How to use health statistical data

Statistical data such as demographics are useful for understanding local characteristics and problems.
How should I use and evaluate the figures in statistical data? Here, I would like to mention the points of interest and precautions when using statistical data.

If the purpose is clear, you can set a policy on what statistical data should be collected and how to evaluate it.

Do you look at the city unit, ward unit, or do you look more closely at regional units? Please determine the evaluation unit for the purpose of using statistical data.

The average life expectancy is longer for women than men, and the gross mortality rate is higher for older people. Thus, statistical figures tend to differ depending on gender and age.
Frequently used death and births can be obtained from demographic statistics.

The larger the population of the region, the more people are likely to get sick. Therefore, it is necessary to consider the impact of statistical data due to differences in population.
Regarding the population, the census population and the estimated population (the census population minus birth, death, transference, etc.) are often used.

Government statistics such as demographic statistics can also be downloaded from the WEB page. (Links to obtain statistical data)

（１） Use real numbers

Is it best to use health indicators such as "rough mortality rate" and "age-adjusted mortality rate" to evaluate statistical data?
Health indicators look good, and indicators that take into account the effects of population, age, etc. have the advantage of being easy to compare between regions and annually.

But can we say that there is no problem in the area just because the health index is low?

For example, in City A with a population of 1 million and City B with 100,000, the number of suicides was 200 in A Ward and 50 in B. In this case, the "rough mortality rate (compared to 100,000 population)" of suicide is 20.0 in A ward and 50.0 in B. Although the frequency of suicide is lower than Ward B, the phenomenon is that the number of suicides is higher. Isn't it a problem that the number of suicides itself is large in A ward?

It depends on the way of thinking of the people who use statistics, "how much is the problem", but it is also important to take a closer look at the figures of the statistical data themselves.

Real numbers can be evaluated by gender and age group, or by looking at secular changes. It's also a good idea to make a graph to make it visually understandable.

（２） Use health indicators

There are a lot of health indicators. Rather than looking for a health index as much as you can imagine, extracting a few indicators according to the purpose, it will save less labor and easier to evaluate.

In order to select the best health indicators, it is important to understand its characteristics (pros and disadvantages).

Some health statistical data have been published in advance for health indicators, such as special reports on demographic statistics and life tables by municipalities. There are some difficulties that it takes time before it is released, but it is also a good idea to use these materials to evaluate it.
（Obtain existing health indicators）

When using existing health indicators, it is necessary to check which statistical data numbers were used and what formulas were used. This is because when looking at aging or comparing between regions, it is not possible to compare honestly unless the health index is used in the same way.

When seeking health indicators for low-frequent events or when using statistical data in regions with a small population, it is necessary to consider accidental errors in numbers.

Statistical methods can be used to reduce accidental errors, but in general, multi-year figures are generally used.

When using statistical data created based on questionnaire surveys, such as the Basic Survey on National Life and the Census, it is necessary to consider grid errors.

This is because the answer may be distorted depending on the nature of the question, such as marriage status and occupation. In addition, it is possible to think that there is almost no systematic error in the number of deaths, births, and census population in the frequently used demographic statistics.

The obtained health indicators can be used by gender and age group, seeing age changes, and compared with other regions. It is also a good idea to make it visually understandable by graphing or map color coding.

Consider a method of evaluation according to the characteristics of the index.