What can alert the general practitioner to people whose common mental health problems are unrecognised?

Mental disorders are common in primary care, with a review of studies using structured case-finding interviews concluding that they were present in a quarter of all presentations.1 Depressive and anxiety disorders were most common, with current depression reported in 17% and generalised anxiety in 10% of all presentations.1 When self-report screening instruments are used, the rates of possible mental illness are even higher, with estimates of a third2 to a half3,4 of all primary care presentations.

The assessment and management of common mental disorders in primary care is thus an essential aspect of health care. Increasing treatment rates over the past 15 years are likely to have played a significant role in the fall in suicide rates in Australia.5 Early intervention for common mental disorders is now widely promoted,6 with an emphasis on increasing detection and treatment rates in young people.7 In primary care settings, depression and anxiety frequently occur in conjunction with other medical conditions. Patients may be more willing to emphasise their somatic symptoms, and, consequently, non-specific symptoms such as insomnia, prolonged fatigue, headache, nausea and musculoskeletal pain are common presentations in people with depression or anxiety.8

The self-report 12-item Somatic and Psychological HEalth REport (SPHERE-12)3 was developed specifically to capture non-specific presentations that may indicate the presence of a psychological disorder, although its efficiency as a screening tool in primary care has been criticised.9 Others have promoted tools such as the Composite International Diagnostic Interview – Auto (CIDI-A),10,11 which generates diagnoses based on criteria of the Diagnostic and statistical manual of mental disorders, fourth edition (DSM-IV) and the International classification of diseases, 10th revision (ICD-10).

In recent years, the rate of psychological disorder diagnoses made by general practitioners has remained relatively steady, at less than 15% of presentations.12 Given that about a third of these will be patients not currently reporting psychological symptoms, we can estimate that GPs detect about half of those with high symptom levels. One view suggests there is little merit in promoting detection of those patients with psychological disorders that GPs do not recognise, in the belief that these are less severe cases or that they will be detected on another occasion. An alternate view is that among those undetected are a significant number of people at considerable risk of ongoing disability or premature death, and that the reasons they are not detected are likely to be due to other demographic, symptom pattern, or medical practice factors.

To explore these competing views in Australian primary care, we compared two very different case-finding techniques (SPHERE-123 and CIDI-A11) with GPs’ detection and treatment rates for psychological disorders.

The rates of depressive and anxiety disorders reported here are comparable with rates detected in primary care,1 although rates of somatisation disorder were higher.14,15

In a large New Zealand primary care study,16 GPs recognised psychological issues in more than half of patients not meeting CIDI-diagnosed DSM-IV disorder criteria, suggesting that GPs use different assessment criteria when assessing psychological problems. In our study, GPs identified psychological issues in 20% of patients overall, but only 30% of these patients met 1-month criteria for at least one of the DSM-IV disorders. This finding underlines the premise that “primary care psychiatry is not specialist psychiatry in general practice”,17 and that GPs are managing different conditions, where the DSM-IV system is not the “gold standard”.16 Because of this difference between GP practice and psychiatric practice, we took GPs’ judgements of the presence of psychological problems as the benchmark for caseness, rather than ICD-10 or DSM-IV psychiatric diagnoses.

Although the structured CIDI/DSM-IV case-finding instrument did not seem useful in guiding management of common mental health conditions in primary care,17 GPs still need knowledge of psychiatric condition diagnoses (such as melancholic depression, bipolar disorder, schizophrenia and severe somatoform disorders) used by psychiatrists, psychologists and mental health teams.

Conversely, the SPHERE-12 captures a much wider range of problems presenting as psychological or somatic distress, and has more in common with GPs’ clinical decision making. Its threshold is designed to identify patients who may require further questioning about common mental health problems, and this appeared to be successful here. The GPs seemed more familiar with identifying patients with depression and anxiety symptoms (ie, PSYCH-6 only, or PSYCH-6 and SOMA-6). However, there is still work to be done in assisting GPs to identify and manage psychosomatic presentations (ie, SOMA-6 only). This finding is in keeping with other reports of lower recognition of somatising patients who present without obvious psychological symptoms.12,14,15 We found that the use of self-report patient information on general health and disability added to the identification of somatic patients whom GPs had overlooked.

Our study has some limitations, most notably in retaining GPs and their patients in what was essentially a naturalistic study. In particular, patients with psychological impairment (indicated by SPHERE-12 caseness) were more likely to be excluded from the study due to missing data or inability to be contacted for telephone interviews. The study sample was also significantly older than the group with incomplete data, and may have had more medical comorbidity. Despite these sampling problems, caseness rates on the SPHERE-12 were similar to those reported elsewhere in large Australian3 and New Zealand16 primary care samples.

The SPHERE-12 was designed as a screening instrument with broad criteria, and subsequently produces high rates of possible cases, particularly those with somatic presentations, as in this study. The somatic/fatigue scale of the SPHERE-12 may assist in identifying patients requiring further consideration of “hidden” and potentially treatable psychiatric disorders.18 A screening instrument is useful if it captures most or all likely cases, including somatic patients, and leads to better and early identification of psychiatric disorders, less distress, better functional outcomes and more efficient use of medical resources. The SPHERE-12 is not a diagnostic system, but a tool for identifying patients with whom GPs could potentially spend more time to assess their emotional distress.3 However, the benefits of such a screening instrument need to be weighed against the potential costs of capturing false positives.

We noted a relationship between perception of poor to fair general health and higher caseness rates using the SPHERE-12 and CIDI/DSM-IV. Health perception describes a patient’s view of his or her health, and a study of adult patients in general practice found similar rates of patients who perceived themselves as being in poor health (21%, compared with 24% here).19 In that study, the authors noted that the patients’ perception of poor health (when contrary to their doctor’s perception) predicted emotional distress, poor social and occupational function, and high service utilisation, and concluded that the patients’ doctors had not appreciated the impact of health-related worry, acute pain, anxiety or depression on the patients’ sense of wellbeing.19 The SPHERE-12 already includes this general health question and a short self-rated disability measure,3 and should prove effective in detecting somatic presentations.

We contend that the SPHERE-12 is a useful screening instrument, particularly if supplemented by a disability measure to increase detection rates of clinically relevant disorders. The addition of the SF-12 MCS disability criterion gave similar sensitivity and specificity rates to those for the CIDI/DSM-IV, and has the potential to identify patients likely to benefit from mental health interventions. There is still a need for more education to assist GPs in conversing with their patients on the issue of medically unexplained symptoms, as increasing recognition will only improve outcomes if GPs have the skills and resources to deliver adequate interventions.