Paul Welaga Cornelius Debpuur Cornelius Debpuur Timothy Awine 10 - - PowerPoint PPT Presentation

Paul Welaga Cornelius Debpuur Cornelius Debpuur Timothy Awine

10th INDEPTH AGM 27th -30th September 2010 Accra

Background

Evidence on the relationship between birth-spacing and

child survival is still not clear

Recommendations for birth spacing are based on

information that was available several years ago (WHO Technical report on birth spacing, 2005)

Some publications recommend waiting at least 2–3 years

between pregnancies to reduce infant and child mortality

Recent studies suggested that longer birth spacing, 3 to 5

years might be more advantageous to child survival (S. O. Rutstein, 2005)

Objectives

To compare infant mortality rates between index children with

preceding birth interval of 3 to 5 and 2 to 3 years

To examine differences in neonatal mortality rates between

index children with preceding birth interval of 3 to 5 and 2 to 3 years

To examine differences in under two mortality rates between To examine differences in under two mortality rates between

index children with preceding birth interval of 3 to 5 and 2 to 3 years

To examine differences in under-five mortality rates between

index children with preceding birth interval of 3 to 5 and 2 to 3 years

Methods

Children of second or higher order singleton live birth

(index children) born in the DSS area between 1st January 2002 to 31st December 2007.

Preceding birth was defined as the difference in months

between the birthdates of the index and immediately preceding child

Preceding birth interval was categorised into four groups;

<24, 24-35, 36-59, and 60+ months

All children registered into the system were taken into

account in calculating preceding birth interval

Data

Data for the analysis come from the Navrongo Health

and Demographic Surveillance System (NHDSS)

NHDSS monitors a population of about 150,000 under

surveillance

• Demographic events such as births, deaths, in and out

migrations, pregnancies are updated every three months

NHDSS also collects data on vaccination, education,

verbal autopsy and socioeconomic status

Study setting – Fertility trends

4.00 5.00 6.00

TFR

0.00 1.00 2.00 3.00 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 TFR

Analysis

To examine the effect of preceding birth interval on post-

neonatal, infant and under-five mortality, Cox proportional regression models were used.

Logistic regression models were used to assess the effect on

neonatal mortality.

Mortality rates were expressed per 1000 live births for neonatal Mortality rates were expressed per 1000 live births for neonatal

mortality and per 1000 child years of risk for post-neonatal, infant and under-five mortality

Potential confounders controlled for: mother’s age at birth, season of

birth, mother’s education, place of delivery, socioeconomic status using wealth index, survival status of immediately preceding sibling, sex, year of birth and birth order

Results

In all 15424 under-five children born between 1st

January 2002 and 31st December 2007 were included in the analysis

1,272 deaths were recorded Infant mortality rate of 63.4 per 1000 person

years

Background Characteristics

Variable N (%) deaths (% dead) Preceding birth interval <24 months 24-35 36-59 60+ 888 (5.8) 3,168 (20.5) 7,894 (51.2) 3,475 (22.5) 82 (9.2) 287 (9.1) 621 (7.9) 282 (8.1) Sex Female Male 7,637 (49.5) 7,787 (50.5) 587 (7.7) 685 (8.8) Mother’s age at birth <20 20-29 30-39 40-49 338 (2.2) 6,637 (43.0) 6,315 (40.9) 2,134 (13.8) 27 (8.0) 488 (7.4) 554 (8.8) 203 (9.5) Mother’s education No education Primary/JSS Secondary+ Missing 9,855 (63.9) 4,050 (26.3) 493 (3.2) 1,026 (6.7) 853 (8.7) 281 (6.9) 19 (3.9) 119 (11.6)

Relative risk for univariate and multivariate models for under two mortality by preceding birth interval

Univariate model Multivariate model Varaible OR [95% CI] (unadjusted) P-value OR [95% CI] (adjusted) P-value Preceding birth interval < 24 months 24-35 1.27 [1.0 – 1.61] 1.18 [1.02 – 1.37] 1 0.053 0.029 1.25 [0.97 – 1.63] 1.18 [1.01 – 1.38] 1 0.089 0.039 24-35 36-59* 60+ 1 1.07 [0.92 – 1.24] 0.403 1 1.09 [0.93 – 1.30] 0.266 Mother’s education None* Primary/Junior Sec Secondary+ missing 1 0.80 [0.69 – 0.93] 0.50 [0.31 – 0.79] 1.49 [1.22 – 1.83] 0.003 0.003 <0.001 1 0.78 [0.67 – 0.92] 0.48 [0.29 – 0.78] 1.21 [0.97 – 1.52] 0.003 0.003 0.094

Relative risk for univariate and multivariate models for infant mortality by preceding birth interval

Univariate model Multivariate model Varaible HR [95% CI] (unadjusted) P-value HR [95% CI] (adjusted) P-value Preceding birth interval < 24 months 24-35 36-59* 1.32 [1.01 – 1.74] 1.18 [1.00 – 1.40] 1 1.11 [0.93 -1.31] 0.043 0.057 0.241 1.29 [0.97 - 1.72] 1.17 [0.98 – 1.40] 1 1.12 [0.92 – 1.35] 0.078 0.092

• 0.257

36-59* 60+ 1.11 [0.93 -1.31] 0.241 1.12 [0.92 – 1.35] 0.257 Mother’s education None* Primary/Junior Sec Secondary+ missing 1 0.79 [0.67 – 0.93] 0.61 [0.37 – 0.98] 1.60 [1.28 – 2.00] 0.005 0.042 <0.001 1 0.76 [0.63 – 0.91] 0.55 [0.33 – 0.92] 1.24 [0.97 – 1.58] 0.004 0.022 0.084

Relative risk for univariate and multivariate models for post neonatal mortality by preceding birth interval

Univariate model Multivariate model Varaible HR [95% CI] (unadjusted) P-value HR [95% CI] (adjusted) P-value Preceding birth interval < 24 months 24-35 36-59* 1.43 [1.03 – 1.99] 1.47 [1.20 – 1.80] 1 1.17 [0.95 -1.44] 0.034 <0.001 0.151 1.41 [0.99 – 2.01] 1.45 [1.17 – 1.78] 1 1.21 [0.96 – 1.53] 0.055 0.001 0.100 60+ Mother’s education None* Primary/Junior Sec Secondary+ missing 1 0.78 [0.63 – 0.95] 0.59 [0.32 – 1.07] 1.75 [1.35 – 2.28] 0.016 0.080 <0.001 1 0.74 [0.60 – 0.92] 0.54 [0.29 – 1.00] 1.38 [1.04 – 1.83] 0.008 0.051 0.027

Odds Ratio for univariate and multivariate models for neonatal mortality by preceding birth interval

Univariate model Multivariate model Varaible OR [95% CI] (unadjusted) P-value OR [95% CI] (adjusted) P-value Preceding birth interval < 24 months 24-35 1.13 [0.70 – 1.83] 0.69 [0.49 – 0.98] 0.620 0.037 1.1 [0.65 – 1.85] 0.69 [0.48 – 1.01] 0.720 0.054 24-35 36-59* 60+ 0.69 [0.49 – 0.98] 1 1.0 [0.75 -1.34] 0.037 0.996 0.69 [0.48 – 1.01] 1 0.93[0.66-1.30] 0.054 0.660 Mother’s education None* Primary/Junior Sec Secondary+ missing 1 0.82 [0.61 – 1.09] 0.65 [0.29 – 1.47] 1.32 [0.86 – 2.01] 0.173 0.298 0.202 1 0.81 [0.58 – 1.13] 0.57 [0.23 – 1.41] 1.02 [0.62 – 1.67] 0.222 0.220 0.941

Relative risk for univariate and multivariate models for under- five mortality by preceding birth interval

Univariate model Multivariate model Varaible N (%) HR [95% CI] (unadjusted) HR [95% CI] (adjusted) P-value Preceding birth interval < 24 months 888 (5.8) 3,168 (20.5) 1.17 [0.93 – 1.48] 1.19 [1.03 – 1.36] 1.23 [0.96 – 1.58] 1.17 [1.01 – 1.36] 0.096 0.038 24-35 36-59* 60+ 7,893 (51.2) 3,475 (22.5) 1 1.05 [0.91 -1.21] 1 1.02 [0.88 – 1.18] 0.809 Mother’s education None* Primary/Junior Sec Secondary+ missing 9,855 (63.9) 4,050 (26.3) 493 (3.2) 2,134 (13.8) 1 0.83 [0.72 – 0.94] 0.47 [0.30 – 0.74] 1.47 [1.22 – 1.78] 1 0.78 [0.67 – 0.90] 0.40 [0.25 – 0.64] 1.24 [1.01 – 1.52] 0.001 <0.001 0.045

Discussion and Conclusion

Children with preceding birth interval of 24 to 35 months

had a higher risk of dying at infancy, post neonatal, under two and under five compared to 36 to 59 months

Preceding birth interval did not significantly affect

neonatal mortality rates. neonatal mortality rates.

The beneficial effect of longer birth spacing becomes

evident after the neonatal stage

Discussion and Conclusion

Birth interval of <24 months is generally associated

with a higher risk of dying compared to 36 to 59 months

Mother’s education contributes significantly in Mother’s education contributes significantly in

reducing child mortality

Recommendation

Generally, preceding birth interval of 36 to 59

months should be encouraged as it appears to be beneficial to child survival

Reducing births associated with preceding birth

interval of <24 months could help reduce child mortality.

Limitations

The HDSS may not have covered a complete birth history

for each woman. This however may form a minority of births for any individual woman and should not undermine the validity of the outcome

Acknowledgement

The chiefs and people of the Kassena-Nankana East and

West districts

The director and staff of the Navrongo Health Research

Centre

The Navrongo Health and Demographic Surveillance

System field and computing teams

INDEPTH Network

Thank you Thank you