferentiated SD (300.81) Pain disorder c (307.8) Hypochondriasis (300.7) Conversion disorder (300.11) Somatoform disorder, not otherwise specified (NOS) (300.81) SD combined Excluding SD NOS Excluding undifferentiated and NOS SD Note: SCAN Schedules of Clinical Assessment in Neuropsychiatry. a An index of definition of 5 or higher on the SCAN interview, excluding the physical section. Percentage of cases. b Criteria for SD, except duration 2years and 6 months. c Duration 6 months. TABLE 2. Sociodemographic characteristics, statistical significance only ICD-10 Somatoform (N 24) a No Somatoform (N 75) a N % N % v 2 Social group Unskilled P Skilled Middle or above Education Primary school or less P Higher Employment Unemployed and pension/retired P Other Economy a Social security, pension, disability benefit P Other Note: ICD-10 International Classification of Diseases-10th Revision. a Missing information on some of the listed variables. Psychosomatics 40:4, July-August

5 visited a psychiatric specialist and one a community psychiatric service (P 0.05). Five of the patients with a somatoform disorder, compared with one of the patients without, had been to a psychologist (P by ICD-10 criteria, P by DSM-IV criteria). GP Recognition Table 4 displays the GP s recognition of somatization and awareness of whether psychosocial factors may influence the patient s condition. The GPs rated that 40.4% of the interviewed patients were more than expected preoccupied with bodily symptoms or that somatization (i.e., illdefined conditions or mental illness with physical symptoms) was the reason for the consultation. Of the patients with an ICD-10 somatoform disorder, the GPs rated 71% as more preoccupied with bodily sensations and illness than expected from their medical condition. In half of the patients with an ICD-10 somatoform disorder, the GPs rated that somatization was the reason for the consultation and in 66.7 % of the cases that psychosocial factors were affecting the patient s condition. The association between the interview-based diagnoses of somatoform disorders and the GP s rating was statistically significant (P 0.01) on all the indicators of somatization. However, the agreement was only fair-to-moderate, with kappa values in the range (Table 4). The agreements between diagnostics interview and GP rating were best when the ICD-10 TABLE 3. Utilization of health care resources the previous year according to the patient ICD-10 Somatoform DSM-IV Somatoform Other Patients M-W Other Patients M-W (N 24) (N 75) U-test (N 59) (N 40) U-test Median Q1 Q3 Total a Median Q1 Q3 Total a P Median Q1 Q3 Total a Median Q1 Q3 Total a P Emergency visit on call GP or visit to general hospital emergency c c Specialist or outpatient service b d d Visits to GP c c c c Inpatient admission to general hospital Nonpsychiatric services combined Note: GP general practitioner; Q1 25th percentile; Q3 75th percentile; M-W U-test Mann-Whitney U-test. a Total for the patient subgroup. b Nonpsychiatric only. c Including one patient who has stated 99 contacts. d Including two patients who have stated 99 contacts. TABLE 4. General practitioner recognition of somatoform disorders Total (N 99) ICD-10 Criteria Somatoform (N 24) Other Patients (N 74) v 2 Kappa DSM-IV Criteria Somatoform (N 59) Other Patients (N 40) v 2 Kappa Psychosocial factor affecting the condition a Somatization as reason for consultation Bodily preoccupation Mental illness Note: ICD-10 International Classification of Diseases-10th Revision. a One patient has missing value. 334 Psychosomatics 40:4, July-August 1999

6 Fink et al. diagnostic criteria were used, compared with the less restrictive DSM-IV criteria. DISCUSSION Depending on the diagnostic criteria used, this study shows that between 22% and 58% of consecutive patients in primary care fulfill the diagnostic criteria for a somatoform disorder. Prior studies have found similar high prevalence (20% 26%) despite not quite similar assessment methods or diagnostic criteria for somatization illness. 3 6 Therefore, the prevalence of somatoform disorders is as high as the overall prevalence between 14% and 36% of mental disorders shown in prior studies done in primary care settings. 1,2 A high comorbidity between somatoform disorders and other mental disorders was found, in concordance with other studies The comorbidity was lowest in the less severe of the somatoform disorders, that is, 20% in the DSM-IV somatoform disorder NOS and highest in the most severe, that is, 84% 100% in the somatization disorder. These findings support prior findings on a correlation between the severity of the somatization illness and the severity of psychological distress/symptoms, that is, the more severely ill a patient is, the more prone he/she is to experience physical and psychological symptoms Prior studies on mental illness in primary care have only insufficiently investigated for somatoform disorders, 1,2 and our finding indicates that about 10% 20% may be added to the overall prevalence figures of mental disorders in primary care if the somatoform diagnostic category is included. The wide range in the overall prevalence of somatoform disorders in this study is caused by the differences between the DSM-IV and the ICD-10. The main differences can primarily be attributed to the high prevalence of DSM-IV NOS diagnosis and to a lesser extent to the undifferentiated somatoform diagnosis, whereas there was a complete concordance between ICD-10 and DSM-IV as to pain disorder and hypochondriasis. One patient fulfills the diagnostic criteria for the somatization disorder diagnosis in the DSM-IV system, compared with four in the ICD-10 system. This difference is difficult to interpret, as each of the four symptom groups in the DSM-IV criteria for a somatization disorder was about equally prevalent, but when combined the prevalence dropped to one patient only. Therefore, it seems to be dubious to use a checklist with predefined specific physical symptoms instead of checking for all kinds of medically unexplained physical symptoms and the number of organ systems these can be referred to. 27,28 The prevalence of somatization disorder may be underestimated in the present study, as the diagnoses are based on interviews, and a review of lifetime medical case notes may have revealed more medically unexplained symptoms. 29 The marked difference in prevalence between DSM-IV and ICD-10 undifferentiated somatoform disorder diagnosis results because only one medically unexplained symptom is required in the DSM-IV criteria, whereas multiple symptoms are required in the ICD-10 criteria. Several studies have addressed the problem on the number of medically unexplained symptoms that are needed for a somatoform diagnosis. Based on a series of studies in community samples using the DIS, Escobar and co-workers and others 32 have suggested an abridged somatization disorder construct requiring at least four symptoms for males and six for females. Based on a study in primary care using the PRIME-MD, Kroenke et al. 33 have suggested that the undifferentiated somatoform disorder diagnosis in DSM is a replacement for a multisomatoform disorder diagnosis. 33 The criteria for this diagnosis is closer to the ICD-10 criteria for the undifferentiated SD diagnosis than the DSM-IV criteria and may be more valid for use in primary care. Generally, the ICD-10 criteria demands a pattern of repeated presentations of medically unexplained physical symptoms, together with persistent requests for medical investigations. These demands are not included in the DSM-IV diagnostic system and could explain the high prevalence of the DSM-IV NOS diagnosis in this study. The importance of the time criteria is underscored by the fact that the most prevalent ICD-10 diagnoses were autonomous dysfunctions applied to 14.1% of the patients, which do not explicitly list any time criteria. If the time criteria are excluded, 26.3% fulfill the criteria for a pain disorder, and 14.1% meet the criteria for hypochondriasis. These findings are in concordance with other studies on hypochondriasis in primary care. 34,35 Thus, other somatoform diagnoses would be at least as frequent as autonomous dysfunction, ignoring the time criteria and the overall prevalence of somatoform disorders by ICD-10 criteria, as would the DSM-IV prevalence if the persistence of the complaints was ignored. Medically unexplained or functional physical symptoms are very frequent in primary care. Kroenke and Mangelsdorff 36 showed that in some of the most common symptoms such as chest pain, fatigue, dizziness, headache, dyspnea, etc., it is only between one-tenth and one-fourth of cases that an organic etiology can be found. 36 In the Psychosomatics 40:4, July-August

7 DSM-IV system, these patients with complaints of no organic pathology will qualify for at least an NOS somatoform disorder diagnosis. But as the symptoms in many cases are transient and the patients are easily reassured by the doctor, and no connection between distressing event or problems is evident, these patients will not qualify for an ICD-10 somatoform disorder diagnosis. One may argue that it is a normal illness behavior to seek reassurance by the doctors for physical symptoms, and this ought not be diagnosed as a disorder. On the other hand, physical sensations and symptoms are very common in the general population, but only a small proportion seek a medical checkup, which is why it may be viewed as abnormal illness behavior This problem of diagnostic labeling is most urgent in primary care settings; generally, the symptoms would be much more troublesome in the patients referred to inpatient clinics or specialized medical settings and will normally qualify for a somatoform disorder diagnosis without debate. The present diagnostic systems reflect that they are based on studies in specialized settings and an adoption to primary care settings, where the patient may be much earlier in the course of illness may be needed. This need is supported by previous reports on a high prevalence of a transient form of hypochondriasis in primary care. 34,35 One way to meet this need would be to introduce a diagnostic category for patients who present transient medically unexplained or functional complaints and are easily reassured that they do not suffer from a genuine physical disease. This step would also counteract the tendency by many doctors to use unfounded odd diagnostic labels that, although unintentional, may indicate to the patient a somatic disease that worries the patient instead of reassures him/her. In the present study, no statistically significant differences as to gender were found in any diagnostic category and, except for autonomic dysfunction, DSM-IV undifferentiated somatoform disorder and the NOS category, the tendency was that the prevalence was equal or higher among males. Other studies have found that hypochondriasis and pain disorders are more or equally prevalent in males, compared with females, 5,40 whereas other studies have found that somatoform disorders are more prevalent among females and furthermore that females in general report more symptoms than males. 3,5,41,42 However, as the figures in our study were small when broken into diagnostic categories, it is impossible to draw any conclusions on the gender distribution in somatization illnesses. As the Danish health care system is a family doctor system and is free of charge for the patients, and since the participating doctors were randomly selected for inclusion, no obvious selection bias as to sociodemographic, cultural, or other variables seems to have affected the results. The results may point to the fact that gender distributions cannot be established definitely on the basis of samples drawn from medical settings, as females and males have a different threshold for seeing a doctor, as reflected in this study in which three-fourths of the sample were females. To avoid confounding by the differences in illness behavior as to gender and sociocultural factors, it is necessary to use general population samples. The GPs rated in 26.3% of the interviewed patients somatization as the reason for the consultation. If also including the response more than expected preoccupied with bodily symptoms as an indicator for somatization, 40.4% of the patients were rated positively by the GPs; thus, the GPs by and large agreed on the prevalence of somatization in primary care. The agreement between the diagnostic SCAN interview ratings and the GP s rating was, however, only modest, which means that the rating, only to some extent, was the same for the patients who were rated as somatizers by the GPs and the SCAN. The concordance was far better in the more restrictive ICD-10 diagnostic criteria excluding the somatoform disorder (unspecified group) than the less restrictive DSM-IV criteria including the NOS category. The GPs recognized from 50% to 71% of the patients with an ICD-10 somatoform disorder, according to the SCAN interview, and from 36% to 48%, according to the DSM-IV criteria. In a study on hypochondriacal patients in primary care, Gerdes et al. 43 found similar incidence figures. Therefore, the recognition rate is at the level of other mental disorders. 1,7 The doctors participating in this study are probably more psychologically minded than average doctors, 2 so the recognition rate may be lower in primary care in general. The hidden cases, that is, patients who have a somatoform disorder, according to the SCAN interview, but are negatively rated by the GPs, may be caused by lack of knowledge among the GPs about somatization or functional disorders or the unfamiliarity of the concept. On the basis of the earlier discussion, another explanation may be that the doctors found it quite normal that a patient seeks medical examination for physical complaints that are worrying him/her and would not stigmatize the patients by labeling them as somatizers, implicating a mental disorder, if the patients otherwise seem healthy. Patients whom the GPs rated as somatizers but who were noncases, according to the SCAN interview, are interesting. Are these patients 336 Psychosomatics 40:4, July-August 1999

8 Fink et al. simulating at the interview or are there other explanations? The results of this study cannot yield the answers. The results show that the somatizing patients use more nonpsychiatric health care resources than the other patients, but these results are not statistically significant when the less restrictive DSM-IV criteria are used. The relevance of comparing somatizing patients with physical-diseased patients may, however, be questioned. A great part of the somatizing patient s overuse of nonpsychiatric health care may be avoided by strengthening the psychosocial component in treatment. 44,45 References 1. Üstün TB, Sartorius N: Mental Illness in General Health Care: An International Study. Geneva, Switzerland, World Health Organization, and New York, John Wiley & Sons, Fink P, Jensen J, Borgquist L, et al: Psychiatric morbidity in primary public health care. A Nordic multicenter investigation. Part I: method and prevalence of psychiaric morbidity. Acta Psychiatr Scand 1995; 92: Bridges KW, Goldberg DP: Somatic presentation of DSM-III psychiatric disorders in primary care. J Psychosom Res 1985; 29: Kirmayer LJ, Robbins JM, Dworkind M, et al: Somatization and the recognition of depression and anxiety in primary care. Am J Psychiatry 1993; 150: Kirmayer LJ, Robbins JM: Three forms of somatization in primary care: prevalence, co- occurrence, and sociodemographic characteristics. J Nerv Ment Dis 1991; 179: Lobo A, Garcia-Campayo J, Campos R, et al: Somatisation in primary care in Spain: I. Estimates of prevalence and clinical characteristics. Working Group for the Study of the Psychiatric and Psychosomatic Morbidity in Zaragoza. Br J Psychiatry 1996; 168: Munk-Jørgensen P, Fink P, Brevik JI, et al: Psychiatric morbidity in primary public health care. A multicentre investigation. Part II: hidden morbidity and choice of treatment. Acta Psychiatr Scand 1997; 95: Fink P: Admission patterns of persistent somatization patients. Gen Hosp Psychiatry 1993; 15: Fink P: The use of hospitalizations by persistent somatizing patients. Psychol Med 1992; 22: Quill TE: Somatization disorder. One of medicine s blind spots. JAMA 1985; 254: Derogatis LR, Lipman RS, Covi L: SCL-90: an outpatient psychiatric rating scale preliminary report. Psychopharmacol Bull 1973; 9: Wing J, Babor T, Brugha T: SCAN. Schedules for Clinical Assessment in Neuropsychiatry. Geneva, Switzerland, World Health Organization, Division of Mental Health, Pilowsky I: Dimensions of hypochondriasis. Br J Psychiatry 1967; 113: Barsky AJ, Wyshak G, Klerman LG: Hypochondriasis. An evaluation of the DSM-III criteria in medical outpatients. Arch Gen Psychiatry 1986; 43: Fink P, Ewald H, Jensen J, et al: Screening for somatization and hypochondriasis in primary care and neurological inpatients. A 7- item scale for hypochondriasis and somatization. J Psychosom Res 1999; 46: Tarnopolsky A, Hand DJ, McLean EK, et al: Validity and uses of a screening questionnaire (GHQ) in the community. Br J Psychiatry 1979; 134: Wing J, Cooper JE, Sartorius N: Measurement and classification of psychiatric symptoms. An instruction manual for the PSE and CA- TEGO program. Cambridge, UK, Cambridge University Press, World Health Organization ( WHO): The ICD-10 Classification of Mental and Behavioural s. Diagnostic Criteria for Research. Geneva, Switzerland, WHO, American Psychiatric Association: Diagnostic and Statistical Manual of Mental s, 4th Edition. Washington, DC, American Psychiatric Association, Norusis MJ: Statistical Package for the Social Sciences (SPSS) for Windows. Release 6.0. Chicago, IL, SPSS, Tomasson K, Kent D, Coryell W: Somatization and conversion disorders: comorbidity and demographics at presentation. Acta Psychiatr Scand 1991; 84: Barsky AJ, Wyshak G, Klerman GL: Psychiatric comorbidity in DSM-III-R hypochondriasis. Arch Gen Psychiatry 1992; 49: Fink P: Psychiatric illness in patients with persistent somatisation. Br J Psychiatry 1995; 166: Simon G, Gater R, Kisely S, et al: Somatic symptoms of distress: an international primary care study. Psychosom Med 1996; 58: Katon W, Lin E, Von Korff M, et al: Somatization: a spectrum of severity. Am J Psychiatry 1991; 148: Kroenke K, Spitzer RL, Williams JB, et al: Physical symptoms in primary care. Predictors of psychiatric disorders and functional impairment. Arch Fam Med 1994; 3: Fink P: Physical complaints and symptoms of somatizing patients. J Psychosom Res 1992; 36: Mayou R, Bass C, Sharpe M: Overview of epidemiology, classification, and aetiology, in Treatment of Functional Somatic Symptoms, edited by Mayou R, Bass C, Sharpe M. Oxford, UK, Oxford University Press, 1995, pp Bass C: Somatization: Physical Symptoms & Psychological Illness. Oxford, UK, Blackwell Scientific, 1990, pp Escobar JI, Burman MA, Karno M, et al: Somatization in the community. Arch Gen Psychiatry 1987; 44: Escobar JI, Golding JM, Hough RL, et al: Somatization in the community: relationship to disability and use of services. Am J Public Health 1987; 77: McKegney FP: The incidence and charateristics of patients with conversion reactions: I. A general hospital consultation service sample. Am J Psychiatry 1967; 124: Kroenke K, Spitzer RL, degruy FV, et al: Multisomatoform disorder. An alternative to undifferentiated somatoform disorder for the somatizing patient in primary care. Arch Gen Psychiatry 1997; 54: Barsky AJ, Wyshak G, Klerman GL: Transient hypochondriasis. Arch Gen Psychiatry 1990; 47: Robbins JM, Kirmayer LJ: Transient and persistent hypochondriacal worry in primary care. Psychol Med 1996; 26: Kroenke K, Mangelsdorff AD: Common symptoms in ambulatory Psychosomatics 40:4, July-August

9 care: incidence, evaluation, therapy, and outcome. Am J Med 1989; 86: Kellner R, Sheffield BF: The one-week prevalence of symptoms in neurotic patients and normals. Am J Psychiatry 1973; 130: Rasmussen NK, Groth MV, Bredkjær SR, et al: Sundhed & Sygelighed i Danmark. En raport fra DIKEs undersøgelse. København, Dansk Institut for Klinisk Epidemiology, 1987, pp Kirmayer LJ, Robbins JM: Functional somatic syndromes, in Current Concepts of Somatization, edited by Kirmayer LJ, Robbins JM. Washington, DC, American Psychiatric Press, 1991, pp Barsky AJ, Frank CB, Cleary PD, et al: The relation between hypochondriasis and age. Am J Psychiatry 1991; 148: Golding JM, Smith GR Jr, Kashner TM: Does somatization disorder occur in men? Clinical characteristics of women and men with multiple unexplained somatic symptoms. Arch Gen Psychiatry 1991; 48: Kroenke K, Spitzer RL: Gender differences in the reporting of physical and somatoform symptoms. Psychosom Med 1998; 60: Gerdes TT, Noyes RJ, Kathol RG, et al: Physician recognition of hypochondriacal patients. Gen Hosp Psychiatry 1996; 18: Smith GR Jr, Rost K, Kashner TM: A trial of the effect of a standardized psychiatric consultation on health outcomes and costs in somatizing patients. Arch Gen Psychiatry 1995; 52: Smith GR Jr, Monson RA, Ray DC: Psychiatric consultation in somatization disorder. A randomized controlled study. N Engl J Med 1986; 314: Psychosomatics 40:4, July-August 1999