|

 

Chestioumal.chestpubs.org

Original Research Timing of Oseltamivir Administration
and Outcomes in Hospitalized Adults
With Pandemic 2009 Infl uenza A(H1N1)
Virus Infection
Diego Viasus , MD ; José Ramón Paño-Pardo , MD , PhD ; Jerónimo Pachón , MD , PhD ; Melchor Riera , MD , PhD ; Francisco López-Medrano , MD , PhD ; Antoni Payeras , MD , PhD ; M. Carmen Fariñas , MD , PhD ; Asunción Moreno , MD , PhD ; Jesús Rodríguez-Baño , MD , PhD ; José Antonio Oteo , MD , PhD ; Lucia Ortega , MD , PhD ; Julián Torre-Cisneros , MD , PhD ; Ferrán Segura , MD , PhD ; and Jordi Carratalà , MD , PhD ; for the Novel Infl uenza A(H1N1) Study Group of the Spanish Network for Research in Infectious Diseases (REIPI) * Background: Data on the clinical effectiveness of oseltamivir in patients with pandemic 2009
infl uenza A(H1N1) (A[H1N1]) virus infection are scarce. We aimed to determine the effect of tim-
ing of oseltamivir administration on outcomes in hospitalized adults with A(H1N1).
Methods: Observational analysis of a prospective cohort of adults hospitalized with laboratory-
confi rmed A(H1N1) was performed at 13 Spanish hospitals. Time from onset of symptoms to
oseltamivir administration was the independent variable. Outcomes were duration of fever, hos-
pital length of stay (LOS), need for mechanical ventilation, and mortality during hospitalization.
Multivariate logistic regression was used to describe the association between the independent
variable and the outcomes.
Results: Five hundred thirty-eight hospitalized patients with A(H1N1) were studied. The median
time from onset of symptoms to oseltamivir administration was 3 days (interquartile range [IQR],
2-5 days). With regard to outcomes, the median duration of fever was 2 days (IQR, 1-3 days), the
median LOS was 5 days (IQR, 3-8 days), 49 patients (9.1%) underwent mechanical ventilation,
and 11 patients (2%) died during hospitalization. In univariate analysis, prolonged duration of
fever (above the median), prolonged LOS (above the median), need for mechanical ventilation,
and mortality all increased with time to oseltamivir administration ( x 2 test for trend P 5 .001,
P ⱕ .001, P 5 .008, and P 5 .001, respectively). After adjustment for confounding factors, time
from onset of symptoms to oseltamivir administration ( 1 1-day increase) was associated with a
prolonged duration of fever (OR, 1.10; 95% CI, 1.02-1.19), prolonged LOS (OR, 1.07; 95% CI,
1.00-1.15), and higher mortality (OR, 1.20; 95% CI, 1.06-1.35).
Conclusions: Timely oseltamivir administration has a benefi cial effect on outcomes in hospital-
ized adults with A(H1N1), even in those who are admitted beyond 48 h after onset of symptoms.

CHEST 2011; 140(4):1025–1032
Abbreviations: A(H1N1) 5 pandemic 2009 infl uenza A(H1N1); IQR 5 interquartile range; LOS 5 length of stay;
RT-PCR 5 reverse-transcription polymerase chain reaction
Pandemic 2009 infl uenza A(H1N1) (A[H1N1]) virus severe A(H1N1) virus infection also has been reported
emerged in Mexico during the spring of 2009 in individuals who are obese (particularly in those and spread rapidly worldwide, 1 resulting in the fi rst with morbid obesity) 3 and among certain indigenous infl uenza pandemic of the current century. The major- populations. 4,5 Conversely, it has been suggested that ity of patients with A(H1N1) virus infection have early antiviral therapy is associated with a lower risk self-limited mild to moderate uncomplicated disease. of complications. 6-9 Risk factors for severe A(H1N1) are similar to those Because no effective vaccine was available dur- identifi ed for seasonal infl uenza. 2 A higher risk for ing the fi rst months of the pandemic, antiviral drugs CHEST / 140 / 4 / OCTOBER, 2011 1025
Downloaded From: http://publications.chestnet.org/ on 10/08/2016
(oseltamivir and zanamivir) were the primary weap- on outcomes in hospitalized adults with laboratory- ons for improving the clinical outcomes of patients confi rmed A(H1N1) virus infection. with A(H1N1) virus infection. The guidelines recom-mend that healthy patients with uncomplicated ill-ness should not be treated with antivirals, but in Materials and Methods patients with initially severe presentations and those who have risk factors for complications or require Study Design and Study Population hospital admission, treatment with antiviral drugs This observational analysis was of a prospective cohort of should commence as soon as possible. 2 adults aged . 15 years hospitalized with confi rmed A(H1N1) However, most available data on the use of antivi- virus infection at 13 Spanish university hospitals between June 12, ral drugs come from randomized trials in healthy 2009, and November 10, 2009. All patients had an infl uenza-like illness with laboratory-confi rmed A(H1N1) virus infection by real- adult patients and outpatients with seasonal infl uenza time reverse-transcription polymerase chain reaction (RT-PCR) or treated within 48 h of symptom onset. 10-12 In addi- viral culture. 16 Cases were identifi ed at the ED by attending physi- tion, two meta-analyses 10,11 recently showed that data cians or investigators when the microbiology results were available on the effectiveness of neuraminidase inhibitors for or by the daily reviewing by investigators of the microbiology- decreasing complications and mortality in patients positive results of the RT-PCR for typing (A/B) and subtyping ( 1 H/ 1 H 2009/H3/H5) infl uenza virus. In these latter cases, the with seasonal infl uenza are unclear. Notably, ran- RT-PCR results were typically available in , 48 h after hospital domized trials of neuraminidase inhibitor treatment admission. Patients with missing data on time from symptom of hospitalized patients with infl uenza are limited. 13 onset to oseltamivir administration were excluded. The study was Additionally, information about antiviral treatment approved by the Institutional Review Board (number PR182/09) and outcomes in hospitalized patients with A(H1N1) of the coordinating center (Hospital Universitari de Bellvitge; Barcelona, Spain), and informed consent was obtained from patients. virus infection are scarce. 6-9,14 Consequently, there is a need for additional clinical, virologic, and timing Clinical Assessment and Follow-up of antiviral administration studies to assess neuramin-idase inhibitor effectiveness for hospitalized patients Hospital and ICU admission criteria and treatment decisions with A(H1N1). 15 The present study aimed to deter- were not standardized and were applied or made by attending physicians. Patients were seen during their hospital stay by mine the effect of timing of oseltamivir administration one or more of the investigators at each participating hospital, who recorded clinical data in a standardized, computer-assisted Manuscript received October 28, 2010; revision accepted March 1, protocol. Data were collected on demographic characteristics, comorbidities, BMI, clinical signs and symptoms, biochemical Affi liations: From the Departments of Infectious Diseases from
analysis, chest radiograph fi ndings, antiviral and antibacterial the Hospital Universitari de Bellvitge-IDIBELL, University of therapy, concomitant and secondary bacterial pneumonia or Barcelona (Drs Viasus and Carratalà), Barcelona; Hospital Universitario La Paz-IDIPAZ (Dr Paño-Pardo), Madrid; Hos- infection, duration of fever during hospitalization, complications, pital Universitario Virgen del Rocío (Dr Pachón), Sevilla; and in-hospital mortality. For time from onset of symptoms to Hospital Universitario Son Dureta (Dr Riera), Palma de Mallorca; antiviral administration, the day of onset of symptoms was con- Hospital Universitario 12 de Octubre (Dr López-Medrano), Madrid; Hospital Son Llàtzer (Dr Payeras), Palma de Mallorca; Hospital Universitario Marqués de Valdecilla (Dr Fariñas), Study Variables and Defi nitions Santander; Hospital Universitario Clinic (Dr Moreno), Barcelona; Hospital Universitario Virgen Macarena (Dr Rodríguez-Baño), Time from onset of symptoms to oseltamivir administration Sevilla; Hospital San Pedro-CIBIR (Dr Oteo), Logroño; was the independent variable. The time interval between the SCIAS-Hospital de Barcelona (Dr Ortega), Barcelona; Hospital Universitario Reina Sofía-IMIBIC, University of Córdoba, fever or malaise onset (as reported by the patient) and the oselta- (Dr Torre-Cisneros) Córdoba; and Hospital Parc Tauli (Dr Segura), mivir administration was calculated for each patient. The out- Sabadell, Spain. comes of interest were duration of fever, hospital length of stay * A list of the study group members is available in e-Appendix 1. (LOS), need for mechanical ventilation, and mortality during Funding/Support: This study was supported by Ministerio
hospitalization. Fever was defi ned as axillary temperature . 37.8°C de Ciencia e Innovación, Instituto de Salud Carlos III, Programa and LOS as discharge date minus admission date. Mechanical de Investigación sobre gripe A/H1N1 (grant GR09/0014), and ventilation was defi ned as need for endotracheal intubation or Ministerio de Ciencia e Innovación, Instituto de Salud Carlos III. noninvasive mechanical ventilation and mortality as death from It was cofi nanced by the European Regional Development Fund any cause during hospitalization. "A way to achieve Europe," Spanish Network for Research in Infectious Diseases (REIPI RD06/0008). Dr Viasus is the Underlying medical conditions were assessed according to recipient of a research grant from the Institut d´Investigació the Charlson Comorbidity Index. 17 Other comorbidities, such as Biomèdica de Bellvitge (IDIBELL ). immunosuppression, neuromuscular disorders, and sickle cell Correspondence to: Jordi Carratalà, MD, PhD, Department of
disease, also were recorded. Morbid obesity was defi ned as a Infectious Diseases, Hospital Universitari de Bellvitge, Feixa BMI ⱖ 40 kg/m 2 or a subjective assessment by the physician if Llarga s/n, 08907, L'Hospitalet de Llobregat, Barcelona, Spain; weight and height data were not available. e-mail: [email protected] Because hospital criteria decisions were not standardized in 2011 American College of Chest Physicians. Reproduction
the present study, it is likely that factors other than disease sever- of this article is prohibited without written permission from the ity may have contributed to site-of-care decisions. Therefore, we defi ned patients with progressive, severe, or complicated illness DOI: 10.1378/chest.10-2792
at hospital admission when any of the following were present on Original Research Downloaded From: http://publications.chestnet.org/ on 10/08/2016
admission 2 : (1) signs and symptoms of lower respiratory tract dis- Comorbidities were present in 304 patients (56.5%) ease (wheezing, intercostal retractions, tachypnea, pneumonia on as mainly chronic pulmonary diseases (29.7%), immu- chest radiographs, Pa o /F io , 300, or arterial saturation , 90%), nossuppression (14.5%), diabetes mellitus (11%), (2) altered mental status, (3) hypotension (systolic BP ⱕ 90 mm Hg), or (4) bacterial coinfection based in laboratory testing. and chronic heart diseases (7.1%). Furthermore, 170 patients (31.6%) were current smokers, and Statistical Analyses 27 (5%) were morbidly obese. Seventy-seven patients (14.3%) were pregnant women. No patient was vac- All proportions were calculated as percentages of the patients cinated for A(H1N1) or received oseltamivir prior to with available data. For continuous variables, the median and inter-quartile range (IQR) are reported. Crude outcomes for time from hospitalization. Pneumonia on chest radiograph was onset of symptoms to oseltamivir administration were reported by found in 224 patients (41.6%) and bacterial coinfec- time-to-treatment groups (group 1, ⱕ 2 days; group 2, 3-4 days; tion in 36 (16%) of them. Regarding treatment, all group 3, 5-6 days; and group 4, ⱖ 7 days). Linear trend analysis patients received oseltamivir (35 patients at a dose was used to account for multiple comparisons of crude outcomes of 150 mg bid), and one patient also received inhaled (univariate analysis). Multivariate logistic regression was used to describe the association between time from onset of symptoms zanamivir; antibacterial treatment was administered ( 1 1-day increase) to oseltamivir administration and each of the to 396 patients (73.6%) (all patients with bacterial clinical outcomes. Signifi cant variables detected in the univariate coinfection received antibacterial treatment) and analysis were entered into the multivariate analysis. Variables corticosteroids to 152 (28.3%). The median duration entered in the univariate analysis were age, male sex, comor-bidities, prior seasonal infl uenza and pneumococcal vaccination, of oseltamivir therapy was 5 days (IQR, 5-5 days). No obesity, morbid obesity, hypotension, tachypnea (respiratory fre- patient received IV peramivir or zanamivir. quency ⱖ 24/min), altered mental status, pleural effusion, pneumo-nia on chest radiographs, bacterial coinfection, and ICU admission. Effect on Outcomes of Time From Onset of Results derived from the logistic regression models were expressed as OR and 95% CI. An a priori subgroup analysis was performed Symptoms to Oseltamivir Administration in patients with progressive, severe, or complicated illness at hos-pital admission. The results were analyzed using SPSS version The median time from onset of symptoms to oselta- 15.0 (SPSS Inc; Chicago, Illinois) statistical software. Statistical mivir administration was 3 days (IQR, 2-5 days). Of signifi cance was established at a 5 .05. All reported P values are the 538 patients, 202 (37.5%) were included in time-to- treatment group 1, 160 (29.7%) in group 2, 87 (16.2%) in group 3, and 89 (16.5%) in group 4. Regarding out-comes, the median duration of fever during hospital- ization was 2 days (IQR, 1-3 days), the median LOS Patient Characteristics was 5 days (IQR, 3-8 days), 49 (9.1%) patients under- Of the 585 hospitalized patients with laboratory- went mechanical ventilation, and 11 (2%) died dur- confi rmed A(H1N1), 47 with missing data on time ing hospitalization. The median time from hospital from symptom onset to oseltamivir administration were admission to death was 5 days (IQR, 3.5-19.5 days). excluded. Excluded patients did not present signifi - Causes of death were respiratory failure or ARDS cant differences with regard to age, signs and symp- (three of 11 patients), shock or multiorgan failure toms, and outcomes compared with patients included (four patients), decompensated comorbid condition in the study. Among the 538 patients analyzed in the (three patients), and nosocomial infection (one patient). present study, the median time from onset of symp- Crude outcomes stratifi ed by time-to-treatment toms to hospital arrival was 3 days (IQR, 2-5 days). groups are shown in Table 1 . In univariate analysis, The median age was 39 years (IQR, 28-52 years). Just duration of fever (above the median . 2 days) during more than one-half (52.4%) of the patients were men. hospitalization, LOS (above the median . 5 days), need Table 1— Crude Outcomes Stratifi ed by Groups of Time From Onset of Symptoms to Oseltamivir Administration
in All Hospitalized Patients With A(H1N1): Univariate Analysis
Time to Antiviral Need for Mechanical Group 1 ( ⱕ 2 d) Group 4 ( ⱖ 7 d) Data are presented as median (IQR) or No. (%). A(H1N1) 5 pandemic 2009 infl uenza A(H1N1); IQR 5 interquartile range; LOS 5 length of stay.
a x 2 test for trend P 5 .001.
b x 2 test for trend P ⱕ .001.
c x 2 test for trend P 5 .008.
d x 2 test for trend P ⱕ .001.
CHEST / 140 / 4 / OCTOBER, 2011 1027
Downloaded From: http://publications.chestnet.org/ on 10/08/2016
Table 2— Factors Associated With Prolonged Duration
( x 2 test for trend P 5 .003, P 5 .005, and P 5 .006, of Fever (Above the Median of 2 d) in Hospitalized
respectively). No association was found between tim- Patients With A(H1N1): Multivariate Analysis
ing of oseltamivir administration and need for mechan- ical ventilation in univariate analysis in this subgroup of patients. In the multivariate logistic regression analysis, time from onset of symptoms to oseltamivir adminis- tration ( 1 1-day increase) was associated with pro- Bacterial coinfection longed duration of fever (OR, 1.11; 95% CI, 1.01-1.20), Time from onset of symptoms prolonged LOS (OR, 1.10; 95% CI, 1.01-1.20), and to oseltamivir administration higher mortality (OR, 1.19; 95% CI, 1.05-1.36). How- ( 1 1-d increase) ever, time from onset of symptoms to oseltamivir See Table 1 legend for expansion of abbreviation.
administration also was not associated with the need for mechanical ventilation in the multivariate analysis for mechanical ventilation, and mortality increased (OR, 1.04; 95% CI, 0.94-1.14). with time to oseltamivir administration ( x 2 test for trend P 5 .001, P ⱕ .001, P 5 .008, and P ⱕ .001, Effect on Outcomes of Time From Arrival to respectively). After adjustment for confounding the Hospital to Oseltamivir Administration factors, time from onset of symptoms to oseltamivir There was concern that previous results may refl ect administration ( 1 1-day increase) was associated with differences in the time from onset of symptoms until prolonged duration of fever ( Table 2 ), prolonged hospitalization. Therefore, a post hoc analysis was LOS ( Table 3 ), and higher mortality during hospital- performed to evaluate the effect on outcomes of ization ( Table 4 ). However, time from onset of symp- delay of oseltamivir administration after arrival to the toms to oseltamivir administration ( 1 1-day increase) hospital ( Table 7 ). Of the 538 hospitalized patients, was not related to the need for mechanical ventila- 127 received oseltamivir treatment after the fi rst tion ( Table 5 ). 24 h after arrival. In these patients, the RT-PCR results were more frequently available after the fi rst Patients With Progressive, Severe, day since the arrival to the hospital (47.2% vs 25.1%; or Complicated Illness P ⱕ .001). Compared with patients who received osel- Three hundred seventy-seven patients had at least tamivir within the fi rst 24 h, the delay of oseltamivir one criterion of progressive, severe, or complicated administration ( . 24 h) was independently associated illness at hospital admission (signs and symptoms of with prolonged duration of fever (adjusted OR, 1.67; lower respiratory tract disease in 371 patients, altered 95% CI, 1.03-2.72), prolonged LOS (adjusted OR, 1.67; mental status in 13, hypotension in 15, and bacte- 95% CI, 1.06-2.63), more-often need for mechanical rial coinfection in 44). Crude outcomes stratifi ed ventilation (adjusted OR, 3.13; 95% CI, 1.56-6.27), by time-to-treatment groups in hospitalized patients and higher mortality (adjusted OR, 4.29; 95% CI, with progressive, severe, or complicated illness are shown in Table 6 . In the univariate analysis, duration of fever (above the median . 2 days), LOS (above the median . 6 days), and mortality during hospitaliza-tion increased with time to oseltamivir administration In this large, prospective cohort study of hospital- ized patients with laboratory-confi rmed A(H1N1) Table 3— Factors Associated With a Prolonged
virus infection, time from onset of symptoms to osel- Length of Hospital Stay (Above the Median of 5 d)
tamivir administration ( 1 1-day increase) was inde- in Hospitalized Patients With A(H1N1):
pendently associated with a prolonged duration of Multivariate Analysis
the fever, prolonged LOS, and higher mortality dur- ing hospitalization. Moreover, we performed a sub- group analysis in patients with progressive, severe, or complicated illness at hospital admission because hospital criteria decisions were not standardized in the present study and because it was likely that mild Bacterial coinfection cases had been admitted due to the uncertainty sur- Time from onset of symptoms to rounding disease progression and prognosis during the initial period of the pandemic. In fact, we found ( 1 1-d increase) that  30% of patients did not have criteria for pro- See Table 1 legend for expansion of abbreviation.
gressive, severe, or complicated illness at hospital Original Research Downloaded From: http://publications.chestnet.org/ on 10/08/2016
Table 4— Factors Associated With Mortality in
including those requiring hospital admission for infl u- Hospitalized Patients With A(H1N1):
enza or its complications. 12 Multivariate Analysis
The present data concur with those from previous observational studies of hospitalized patients with A(H1N1) virus infection, which suggests that early antiviral treatment ( , 48-72 h) may reduce disease Time from onset of symptoms to severity and mortality. However, to our knowledge, this study is the fi rst to address the effect of each ( 1 1-d increase) day of delay in the administration of oseltamivir after See Table 1 legend for expansion of abbreviation.
symptom onset and after arrival to the hospital on clinical outcomes in hospitalized adults. In a study in the United States carried out between April 2009 admission. Similar benefi cial results of timely oselta- and June 2009, early antiviral treatment ( ⱕ 2 days) mivir administration were obtained in these patients was signifi cantly associated with a decreased risk of with progressive, severe, or complicated illness at ICU admission or mortality. 6 Similarly, in a previous hospital admission. study, we reported that oseltamivir treatment within In addition, in a further analysis, a delay ( . 24 h) in 72 h was independently associated with a lower the administration of oseltamivir after arrival to the risk of severe disease. 9 In addition, among pregnant hospital was also associated with higher morbidity women, those who received antiviral treatment . 48 h and mortality independently of time from symptom after symptom onset were more likely to be admitted onset to hospital admission. Further, although the to an ICU, 7 and Chien et al 8 found that late oseltami- specifi c reasons for delay of oseltamivir adminis- vir treatment ( . 48 h) was associated with a higher tration were not recorded, we found that RT-PCR risk of developing respiratory failure among hospi- results were more frequently available after the fi rst talized patients with pneumonia. However, a recent day of hospitalization in these patients. Thus, labora- multicenter, retrospective study in China 14 found no tory confi rmation of A(H1N1) virus infection should association between early oseltamivir treatment and not delay the initiation of empirical treatment in improvement outcome in hospitalized patients with hospitalized patients. During the pandemic, it was recommended that The present results are also consistent with data antiviral treatment be administered to patients at a from studies that evaluated the role of neuraminidase high risk for infl uenza complications, in need of hos- inhibitors in patients with seasonal infl uenza requir- pitalization, and with severe or complicated infl uenza ing hospitalization. In a prospective cohort study, illness. 2 However, most available data on the use of McGeer et al 18 found that treatment with oseltamivir antiviral drugs come from several prospective studies (independently of time between symptom onset to in ambulatory healthy adult and pediatric patients antiviral administration) was associated with a reduc- with seasonal infl uenza treated within 48 h of symp- tion in mortality. Similarly, in a retrospective study tom onset. 10-12 Signifi cantly, few clinical data have been to examine factors associated with earlier hospital published on the clinical effectiveness of antiviral drugs discharge of adult patients with severe infl uenza ill- in hospitalized patients with seasonal infl uenza 18-20 or ness, Lee et al 19 found that oseltamivir initiated within A(H1N1). 6-9,14 Therefore, little information is cur- 2 days of onset and infl uenza vaccination within rently available to assess whether there is a benefi t 6 months were independently associated with shorter on prognosis in treating patients with severe illness, LOS. In a further report, 20 these researchers also found that oseltamivir initiated within 96 h after illness onset was independently associated with a Table 5— Factors Associated With the Need for
decreased mortality. Mechanical Ventilation in Hospitalized Patients With
A(H1N1): Multivariate Analysis
Another important fi nding in the present study was that the patients appeared to benefi t from oseltamivir therapy initiated . 48 h after symptom onset. In this regard, Lee et al 21 and Leekha et al 22 reported that patients hospitalized with severe infl uenza have more active and prolonged viral replication. Major comorbidities, immunosuppression, and systemic corti- Bacterial coinfection Time from onset of symptoms to costeroid use for asthma or COPD exacerbations have been related to slower viral clearance. In addition, ( 1 1-d increase) observational studies showed that the A(H1N1) virus can See Table 1 legend for expansion of abbreviation.
be shed for a longer period than seasonal infl uenza. 23 CHEST / 140 / 4 / OCTOBER, 2011 1029
Downloaded From: http://publications.chestnet.org/ on 10/08/2016
Table 6— Crude Outcomes Stratifi ed by Groups of Time From Onset of Symptoms to Oseltamivir Administration
in Hospitalized Patients With Progressive, Severe, or Complicated Illness: Univariate Analysis
Time to Antiviral Group 1 ( ⱕ 2 d) Group 4 ( ⱖ 7 d) Data are presented as median (IQR). See Table 1 legend for expansion of abbreviations.
a x 2 test for trend P 5 .003.
b x 2 test for trend P 5 .005.
c x 2 test for trend P 5 .07.
d x 2 test for trend P 5 .006.
Prolonged viral shedding in hospitalized patients at present that they have any effect on complica- suggests that late onset of antiviral therapy ( . 48 h) tions of the lower respiratory tract, antibiotic use, may decrease morbidity. Thus, Adisasmito et al 24 or admissions to the hospital. Nevertheless, most reported that patients with A(H5N1) virus infection of the trials studied in these meta-analyses evalu- presented a survival benefi t when prescribed antiviral ated only healthy adult patients and did not include treatment, preferably within 2 days of symptom onset, hospitalized patients. Notably, the only published although the benefi t remained even with treatment randomized placebo-controlled study regarding the initiation up to 5 to 6 days after onset. Moreover, use of antivirals (zanamivir) in hospitalized patients oseltamivir treatment has been associated with an with seasonal infl uenza was terminated early and accelerated decrease in viral load. 21,25,26 Similarly, viral included a small number of patients. Thus, the RNA clearance has been related to a shorter hospital authors pointed out that further investigations are needed to assess the effi cacy of antiviral drugs in this Signifi cantly, however, two recent meta-analyses 10,11 reported that the effectiveness of antiviral treatment The strengths of the present study are its pro- in preventing serious complications and mortality in spective and multicenter design, the large number of patients with seasonal infl uenza is still unsatisfactory. patients included, and the comprehensive clinical Falagas et al 10 reported that neuraminidase inhibitors data collection. In addition, we adjusted for con- are effective in reducing infl uenza complications, founding factors to determine the relation between such as acute otitis media or pharyngitis but not timing of oseltamivir administration and outcomes. pneumonia. In addition, Jefferson et al 11 found that However, our study has several limitations that should neuraminidase inhibitors are effective in reducing be acknowledged. It was not a randomized trial and, the symptoms of infl uenza, but there is little evidence as is the case with any observational study, there is Table 7— Effect on Outcomes of Delay of Oseltamivir Administration After Arrival to the Hospital
in All Hospitalized Patients With A(H1N1)
Oseltamivir Administration After Arrival to the Hospital, % Crude OR (95% CI) Adjusted OR (95% CI) Fever above the median (2 d) 1.67 (1.03-2.72) a LOS above the median (5 d) 1.67 (1.06-2.63) b Need for mechanical 3.13 (1.56-6.27) c ventilationMortality 4.29 (1.25-14.6) d See Table 1 legend for expansion of abbreviations.
a The multivariate logistic regression included timing of oseltamivir administration after arrival to the hospital, time from symptom onset to hospital admission, age ( , 50 y), pleural effusion, pneumonia, and bacterial coinfection.
b The multivariate logistic regression included timing of oseltamivir administration after arrival to the hospital, time from symptom onset to hospital admission, age ( , 50 y), comorbidities, pleural effusion, pneumonia, bacterial coinfection, and ICU admission.
c The multivariate logistic regression included timing of oseltamivir administration after arrival to the hospital, time from symptom onset to hospital admission, comorbidities, morbid obesity, pleural effusion, pneumonia, and bacterial coinfection.
d The multivariate logistic regression included timing of oseltamivir administration after arrival to the hospital, age ( , 50 y), and comorbidities.
Original Research Downloaded From: http://publications.chestnet.org/ on 10/08/2016
potential for residual confounding. However, a ran- domized trial that intentionally delays treatment 1 . Dawood FS , Jain S , Finelli L , et al ; Novel Swine-Origin would present signifi cant ethical challenges. In addi- Infl uenza A (H1N1) Virus Investigation Team . Emergence of a novel swine-origin infl uenza A (H1N1) virus in humans . tion, because of the small sample size of patients N Engl J Med . 2009 ; 360 ( 25 ): 2605 - 2615 . who died in the present study, our data for mortality 2 . Centers for Disease Control and Prevention . Updated interim should be interpreted with caution. Finally, we did recommendations for the use of antiviral medications in not evaluate the response of viral shedding to oselta- the treatment and prevention of infl uenza for the 2009-2010 mivir treatment and its relation with outcomes. season. Centers for Disease Control and Prevention Web site. Accessed October 5, 2010. 3 . Morgan OW , Bramley A , Fowlkes A , et al . Morbid obesity as a risk factor for hospitalization and death due to 2009 pandemic infl uenza A(H1N1) disease . PLoS ONE . 2010 ; 5 ( 3 ): e9694 . The results of this study suggest that timely oselta- mivir administration has a benefi cial impact on dura- tion of fever, LOS, need for mechanical ventilation, 4 . La Ruche G , Tarantola A , Barboza P , Vaillant L , Gueguen J , Gastellu-Etchegorry M ; Epidemic Intelligence Team at InVS . and mortality in hospitalized patients with A(H1N1) The 2009 pandemic H1N1 infl uenza and indigenous pop- virus infection. Consequently, treatment with oselta- ulations of the Americas and the Pacific . Euro Surveill . mivir should be administered as soon as possible to 2009 ; 14 ( 42 ). pii: 19366. improve the clinical outcomes, even in patients who seek medical advice at a relatively late stage in the 5 . Centers for Disease Control and Prevention (CDC) . Deaths course of the illness. related to 2009 pandemic infl uenza A (H1N1) among American Indian/Alaska Natives-12 states, 2009 . MMWR Morb Mortal Wkly Rep . 2009 ; 58 ( 48 ): 1341 - 1344 . Acknowledgments 6 . Jain S , Kamimoto L , Bramley AM , et al ; 2009 Pandemic Infl uenza A (H1N1) Virus Hospitalizations Investigation Author contributions: Drs Viasus and Carratalà had full access
to all of the data in the study and take responsibility for the integ-
Team . Hospitalized patients with 2009 H1N1 infl uenza in the rity of the data and the accuracy of the data analysis. United States, April-June 2009 . N Engl J Med . 2009 ; 361 ( 20 ): Dr Viasus: contributed to the protocol design, data collection and analysis, writing and fi nal approval of the manuscript. 7 . Louie JK , Acosta M , Jamieson DJ , Honein MA ; California Dr Paño-Pardo: contributed to data collection and analysis and Pandemic (H1N1) Working Group . Severe 2009 H1N1 infl u- review and fi nal approval of the manuscript. enza in pregnant and postpartum women in California . N Engl Dr Pachón: contributed to obtaining funding, data collection, and review and fi nal approval of the manuscript. 8 . Chien YS , Su CP , Tsai HT , et al . Predictors and outcomes of Dr Riera: contributed to data collection and analysis and review respiratory failure among hospitalized pneumonia patients with and fi nal approval of the manuscript. Dr López-Medrano: contributed to data collection and review and 2009 H1N1 infl uenza in Taiwan . J Infect . 2010 ; 60 ( 2 ): 168 - 174 . fi nal approval of the manuscript. 9 . Viasus D , Paño-Pardo JR , Pachón J , et al. Factors associ- Dr Payeras: contributed to data collection and analysis and review ated with severe disease in hospitalised adults with pandemic and fi nal approval of the manuscript. (H1N1) 2009 in Spain . Clin Microbiol Infect. 2010 ; 17 ( 5 ): Dr Fariñas: contributed to data collection and review and fi nal approval of the manuscript. 10 . Falagas ME , Koletsi PK , Vouloumanou EK , Rafailidis PI , Dr Moreno: contributed to data collection and review and fi nal Kapaskelis AM , Rello J . Effectiveness and safety of neuramin- approval of the manuscript. idase inhibitors in reducing influenza complications: a Dr Rodríguez-Baño: contributed to data collection and analysis and review and fi nal approval of the manuscript. meta-analysis of randomized controlled trials . J Antimicrob Dr Oteo: contributed to data collection and review and fi nal Chemother . 2010 ; 65 ( 7 ): 1330 - 1346 . approval of the manuscript. 11 . Jefferson T , Jones M , Doshi P , Del Mar C . Neuraminidase Dr Ortega: contributed to data collection and review and fi nal inhibitors for preventing and treating infl uenza in healthy approval of the manuscript. adults: systematic review and meta-analysis . BMJ . 2009 ; Dr Torre-Cisneros: contributed to data collection and review and fi nal approval of the manuscript. Accessed January 15, 2010. Dr Segura: contributed to data collection and review and fi nal 12 . Harper SA , Bradley JS , Englund JA , et al ; Expert Panel of the approval of the manuscript. Infectious Diseases Society of America . Seasonal infl uenza Dr Carratalà: contributed to obtaining funding, protocol design, analysis, writing and fi nal approval of the manuscript, and study in adults and children—diagnosis, treatment, chemoprophy- laxis, and institutional outbreak management: clinical prac- Financial/nonfi nancial disclosures: The authors have reported
tice guidelines of the Infectious Diseases Society of America . to CHEST that no potential confl icts of interest exist with any Clin Infect Dis . 2009 ; 48 ( 8 ): 1003 - 1032 . companies/organizations whose products or services may be dis- 13 . Ison MG , Gnann JW Jr , Nagy-Agren S , et al ; NIAID cussed in this article . Collaborative Antiviral Study Group . Safety and effi cacy Role of sponsors: The funding sources had no role in the study
of nebulized zanamivir in hospitalized patients with serious design, data collection, data analysis and interpretation, or writing infl uenza . Antivir Ther . 2003 ; 8 ( 3 ): 183 - 190 . of the manuscript. 14 . Cui W , Zhao H , Lu X , et al . Factors associated with death Additional information: The e-Appendix can be found in the

in hospitalized pneumonia patients with 2009 H1N1 infl u- enza in Shenyang, China . BMC Infect Dis . 2010 ; 10 : 145 . CHEST / 140 / 4 / OCTOBER, 2011 1031
Downloaded From: http://publications.chestnet.org/ on 10/08/2016
http://www.biomedcentral.com/1471-2334/10/145 . Accessed 21 . Lee N , Chan PK , Hui DS , et al . Viral loads and duration of viral shedding in adult patients hospitalized with infl uenza . 15 . Uyeki T . Antiviral treatment for patients hospitalized with J Infect Dis . 2009 ; 200 ( 4 ): 492 - 500 . 2009 pandemic infl uenza A (H1N1) . N Engl J Med . 2009 ; 22 . Leekha S , Zitterkopf NL , Espy MJ , Smith TF , Thompson RL , Sampathkumar P . Duration of infl uenza A virus shedding in 16 . World Health Organization . Interim WHO guidance for hospitalized patients and implications for infection control . the surveillance of human infection with swine infl uenza Infect Control Hosp Epidemiol . 2007 ; 28 ( 9 ): 1071 - 1076 . A(H1N1) virus, 27 April 2009. World Health Organization 23 . Yu H , Liao Q , Yuan Y , et al . Effectiveness of oseltamivir on disease progression and viral RNA shedding in patients Accessed June 1, 2009. with mild pandemic 2009 infl uenza A H1N1: opportunistic 17 . Charlson ME , Pompei P , Ales KL , MacKenzie CR . A new retrospective study of medical charts in China . BMJ . 2010 ; method of classifying prognostic comorbidity in longitu- 341 : c4779 . http://www.bmj.com/content/341/bmj.c4779 . dinal studies: development and validation . J Chronic Dis . Accessed October 10, 2010. 24 . Adisasmito W , Chan PK , Lee N , et al . Effectiveness of antivi- 18 . McGeer A , Green KA , Plevneshi A , et al ; Toronto Invasive ral treatment in human infl uenza A(H5N1) infections: anal- Bacterial Diseases Network . Antiviral therapy and outcomes ysis of a Global Patient Registry . J Infect Dis . 2010 ; 202 ( 8 ): of infl uenza requiring hospitalization in Ontario, Canada . Clin Infect Dis . 2007 ; 45 ( 12 ): 1568 - 1575 . 25 . Li IW , Hung IF , To KK , et al . The natural viral load profi le 19 . Lee N , Chan PK , Choi KW , et al . Factors associated with of patients with pandemic 2009 infl uenza A(H1N1) and the early hospital discharge of adult infl uenza patients . Antivir effect of oseltamivir treatment . Chest . 2010 ; 137 ( 4 ): 759 - 768 . Ther . 2007 ; 12 ( 4 ): 501 - 508 . 26 . Cao B , Li XW , Mao Y , et al ; National Infl uenza A Pandemic (H1N1) 2009 Clinical Investigation Group of China . Clin- Antiviral treatment for patients hospitalized with severe infl u- ical features of the initial cases of 2009 pandemic infl uenza enza infection may affect clinical outcomes . Clin Infect Dis . A (H1N1) virus infection in China . N Engl J Med . 2009 ; Original Research Downloaded From: http://publications.chestnet.org/ on 10/08/2016

Source: http://chestioumal.chestpubs.org/data/Journals/CHEST/23323/102792.pdf