15

would be possible to quantify, not all subjects are travelling with the same speed (depending on external factors like mode of transportation, weather or administrative regulations). Shannon and colleagues argue that what actually counts is the total effort made to reach a health facility and suggest that perceived travel time could be a good approximation of this effort [97]. Travel time is then directly related to the second delay while the perception of the effort, including all factors influencing travel, is significant for the first delay of care-seeking. Living area has also been suggested to be a part of physical accessibility [77]. Usually comparisons between urban and rural areas are made, but also differences between regions have been investigated. Most common is to find a higher use of health services in urban areas [106, 107], much explained by the distance-decay effect, but also as a function of demographic distributions of education, wealth and ethnicity. Geographical Information Systems (GIS) can be used to investigate the geographical components of physical accessibility, such as distance and area. The use of GIS is becoming more common in health research as the technique develops [108]. It can be used both for surveillance, for example the monitoring of such infectious diseases as malaria and acute respiratory infections (ARI) [108-110], and for planning, using cluster analyses [111] and other mapping techniques.The UNICEF Innocenti Research Centre in Florence, Italy, was established in 1988 to strengthen the research capability of the United Nations Children's Fund (UNICEF) and to support its advocacy for children worldwide. The Centre (formally known as the International Child Development Centre) helps to identify and research current and future areas of UNICEF's work. Its prime objectives are to improve international understanding of issues relating to children's rights and to help facilitate the full implementation of the United Nations Convention on the Rights of the Child in both industrialized and developing countries. The Centre's publications are contributions to a global debate on child rights issues and include a wide range of opinions. For that reason, the Centre may produce publications that do not necessarily reflect UNICEF policies or approaches on some topics. The views expressed are those of the authors and are published by the Centre in order to stimulate further dialogue on child rights. The Centre collaborates with its host institution in Florence, the Istituto degli Innocenti, in selected areas of work. Core funding for the Centre is provided by the Government of Italy, while financial support for specific projects is also provided by other governments, international institutions and private sources, including UNICEF National Committees. The opinions expressed in this paper are those of the authors and do not necessarily reflect the policies or views of UNICEF. 1 Abstract This paper examines the measurement of infant mortality in the countries of Central and Eastern Europe and the Commonwealth of Independent States. There are worrying indications that official infant mortality counts, based on administrative data, may understate the true gravity of the problem in 15 countries in the region, including 11 out of 12 CIS countries, and four countries in South Eastern Europe. In the case of eight CIS countries in the Caucasus and Central Asia plus Romania, the evidence is strongest, not least because surveys that allow independent estimation of infant mortality have been carried out there. In the case of the remaining six countries, the evidence is more circumstantial, and based on inconsistencies within the official data themselves, combined with information on how live births are defined. However, the paper also finds that surveys are rather blunt instruments, and that the confidence intervals that surround estimates from these surveys are often large. Further work needs to be done, not only to improve the collection of official statistics on births and infant deaths in many countries across the region, but also to improve the effectiveness of surveys as a measurement tool. Introduction As the communist period recedes further in time, it is now clear that many of the 27 post-communist countries of Central and Eastern Europe and the Commonwealth of Independent States face continued challenges to improve the well-being of their citizens. Yet even as national incomes have fallen and poverty has increased, official data in nearly every CEE/CIS country suggest that infant mortality rates decreased through the 1990s, in many cases to within shouting distance of rates in Western Europe. Georgia, for example, posted an official infant mortality rate of 12 deaths per 1,000 live births in 2001, similar to that in Latvia, and only 5 deaths per 1,000 more than in Portugal. Is this credible? For many countries, the short answer is ‘no’. This paper identifies 15 countries – 11 out of 12 CIS countries, plus 4 countries in South Eastern Europe – where official infant mortality rates appear to understate the true picture. In some of these countries (including Georgia), survey data suggest that real rates may be more than twice the official figure. This is a worrying finding. The infant mortality rate is a key indicator of progress towards Millennium Development Goals.1 Therefore, its accurate measurement is important. Moreover, evidence that official data understate the true picture must be taken as a sign that this international development target is in danger of not being reached. 1 There are eight major Millennium Development Goals, the fourth one of which is to reduce under five mortality by two thirds between 1990 and 2015. In general, four in five deaths of children aged under 5 are in fact infant deaths. 2 Interrogation of official data in CEE/CIS countries is not new. The research in this paper builds on quite a large body of literature on infant mortality statistics in the Soviet Union, starting with Davis and Feshback (1980), running through to Anderson and Silver (1986), Velkoff and Miller (1995), Kingkade and Sawyer (2001) and UNICEF (2003). This existing literature lacks two elements: first, it does not present a clear picture of underestimation of infant mortality across the entire CEE/CIS region in the post-communist era, or a clear examination of the reasons that might lie behind it. Second, a rigorous analysis of the reliability of the major source of alternative estimates of infant mortality in the region – household surveys where women are asked about their reproductive histories – is also missing. It is in these two areas that this paper makes a contribution. Why do official data understate the true picture in 15 countries? In most of these countries, the way a live birth is currently defined may reduce the infant mortality count. Mis-reporting, for