Watson, Samuel I. and Lilford, Richard (2017) Mortality decrease according to socioeconomic groups. The Lancet, 389 (10081). pp. 1794-1795. doi:10.1016/S0140-6736(17)31157-1 ISSN 0140-6736.

Preview

PDF WRAP-correspondance-Watson-2017.pdf - Accepted Version - Requires a PDF viewer. Download (362Kb) | Preview

Request Changes to record.

Abstract

Enrique Regidor and colleagues' Article analysed the effect of the economic crisis in Spain on age-adjusted mortality for different socioeconomic groups. We wish to point out that the results are consistent with a hypothesis of no difference between people of different socioeconomic groups, despite the authors' conclusions.

First, the economic crisis might have led people to change socioeconomic status. The authors used household floor space and the number of cars as proxies for socioeconomic status. However, the crisis might have caused households to downsize or to sell a car, of which the expected effect would be a reduction in mortality in low socioeconomic groups.

Second, in relative terms, the reduction in all-cause mortality was approximately the same between socioeconomic groups, at around 15%. Different absolute rates between socioeconomic groups might have simply reflected differences in the distribution of disease burden.

Third, in Spain, average health-care expenditure per capita during 2004–07 was about US$2300, whereas average health-care expenditure per capita during 2008–11 was $3000. This increase might explain the observed declines in all-cause mortality, especially when our second point is taken into consideration.

Fourth, a common statistical error is to assume that the difference between significant and not significant is itself statistically significant. Although within-group changes were tested, no hypothesis tests were done to compare changes between different groups. Without these comparisons, conclusions cannot be made without the risk of falling foul of the error of inferring that a difference in significance is itself significant.

To make accurate inferences from observational data, appropriate caution must be taken to select, test, and compare appropriate models. Without this rigour, few conclusions are tenable.

Second, in relative terms, the reduction in all-cause mortality was approximately the same between socioeconomic groups, at around 15%. Different absolute rates between socioeconomic groups might have simply reflected differences in the distribution of disease burden.

Third, in Spain, average health-care expenditure per capita during 2004–07 was about US$2300, whereas average health-care expenditure per capita during 2008–11 was $3000.2 This increase might explain the observed declines in all-cause mortality, especially when our second point is taken into consideration.

Fourth, a common statistical error is to assume that the difference between significant and not significant is itself statistically significant.3 Although within-group changes were tested, no hypothesis tests were done to compare changes between different groups. Without these comparisons, conclusions cannot be made without the risk of falling foul of the error of inferring that a difference in significance is itself significant.

To make accurate inferences from observational data, appropriate caution must be taken to select, test, and compare appropriate models. Without this rigour, few conclusions are tenable.

Item Type:	Journal Article
Subjects:	H Social Sciences > HB Economic Theory
Divisions:	Faculty of Science, Engineering and Medicine > Medicine > Warwick Medical School > Health Sciences > Population, Evidence & Technologies (PET) Faculty of Science, Engineering and Medicine > Medicine > Warwick Medical School
Library of Congress Subject Headings (LCSH):	Mortality -- Economic aspects -- Spain, Social status -- Health aspects
Journal or Publication Title:	The Lancet
Publisher:	Lancet Publishing Group
ISSN:	0140-6736
Official Date:	6 May 2017
Dates:	Date Event 6 May 2017 Published
Volume:	389
Number:	10081
Page Range:	pp. 1794-1795
DOI:	10.1016/S0140-6736(17)31157-1
Status:	Not Peer Reviewed
Publication Status:	Published
Access rights to Published version:	Open Access (Creative Commons)
Date of first compliant deposit:	19 February 2019
Date of first compliant Open Access:	20 February 2019

Request changes or add full text files to a record

Repository staff actions (login required)

View Item

Downloads