Correspondence 200 www.thelancet.com Vol 381 January 19, 2013 1 Bainbridge D, Martin J, Arango M, Cheng D, for the Evidence-based Peri-operative Clinical Outcomes Research (EPiCOR) Group. Perioperative and anaesthetic-related mortality in developed and developing countries: a systematic review and meta-analysis. Lancet 2012; 380: 1075–81. vaccination is not expected to affect the number of dog bites. In conclusion, we support Depani and colleagues’ contention that rabies is an important, but neglected disease. Additionally, however, we emphasise the need for the research community to understand the systemic challenges facing the development of effective and practical solutions for rabies. We declare that we have no conflicts of interest. Syed Shahid Abbas, *Manish Kakkar manish.kakkar@phfi.org Public Health Foundation of India, Disease Surveillance Systems, ISID Complex, New Delhi, Delhi 110070, India 1 Depani S, Mallewa M, Kennedy N, Molyneux E. World Rabies Day: evidence of rise in paediatric rabies cases in Malawi. Lancet 2012; 380: 1148. 2 Kakkar M, Venkataramanan V, Krishnan S, Chauhan RS, Abbas SS. Moving from rabies research to rabies control: lessons from India. PLoS Neglect Trop Dis 2012; 6: e1748. 3 Abbas SS, Venkataramanan V, Pathak G, Kakkar M. Rabies control initiative in Tamil Nadu, India: a test case for the “One Health” approach. Int Health 2011; 3: 231–39. principle of a new four-site intradermal post-exposure vaccine regimen, 2 which is a simplified version of the original eight-site regimen but useable with all WHO-accredited rabies vaccines. It is as immunogenic as the standard five-dose intramuscular course and very economical. 3 The larger volumes of vaccine mean that it does not rely on expert intradermal injection technique. There are three clinic visits: four intradermal injections are given on day 0, two doses on day 7, and one on day 28. (An intradermal dose is 0·1 mL for Vero cell vaccine containing 0·5 mL per vial, or 0·2 mL for vaccines of 1·0 mL per vial.) Any vaccine left over can be used as intradermal pre- exposure vaccination for relatives or others. Dog vaccination and sterilisation remain the only effective means of preventing human rabies in the long term and can reduce the incidence of dog bite. 4 Implementation is challenging, but difficulties can be overcome. 5 This is important in settings such as Malawi where many dog-bite victims do not seek or do not have access to post-exposure prophylaxis. We agree that there is failure to implement rabies control, but that the responsibility is shared by government and non-governmental health agencies as well as the research community. The studies quoted have indicated practical cost-effective methods. The challenge now is to tackle the systemic obstacles, but at the same time apply the best scientific principles and translate research into local health care. We declare that we have no conflicts of interest. *Sarita Depani, Macpherson Mallewa, Neil Kennedy, Elizabeth Molyneux, Mary Warrell saritadepani@yahoo.co.uk Queen Elizabeth Central Hospital, College of Medicine, Blantyre 3, Malawi (SD, MM, NK, EM); and Oxford Vaccine Group, University of Oxford, Centre for Clinical Vaccinology and Tropical Medicine, Oxford, UK (MW) 1 Gongal G, Wright AE. Human rabies in the WHO Southeast Asia Region: forward steps for elimination. Adv Prev Med 2011; 2011; 383870. Systems thinking needed for rabies control Sarita Depani and colleagues (Sept 29, p 1148) 1 admirably highlight the importance of improved rabies control. However, as is the case with most rabies research, 2 they focus on biomedical solutions to complex systemic challenges. For instance, Depani and colleagues propose intradermal vaccines as cost-effective interventions, but our discussions with programme managers 3 revealed that uniform recommendations such as these are not always practical. Intradermal vaccination is most effective and efficient in hospitals with trained staff and a high daily throughput of cases. In peripheral facilities, with fewer trained caregivers and low throughput of patients, doctors preferred to give the entire vial through the simpler intramuscular route rather than waste unused vaccine and administer it intradermally. This finding allowed state managers to better focus their resources on training and supervision of vaccination in central facilities and to procure suitable vaccine formulations. Similarly, Depani and colleagues suggest dog vaccination as a feasible strategy for rabies control. Although this might be a less costly approach than animal population control, real cost-effectiveness data have been difficult to come by from any field study. Our own analysis (unpublished) has shown that to control dog populations or to vaccinate them will be much more expensive than will human vaccination. Moreover, any conclusions about the financial sustainability of such an intervention would have to take into account long-term expenditure, since human Authors’ reply We agree with Syed Abbas and Mannish Kakkar that rabies control is difficult and complex and that biomedical solutions are hard to translate into practice. Implementation of rabies control measures involves financial, strategic, and educational challenges including changing existing practice. Support and research are therefore vital. It is also clear that current policy in many countries is inadequate, ad-hoc, and reliant on outdated practices which should not continue to be promoted. Use of intradermal regimens has been shown to be effective and cost-saving when several patients are treated on the same day, but we agree with Abbas and Kakkar that intradermal post-exposure vaccination with the two-site intradermal regimen, as used in India, is not suitable for use in rural areas and 25% of vaccine is wasted. 1 An alternative intradermal vaccine regimen that uses an entire vial of vaccine given intradermally on the first crucial day is more suitable in peripheral health facilities. This is the