GENERAL practitioners are used to filling gaps in the health system. Over our careers, we have lived through times where we are seen as underqualified and then essential to a range of services, including maternity care, dermatology, sexual health and more recently, urgent care, infectious disease and psychiatry.

Justine is well known to you and presents with ear and throat pain.  Over the course of the consultation, she becomes distressed and tells you it is the anniversary of her mother’s death. She says she’s also “not coping well” at the moment, because her partner is verbally abusive, and getting “paranoid”.

Jarra is a young man with known psychosis and substance abuse living in a small town where you are the local GP. He presents with florid psychosis. You know that the RFDS will need to sedate and intubate if you transfer him to Darwin for tertiary care. You need to assess whether treating locally or transferring has the greatest benefit and lowest harm.

Anton is 15 and you’ve seen him a lot in childhood, because he was treated for leukaemia. Recently you’ve noticed that he has been wearing more feminine clothing, and makeup. Today he tells you he is feeling “really anxious” and not doing well at school. He has always been shy, and asks if perhaps he might have autism.

This ability to flex with community need is one of the core capacities of generalists, and enables the health system to rapidly adapt to changing community need.

In mental health, we see the system as a Venn diagram, with specialised primary, secondary and tertiary services creating a series of overlapping circles. Patients move in and out of these circles, but we occupy the space that none of them reach. Our patients are often not ready, willing or able to access other services. They may not accept that they have a mental health condition, or not feel able to admit they need mental health care, so we provide the care we can, often without declaring it to be mental health care at all.

This capacity GPs have to cross the artificial boundary between mind and body allows us to manage emotional precontemplators, who are not yet ready to seek more specialised care. Our rural colleagues are particularly adept at engaging people who have high self-stigma, including rural men.

Our patients may be rejected or ejected from other forms of care, or there may simply be no other option they can access or afford. Many of these patients will never receive any other mental health care than us, even if we do our very best to “coordinate” care, in an environment where such care will never exist.

In order to understand this population, it is essential to respect, engage and listen to the people who work with them. The professionals inside the “circles” have a critical role, but are not experts in the population they will never see. If we are to design appropriate mental health policy for all Australians, it is essential that we understand the space outside the circles of care currently supported.

Value-based healthcare and segments in mental health

In recent decades, there has been a drive to simplify and standardise mental health diagnosis and treatment. Given the numbers of people experiencing mental ill-health, it seems logical to streamline diagnosis, and prescribe appropriate evidence-based treatment using algorithms based on symptom lists. The Better Access initiative was built on this assumption: make a diagnosis of a common mental health condition, prescribe an evidence-based therapy and measure the outcome. This process is enshrined in the MBS, where we are required to diagnose a mental disorder and refer to evidence based treatment, if we are to provide “Better Access” to mental health services.

Value-based healthcare principles would suggest that the best way of achieving better outcomes in healthcare is to define a relatively homogenous group of patients, develop a series of outcomes that matter to patients, and continue quality improvement to get the best value care for the lowest cost. The model has merit for many patients with mental health conditions. Face to face services and online programs for disorders such as depression, anxiety and eating disorders have enabled more care to occur outside of the health system, particularly for those with mild to moderate symptoms. However, mental health care is broader than the treatment of disorders, and mental health policy needs to manage that breadth.

Often, the teams driving value-based healthcare interventions  for mental disorders discuss unacceptable variation in health care delivery, and burgeoning “low value care”. The idea is that more industrialised models of standardised delivery will deliver better outcomes for lower cost. Implicit in these models is the assumption that GPs provide “something unspecified” for patients when they should be using technological solutions and multidisciplinary teams. Proponents of this model assert that this sort of model provides “right care, first time”, reducing inequity, enhancing access and improving outcomes.

However, in general practice, we do not only see patients within these defined segments. We see 85% of the population every year, giving us a unique opportunity to implement primary, secondary and tertiary mental health promotion strategies for patients outside the segments of care for which there is clear evidence and generalisable outcomes. This is not surprising, given we work cradle to grave, prevention to palliation, maintenance to crisis, undifferentiated to defined, remote to urban and across the socioeconomic spectrum.

This is why GPs are in an ideal position to inform policy decisions across the breadth of mental health needs in the diverse communities we serve.

Nature of generalism and the breadth and depth of psychiatric diagnosis

General practice is one of the few parts of the health system where people who are, or believe themselves to be, unwell can access care without having to meet referral criteria. People attend with symptoms, feelings, sensations and emotions, and usually present with thoughts and ideas about the source and the meaning of their discomfort. In all the talk about “evidence-based medicine” it can be easy to underestimate the complexity and importance of partnering with patients to make a good diagnosis. The time “before” diagnosis is important.

Good diagnosis is not just about sorting people into categories, any more than good management is just about following protocols. Good diagnosis involves categorical classification and a more qualitative narrative, what psychiatrists would call “the formulation”. Together, they provide a layered understanding of “what is going on”. Much of our evidence is based on categorical thinking, but many would argue that there is considerable diversity in the experience of patients who meet the diagnostic threshold for any mental disorder.

Symptoms can be a sign of a specific mental illness, but can also be a description of stressful circumstances, or mental angst or anguish that doesn’t reach the level of a clinical mental health diagnosis. Frequently those symptoms can point to another physical condition, such as hypothyroidism or anaemia, or even herald rare diseases that might take years to identify, including autoimmune or neurological disease. Patients who live with other chronic illnesses, disability, cultural and linguistic diversity, difficult psychosocial circumstances, or even a non-neurotypical brain can experience and express their distress in ways that are harder to untangle and understand.

GPs are highly trained to sort through undifferentiated symptoms to make a diagnosis, where there is one, and to consider the breadth of conditions behind a patient’s distress. Premature closure at this point can cause harm because it prevents optimal treatment and diminishes patient trust. Diagnostic checklists at this point can be misleading, and perpetuate harm by offering the wrong diagnosis and the wrong treatment at the wrong time.

Nature of mental health diagnoses

In some fields of medicine, diagnosis involves clear, categorical thinking with well-defined taxonomies. A leg is broken, or it isn’t. A person reaches the diagnostic threshold for hypertension, or they don’t.

Psychiatric diagnosis is inherently complex and qualitatively different. There is considerable co-morbidity, overlap in diagnostic categories, and change in diagnostic descriptions over time. Psychological distress is not always caused by psychiatric disorders, and in general practice, many of our patients with mental ill-health will never receive a psychiatric diagnosis.

One of the criticisms of the DSM 5 is its tendency to try to separate behavioural indicators from the sociocultural context of the person, and to diminish the qualitative experience of mental illness. Critics argue that patients may well fulfill the diagnostic criteria of major depression, without having mental illness at all. Our psychiatrist colleagues give us essential support in this endeavour, using their extensive training to provide a layered understanding of a person’s mental health that enables us to manage their care effectively.

To understand this complexity, consider the diagnostic criteria for “major depression”. The symptom lists below could easily apply where the patient has lupus, or a pituitary adenoma, or another mental illness such as schizophrenia, substance misuse or eating disorder, hence the caveat at the end of the criteria that the episode “must not be caused by another medical or psychiatric condition or substance abuse”. Before we involve patients in a segment-based treatment program, it is essential that we ensure we are choosing the right segment. Providing inappropriate treatment is not only wasteful, it can also cause harm.


DSM 5 criteria for Major Depression


Five or more of the following symptoms have been present during the same 2 week period and represent a change from previous functioning. Symptoms 1 and/or 2 must be present:


1.      Depressed mood most of the day, nearly every day

2.      Markedly diminished pleasure in activities most of the day, nearly every day

3.      Significant weight loss or change in appetite

4.      Insomnia/hypersomnia

5.      Psychomotor agitation/retardation

6.      Fatigue

7.      Diminished ability to think or concentrate

8.      Recurrent thoughts of death and/or suicidal ideation.


The symptoms must cause impairment in functioning

The episode must not be caused by another medical or psychiatric condition or substance use.



Nature of distress and emotional responses to social circumstances

Almost all GP consultations have an element of mental health, though not all have an element of mental illness. People present to a GP with a symptom when their usual methods of managing their health are not sufficient. A GP – especially if the patient has developed a trusted relationship – is often the only professional someone can present to when they feel upset. Conversations about physical illness can often expand to incorporate mental health concerns, including distress arising from loneliness, from violence at the hands of a partner, from not being able make ends meet week to week, to the profound long-term effects of a childhood of abuse or trauma.

Patients with a history of trauma that frequently shows itself in a personality deemed as difficult, are often refused care in other health services, or may receive care that is unhelpful or discriminatory. The policy question linked to this situation is whether patients with mental ill-health and no obvious categorical diagnosis deserve as much care as those who have a mental “disorder”. Patients who are suffering deserve care, whether their symptoms fit within a diagnostic algorithm or not.  Except in situations where patients lack capacity (for example, in psychosis) applying a diagnostic label without patient consent is unhelpful and potentially damaging. Patients will not engage in therapy if they feel the diagnosis is wrong.

For many patients, the cause of their emotional distress is external: they are not “disordered”, they are survivors of various forms of trauma. When a patient is injured in a motor vehicle accident, we do not say they have “MVA disorder”. In a similar vein, patients who have lived with deep emotional trauma may resist the label of a psychiatric disorder, because they rightly feel this misrepresents their trauma experience. This is not an academic problem. As GPs, we are well familiar with the tactic used by abusers in the family court, casting the victim of domestic violence as “disordered” and therefore arguing for custody of the children.

Nature of mental health treatment

Once a diagnosis is made in general practice, the GP and patient decide on the next steps, which might include referral to smartphone apps, websites, psychology, psychiatry, medication, social services and agencies, but will always include care (in all the senses of that word) including validation of the patient’s story and being witness to their circumstances, and often a rebuilding of the patient’s story. Coordinating care involves understanding what services are available, and what the patient can access or afford. As GPs, we need to be prepared to “fill the gaps” when the services reject our referrals because the patient sits outside of their segment, or services are simply inaccessible, due to geography, cost or cultural fit.

Guidelines and evidence heavily inform the approach to get the best outcomes for patients within segments of care, but they are not followed uncritically by GPs for patients who sit outside the well-defined segments. This is an example where evidence generated inside segments of care may not be helpful or relevant to the population outside of them. The individual circumstances of the patient are important. Failure to respect context is not only disrespectful, it is also scientifically naïve. While we may treat a broken leg with the same guidelines in Bondi or Broome, mental health is much more context dependent. Implementing “evidence-based” techniques developed in Bondi with urban patients with generalised anxiety may not be appropriate when transplanted into Broken Hill.

For instance, while we may use standardised measures like the K10 to quantify distress, it is important to recognise the limits of these tools. They are not objective, and they are not culturally neutral. Even if translated into the preferred language of our patients, the tools rely on words. Complex concepts like “hopelessness” or “worthlessness” are culturally bounded, and in some cases, can even be seen as offensive. Therefore, outcome tools of great value in specialist contexts may not be helpful in ours.

Patients with “depression” are not homogenous and therefore it is understandable that the one treatment may not be relevant to all of them. When we study interventions like cognitive behavioural therapy (CBT) for instance, we may see an average benefit. However, if we gave paracetamol to everyone in accident and emergency, we would see an average benefit. Blindly applying standard approaches is not what evidence-based medicine was meant to be and yet under Better Access requirements, the MBS specifies which focussed psychological strategies we must use once a disorder is diagnosed. Whilst this is appropriate for patients who sit within the segments of care defined by mental health diagnosis, it is not appropriate across the range of patients we manage.

Similarly, treatments should not only be symptom based. Although psychiatric classification systems have attempted to sidestep causal explanations, and rely solely on observable symptoms, this leads to issues in choosing therapy. We do not treat the bone pain of rickets or multiple fractures using guidelines developed for osteoarthritis. We should not treat the emotional pain of childhood neglect or the trauma of domestic violence with guidelines developed for depression, regardless of whether the symptom lists of DSM 5 are ticked.

One of the privileges of general practice is seeing patients over long periods of time. Patients can have a lifetime of fluctuating symptoms, and need more than treatments developed for first episode illness.

There has been considerable criticism around the cost and lack of efficacy of “endless therapy”. However, we do not have similar difficulties with endless insulin or endless statins. If mental illnesses are chronic, we should expect that treatment will need to continue and adapt over the course of the illness, from prevention to palliation, and that treatment will be negotiated across the trajectory of illness. Clinicians should not see maintenance as a poor use of public funding.

Nature of patient-centredness

The Australian health system aims to be patient-centred, and most health professionals, would hope to practice in a patient-centred way at an individual level. However, the nature of patient-centred health care is complex to deliver, and subject to competing priorities. How can patient-centred care be offered in a system with out-of-pocket costs for psychology and psychiatry too high for many to access, geographic distance too far, or wait times in the public system too long, or apps and websites that require levels of literacy and technical ability that are beyond them?

A system is not patient-centred if it offers only checklist-based assessments to people, or if treatment is framed like a mental health procedure, without consideration of that person’s perspectives, values or context. What’s more, patients with symptoms that don’t fit the diagnostic box, or with normal unpleasant responses to adverse circumstances still require care, even if there’s no diagnosis or evidence-based management. “We’ll see this through together” is patient-centred care.

For those with diagnoses – frequently more than one – one of the roles of GPs is to provide this flexibility to provide patient-centred care within a health system. The evidence shows that this improves outcomes – the so-called paradox of primary care. This paradox states that while GPs follow guidelines less often than their non-GP specialist colleagues, they achieve better, more equitable outcomes.

There have been attempts to reduce cost and increase access by moving away from relationship-centred care, reducing conversation, discussion, negotiation and ultimately patient control over their own narrative. This includes digital mental health services, but also services that use large teams to care for single patients. Given the impact of the therapeutic alliance, even in highly directive therapies, diluting the core therapeutic relationship seems unwise, or at least premature.

Policy implications

If we are to understand and respond to the breadth and depth of mental health issues in the community, we need to think beyond simplistic views of episodic “disorders”. General practice mental health care ranges from disorder management, to prevention, to individual trauma (domestic violence, sexual abuse, medical trauma), to crisis (natural disasters, major medical illness) to life stressors (eg grief, suicide postvention) to social harms (discrimination, harassment) to existential crises (infertility, death and dying).

If we are to understand what is necessary to manage the population that other services cannot treat, we need recognise and respond the work that GPs do outside of Better Access and other intake criteria. At present, we represent GP work by counting MBS Better Access item numbers, which not only under-represents the volume of our work, it mis-characterises the type of work we do.

If we are to develop equitable mental health policy, it is time we understood the breadth of care provided across Australian communities in general practice to better understand the unmet needs of the populations we serve.

Louise Stone is a GP with clinical, research, teaching and policy expertise in mental health. She is Associate Professor in the Social Foundations of Medicine group, ANU Medical School and works in youth health.

Tim Senior is a GP at the Aboriginal Community Controlled Health Service in South West Sydney, and a Senior Lecturer in General Practice and Indigenous Health at Western Sydney School of Medicine



The statements or opinions expressed in this article reflect the views of the authors and do not necessarily represent the official policy of the AMA, the MJA or InSight+ unless so stated.

Subscribe to the free InSight+ weekly newsletter here. It is available to all readers, not just registered medical practitioners.

If you would like to submit an article for consideration, send a Word version to



Women should be notified if they have dense breasts and offered supplemental screening
  • Strongly agree (74%, 202 Votes)
  • Agree (19%, 51 Votes)
  • Neutral (3%, 9 Votes)
  • Strongly disagree (3%, 7 Votes)
  • Disagree (2%, 5 Votes)

Total Voters: 274

Loading ... Loading ...

5 thoughts on “General practice: the liquid in the mental health system

  1. Eszter Fenessy - General Practitioner Sydney says:

    Such a clear and eloquently presented outline of one facet of the work of the General Practitioner. Thank you Louise and Tim. I hope those in decision making positions read, reflect and acknowledge the truths in this article, and that it informs their decisions.

  2. Sharee Johbson says:

    A terrific thoughtful description of general practice in the ideal world. I support this vision wholeheartedly and will support those working to achieve it. In reality many of the general practitioners and rural generalists I meet in coaching feel unable to deliver this kind of medicine. Some of their discontent is brought about because of the gap between what they want to provide and what is actually possible. As a psychologist who has closed her counselling practice in a regional centre i applause this effort to describe an effect response to mental health. Every consultation must consider mental health, it’s part of our overall health as long as we are breathing. Thank you Tim and Louise

  3. Dr Aletia Johnson says:

    Thank you for this wonderfully written, well presented article. When assessing the outcome of a product or intervention in any clinical trial, the number of measured outcomes is, by necessity, small. The more outcomes being measures, the larger the population needs to be for the results to have statistical power. For a disease like community acquired pneumonia, this is not a problem, as there are a limited number of variables to the outcome: did the patient live or die?; did they require hospital admission?; if so, how long did they require admission for?; did they require oxygen?; etc. In mental health care, the outcomes are multifactorial and complex. There are the obvious ones: did the patient commit suicide, need hospital admission etc., but this does not tell the whole story with mental health. KPI’s in clinical work are like outcome interventions in a medical trial, but there are literally thousands of KPI’s to determine how well a mental health intervention has worked. Mental health (or lack of it) is a lifelong condition. The trial only ends when a patient dies. In fact, it does not even end there, because HOW a person dies (and lives) will have implications for people long after they have died. Two people with ‘complex PTSD’ might have two similar initial presentations, but they will exist in very different environments. They will have very different experiences, both with life, and with their health care. One may have a story where they have an easily accessible GP, high intelligence, good health literacy, financial stability and appropriate interventions when required, high levels of resilience and strong family supports. Another may be homeless, living with an acquired brain injury, limited access to health care and multiple undiagnosed medical conditions. They will have different outcomes. Under the current funding models, they would both technically have access to the same treatments. The second patient would use more hospital services, have more admissions, spend more time in the Emergency Department. If one were so inclined, one could work out the cost benefit ratio for the second patient’s care quite easily. The first patient would be more complex though. When one measures outcomes, not everything is measured, especially if it comes out of ‘someone else’s budget’. The cost to the health system may be low for the first patient if they have access to long term, properly funded GP treatment. No hospital admissions, some short term psychological interventions, perhaps one or two psychiatrist reviews when required. However, because the first patient doesn’t require the ‘expensive’ interventions like the second patient, frequent GP visits and counselling sessions would be considered ‘low value care’ in this situation because the patient’s symptoms ‘aren’t that bad’. This decicion will be made by someone who doesn’t know the patient or their story. This is what some Psychiatrists, Economists and Politicians don’t understand or recognise. We GP’s are not ‘gatekeepers’. We are Gurus, Grunts and Oracles. We are expected to know everything, do all of the heavy lifting and also see the future. And we can do it too! Without GP early intervention and ongoing mental and management, the first patient becomes the second. With our input, the patient has much better outcomes. Harm is prevented, but because if this, it is not measured and so our efforts are not considered ‘valuable’. It is time for this to change. We Generalists are not ‘less than’ our more compartmentalised Specialist colleagues, we are more like ‘Long term interventionists’. If doctors were clinical trials, other specialists would be Randomised Controlled Trials and GP’s and Rural Generalists would be Large Population Based Longitudinal Studies. Both are valuable, but only the second has the power to deal with patients with multiple complex variables and outcomes over time.

  4. Anonymous says:

    Brilliant article. As a psychiatrist, I have utmost respect for the under-appreciated and ‘invisible’ (to the MBS) work GPs perform in mental health.

  5. Caroline West says:

    Thank you Louise and Tim for such a comprehensive article . You are spot on . GPs need to be recognised for their central role in helping people manage mental health issues . As the saying goes “it’s complicated “ and GPs are in an excellent position to understand and flex according to a persons needs for management .

Leave a Reply

Your email address will not be published. Required fields are marked *