A recent paper in Archives of Internal Medicine examined how often established medical advice is reversed by new evidence. It found that about 13 per cent of papers published in New England Journal of Medicine during the course of a year amounted to reversals of previously accepted belief. It usually takes about ten years for advice to be reversed.

The authors ought to have spoken to the Joint Committee on Vaccination and Immunisation, which has reversed its policy on giving the pneumococcal vaccine to over 65s twice in the past six months. The latest reversal, shockingly, is based on studies from the Health Protection Agency which remains unpublished.

Pneumococcal vaccine is meant to protect against an infection that causes pneumonia. It was launched for the over-65s in 2005, although there were voices even than suggesting that it might not be as effective as the JCVI believed. For example, Norman Waugh of the Department of Public Health at Aberdeen, writing in International Journal of Epidemiology, said that government advice in the UK takes “an unjustifiably favourable view of the efficacy of the vaccine” – basing its advice on observational studies and ignoring randomised control trials that showed no protection.

The JCVI itself had doubts, agreeing to recommend an immunisation programme only by a majority vote and acknowledging that the evidence was not as robust as it would have liked. (JCVI Annual Report, 2003, section 3.2)

In March this year, after a programme involving millions of vaccinations at a cost of up to £100 million, the JCVI changed its mind. The programme had “no discernible impact” on the incidence of pneumococcal disease in older people, it said. By this time, data from the Medicines and Healthcare products Regulatory Agency showed that the vaccination had been linked to 30 deaths and more than 3,300 reported side effects, including heart disorders and joint and muscle plain. 

That seemed clear, albeit a lot of money had been wasted and many patients had suffered side-effects. But yesterday the JCVI reversed its position yet again.

This isn’t because of pressure from the manufacturers. A representation by the Sanofi Pasteur defending its product is dismissed by the JCVI as a “selective interpretation of the evidence”. Evidence from the Health Protection Agency included a statistical analysis showing no observable effect of the vaccine at the population level in England & Wales between 1998-99 and 2007-08. But a second analysis by the HPA using the indirect cohort method did, apparently, show “moderate” effectiveness in people aged between 65 and 74 without clinical risk factors for pneumococcal disease. Neither of these analyses has been published.

On the basis of the second study, the JCVI reversed its advice to the Government, recommending that the programme continue, as it is more cost-effective to vaccinate all over-65s at the time they get the seasonal flu vaccination than to vaccinate simply those deemed by GPs to be at higher risk.

Since the beginning of 2011 the JCVI has therefore reversed its advice twice, recommending the vaccine up to March 16 , then not recommending it (March 16 to July 20), and now recommending it again. Medical knowledge is uncertain, but surely not as uncertain as that.

What is reprehensible is that the latest reversal is based on evidence from HPA that is unpublished. Why? It is exists it should be made available at the same time as the advice based upon it. In its most recent set of draft minutes the JCVI “encourages” the HPA to publish the new studies as quickly as possible. It ought to have insisted they be published at once, if only so that GPs puzzled by the double U-turn could make their own judgement of whether to advise the vaccine to their older patients.