- Split View
-
Views
-
Cite
Cite
Maria Pia Fantini, Elisa Stivanello, Laura Dallolio, Marzia Loghi, Elena Savoia, Persistent geographical disparities in infant mortality rates in Italy (1999–2001): comparison with France, England, Germany, and Portugal, European Journal of Public Health, Volume 16, Issue 4, August 2006, Pages 429–432, https://doi.org/10.1093/eurpub/ckl009
- Share Icon Share
Abstract
Background: Infant mortality rate (IMR) is used as a population health indicator. We provide an updated description of temporal and geographical trends of IMR in Italy. Methods: Regional data on infant deaths and live births were available for France, Germany, England, Portugal (1999–2000), and Italy (1990–2001). Mortality rates including 95% CIs and time-trends were computed. Results: IMR was 4.5 per 1000 live births in 1999–2001. Between 1999–2001 and 1990–1992 both neonatal and post-neonatal mortality rates declined (P < 0.05) but not the North/South ratio. In 1999–2000 the regional variability in IMR was higher in Italy than in other European countries. Conclusion: Despite progresses in reducing IMR, geographical disparities persist within Italy.
The infant mortality rate (IMR) has been widely used as a highly sensitive (proxy) measure of population health and as a geo-political measure of environmental and socio-economic development.1–3 Infants are expected to have the most pronounced effects of low socio-economic status, as the precariousness of their resistance capacity makes them particularly susceptible to the influence of standard of living on survival. However, the validity of IMR as an indicator of population health has been debated.1,4 Multiple factors, as the rise in pre-term, low weight, and plural births, the increase in assisted reproductive technologies and changes in reporting may alter its power as health indicator.4–11 During the 20th century the IMR declined dramatically in all the developed countries.12 Italy, too, registered a remarkable reduction. However, at the beginning of the 1990 Italy still ranked tenth among the twelve countries of the European Community.13 The reduction occurred at different rates in different geographical areas within Italy generating great disparities between the North and South of the nation.14–16 The decline was attributable to both components of IMR: neonatal mortality and post-neonatal mortality. Changes in the population lifestyle occurring in the last decade may have influenced the IMR distribution across the country. It is important to provide an up to date description according to the most recent available data, because it is useful to identify population needs for health care services.1
To provide an up to date description of infant, neonatal, and post-neonatal mortality rates in Italy. To assess geographical disparities, temporal trends, and differences with other European countries.
Methods
The Italian Institute of Statistics provided the vital statistics by region and year for the French, the Portuguese, the German, and the English National Institutes of Statistics from 1999 to 2000. IMRs were calculated as the number of deaths under 1 year of age per 1000 live births during the same period. Neonatal mortality and post-neonatal rates were calculated as the number of deaths within the first four weeks and the number of subsequent deaths within the first year of life per 1000 live births. Rates were calculated for the two periods: 1990–1992 and 1999–2001, by region and by three geographical areas: North (Piemonte, Valle D'Aosta, Lombardia, Trentino Alto Adige, Veneto, Friuli Venezia-Giulia, Liguria, and Emilia-Romagna), Centre (Toscana, Marche, Umbria, and Lazio), and South (Abruzzo, Molise, Campania, Basilicata, Puglia, Calabria, Sardegna, and Sicilia). We calculated 95% confidence intervals (CIs) assuming a Poisson distribution to compare regional rates to the national rate.17,18
Geographical variability was assessed by range, standard deviation (SD), and North/South ratio. Subsequently, we computed the number of lives that could have been saved assuming the lowest regional IMR in the country.
Temporal trends were assessed using the test for trend. The variable ‘year’ was fitted into a negative binomial model and the likelihood ratio test was performed.19
The statistical package used was STATA 8.0 (Stata Corp, TX, USA).
Results
In 1999–2000 the IMR was 5.6 in England and Portugal, 4.6 in Italy and 4.4 per 1000 live births in Germany and France. Portugal and Italy had the largest variability within regions (SD 1.4 and 1.0; ranges 4.3 and 3.8 per 1000 live births, respectively). France and Germany showed the smallest variability (SD 0.8 and 0.5; ranges 1.8 and 2.1 per 1000 live births, respectively) (table 1). The neonatal mortality rate was 3.9 in England, 3.5 in Portugal, 3.4 in Italy, and 2.8 per 1000 live births in France and Germany. The highest and smallest rates were registered in Italy showing the greatest variability among the studied countries (table 1). The post-neonatal mortality rate was 2.1 in Portugal, 1.8 in England, 1.7 in Germany, 1.6 in France, and 1.2 in Italy per 1000 live births. The greatest post-neonatal mortality rate within the country variability was registered in Portugal.
Country . | Measure . | IMR . | NMR . | PNM . |
---|---|---|---|---|
England | National rate × 1000 (95% CI) | 5.64 (5.53–5.77) | 3.88 (3.78–3.99) | 1.77 (1.70–1.83) |
Ranges | 4.52–6.85 | 3.01–4.88 | 1.48–2.27 | |
Mean | 5.68 | 3.92 | 1.76 | |
SD | 0.84 | 0.63 | 0.26 | |
France | National rate × 1000 (95% CI) | 4.41 (4.30–4.51) | 2.84 (2.75–2.92) | 1.57 (1.51–1.63) |
Ranges | 3.40–5.25 | 1.85–3.38 | 0.95–2.03 | |
Mean | 4.17 | 2.75 | 1.54 | |
SD | 0.77 | 0.35 | 0.28 | |
Germany | National rate × 1000 (95% CI) | 4.46 (4.40–4.61) | 2.79 (2.74–2.91) | 1.66 (1.62–1.75) |
Ranges | 3.55–5.67 | 2.06–3.53 | 1.37–2.14 | |
Mean | 4.46 | 2.77 | 1.68 | |
SD | 0.52 | 0.41 | 0.20 | |
Italy | National rate × 1000 (95% CI) | 4.58 (4.45–4.71) | 3.37 (3.26–3.48) | 1.20 (1.11–1.27) |
Ranges | 2.62–6.45 | 1.74–4.95 | 0.30–1.63 | |
Mean | 4.34 | 3.22 | 1.12 | |
SD | 1.03 | 0.86 | 0.28 | |
Portugal | National rate × 1000 (95% CI) | 5.58 (5.28–5.89) | 3.50 (3.27–3.75) | 2.08 (1.90–2.27) |
Ranges | 4.53–8.79 | 2.54–4.64 | 1.49–4.40 | |
Mean | 5.84 | 3.52 | 2.33 | |
SD | 1.41 | 0.74 | 0.87 |
Country . | Measure . | IMR . | NMR . | PNM . |
---|---|---|---|---|
England | National rate × 1000 (95% CI) | 5.64 (5.53–5.77) | 3.88 (3.78–3.99) | 1.77 (1.70–1.83) |
Ranges | 4.52–6.85 | 3.01–4.88 | 1.48–2.27 | |
Mean | 5.68 | 3.92 | 1.76 | |
SD | 0.84 | 0.63 | 0.26 | |
France | National rate × 1000 (95% CI) | 4.41 (4.30–4.51) | 2.84 (2.75–2.92) | 1.57 (1.51–1.63) |
Ranges | 3.40–5.25 | 1.85–3.38 | 0.95–2.03 | |
Mean | 4.17 | 2.75 | 1.54 | |
SD | 0.77 | 0.35 | 0.28 | |
Germany | National rate × 1000 (95% CI) | 4.46 (4.40–4.61) | 2.79 (2.74–2.91) | 1.66 (1.62–1.75) |
Ranges | 3.55–5.67 | 2.06–3.53 | 1.37–2.14 | |
Mean | 4.46 | 2.77 | 1.68 | |
SD | 0.52 | 0.41 | 0.20 | |
Italy | National rate × 1000 (95% CI) | 4.58 (4.45–4.71) | 3.37 (3.26–3.48) | 1.20 (1.11–1.27) |
Ranges | 2.62–6.45 | 1.74–4.95 | 0.30–1.63 | |
Mean | 4.34 | 3.22 | 1.12 | |
SD | 1.03 | 0.86 | 0.28 | |
Portugal | National rate × 1000 (95% CI) | 5.58 (5.28–5.89) | 3.50 (3.27–3.75) | 2.08 (1.90–2.27) |
Ranges | 4.53–8.79 | 2.54–4.64 | 1.49–4.40 | |
Mean | 5.84 | 3.52 | 2.33 | |
SD | 1.41 | 0.74 | 0.87 |
Country . | Measure . | IMR . | NMR . | PNM . |
---|---|---|---|---|
England | National rate × 1000 (95% CI) | 5.64 (5.53–5.77) | 3.88 (3.78–3.99) | 1.77 (1.70–1.83) |
Ranges | 4.52–6.85 | 3.01–4.88 | 1.48–2.27 | |
Mean | 5.68 | 3.92 | 1.76 | |
SD | 0.84 | 0.63 | 0.26 | |
France | National rate × 1000 (95% CI) | 4.41 (4.30–4.51) | 2.84 (2.75–2.92) | 1.57 (1.51–1.63) |
Ranges | 3.40–5.25 | 1.85–3.38 | 0.95–2.03 | |
Mean | 4.17 | 2.75 | 1.54 | |
SD | 0.77 | 0.35 | 0.28 | |
Germany | National rate × 1000 (95% CI) | 4.46 (4.40–4.61) | 2.79 (2.74–2.91) | 1.66 (1.62–1.75) |
Ranges | 3.55–5.67 | 2.06–3.53 | 1.37–2.14 | |
Mean | 4.46 | 2.77 | 1.68 | |
SD | 0.52 | 0.41 | 0.20 | |
Italy | National rate × 1000 (95% CI) | 4.58 (4.45–4.71) | 3.37 (3.26–3.48) | 1.20 (1.11–1.27) |
Ranges | 2.62–6.45 | 1.74–4.95 | 0.30–1.63 | |
Mean | 4.34 | 3.22 | 1.12 | |
SD | 1.03 | 0.86 | 0.28 | |
Portugal | National rate × 1000 (95% CI) | 5.58 (5.28–5.89) | 3.50 (3.27–3.75) | 2.08 (1.90–2.27) |
Ranges | 4.53–8.79 | 2.54–4.64 | 1.49–4.40 | |
Mean | 5.84 | 3.52 | 2.33 | |
SD | 1.41 | 0.74 | 0.87 |
Country . | Measure . | IMR . | NMR . | PNM . |
---|---|---|---|---|
England | National rate × 1000 (95% CI) | 5.64 (5.53–5.77) | 3.88 (3.78–3.99) | 1.77 (1.70–1.83) |
Ranges | 4.52–6.85 | 3.01–4.88 | 1.48–2.27 | |
Mean | 5.68 | 3.92 | 1.76 | |
SD | 0.84 | 0.63 | 0.26 | |
France | National rate × 1000 (95% CI) | 4.41 (4.30–4.51) | 2.84 (2.75–2.92) | 1.57 (1.51–1.63) |
Ranges | 3.40–5.25 | 1.85–3.38 | 0.95–2.03 | |
Mean | 4.17 | 2.75 | 1.54 | |
SD | 0.77 | 0.35 | 0.28 | |
Germany | National rate × 1000 (95% CI) | 4.46 (4.40–4.61) | 2.79 (2.74–2.91) | 1.66 (1.62–1.75) |
Ranges | 3.55–5.67 | 2.06–3.53 | 1.37–2.14 | |
Mean | 4.46 | 2.77 | 1.68 | |
SD | 0.52 | 0.41 | 0.20 | |
Italy | National rate × 1000 (95% CI) | 4.58 (4.45–4.71) | 3.37 (3.26–3.48) | 1.20 (1.11–1.27) |
Ranges | 2.62–6.45 | 1.74–4.95 | 0.30–1.63 | |
Mean | 4.34 | 3.22 | 1.12 | |
SD | 1.03 | 0.86 | 0.28 | |
Portugal | National rate × 1000 (95% CI) | 5.58 (5.28–5.89) | 3.50 (3.27–3.75) | 2.08 (1.90–2.27) |
Ranges | 4.53–8.79 | 2.54–4.64 | 1.49–4.40 | |
Mean | 5.84 | 3.52 | 2.33 | |
SD | 1.41 | 0.74 | 0.87 |
From 1990–1992 to 1999–2001, the Italian IMR decreased by 44% from 8.0 (95% CI 7.85–8.12) to 4.5 (95% CI 4.41–4.62) per 1000 live births. In 1990–1992 the IMR ranged from 4.7 (95% CI 3.91–5.62) in Friuli Venezia-Giulia to 10.3 (95% CI 9.90–10.80) in Sicilia. In 1999–2001, the lowest IMR was 2.6 (95% CI 2.06–3.29) in Friuli Venezia-Giulia, while the highest was 6.4 (95% CI 6.07–6.86) in Sicilia. The reduction appeared in all geographical areas (figure 1) and regions. The test for trend showed a significant decline (P < 0.05) in all regions except Valle D'Aosta (P = 0.35). Variability measured by the standard deviation decreased by 39% from 1.7 to 1.0; however, the North/South ratio remained unchanged (0.6). The number of lives that could have been saved assuming all regions having the lowest rate registered in the country (2.6 per 1000 live births) is 3067.
During 1990–1992, neonatal mortality occurred at a rate of 6.2 (95% CI 6.06–6.30) per 1000 live births ranging from 3.3 (95% CI 2.67–4.11) in Friuli Venezia-Giulia to 8.13 (95% CI 7.74–8.54) in Sicilia. In 1999–2000 the rate declined from 46 to 3.3% (95% CI 3.25–3.43) representing 77% of IMR and ranging from 1.7 (95% CI 1.28–2.29) in Friuli Venezia-Giulia to 4.9 (95% CI 4.61–5.30) in Sicilia, per 1000 live births. The reduction in neonatal mortality rate was registered in all geographical areas (figure 1) and regions, with a significant time trend (P > 0.05) reported in all regions except Umbria and Valle D'Aosta. The standard deviation decreased by 41% from 1.42 to 0.84 while the North/South ratio remained unchanged (ratio = 0.6).
The post-neonatal mortality rate was 1.8 (95% CI 1.79–1.92) per 1000 live births in 1990–1992 ranging from 1.14 (95% CI 0.82–1.54) in Marche to 2.39 (95% CI 1.98–2.85) in Sardegna. The overall reduction was 57% achieving a rate of 1.2 (95% CI 1.13–1.23) per 1000 in the period 1999–2001, varying from 0.3 (95% CI 0.001–1.64) in Valle D'Aosta to 1.6 (95% CI 1.32–1.7) in Puglia. The standard deviation of the distribution of rates across regions was 0.36 in 1990–1992 and 0.29 in 1999–2001 and the North/South ratio was 0.8 and 0.7, respectively.
Discussion
This study provides an up to date description of infant, neonatal, and post-neonatal mortality rates throughout Italy and comparisons over a 12 year period. Despite most regions registering a dramatic decrease in IMR from 1990 to 2001, great geographical disparities persist in Italy. The North/South ratio has remained substantially unchanged since the 1960s.13 We estimated that 3067 children could have been saved given that all Italian regions have the lowest rate registered in the country. It is evident that there is room for improvement. Contrary to previous studies our findings show that the regional variability in infant mortality is largely determined by the variability in neonatal mortality and not by the variability in post-neonatal mortality.15,20 The remarkable and stable geographical pattern cannot be attributed to biological factors; beyond environmental deteminants, differences in maternal and neonatal care should be carefully considered at the root of the inequalities in infant mortality, and in neonatal mortality particularly.21–24 We invited regional health authorities and public health officials to join a preliminary discussion on the topic that was held in Bologna in April 2005. Differences between North and South in the distribution and organization of intensive care units and in the organization of prenatal and perinatal care were discussed and identified as likely determinants of disparities. However, the analyses of the inequalities in IMR is beyond the scope of our study and with the available data it is not possible to understand how far infant health is affected by these and other more distal determinants. Health inequalities between Northern and Southern regions exist for other problems as well. Explanations for such inequalities are complex and involve multiple factors. Strong historical and structural differences exist within Italian regions and they involve social, cultural, and economic aspects; the North is performing better than the South according to various health status indicators and the Centre is in an intermediate position. Various health service indicators show also a geographical gradient with respect to quality and appropriateness of care.25,26 Future research in this field should focus on the identification of the determinants of IMR. To achieve this scope, observational studies need to be conducted to assess the association between health care policies and programmes and infant death.
In conclusion, there remains a significant challenge for public health and health care officials to reduce disparities of infant mortality across Italian regions. According to our study the room for improvement in reducing IMR relies mainly on the prevention of neonatal mortality.
Temporal and geographical trends of IMR in Italy are described and compared with other European countries.
All Italian regions registered a remarkable decline in IMR during the period 1990–2001.
Interregional disparities persist and are larger than in other European countries with neonatal mortality accounting for most of the disparities.
There remains an important challenge to reduce disparities of infant mortality across Italian regions.
Room for further reduction in IMR relies mainly on the prevention of neonatal mortality.
The authors wish to thank Pietro Folino for his contribution and comments during the study and John Wilkinson, Helmut Brand, Marinho Falcao, and Bernard Ledesert who provided data from England, Germany, Portugal, and France, respectively.
References
Murray CJL, Salomon JA, Mathers C. A critical examination of summary measures of population health.
Reidpath DD, Allotey P. Infant mortality rate as an indicator of population health.
Murray CJL, Rethinking DALYs. In: Murray CJL, Lopez AD, editors. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Disease, Injuries, and Risk Factors in1990 and Projected to 2020. Cambridge, MA: Harvard School of Public Health,
Kochanek KD, Martin JA. Supplemental analyses of recent trends in infant mortality.
Fiscella K. Does prenatal care improve birth outcomes? A critical review.
Reynolds MA, Schieve LA, Martin JA, et al. Trends in multiple births conceived using assisted reproductive technology, United States, 1997–2000.
Tucker J, McGuire W. Epidemiology of preterm birth.
McDorman MF, Martin JA, Mathews TJ, et al. Explaining the 2001–2002 infant mortality increase: data from the linked birth/infant death data set.
Gisselmann MD. Education, infant mortality, low birth weight in Sweden 1973–1990: emergence of the low birth weight paradox.
Parazzini F, Levi F, Lucchini F, La Vecchia C. Trends in neonatal and infant mortality in five continents.
Piccardi P, Cattaruzza MS, Osborn JF. A century of infant mortality in Italy: the years 1870–90.
Osborn JF, Spinelli A, Cattaruzza MS. A novel method of comparing regional trends in infant mortality applied to Italy 1950–89. In: Zaba B, Blacker J, editors. Brass Tacks: Essays in Medical Demography. A tribute to the Memory of Professor William Brass. London: Athlone Press,
Parazzini F, La Vecchia C, Negri E. Descriptive epidemiology of perinatal and infant mortality in various Italian geographic areas.
Lauria L, De Stavola BL. A district-based analyses of stillbirth and infant mortality rates in Italy: 1989–93.
Lack N, Zeitlin J, Krebs L, et al. Methodological difficulties in the comparison of indicators of perinatal health across Europe.
Martuzzi M, Hills M. Estimating the degree of heterogeneity between event rates using likelihood.
Nolte E, Koupilovà I, McKee M. Temporal and spatial pattern of infant mortality in Germany after unification.
Lemma P, Costa G, Demarca M, et al. Social differences in infant mortalità in a longitudinal Turin study.
Hack M, Fanaroff AA. Outcomes of children of extremely low birthweight and gestational age in the 1990's.
Parazzini F, Pirotta N, La Vecchia C, et al. Determinants of perinatal and infant mortality in Italy.
Bonati M, Campi R. What can we do to improve child health in Southern Italy?
ISTAT Istituto Nazionale di Statistica. Rapporto annuale. 2004: Dati statistici. Rome: Istituto Nazionale di Statistica,
Comments