Published 6 July 2009, doi:10.1136/bmj.b2698
Cite this as: BMJ 2009;339:b2698

Editorials

Who should receive Tamiflu for swine flu?

People with severe disease or risk factors, until the population is vaccinated

The response to the arrival of swine flu in the United Kingdom is unprecedented. In the past, epidemics of infectious diseases have been countered by mass vaccination, never before by mass chemotherapy. Detailed instructions that are modified daily on testing, prophylaxis, and treatment have come from the national Health Protection Agency (HPA) to primary care practitioners via the internet. This strategy could not have been attempted without the internet, and a substantial proportion of the population had to be sufficiently afraid of swine flu for people to readily accept, indeed often demand, a medicine that its prescribers had no previous experience of.

Chemoprophylaxis makes sense when the prophylactic agent is both efficacious and safe, when the disease it is intended to prevent can have serious consequences, and when the epidemic is at a stage when prophylaxis would prevent infection in most of the population until an effective vaccine became available; otherwise it just delays the inevitable. Does experience of the epidemic in the UK support the chemoprophylaxis strategy or is it better to revert to the more conventional one of treating cases as they arise?

Randomised controlled trials provide evidence on the safety and efficacy of neuraminidase inhibitors in the prophylaxis of seasonal flu. Two forms of prophylaxis were tested—prolonged daily dosing regardless of exposure taken throughout the risk period and prophylaxis after household contact with a known case. The reported efficacy was 74% (95% confidence interval 53% to 88%) when using oseltamivir daily as prophylaxis over a six week period and 68% (34.9% to 84.2%) and 89% (67% to 97%) in two studies when it was used as post-exposure prophylaxis for household contacts of primary cases.^{1 2 3} Subsequent reports and retrospective case-control trials show that oseltamivir given after exposure can halt or limit outbreaks of flu in residential institutions.^{4 5 6} In all of these studies, oseltamivir was generally well tolerated, with reports of nausea and vomiting being slightly higher in the treatment group than in the placebo group in the randomised controlled trials in adults. Withdrawal rates as a result of adverse effects were similar in the treatment and placebo groups in all three trials. The product literature describes reports of skin reactions including Stevens-Johnson reaction, but these have been too infrequent to quantify. Reports from Japan, where oseltamivir is used much more often than elsewhere, have suggested a possible association with neuropsychiatric symptoms; however, this occurred only in those on treatment, not prophylaxis, and it may be disease related rather than drug related.⁷

The problem is that chemoprophylaxis has not been tested in an epidemic or pandemic, or against the current epidemic strain. Post-exposure prophylaxis for all contacts of probable cases was the approach advised by the HPA at the start of the current epidemic. However, in the face of widespread transmission in the community—as is now happening in Birmingham, London, Glasgow, and elsewhere—the strategy is no longer practicable because people are likely to be exposed repeatedly.

A policy of treatment as opposed to post-exposure prophylaxis rests on the answers to other questions: how identifiable is early swine flu and how rapidly effective is early treatment, especially in severe cases? In summary, can we readily identify the minority of patients who will develop severe disease and treat them in time to stop progression? If so, we can safely abandon chemoprophylaxis when cases in a given community are arising unpredictably, with most having no known exposure to a contact—the point when containment is no longer feasible.

Early experience in Birmingham suggests some answers. Most people admitted to hospital with swine flu are children and adults under 30, or they come from three groups—those with asthma severe enough to require medication, obese people, and pregnant women—who are evidently at high risk. Previously fit patients with severe disease have usually been unwell for 24 hours, rarely more than two days, with symptoms typical of flu—fever, headache, sore throat, and coryza in most, with diarrhoea or vomiting (or both) in a quarter. Symptoms indicating severity are hypoxia—often with a normal chest radiograph—hypotensive shock, or confusion. Patients with severe symptoms who were previously in good health improve rapidly in one to two days. Those with pre-existing risk factors may need mechanical ventilation within hours of presentation, despite prompt administration of a neuraminidase inhibitor. They may continue to need support for several days, and this will probably exceed the capacity of intensive care facilities.

We have seen patients from high risk groups who have developed swine flu despite chemoprophylaxis; in these cases we have doubled the dose on the grounds that some resistance should be presumed. We are also starting treatment in sick patients and those from high risk groups even when they have been symptomatic for longer than 48 hours because the suggested restriction was based on experience with seasonal flu.

The neuraminidase inhibitors licensed for swine flu treatment and prophylaxis are oral oseltamivir and inhaled zanamivir. We have few data on their use in pregnancy, and zanamivir is recommended in pregnant women largely because blood concentrations are lower after inhalation than after oral administration. Given the apparent early success in Birmingham of preventing progression in people with indicators of severity but without risk factors—by prompt treatment with oseltamivir on presentation, with doses of 75 mg twice daily prescribed if the diagnosis is then virologically confirmed (polymerase chain reaction turnover is 24 hours)—we can safely abandon prophylaxis except, arguably, in patients from high risk groups, which can probably be more narrowly defined than has been attempted so far. Until a vaccine becomes available, we propose that chemoprophylaxis should still be considered for people on medication for asthma, obese people, and pregnant women who have been in close contact with probable cases.

Previously healthy children generally have mild illness, and about a quarter have nausea and vomiting while taking prophylaxis.³ A policy of early treatment for those with defined severe features therefore seems appropriate for children also. This would allow the development of appreciable herd immunity, which might delay and minimise the second wave of cases that is likely this winter if the 2009 UK epidemic follows the pattern of its predecessors.⁸

A statement to parliament by secretary of state, Andy Burnham, on 2 July confirmed that the move from containment and prophylaxis to the "treatment phase" would be led by primary care. Sixty million doses of vaccine are expected to be available by the end of the year, at which point we can expect the epidemic to come to a rapid conclusion.

Our local experience suggests that people in the high risk groups and their healthcare providers are not always aware that they are at a significantly greater risk of severe disease. Primary and secondary care, including emergency departments, medical and surgical admitting units (abdominal pain and gastrointestinal symptoms have been a feature), and obstetric units need to be fully aware of the features of severity and those at increased risk (box). Given that we have a safe and effective treatment, immediate treatment and swabbing for people with features of severe disease or risk factors remains paramount until the population is protected by vaccination.

Risk groups and features of severity Risk factors for severe disease Asthma requiring medication Obesity Pregnancy Features of severity Hypoxia (often with a normal chest radiograph) Hypotensive shock Confusion

Cite this as: BMJ 2009;339:b2698

¹ Department of Infection and Tropical Medicine, Heartlands Hospital, Birmingham B9 5SS

---

Competing interests: None declared.

Provenance and peer review: Commissioned; not externally peer reviewed.

References

1. Hayden FG, Atmar RL, Schilling M, Johnson C, Peretz D, Paar D, et al. Use of the selective oral neuraminidase inhibitor oseltamivir to prevent influenza. N Engl J Med 1999;341:1336-43.[Abstract/Free Full Text]
2. Welliver R, Monto AS, Carewicz O, Scatteman E, Hassman M, Hedrick J, et al. Effectiveness of oseltamivir in preventing influenza in household contacts. JAMA 2001;285:748-54.[Abstract/Free Full Text]
3. Hayden FG, Belshe R, Villanueva C, Lanno R, Hughes C, Small I, et al. Management of influenza in households: a prospective, randomised comparison of oseltamivir treatment with or without post exposure prophylaxis. J Infect Dis 2004;189:440-9.[CrossRef]
4. Vu D, Peck AJ, Nichols WG, Varley C, Englund JA, Corey L, et al. Safety and tolerability of oseltamivir prophylaxis in haemopoietic stem cell transplant recipients: a retrospective case-control study. Clin Infect Dis 2007;45:187-93.[CrossRef][Web of Science][Medline]
5. Chang YM, Li WC, Huang CT, Huang CG, Tsao KC, Cheng YH, et al. Use of oseltamivir during an outbreak of influenza A in a long term care facility in Taiwan. J Hosp Infect 2008;68:83-7.[CrossRef][Web of Science][Medline]
6. Riseborough NA, Bowles SK, Simor AE, McGeer A, Oh PI. Economic evaluation of oseltamivir phosphate for post exposure prophylaxis of influenza in long term care facilities. Am Soc Geriatr 2005;53:444-51.[CrossRef]
7. Fuyuno I. Tamiflu side effects come under scrutiny. News 2007;446:358-9.
8. Miller MA, Viboud C, Balinska M, Simonsen L. The signature features of influenza pandemics—implications for policy. N Engl J Med 2009;360:2595-8.[Free Full Text]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    StumbleUpon    Technorati    What's this?

This article has been cited by other articles:

• Ajlan, A. M., Quiney, B., Nicolaou, S., Muller, N. L. (2009). Swine-Origin Influenza A (H1N1) Viral Infection: Radiographic and CT Findings. Am. J. Roentgenol. 193: 1494-1499 [Abstract] [Full text]  
• Barlow, G. D., on behalf of the BSAC Council, (2009). Swine flu and antibiotics. J Antimicrob Chemother 64: 889-894 [Abstract] [Full text]  

Rapid Responses:

Read all Rapid Responses

Pandemic influenza and antibiotics

Matthew Dryden, et al.
bmj.com, 23 Jul 2009 [Full text]

Hypocrisy over antimicrobial resistance threat

David J Bartram
bmj.com, 23 Jul 2009 [Full text]

Experience in Sri Lanka

Kamal Abdul Naser
bmj.com, 24 Nov 2009 [Full text]

Online poll

Find out more on doc2doc >>