Research and Methods in Health Statistics

archived

Research and Methods in Health Statistics Purpose This initiative invites investigator-initiated research grant applications for projects involving the development and testing of statistical and survey methodology relevant to the conduct, analysis and reporting of health surveys and vital records. Existing NCHS data sets alone or in conjunction with other data sets may be used to develop and test new survey methodology, statistical analytical approaches or methods of displaying data. Individual stand-alone projects relevant to the collection, analysis, or display of data are acceptable under this announcement. The R03 mechanism is intended to support small research projects that can be carried out in a short period of time with limited resources. It can be used to support different types of projects including: secondary analysis of existing data; small, self-contained research projects; pilot and feasibility studies; development and testing of statistical and survey methodology; and the development and testing of new survey technology. SURVEY AND STATISTICAL METHODOLOGY: Health research using individual and provider surveys and vital statistics is facing both new and ongoing challenges. In recent years, survey response rates have been falling raising concerns about possible bias in the resulting data. New concerns about confidentiality have been raised which may influence cooperation in survey response. Technological developments have opened up new data collection opportunities but questions remain about the validity and reliability of the data obtained through these new approaches. The increasingly rapid availability of health-related data from non-traditional sources such as website visits and other types of transactions has spurred the demand for rapid information dissemination but what is the accuracy and reliability of data from these sources? The expense and time involved in collecting new data argue for exploring ways to link data sets to obtain maximum use of existing data. Researchers have long been concerned about the accuracy of self-reported data. New methods of collecting biomarker data have made it possible to include specimen collection in traditional surveys, but how accurate are these new methods? How do they relate to self-reported data? Increasingly, there is a demand for providing data in easily understood, eye-catching possibly interactive formats. Examples of topics of interest under this announcement include but are not limited to: o the development and refinement of innovative techniques for measurement of biomarkers in survey research including the collection of biological specimens such as urine or blood, or other physical measures such as heart rate, senses, blood pressure, height and weight; o the development and refinement of summary measures of health; o the development and refinement of measures of health insurance access and use; o the development and improvement of sampling strategies for subpopulations of interest including minority populations, people with specific rare diseases or conditions, specific socioeconomic statuses, or people with only cell phones; o the development and improvement of methods of analysis for small areas or small subpopulations; o the development and improvement of techniques to avoid disclosure of confidential data in public use data sets; o the examination of characteristics of the interview situation that may affect the collection of unbiased data, the assessment of the sources of bias and the impact of biased data on results; o the validation of various sources of data including vital records and electronic health records; o new ways of presenting analytical results to a variety of audiences; o research exploring new data sets available from NCHS including the NHANES National Youth Fitness Survey (NNYFS) http://www.cdc.gov/nchs/nnyfs.htm; NHANES oversample of Asian Americans http://www.cdc.gov/nchs/nhanes/nhanes_questionnaires.htm; o NHIS sexual orientation data http://www.cdc.gov/nchs/nhis/nhis_questionnaires.htm http://www.cdc.gov/nchs/data/nhsr/nhsr077.pdf; o U.S. fetal death data http://www.cdc.gov/nchs/data_access/Vitalstatsonline.htm http://www.cdc.gov/nchs/fetal_death.htm; o birth data files including multistate data for selected items exclusive to the 2003 revision of the U.S. Standard Certificate of Live Birth http://www.cdc.gov/nchs/data/nvsr/nvsr62/nvsr62_04.pdf http://www.cdc.gov/nchs/data_access/Vitalstatsonline.htm; o Native Hawaiian Pacific Islander NHIS survey http://www.cdc.gov/nchs/nhis/nhpi.html ; o NCHS linked data sets http://www.cdc.gov/nchs/data_access/data_linkage_activities.htm; o NHANES genetic data http://www.cdc.gov/nchs/nhanes/genetics/genetic_types.htm#restricted . Research proposals need not be limited to the examples given above. It is anticipated that many questions will derive from previous or current research. It may be cost-effective to design methodological studies to be carried out in conjunction with ongoing research projects. Applicants are referred to the NCHS website for information on the many other data sets available from NCHS. http://www.cdc.gov/nchs/

High