[Antidepressants are frequently prescribed but still critized; a perspective on causes and solutions]. / Antidepressiva: al decennia voor­geschreven én toch nog steeds bekritiseerd; een perspectief op oorzaken en oplossingen.

BACKGROUND: From around 1980, antidepressants (ad) have increasingly been prescribed, for longer periods of time, especially selective serotonin reuptake inhibitors (ssris). Paradoxically, their effectiveness is still doubted, especially outside the psychiatric profession.
AIM: To explain increase and offer a perspective on causes and solutions, and to indicate how to reach consensus.
METHOD: Position paper with critical analysis and synthesis of relevant literature.
RESULTS: The rise in AD prescriptions results from: 1. increased safety and ease of prescribing, 2. increased presentation and recognition of depression in primary care, 3. extension of indication criteria, 4. effective marketing strategies, and 5. effectiveness in acute phase (aad) and of relapse/recurrence prevention in continuation/maintenance phases (coad).Critics point to: 1. low added value of aad relative to placebo, 2. many drop-outs and non-responders, 3. relapse/recurrence prevention with coad works only for responders to aad, 4. relapse/recurrence after AD discontinuation often involves withdrawal symptoms, and 5. publication bias, selective reporting, selective patient selection, and suboptimal blinding, resulting in overestimated effectiveness and underestimated disadvantages.Factors that keep fueling the controversy are: 1. critics stress the net effectiveness of AD whereas proponents point at gross effectiveness which includes spontaneous recovery and placebo effect; 2. persistence of distrust in industry-funded rcts; 3. ideological positions, reinforced by conflicts of interest and selective citations; 4. lack of rcts with relevant long-term outcome measurements.
CONCLUSION: Although consensus is difficult to achieve given the ideological component, there are options. Three factors are critically important: confer to establish which data convince the opposition, response prediction (what works for whom), and rcts with long-term functional outcomes.

The introduction of the practice nurse mental health in general practices in the Netherlands: effects on number of diagnoses of chronic and acute alcohol abuse.

BACKGROUND: Since 2008 mental health practice nurses have been gradually introduced in general practices in the Netherlands as part of health policy aiming to improve early identification and treatment of mental health problems in primary care. This study aims to investigate the effect of the introduction of the practice nurse mental health in general practices in the Netherlands on the number of diagnoses of chronic and acute alcohol abuse. METHODS: The Netherlands Institute for Health Services Research (NIVEL) retrieved data of a representative sample of general practices (n = 155) for this study. Data were aligned at the starting point of the implementation of the PN-MH to compare the practices on our outcome measures after implementation of the PN-MH. Multilevel regression analyses were conducted to investigate differences in average number of chronic and acute alcohol abuse diagnoses between practices with a practice nurse mental health and control practices (without a practice nurse mental health and without a primary care psychologists). RESULTS: A significant decrease over time of chronic alcohol abuse diagnoses was observed (ß = -.52, p < 0.05) as well as a significant decrease over time of acute alcohol abuse diagnoses (ß = -.06, p < 0.05). After adjustment for multiple comparisons, no significant differences were found between practices that implemented a practice nurse mental health or only have a primary care psychologist and control practices. Practices that implemented a practice nurse mental health and have a primary care psychologist, had a higher mean of chronic and acute alcohol abuse diagnoses than control practices during all periods, but the differences between these groups were not statistically significant. CONCLUSIONS: Based on the results of this study it seems that the introduction of practice nurses mental health in general practices is not associated with increased diagnoses of chronic or acute alcohol abuse. Potential explanations are barriers experienced by practice nurses to addressing alcohol use with patients and prioritization of other mental health issues over alcohol abuse. In order to improve the management of alcohol abuse by practice nurses, more research is needed on how practice nurses can be involved in diagnosing and treatment of patients with alcohol abuse.

[Developments in mental health care in Dutch general practices: an overview of recent studies]. / Ontwikkelingen in de Nederlandse huisartsenzorg voor psychische problemen: een overzicht van recente studies.

BACKGROUND: Dutch policy aims to strengthen mental health care in general practices, to keep health care affordable. Recently, a new function (mental health nurses) and a new referral model for patients with mental health problems were introduced.
AIM: To explore to what extent the volume of mental health care in Dutch general practices has increased and to what extent the content changed in the period 2010-2015.
METHOD: This study employed: 1. analyses of medical records, and 2. a case study in a primary health care centre.
RESULTS: The number of general practices with at least one mental health nurse increased from 20% in 2010 to almost 90% in 2015. In the period 2010-2014, general practitioners (gps) and mental health nurses treated increasing numbers of patients with mental health problems. No task shifting from gps to mental health nurses was observed. In the period 2011-2015, the number of antidepressant prescriptions increased slightly. In 2014, gps in a well-prepared primary care centre allocated 87% of their patients with mental health problems to a treatment setting in line with the referral model.
CONCLUSION: Dutch general practices have recently provided more mental health care, thereby emphasising their important role in the mental health care system.

Prevalence, presentation and prognosis of delirium in older people in the population, at home and in long term care: a review.

OBJECTIVE: The aim of this study is to provide an overview of prevalence, symptoms, risk factors and prognosis of delirium in primary care and institutionalized long-term care. DESIGN: The method used in this study is a systematic PubMed search and literature review. RESULTS: The prevalence of delirium in the population among the elderly aged 65+ years is 1-2%. Prevalence rises with age: 10% among a "general" population aged 85+ years. Prevalence rises up to 22% in populations with higher percentages of demented elder. In long-term care, prevalence ranges between 1.4% and 70%, depending on diagnostic criteria and on the prevalence of dementia. There is a significant increase of the risk of delirium with age and cognitive decline in all groups. Concerning prognosis, most studies agree that older people who previously experienced delirium have a higher risk of dementia and a higher mortality rate. Population and long-term care studies show the same tendency. CONCLUSIONS: Delirium in a non-selected population aged 65+ years is uncommon. However, prevalence rises very quickly in selected older groups. Primary care doctors should be aware of a relatively high risk of delirium among the elderly in long-term care, those older than 85 years and those with dementia.

Symptoms in long-term breast cancer survivors: A cross-sectional study in primary care.

PURPOSE: Various long-term symptoms can manifest after breast cancer treatment, but we wanted to clarify whether these are more frequent among long-term breast cancer survivors than matched controls and if they are associated with certain diagnoses. METHODS: This was a cross-sectional, population-based study of 350 breast cancer survivors treated with chemo- and/or radiotherapy ≥5 years (median 10) after diagnosis and 350 women without cancer matched by age and primary care physician. All women completed a questionnaire enquiring about symptoms, underwent echocardiography to assess the left ventricle ejection fraction, and completed the Hospital Anxiety and Depression Scale. Cardiovascular diseases were diagnosed from primary care records. In a multivariable logistic regression analysis, symptoms were adjusted for the long-term effects and compared between cohorts and within the survivor group. RESULTS: Concentration difficulties, forgetfulness, dizziness, and nocturia were more frequent among breast cancer survivors compared with controls, but differences could not be explained by cardiac dysfunction, cardiovascular diseases, depression, or anxiety. Intermittent claudication and appetite loss were more frequent among breast cancer survivors than controls and associated with cardiac dysfunction, depression, and anxiety. Breast cancer survivors treated with chemotherapy with/without radiotherapy were at significantly higher odds of forgetfulness and nocturia, but significantly lower odds of dizziness, compared with breast cancer survivors treated with radiotherapy alone. CONCLUSIONS: Intermittent claudication and appetite loss are common among breast cancer survivors and are associated with cardiac dysfunction and mood disorders. Other symptoms varied by whether the patient underwent chemotherapy with/without radiotherapy (forgetfulness and nocturia) radiotherapy alone (dizziness).

Long-term psychological distress in breast cancer survivors and their matched controls: A cross-sectional study.

INTRODUCTION: Breast cancer survivors often experience psychological distress shortly after diagnosis. Long-term psychological effects, however, have not been clearly demonstrated. METHODS: This cross-sectional cohort study included 350 breast cancer survivors and 350 age-matched and general-practitioner-matched women. The median follow-up was 10 years. Using logistic regression we compared breast cancer survivors with controls on having (severe) symptoms of depression and/or anxiety, as measured with the Hospital Anxiety and Depression Scale. In multivariable logistic regression, we adjusted the results for a history of depression or prescription of antidepressants. RESULTS: Larger proportions of breast cancer survivors experienced symptoms of depression (10.6%) compared with controls (4.9%) and symptoms of anxiety (18.6%) compared with controls (16.3%). The odds of symptoms of depression (OR 2.3, 95%CI 1.3-4.2), severe symptoms of depression (OR 3.3, 95%CI 1.1-10.3) and severe symptoms of anxiety (OR 2.1, 95%CI, 1.1-4.0) were significantly higher for breast cancer survivors than for controls, even after adjusting for history of depression or prescription of antidepressants. No significant difference was seen for mild symptoms of anxiety. CONCLUSIONS: Breast cancer survivors have an increased risk of symptoms of depression, including severe symptoms, and severe symptoms of anxiety compared with controls, for up to at least 10 years after diagnosis.

Development and external validation of a prediction rule for an unfavorable course of late-life depression: A multicenter cohort study.

BACKGROUND: Given the poor prognosis of late-life depression, it is crucial to identify those at risk. Our objective was to construct and validate a prediction rule for an unfavourable course of late-life depression. METHODS: For development and internal validation of the model, we used The Netherlands Study of Depression in Older Persons (NESDO) data. We included participants with a major depressive disorder (MDD) at baseline (n = 270; 60-90 years), assessed with the Composite International Diagnostic Interview (CIDI). For external validation of the model, we used The Netherlands Study of Depression and Anxiety (NESDA) data (n = 197; 50-66 years). The outcome was MDD after 2 years of follow-up, assessed with the CIDI. Candidate predictors concerned sociodemographics, psychopathology, physical symptoms, medication, psychological determinants, and healthcare setting. Model performance was assessed by calculating calibration and discrimination. RESULTS: 111 subjects (41.1%) had MDD after 2 years of follow-up. Independent predictors of MDD after 2 years were (older) age, (early) onset of depression, severity of depression, anxiety symptoms, comorbid anxiety disorder, fatigue, and loneliness. The final model showed good calibration and reasonable discrimination (AUC of 0.75; 0.70 after external validation). The strongest individual predictor was severity of depression (AUC of 0.69; 0.68 after external validation). LIMITATIONS: The model was developed and validated in The Netherlands, which could affect the cross-country generalizability. CONCLUSIONS: Based on rather simple clinical indicators, it is possible to predict the 2-year course of MDD. The prediction rule can be used for monitoring MDD patients and identifying those at risk of an unfavourable outcome.

Does training general practitioners result in more shared decision making during consultations?

OBJECTIVE: We conducted a clustered randomised controlled trial to study the effects of shared decision making (SDM) on patient recovery. This study aims to determine whether GPs trained in SDM and reinforcing patients' treatment expectations showed more trained behaviour during their consultations than untrained GPs. METHODS: We compared 86 consultations conducted by 23 trained GPs with 89 consultations completed by 19 untrained GPs. The primary outcomes were SDM, as measured by the OPTION scale, and positive reinforcement, as measured by global observation. Secondary outcomes were the level of autonomy in decision making and the duration of the consultation. RESULTS: Intervention consultations scored significantly higher on most elements of the OPTION scale, and on the autonomy scale; however, they were three minutes longer in duration, and the mean OPTION score of the intervention group remained below average. CONCLUSION: Training GPs resulted in more SDM behaviour and more autonomy for the patient; however, this increase is not attributable to the adoption of a patient perspective. Furthermore, while we aimed to demonstrate that SDM facilitates the reinforcement of patients' positive expectations, the measurement of this behaviour was not reliable. PRACTICE IMPLICATIONS: In supporting SDM, professionals should give greater attention to patients' treatment expectations.

[Do mental health nurses take over care from general practitioners?] / Verlicht de POH-GGZ de werkdruk van de huisarts?

OBJECTIVE: The objective of this study was to examine mental health care provided by general practitioners and by mental health nurses working in general practices. DESIGN: Observational research. METHOD: We analysed how many consultations with patients with mental health problems were recorded in Dutch general practices in the period 2010-2014. General practices with and without a mental health nurse were compared, and we investigated which patients were mainly treated by mental health nurses. RESULTS: An increasing number of patients visited the GP for mental health problems in the period 2010-2014. GPs collaborating with a mental health nurse recorded a somewhat higher number of patients with mental health problems than GPs without a mental health nurse, but used as many consultations per patient. Mental health nurses mainly treat females, adult patients, and patients with common mental health problems. CONCLUSION: Mental health nurses do not take over care from GPs, but provide additional mental health care to patients with mental health problems. Collaborating with a mental health nurse might increase GPs' alertness to record mental health problems.

The effect of pain on presence and severity of depressive disorders in older persons: The role of perceived control as mediator.

BACKGROUND: The relation between pain and depression is reported repeatedly. It is suggested that pain by itself is not sufficient for the development of depression. We aim to study the role of perceived control as mediating factor in the relation between pain and depressive disorders at old age. METHODS: Baseline data of the Netherlands Study of Depression in Older Persons (NESDO) were used, including 345 persons with DSM-IV depressive disorders (CIDI) and 125 control persons without depressive disorders, aged 60 years and over. Measures included severity of depression (Inventory of Depressive Symptomatology), presence and intensity of pain and pain-related disability (Chronic Graded Pain scale), and a general measure of perceived control over life (Pearlin Mastery Scale). In mediation analyses direct and indirect effects were estimated. RESULTS: Older persons with depressive disorders reported pain more frequently with higher intensity than controls. After controlling for confounding, the direct effect of pain intensity and the indirect effect through perceived control on depression were OR=1.10 (CI 95% .98;1.25) and OR=1.24 (1.15;1.35). For pain-related disability these were OR=1.14 (1.02;1.29) and OR=1.21 (1.13;1.29). In depressed persons there was a strong direct effect of pain intensity and disability and a smaller indirect effect through perceived control on severity of depressive symptoms. LIMITATIONS: This cross-sectional study cannot give evidence on causal direction. CONCLUSIONS: Perceived control plays an important role as mediator in the association between pain and presence of depression. In depressed persons however, the direct role of pain seems more important in the association with depression severity.

The prevalence of long-term symptoms of depression and anxiety after breast cancer treatment: A systematic review.

OBJECTIVES: It is unclear whether breast cancer survivors have a higher risk of long-term symptoms of depression or anxiety. The aim of this study was to systematically review the evidence about long-term symptoms of depression and anxiety in breast cancer survivors. STUDY DESIGN: Systematic review. MAIN OUTCOME MEASURES: PubMed, Embase, Cochrane and PsycINFO were searched for studies with at least 100 survivors ≥1 year after diagnosis, and which used common questionnaires measuring symptoms of depression or anxiety, by two independent reviewers. The quality was assessed with the NIH 'Quality Assessment Tool' checklist. Prevalence of symptoms of depression and anxiety was compared to time since diagnosis, available control groups and a general female population. RESULTS: Seventeen articles were included in this review with an average quality score of 57% (range 38-86%). The prevalence of symptoms of depression varied from 9.4% to 66.1% and of anxiety from 17.9% to 33.3%. The results on the depression scale suggested an increase in risk of symptoms of depression for breast cancer survivors at one year after diagnosis, which decreases over the ensuing years. Symptoms of anxiety were not more prevalent among the women with early stage breast cancer. CONCLUSIONS: This review suggests a higher prevalence of symptoms of depression among breast cancer survivors than among the general female population, persistent over more than 5 years after diagnosis. Health care providers should be aware of this. There was no indication for an increased prevalence of symptoms of anxiety among breast cancer survivors.

Electronic diary assessment of pain, disability and psychological adaptation in patients differing in duration of pain.

Computerized diary measurement of pain, disability and psychological adaptation was performed four times a day for 4 weeks in 80 patients with various duration of unexplained pain. Reported are (1) the temporal characteristics and stability of pain report during the 4-week measurement period, (2) the association between pain duration and pain report, disability and general psychopathology, and (3) the accordance between diary assessment versus questionnaire assessment of pain, disability and psychological adaptation. No evidence of instrument reactivity was found: pain report was stable across the 4-week period. However, pain report appeared to be highly variable both between and within days. About half the patients showed a clear increasing trend in pain during the day. Several differences were found between subgroups of patients varying in pain duration. Patients with less than 6 months of pain reported significantly less pain intensity, disability and fatigue than patients whose pain persisted for more than 6 months. Pain coping and responses to pain behaviors by the spouse also differed for the subgroups: longer pain duration was associated with increased catastrophizing and solicitous responses from the spouse. Comparison of scores obtained with diary versus questionnaire assessment indicated moderate correlations for most variables. Retrospective (questionnaire) assessment of pain intensity yielded significantly higher pain scores than diary assessment.

Unexplained severe chronic pain in general practice.

The aim of this study was to estimate the prevalence of unexplained severe chronic pain (USCP) in general practice and to report medical as well as psychological descriptions of patients suffering from this condition.A total of 45 GPs in 35 different practices included patients throughout the year 1996. Patients were included according to the following criteria: between 18 and 75 years of age; pain which had lasted at least 6 months; pain is the most prominent aspect in the clinical presentation; pain is serious enough to justify clinical attention; pain has led to obvious discomfort and disability in daily life for at least for 1 month. Medical aspects were measured with the IASP taxonomy while psychological aspects were derived from the MPI. The overall prevalence of USCP was 7.91 per 1000 enlisted patients. Estimates ranged between 1.87 in the youngest age group and 13.50 in the 55-59 age category. The lower back and lower limbs were most frequently affected and 31% of the patients had pain in more than three major body sites. Pain was most frequently associated by the musculoskeletal system and most often (nearly) continuous. Mean severity of current pain was 3.7 on a scale from 0 (indicating no pain) to 6 (indicating a lot of pain). Mean rating of 'average pain in the last week' was 4.1. Regarding the psychosocial and behavioural aspects of pain, 27% of the patients could be described as perceiving severe pain while gaining social support for it. Fourteen per cent felt in the category 'pain combined with affective and relational distress' and 10% was classified as 'coping well with pain intensities lower than those of the other groups'. The other half of the patients were on average or not classifiable on these aspects. Unexplained severe chronic pain lasting more than 6 months had on overall prevalence of 7.91 per 1000 enlisted patients, ranging from 1.87 in the youngest to 13.50 in the oldest patients in these 35 general practices in The Netherlands. Our prevalence estimate of USCP is low compared to other studies on chronic pain. Probably for three reasons: Firstly, our study was confined to unexplained pain and not all chronic pain. Secondly, our inclusion criteria focused the attention of very severe chronic pain patients, and thirdly, we have defined 'chronic' as more than 6 months, while others have been using shorter time spans.

Variations in the diagnosis of psychosocial disorders: a general practice observation study.

The factors that influence the interpretation and treatment of psychosocial complaints by general practitioners are discussed. The assessment of complaints differs considerably from one GP to another, in the sense that one will attach significance to psychological and sociological factors in many more cases than another. We investigate the effect of physician characteristics and their styles of communication on their bias over psychosocial assessments and treatment, and the way these effects are interrelated. The interpretation and eventually the treatment of complaints by 30 GPs (complaints presented at approx. 50 consultations per GP) were studied. Data of treatment and communication were collected from observation of the videotaped consultations, data of interpretation were collected on questionnaires for each consultation; doctor characteristics were inventarized by questionnaire. The following results can be reported. When a doctor communicates with a patient in an open manner, more complaints are interpreted as 'non-somatic' and treated as such. The same is the case among doctor's with a 'general medical' approach, rather than a 'clinical' one, when we look at interpretation. The effect on treatment is less marked. Practice characteristics and a GP's subjective feeling of competence have greater consequences for treatment than for judgement. There is not much interaction between the discerned effects.

Psychosocial problems in primary care: some results from the Dutch National Study of Morbidity and Interventions in General Practice.

According to standardized screening instruments, mental distress is a common phenomenon among many patients who visit their general practitioner. However, a number of patients who seem to be in need of mental help do not put forward such a demand for help, whereas other patients who express psychosocial problems to their GP are not considered to be in need, according to a standardized measure. In this paper, a distinction has been made between the objectified needs of the patient as expressed by a standardized assessment, and the demands of the patient, expressed by the Reason for Encounter, stated during their visit at the GP. Results of a follow-up study of two cohorts of patients have been presented: one cohort presented during a 3 month period at least one articulated demand for psychosocial help, a second cohort presented at least one somatic complaint, considered by the GP as being psychological by character, without presenting any psychosocial complaint in that period. Objective needs for mental help of patients in both cohorts were assessed by means of the General Health Questionnaire. During one year all consultations of these two cohorts were registered. The following questions have been put forward: what demands for help have been put forward by the patients, what treatment have these patients got, and what has been the course of the problems during one year of patients with different needs and demands. From the results the following conclusions may be drawn: many patients with a probable mental illness (according to their objective need) present only physical symptoms.(ABSTRACT TRUNCATED AT 250 WORDS)

Byrne and Long: a classification for rating the interview style of doctors.

In order to assess the doctor's verbal behaviour in the consultation, a classification instrument is needed that will enable us to summarize the doctor's interview style. Such an instrument, developed by Byrne and Long, is evaluated in this study. After presenting the procedure as it has been worked out by Byrne and Long, some points of criticism are formulated. On account of our criticism, some changes have been introduced into the instrument. To test the relevance and reliability of the instrument, 36 consultations have been classified according to the adjusted system. This procedure showed that it was possible to represent a consultation with the 45 categories the system consisted of (although only 15 categories occurred in more than half of the consultations). Those categories that occurred often enough, could be measured with sufficient reliability, with inter-observer reliability coefficients mostly above 0.70. An example shows that the procedure is extremely useful to indicate certain differences between doctors and types of consultations.

Somatic disease and psychological disorder.

The association between physical and psychological disorders has been demonstrated repeatedly. There are a number of explanations for this association, each of them pointing to specific diseases and operationalizations of mental distress. In this article, the relationship between various somatic diseases and a number of indices for psychological distress was investigated. Within one study population, patients with different somatic diseases were identified, and their experience with mental distress, their requests for help from their GP during consultations, and their GPs' diagnoses were registered and compared with the total study population: It appears that relationships could be demonstrated between experience of distress and presentation of psychological symptoms during consultations, on the one hand, and common physical disorders, on the other. Patients with neurological diseases (Parkinson's, epilepsy, multiple sclerosis) and gastric ulcers showed the same relationships, but were also more frequently diagnosed by the GP as having psychological disorders. Patients with a number of other serious somatic diseases, such as diabetes, cancer, and arthritis, did not distinguish themselves in a positive way on one of indices for psychological distress.

Analysis of referrals of mental health problems by general practitioners.

The majority of people in the community who have a psychiatric disorder will consult their general practitioner. Referrals from general practice to specialist services are, however, relatively rare. The filter between primary care and specialist care has been characterized by Goldberg and Huxley as the least permeable of the filters separating psychiatrists and other specialists from the populations they serve. These referrals form the subject of this study in the Netherlands. Using a large database of doctor-patient contacts, the proportion of mental health disorders resulting in a referral and the characteristics of the patient and general practitioner that are involved in such a referral have been determined. In addition, the type of mental health institution or specialist to which referrals were directed and the characteristics influencing this choice were examined. Only 6% of patients presenting with a psychiatric disorder during surgery hours were referred to specialist care. Younger patients, male patients and patients with severe diagnoses had a greater probability of being referred. The percentage of patients referred was higher in urban areas than in rural areas. Doctors with a limited task perception regarding mental treatment tended to refer more often. Although the diagnosis did have some relationship with the institutions to which patients were referred (psychotic conditions to psychiatric services and social/material problems to social workers), the most prevalent diagnoses (neurotic conditions and relationship problems) seemed to be more or less randomly distributed over the various possibilities. Preferences appeared to be related to the existence of regular meetings between general practitioners and specialists and a positive evaluation by general practitioners of the institution concerned.

GPs' referral to mental health care during the past 25 years.

Previous research has shown that mental disorder in the community has remained fairly constant over the past 30 years. As a result there has been a shift in mental health care from primary care to specialised mental health care. This shift should be visible in higher referral figures from general practice. In this longitudinal analysis of mental health referrals (1971 to 1997), the authors aimed to answer whether these higher referral rates have occurred, whether there are increases in referral for specific groups, and whether the referral pattern has changed. The results demonstrate an increase in referral rate with a factor of 4.5. It is concluded that we are witnessing a pull from mental health care together with a push from general practice, thus reinforcing each other.

Communication in general practice: recognition and treatment of mental illness.

From previous studies there is a lot of evidence that in primary care settings, many patients tend to express their mental problems in terms of physical symptoms. Therefore, the general practitioner (GP) needs to recognize mental problems at an early stage. Early recognition allows for adequate treatment that might speed up recovery. The present article reports on a study exploring the GP's ability to recognize mental illness, the communication style that is supposed to support this ability, the subsequent treatment of mental problems, and the patient's recovery. Two databases were used. First, an observation study, involving 351 videotaped consultations held by 15 GPs, yielded information on communication style and recognition abilities. Patients in this study were selected randomly. The second database obtained treatment data and measures of patient recovery from a 1-year follow-up study dealing with the treatment and course of mental illness. Patients in this study were selected because their GPs considered their problems "mainly psychosocial by nature". Half of them were categorized within psychological and social diagnostic categories of the International Classification for Primary Care (ICPC), the other half were categorized within physical disease categories, with an assessment by the GP that the complaints were mainly psychosocial. Results showed no significant relationships between the recognition of mental illness and nine communication features supposed to induce these abilities. There was a tendency however, for a positive association between recommended communicative behaviour of the GP and his or her tendency to give frequently psychosocial evaluations of the patient's complaints. Also, there was a negative tendency between this recommended behaviour and the degree of agreement between the GP's evaluation and the score on a psychiatric screening questionnaire. This agreement is called "accuracy". Frequent psychosocial evaluations were related to exploring behaviour and mental health referral in case of psychosocial complaints. Further, relationships between the GPs' recognition ability and various measures of patients' recovery did not prove univocal. Both positive, negative and absent relationships were found.