The legally binding text is the original French version TRANSPARENCY COMMITTEE OPINION. 16 April 2008

Transcription

1 The legally binding text is the original French version TRANSPARENCY COMMITTEE OPINION 16 April 2008 TAMIFLU 12 mg/ml, powder for oral suspension One bottle of 30 g (CIP: ) TAMIFLU 75 mg, hard capsule Pack of 10 capsules (CIP: ) TAMIFLU 30 mg, hard capsule Pack of 10 (CIP: ) TAMIFLU 45 mg, hard capsule Pack of 10 (CIP: ) Applicant: ROCHE REGISTRATION LTD oseltamivir phosphate List I Date of Marketing Authorisation: TAMIFLU 12 mg/ml, powder for oral suspension and 75 mg capsule: 20/06/2002 TAMIFLU 30 mg and 45 mg capsule: 19 September 2007 Reason for request: Reassessment of Actual Benefit at the request of the General Directorate of Health and Social Security Directorate Medical, Economic and Public Health Assessment Division 1

2 1 CHARACTERISTICS OF THE MEDICINAL PRODUCT 1.1. Active ingredient Oseltamivir phosphate 1.2. Indications Treatment of influenza In patients one year of age and older who present with typical symptoms of influenza, when influenza virus is circulating in the community. Efficacy has been demonstrated when treatment is initiated within two days of first onset of symptoms. This indication is based on clinical studies of naturally occurring influenza in which the predominant infection was influenza A. Prevention of influenza - Post-exposure prevention: in individuals one year of age or older following contact with a clinically diagnosed influenza case when influenza virus is circulating in the community. - The appropriate use of Tamiflu for prevention of influenza should be determined on a case by case basis by the circumstances and the population requiring protection. In exceptional situations (e.g., in case of a mismatch between the circulating and vaccine virus strains, and a pandemic situation) seasonal prevention could be considered in individuals one year of age or older. TAMIFLU is not a substitute for influenza vaccination. The use of antivirals for the treatment and prevention of influenza should be determined on the basis of official recommendations, variability in the epidemiology, and the impact of the disease in different geographical areas and patient populations Dosage Treatment: Treatment should be initiated as soon as possible within the first two days of onset of symptoms of influenza. For adolescents (13 to 17 years) and adults: the recommended oral dose is 75 mg oseltamivir twice daily for 5 days. For infants older than 1 year of age and for children 2 to 12 years of age: TAMIFLU 30 mg and 45 mg capsules and oral suspension are available. The following weight-adjusted dosing regimens are recommended: Body weight Recommended dose for 5 days 15 kg 30 mg twice daily > 15 kg to 23 kg 45 mg twice daily > 23 kg to 40 kg 60 mg twice daily > 40 kg 75 mg twice daily Children weighing over 40 kg and who are able to swallow capsules may receive treatment with the adult dosage of 75 mg capsules twice daily for 5 days as an alternative to the recommended dosage of TAMIFLU suspension. 2

3 Prevention of influenza : For adolescents (13 to 17 years) and adults : the recommended dose of oseltamivir for prevention of influenza following close contact with an infected person, is 75 mg once daily for 10 days. Therapy should begin as soon as possible,within two days of exposure to an infected individual For infants older than 1 year of age and for children 2 to 12 years of age: TAMIFLU 30 mg and 45 mg capsules and oral suspension are available. The recommended post-exposure prevention dose of TAMIFLU is: Body weight Recommended dose for 10 days 15 kg 30 mg once daily > 15 kg to 23 kg 45 mg once daily > 23 kg to 40 kg 60 mg once daily > 40 kg 75 mg once daily Children weighing over 40 kg and who are able to swallow capsules may receive prevention with a 75 mg capsule once daily for 10 days as an alternative to the recommended dosage of TAMIFLU suspension. Prevention during an influenza epidemic: The recommended dose for prevention of influenza during a community outbreak is 75 mg oseltamivir once daily for up to 6 weeks. Specific populations: - Impaired hepatic function: no dose adjustment is required either for treatment or for prevention in patients with impaired hepatic function. - Impaired renal function: adjust the dose from a creatinine clearance of 30 ml/min or less. Not recommended for a clearance of less than 10 ml/min and for dialysed patients. - Elderly: no dosage adjustment is necessary, except in the case of severe renal impairment. 2 PREVIOUS COMMITTEE OPINIONS AND LISTING CONDITIONS Post-exposure prophylaxis treatment of influenza Opinion of February 11 th, 2004) In subjects from 13 to 64 years without comorbidity, the actual benefit is insufficient to justify reimbursement. In At risk patients : adolescents and adults from 13 to 64 years with comorbidity, adults over 65 years, the Committee considers the actual benefit to be not important and describe it as low. (à verifier) In the specific case of subjects at risk (institutionalised patients, contraindication to the vaccine, immunocompromised subjects (in particular AIDS patients, transplant recipients, patients receiving immunosuppressive drugs), incomplete vaccine protection compared to the circulating strain): the Committee considers the actual benefit to be moderate. IAB III relative to Mantadix Opinion of June 21 th, 2006 (extension of indication in children) Same formulation for the actual benefit in the extension of indication to children older than 1 year. IAB V in the management of post-exposure prophylaxis in children from 1 to 12 years. 3

4 Opinion of January 3 rd, 2007 (extension of reimbursement) Low actual benefit in the following populations unable to undergo influenza vaccination for medical reasons: - Patients with asthma and COPD, - Persons of whatever age in medium or long-stay health institutions - Children and adolescents (from 1 to 18 years) whose health status requires prolonged treatment by acetyl-salicylic acid. Opinion of December 5 th, 2007 (listing of 30-mg and 45-mg capsules) In adults and children In subjects without comorbidity, the actual benefit is insufficient to justify reimbursement. In high-risk populations: children from 1 year and adults from 13 to 64 years with comorbidity and adults over 65 years, the actual benefit is low. In the specific case of subjects at risk: - Subjects living in institutions (institutionalised patients) - Subjects with a contraindication to the vaccine - Immunocompromised subjects (in particular AIDS patients, transplant recipients or patients receiving immunosuppressive drugs) - Situations in which the vaccine only provides limited protection against circulating virus: the actual benefit is moderate Curative treatment of influenza Opinion of February 11 th, 2004, confirmed on June 21th, 2006 Children The actual benefit of TAMIFLU is insufficient to justify reimbursement. Adults from 13 to 64 years of age with or without comorbidity: The actual benefit of TAMIFLU is insufficient to justify reimbursement. Adults over 65 years-old The actual benefit of TAMIFLU is insufficient to justify reimbursement. 5 December 2007 (listing of 30-mg and 45-mg capsules) The actual benefit of TAMIFLU is insufficient to justify reimbursement. 3 SIMILAR MEDICINAL PRODUCTS 3.1. ATC Classification (2008) J : ANTIINFECTIVE FOR SYSTEMIC USE J05 : ANTIVIRAL FOR SYSTEMIC USE J05A : DIRECT-ACTING ANTIVIRAL AGENT J05AH : NEURAMINIDASE INHIBITORS J05AH02 : oseltamivir 3.2. Medicines in the same therapeutic category Comparator medicines Neuraminidase inhibitors: RELENZA 5mg/dose, (zanamivir) inhalation powder, in single-dose container 4

5 3.3. Medicines with a similar therapeutic aim Preventive treatment: - Influenza vaccines - Another antiviral, MANTADIX 100 mg capsule (amantadine), is indicated for prophylaxis of influenza exclusively due to influenza virus A. Non-specific symptomatic medications: analgesics and antipyretics 4 UPDATING WITH DATA OBTAINED SINCE THE PREVIOUS OPINION 4.1. COSMOS Study (cf. appendix) A post-listing study for TAMIFLU was requested by the Transparency Committee in its opinion of 11 February Purpose This multicentre, prospective, longitudinal, observational study over the epidemic period planned the participation of 120 physicians in 120 old people s care facilities (EHPA) and a total of 6500 residents. Endpoint Primary efficacy endpoint: mortality Secondary endpoints: hospitalisations and composite endpoint death + hospitalisations Results 98 EHPA participated to the study and included 8042 patients. Follow-up data were analysed for 6989 residents, in 86 EHPA of these patients were exposed to influenza virus. Only 270 residents were treated by TAMIFLU: 164 for curative treatment and 106 for prophylactic treatment (versus 2000 planned in the protocol). During the epidemic period, 53 EHPA (54%) reported at least one flu-like syndrome, corresponding to a total of 600 residents (9% of the total follow-up population). Among residents not treated by TAMIFLU, 10% (422/4206) presented at least one flu-like syndrome versus 15% in residents receiving TAMIFLU prophylaxis (16/106) ; this difference was not statistically significant. No statistically significant difference was observed between residents receiving TAMIFLU and those not receiving it, both from the primary endpoint (mortality) and the secondary endpoints (hospitalisations), or the composite death + hospitalisations endpoint Other data provided by the applicant The company submitted: - two survey in the curative treatment of influenza, - two surveys in post-exposure prophylaxis treatment the second of which was already presented in the initial dossier (2004) and is not therefore reviewed here, - French epidemiological data on influenza derived from BEH. 5

6 4.1.2 Epidemiological studies in the curative treatment of influenza Prospective cohort study performed in Canada 1. The objective of this study was to evaluate the impact of a specific influenza antiviral on mortality and hospital length of stay, in patients with laboratory confirmed influenza and admitted to one of the 21 voluntarily participating hospitals. The primary efficacy endpoint was mortality during the 15 days after onset of symptoms. Results: 541 patients were eligible, the median age was 77 years and oseltamivir 75mg BID for 5 days was prescribed to 103 patients. By univariate analysis, 27 of the 327 patients in the study deceded within 15 days (8.3%): 4 (3.9%) in the oseltamivir group and 22 (10.0%) in the group not receiving it. The difference was not statistically significant (p=0.08). Moreover, there was no significant difference for the median hospital length of stay in survivors: 7.5 days in the oseltamivir group versus 6 days in the other group (p=0.07). In the final multivariate model, oseltamivir treatment was associated with a reduced risk of death (OR = 0.21, 95% CI = [0.06; 0.80], p = 0.02), but not with a reduction in the median hospital length of stay (p=0.35). There were numerous methodological limitations: - Voluntary recruitment of participating hospitals - Doubts about extrapolating Canadian results to France - Lack of information about the construction of the multivariate model - Potential confounding factors that were not taken into account (antibiotic treatment, underlying lung disease) - Arguable primary endpoint (the 15-day period for recording mortality seems too short), - Potential treatment bias (management with oseltamivir may be a sign of better overall management) - Small patient sample size Because of discrepant results between the univariate and multivariate analysis and the many methodological limitations, this study does not show that TAMIFLU has a significant impact on mortality due to complications of influenza. Retrospective cohort study in Hong-Kong 2 Purpose: to determine the factors affecting hospital length of stay for influenza. Design: retrospective cohort study Population: adults aged over 18 years old with virologically confirmed influenza, admitted in 2004 and 2005 to a Hong-Kong hospitalward. Patients who had had symptoms for under 48h received an antiviral. Primary endpoint: total hospital length of stay. Results: 356 patients were included, in 2/3 of cases; patients were more than 70 years old. Nearly 70% of cases had at least one underlying conditions and 69% presented a cardiovascular or respiratory complication. Oseltamivir was prescribed to 257 patients (72.2%), including 161 (45.2%) within two days and 96 after 2 days. Other antiviral agents were rarely prescribed. By univariate analysis, age over 70 years, underlying conditions or complications and oseltamivir taken more than2 days after onset of symptoms were all associated with a longer hospital length of stay. The median hospital length of stay for patients who received oseltamivir after 48 hours was identical to that of patients who did not receive it (6 days in the two groups, p=0.431). 1 Mc Geer and al. Antiviral therapy and outcomes of influenza requiring hospitalization in Ontario, Canada. Clinical infectious 45 (2007: ). 2 Lee and al. Use of oseltamivir during outbreak of influenza A in a long-term acre facility in Taiwan. Journal of hospital infectious (2007 :1-5). 6

7 Remarks: - Previous studies already showed that oseltamivir given within 48 hours reduced the duration of flu symptoms. - Few explanatory variables were studied so the existence of a confounding factor cannot be excluded. Influenza vaccination, for example, was only studied on a subsample of patients. - The results of this study may not be representative or transposable to the French population Prophylaxis studies in influenza Pragmatic study in Taiwan 3 Purpose: to determine the efficacy of TAMIFLU as preventive treatment after prescription to the population of a long-term care unit. Design: open observational study on 41 residents and 14 care-givers in an rest home during the influenza season in Taiwan. Seven out of 14 staff members and 14 out of 41 residents developed a respiratory disorder during a 14-day period (end January - beginning of February 2004). Curative treatment by oseltamivir was administered to 1 staff member and 7 residents with flu and prophylactic treatment was given to 12 staff members and 30 residents who were either well or had no symptoms suggesting diagnosis of influenza like illness. None of the subjects in the prophylaxis group developed flu infection. Methodological limitations: - No comparator arm: the reduction in the number of recorded cases may simply have been due to the natural course of the disease, more especially as treatment was generally administered late after the onset of symptoms, - The representativity and transposability of the results are uncertain Conclusion The Cosmos observational study on prophylactic treatment of influenza in old people s care facilities did not show that prophylactic treatment by TAMIFLU had a significant impact on morbidity and mortality. In this study, the low number of patients treated by TAMIFLU showed the difficulties of prescription and access to treatment in a real situation. A prospective cohort study performed in Canada in prophylactic treatment of influenza with numerous methodological limitations found that TAMIFLU had no significant impact on mortality. A pragmatic open-label study of prophylactic treatment of influenza in a long-term care facility in Taiwan showed that TAMIFLU had an impact though this study had methodological limitations making interpretation difficult. In a retrospective cohort study of curative treatment in Hong-Kong, the administration of oseltamivir within 48 hours of the onset of symptoms was statistically associated with a reduced hospital length of stay. This study has methodological limitations and confirmed data already examined by the Transparency Committee. To conclude, these data do not modify the Transparency Committee conclusions. 3 Chang and al. Use of oseltamivir during an outbreak of influenza in a long term care facility in Taiwan. Journal of hospital infectious : 1-5 7

8 5 DRUG USAGE DATA According to the data extracted from the IMS/Dorema database (CMA nov 07), 93,000 prescriptions of TAMIFLU are made per year. In 89% of cases, the reason for prescription is influenza. The mean duration of prescription is 5.2 days and the usual dosage is 1.9 tablets per day. These data suggest that TAMIFLU is mainly prescribed for curative treatment Reassessment of actual benefit Prior definitions: 6 TRANSPARENCY COMMITTEE CONCLUSIONS Subjects at risk are defined as subjects aged over 65 years or subjects belonging to the one of 12 categories below (9 long-term conditions and 3 other categories) entitled to reimbursement of influenza vaccine: - Insulin-dependent diabetes mellitus, non-insulin-dependent diabetes mellitus not controlled by dietary measures alone - Invalidating stroke - Severe chronic renal disease and primary nephrotic syndrome - Severe neuromuscular disorders (including myopathy) - Cystic fibrosis - Poorly tolerated congenital heart disease, severe cardiac insufficiency or heart failure, severe heart valve impairement. - Serious chronic respiratory disease (including asthma on the long-term condition list) -Serious primary immune deficiency disorder requiring prolonged treatment, infection by human immunodeficiency virus (for HIV-infected subjects the most recent studies show that vaccination may cause a transient increase in viral load so that vaccination should not be systematically recommended) - Homozygous sickle cell anaemia (congenital haemolytic anaemia due to haemoglobinopathy); - Patients with asthma and COPD, - Persons staying in medium or long-term stay facilities of whatever age, - Children and adolescents (aged from 1 to 18 years) whose condition requires prolonged treatment by acetyl-salicylic acid. Influenza is a highly contagious, acute viral disease which, in most cases, is not serious and spontaneously resolves in about 1 week. However, in certain subjects, influenza complications may be serious and life-threatening. In the child, the clinical aspect of influenza is more subtle when the child is young. Children are the first to be affected during an epidemic. The complications of influenza are particularly serious in infants aged less than 1 year and children with underlying condition (asthma in particular). There are usually no serious complications in healthy children over 1 year of age. The most frequent complications are respiratory and ENT disorders and otitis in particular. The incidence of these complications decreases with age. Vaccination against influenza is the cornerstone of management of this disease. It is particularly recommended in subjects at high risk of complications and in health care providers. TAMIFLU is an antiviral treatment given for curative treatment and prevention (post-exposure prophylaxis). Oseltamivir treatment only has a limited advantage for the following reasons: - The probabilistic nature of influenza diagnosis in particular when the subject is young, - The need to administer treatment within 48 hours of first symptoms for it to be effective, However, data show that problems of health care organisation make this objective difficult to achieve in practise. 8

9 Actual benefit in curative treatment of influenza Children ( 1 year) and adults (< 65 years) not belonging to a high risk population: The new data submitted are insufficient to change the committee s previous opinion. The actual benefit of TAMIFLU for curative treatment in children ( 1 year) and adults (< 65 years) remains insufficient. Children ( 1) and adults (< 65 years) with comorbidity: The epidemiological studies provided do not show a significant reduction in mortality or length of hospital stay under real conditions of use. The actual benefit of TAMIFLU for curative treatment in children ( 1 year) and adults (< 65 years) with comorbid disease is still considered insufficient. Elderly (> 65 years): The actual benefit of TAMIFLU for curative treatment in the elderly (>65 years) is still considered insufficient. Actual benefit in post-exposure prophylaxis in adults and children from 1 year TAMIFLU may be used for influenza prophylaxis from 1 year after close contact with a clinically diagnosed case in the household. Prevention must begin as soon as possible, within 48 hours of the onset of symptoms. Protection only lasts for the duration of treatment. Population without risk for complications: The new data submitted are not sufficient to change the committee s previous opinion. Taking into account the fact that, in this population, influenza is usually benign and vaccine efficacy high, the actual benefit of TAMIFLU for influenza prophylaxis in subjects without comorbid diseases is insufficient. Population at high risk for complications: The new data submitted are not sufficient to change the committee s previous opinion. Taking into account: - The risk of serious complications in these populations, - The lower efficacy of the vaccine in adults aged over 65 years, the actual benefit of TAMIFLU for post-exposure prophylaxis in high risk populations is low. In addition, the actual benefit of TAMIFLU in post-exposure prophylaxis is moderate for subjects at risk, in the following specific cases: - Subjects living in institutions (institutionalised patients) - Subjects with a contraindication to the vaccine. - Immunocompromised subjects (in particular AIDS patients, transplant recipients or patients receiving immunosuppressive drugs) - Situations in which the vaccine only provides partial protection from the circulating strain. Taking into account the risk of potentially serious complications in these populations, the committee considers that TAMIFLU may have an advantage in these subjects Therapeutic use Because of epidemiological evidence suggesting a reduction in influenza complications, hospital admissions and deaths of vaccinated subjects compared to unvaccinated subjects, influenza vaccination is the reference strategy for the management of influenza to protect in high-risk groups. In patients with a flu-like syndrome, the reference symptomatic treatment is nonspecific and is based on a combination of antipyretics and analgesics. The place of influenza antiviral agents (oseltamivir and zanamivir) for symptomatic treatment of influenza in a normal epidemic situation is limited. 9

10 During an epidemic, influenza antiviral agents (oseltamivir and zanamivir) may be used after contact with a subject with flu-like syndrome. Treatment must be initiated as early as possible after exposure and not later than 48 after the first symptoms in subjects with influenza symptoms. The protective effect lasts for the duration of treatment. Prophylactic treatment is short lasting (maximum 6 weeks for oseltamivir and 4 weeks for zanamivir). Late vaccination is therefore still recommended during an epidemic. In this situation, following contact with a subject with flu symptoms, the prophylactic use of zanamivir or oseltamivir is recommended in particular in subjects at high risk for complications who are only partially protected or not protected by the vaccine: - Subjects aged over 65 years, - Subjects vaccinated for less than 15 days, - Subjects in whom vaccination is contra-indicated, - Mismatch between the vaccine strain and the circulating viral strain. The following recommendations should be respected when an epidemic occurs in institutions including healthcare facilities caring for high-risk subjects 4 while influenza virus is circulating in the community and has been documented. - Post-exposure prophylaxis should be initiated within 48 hours of contact with a person presenting influenza symptoms in all high-risk subjects aged from 1 year, whether they have been vaccinated or not, - This treatment should be prescribed for up to 7 days after the onset of symptoms in the last case. The choice of antiviral agent must take into account the pharmaceutical form: capsule or inhalation. 4 According to the opinion of French Higher Council of Public Health (CSHPF) concerning prophylaxis for persons at risk during an influenza outbreak in an institution when the virus is circulating in the community. Session of 16 January

11 APPENDIX OPINION OF THE PUBLIC HEALTH BENEFIT GROUP ON THE FINAL RESULTS (29/03/2007) OF THE POST-LISTING TAMIFLU STUDY (COSMOS Study) PROTOCOL: COSMOS Study: Impact of post-exposure oseltamivir (TAMIFLU) prophylaxis of influenza on the mortality and morbidity in institutionalised elderly subjects VERSION: Final report of March 29 th, 2007 PRODUCT: TAMIFLU COMPANY: Roche DATE OF OPINION: ISPm Group of 15/05/2007 I. REMARKS ON THE METHODOLOGY USED On 12/12/2006, the company provided a final report of the post-registration study of the proprietary medicine TAMIFLU (Cosmos Study), during the epidemic. This multicentre, prospective, longitudinal, observational study was requested by the Transparency Committee in its opinion of February 11 th, A protocol was validated by the Committee Public health impact Working Group on February 8 th, The protocol planned the participation of 120 physicians in 120 different old people s care facilities (EHPA) with inclusion of a total of 6500 residents. 98 EHPA finally took part and included 8042 patients. Follow-up data were analysed for 6989 residents in 86 EHPA of these patients were exposed to influenza. Only 270 residents were treated by TAMIFLU: 164 for curative treatment and 106 for prophylactic treatment (versus 2000 planned in the protocol). The ISP Group examined a first report, which did not correspond to the analysis plan in the protocol validated by the ISP group on 28/12/2004. The company was therefore asked to provide additional data and it sent a new version of the report of results on 05/04/2007. II. RESULTS PRESENTED MAIN POINTS The results presented showed that there was no statistically significant difference between residents who received TAMIFLU and those who did not receive it either for the primary endpoint (mortality) or the secondary endpoint (hospitalisations), or the composite endpoint of death + hospitalisations. Aboutmortality, 13 deaths were observed among the 270 residents who received TAMIFLU (i.e. 4.8%) versus 247 deaths in 4206 residents who did not receive TAMIFLU (5.9%). The relative risk is 0.82 [0.476; 1.413]. Predictors of all-cause death in the population of patients exposed to influenza were the patient's age, GIR geriatric autonomy score and the presence of respiratory disease; exposure to TAMIFLU during the epidemic period did not appear to be an explanatory factor (OR = 0.71 [0.39; 1.28], p = ). There was no significant difference for the secondary endpoint (hospitalisations). About all-cause hospitalisation, 8.9% of TAMIFLU-treated patients were admitted to hospital versus 9.3% of patients who did not receive TAMIFLU. The relative risk (RR) was [0.647; 1.421]. About the composite endpoint (death + hospitalisation), the RR of exposure to TAMIFLU was [0.644; 1.257]. In the population exposed to influenza, the proportion of residents who died or were admitted to hospital was 11.9% for those who received TAMIFLU versus 13.2% for those who did not; this difference was not statistically significant (p = 0.5). Regarding the conditions of use of TAMIFLU, the only variable studied was the duration of treatment, which, on average, complied with recommendations. 11

12 Conclusion: taking into account the very small number of TAMIFLU prescriptions and the results not demonstrating effect on morbidity and mortality, the TAMIFLU study report provides no evidence for a public health benefit in a population of residents of EHPA old people's homes. DETAILED REPORT 1. CHARACTERISTICS OF EHPA The EHPA (old peoples homes) taking part in the study had a similar distribution between regions as all French EHPA (The Ile de France, PACA and Rhone-Alps regions were the best represented). The 98 EHPA taking part in the study declared a mean number of 83 residents, which was close to the mean value of 79 places declared in the FINESS 2004 file. In the 98 EHPA, the mean proportion of deaths was 22% in 2001, 23% in 2002 and 25% in The influenza vaccination rate of nursing and boarding staff was 35% for the winter period. 2. CHARACTERISTICS OF RESIDENTS The 8041 residents included in the study were 85 years old, 76.2% were women and 81% had health insurance. The mean autonomy score of residents was 3, evaluated using the Iso-resources Group scale (GIR). 83% of residents had at least one long-term condition (ALD). This was usually psychosis or a serious personality disorder (21% of residents). The main comorbidities reported in the study population were cardiovascular (73%), osteoarticular (56%), neurodegenerative (54%), psychiatric (46%) and metabolic disorders (38%). The most frequent underlying disorders were high blood pressure (50%), osteoarthritis (41%) and Alzheimer's disease (26%). During the season, 93% of residents were vaccinated against influenza. Residents exposed to TAMIFLU were more frequently vaccinated than the unexposed population (97% versus 93%, p=0.01) and more often had a long-term condition (91.5% versus 84.3%, p<0.001). On the contrary, the proportion of vaccinated residents with and without influenza was the same. 3. FLU-LIKE SYNDROMES AND TAMIFLU TREATMENT DURING THE EPIDEMIC PERIOD During the epidemic period, 53 EHPA (54%) reported at least one flu-like syndrome, i.e. a total of 600 residents (9% of the total follow-up population). Among residents not treated by TAMIFLU, 10% (422/4206) presented at least one flu-like syndrome versus 15% in residents receiving TAMIFLU prophylaxis (16/106); this difference was not statistically significant. In most cases, physicians qualified the influenza episode as mild (40%) or moderately (40%) severe. A virological test was only carried out in exceptional cases (3.5% of residents with flu-like syndrome, i.e. 21 tests) and 43% (n=9) were found to be positive. Among the 270 patients who received TAMIFLU (in 23 EHPA, i.e. 24% of the study EHPA), 106 received preventive and 164 curative treatment. The mean length of TAMIFLU treatment was 8.5 days and 5.8 days and this treatment complied with recommendations for 87% and 95% of residents receiving curative and preventive treatment respectively. 4. DEATHS DURING THE EPIDEMIC PERIOD The 367 deaths listed during the community outbreak concerned 5.3% of the study follow-up population. The main causes of death were cardiac (142 deaths, 38.7% of cases), respiratory (83 deaths, 22.6%), neuropsychological (51 deaths, 13.9%), neoplastic (34 deaths, 9.3%) and metabolic disorders (32 deaths, 8.7%). Residents exposed to influenza had a relative risk of all-cause death of [1.095; 1.700] compared to those not exposed to influenza. About deaths due to infection, exposure to influenza led to an excess mortality in EHPA residents (RR = [1.473; 4.333]). The analysis did not show a statistically significant difference between residents who received TAMIFLU and those who did not (RR = [0.337 ; 2.493]). Exposure to influenza was also found to be a factor of excess mortality for deaths due to chronic decompensation (RR = [1.097; 1.876]). 12

13 There was no statistically significant difference between subjects receiving or not receiving TAMIFLU for all-cause death (RR = [0.476; 1.413]), deaths due to infection (RR = [0.337; 2.493]) or decompensation of an underlying disease (RR = [0.683; 2.053]) or cause-specific death. The following predictors (of all-cause death) were identified in the total study population: elderly age, high level of dependency, existence of a cardiovascular, onco-hematological or respiratory disease. Exposure to influenza was only identified as an explanatory factor for death due to infection or respiratory disease. In addition, there was no statistically significant difference in terms of death in residents who received TAMIFLU compared to those who did not receive it (RR = [0.476 ; 1.413]). 5. HOSPITALISATIONS DURING THE EPIDEMIC PERIOD A total of 601 residents were admitted to hospital at least once (i.e. 8.6% of the study population and 670 hospitalisations) during the epidemic period. Residents exposed to influenza had a relative risk of all-cause hospitalisation of [1.053; 1.467] and a relative risk of [1.016; 1.955] compared to those not exposed to influenza. Overall, 304 hospitalisations for decompensation of a chronic disease (51% of analysed hospitalisations) and 172 hospitalisations for an infection (29% of hospitalisations) were reported. The four main causes of admission to hospital were cardiovascular (n=161, 27%), respiratory (n=135, 23%), orthopedic-traumatological (n=124, 21%) and neuropsychological disorders (n=89, 15%). The following predictors of hospital admission were identified in the total study population: elderly age, high level of dependency, existence of a cardiovascular, onco-hematological or respiratory disease. Exposure to influenza or TAMIFLU were not found to be explanatory factors for hospitalisation. 6. DEATHS AND/OR HOSPITALISATIONS DURING THE EPIDEMIC PERIOD Analysis of the composite endpoint (death and/or hospitalisation) showed that exposure to influenza was an explanatory factor (RR = [1.115 ; 1.470]). There was no statistically significant difference according to exposure to TAMIFLU or not (RR = [0.644 ; 1.257]). 7. DISEASE EVENTS DURING THE EPIDEMIC PERIOD A total of 1751 infuenza-related disease events were reported, concerning 21% of EHPA residents. These were usually respiratory events (affecting 67.3% of residents), loss of autonomy (35.7%), mental confusion (20.1%), cardiovascular events (16.8%) and dehydration (15.4%). 5 In the case of exposure to influenza, the residents more frequently suffered respiratory events (16.2% vs 12.2%, p<0.001), mental confusion (4.9% vs 3.7%, p=0.01) and dehydration (3.9% vs 2.7%, p=0.01). The following events were more frequent in TAMIFLU-treated patients: worsening or onset of mental confusion (9% vs 5%, p = 0.01), dehydration (8% vs 4%, p = 0.002) and loss of autonomy (13% vs 7%, p = 0.003) CONCLUSION Taking into account the very small number of TAMIFLU prescriptions and the results showing no effect on morbidity and mortality, the TAMIFLU study report provides no evidence for a public health benefit in a population of residents of EHPA old people's homes. 5 One resident may have presented several pathological events. 6 Residents exposed totamiflu more often presented a long-term condition (91.5% versus 84.3%, p<0.001) than other residents, but the report includes no analysis taking this confounding factor into account. 13

14 On the contrary, this study confirmed the impact of influenza in terms of morbidity and mortality among institutionalised elderly people. It also described the characteristics of residents treated by TAMIFLU and showed that the duration of TAMIFLU treatment (for both post-exposure prophylaxis and curative treatment) complied with the official recommendations. 14