The following Table revises Table 3, which appeared on page 153 of our recent report.1 Table presented here incorporates two sets of changes: 1) We corrected an error in the rural-urban category that had affected the results for that variable and the model. 2) In a letter to the editor, Hendryx and Ahern challenged our findings and they also posted their dataset for these variables.2 As discussed in our reply to their letter, we determined that there were differences between our datasets for seven other variables, of which four (adjusted mortality rate; high school education rate; college education rate; and coal mining (yes/no)) were significantly different.3 We have not determined which of the datasets, ours or theirs, is “more correct” for six of those variables, but their coal mining variable is clearly incorrect. To avoid further disputes, the following Table incorporates the WVU data for adjusted mortality, high school education and college education rates.
The analytical results are quantitatively different from our original report, but qualitatively unchanged. Coal mining remains not significantly associated with mortality. We regret any inconvenience caused by our error in assigning the Rural-Urban category variables.
1. Borak J, Saipante-Zaidel C, Slade MD, Fields CA. Mortality disparities in Appalachia: reassessment of major risk factors. J Occup Environ Med. 2012;54:146–156.
2. Hendryx M, Ahern M. Reply to Borak et al: mortality disparities in Appalachia major risk factors. J Occup Environ Med. 2012;54:768–769.
3. Borak J, Slade MD, Allen RR, Saipante-Zaidel C, Fields CA. Ecological bias and data entry errors: a reply to Hendryx and Ahern. J Occup Environ Med. 2012;54:770–773.