Article Text

Download PDFPDF

  1. Richard Lehman, MRCGP, MA
  1. Department of Primary Care, Oxford University, Oxford, UK

    Statistics from

    Eat less and take more exercise: ancient wisdom, seldom heeded by the rich, or even by those giving it. But it’s one way to prevent diabetes, as the Diabetes Prevention Program showed; the other way is to take metformin. The costs and benefits of the 2 approaches are modelled in

    : lifestyle modification is definitely the better bet. Once the diabetic threshold has been crossed, the risk of cardiac death is markedly increased, and the best marker seems to be QT interval abnormality rather than ambulatory blood pressure or ankle/brachial pressure index (–50). There has been a lot debate about the value of self monitoring of blood glucose in type 2 diabetes treated without insulin, and the Cochrane Review CD005060 provides a welcome analysis—some evidence of benefit, though we still need a large prospective study to decide the issue. Crossing the insulin threshold brings further problems—many people are needle-phobic to start with, so it’s tempting to offer a once daily regime, such as insulin glargine; but the INITIATE study (
    ) achieved better control with a twice daily biphasic (70/30) insulin regime. As for the alleged superiority or otherwise of “human” insulin over animal-derived products, a big Cochrane Review (CD003816) concludes that most of it was a triumph of hype over evidence: there is little to choose between the 2. Hype has certainly surrounded the arrival of inhaled insulin as an alternative to preprandial short acting injections—the two achieved similar HbA1c results in type 1 patients over 6 months (
    ). But in type 2 diabetes, would we make a bigger difference if we used a computer assisted intervention to guide our management? Certainly some, in terms of protocols followed and tests done: but not at all in patient outcomes (
    ). And this echoes the evidence about computer assisted interventions generally—100 were looked at in a systematic review in , showing that computers help clinicians follow pathways, but there is little evidence of patient benefit. A major challenge to both the guideline and IT industry, I think. Acute low back pain gets better in three quarters of people within a month or 2, but that still leaves a quarter in pain at 3 months. What features predict this? A Norwegian study () looked at 123 patients presenting with acute LBP in primary care: lack of recovery was predicted by neurological signs and/or high somatic awareness, though the traditional “fear avoidance” model did not seem to apply. In chronic back pain, however, a therapeutic approach based on overcoming fear and promoting activity has shown lasting results in a randomised trial (). It has taken a long time for doctors to accept the idea that β blockers are mandatory for heart failure of all grades, unless there are very good grounds to avoid them. Age > 70 years is certainly not a barrier, as shown by a meta-analysis of 5 large RCTs (
    ). It is less clear which might be the best β blocker to use amongst this heterogeneous group of drugs, though carvedilol seems the best at reducing levels of B type natriuretic peptide and has powerful antiarrhythmic effects after myocardial infarction (
    ). Most readers will be familiar with JAMA’s indispensable Rational Clinical Examination series, which takes a robustly clinical approach to evidence about diagnosis in a wide variety of conditions. Influenza () and myasthenia gravis () are recent examples. We are all expected to diagnose influenza thousands of times in a working lifetime, which can only be done on the basis of local epidemiology and non-specific symptoms, whereas it’s considered far cleverer to diagnose myasthenia once, on the basis of highly specific motor tests and a spot of edrophonium. Cognitive behavioural therapy (CBT) is the best treatment for panic disorder, but many of us in primary care have very limited access and no CBT skills of our own. A study in
    used a complex design involving CBT but also drug treatment by protocol, mostly administered by non-doctor professionals: the results were good. Don’t panic—but you may need to do some personal learning and service planning before you can throw away your paper bag. Recurrent miscarriage causes much distress, and over the last 2 decades antiphospholipid antibody and lupus anticoagulant measurements have identified a group of women at particular risk. But is there any effective treatment? Happily, a Cochrane Review (CD002859) concludes that heparin and aspirin can reduce the rate by half. Butterbur (Petasites hybridus) is an invasive weed in northern temperate countries, with fragrant flowers in winter. Its root has been shown to be effective for some kinds of allergy, and now it seems to prevent migraine (
    ). Music at bedtime can help older people (age 60–83 y) go to sleep, according to a study in
    . Who knows, for those still young at heart, it may even be the food of love.

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.