Change search
ReferencesLink to record
Permanent link

Direct link
Demonstrating the robustness of population surveillance data: implications of error rates on demographic and mortality estimates
Umeå University, Faculty of Medicine, Department of Public Health and Clinical Medicine, Epidemiology and Global Health.
Umeå University, Faculty of Medicine, Department of Public Health and Clinical Medicine, Epidemiology and Global Health.ORCID iD: 0000-0001-5474-4361
Addis Continental Inst Publ Hlth, Addis Ababa, Ethiopia.
2008 (English)In: BMC Medical Research Methodology, ISSN 1471-2288, Vol. 8, Article nr 13- p.Article in journal (Refereed) Published
Abstract [en]

BACKGROUND: As in any measurement process, a certain amount of error may be expected in routine population surveillance operations such as those in demographic surveillance sites (DSSs). Vital events are likely to be missed and errors made no matter what method of data capture is used or what quality control procedures are in place. The extent to which random errors in large, longitudinal datasets affect overall health and demographic profiles has important implications for the role of DSSs as platforms for public health research and clinical trials. Such knowledge is also of particular importance if the outputs of DSSs are to be extrapolated and aggregated with realistic margins of error and validity.

METHODS: This study uses the first 10-year dataset from the Butajira Rural Health Project (BRHP) DSS, Ethiopia, covering approximately 336,000 person-years of data. Simple programmes were written to introduce random errors and omissions into new versions of the definitive 10-year Butajira dataset. Key parameters of sex, age, death, literacy and roof material (an indicator of poverty) were selected for the introduction of errors based on their obvious importance in demographic and health surveillance and their established significant associations with mortality. Defining the original 10-year dataset as the 'gold standard' for the purposes of this investigation, population, age and sex compositions and Poisson regression models of mortality rate ratios were compared between each of the intentionally erroneous datasets and the original 'gold standard' 10-year data.

RESULTS: The composition of the Butajira population was well represented despite introducing random errors, and differences between population pyramids based on the derived datasets were subtle. Regression analyses of well-established mortality risk factors were largely unaffected even by relatively high levels of random errors in the data.

CONCLUSION: The low sensitivity of parameter estimates and regression analyses to significant amounts of randomly introduced errors indicates a high level of robustness of the dataset. This apparent inertia of population parameter estimates to simulated errors is largely due to the size of the dataset. Tolerable margins of random error in DSS data may exceed 20%. While this is not an argument in favour of poor quality data, reducing the time and valuable resources spent on detecting and correcting random errors in routine DSS operations may be justifiable as the returns from such procedures diminish with increasing overall accuracy. The money and effort currently spent on endlessly correcting DSS datasets would perhaps be better spent on increasing the surveillance population size and geographic spread of DSSs and analysing and disseminating research findings.

Place, publisher, year, edition, pages
BioMed Central, 2008. Vol. 8, Article nr 13- p.
Keyword [en]
Age Distribution, Analysis of Variance, Demography, Epidemiologic Methods, Ethiopia/epidemiology, Female, Humans, Male, Mortality, Poisson Distribution, Population Surveillance, Regression Analysis, Sex Distribution, Software
National Category
Public Health, Global Health, Social Medicine and Epidemiology
URN: urn:nbn:se:umu:diva-10416DOI: 10.1186/1471-2288-8-13ISI: 000254661800001PubMedID: 18366742OAI: diva2:150087
Available from: 2008-09-08 Created: 2008-09-08 Last updated: 2015-04-29Bibliographically approved
In thesis
1. Dying to count: mortality surveillance methods in resource-poor settings
Open this publication in new window or tab >>Dying to count: mortality surveillance methods in resource-poor settings
2008 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]


Mortality data are critical to understanding and monitoring changes in population health status over time. Nevertheless, the majority of people living in the world’s poorest countries, where the burden of disease is highest, remain outside any kind of systematic health surveillance. This lack of routine registration of vital events, such as births and deaths, constitutes a major and longstanding constraint on the understanding of patterns of health and disease and the effectiveness of interventions. Localised sentinel demographic and health surveillance strategies are a useful surrogate for more widespread surveillance in such settings, but rigorous, evidence-based methodologies for sample-based surveillance are weak and by no means standardised. This thesis aims to describe, evaluate and refine methodological approaches to mortality measurement in resource-poor settings.


Through close collaboration with existing community surveillance operations in a range of settings, this work uses existing data from demographic surveillance sites and community-based surveys using various innovative approaches in order to evaluate and refine methodological approaches to mortality measurement and cause-of-death determination. In doing so, this work explores the application of innovative techniques and procedures for mortality surveillance in relation to the differing needs of those who use mortality data, ranging from global health organisations to local health planners.


Empirical modelling of sampling procedures in community-based surveys in rural Africa and of random errors in longitudinal data collection sheds light on the effects of various data-capture and quality-control procedures and demonstrates the representativeness and robustness of population surveillance datasets. The development, application and refinement of a probabilistic approach to determining causes of death at the population level in developing countries has shown promise in overcoming the longstanding limitations and issues of standardisation of existing methods. Further adaptation and application of this approach to measure maternal deaths has also been successful. Application of international guidelines on humanitarian crisis detection to mortality surveillance in Ethiopia demonstrates that simple procedures can and, from an ethical perspective, should be applied to sentinel surveillance methods for the prospective detection of important mortality changes in vulnerable populations.


Mortality surveillance in sentinel surveillance systems in resource-poor settings is a valuable and worthwhile task. This work contributes to the understanding of the effects of different methods of surveillance and demonstrates that, ultimately, the choice of methods for collecting data, assuring data quality and determining causes of death depends on the specific needs and requirements of end users. Surveillance systems have the potential to contribute substantially to developing health care systems in resource-poor countries and should not only be considered as research-oriented enterprises.

Place, publisher, year, edition, pages
Umeå: Epidemiologi och folkhälsovetenskap, 2008. 66 p.
Umeå University medical dissertations, ISSN 0346-6612 ; 1152
mortality, surveillance, verbal autopsy, survey methods
National Category
Public Health, Global Health, Social Medicine and Epidemiology
urn:nbn:se:umu:diva-1544 (URN)978-91-7264-500-4 (ISBN)
Public defence
2008-02-29, sal 135, 9A, Norrlands Universitetssjukhus, Umeå, 13:00 (English)
Available from: 2008-02-14 Created: 2008-02-14 Last updated: 2010-01-11Bibliographically approved

Open Access in DiVA

Demonstrating the robustness of population surveillance data: implications of error rates on demographic and mortality estimates(359 kB)46 downloads
File information
File name FULLTEXT02.pdfFile size 359 kBChecksum SHA-512
Type fulltextMimetype application/pdf

Other links

Publisher's full textPubMed

Search in DiVA

By author/editor
Fottrell, EdwardByass, Peter
By organisation
Epidemiology and Global Health
In the same journal
BMC Medical Research Methodology
Public Health, Global Health, Social Medicine and Epidemiology

Search outside of DiVA

GoogleGoogle Scholar
Total: 46 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Altmetric score

Total: 65 hits
ReferencesLink to record
Permanent link

Direct link