A misunderstanding of the time series of the healthy life years from Eurostat: a break in the data source rather than a epidemiological dramatic trend

Paolo Piergentili — 2013-01-10T10:00:34Z

Trends in population data are very stable, and unexpected changes are very rare and should be always arise doubts. The most usual reason for oscillations in data values are changes in the Information system. This is the case also for Healthy Life Years (HLY). The document of SANCO, the General Directorate of the European Commission which is devoted to oversee public health issues, explains as in 2004 a break occurred in the HLY time series since the way of calculating them has changed (1). The LHY are calculated according to the Sullivan guidelines (2), and are based on the Life Expectancy statistics, which are well established and easy to calculate, and the on prevalence measures of the age specific proportion of population with and without disabilities, which is very difficult to estimate. Such estimate is derived from the activity limitation data assessed by health surveys. Since 2004 activity limitation data come from the Eurostat survey Statistics on Income and Living Conditions (EU-SILC http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2003:298:0001:0022:EN:PDF ), that contains a health module including a global question on activity limitation, defined as the General Activity Limitation Indicator (GALI). Before 2004, HLY were calculated with data on self-perceived disability coming from the Eurostat survey European Community Household Panel (ECHP). This change of the data source for calculating the prevalence of disability in 2004 created a break in series. In fact, the value of data regarding Italy since 2004 oscillates, meaning that the new data source is not stable yet. The value of the indicator for females in the 2001 ¿ 2003 years was respectively 73.0, 73.9 and 74.4, but the table warns that the figures are an estimate and not an actual data (3). From 2004 on the figures are: 71.0, 67.8, 64.7, 62.5, 61.9, 62.6 and 67.3 in 2010. There is a warning of break in the table for year 2004, 2006, 2007 and 2010 (4) and another warning is issued in the metadata page accompanying the table of the SILC data (5). As a consequence, data since 2004 are not a reliable source of information, and probably a few years are needed before the trend becomes stable again.Thus, the public health and social problems outlined by the authors may be possibly true, but should not be based on HLY statistics. It is important to underpin this concept since the article by Gennaro at coll. has been a source of information for, and supported strong polemics in a number of websites (6-9) (even the one of a political movement10).

correction

valerio gennaro — 2013-01-10T09:59:51Z

Dear Editor,
please, note that in our paper is necessary to replace "2005" with "2008" in the title and in the abstract as reported in the original file and also documented in the figures.
Sorry for this problem.

Best wishes.
Valerio Gennaro

comment

Federico Allemand — 2011-01-12T16:45:48Z

Reference number 3 is:
Perlman JM: Summary proceedings from the neurology group on hypoxic-ischemic encephalopathy.Pediatrics 2006 Mar;117(3 Pt 2):S28-33
Regards
The Authors

minor corrections for tables 4 and 5

PEACE IBO OPARA — 2011-01-12T16:45:18Z

Table 4:
figures for hypoglycemia should be 30 and 3 not 26 and 3
figures for neonatal jaundice should be 27 and 4 not 25 and 4
Table 5:
figures for hypoglycemia should be 9 and 18 not 8 and 18
figures for jaundice should be 11 and 16 not 11 and 14

Italian Journal of Pediatrics - Latest Comments

A misunderstanding of the time series of the healthy life years from Eurostat: a break in the data source rather than a epidemiological dramatic trend

correction

comment

minor corrections for tables 4 and 5