Accuracy of influenza vaccination rate estimates in United States nursing home residents

J. M. Grosholz, S. Blake, J. D. Daugherty, E. Ayers, S. B. Omer, L. Polivka-West, D. H. Howard

Research output: Contribution to journalArticlepeer-review

5 Scopus citations


The US Center for Medicare and Medicaid Services (CMS) requires nursing homes and long-term-care facilities to document residents' vaccination status on the Resident Assessment Instrument (RAI). Vaccinating residents can prevent costly hospital admissions and deaths. CMS and public health officials use RAI data to measure vaccination rates in long-term-care residents and assess the quality of care in nursing homes. We assessed the accuracy of RAI data against medical records in 39 nursing homes in Florida, Georgia, and Wisconsin. We randomly sampled residents in each home during the 2010-2011 and 2011-2012 influenza seasons. We collected data on receipt of influenza vaccination from charts and RAI data. Our final sample included 840 medical charts with matched RAI records. The agreement rate was 0·86. Using the chart as a gold standard, the sensitivity of the RAI with respect to influenza vaccination was 85% and the specificity was 77%. Agreement rates varied within facilities from 55% to 100%. Monitoring vaccination rates in the population is important for gauging the impact of programmes and policies to promote adherence to vaccination recommendations. Use of data from RAIs is a reasonable approach for gauging influenza vaccination rates in nursing-home residents.

Original languageEnglish (US)
Pages (from-to)2588-2595
Number of pages8
JournalEpidemiology and Infection
Issue number12
StatePublished - Sep 1 2015
Externally publishedYes


  • Analysis of data
  • health statistics
  • influenza vaccines
  • vaccine policy development
  • vaccines

ASJC Scopus subject areas

  • Epidemiology
  • Infectious Diseases


Dive into the research topics of 'Accuracy of influenza vaccination rate estimates in United States nursing home residents'. Together they form a unique fingerprint.

Cite this