Sources of Data, Population at Risk, Standardization

Sources of Data

The principle sources of epidemiologic data are:

1. Census

Census includes a complex set of information concerning household and details on individual in each household. The information is compiled into national database. Census is under the responsibility of National Statistical Office (NSO). In Nepal, Central Bureau of Statistics conducts census in every ten years.

2.Registration for vital events

The vital statistics are recorded and processed by Ministry of Development and Ministry of Health and Population. Both two agencies get the information from death certificates, which are filled up by health personnel (if the death occurs in hospitals or health care facilities), and by health workers (if the death occurs in village). Death certificates comprise three identical parts.the first is the given information; the second part is sent to local district registrar office and then forwarded to Central Registrar Division, Ministry of Local Development. The third part is sent to local district health office and then forwarded to Health Statistics Division, Ministry of Health and Population. Central Bureau of Statistics produces annual reports concerning number of births, deaths, migrations and nuptial statistics annually.

Ministry of Health produces annual reports concerning births and deaths classified by various characteristics such as age of mothers, infant deaths, cause of death, age o death, and so on; by region and by districts.

3. Notifiable disease statistics

List of the disease under surveillance differs from place to place. WHO had set up a list of disease required to be reported by every country as:

• Cholera
• Plague
• Yellow fever
• AIDS

4. Morbidity and mortality statistics

The information on morbidity and mortality in Nepalese population; by cause, age, sex and district is provided by Management Division, Health Management Information System, Epidemiology and Disease Control Division, Department of Health Services, Ministry of Health and Population. The causes of deaths are grouped according to International Classification of Diseases (ICD) 10^th revision, using basic tabulation list which covers 56 groups of diseases. The morbidity information is obtained from out patient and in patient departments of all hospitals over the country. The data from out patient department is classified into 17 groups and the data from in patient department is classified into 65 groups of illness. The data is sent from all hospitals and health facilities of all over the country. For calculation of morbidity rates, the population enumeration is based on those collected by Central Bureau of Statistics. Health Management Information System (HMIS) and early Warning Reporting and Response System (EWARRS) have been developed under the Department of Health Services. Both of these systems are functioning to get the regular institutional based health information. Annual morbidity and mortality statistics are published.

5. Surveys

In Nepal, most of the nationwide surveys are done by the Central Bureau of Statistics, such as survey of population change, which is conducted in every 10 years, national health surveys and economic surveys, demographic health survey.

Population at Risk

The population at risk is a group of persons who is exposed to the risk and is capable of contracting the disease. The population at risk (denominator) or a rate measuring a disease occurring during a given time interval is the average number of the population at risk. For an annual rate, the midyear population is used as risk population. In the case of dynamic population, where it cannot be assumed nearly every person is present and at a risk for the whole year, the exposure may be obtained by summing up, for members. This can be very tedious, particularly in a large population, but it gives a good approximation of incidence density.

Standardization

To compare the incidence, prevalence or mortality of some conditions in two or more populations, it is important to ensure that rates are correctly calculated and calculated from data, which are complete and accurate. To compare rates and to get reliable results between different population or geography, other factors such as age, sex, and other variables should be considered. If morbidity and mortality statistics are affected by such characteristics, the calculated crude morbidity or mortality will give wrong implications. If we compare age or sex or other characteristic specific rates of such population, it will inform better about what we calculated. Single figure per population is important, and for this, we should standardize for the characteristic responsible for the difference. This is called adjusted rate. The choice of the characteristic to be standardized for depends on the research question being answered. The characteristic to be standardized for is known to differ in distribution within the population being compared.

The morbidity of HIV/AIDS in males among total cases is 73.3 percent which is almost three times as compared to case number of the females. It is also noted that morbidity is related with age. The cases distribution according to age structure of the two populations indicates that the numbers of cases are high in 15 to 49 years age in both the groups. However, the proportion of population over 50 years of age is higher in male as compared to female. It is opposite in age group below 15 years of age. These data give different weight to the crude morbidity rates. At this stage, we might consider the use of more reliable single figure as an index for morbidity level, since it comes out as if we should not compare mortality rates among a single figure index of morbidity, we must, therefore take account for the different population structures.

Sex and age specific morbidity of HIV/AIDS in Nepal in 2004

Percent distribution of cases by age

There are two common methods o deriving standardized rates, known as direct standardization and indirect standardization.

1. Direct standardization

To compare the population of different structures, a standard population which may or may not be one of the populations being compared is used, and specific rates of the study populations are applied or mortality rate assuming that the number of person in each age category is similar in the two populations. These standardized expected number of deaths are summed up and divided by total expected cases or population to get overall standardized mortality or morbidity.

Although standardized rates are commonly calculated and can be effectively compared among different population, their actual values have no meaning since it depends on choice of standard population being used.

2.Indirect standardization

In some cases, an indirect standardization based on the standard morbidity or mortality rates to the actual population may be preferable than direct method. The age-specific rates in study population may be not available (however, total number of deaths or cases must be available). Because of the small number of death or case at each age group, age-specific mortality or morbidity rates of the study population used in indirect method may be subjected to random error, and may result different adjusted morbidity or mortality rate. If the study population is a subset of standard population, indirect standardization is suitable to use. For example, if we categorize population as government employees, farmers, laborers and use male population, indirect standardization is suitable.

References

Atlas, RM and R Bartha. Microbial Ecology:Fundamentals and Applications. The Benjamin Cummins Publication co. Inc., 1998.

Gordis, L. Epidemiology. third edition. 2004.

Maier, RM, IL Pepper and CP Gerba. Environmental Microbiology. Academic press Elsevier Publication, 2006.

park, K. Park's Text Book of social and prevention Medicine. 18th edition. 2008.