Evaluation of pandemic and seasonal influenza vaccination surveys Público
Singleton, James Andrew (2012)
Abstract
To monitor 2009-10 monovalent pH1N1 and seasonal vaccination,
the Centers for Disease Control and Prevention conducted the
National 2009 H1N1 Flu Survey (NHFS). providing weekly estimates of
pH1N1 and seasonal influenza vaccination. Evaluation of the
validity of findings from the NHFS is needed to improve design,
implementation, analysis and interpretation of future pandemic and
inter-pandemic influenza vaccination surveys. Three questions about
systematic error in measures of frequency based on influenza
vaccination surveys were addressed:
1. Can a quicker and cheaper telephone survey be conducted without
introducing too much additional selection bias?
2. How much selection bias is incurred by conducting a telephone
survey compared to the National Health Interview Survey (NHIS), a
less timely data but possibly more valid data source?
3. How accurate is parental report of young children's influenza
vaccination status, compared to provider reported status in the
National Immunization Survey (NIS)?
From analysis of NHFS based on interviews conducted within the
first two weeks compared with the full sample recruited over five
weeks, there was little difference in influenza vaccination
estimates. However, estimates from the full NHFS were 6-9
percentage points higher than estimates from the NHIS. Comparison
of parental to provider reported influenza vaccination status for
children aged 10-37 months indicated that vaccination prevalence
based on parental report was 5-12 percentage points higher. This
evaluation quantified levels of potential selection and
misclassification bias incurred by telephone surveys of influenza
vaccination. Telephone surveys to collect influenza vaccination
data by parental and self report remain a timely and efficient
approach for surveillance of vaccination programs at the national
and state levels. The attributes of ongoing surveillance systems
must be monitored to ensure they are meeting the needs of intended
use and are correctly interpreted.
Evaluation of pandemic and seasonal influenza vaccination
surveys
By
James A. Singleton
B.S., University of California, Davis, 1983
M.S., University of California, Davis, 1985
Advisor: Carolyn Drews-Botsch, Ph.D., M.P.H.
A dissertation submitted to the Faculty of the
James T. Laney School of Graduate Studies of Emory University
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
in Epidemiology
2012
Table of Contents
TABLE OF CONTENTS
Table of Contents...vii
List of Tables...ix
List of Figures...xii
Chapter 1 Introduction...1
Surveillance of influenza vaccination...1
Sources of systematic error in surveys of influenza
vaccination...3
Selection bias...6
Information bias...9
Dissertation aims...12
Chapter 2 Methods...13
Data sources...14
National 2009 H1N1 Flu Survey (NHFS)...14
National Health Interview Survey (NHIS)...16
National Immunization Survey (NIS)...17
Survey weighting and analysis...19
Estimating prevalence of influenza vaccination...21
Protection of human subjects...25
Chapter 3 Design of health surveys for public health emergencies: early responder bias in the National 2009 H1N1 Flu Survey (NHFS)...26
Abstract...26
Introduction...27
Methods...29
Data sources...29
Analysis...29
Nonresponse bias assessment...30
Effect of restriction to early respondents on vaccination
prevalence rates...34
Results...31
Nonresponse bias assessment...33
Effect of restriction to early respondents on vaccination
prevalence ratios...34
Discussion...35
Limitations...37
Conclusions...37
Chapter 4 Evaluating nonresponse and noncoverage bias in a telephone survey of Influenza vaccination by comparison to the National Health Interview Survey (NHIS)...52
Abstract...52
Introduction...53
Methods...55
Data Sources...55
National 2009 H1N1 Flu Survey (NHFS)...55
National Health Interview Survey (NHIS)...55
Analysis...56
Comparison of characteristics across sample
sources and surveys...58
Prevalence of selected influenza-related outcomes to NHFS sample
sources...59
Time-to-event analysis of influenza vaccination coverage...60
Results...61
Discussion...66
Limitations...70
Conclusions...72
Appendix 4-1 Undercoverage bias...82
Chapter 5 Validity of parental report of 2009-10 influenza vaccination in young children in a population-based national survey...83
Abstract...83
Introduction...84
Methods...86
Data Source - the National Immunization Survey
(NIS)...86
Determination of household reported influenza vaccination
status...87
Determination of provider reported influenza vaccination
status...89
Definition of validity parameters...89
Primary validity analysis...90
Accuracy of joint distribution of household reported pH1N1 and
seasonal influenza vaccination status...92
Accuracy of household reported month of vaccination...92
Accuracy of household reported type of pH1N1 and seasonal influenza
vaccination...93
Completeness of provider reported influenza vaccinations...93
Scenarios to illustrate potential implications of imperfect
recall...94
Results...95
Characteristics of study population...95
Primary validity results...95
Characteristics associated with sensitivity and
specificity...97
Joint distribution of household reported pH1N1 and seasonal
influenza vaccination status...99
Accuracy of household reported month of vaccination...99
Accuracy of household reported type of pH1N1 and seasonal influenza
vaccination...101
Effects of under-ascertainment of provider reported vaccinations on
validity Parameter estimates...102
Scenarios to illustrate potential implications of imperfect
recall...104
Discussion...106
Limitations...112
Conclusions...113
Appendix 5-1 Definitions of validity parameters...147
Appendix 5-2 Model for misclassification of provider reported
vaccination status...149
Appendix 5-3 Estimating actual vaccination prevalence using
validity studies...152
Chapter 6 Conclusions...154
Contributions to public health and survey
methodology...155
Limitations...157
Conclusions...159
Literature Cited...162
LIST OF TABLES
Table 3-1 Comparison of respondent characteristics between early
and later responders, by sample source, National 2009 H1N1 Flu
Survey (standalone component)...38
Table 3-2a Comparisons of adjusted prevalence (%) of
influenza-related outcomes between early and later responders,
National 2009 H1N1 Flu Survey (standalone component)...40
Table 3-2b Comparisons of adjusted prevalence (%) of
influenza-related outcomes between 1st week and later responders,
National 2009 H1N1 Flu Survey (standalone component)...41
Table 3-3a Influenza vaccination coverage - pH1N1 - through May
2010 for early vs. later and early weighted vs. all responders, by
vaccine and selected respondent characteristics, National 2009 H1N1
Flu Survey (standalone component)...42
Table 3-3b Influenza vaccination coverage - trivalent seasonal -
through May 2010for early vs. later and early weighted vs. all
responders, by vaccine and selected respondent characteristics,
National 2009 H1N1 Flu Survey (standalone component)...43
Table 3-3c Influenza vaccination coverage - pH1N1 - through May
2010 for 1st week vs. later and 1st week weighted vs. all
responders, by vaccine and selected respondent characteristics,
National 2009 H1N1 Flu Survey (standalone component)...44
Table 3-3d Influenza vaccination coverage - trivalent seasonal -
through May 2010for 1st week vs. later and 1st week weighted vs.
all responders, by vaccine and selected respondent characteristics,
National 2009 H1N1 Flu Survey (standalone component)...45
Table 3-4a Comparison of influenza vaccination prevalence ratios
from main effects logistic regression models, child overall sample,
early weighted respondents, and 1st week weighted respondents,
National 2009 H1N1 Flu Survey (standalone component)...46
Table 3-4b Comparison of influenza vaccination prevalence ratios
from main effects logistic regression models, adult overall sample,
early weighted respondents, and 1st week weighted respondents,
National 2009 H1N1 Flu Survey (standalone component)...47
Table 3-4c Comparison of influenza vaccination prevalence ratios
from logistic regression models with opinion covariates, adult
overall sample, early weighted respondents, and 1st week weighted
respondents, National 2009 H1N1 Flu Survey (standalone
component)...49
Table 4-1 Comparison of adult respondent characteristics by sample
source, National 2009 H1N1 Flu Survey (NHFS) (standalone component)
and National Health Interview Survey (NHIS), January-June 2010
interview data...74
Table 4-2 Comparison of child characteristics by sample source,
National 2009 H1N1 Flu Survey (NHFS) (standalone component) and
National Health Interview Survey (NHIS), January-June 2010
interview data...76
Table 4-3 Comparisons of unadjusted and adjusted prevalence (%) of
influenza-related outcomes between landline and cell only/mainly
adult respondents, National 2009 H1N1 Flu Survey (standalone
component)...77
Table 4-4 Comparisons of unadjusted and adjusted prevalence (%) of
influenza-related outcomes for children between landline and cell
only/mainly respondents, National 2009 H1N1 Flu Survey (standalone
component)...77
Table 4-5 Influenza vaccination coverage - pH1N1 - among adults
through May 2010, by selected respondent characteristics,
January-June 2010 interviews from the National 2009 H1N1 Flu Survey
(standalone component) and National Health Interview
Survey...78
Table 4-6 Influenza vaccination coverage - Seasonal - among adults
through May 2010, by selected respondent characteristics,
January-June 2010 interviews from the National 2009 H1N1 Flu Survey
(standalone component) and National Health Interview
Survey...79
Table 4-7 Influenza vaccination coverage - pH1N1 - among children
through May 2010, by selected respondent characteristics,
January-June 2010 interviews from the National 2009 H1N1 Flu Survey
(standalone component) and National Health Interview
Survey...80
Table 4-8 Influenza vaccination coverage - Seasonal - among
children through May 2010, by selected respondent characteristics,
January-June 2010 interviews from the National 2009 H1N1 Flu Survey
(standalone component) and National Health Interview
Survey...81
Table 5-1 Weighted distribution of study sample by shot card
status† and selected characteristics, National Immunization
Survey (NIS), Quarter 4 2009 - Quarter 2 2010...115
Table 5-2 Estimated validity of household as compared to provider
reported seasonal influenza vaccination status (one or more doses
in past 12 months) by shot card status and month of interview,
National Immunization Survey (NIS), Quarter 4 2009 - Quarter 2
2010...117
Table 5-3 Estimated validity of household as compared to provider
reported pH1N1influenza vaccination status (one or more doses since
October 2009) by shot card status and month of interview, National
Immunization Survey (NIS), Quarter 4 2009 - Quarter 2
2010...118
Table 5-4 Estimated validity of household as compared to provider
reported seasonal influenza vaccination status (two or more doses
in past 12 months) by shot card status and month of interview,
National Immunization Survey (NIS), Quarter 4 2009 - Quarter 2
2010...119
Table 5-5 Estimated validity of household as compared to provider
reported pH1N1influenza vaccination status (two or more doses in
since October 2009) by shot card status and month of interview,
National Immunization Survey (NIS), Quarter 4 2009 - Quarter 2
2010...120
Table 5-6 Adjusted estimates (%) of sensitivity and specificity of
parental report of child's seasonal influenza vaccination status
(one or more doses in past 12 months) as of date of interview using
shot card and recall (shot card group), by selected
characteristics, National Immunization Survey (NIS), Quarter 4 2009
- Quarter 2 2010...121
Table 5-7 Adjusted estimates (%) of sensitivity and specificity of
parental report of child's seasonal influenza vaccination status
(one or more doses in past 12 months) as of date of interview using
recall only (recall only group), by selected characteristics,
National Immunization Survey (NIS), Quarter 4 2009 - Quarter 2
2010...123
Table 5-8 Adjusted estimates (%) of sensitivity and specificity of
parental report of child's pH1N1 influenza vaccination status
(since October 2009) as of date of interview using shot card and
recall (shot card group), by selected characteristics, National
Immunization Survey (NIS), Quarter 4 2009 - Quarter 2
2010...125
Table 5-9 Adjusted estimates (%) of sensitivity and specificity of
parental report of child's pH1N1 influenza vaccination status
(since October 2009) as of date of interview using recall only
(recall only group), by selected characteristics, National
Immunization Survey (NIS), Quarter 4 2009 - Quarter 2
2010...127
Table 5-10 Joint distribution of seasonal and pH1N1 influenza
vaccination status by provider and household report, by shot card
status, National Immunization Survey (NIS), Quarter 4 2009 -
Quarter 2 2010...129
Table 5-11 Percent of children with provider reported seasonal
influenza vaccination in a given month with household reported
seasonal influenza vaccination in each month, National Immunization
Survey (NIS), Quarter 4 2009 - Quarter 2 2010...130
Table 5-12 Percent of children with provider reported pH1N1
influenza vaccination in a given month with household reported
pH1N1 influenza vaccination in each month, National Immunization
Survey (NIS), Quarter 4 2009 - Quarter 2 2010...130
Table 5-13 Sensitivity and specificity of parental report of
child's seasonal influenza vaccination status by date of interview
using shot card and recall (shot card group), by monthly
vaccination periods October-December 2009 and month of survey
interview, National Immunization Survey (NIS), Quarter 4 2009 -
Quarter 2 2010...131
Table 5-14 Sensitivity and specificity of parental report of
child's seasonal influenza vaccination status by date of interview
using recall only (recall only group) , by monthly vaccination
periods October-December 2009 and month of survey interview,
National Immunization Survey (NIS), Quarter 4 2009 - Quarter 2
2010...132
Table 5-15 Sensitivity and specificity of parental report of
child's pH1N1 influenza vaccination status by date of interview
using shot card and recall (shot card group), by monthly
vaccination periods October-December 2009 and month of survey
interview, National Immunization Survey (NIS), Quarter 4 2009 -
Quarter 2 2010...133
Table 5-16 Sensitivity and specificity of parental report of
child's pH1N1 influenza vaccination status by date of interview
using recall only (recall only group) , by monthly vaccination
periods October-December 2009 and month of survey interview,
National Immunization Survey (NIS), Quarter 4 2009 - Quarter 2
2010...134
Table 5-17 Distributions of types of seasonal and pH1N1 influenza
vaccinations received by provider and household report among
children with influenza vaccination reported by both sources, by
shot card status and age as of November 2009, National Immunization
Survey (NIS), Quarter 4 2009 - Quarter 2 2010...135
Table 5-18 Provider ascertainment of influenza vaccinations
reported by survey respondents from vaccination records, by
provider response category, National Immunization Survey (NIS),
Quarter 4 2009 - Quarter 2 2010...136
Table 5-19 Effect of incomplete of provider ascertainment of
influenza vaccinations status on validity parameters, assuming
non-differential under-ascertainment by household reported
influenza vaccination status...137
Table 5-20 Effect of incomplete of provider ascertainment of
influenza vaccinations status on validity parameters, assuming
higher under-ascertainment when the household reports that the
child has not received influenza vaccination...138
About this Dissertation
| School | |
|---|---|
| Department | |
| Degree | |
| Submission | |
| Language |
|
| Research Field | |
| Palabra Clave | |
| Committee Chair / Thesis Advisor | |
| Committee Members |
Primary PDF
| Thumbnail | Title | Date Uploaded | Actions |
|---|---|---|---|
|
|
Evaluation of pandemic and seasonal influenza vaccination surveys () | 2018-08-28 16:29:29 -0400 |
|
Supplemental Files
| Thumbnail | Title | Date Uploaded | Actions |
|---|