On The Pulse - 1st June 2012
New therapy for tinnitus
Many treatments have been suggested for tinnitus, but few are effective. A Dutch study in The Lancet randomizes 492 patients to usual care or cognitive behaviour therapy with sound-focused tinnitus retraining therapy, with stratification by tinnitus severity and hearing ability. At 12 months, the intervention was superior to usual care in health-related quality of life, tinnitus severity (measured by a tinnitus-specific questionnaire) and tinnitus impairment. An accompanying Comment thinks this should dispel ideas that nothing can be done for tinnitus, or that only pharmacological interventions can be rigorously tested, but admits that this therapy aims at habituation to tinnitus, rather than a cure.
Obstructive sleep apnoea and hypertension
Findings published in JAMA from the Zaragoza Sleep Study, a longitudinal study of 2000 patients referred for polysomnography, suggest that the well-recognized association between obstructive sleep apnoea and hypertension is causal. People who declined continuous positive airway pressure (CPAP) to treat their sleep apnoea were at increased risk of new-onset hypertension, while those receiving it had a reduced risk. There’s an alternative explanation of course: people willing to adhere to CPAP therapy may be more health-conscious and more likely to take other measures to reduce cardiovascular risk.
Back to square one in septic shock
In 2001, recombinant human activated protein C (drotrecogin alfa), was approved for treating severe sepsis in those at ‘high risk of death’ on the basis of the PROWESS study, but subsequent data have not been supportive. The PROWESS-SHOCK study, published in the NEJM, randomizes 169 patients with infection, systemic inflammation and shock to drotrecogin alfa or placebo for 96 h. There was no difference in mortality at 28 or 90 days, with the trend slightly favouring placebo. An accompanying Editorial surveys a sad story of too much trust in a single trial and dubious promotion by the study sponsor.
Scattered research reports
A survey in the BMJ explores the problem of information overload. It surveys RCTs and systematic reviews and finds that, despite some variation between specialties, these articles were published in a huge number of different journals. In neurology, for example, 2770 trials were published in 896 journals in 2009 alone. Obviously, no doctor can make the time to scan so many journals but, apart from speculating that social media tools might help alert clinicians to important new research, the authors don’t offer much by way of a solution.
Loss to follow up
One question that the above study fails to ask is whether all these trials and reviews are worth reading. A few years ago, an article in PloS Medicine argued that published research findings were more likely to be false than true. This week, a paper in the BMJ reckons that loss to follow-up in trials often leads to misleading claims. The investigators took 235 trials reported in five top medical journals and modelled what might have happened if information had been available on those who dropped out. Although the precise answer depends on the assumptions made about likely event rates in those who dropped out, the results of around a third of trials lost their significance.
Magnesium for aneurysmal subarachnoid haemorrhage
Aneurysmal subarachnoid haemorrhage has high morbidity and mortality in young adults, partly because of delayed cerebral ischaemia, which usually develops within 2 weeks, and calcium-channel blockers have only a modest benefit. A phase III trial in The Lancet randomizes 1204 adult patients admitted within 4 days of haemorrhage to intravenous magnesium sulphate or placebo for 20 days or until discharge. Unfortunately there was no reduction in disability at 3 months, including in the prespecified subgroup analyses. An accompanying Comment wonders why the incidence of delayed cerebral ischaemia was not reported, but agrees that routine magnesium infusion cannot be recommended.
Progressive memory loss, confusion, gait disturbance
A 62-year-old woman presents with confusion, memory loss and difficulty walking. She denies nausea, vomiting, lethargy, headache, fevers or night sweats, but smokes cigarettes and has a family history of breast cancer. The only abnormality on neurological examination is instability when walking heel to toe. A CT scan shows a mass in the third ventricle and enlarged lateral ventricles. What’s the next step, asks this clinical challenge in JAMA ? Lumbar puncture to treat the hydrocephalus and obtain cerebrospinal fluid for analysis? IV steroids? Further investigation with CT of chest and abdomen? Or should you consult a neurosurgeon?
Performance measures in hypertension
The problem with measuring performance is that you tend to get only what you measure, often at the expense of other valuable outcomes. A study in Archives of Internal Medicine examines blood pressure in nearly a million US veterans, and concludes that while 94% met their ‘action measure’, (acceptable blood pressure or appropriate clinical action), almost as many were potentially overtreated as undertreated, with possible adverse consequences. An accompanying Commentary likes the approach, but points out that physicians’ actions were probably consonant with guidelines at the time – evaluating physician performance and evaluating patient care are different concerns.