4 DISCUSSION
Influenza remains a heavy clinical burden worldwide. Recent reports have
suggested estimated annual hospitalization rates in various countries
varying from 13 to 141 per 100,000 population 5-10. In
Japan, 1.0% of influenza-positive patients aged under 75 years were
hospitalized between 2012 and 2016 11. However, those
reports included pediatric patients, and focused mainly on respiratory
complications such as pneumonia.
The JRS conducted a surveillance of adults and adolescent patients with
influenza requiring hospitalization. All causes for hospitalization
related to influenza infection were enrolled. For this study, we
introduced a new surveillance system that worked via the
internet. By using this internet-based system, prospective real-time
data can be registered and renewed rapidly from many institutions
throughout Japan, and reminder mails can be sent.
In our study, most cases of influenza-related hospitalization in Japan
occurred in the elderly, as has been previously in papers from various
countries 7, 10, 12-14. However, the age distribution
during the 2015-2016 season was significantly younger than those of
other seasons. This distribution was thought to reflect the fact that
the predominant influenza virus strain in the 2015-2016 season was A
(H1N1) pdm. This observation confirms a previous report indicating that
the median age of hospitalized cases in the 2009 A (H1N1) pdm pandemic
was younger than is typically seen for seasonal influenza14. A considerable number of younger adults who were
naïve for A (H1N1) pdm in the preceding seasons might have been infected
for the first time during the 2015-2016 season, and some of these
individuals presumably were hospitalized due to complications.
The most common cause of hospitalization was pneumonia, an observation
that is consistent with the results of another Japanese study by
Maruyama and colleagues 15. Primary viral pneumonia
was recognized in only 7.6% of all pneumonia cases. However, 68% of
those primary viral pneumonia cases were reported in the 2015-2016
season. This pattern also is thought to be attributable to the
predominance of A (H1N1) in the 2015-2016 season. Diagnosis of primary
viral pneumonia seems to depend on attending physicians’ decisions based
on clinical findings. However, some cases were diagnosed via the
PCR method using bronchoalveolar lavage fluid (BALF). Recently, a
multiplex real-time PCR method has been introduced for diagnosis of CAP16, which might lead to an apparent increase in the
frequency of influenza viral pneumonia. In the present study, the major
pathogens in secondary bacterial pneumonia were S. pneumoniae, H.
influenzae, and S. aureus consistent with previous report17. However, none of these bacterial species appeared
to be an independent risk factor for mortality.
The second-most common cause of hospitalization in our study was “poor
general condition”. Most of our patients were elderly people who had
multiple medical conditions, and thus were fragile in the face of
influenza infection. In Japan, the threshold of hospitalization
generally is lower than in other countries. Therefore, in Japan, the
elderly were more typically hospitalized when suffering from influenza.
Our study enrolled patients over 15 years old. Thus, there were few
cases of encephalopathy with influenza. Additionally, our cases included
few pregnant women, though such patients have been defined as having a
high risk of complication from influenza 18. In Japan,
there have not been many severe cases of pregnant women with influenza,
and no fatal cases of pregnant women with influenza were reported during
the pandemic of 2009 A (H1N1) 19.
Most patients (97%) received antiviral therapy from the early stage of
the illness in this study. In Japan, those who are suspected to have
influenza undergo rapid diagnostic antigen testing, and antiviral
therapy is initiated for those with positive results or on a clinical
basis as a standardized strategy 20, 21. Peramivir was
the most-commonly used antiviral agent because this medication is
provided as an injection drug, a format that is useful for treatment of
inpatients. In Japan, baloxavir marboxil, a novel cap-dependent
endonuclease inhibitor, was introduced to the market in March, 2018. An
effect of baloxavir marboxil is expected because this medication has an
antiviral mechanism that is distinct from that of neuraminidase
inhibitors (NAIs) 22, 23. However, there is some
concern that viral strains with reduced baloxzavir susceptibility will
emerge especially in younger children, seriously ill patients, or the
immunocompromised 24. Few patients underwent antiviral
therapy with baloxavir in the present study. It will be necessary to
monitor the future trends of baloxavir use. Treatment of hospitalized
patients with the combination of barloxavir and NAIs has been reported
to be effective 25. This combination treatment is
thought to be a possible strategy that deserves further examination in
future studies.
In total, 4.8% of the hospitalized cases were fatal. This rate was
lower than that reported in other countries 8, 14,26, 27. The lower value was thought to reflect the
generally lower threshold for hospitalization in Japan compared to other
countries. As mentioned above, early diagnosis and treatment are the
essential strategies against influenza in Japan. In our study, pneumonia
was overwhelmingly the leading cause of death associated with influenza
as reported previously 8, 11, 12, 14, 15. Secondary
bacterial pneumonia following influenza infection remains a serious
problem, especially in the elderly. Our multivariate analysis proved
that nursing home resident and obesity were independent risk factors for
mortality. Obesity has been identified as a prognostic factor for
influenza 28.29. Recently, the number of obese people
has increased in Japan, which has lead in turn to several clinical
problems including increased susceptibility to infection.
This study had several limitations. First, the number of participating
institutions was small because this research was conducted as a pilot
study. Moreover, about one third of the enrolled cases came from a
single hospital (Kurashiki Central Hospital), which might lead to
institutional bias. Second, the scale of the participating institutions
varied from primary physician’s clinics to educational hospitals; thus,
clinical levels were not uniform. Sufficient examinations could not be
done in all of the cases. The detailed data obtained from a single
center will be reported as a supplement to this report. Third, the
influence of vaccination could not be analyzed because vaccination
status was not obtained for many patients despite of a prospective
nature of this study. Fourth, detailed clinical data for each case were
not obtained because of the limits of the internet-surveillance system.
In conclusion, complications of influenza infection remain a heavy
burden especially among the elderly. Continuous nationwide surveillance
will be required to clarify the actual situation of influenza.