CLINICAL Hospitalized Adult Patients with 2009 Pandemic Influenza A (H1N1) in Kaunas, Lithuania

Summary. The objective of this study was to identify case characteristics and clinical course of the disease in patients hospitalized with 2009 pandemic influenza A (H1N1) infection during the first wave of the pandemic and to identify risk factors associated with the complicated course of illness. Material and methods. A retrospective study of adult cases of the laboratory-confirmed 2009 pandemic influenza A (H1N1) virus admitted to three hospitals in Kaunas between November 1, 2009, and March 15, 2010, was carried out. The main outcome measures were clinical characteristics, risk factors for complicated disease, treatment, and clinical course of the disease. Results. The study enrolled 121 of the 125 patients hospitalized due to 2009 pandemic influenza A (H1N1) virus infection. The median age was 31 years (range, 18–83); 5% of the patients were aged more than 65 years. Pregnant and postpartum women comprised 26% of all hospitalized cases. Nearly half (49.5%) of those who underwent chest radiography had findings consistent with pneumonia, which was bilateral in one-third of cases. The risk to have pandemic influenza complicated by pneumonia increased significantly with one-day delay from symptom onset to antiviral treatment (OR, 2.241; 95% CI, 1.354–3.710). More than half (57%) of the patients received antiviral treatment. In 45% of the treated patients, antiviral drugs were administered within 48 hours from symptom onset. Intensive care was required in 7.4% of the cases. The overall mortality was 5% (6/121). The median age of the patients who died was 43.5 years (range, 23–62); 4 patients had been previously healthy, 1 patient suffered from chronic lympholeukemia, and 1 patient was a pregnant woman. Conclusion. The 2009 pandemic influenza A (H1N1) caused considerable morbidity in a signifi-cant proportion of hospitalized adults. The main risk factor associated with the complicated course of illness was delayed antiviral treatment.


Introduction
The novel infl uenza A (H1N1) virus was fi rst identifi ed in Mexico in April 2009. Different from the strains in the past and containing a unique combination of gene segments from swine, avian, and human lineages, this new virus appeared to be able to spread among human beings leading to infl uenzalike symptoms and progressing in a few cases to viral pneumonia, respiratory failure, and death. On June 11, 2009, the World Health Organization (WHO) raised the level of pandemic alert to phase 6, indi-cating that a global pandemic had begun (1).
In Lithuania, the fi rst case of pandemic infl uenza was diagnosed on June 26, 2009, in a traveler who had returned from India. Since the beginning of November, a sustained local spread of the pandemic infl uenza virus within the country was documented. A total of 69 000 infl uenza cases were registered during the fi rst wave of the pandemic. The laboratory-confi rmed pandemic infl uenza was reported in 810 cases resulting in 23 deaths (2).
Pandemic infl uenza seems to be more severe than the previous seasonal infl uenza A (H3N2), which has dominated since 1968 (3,4). The 2009 infl uenza A (H1N1) virus has a higher affi nity for α2, 3 sialic acid receptors in the lungs compared with A (H3N2), which explains why the risk of developing virus pneumonia is higher with the new pandemic virus (5). Hereby, we report the main clinical fi ndings and risk factors for complicated disease in adult patients hospitalized due to 2009 pandemic infl uenza A (H1N1) in Kaunas, Lithuania.

Material and Methods
The study retrospectively reviewed adult (≥18 years of age) cases of the pandemic 2009 infl uenza A (H1N1) virus admitted to three hospitals located in Kaunas ( (6). Testing was performed at the Virology Laboratory of the Center for Communicable D iseases and AIDS of Lithuania, a statewide service.
Data on demographics, clinical signs and symptoms, comorbid conditions, selected laboratory tests, radiographic fi ndings, treatment course, and outcomes were extracted from the medical records using the same standardized case report form by two investigators. Comorbidities were considered absent in cases for which records stated that a patient had been previously healthy or had no underlying medical conditions or when there was no direct reference to the condition. Body mass index (BMI) was calculated for the patients whose height and weight were available. Obesity was defi ned as BMI of ≥30 kg/m 2 . BMI was not calculated for pregnant women. Fever was defi ned as documented or self-reported temperature of ≥37.7°C.
Complicated pandemic infl uenza was defi ned in the presence of radiologically confi rmed pneumonia, septic shock, multiorgan failure, central nervous system (CNS) involvement, etc. or secondary bacterial infections of the upper respiratory tract defi ned and diagnosed by standard methods. A severe case was defi ned as requiring admission to an intensive care unit (ICU) or death.
Data on Kaunas city and district demographics were obtained from the Department of Statistics under the Government of the Republic of Lithuania (7). Statistical Analysis. Missing or unknown information was excluded from all calculations. Patients who had a noncomplicated course of infl uenza were compared with those who experienced complications. Nonsevere cases were compared with the severe ones. A comparison of pregnant women with nonpregnant women of childbearing age was also made.
Statistical analysis included descriptive statistics with frequency analysis (percentages) for categorical variables and median values with ranges and/or means with standard deviations (SDs) for continuous variables. To test for differences between the compared groups, the two-sample t or Mann-Whitney tests were used for continuous variables as appropriate, and the χ 2 test was used for categorical variables.
Multivariate logistic regression models were employed to investigate associations of risk factors with the complicated illness. Results of the logistic regression analysis are summarized by odds ratios (OR) and 95% confi dence intervals (CI) with OR greater than 1.0 signifying greater risk of complications.
Receiver operating characteristic (ROC) analysis was used to calculate cut-off points and their sensitivity as well as specifi city for age, BMI, time from symptom onset to hospitalization, and C-reactive protein (CRP) level.
P values were two-sided and considered to be statistically signifi cant if <0.05 (SPSS Statistics for Windows software, version 13.0).

Demographic Data and Clinical Characteristics.
The study involved three settings: the tertiary hospital, the university-associated city hospital with the ward specially designed for the treatment of infectious diseases patients, and the regional hospital. These hospitals admitted 124 of the 125 laboratory confi rmed pandemic infl uenza cases hospitalized in Kaunas. Hospital records of 3 eligible cases were not available for the review; therefore, these cases were not included into the study. Overall, 121 patients were examined: 77 cases (64%) were treated in Kaunas 2nd Clinical Hospital (university-associated), 37 cases (30%) in the Hospital of the Lithuanian University of Health Sciences (tertiary institution), and 7 cases (6%) in Kaunas Hospital (regional).
The median age of the study participants was 31 years (range,   Fig. 2.
Demographic data, risk conditions, and clinical symptoms are shown in Table 1.
The median time from illness onset to hospital admission was 2 days (range, 0-7).
On admission, 64 patients (53%) reported symptoms, which met the WHO clinical case defi nition of uncomplicated pandemic infl uenza (fever with cough, sore throat, rhinorrhea, headache, muscle pain, or malaise, but no shortness of breath and dyspnea), and 46 patients (38%) had symptoms compatible with the defi nition of complicated infl uenza (fever and dyspnea and/or hypoxia) or presenting secondary complications such as septic shock, multiorgan failure, CNS involvement, etc. Eleven patients (9%) could not be classifi ed into any of the two groups due to the lack of the information regarding symptoms and signs described in the defi nition.
Of the 64 patients with the symptoms of uncomplicated infl uenza on admission, 38 (59.4%) had risk factors for severe disease from pandemic A (H1N1) virus infection according to the WHO: comorbidity was observed in 15, pregnancy in 23, obesity in 6, and 1 patient was aged more than 65 years.
Diagnostic Findings. Of the 97 patients who underwent chest radiography on admission, 48 (49.5%) had fi ndings consistent with pneumonia. Radiographic fi ndings included bilateral infi ltrates in 35.4% (17/48) of the patients; the remaining had unilateral infi ltration. Chest x-ray abnormalities on admission were detected in 14 (31.8%) of the 44 patients and 30 (69.8%) of the 43 patients with the clinical symptoms compatible with uncomplicated and complicated infl uenza, respectively, corresponding to the WHO defi nition. Both sensitivity and specifi city of the defi nition of uncomplicated and complicated infl uenza were 0.7.
Of the 34 patients whose blood cultures or lower respiratory tract specimens were taken, 10 (29%) had microbiologic evidence of the secondary bacterial infection. The most common isolates were Klebsiella (n=3) and Pseudomonas aeruginosa (n=3).
The rest of the baseline laboratory fi ndings are presented in Table 1.
Treatment. The data regarding the use of antiviral drugs were available from all patients. Antiviral treatment was given to 69 patients (57%). All patients were treated with oseltamivir. Oseltamivir at a standard dosage of 75 mg twice a day was given to 66 patients. An increased dosage (150 mg twice a day) was used in 3 patients. The median time from illness onset to initiation of antiviral therapy was 3 days (range, 0-10). The mean time from symptom onset to oseltamivir therapy in patients with risk conditions was 4.0 days (SD, 2.74 days) and  in those without risk conditions was 3.14 days (SD, 2.55 days) (P=0.221). In these groups of patients, antiviral therapy was started within 48 hours in 22 (50%) and 8 (36.4%) cases, respectively (P=0.294).
Antiviral drugs were administered to 34 (71%) of the 48 patients with radiographic fi ndings consistent with pneumonia. At the same time, all patients (100%) were treated with antibiotics. Clinical Course. Complicated vs. Noncomplicated Cases. In total, complicated infl uenza was diagnosed in 52 (43%) of the 121 cases. Radiologically confi rmed pneumonia was the main complication observed in 48 (92.3%) of the 52 complicated cases. Otitis (n=1), sinusitis (n=2), and tonsillitis (n=1) were diagnosed in 7.7% of cases.
Comparison of the baseline characteristics of patients with complicated vs. noncomplicated infl uenza A (H1N1) is presented in Table 1.
Independent predictors of the complicated course of illness were assessed using the multivariate logistic regression model. Only factors that were statistically signifi cant were included in the multivariate analysis, i.e., age, sex, CRP level, and time from symptom onset to antiviral treatment. Since age and presence of comorbidities, time from symptom onset to hospitalization, and initiation of antiviral treatment as well as shortness of breath and CRP levels were interrelated, only one of the two interrelated variables was included in the logistic regression model.
Severe cases. Of the 121 patients examined, 9 (7.4%) were admitted to an ICU. The median age of those who were admitted to the ICU was 43 years (range, 23-62 years). Of the 9 ICU patients, 3 had underlying medical conditions, including 1 patient with chronic lympholeukemia, 1 pregnant woman (25-week pregnancy), and 1 obese patient with cardiovascular pathology. Acute respiratory distress syndrome was diagnosed in 7 ICU patients; 5 had a clinical diagnosis of septic shock. Mechanical ventilation was urgent in 8 cases; meanwhile, extracorporeal membrane oxygenation was necessary in 2 cases. Secondary bacterial pneumonia was diagnosed in 3 ICU patients, in all of which it was ventilation-associated pneumonia caused by Pseudomonas (n=2) and Acinetobacter (n=1). All patients received antiviral drugs and antibiotics.
Of the 121 study patients, 6 (5%) died; all were ICU patients. The median age of the patients who died was 43.5 years (range, 23-62). A pregnant woman and a patient with lympholeukemia were among those who died. The remaining 4 patients who died had been previously healthy.
The median time from illness onset to initiation of antiviral therapy in both the groups of ICU patients who survived and those who died was 5 days (range, 0-10 days). Antiviral drugs had been given to 22% (2/9) of the ICU patients and 17% (1/6) of the patients who died within 48 h after illness onset.
The mean hospital stay of noncomplicated, complicated, and severe cases was 4.8±2.3, 12.2±13, and 33±32.5 days, respectively. The median time from illness onset to death was 21.5 days (range, 7-49 days). Table 2 represents comparison of severe with nonsevere cases.
Pregnant vs. Nonpregnant Women of the Reproductive Age. Of the 55 women in the study cohort who were of childbearing age (18-44 years old), 30 (54.5%) were pregnant and 1 immediately postpartum (1.8%). Of the 30 pregnant women, 2 (7%) were in the fi rst, 15 (50%) in the second, and 13 (43%) in the third pregnancy trimester. No obstetric complications in the group of pregnant women were reported during hospitalization. Comparison of pregnant and nonpregnant women of childbearing age is presented in Table 3.

Discussion
In this study, we have shown that pandemic infl uenza caused considerable morbidity in a signifi cant proportion of hospitalized adults and that the use of antiviral drugs was benefi cial in hospitalized patients, especially when such therapy was initiated early.
The majority of information on clinical presen-  (6,7). We believe that it strengthens the validity of our fi ndings despite the retrospective nature of this study. The distribution of infl uenza A (H1N1) cases by age, characterized by the concentration of cases in 18-40-year-old individuals, was similar to the distribution of cases observed worldwide (8)(9)(10)(11). Only 5% of our patients were aged more than 65 years. Such fi ndings suggest the possibility of varying levels of immunity in this age group or a reduced penetration of the virus into the elderly community in the fi rst wave of the pandemic.
The majority of hospitalized patients in our study had the recognized risk factors for severe infl uenza. Obesity and cardiovascular pathology were the most common risk conditions. Arterial hypertension was the main illness observed in 69% of cases with cardiovascular pathology (data not shown). This refl ects a higher prevalence of arterial hypertension in relation to age rather than infl uence on the course of disease (3,4,9,10). In contrast, recent reports have reported high proportions of severe or fatal pandemic infl uenza cases among obese patients (8)(9)(10)(11). We did not, however, observe signifi cant differences in the prevalence of any obesity among hospitalized patients who died and those who survived due to the limited number of patients includ-ed in our study. However, taking into consideration its prevalence in our cohort and the fact that obesity is usually associated with other conditions such as diabetes, obesity should be viewed as an important risk factor when treatment and prevention measures of infl uenza are considered.
Pneumonia was the main complication observed in 39.7% of hospitalized patients. Complicated disease was associated with a longer interval from symptom onset to antiviral treatment and increased age. Shortness of breath was signifi cantly more often observed in the group of complicated cases and can be considered as a highly predictive symptom of possible involvement of the lower respiratory tract. However, 1 of the 3 patients with clinical symptoms compatible with the WHO defi nition of uncomplicated pandemic infl uenza had radiologically confi rmed pneumonia on admission. Moreover, more than one-third of complicated cases (38.8%) were hospitalized within 48 hours from symptom onset. These fi ndings suggest that involvement of the lower respiratory tract can be rapid in a signifi cant proportion of patients with pandemic infl uenza and that sensitivity of subjective complaints is rather low to rely on it.
A signifi cantly greater proportion of patients who received antiviral treatment within 48 h from symptom onset of noncomplicated compared with complicated cases (69.7% vs. 21.2%, P>0.001) suggest that early antiviral treatment protects against the development of pneumonia. However, the role of neuraminidase inhibitors in recovery from pandemic infl uenza complicated by pneumonia remains unclear, since 30% of patients with radiologically confi rmed pneumonia recovered without antiviral treatment. Similar fi ndings were reported in other studies (12)(13)(14)(15). It is diffi cult to determine precisely the cause of pneumonia based on radiographs only,  and further studies are necessary to correlate radiographic fi ndings with the cause of pneumonia during an infl uenza outbreak. In the absence of accurate diagnostic methods, treatment with both antibiotics and antiviral drugs should be indicated for the patients who are hospitalized with suspected infl uenza and lung infi ltrates on chest radiography. In our series, severe cases requiring management in the ICU occurred in all age groups, mostly in previously healthy individuals without any known risk factors predisposing to complicated infl uenza. Death occurred in 5% of patients in our study, which corresponds to the fi ndings reported elsewhere (10,15,16). The risk of death from seasonal infl uenza is highest among people aged 65 years and more. In the case of pandemic infl uenza, severe outcomes occurred in a much younger population than typically it is in the case of seasonal infl uenza (10,(17)(18)(19). The results of our study confi rm this observation.
A signifi cant proportion -a quarter of all hospitalized patients -in this cohort was comprised of pregnant or postpartum women. The majority of pregnant women (90.3%) had uncomplicated infl uenza and were admitted due to pregnancy as a major risk factor for possible aggravated course of illness. Pregnant women did not differ from nonpregnant with regard to age, comorbidity, and use of antivirals. Only 1 of the 31 pregnant and postpartum patients had a severe course and died, which differs from the reports of increased mortality rates for this group of patients in other countries (20)(21)(22)(23)(24)(25)(26). Since time from symptom onset to hospitalization and initiation of antiviral treatment was signifi cantly shorter in pregnant women compared to nonpregnant, it is tempting to conclude that pregnant women ben-efi t from early antiviral therapy. Taken together, we attribute our fi ndings of a less complicated course of illness in pregnant women to early presentation, diagnosis, and prompt specifi c treatment.
Our study has several limitations, which make our fi ndings suggestive rather than defi nitive. We evaluated only the patients with confi rmed pandemic 2009 infl uenza A (H1N1) virus infection; therefore, our sample may not be representative of the hospitalized patients who may not have been tested. In addition, the criteria for hospitalization and the algorithms for the use of diagnostic tests were not uniform in all three participating centers. Finally, despite the use of a standardized data collection form, not all information was collected for all patients.