key: cord-0847756-ge8d98wt
authors: Zamboni, Lucila M.; Martin, Erika G.
title: Association of US Households’ Disaster Preparedness With Socioeconomic Characteristics, Composition, and Region
date: 2020-04-27
journal: JAMA Netw Open
DOI: 10.1001/jamanetworkopen.2020.6881
sha: 12085a1cb12d08c6c7d4b5573ba6cdb4717d2957
doc_id: 847756
cord_uid: ge8d98wt

IMPORTANCE: Since the terrorist attacks on September 11, 2001, the US government has promoted household disaster preparedness, but preparedness remains low. OBJECTIVE: To identify disparities in disaster preparedness among US households. DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study used data from a nationally representative sample of US households from the 2017 American Housing Survey’s topical section on preparedness to assess associations of disaster preparedness with households’ socioeconomic characteristics, composition, and region. Logistic regressions were used to assess associations of household characteristics with overall preparedness, resource- and action-based preparedness, and specific preparedness items. Data analyses were completed on March 27, 2020. EXPOSURES: Combined household income, head of household’s education level, race/ethnicity, marital status of head of household, head of household aged 65 years or older, presence of children or a household member with a disability, and region. MAIN OUTCOMES AND MEASURES: Nine actionable preparedness items, such as having an emergency carry-on kit (resource), food and water stockpiles (resource), and alternative communication plans and meeting locations (action). Items were summed for the measures of overall, resource-based, and action-based preparedness, with preparedness defined as meeting at least half of the criteria. RESULTS: Among 16 725 included households, 9103 household heads were men (54.4%), 11 687 were married (69.9%), and 10 749 (66.1%) had some college education or higher. In all, 1969 household heads (11.8%) were black, while 2696 were Hispanic/Latino (16.1%); 3579 household heads (21.4%) were 65 years or older. A total of 7163 households (42.8%) included children, and 3533 households (21.2%) included a person with a disability. Households were more likely to fulfill at least half of the criteria for resource-based preparedness (10 950 households [65.5%]) than for action-based preparedness (6876 households [41.1%]). Wealthy households and those with household heads aged 65 years or older were more likely to fulfill at least half of resource-based items (wealthy households: adjusted odds ratio [aOR] by logged income, 1.18 [95% CI, 1.13-1.22]; household heads age ≥65 years: aOR, 1.42 [95% CI, 1.29-1.55) but less likely to fulfill action-based items (wealthy households: aOR: 0.96 [95% CI, 0.93-0.99]; household heads age ≥65 years: aOR, 0.92 [95% CI, 0.84-0.99]). Households with black household heads were more likely to fulfill items directly related to emergencies (carry-on emergency kit: aOR, 1.26 [95% CI, 1.14-1.39]; alternative communication plan: aOR, 1.55 [95% CI, 1.39-1.72]; alternative meeting location: aOR, 1.18 [95% CI, 1.07-1.31]) but less likely to fulfill resource-based items (at least half of resource items: aOR, 0.89 [95% CI, 0.80-0.99]). CONCLUSIONS AND RELEVANCE: These findings suggest that types of preparedness vary by household characteristics. Targeted strategies are needed to promote preparedness across communities.

For the OLS regression, * P<0.01, ** P<0.001. Outcome is overall preparedness, on a 0 to 9 scale. The logistic regression specification, used in the manuscript, represents fulfilling 5 or more (at least half) of items. The ordinary least squares specification uses overall preparedness as a continuous outcome and includes P-values to denote statistical significance. The multinomial logistic regression examines three categories of preparedness (low (0-3 items), medium (4-6 items), and high (7-9 items)), with low as the reference category. For each coefficient, there are two odds ratios corresponding to level 2 versus level 1 (i.e., medium versus low) and level 3 versus level 1 (i.e., high versus low). Abbreviations: adjusted odds ratio (aOR), confidence interval (CI). Preparedness is defined by fulfilling at least half of items. The study sample is limited to participants who responded to the 9 questions on specific preparedness items. To adjust for possible nonresponse bias, this replicates Table 2 from the manuscript using a logistic regression weighting method. Notes: Pearson correlation coefficients. P-values are not provided because with such a large sample, many associations are statistically significant although not meaningful qualitatively.

Median (interquartile range) 60,000 (80,640) disability, N (%) 5,971 (22.0) Census region: South, N (%)

Household availability of resource-based disaster preparedness items Vehicle available for evacuation

Action-based preparedness

For fulfillment of overall preparedness, resource-based items, and actionbased items, preparedness is defined as meeting at last half of the items in that category. The study sample is limited to participants who responded to the 9 questions on specific preparedness items. To adjust for possible nonresponse bias