Guidelines for GPs are often based on evidence and studies performed outside of general practice itself. This affects doctors’ judgements of the risk-to-benefit ratio of treatments and the performance of diagnostic tests.
As Tania Winzenberg, Professor of chronic disease management at the University of Tasmania, and I write in the Medical Journal of Australia today, interventions (such as medications and other treatments) are often tested in hospitals and other settings where the patient cohort is markedly different to the one seen in primary care.
A study from the United Kingdom indicates GPs ignore advice from guidelines not relevant to general practice. This means GPs could ignore potentially good advice, disadvantaging their patients and increasing costs.
General practice provides the majority of medical primary health care for Australians. About 85% of the Australian population see their GP at least once a year.
More than 137 million GP medical services were delivered by 33,279 general practitioners in 2014-15. In 2011-12, general practice spending was A$14.2 billion, equating to 9.7% of total health expenditure. GPs, the taxpayer and patients would all wish to believe this expenditure provided effective, safe and high-value services.
The reality is somewhat different, for several reasons.
Unlike many specialists, GPs see a wide range of medical conditions. Australian data from the BEACH study show 167 conditions form 85% of a GP’s work compared to five that form the same proportion of a specialist’s.
In half of all GP consultations, it is not possible to make a diagnosis that fits with the diagnostic criteria of the 10th revision of the International Statistical Classification of Diseases and Related Health Problems (ICD 10), the current international classification of disease. This is not because GPs are bad at diagnosis but because much of what they see is mild illness; the symptoms of which are vague and fleeting and that often resolve on their own.
So one the of primary roles of GPs is to rule out significant illness; and if they suspect something sinister, to take appropriate action which may result in hospital or specialist referral.
The Australian health care system requires patients be reviewed and assessed by a GP first before they can access Medicare subsidies for specialist treatment. This ensures those referred for specialist review have a high probability of significant illness. This in turn keeps the cost of specialist care down and ensures the patient sees the most appropriate specialist.
Many patients, particularly as they age, have multiple problems (multimorbidity), which complicates diagnosis and treatment. To ensure GPs provide consistent care among themselves, that patients are referred in an appropriate manner to specialists and hospitals, and that care is as safe as possible, guidelines have been developed to assist GPs.
These work a bit like a GPS in your car. They offer advice on how to get to where you need to be but are not infallible and it is sometimes wise to ignore their advice.
The National Health and Medical Research Council has developed protocols for guideline development that rely on what are called levels of evidence. There are four levels of evidence and these are listed in order of reliability:
- systematic reviews of multiple controlled trials (studies where one group of people receives the intervention tested while the other receives the comparison, such as a placebo)
- a single controlled trial
- observational studies (where trends are observed over time)
- expert opinion.
Where possible, we’d like to have the highest level of evidence – expert opinion is often wrong.
A good example from Australia is the belief from doctors and patients that skin wounds need to be kept dry to prevent infection developing. But a 2011 controlled trial showed showering does not increase wound infection rates.
But then, even when there is controlled trial evidence available, it may not be appropriate for primary care. Patients who participate in controlled trials are often different to those seen by GPs.
Children, women of child bearing age and older people are less frequent participants in trials. Trial subjects tend not to have chaotic lifestyles, do not have literacy or language problems, or rarely have several conditions at once.
Depression is a common condition in Australia. Around one-third of the population will suffer at least one episode in their life time. The type of depression seen by GPs is very different from that seen by psychiatrists and psychiatric inpatient units.
GPs treat milder disease that responds to counselling and psychological therapies. Psychiatrists see severe cases where physical treatments such as medication or electroconvulsive therapy (where an electrical current introduced into a patient’s scalp causes a mild seizure) may be needed to improve the patient’s condition.
If GPs follow advice based on evidence in patients seen by psychiatrists to treat their patients with milder depression, the GP could over-treat or even harm their patient through inappropriate therapies.
Yet an Australian study of international general practice depression guidelines found most are derived from trials conducted in psychiatric settings rather than general practice settings.
How can this change?
There is a need to improve the research capabilities of Australian general practice. It is particularly difficult to get funding for controlled trials of non-medicine therapies or for medicines that are off-patent.
We lack the required trained general practice research workforce, research infrastructure and some certainty of ongoing support. Our current general practice researchers are good, and have high international recognition, but we just do not have enough to address the urgent needs of Australian GPs and their patients.
Gerard Gill received funding from NHMRC. He is a Fellow of the RACGP and previously a member of the RACGP national research council.