[MUSIC] Welcome to Routine and Public Health Information Systems Data for Public Health Practice. I'm Keri Althoff from the John's Hopkins Bloomberg school of public health. The objectives are to introduce the principles and technical content of routine health information systems or RHIS and Public Health Information Systems or PHIS. We'll describe these systems and give a few examples. We will also have a brief exercise to evaluate the vital registration systems in countries in North, Central, and South America. The supporting reading for today is by Mahapatra and colleagues, published in the Lancet in 2007. Although this article is a bit dated, it provides a thoughtful evaluation of probably the most important routine health information system, civil registrations systems. So, let's begin with Routine Health Information Systems. David Ogilvy was an English man who in 1935 wrote what Fortune magazine called the best sales manual ever written. As his career continued, he had great influence in the field of advertising. I like this quote, because routine health information system are not remarkable, they're pretty ordinary, but the data collected by these systems informs mortality in a population. However, routine health information systems are lacking when it comes to providing data for other measurements of population health. That's where Public Health Information Systems and other specifically designed studies provide better benefit when solving public health problems. Routine health information systems or RHIS can produce timely information that can be used in decision-making. For example, data from routine health information systems are often used to monitor progress towards important public health goals. Such goals are things like the Millennium Development Goals, that have now extended past 2015. And the healthy people goals, which are US-based. Routine health information systems are often of three types. Civil or vital registration systems, medical record systems, and administrative record systems like the Census or Medicare. We'll start by describing the most commonly used routine public health information system. Civil or vital registration systems. Civil registration systems are systems that record births, marriages, divorces, and deaths. Civil registration systems date back thousands of years. The Zhou dynasty had a household registration as early as 1100 BC. France's system was originally rooted in the church but became secular in 1792. England and Wales had the Bills of mortality, which contained weekly mortality statistics to monitor burials. From this system, a full civil registration system was developed. And finally, vital statistics became available from the data collected. The US established a civil registration system in 1641, but it took almost 300 years to ensure the system was complete in all states. The US vital statistics system was created in 1950 and is still operating. You will hear civil and vital used interchangeably to describe these systems. This is because civil and vital registration systems collect common data elements for health purposes namely births and deaths. Vital registration systems are systems developed specifically to capture data in which vital statistics are estimated. This occurs most frequently in the absence of the civil registration system. So really the difference in Civil and Vital registration systems is rooted in the original purpose of the system. Additionally, the common data manifested from these systems, namely births and deaths, may have differences in the level of health-related details. The more common terminology now used is civil registration system and vital statistics because the system is the mechanism to collect the data from which vital statistics can be estimated. A more common terminology now used is civil registration system and vital statistics because the system is the mechanism to collect the data from which vital statistics can be estimated. When there is no functioning civil registration system, a vital registration system is established to collect important birth and death information. But there is a spectrum of systems that are considered vital registration systems. In some parts of the world, there's no system available. Then there are places where verbal autopsy is the mechanism through which death information is collected. Verbal autopsies are standardized surveys that are given to caregivers, family, or friends of the deceased. This survey questions ask about symptoms in the disease prior to death. This is a crude way of gathering information about the death. Sentinel registration system also exist whereas a single or group of diseases or causes of death are monitored. For example, some sentinel registration systems only collect information pertaining to maternal mortality. Next are demographic surveillance systems. These systems exist within a subpopulation in which they develop test and validate the vital registration system. Sample registration systems are also limited in their capture. A nationally representative sample of the population is monitored by the vital registration system. This is usually an intermediary step when building a full registration system, because the sample registration system is a platform to establish the protocols that work most efficiently. Partial vital registration systems are systems that exist, but only where the necessary infrastructure exists. Often times rural areas are not included in partial vital registration systems. Finally all vital registrations systems are shown to capture at least 90% of the country's deaths and births. The purpose of all these various forms of vital registration systems is to collect and maintain vital records. Vital record are collected through the civil and vital registration systems. Porter describes the vital records are records to do with the living including certificates of birth, death, marriage and divorce. The US has a full vital registration system so at least 90% of births and deaths are believed to be captured in the US Vital Statistics system. The legal responsibility for keeping these records is at the state level for the 50 states. Washington DC and New York City have separate legal responsibility to keep vital records. Five US territories also keep vital records. Additionally, other important information, like race, sex, and smoking status can be collected through these vital records. These states, cities, and territories, then submit some of the information collected to the federal government. The CDC's National Center for Health Statistics then analyze this data and create reports. One of the most important reports is of the mortality rates and the top causes of death in the population. Worldwide vital records can be kept at the local state or federal level. Each country has their own rules and protocols for handling vital records. So let's dive into the data that are captured in vital records. In order for a birth certificate, data are collected via standardized form. Basic data that are collected include the date and place of birth, as well as the sex of the baby. Information about the baby's mother and father are also collected including their data of birth, race ethnicity, and residence. Some countries collect much more information before issuing a birth certificate such as the baby's birth rate and gestational age at delivery or whether the mother smoked or had prenatal care. The fundamental data collected for issuing a death certificate include the immediate cause of death. Defined as the disease or condition directly leading to death as well as three anteceding causes of death which are typically referred to as underlying causes of death. The WHO defines underlying cause of death as the disease or injury that initiated the train of morbid events leading directly to death. Or the circumstances of accident or violence that produce the fatal injury. There is an algorithm that is then used to select one underlying cause of death for each recorded death. It is that underlying cause of death that is used when producing mortality statistics from these vital records data. Finally, other significant conditions contributing to death are collected. These causes of death are coded using the international classification of disease version 10 or ICD-10. This coding system allows for standardization in the collection of the cause of death information. So that the mortality and morbidity statistics calculated from these data are comparable over time, and between regions of a single country, as well as across countries. This is fundamental data collected. Some countries also collect whether the deceased used tobacco. Every good epidemiologist knows the limitations of their data and we also live by the golden rule of garbage in, garbage out. That is why it is essential to evaluate the strengths and weaknesses of civil and vital registration systems, so that we know how the weaknesses influence our mortality and morbidity statistics. The article written by Mahapatra clearly defines common strengths and weaknesses of civil and vital registration systems. The article describes key areas that are needed to evaluate the quality of the vital records data from civil and vita registration systems. Coverage is the understanding of the proportion of the population that had medical certification of the cause of death. Completeness refers to the proportion of deaths that have a reported medically certified cause of death. Missing data refers to the additional standard data captured on the application for the death certificate. Estimating the proportion of those with submitted cause of death information who have missing age, sex, or race data is one way to identify the extent of missing data. Next, the quality of the cause of death data is evaluated, by looking at the proportion of deaths classified under various miscellaneous and ill-defined categories. The more specific the cause of death is, the better. Improbable classifications of death is another evaluation of the quality of the cause of death data from civil or vital registration systems. The number of deaths assigned to an improbable age or sex category per 100,000 coded deaths is the metric for understanding the extent to which improbable classifications of death are noted. And finally, consistency between the cause of death and predictions of general mortality can be made to evaluate the accuracy of the data from the civil registration system. Let's take a break before moving onto our next example of routine health information systems, medical records.