Immunization Registry and ProviderReported Vaccination Histories: Assessing Missing Vaccinations

1
Immunization Registry and Provider-Reported
Vaccination HistoriesAssessing Missing
Vaccinations

• Linda Piccinino (Abt Associates), Meena Khare
  (CDC), Mike Battaglia (Abt Associates), Diana
  Bartlett (CDC), Lawrence Barker (CDC)

2
NIS REGISTRY DATA QUERY STUDY

• Study integrated registry query consent question
  into standard National Immunization Survey (NIS)
  interview
• Four mature registry sites participated in the
  study
• Each site had one quarter of data collection (one
  site in Q3/2002, three sites in Q4/2002)
• Vaccination history data obtained from NIS
  provider reports and from state/local registries
  for children with parental consent

3
RESEARCH PROBLEM

• Need a way to assess the quality of vaccination
  data for NIS children with data available from
  both NIS provider reports and immunization
  registries

4
OBJECTIVES OF THIS ANALYSIS

• To compare vaccination histories obtained from
  immunization providers and registries for a
  sample of NIS children 19-35 months old
• To examine the pattern of missing vaccination
  data by vaccination type, dose and source of data
• To help identify subgroups of children where
  immunization data were more likely to be absent,
  by characteristics associated with the children
  and their provider types

5
ANALYSIS GROUPS

• Sample consisted of NIS children with both
  registry and provider data
• Data were combined for four sites because of
  sample size and confidentiality concerns
• For each vaccine type, children were grouped into
  two subgroups of children with
• provider-reported data but had some missing
  registry data
• registry-reported data but had some missing
  provider data

6
METHODOLOGY FOR IDENTIFYING MISSING VACCINATIONS

• 323 children with registry vaccination data (at
  least one shot) and adequate provider data in the
  NIS
• Focus on DTaP vaccinations present in one source
  but missing from the other source
• Children UTD with 4 doses of DTaP in the provider
  data (N 245) but fewer than 4 doses in the
  registry data (N 75)
• Children UTD with 4 doses of DTaP in the registry
  data (N 180) but fewer than 4 doses in the
  provider data (N 16)
• Determine dose number (order) of missing
  vaccination(s)

7
DETERMINING THE DOSE NUMBER OF MISSING
VACCINATIONS

• Provider data
• (1) 2/15/2001 (2) 8/12/2001 (3) 1/14/2002 (4)
  9/03/2002
• Registry data
• 2/15/2001 1/16/2002 8/30/2002
• For each registry vaccination date, determined
  which provider vaccination date was closest in
  time
• DOSE 1 MISSING DOSE 3 DOSE 4

8
CHILDREN WITH MISSING DTaP DOSES

• REGISTRY 75 CHILDREN (30.6)
  PROVIDER 16 CHILDREN (8.9)

9
CHILDREN WITH MISSING DTaP DOSES, contd

• REGISTRY 75 CHILDREN (30.6)
  PROVIDER 16 CHILDREN (8.9)

10
SUMMARY OF MISSING DOSES Dose Order

11
SUMMARY OF MISSING DOSES Dose Frequency

12
REASONS FOR MISSING PROVIDER DOSES

• 23 children had more registry doses than provider
  doses when all vaccines were examined We
  reviewed the Immunization History Questionnaires
  (IHQs) and recontacted the providers when
  necessary. Extra registry vaccination dose was
• very close to another registry vaccination dose
  date
• the 4th dose omitted from the IHQ or that dose
  was administered by a provider who did not
  respond to the NIS
• not the last dose and was inadvertently omitted
  when the IHQ was filled out
• for a child with multiple providers but household
  respondent only identified one provider
• for a child for whom not all identified providers
  responded or returned the IHQ with vaccination
  data

13
CHARACTERISTICS OF CHILDREN WITH MISSING DTaP
DOSES
14
IMPLICATIONS FOR DATA QUALITY

• INCOMPLETE VACCINATION HISTORIES
• The vaccination histories were not always
  complete.
• For children with 4 provider doses of DTaP, about
  31 of children had one or more doses missing in
  the registry data.
• For children with 4 registry doses of DTaP, only
  about 9 of children had one or more doses
  missing from the provider reports.
• These suggest that registry histories tended to
  be more incomplete than provider histories.

15
IMPLICATIONS FOR DATA QUALITY

• MIXED PATTERN OF MISSING DATA
• The frequency and order of the missing doses for
  DTaP varied by source of data.
• Therefore, source of data should be considered
  when interpreting results and making estimates.
• The fourth DTaP dose was the most likely to be
  missing in both the provider and registry data,
  although the reasons this 4th dose was missing
  might differ by data source.

16
IMPLICATIONS FOR DATA QUALITY

• SPECIAL SUBGROUP ISSUES
• Children with multiple providers were more likely
  to have missing provider doses
• Children with only private providers were more
  likely to have missing registry doses
• Providers that were not identified by respondents
  might also be responsible for some of the missing
  data
• A higher percentage of children with missing
  registry data had shot card data than for
  children with missing provider data

17
CONCLUSION

• Results may not be conclusive due to the small
  sample size.
• Results showed a mixed pattern of missing doses.
• Data quality concerns suggest that building
  combined best value estimates may be difficult
  since the quality and completeness of registry
  data are lacking.
• Although there are data quality problems, some
  registries were more complete than others.
• Results strongly support further research on
  supplementing provider-reported immunization
  histories with registry data in order to improve
  vaccination coverage estimates.

18
NEXT STEP

• Future research plans include looking at the
  quality of registry data compared to
  household-reported vaccination data from shot
  cards