The Impact of Rotavirus Mass Vaccination
The Impact of Rotavirus Mass Vaccination
The retrospective evaluation focused on all cases of RV-GE hospitalized between 1 January 2002 and 31 December 2009 at the Department of Pediatrics, Medical University Innsbruck, a tertiary hospital with additional primary and secondary care functions, covering the area of Tyrol with an average population of 1,277,775 inhabitants between 2002 and 2009. For our analysis, all RV positive stool samples from the Division of Hygiene and Medical Microbiology and the Routine Laboratory of the Department of Pediatrics were matched with the data from the hospital discharge records using the International Classification of Diseases, 10 edition (ICD-10) with search terms "gastroenteritis" (K52.9), "rotavirus" (A08.0), "exsiccosis" (A09) and "blood stream infection" (A41.9) as primary and secondary diagnosis to identify potential RV cases. For inclusion into the study, clinical diagnosis of RV infection had to be confirmed by laboratory results. For included patients a chart review was performed.
Groups were separated into a "pre-vaccination period" (January 2002 to December 2005), a "recommended and early funded vaccination period" (intermediate period) (January 2006 to December 2007) and a "funded vaccination period" (January 2008 to December 2009). For determination of the influence of UMV on the age distribution and the hospitalization rates in children with RV infections, children were separated into 5 different age groups according to epidemiological data: infants aged 0–11 months, toddlers aged 12–23 months, children aged 2–5 years, school children aged 6–10 years and children 11–18 years of age. The reasons for age classification are: Most children acquire their first RV-GE before the age of 5 years. Severe RV-GE is largely limited to children aged <24 months. Vaccinated children aged between 0 and 11 months should have received at least one dose of the vaccine. Toddlers aged between 7 and 24 months are most likely to have been fully vaccinated during UMV. The RV vaccination status could not be collected by chart review. Double-counters and patients admitted more than once due to RV-GE were excluded.
The study was performed according to the principles of the declaration of Helsinki 2008 and the local ethics committee of the Innsbruck Medical University.
Clinical data from patients included into the study were collected by chart analysis. Cases were defined as nosocomial (hospital-acquired) RV infections if the onset of gastroenteritis-specific symptoms (diarrhea, vomiting) was at least 48 hours after admission to hospital considering an incubation time for RV of 18–36 hours and an admission diagnosis that was not "gastroenteritis". Gastroenteritis was defined by more than 3 loose stools or watery diarrhea within 24 hours with or without vomiting (particularly in older children), fever and dehydration according to signs given in a scoring system for RV GE. The duration of the hospital stay was defined as the time span between the day of admission and the day of discharge. Nosocomial RV-GE cases were excluded from the analysis of hospital stay durations.
A blood stream infection (BSI) was defined as at least one of the following features: first, a blood culture positive for a pathogen; second, a common pathogen of human skin cultured from two or more blood cultures, both drawn on separate occasions; or third, a common pathogen of the human skin cultured from at least one blood culture in association with signs of a systemic inflammatory response syndrome (SIRS). A SIRS was defined by at least two of the following criteria: elevated body temperature >38°C or hypothermia <36°C, tachycardia, tachypnea, leucocytosis, leukopenia or >10% immature neutrophils according to age ranges. A secondary BSI was defined as a BSI following the clinical symptoms of gastroenteritis associated with laboratory confirmed RV infection more than 48 after onset of RV-associated disease. The present study focused exclusively on secondary BSI.
Hygienic regulations for prevention of nosocomial infections on the ward were not changed during 2002 to 2009. All nurses and doctors with direct contact to the RV-infected patient had to wear over-coats and had to follow a three minutes long hand washing program with Bode Sterillium® Virugard (Paul Hartmann AG, Telgte, Germany) disinfectant solution after patient contact. RV-positive patients and their accompanying persons were cohorted on the ward in separate rooms with own bath rooms and were prohibited to use any facilities on the ward which may have offered the possibility to get in contact with other patients.
Between 2002 and 2009, all hospitalized patients with gastroenteritis were screened for RV antigen in their stools. For detection of human RV antigen in stool specimen the Pathfinder Direct Antigen Detection System (Kallestad Laboratories, Inc. Austin, Texas) was used till 2005. The sensitivity and specificity of this test system is 84% and 98%, respectively. After 2005, the CerTestRota Card (Biotec, Zaragoza, Spain), an immunochromatographic test for Rotavirus detection in stool specimen was used for routine testing. The sensitivity and specificity of this test system is >99% and 98%, respectively. Routinely, stools were also investigated for additional viral pathogens (Norovirus, Adenovirus) via antigen detection and for bacterial pathogens (Salmonella, Yersinia, Shigella, Campylobacter spp. and enterohemorrhagic Escherichia coli) via stool cultures according to standard procedures.
From patients showing at least two signs of SIRS, an average of 4 ml blood was drawn for detection of bloodstream pathogens and inoculated into a BacT/Alert PF Pediatric FAN bottle (BioMérieux, Durham, USA). This procedure did not change over the observation period. The bottles were then loaded into a BacT/Alert 3D automated blood culture system (BioMérieux) for a five day protocol with monitoring of carbon dioxide production within each bottle every 10 min. All bottles marked positive were removed from the instrument, and an aliquot was taken for Gram staining and culture on solid media for subsequent analysis. Pathogen identification was performed according to standardized microbial procedures and by VITEK 2 system (BioMérieux).
The direct hospitalization costs for children and their accompanying persons are based on the accounts provided by the clearing office of the "Tiroler Landeskrankenanstalten" (TILAK) holding. For the pre-vaccination period the mean of the costs were 870 € per patient per day and for the accompanying person 35 € per day given a proportion of 87.2% accompanying persons in children <6 years (funded vaccination period: 930 €, 38 € and 90.8%, respectively). Children ≥6 years of age are usually not accompanied due to the insurance system which does not refund parents' costs for the hospital stay together with their child. Nosocomial infections were excluded from cost calculations because of co-morbidity-associated costs which do not allow an approximation of hospitalization costs. Estimated mean of costs per year for community-acquired RV-GE cases were calculated by multiplication of the mean hospital duration (days) (Table 1), mean number of patients per year per age-group (Table 2), mean of direct costs (€) and proportion of accompanying persons in age groups <6 years. Costing procedures have not changed during the study period.
The Innsbruck hospital covers approximately 70% of all pediatric RV-GE hospitalizations in Tyrol. estimated hospitalization rates were calculated for the pre-vaccination, the intermediate and funded vaccination period using the draw area of Tyrol for patients 0–18 years for the different periods (pre-vaccination period: 652,557; intermediate period: 319,825; funded vaccination period: 305,393) as denominator for calculation of incidence rates.
Statistical analysis was performed with SPSS Version 18.0 (Chicago, IL). Non-parametric Mann-Whitney-U test was used to compare mean hospitalizations per year between the pre-vaccination and the vaccination period. Pearson's Chi-square test was used to analyze difference in dichotome variables. A p < 0.05 was defined statistically significant.
Methods
Study Design and Study Population
The retrospective evaluation focused on all cases of RV-GE hospitalized between 1 January 2002 and 31 December 2009 at the Department of Pediatrics, Medical University Innsbruck, a tertiary hospital with additional primary and secondary care functions, covering the area of Tyrol with an average population of 1,277,775 inhabitants between 2002 and 2009. For our analysis, all RV positive stool samples from the Division of Hygiene and Medical Microbiology and the Routine Laboratory of the Department of Pediatrics were matched with the data from the hospital discharge records using the International Classification of Diseases, 10 edition (ICD-10) with search terms "gastroenteritis" (K52.9), "rotavirus" (A08.0), "exsiccosis" (A09) and "blood stream infection" (A41.9) as primary and secondary diagnosis to identify potential RV cases. For inclusion into the study, clinical diagnosis of RV infection had to be confirmed by laboratory results. For included patients a chart review was performed.
Subgroups of the Study Population
Groups were separated into a "pre-vaccination period" (January 2002 to December 2005), a "recommended and early funded vaccination period" (intermediate period) (January 2006 to December 2007) and a "funded vaccination period" (January 2008 to December 2009). For determination of the influence of UMV on the age distribution and the hospitalization rates in children with RV infections, children were separated into 5 different age groups according to epidemiological data: infants aged 0–11 months, toddlers aged 12–23 months, children aged 2–5 years, school children aged 6–10 years and children 11–18 years of age. The reasons for age classification are: Most children acquire their first RV-GE before the age of 5 years. Severe RV-GE is largely limited to children aged <24 months. Vaccinated children aged between 0 and 11 months should have received at least one dose of the vaccine. Toddlers aged between 7 and 24 months are most likely to have been fully vaccinated during UMV. The RV vaccination status could not be collected by chart review. Double-counters and patients admitted more than once due to RV-GE were excluded.
The study was performed according to the principles of the declaration of Helsinki 2008 and the local ethics committee of the Innsbruck Medical University.
Data Collection and Definitions
Clinical data from patients included into the study were collected by chart analysis. Cases were defined as nosocomial (hospital-acquired) RV infections if the onset of gastroenteritis-specific symptoms (diarrhea, vomiting) was at least 48 hours after admission to hospital considering an incubation time for RV of 18–36 hours and an admission diagnosis that was not "gastroenteritis". Gastroenteritis was defined by more than 3 loose stools or watery diarrhea within 24 hours with or without vomiting (particularly in older children), fever and dehydration according to signs given in a scoring system for RV GE. The duration of the hospital stay was defined as the time span between the day of admission and the day of discharge. Nosocomial RV-GE cases were excluded from the analysis of hospital stay durations.
A blood stream infection (BSI) was defined as at least one of the following features: first, a blood culture positive for a pathogen; second, a common pathogen of human skin cultured from two or more blood cultures, both drawn on separate occasions; or third, a common pathogen of the human skin cultured from at least one blood culture in association with signs of a systemic inflammatory response syndrome (SIRS). A SIRS was defined by at least two of the following criteria: elevated body temperature >38°C or hypothermia <36°C, tachycardia, tachypnea, leucocytosis, leukopenia or >10% immature neutrophils according to age ranges. A secondary BSI was defined as a BSI following the clinical symptoms of gastroenteritis associated with laboratory confirmed RV infection more than 48 after onset of RV-associated disease. The present study focused exclusively on secondary BSI.
Hygienic regulations for prevention of nosocomial infections on the ward were not changed during 2002 to 2009. All nurses and doctors with direct contact to the RV-infected patient had to wear over-coats and had to follow a three minutes long hand washing program with Bode Sterillium® Virugard (Paul Hartmann AG, Telgte, Germany) disinfectant solution after patient contact. RV-positive patients and their accompanying persons were cohorted on the ward in separate rooms with own bath rooms and were prohibited to use any facilities on the ward which may have offered the possibility to get in contact with other patients.
Laboratory Confirmation of RV and Other Pathogens From Stool Samples
Between 2002 and 2009, all hospitalized patients with gastroenteritis were screened for RV antigen in their stools. For detection of human RV antigen in stool specimen the Pathfinder Direct Antigen Detection System (Kallestad Laboratories, Inc. Austin, Texas) was used till 2005. The sensitivity and specificity of this test system is 84% and 98%, respectively. After 2005, the CerTestRota Card (Biotec, Zaragoza, Spain), an immunochromatographic test for Rotavirus detection in stool specimen was used for routine testing. The sensitivity and specificity of this test system is >99% and 98%, respectively. Routinely, stools were also investigated for additional viral pathogens (Norovirus, Adenovirus) via antigen detection and for bacterial pathogens (Salmonella, Yersinia, Shigella, Campylobacter spp. and enterohemorrhagic Escherichia coli) via stool cultures according to standard procedures.
Identification of Pathogens in Blood Culture
From patients showing at least two signs of SIRS, an average of 4 ml blood was drawn for detection of bloodstream pathogens and inoculated into a BacT/Alert PF Pediatric FAN bottle (BioMérieux, Durham, USA). This procedure did not change over the observation period. The bottles were then loaded into a BacT/Alert 3D automated blood culture system (BioMérieux) for a five day protocol with monitoring of carbon dioxide production within each bottle every 10 min. All bottles marked positive were removed from the instrument, and an aliquot was taken for Gram staining and culture on solid media for subsequent analysis. Pathogen identification was performed according to standardized microbial procedures and by VITEK 2 system (BioMérieux).
Cost Calculations
The direct hospitalization costs for children and their accompanying persons are based on the accounts provided by the clearing office of the "Tiroler Landeskrankenanstalten" (TILAK) holding. For the pre-vaccination period the mean of the costs were 870 € per patient per day and for the accompanying person 35 € per day given a proportion of 87.2% accompanying persons in children <6 years (funded vaccination period: 930 €, 38 € and 90.8%, respectively). Children ≥6 years of age are usually not accompanied due to the insurance system which does not refund parents' costs for the hospital stay together with their child. Nosocomial infections were excluded from cost calculations because of co-morbidity-associated costs which do not allow an approximation of hospitalization costs. Estimated mean of costs per year for community-acquired RV-GE cases were calculated by multiplication of the mean hospital duration (days) (Table 1), mean number of patients per year per age-group (Table 2), mean of direct costs (€) and proportion of accompanying persons in age groups <6 years. Costing procedures have not changed during the study period.
Calculation of Hospitalization Rates
The Innsbruck hospital covers approximately 70% of all pediatric RV-GE hospitalizations in Tyrol. estimated hospitalization rates were calculated for the pre-vaccination, the intermediate and funded vaccination period using the draw area of Tyrol for patients 0–18 years for the different periods (pre-vaccination period: 652,557; intermediate period: 319,825; funded vaccination period: 305,393) as denominator for calculation of incidence rates.
Statistical Analysis
Statistical analysis was performed with SPSS Version 18.0 (Chicago, IL). Non-parametric Mann-Whitney-U test was used to compare mean hospitalizations per year between the pre-vaccination and the vaccination period. Pearson's Chi-square test was used to analyze difference in dichotome variables. A p < 0.05 was defined statistically significant.
Source...