Azithromycin and Quality Care

This is my first blog that hopes to distill and appraise some of the recent relevant and practical research for training and practicing medics in developing countries. In it, I will try and select just a couple of recent papers and pick out some of the key areas of discussion. It goes without saying – anything not open access will not be discussed.

I will kick off with one of the more high profile studies of recent weeks, a cluster-randomized, double blinded control trial of azithromycin, given every 6 months for 2 years, to children aged 1 month – 5 years in Malawi, Tanzania and Niger, published in the New England Journal of Medicine.

To start, what a huge undertaking, almost 200,000 children across 1533 communities were identified  in the baseline survey. Why the need for such a large sample size? Essentially, a mixture of factors – reduced child mortality means a lower rate of primary outcome, relatively large variation between clusters, and an estimated effect size of 10%.

The results show a statistically significant reduction of 13.5% in under 5 mortality, with the majority (and indeed only significant) benefit seen in the largest site – Niger. In subgroup analyses of different ages, the youngest (especially aged 1-5 months) have the greatest benefit. Interestingly, as time progressed over the 2 years, there appeared to be an increasing effect of azithromycin.

Facing the current global antimicrobial resistance threat and evidence from trachoma mass drug administration of azithromycin resistance (as described in the trial), the concept of mass antimicrobial prescription is tricky to reconcile. Furthermore, outside of specific subgroups, the only real benefit was seen in Niger, where child mortality is highest and pneumococcal outbreaks common with limited (as yet) vaccine coverage. In Malawi and Tanzania, only a modest trend was seen (perhaps with greater numbers significance would have been reached). Finally, data collection was census based and relatively minimal, not providing enough information about to understand how and why the intervention seemingly worked. Do we change practice because of the trial? I would argue not without more information; better vaccination may be more targeted and longer term benefit may be more likely with a health systems approach. This intervention seems like a temporary bridge to quality care.

On the subject of quality care, a vaguely defined evidence synthesis (non-systematic review?) joint effort by the World Bank, Harvard University and the BMJ, found that universal care does not automatically equate to quality care. Citing examples such as 90% of angina cases misdiagnosed in Nairobi, they argue that improving quality should be a pre-requisite for access, as without it – what is the point in access? Qualification do not equate to clinical knowledge – a large proportion of nurses in Kenya outperform doctors on clinical vignette testing, and finally, that often even though staff know what to do – because of short consultations and perverse prescribing incentives – they don’t do it. Despite the limitations in methodology, the authors make an important point, we need new approaches to tackle low qualitt training, under-motivation and under-utilization of healthcare professionals. But do they need to be “new”? Globally, there is a wealth of experience on these areas, we just need to implement it.

Nb. One acronym used for the needs of healthcare providers by the Health Information For All (HIFA) group is SEISMIC: 

- Skills
- Equipment
- Information
- Structural support
- Medicines
- Incentives
- Communication facilities.

Acknowledgment also goes to Dr B Sossen, for presenting the NJEM paper last week.