Measuring coverage of maternal and child health services using routine health facility data: a Sierra Leone case study.
Data quality
Maternal and child health
Measuring coverage
RHIS data
Sierra Leone
Journal
BMC health services research
ISSN: 1472-6963
Titre abrégé: BMC Health Serv Res
Pays: England
ID NLM: 101088677
Informations de publication
Date de publication:
13 Sep 2021
13 Sep 2021
Historique:
received:
11
05
2021
accepted:
13
05
2021
entrez:
13
9
2021
pubmed:
14
9
2021
medline:
15
9
2021
Statut:
epublish
Résumé
There are limited existing approaches to generate estimates from Routine Health Information Systems (RHIS) data, despite the growing interest to these data. We calculated and assessed the consistency of maternal and child health service coverage estimates from RHIS data, using census-based and health service-based denominators in Sierra Leone. We used Sierra Leone 2016 RHIS data to calculate coverage of first antenatal care contact (ANC1), institutional delivery and diphtheria-pertussis-tetanus 3 (DPT3) immunization service provision. For each indicator, national and district level coverages were calculated using denominators derived from two census-based and three health service-based methods. We compared the coverage estimates from RHIS data to estimates from MICS 2017. We considered the agreement adequate when estimates from RHIS fell within the 95% confidence interval of the survey estimate. We found an overall poor consistency of the coverage estimates calculated from the census-based methods. ANC1 and institutional delivery coverage estimates from these methods were greater than 100% in about half of the fourteen districts, and only 3 of the 14 districts had estimates consistent with the survey data. Health service-based methods generated better estimates. For institutional delivery coverage, five districts met the agreement criteria using BCG service-based method. We found better agreement for DPT3 coverage estimates using DPT1 service-based method as national coverage was close to survey data, and estimates were consistent for 8 out of 14 districts. DPT3 estimates were consistent in almost half of the districts (6/14) using ANC1 service-based method. The study highlighted the challenge in determining an appropriate denominator for RHIS-based coverage estimates. Systematic and transparent data quality check and correction, as well as rigorous approaches to determining denominators are key considerations to generate accurate coverage statistics using RHIS data.
Sections du résumé
BACKGROUND
BACKGROUND
There are limited existing approaches to generate estimates from Routine Health Information Systems (RHIS) data, despite the growing interest to these data. We calculated and assessed the consistency of maternal and child health service coverage estimates from RHIS data, using census-based and health service-based denominators in Sierra Leone.
METHODS
METHODS
We used Sierra Leone 2016 RHIS data to calculate coverage of first antenatal care contact (ANC1), institutional delivery and diphtheria-pertussis-tetanus 3 (DPT3) immunization service provision. For each indicator, national and district level coverages were calculated using denominators derived from two census-based and three health service-based methods. We compared the coverage estimates from RHIS data to estimates from MICS 2017. We considered the agreement adequate when estimates from RHIS fell within the 95% confidence interval of the survey estimate.
RESULTS
RESULTS
We found an overall poor consistency of the coverage estimates calculated from the census-based methods. ANC1 and institutional delivery coverage estimates from these methods were greater than 100% in about half of the fourteen districts, and only 3 of the 14 districts had estimates consistent with the survey data. Health service-based methods generated better estimates. For institutional delivery coverage, five districts met the agreement criteria using BCG service-based method. We found better agreement for DPT3 coverage estimates using DPT1 service-based method as national coverage was close to survey data, and estimates were consistent for 8 out of 14 districts. DPT3 estimates were consistent in almost half of the districts (6/14) using ANC1 service-based method.
CONCLUSION
CONCLUSIONS
The study highlighted the challenge in determining an appropriate denominator for RHIS-based coverage estimates. Systematic and transparent data quality check and correction, as well as rigorous approaches to determining denominators are key considerations to generate accurate coverage statistics using RHIS data.
Identifiants
pubmed: 34511135
doi: 10.1186/s12913-021-06529-7
pii: 10.1186/s12913-021-06529-7
pmc: PMC8435364
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
547Informations de copyright
© 2021. The Author(s).
Références
PLoS Med. 2013;10(5):e1001416
pubmed: 23667336
Adv Health Care Manag. 2012;12:25-58
pubmed: 22894044
Health Policy Plan. 2009 May;24(3):217-28
pubmed: 19304786
Lancet. 2003 Sep 27;362(9389):1022-7
pubmed: 14522532
PLoS One. 2019 Jan 25;14(1):e0211265
pubmed: 30682130
Glob Health Action. 2013 Feb 13;6:20001
pubmed: 23406921
Bull World Health Organ. 2012 May 1;90(5):379-84
pubmed: 22589572
Bull World Health Organ. 2017 Oct 1;95(10):683-694
pubmed: 29147041
Lancet. 2020 Feb 22;395(10224):605-658
pubmed: 32085821
J Infect Dis. 2014 Nov 1;210 Suppl 1:S523-30
pubmed: 25316876
Bull World Health Organ. 2005 Aug;83(8):578-83
pubmed: 16184276
Malar J. 2017 Aug 17;16(1):344
pubmed: 28818071
BMC Med Inform Decis Mak. 2014 May 13;14:40
pubmed: 24886567
BMJ Glob Health. 2019 Sep 29;4(5):e001849
pubmed: 31637032
Lancet Glob Health. 2015 Jul;3(7):e396-409
pubmed: 26087986
BMJ Glob Health. 2018 Oct 2;3(5):e000807
pubmed: 30294456
Health Syst Reform. 2017 Apr 3;3(2):74-79
pubmed: 31514675