Mortality measures are essential for assessing the overall health status of the populations. Measures such as age-specific death rates, life expectancy, leading causes of death, and cause-specific death rates provide baseline indicators from which national and sub-national health profiles can be constructed, to guide health policy formulation. In Oman, for the first two decades following the emergence of the nation state in 1970, population surveys served as the only data source for mortality measurement and estimation. The first population census, which also provided information on mortality, was conducted in 1993. It was only in 2004 that civil registration (including deaths) operated by the Omani Ministry of Health and the Directorate General of Civil Status at the Royal Oman Police was established. Until now, there has been no in-depth study to comprehensively assess the quality of registration data, either in terms of completeness of registration or the quality of registered causes of death. This research has been designed to conduct a detailed data quality assessment, as well as derive the first-ever local estimates of levels of mortality and causes of death in Oman. Aims and objectives: The primary aims of this study focus on the assessment of the completeness and quality of mortality data from the vital registration system, and the need to ascertain current levels and cause of death patterns, for health policy, evaluation and planning in Oman. These aims were achieved through five objectives: (i) to critically evaluate of available mortality data in order to assess biases in the data, (ii) to test and apply methods to correct the biases, (iii) to derive adjusted mortality rates by age, sex and cause, (iv) to assess mortality patterns in the light of demographic and epidemiological trends in Oman, and (v) to assess the implications of mortality indicators for national health policy and planning. Methods A number of methods have been applied to measure and achieve the study objectives. A systematic review of mortality data sources for Oman since 1970, to document the history or mortality measurement over the past four decades. Indirect and direct techniques were used to assess completeness of death registration. Two indirect methods were used: “Brass growth balance method” and “Preston and Coale method”. This research also conducted a comprehensive assessment of completeness using the direct method which involves record matching between two independent data sources that are each designed to record the same set of events. Matching identifies records captured by both sources, as well as those in one but not ii the other, and vice versa, and this information is used to estimate completeness of record capture in either source. In this research, the two data sources used were the national death registration database for 2010, and the Census 2010 mortality records for deaths reported to have occurred during the calendar year. (iii) The cause recorded in the registration database was validated against the reference cause derived from available medical records, retrieved from the health institutions where the event was registered in 2010. Matching of causes for each death was used to assess validity of registration records, and to adjust for identified biases to generate corrected cause-of-death profiles from the database. (iv) Findings from the completeness component were used to adjust age-sex specific death rates and to derive adjusted life tables, as well as summary mortality indicators. Corrected cause of death data were used to derive final estimates of leading causes of death, and age-standardized death rates and Years of Life Lost by cause. Results: The historical review identified a sharp decline in under-five mortality rates over the past four decades. The findings from the completeness assessment demonstrate that reporting of deaths in Oman from Birth and Death Notification System is near complete among entire population (about 90%), and this is a reassuring result in terms of the overall quality of death registration in the country. Completeness reporting rate in deaths below 15 years was somewhat lower, at about 70%. Adjusted life expectancy at birth was 73.7 years for males, and 80.0 for females. These estimates were higher than those reported from other international sources as derived from model life tables. The cause-of-death validation study resulted in considerable decline in proportion of deaths assigned ill-defined causes, from 35% to 10%, with re-classification to the revised leading causes of death including ischaemic heart disease, road traffic accidents, stroke, diabetes and cancers in both males and females. Congenital malformations and perinatal conditions were also among the leading causes of death in Oman. Analysis of YLLs identified accidental drowning, falls and breast cancer among the priority conditions causing premature mortality. Age-standardised death rates from ischaemic heart disease, road traffic accidents, diabetes, stroke and cancers estimated in this study were substantially different from modelbased estimates published by other international data sources. Conclusions This thesis documents the first empirical estimates on mortality by age, sex and cause for Oman. The findings from this study have implications for national health policy, the national health information system, and national and international epidemiology. The magnitude of mortality from road traffic accidents and cardiovascular disease indicate the need for additional research into the causes, management and potential for prevention of such conditions. The field activities conducted as part of this research has developed the methodology for strengthening data quality in the birth and death notification system, as well as a critical mass of human resources to implement interventions to achieve system improvement. Similar research to evaluate data quality and guide data utilisation should be conducted on a periodic basis in Oman, as well as in other Gulf Cooperation Council states which share similar health system characteristics.

Mortality measures are essential for assessing the overall health status of the populations. Measures such as age-specific death rates, life expectancy, leading causes of death, and cause-specific death rates provide baseline indicators from which national and sub-national health profiles can be ...