Journal of Psychosomatic Research 76 (2014) 200–206
Contents lists available at ScienceDirect
Journal of Psychosomatic Research
The PROSPECTS study: Design of a prospective cohort study on prognosis and perpetuating factors of medically unexplained physical symptoms (MUPS) Nikki van Dessel a,⁎, Stephanie S. Leone b, Johannes C. van der Wouden a, Joost Dekker a,c,d, Henriëtte E. van der Horst a a
Department of General Practice and Elderly Care Medicine, EMGO Institute, VU University Medical Center Amsterdam, The Netherlands Department of Public Mental Health, Netherlands Institute of Mental Health and Addiction, Utrecht, The Netherlands c Department of Rehabilitation Medicine and Department of Psychiatry, VU University Medical Center Amsterdam, The Netherlands d Reade, Centre of Rehabilitation and Rheumatology, Amsterdam, The Netherlands b
a r t i c l e
i n f o
Article history: Received 17 October 2013 Received in revised form 27 December 2013 Accepted 28 December 2013 Keywords: Cohort study Medically unexplained physical symptoms (MUPS) Prognosis Prediction Somatoform disorders Somatization
a b s t r a c t Objective: This paper describes the rationale and methodology of the PROSPECTS study, a study which aims to assess the course and prognosis of medically unexplained physical symptoms (MUPS), in terms of symptom severity and physical and social functioning. Additionally, it aims to identify different course types and to determine which factors are associated with these course types. Based on these factors, one or more prediction models will be developed. Methods: This study is a prospective, multicenter longitudinal cohort study with 1 baseline and 4 follow-up measurements during a 3 year period. 450 MUPS patients (age 18–70 years) will be included, divided over a primary care group, recruited in general practices, and a secondary/tertiary care group, recruited in specialized MUPS care programs. Main study parameters/endpoints: Primary outcome measures are severity of symptoms and degree of functional impairment. Secondary outcome measures are health care consumption and level of depressive symptoms and anxiety. Potential predictors are based on current theoretical models describing the perpetuation of MUPS and include somatic, psychological and social factors. Latent Class Growth Mixture Modeling will be used to identify distinct course types. Logistic regression analysis will be used to identify risk factors associated with these course types. Finally, one or more multivariate prediction models for the course of MUPS will be developed and tested. Conclusion: The PROSPECTS study aims to enhance our insight into the course of MUPS, thus contributing to better recognition of future patients at risk for persistent MUPS. © 2014 Elsevier Inc. All rights reserved.
Background In all health care settings patients present with physical symptoms, such as fatigue, dizziness and pain, for which no sufﬁcient somatic explanation is found after proper medical examination. Such symptoms are called medically unexplained physical symptoms (MUPS). MUPS are very common, especially in primary care. Around 25–50% of the complaints that patients present to their general practitioner (GP) remain unexplained . Fortunately, almost 80% of these episodes remain restricted to one doctor–patient contact , indicating that most MUPS ⁎ Corresponding author at: Department of General Practice and Elderly Care Medicine, EMGO Institute, VU University Medical Center, PO Box 7700, 1100 SN Amsterdam, The Netherlands. Tel.: +31 20 4449678, +31 6 45213235 (mobile); fax: +31 20 4448361. E-mail address: [email protected]
(N. van Dessel). 0022-3999/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpsychores.2013.12.011
may be self-limiting. However, around 20–30% of patients develop persisting symptoms, which can be severe and disabling [3,4]. Due to the variation in presentation and duration, MUPS can be regarded as a continuum ranging from mild (and often self-limiting) symptoms, to chronic severe symptoms, which are also seen in functional somatic syndromes such as Irritable Bowel Syndrome and Chronic Fatigue Syndrome. At the severe end of the continuum, symptoms are often more numerous and psychiatric co-morbidity often occurs . The diversity in the nature and severity of symptoms creates challenges in deﬁning and describing the various forms of MUPS . In this study we choose to use the term ‘MUPS’ as a general description of the entire spectrum of physical symptoms, which last at least several weeks, and for which no sufﬁcient explanation can be found after proper medical examination. This is in line with the recently published Dutch ‘Multidisciplinary Guideline for MUPS and Somatoform Disorders’  and gives a
N. van Dessel et al. / Journal of Psychosomatic Research 76 (2014) 200–206
neutral description of the patient's symptoms, without suggesting a (physical or psychological) causal explanation. Costs and societal relevance of MUPS Patients with persistent MUPS have great risk of functional impairment and experience high levels of psychological distress [7–9]. Additionally, persistent MUPS are associated with high costs . It is known that MUPS patients are often exposed to unnecessary diagnostic procedures and may use superﬂuous medication [11–13]. Outpatient and inpatient medical care utilization is approximately twice as high in patients suffering from severe MUPS, when compared to patients without MUPS . Total costs for this patient group are even higher, due to work- and insurance related costs . In the Netherlands, it is estimated that 30% of long term absence of work is caused by MUPS . Mechanisms contributing to MUPS Several theories are available for the mechanisms that play a role in the development and persistence of MUPS. The cognitive behavioral model is seen as a meta-model, incorporating many of these theories [17,18]. It provides explanations for physical symptoms in different domains, including somatic causes, illness perceptions, illness behavior and illness predispositions. One of the theories incorporated in the model is the somatosensory ampliﬁcation theory, which suggests that a physical sensation leads to increased attention to this sensation, which in turn leads to (faulty) attributions and cognitions about it. This creates a vicious circle, as it ampliﬁes the symptom perception . A second theory incorporated in the model is the sensitivity theory. This theory suggests that some individuals are more vulnerable for developing or maintaining physical symptoms than others. Factors that have been found to be related to this vulnerability are personality traits, such as neuroticism, catastrophic thinking and traumatic experiences in early childhood . A third theory is based on the fact that stress (physical or psychological) inﬂuences the bodily hormonal stress system: the hypothalamic pituitary adrenal axis (HPA axis). Prolonged stress may lead to HPA axis down regulation and reduced cortisol production. As a result stress sensitivity increases . This theory reﬂects the interplay between body and mind and may therefore provide a concrete link between psychological burden and physical symptoms. A narrative review showed some evidence for the inﬂuence of these and other theoretical elements on MUPS . However, it is unknown which elements play the most important role in the persistence of MUPS. Better insight in these contributing factors might provide clues for treatment of MUPS. Prognosis and treatment of MUPS Little is known about the course of MUPS. In a recent review, Olde Hartman et al. summarized cohort studies about the course and prognosis of MUPS . They found that very few, highly heterogenic, studies have been performed. Additionally, included studies had methodological ﬂaws. Baseline duration of symptoms was often unknown, possible treatments were not described and duration of follow-up was generally short (6 to 15 months). They concluded that although 50% of patients improve or recover completely, the symptoms of 10–30% of patients with MUPS deteriorate or become chronic. Due to the heterogeneity in the studies, they could not identify prognostic factors for the course of MUPS. Due to the short follow-up, no conclusions could be drawn about the stability of the short term outcomes (e.g. the long term symptom recurrence rate after an initial recovery is currently unknown). Interventions used to treat persistent MUPS have shown disappointingly little effect. For cognitive behavioral therapy (CBT) empirical
support has been found in some studies, but effect sizes were small. A possible explanation could be the lack of focus on a well deﬁned target population. Identiﬁcation of prognostic subgroups would make it possible to offer CBT to those patients who actually need and may beneﬁt from it. Additionally, it is unclear which components of CBT are effective. This may be caused by the lack of knowledge about factors playing the most important role in inﬂuencing the course of MUPS [22–24]. Based on current knowledge, we can say that MUPS is mostly selflimiting, but when MUPS persist, they have a great personal and societal impact. We know very little about the long-term course of MUPS. It is unclear which percentage of patients develops persistent MUPS in different health care settings and which factors contribute to the persistence of MUPS. Given these gaps in current knowledge, the PROSPECTS study has 3 main study objectives: Aims of this study 1. To assess the long-term course of MUPS presented in different settings (primary care and secondary/tertiary care) in terms of severity of symptoms and functional impairment. 2. To identify different course types, based on the course of symptom severity and functional impairment. 3. To determine which baseline characteristics are associated with favorable and unfavorable course types and to develop a multivariate prediction model for the course of MUPS. Methods Study design This is a prospective cohort study of patients with MUPS in multiple health care settings. The duration of follow-up will be 3 years. After baseline measurement, follow-up measurements will take place after 6, 12, 24 and 36 months. Information will be collected through questionnaires and saliva samples. The Medical Ethics Committee of the VU University Medical Center approved the study protocol (May 10th 2013). Written informed consent will be obtained from all study participants. Health care settings In primary care, the study will be carried out in general practices linked to the VU University Medical Center. These practices are located in urban as well as rural areas across the Netherlands. Approximately 50 general practitioners will be included, who have experience with the electronic ICPC (International Classiﬁcation of Primary Care) coding system . The structure of the ICPC coding system is, though less detailed, comparable with the ICD classiﬁcation (the International Classiﬁcation of Diseases) . In secondary and tertiary care the study will be conducted in organizations, which are participating in an integrated care program for MUPS patients in greater Amsterdam. These are the psychiatric department of the VU University Medical Center (VUmc); GGZ Ingeest, a secondary mental health care organization; and Reade, a center for rehabilitation medicine. Patients Inclusion criteria The study population will consist of patients between 18and 70 years old, suffering from MUPS. We deﬁne MUPS as the presence of physical symptoms, which have lasted at least several weeks and for which no sufﬁcient explanation has been found after proper medical examination by a physician. Additionally, the patient has to have a score of 2 for at least one symptom of the PHQ-15 questionnaire (indicating that the symptom is bothering a lot).
N. van Dessel et al. / Journal of Psychosomatic Research 76 (2014) 200–206
The PHQ-15 [27,28] is a frequently used and validated questionnaire about physical symptoms. In the original validation study, cut offs of 5 and 10 were suggested, based on the correlation of these scores with the presence of a somatoform disorder according to the PRIME-MD diagnostic interview for common mental disorders . However, as our deﬁnition of MUPS only requires the existence of at least one bothering symptom as a criterion, a lower cut off score of 2 for at least one symptom will be used in this study. By using this threshold, we aim to select a population which reﬂects the entire spectrum of MUPS, covering patients with mild symptoms, as well as patients with severe symptoms. Exclusion criteria − A sufﬁcient medical explanation for the symptoms, according to the physician. − Incomplete diagnostic evaluation of the symptoms, according to the physician. − Insufﬁcient command of the Dutch language. − A cognitive or visual impairment that prohibits participating in a questionnaire survey. − Severe psychopathology (e.g. psychotic disorder, bipolar disorder). − Pregnancy. − Cancer diagnosed in 5 years prior to inclusion. − Another life threatening condition or a short life expectancy. Inclusion procedure A ﬂowchart of the inclusion methods, both in primary and in secondary/tertiary care, is given in Fig. 1.
In primary care, patients at risk for MUPS will be searched, using an electronic database search, based on a list of 23 unexplained physical complaints composed by Robbins et al. . The symptoms on this list are associated with functional somatic syndromes such as Chronic Fatigue Syndrome or Irritable Bowel Syndrome. Symptoms are likely to be unexplained when they are on the ‘Robbins list’ and a matching diagnosis ICPC code (meaning an ICPC code N70) is lacking. Patients who visited their general practitioner (GP) twice or more in the last 3months with one or more of these symptoms (without a matching ICPC diagnosis code N70 in the patient's electronic ﬁle) will be selected. We use 2 visits as a cut off, as the presentations of symptoms which are limited to one doctor–patient contact are not the subject of this study. However, as MUPS patients can suffer from multiple symptoms, symptoms presented in these 2 visits are allowed to be different. Selected patients are checked for exclusion criteria by their own GP. The ﬁrst 2 exclusion criteria (a sufﬁcient medical explanation for the symptoms and incompleteness of diagnostic evaluation) are based on the criteria for MUPS as described in the Background section. Patients without exclusion criteria receive the PHQ-15 questionnaire by mail. Patients who return the questionnaire and have a score of 2 for at least one symptom are eligible and will be approached for informed consent and inclusion. In secondary and tertiary care, all newly referred patients with MUPS as the reason for referral are screened for in- and exclusion criteria by the physician performing the intake consultation. The same criteria as in primary care are used. Selected patients receive the PHQ-15 by mail. Patients who return the questionnaire and have score of 2 for at least one symptom are eligible and will be approached for informed consent.
Fig. 1. Flowchart inclusion process PROSPECTS study.
N. van Dessel et al. / Journal of Psychosomatic Research 76 (2014) 200–206
Outcomes Patient characteristics We will mail questionnaires at baseline to assess general characteristics (i.e. gender, age, length, weight, country of origin, education level, occupation) and medical characteristics (medical history, chronic medical conditions and life style parameters). Additionally, a more extensive overview of number and severity of physical symptoms will be created using the PSQ questionnaire, an extensive Dutch questionnaire on 51 physical symptoms . Primary and secondary outcomes All questionnaires used to assess the outcomes were selected based on their validity, previous use in this speciﬁc population, availability of reference values and questionnaire length. Severity of symptoms and functional impairment at follow-up are the primary outcomes, measured respectively with the Patient Health Questionnaire-15 [27,28] and the RAND 36-item health survey questionnaire . Secondary outcomes are health care use (Trimbos/ iMTA questionnaire for Costs associated with Psychiatric Illness ) and level of depressive symptoms and anxiety (Quick Inventory of Depressive Symptomatology  and Beck Anxiety Inventory [34,35] questionnaires). These outcomes will be measured at baseline and at all follow-up moments. Potential predictors The choice of potential predictors of the course of MUPS is based on a number of theories incorporated in the cognitive behavioral model. Therefore, the theories mentioned in the Introduction and most other theories in the meta-model will be reﬂected. Based on the somatosensory ampliﬁcation theory the concepts of hypervigilance (SomatoSensory Ampliﬁcation Scale [36,37]), illness cognitions, causal attributions and coping (Cognitive and Behavioural Responses to symptoms Questionnaire ) and illness perception (Illness Perception Questionnaire [39,40]) were selected.
The sensitivity theory will be covered by the incorporation of questionnaires about personality (NEO Personality questionnaire — Five Factor Inventory [41,42]), perfectionism (Multi-dimensional Perfectionism Scale [43,44]), psychiatric co-morbidity (medical chart, Quick Inventory of Depressive Symptomatology , Beck Anxiety Inventory [34,35] and Whitely Index for hypochondria [45,37]), positive affect (subscale of Positive And Negative Affect Schedule ), life events (Life Events Questionnaire ), social support (Social Support scale ) and physical activity (International Physical Activity Questionnaire [49,50]). An overview of all questionnaires and moments of administration is given in Table 1. Follow-up measurements will take place after 6, 12, 24 and 36 months. At baseline and follow-up all relevant questionnaires will be presented to participants as one instrument (one booklet), for ease of use. The baseline measurement consists of questionnaires regarding a comprehensive set of possible predictors. Only the questionnaires measuring perfectionism and personality will be postponed until the ﬁrst follow-up measurement (T1), in order to reduce participant burden at baseline. In adults these factors are considered to be relatively stable over time . Potential predictors — physiological assessments The theory of endocrine dysregulation (see Introduction) will be tested by measurement of free salivary cortisol levels. Salivary cortisol levels correspond well with cortisol levels in plasma . As solitary cortisol measurements have low intra-individual stability, in this study cortisol levels in response to a stressor will be investigated. As a measure of a natural stress response of the HPA axis, the morning Cortisol Awakening Response (CAR) will be assessed at baseline and after 12 months of follow-up. All patients will collect saliva samples at awakening time (T0), and 30 (T1) and 60 min (T2) afterwards, as it is known that cortisol levels rise during the ﬁrst 30 min after awakening and remain elevated for at least 60 min . They collect saliva at home using Salivettes® (Sarstedt, Etten-Leur, The Netherlands), according to the guideline of the manufacturer. Samples will be stored in home refrigerators and returned by mail as quick as possible. Returned swabs will be centrifuged and analyzed . The area under the curve with
Table 1 Used questionnaires and time points for administration Instrument
Outcome measures: Number/severity of symptoms Number of symptoms Functional impairment Health care use Depression Anxiety
PHQ-15 [27,28] PSQ  RAND36  Subscale TIC-P  QIDS-SR  BAI [34,35]
x x x x x x
Potential predictors: Perfectionism Personality (neuroticism, extraversion) Cognitions and coping Positive affect Hypervigilance Illness perception Causal attributions Depression Anxiety Hypochondria Life events Social support Physical activity
MDPS [43,44] Subscales NEO-FFI [41,42] CBRQ  Subscale PANAS  SSAS [36,37] IPQ-K [39,40] IPQ-K and CBRQ QIDS-SR  BAI [34,35] WI [45,37] LEQ  SoS  IPAQ [49,50]
General characteristics: Patient characteristics (demography/life style) History and chronic medical conditions Received diagnostics/treatments for MUPS
x x x x x x x x x x x
6 mo (T1)
12 mo (T2)
24 mo (T3)
36 mo (T4)
x x x x
x x x x
x x x x x
x x x x x
x x x x x x
N. van Dessel et al. / Journal of Psychosomatic Research 76 (2014) 200–206
respect to the ground (AUCg) and the area under the curve with respect to the increase (AUCi) will be calculated using the formulas described by Pruessner et al. in 2003 . The AUCg is an estimate of the total cortisol secretion over the ﬁrst hour after awakening, whereas the AUCi is a measure of its time-dependent change.
We will use LCGMM, as it allows us to create subgroups based on registered characteristics (and change of characteristics). As a result, identiﬁed subgroups will form a closer approximation of the complex reality, compared to subgroups based on predeﬁned (subjective) questionnaire cut offs.
Treatments In this study the natural course of MUPS is investigated, therefore no speciﬁc treatment is offered to participants. However, during follow-up patients might be treated by their GP or specialist. As treatment may inﬂuence the course of MUPS, we will ask about received treatments at follow-up. Speciﬁc questions will be asked about treatments performed by for example the GP, physiotherapists, psychologists, secondary care somatic specialists or alternative caregivers. Received treatments will be considered as a covariate in the statistical analysis.
Multivariate prediction model As a second step, logistic regression will be used to identify the combinations of risk factors which are associated with the identiﬁed course types. Depending on the number of subgroups resulting from the latent class analysis, binary logistic regression or ordinal regression will be used. In this analysis all baseline characteristics will be taken into account as potential predictors. Results will be presented as odds ratios with 95% conﬁdence intervals. The aim is to create one or more multifactor prediction models, consisting of approximately 6 determinants predicting the identiﬁed course types of MUPS. Predictors with a p ≤ 0.157 will remain in the ﬁnal prediction model(s) . Reliability of the prediction model will be tested through a calibration plot, in which the observed frequency of all course types is plotted against the predicted chance of this course. The discriminating value will be investigated by calculating the area under the ROC-curve (AUC). Internal validity of the model will be investigated through a bootstrapping procedure. Data analysis will be performed with software packages of SPSS and Mplus. Model assumptions will be checked prior to the analysis.
Power calculation Our sample size calculation is based on the prediction model for the course of MUPS that we aim to develop. A rule of thumb states that the number of ‘events’ should be ≥10 for every variable in a multivariate prediction model . We plan to develop a model with approximately 6 variables. Hence at least 6 × 10 = 60 events will be needed. Possible ‘events’ according to the course of MUPS are recovery, chronicity and recurrence (i.e. after a period of absence of symptoms). Of these, chronicity is the most restrictive. According to a Dutch review deterioration of physical symptoms (and thus chronicity) occurs in 10–30% of MUPS patients . As this percentage does not include chronic MUPS patients with a stable level of physical symptoms, we estimate the incidence of chronicity to be at least 30%. Therefore we need 200 evaluable patients to expect 60 events (chronicity) in both primary and secondary care study groups. Taking into account an expected loss to follow up of 10%, we aim at a study sample of 450 patients, divided over the 2 study groups (225 patients per group). For the other study aims, this number of participants is more than sufﬁcient.
Bias handling Selection bias Hypothetically, the prevalence of severe MUPS may be higher among non-responders, as patients with severe complaints may not have the energy to participate. Other groups that may show a lower response (for the same reason) are the elderly and patients with psychiatric co-morbidity. To minimize this effect, we will stimulate all eligible patients to participate, also when they have more severe symptoms. Additionally, we will ask non-responders a few questions about the reason for not participating. A non-response analysis will be performed.
Statistical analysis Descriptive statistics (e.g. frequencies and percentages or mean ± SD) will be used to summarize the demographic characteristics and primary outcomes of the study population in all settings and at all follow-up moments. Cross-sectional relations between severity of MUPS and functional impairment or health care use at baseline will be analyzed, corrected for age, gender and co-morbidity. Latent Class Growth Mixture Modeling One of the aims of our study is to identify distinct course types of severity of symptoms, functional outcome and health care use. These outcome measures are registered at baseline and 4 times during a 3 year follow-up, leading to 4 estimates of changes in outcomes. As a result, longitudinal patterns (trajectories) of these outcomes can be evaluated over time. It is expected that these trajectories vary across participants. Therefore, Latent Class Growth Mixture Modeling (LCGMM) will be used to form a smaller amount of distinct clusters (latent trajectory classes), based on the outcome measures . We will use LCGMM instead of Latent Class Growth Analysis (another modeling method to statistically derive distinct subgroups) as it allows a certain level of variation intercept and slope in one or more classes, leading to larger within class heterogeneity. The optimal number of clusters will be determined using statistical parameters, including Bayesian Information Criterion  and the bootstrapped likelihood ratio test . Additionally, clinical interpretation of the clusters will guide the ﬁnal classiﬁcation, in order to avoid clinically uninterpretable clusters. Patients will be assigned to speciﬁc clusters based on the posterior probabilities to ﬁt in the cluster, using Bayesian statistics .
Attrition bias It is conceivable that patients with severe complaints or comorbidities may prematurely end study participation, as participation costs energy, but has no apparent personal advantages. To stimulate continuation of participation, small incentives (e.g. gift coupons) will be provided during follow-up. In case of discontinuation, patients will be stimulated to resume participation. All reasons for discontinuation will be collected, analyzed and reported. In addition, baseline characteristics, including severity of symptoms, will be compared between participants and drop-outs. In case of missing follow-up data imputation techniques will be considered. Discussion Health care workers in primary care and secondary care are confronted with patients with MUPS on a daily basis. MUPS patients, especially those with persistent MUPS face various insecurities, as their symptoms are not (fully) medically explained. Current knowledge about prognosis is limited and treatment options are scarce. This study will add to earlier studies, as it will have a long follow-up period. Additionally, we will register possible treatments, have an additional focus on functional impairment and identify predictors of the course of MUPS. When designing the study, the greatest challenge concerned the inclusion procedure, especially in primary care. In various previous studies, selection of MUPS patients was based on searches in GPs' electronic databases. Unfortunately, a speciﬁc ICPC code for MUPS is lacking. Therefore, criteria such as the Robbins criteria have often been used as a
N. van Dessel et al. / Journal of Psychosomatic Research 76 (2014) 200–206
proxy to identify possible MUPS cases [3,61]. As Robbins' list symptoms are not always unexplained, a disadvantage of this method can be the selection of patients with symptoms that do not ﬁt the criteria for MUPS (false positives). Furthermore, as selection results depend on proper registration by GPs, possible MUPS cases might be missed in case of suboptimal registration of consultations (false negatives). In other studies questionnaires such as the PHQ-15 have been used as a screening instrument for possible MUPS patients . The PHQ-15 is a validated and widely used instrument for this aim, evaluating 15 symptoms that account for more than 90% of all reported physical symptoms. The PHQ-15 reﬂects the symptoms reported by the patient. However, the scale only includes a somatic symptom count. Therefore, an additional evaluation of the nature of the symptoms by a physician is needed to conﬁrm the existence of MUPS. In this study, we will combine both methods. A digital search strategy based on the Robbins list will be used as a ﬁrst step. Afterwards, the selected patients will be screened for in- and exclusion criteria by the GP. As a third step, selected patients will be asked to ﬁll in the PHQ-15 questionnaire, to objectify the current existence of symptoms. This combination will lead to a ﬁnal selection of ‘true’ MUPS patients. This method has the disadvantage that MUPS patients may be missed, for example due to incorrect registration by GPs or a coincidentally low number of doctor visits in the last 3 months. Although we do believe that in this study it is not essential to select all MUPS patients in our population (as we do not study the prevalence of MUPS), we will try to overcome this disadvantage as much as possible, as a representative sample is essential. We will do this by selecting GPs with ICPC coding experience, using relatively low-threshold selection criteria and by the inclusion of patients in different health care settings. In secondary care, we will recruit patients in specialized MUPS care settings. Therefore, it is a limitation to this study that the results in this study group are not generalizable to the complete population as seen by somatic specialists in secondary care. Despite the fact that we do not include patients in these settings, we do believe that the described group will not be completely missed. At some point patients with MUPS who are seen by somatic specialists will be referred back to their GP (as no diagnosis can be established) or to specialized MUPS care. As a result, an unknown percentage of these patients will be included via the included study groups. The aim of this study is restricted to patient related factors inﬂuencing the course of MUPS. As a result, other potential inﬂuencing factors, such as the doctor–patient relationship or communication will not be taken into account in this study. We believe that study results will provide extensive information about course and prognosis. This information can be used directly in clinical practice, for example in patient education. Additionally, identiﬁcation of predictors of the course of MUPS may lead to better recognition of patients at risk for persistent symptoms, who may beneﬁt from treatment. It may also lead to identiﬁcation of relevant treatment targets. Taking together all these practical implications, we do believe that this study has great relevance for clinical practice. We aim to complete the study in 2016. List of abbreviations AUC area under the (ROC-)curve BAI Beck Anxiety Inventory CAR Cortisol Awakening Response CBT cognitive behavioral therapy HPA axis hypothalamic pituitary adrenal axis ICD International Classiﬁcation of Diseases ICPC International Classiﬁcation of Primary care Code GP(s) general practitioner(s) LCGMM Latent Class Growth Mixture Modeling MUPS medically unexplained physical symptoms PHQ-15 Patient Health Questionnaire-15
PRIME-MD Primary care Evaluation of Mental Disorders PROSPECTS PROSpective cohort study on prognosis and PErpetuating faCTors of medically unexplained physical symptoms (MUPS) QIDS-SR Quick Inventory of Depressive Symptomatology — Self Report RAND 36 RAND 36-item health survey ROC curve Receiver Operating Characteristic curve SPSS Statistical Package for the Social Science TIC-P Trimbos/iMTA questionnaire for Costs associated with Psychiatric Illness VUmc Vrije Universiteit medisch centrum (VU University Medical Centre)
Competing interest statement The authors declare that they have no competing interests to report. Author contributions ND and SL developed the basis of the study protocol. ND drafted the manuscript and will coordinate data collection, statistical analysis and reporting of study results. HH, JD and JW participated in the design, helped to draft the manuscript and will support in data collection, statistical analysis and reporting of the results. All authors have read and approved the manuscript. Acknowledgments This project is funded by a grant of the SBOH, the Dutch employer of GP trainees, and the Stoffels-Hornstra Foundation, a non-proﬁt organization supporting research in primary care. References  Barsky A, Borus J. Somatization and medicalization in the era of managed care. JAMA 1995;274:1931–4.  Van der Linden M, Westert G, de Bakker D, Schellevis F. Tweede Nationale Studie Naar Ziekten En Verrichtingen in de Huisartspraktijk. Second national study of diseases and proceedings in general practice. Utrecht/Bilthoven, the Netherlands: NIVEL/RIVM; 2004.  Verhaak PFM, Meijer SA, Visser AP, Wolters G. Persistent presentation of medically unexplained symptoms in general practice. Fam Pract 2006;23:414–20.  Jackson JL, Passamonti M. The outcomes among patients presenting in primary care with a physical symptom at 5 years. J Gen Intern Med 2005;20:1032–7.  Landelijke Stuurgroep Multidisciplinaire Richtlijnontwikkeling in de GGZ. Multidisciplinaire Richtlijn Somatisch Onvoldoende Verklaarde Lichamelijke Klachten En Somatoforme Stoornissen (Multidisciplinary Guideline of MUPS and Somatoform Disorders). Utrecht, the Netherlands: Trimbos Instituut; 2010.  Carson AJ, Best S, Postma K, Stone J, Warlow C, Sharpe M. The outcome of neurology outpatients with medically unexplained symptoms: a prospective cohort study. J Neurol Neurosurg Psychiatry 2003;74:897–900.  Zoccolillo M, Cloninger CR. Somatization disorder: psychologic symptoms, social disability, and diagnosis. Compr Psychiatry 1986;27:65–73.  Escobar J, Burnam M. Somatization in the community. Arch Gen Psychiatry 1987;44:713–8.  Gureje O, Simon G, Ustun T, Goldberg D. Somatization in cross-cultural perspective: a World Health Organisation study in primary care. Am J Psychiatry 1997;154:989–95.  Konnopka A, Schaefert R, Heinrich S, Kaufmann C, Luppa M, Herzog W, et al. Economics of medically unexplained symptoms: a systematic review of the literature. Psychother Psychosom 2012;81:265–75.  Katon W, Walker E. Medically unexplained symptoms in primary care. J Clin Psychiatry 1998;59:15–21.  Smith G, Monson R, Ray D. Patients with multiple unexplained symptoms: their characteristics, functional health, and health care utilization. Arch Intern Med 1986;146:69–72.  Kroenke K, Mangelsdorff A. Common symptoms in ambulatory care: incidence, evaluation, therapy, and outcome. Am J Med 1989;86:262–6.  Barsky AJ, Orav EJ, Bates DW. Somatization increases medical utilization and costs independent of psychiatric and medical comorbidity. Arch Gen Psychiatry 2005;62:903–10.  Bermingham S, Cohen A, Hague J, Parsonage M. The cost of somatisation among the working‐age population in England for the year 2008–2009. Ment Health Fam Med 2010;7:71–84.
N. van Dessel et al. / Journal of Psychosomatic Research 76 (2014) 200–206
 Brenninkmeijer V, Lagerveld SE, Blonk RWB. Moeilijk objectiveerbare klachten in de praktijk van de bedrijfs- en verzekeringsarts (Symptoms which are difﬁcult to objectify in clinical practice of occupational and insurance physicians). Tijdschrift voor Bedrijfs- en Verzekeringsgeneeskunde 2006;14:424–30.  Van Ravenzwaaij J, Olde Hartman T, van Ravesteijn H, Eveleigh R, van Rijswijk E, Lucassen P. Explanatory models of medically unexplained symptoms: a qualitative analysis of the literature. Ment Health Fam Med 2010;7:223–31.  Deary V, Chalder T, Sharpe M. The cognitive behavioural model of medically unexplained symptoms: a theoretical and empirical review. Clin Psychol Rev 2007;27:781–97.  Barsky A, Wyshak G. Hypochondriasis and somatosensory ampliﬁcation. Br J Psychiatry 1990;157:404–9.  Fries E, Hesse J. A new view on hypocortisolism. Psychoneuroendocrinology 2005;30:1010–6.  Olde Hartman TC, Borghuis MS, Lucassen PLBJ, van de Laar FA, Speckens AE, van Weel C. Medically unexplained symptoms, somatisation disorder and hypochondriasis: course and prognosis. A systematic review. J Psychosom Res 2009;66:363–77.  Kroenke K. Efﬁcacy of treatment for somatoform disorders: a review of randomized controlled trials. Psychosom Med 2007;69:881–8.  Kleinstäuber M, Witthöft M, Hiller W. Efﬁcacy of short-term psychotherapy for multiple medically unexplained physical symptoms: a meta-analysis. Clin Psychol Rev 2011;31:146–60.  Sumathipala A. What is the evidence for the efﬁcacy of treatments for somatoform disorders? A critical review of previous intervention studies. Psychosom Med 2007;69:889–900.  Lamberts H, Wood M. ICPC: International Classiﬁcation of Primary Care. Oxford: Oxford University Press; 1987.  www.who.int/en/.  Kroenke K, Spitzer RL, Williams JBW. The PHQ-15: validity of a new measure for evaluating the severity of somatic symptoms. Psychosom Med 2002;64:258–66.  Van Ravesteijn H, Wittkampf K, Lucassen P, van de Lisdonk E, van den Hoogen H, van Weert H, et al. Detecting somatoform disorders in primary care with the PHQ-15. Ann Fam Med 2009;7:232–8.  Spitzer R, Williams J. Utility of a new procedure for diagnosing mental disorders in primary care. JAMA 1994;272:1749–56.  Robbins J, Kirmayer L, Hemami S. Latent variable models of functional somatic distress. J Nerv Ment Dis 1997;185:606–15.  Van der Zee KI, Sanderman R. Het meten van de algemene gezondheidstoestand met de RAND-36, een handleiding (Measurement of general health with the RAND-36, a manual). Groningen, the Netherlands: Noordelijk Centrum voor Gezondheidsvraagstukken; 1993.  Hakkaart-van Roijen L. Trimbos/iMTA questionnaire for Costs associated with Psychiatric Illness (TiC-P). , 02Rotterdam, the Netherlands: Institute for Medical Technology Assesment; 2002 61.  Rush A, Trivedi M, Ibrahim H. 16-Item Quick Inventory of Depressive Symptomatology (QIDS), clinician rating (QIDS-C), and self-report (QIDS-SR): a psychometric evaluation in patients with chronic. Biol Psychiatry 2003;54:573–83.  Beck AT, Epstein N, Brown G, Steer RA. An inventory for measuring clinical anxiety: psychometric properties. J Consult Clin Psychol 1988;56:893–7.  Muntingh ADT, van der Feltz-Cornelis CM, van Marwijk HWJ, Spinhoven P, Penninx BWJH, van Balkom AJLM. Is the Beck Anxiety Inventory a good tool to assess the severity of anxiety? A primary care study in the Netherlands Study of Depression and Anxiety (NESDA). BMC Fam Pract 2011;12:66.  Barsky a J, Wyshak G. Hypochondriasis and somatosensory ampliﬁcation. Br J Psychiatry 1990;157:404–9.  Speckens a E, Spinhoven P, Sloekers PP, Bolk JH, van Hemert a M. A validation study of the Whitely Index, the Illness Attitude Scales, and the Somatosensory Ampliﬁcation Scale in general medical and general practice patients. J Psychosom Res 1996;40:95–104.  Skerrett TN, Moss-Morris R. Fatigue and social impairment in multiple sclerosis: the role of patients' cognitive and behavioral responses to their symptoms. J Psychosom Res 2006;61:587–93.  Broadbent E, Petrie KJ, Main J, Weinman J. The brief illness perception questionnaire. J Psychosom Res 2006 Jun:631–7.
 De Raaij EJ, Schröder C, Maissan FJ, Pool JJ, Wittink H. Cross-cultural adaptation and measurement properties of the Brief Illness Perception Questionnaire — Dutch Language Version. Man Ther 2012;17:330–5.  Costa PT. McCrae RR: Revised NEO Personality Inventory (NEO-PI-R) and the Five Factor Inventory (NEO-FFI): Professional Manual. Odessa, Florida USA: Psychological Assessment Resources Inc.; 1992.  Hoekstra H, Ormel J, de Fruyt F. NEO-PI-R en NEO-FFI: Big Five Persoonlijkheidsvragenlijsten: Handleiding (NEO-PI-R and NEO-FFI: Big Five personality questionnaires: a manual). Amsterdam, the Netherlands: Hogrefe Uitgevers; 1996.  Frost RO, Marten P, Lahart C, Rosenblate R. The dimensions of perfectionism. Cogn Ther Res 1990;14:449–68.  Flos CHGM, Peeters FPML, van Eiisden M. The “Multidimensional Perfectionism Scale”; reliability and validity of the Dutch translation. Maastricht, the Netherlands: Maastricht University; 2000.  Pilowsky I. Dimensions of hypochondriasis. Br J Psychiatry 1967;113:89–93.  Watson D, Clark L, Carey G. Positive and negative affectivity and their relation to anxiety and depressive disorders. J Abnorm Psychol 1988;97:346–53.  Garnefski N, Kraaij V. Levensgebeurtenissen Vragenlijst (Life Event Questionnaire). Leiden, the Netherlands: Leiden University; 2001.  Feij J, Doorn C, van Kampen D, van den Berg P. Sensation seeking and social support as moderators of the relationship between life events and physical illness/psychological distress. In: Winnubst J, Maes S, editors. Lifestyles stress and health. Leiden: DSWO press; 1992. p. 285–302.  Ainsworth BE, Bassett DR, Strath SJ, Swartz A, O'Brien W, Thompson R, et al. Comparison of three methods for measuring the time spent in physical activity. Med Sci Sports Exerc 2000;32:S457–64.  Vandelanotte C, De Bourdeaudhuij I, Philippaerts R, Sjöström M, Sallis J. Reliability and validity of a computerized and Dutch version of the International Physical Activity Questionnaire (IPAQ). J Phys Act Health U S Hum Kinet 2005;2:63–75.  McCrae RR, Costa PT, Ostendorf F, Angleitner A, Hrebícková M, Avia MD, et al. Nature over nurture: temperament, personality, and life span development. J Pers Soc Psychol 2000;78:173–86.  Kirschbaum C, Hellhammer DH. Salivary cortisol in psychoneuroendocrine research: recent developments and applications. Psychoneuroendocrinology 1994;19:313–33.  Fekedulegn DB, Andrew ME, Burchﬁel CM, Violanti JM, Hartley TA, Charles LE, Miller DB. Area under the curve and other summary indicators of repeated waking cortisol measurements. Psychosom Med 2007;69:651–9.  Heijboer AC, Martens F, Blankenstein MA. Measuring salivary cortisol with the Architect i2000 random access analyser. Ann Clin Biochem 2009;46:261–5.  Pruessner JC, Kirschbaum C, Meinlschmid G, Hellhammer DH. Two formulas for computation of the area under the curve represent measures of total hormone concentration versus time-dependent change. Psychoneuroendocrinology 2003;28:916–31.  Altman D. Practical statistics for medical research. London, Great Britain: Chapman & Hall; 1991.  Jung T, Wickrama K. An introduction to latent class growth analysis and growth mixture modeling. Soc Pers Psychol Compass 2008;2:302–17.  Schwarz G. Estimating the dimension of a model. Ann Stat 1978;6:461–4.  McLachlan G, Peel D. Finite mixture models. New York, USA: John Wiley & Sons; 2000.  Quinn G, Keough M. Bayesian hypothesis testing. Experimental design and data analysis for biologists. Cambridge, Great-Brittain: Cambridge University Press; 2002. p. 54–7.  Smits F, Brouwer H, van Weert H, Schene A, ter Riet G. Predictability of persistent frequent attendance: a historic 3-year cohort study. Br J Gen Pract 2009, Feb:e44–50.  Steinbrecher N, Hiller W. Course and prediction of somatoform disorder and medically unexplained symptoms in primary care. Gen Hosp Psychiatry 2011;33:318–26.  Van Hemert N. Lichamelijke Klachten Vragenlijst (Physical Symptoms Questionnaire). Leiden, the Netherlands: Leiden University; 2003.  Watson D, Clark L, Carey G. Positive and negative affectivity and their relation to anxiety and depressive disorders. J Abnorm Psychol August 1988;97:346–53.