Action Alert: Proposed questionnaire redesign for the National Health Interview Survey
Crucial Content May Be Cut from the National Health Interview Survey
On February 22, 2016 the National Center for Health Statistics announced a plan to dramatically overhaul the NHIS questionnaire and is requesting comments on these changes. They have only provided a 21-day period for public comment, so you must act quickly to make your voice heard. Responses to the proposed plan are due March 15, 2016. To provide comments, please send an email to healthsurveys@cdc.gov.
The most significant proposed change is the elimination of the family questionnaire that collects basic demographic, socioeconomic, health status, disability, and health insurance information about everyone in the sampled household. The revised survey will only collect this information for one sampled adult and one sampled child per household. It is also unclear whether the collection of family income and poverty information will also be discontinued.
The proposed termination of the family survey signals a dramatic departure from the previous fifty years of NHIS data collection and greatly reduces the utility of the NHIS to answer today's pressing health questions, in two ways. First, we will no longer be able to use the NHIS to study the effects of family context on health. Second, the use of the NHIS to study health disparities will be severely curtailed because the new design will be less likely to capture large enough samples of special populations like immigrant families and children, people with disabilities, or persons in same-sex relationships.
We at the MPC urge you to take action. Please let the NHIS know that this data is too important to lose. Submit a letter to healthsurveys@cdc.gov before March 15, 2016. Provide specific details on the damage that will result if these survey questions are eliminated.
We have created a document with sample text to get you started.
Copy ipums@umn.edu on your email, and we will gather a list of the valuable research projects that will be hindered by these changes. Read our response to the announced changes.
Background
Household-based surveys like the NHIS form the backbone of U.S. population data infrastructure. While other U.S. household surveys focus on measurement of labor market outcomes, the NHIS is exceptional in collecting household information alongside a wealth of health and health care outcomes.
Read our bibliography to find studies using NHIS to study these important topics.
The Influence of Family Context on Child Health
The unique co-occurrence of this information in a single data set has contributed to the knowledge that family context is incredibly important for child health. Three areas in particular are that:
- Poverty holds severe consequences for child development and health;
- Maternal education and employment exert great influence on child health outcomes; and
- Children living in families headed by a single parent have dramatically different health outcomes compared with children in nuclear families.
Researchers recognize the importance of family context to child health: Google Scholar citations to studies using the NHIS to study the effect of parent characteristics on child health number over 22,000, with more than half of those studies published in the last 5 years alone ("National Health Interview Survey" & parent, search conducted 2/29/16).
Monitoring Health Disparities
The proposed changes will undermine our ability to use the NHIS to monitor health disparities. Some special populations are identified using information on all household members. These people include:
- Same-sex unions
- Mixed-race unions
- Blended and extended families
- U.S.-born children of immigrants
Limiting the sample to one selected adult and child greatly reduces the number of cases of interest for key health topics. To illustrate how these changes will harm studies of health disparities, we need only highlight that membership in a special population affects whether a person is selected as the sample adult or sample child. For example, sample adults are significantly more likely to be born in the U.S. than adult household members not selected to be the sample adult (82% vs. 77% over the 1997-2014 period). Using the projected sample sizes announced at the 2015 PAA meetings, limiting the NHIS only to the sample adult translates into an estimated NHIS sample size reduction of 5,473 immigrant adults per year, or a 37% annual cut. See how we generated these estimates in Stata.
Evaluating the Impact of the ACA
The importance of family context is not limited to its effects on child health. It is impossible to evaluate some of the recent health care reforms enacted under the Affordable Care Act without information about relationships between co-resident family members because health insurance coverage, whether through an employer or a public program, is often based on family relationships (SHADAC 2012).
The Impact of Family Context on Adult Health
Family and household arrangements impact the risk of suicide for American adults, where persons living with unrelated adults are significantly more likely to commit suicide (Denney 2010). The living arrangements of older adults significantly impact their self-rated quality of life, psychological distress, disability onset, and mortality outcomes (Henning-Smith 2016; Scott et al. 1997).
A disturbing gap in public health surveillance and population data infrastructure
The importance of maintaining data collection about all members of the household is clear. The current system of collecting individual information for the entire household works because it is flexible, allowing researchers to select the family characteristics most relevant to their studies. Removing the measurement of family characteristics from the most important national survey of U.S. health will leave a disturbing gap in public health surveillance and population data infrastructure and will harm the state of health knowledge for years to come.