The senic sampling process: Design for choosing hospitals and patients and results of sample selection

Dana Quade, David H. Culver, Robert W. Haley, Fredrick S. Whaley, William D. Kalsbeek, C. David Hardison, Robert E. Johnson, Richie C. Stanley, Richard H. Shachtman

Research output: Contribution to journalArticlepeer-review

23 Scopus citations

Abstract

To achieve its primary objectives, the Study on the Efficacy of Nosocomial infection Control (SENIC Project) focused its attention on a target population of patients referred to as SENIC-eligible admissions in a target population of hospitals referred to as the "SENIC Universe," SENIC thus required a design for sampling hospitals and patients within these hospitals and a valid procedure for projecting sample results to the target population. This paper presents the details of the sampling design used, describes the actual process of selecting hospitals and patients for the surveys, explains the procedure used to project sample results to the target population, and examines the possibility of bias in the design and hospital selection process. As with most large-scale sample surveys, the design and sample selection processes for the surveys in Phases II and III of SENIC were complicated by incomplete frame, nonresponse and measurement problems. Nevertheless, adjustments to reduce the effects of some of these problems have been made through the development of a valid procedure for projecting sample results to the target popula tion, and it appears unlikely that practically important nonsampling biases will result from the estimation procedures applied to this sample of hospitals.

Original languageEnglish (US)
Pages (from-to)486-502
Number of pages17
JournalAmerican Journal of Epidemiology
Volume111
Issue number5
DOIs
StatePublished - May 1980

Keywords

  • Epidemiologic
  • Health surveys
  • Sampling studies
  • Statistics

ASJC Scopus subject areas

  • Epidemiology

Fingerprint

Dive into the research topics of 'The senic sampling process: Design for choosing hospitals and patients and results of sample selection'. Together they form a unique fingerprint.

Cite this