|
nif:isString
|
-
Ethical approval and signed patient consent forms were not required for our study according to the Guideline for Epidemiological Studies [19], which was established by the Ministry of Health, Labor and Welfare and the Ministry of Education, Culture, Sports, Science and Technology of Japan in accordance with the World Medical Association's Declaration of Helsinki and Japan's Act on the Protection of Personal Information and other related acts. Specifically, (1) all individual data were collected by law and authorized to be utilized for academic purposes [20], and (2) patients could not be identified, as all data were de-identified; i.e., stripped of personal identifiers.
Study population and data sources: Japan has an active infectious disease surveillance system. Since 1999, the National Institute of Infectious Disease (NIID; Tokyo, Japan) has collected reports of patients with various infectious diseases, and the data have been reported in sex and age groups (National Epidemiological Surveillance of Infectious Diseases, NESID) [21]. Diseases of interest in the present study were selected from those reported in NESID. Viral infectious diseases without availability of vaccination reported from the pediatric sentinel points: Five major viral diseases are reported from the pediatric sentinel points of NESID: pharyngoconjunctival fever (PCF), herpangina, hand-foot-and-mouth disease (HFMD), EI, and ES. No vaccinations are available for these diseases in Japan. Data were collected from approximately 3000 pediatric sentinel points all over Japan between 2000 and 2009 (Table 1). The number of the sentinel points represents approximately 10% of the pediatric facilities in Japan, and the average number of sentinel points in 2009 was 3,022. As shown in Table 1, the numbers of male and female cases from the sentinel points were reported across 13 age groups, and adult cases were also reported.
Table data removed from full text. Table identifier and caption: 10.1371/journal.pone.0042261.t001 Numbers of cases and male-to-female morbidity ratios of viral infections reported from the pediatric sentinel points from 2000 to 2009 in Japan. Vaccine-preventable viral infectious diseases reported from the pediatric sentinel points: Two vaccine-preventable viral infectious diseases, mumps and varicella, were studied using reports from the pediatric sentinel points of NESID. Vaccinations for mumps and varicella are optional for children older than 1 year under the Japanese law with vaccination rates being 23.2% and 21.3% against mumps and varicella, respectively [22]. The vaccination rates for those two infections are only available for combined males and females - vaccination rates for males and females were not available separately. However, it is assumed that there are no differences in the vaccination rates between the sexes, based on data from vaccination rates for the measles-rubella combination vaccine in Japan where rates are available for each sex, and there is no difference in the vaccination rates between the sexes [10], [23].
Data for four bacterial infections were available from NESID for the present study (Table 2); Group A streptococcal pharyngitis (GAS), pertussis, enterohemorrhagic Escherichia coli (EHEC), and Mycoplasma pneumonia (MP). GAS and pertussis were reported from the pediatric sentinel points. EHEC cases must, by law, be reported by all clinical facilities in Japan and archived in NESID, while the data for MP were collected from approximately 470 NESID sentinel points. Of these, only pertussis is vaccine-preventable; the vaccine is generally provided four times between 3 months and 7ยท5 years as a component of the diphtheria, tetanus and pertussis combined vaccine which is recommended under Japanese law with a vaccination rate of 95.8% [22].
Table data removed from full text. Table identifier and caption: 10.1371/journal.pone.0042261.t002 Numbers of cases and male-to-female morbidity ratios of bacterial infections reported from the sentinel points from 2000 to 2009 in Japan. *identical to the bottom row of the Table 1.
Statistical model and data analysis: Male-to-female morbidity ratios of infectious diseases without vaccine availability: Morbidities of males and females in each age group were compared through the male-to-female morbidity (MFM) ratios [10], statistics similar to ones used by Green [24] and Reller et al [25]. Since the present sampling is based on the data reported from the sentinel points, the sampling is viewed as a Poisson sampling. The morbidities (symptomatic incidence) at a current time, and , cannot be estimated from the observational patient data. Let and be the probabilities that male and female patients in the age group visit the sentinel points, respectively. From the present sampling from the sentinel points, the ratio can be estimated by maximum likelihood estimator, where NM and NF are the subpopulations of males and females in an age group in Japanese population; i.e. fixed values, and nM, and nF are the random variables that describe the numbers of male and female patients. The ratio is referred to as the apparent MFM ratio. If , then, is the true MFM ratio. For large and , is asymptotically normally distributed with mean and variance . In order to make multiple tests of MFM ratios in age groups, the Bonferroni method [26] is employed, and the Bonferroni 95% joint confidence intervals of MFM ratios are constructed. In order to estimate MFM ratios, ratios of male and female population sizes in age groups should be paid attention, as explained above. The male-to-female population ratios in age were almost constant from 2000 to 2009. Ratios of average subpopulations of males and females from 2000 to 2009 were used in this study (Table 1).
Male-to-female morbidity ratios of vaccine-preventable pediatric infectious diseases: Let and be the probabilities that male and female patients in the age group get vaccinated and let and be the probabilities that vaccinated male and female patients in the age group get immunized, respectively. From the present sampling from sentinel points, the ratio can be estimated by maximum likelihood estimator. If and , then, is the true MFM ratio.
|
![[This page as RDF]](http://data.gesis.org/somesci/static/rdf-icon.gif){kind=icon}