Data Entry: Please note that the research database will be replaced by UNIverse by the end of October 2023. Please enter your data into the system https://universe-intern.unibas.ch. Thanks
Improving quality of medical certification of causes of death in health facilities in Tanzania 2014-2019
Journal
BMC Health Serv Res
Volume
21
Number
Suppl 1
Pages / Article-Number
214
Keywords
Cause of Death; Certification; *Data Accuracy; *Health Facilities; Humans; Tanzania/epidemiology; Causes of death; Civil Registration and vital statistics; Continuing professional development; Dhis2; Data quality assessment; Icd-10; Medical certification; Mortality data; Start-up mortality list; Tanzania; eHealth
Mesh terms
Cause of Death; Certification; Data Accuracy; Health Facilities; Humans; Tanzania, epidemiology
Abstract
BACKGROUND: Monitoring medically certified causes of death is essential to shape national health policies, track progress to Sustainable Development Goals, and gauge responses to epidemic and pandemic disease. The combination of electronic health information systems with new methods for data quality monitoring can facilitate quality assessments and help target quality improvement. Since 2015, Tanzania has been upgrading its Civil Registration and Vital Statistics system including efforts to improve the availability and quality of mortality data. METHODS: We used a computer application (ANACONDA v4.01) to assess the quality of medical certification of cause of death (MCCD) and ICD-10 coding for the underlying cause of death for 155,461 deaths from health facilities from 2014 to 2018. From 2018 to 2019, we continued quality analysis for 2690 deaths in one large administrative region 9 months before, and 9 months following MCCD quality improvement interventions. Interventions addressed governance, training, process, and practice. We assessed changes in the levels, distributions, and nature of unusable and insufficiently specified codes, and how these influenced estimates of the leading causes of death. RESULTS: 9.7% of expected annual deaths in Tanzania obtained a medically certified cause of death. Of these, 52% of MCCD ICD-10 codes were usable for health policy and planning, with no significant improvement over 5 years. Of certified deaths, 25% had unusable codes, 17% had insufficiently specified codes, and 6% were undetermined causes. Comparing the before and after intervention periods in one Region, codes usable for public health policy purposes improved from 48 to 65% within 1 year and the resulting distortions in the top twenty cause-specific mortality fractions due to unusable causes reduced from 27.4 to 13.5%. CONCLUSION: Data from less than 5% of annual deaths in Tanzania are usable for informing policy. For deaths with medical certification, errors were prevalent in almost half. This constrains capacity to monitor the 15 SDG indicators that require cause-specific mortality. Sustainable quality assurance mechanisms and interventions can result in rapid improvements in the quality of medically certified causes of death. ANACONDA provides an effective means for evaluation of such changes and helps target interventions to remaining weaknesses.
