MORTALITY STATISTICS Death records are likewise secured by registration. On the other hand, it is fairly easy to secure complete returns if a death certificate is required before a burial permit is issued. The death certificate (Fig. 128), a portion of which dealing with the cause of the death is filled out by the attending physician, is filed with the proper authorities by the undertaker.
Errors are not uncommon on these certificates and the authorities should require their correction before certificates are accepted. The original data may be incorrect, due to mis statements by members of the decedent's family. The state ment of the cause of death given by the physician may be inaccurate due to an erroneous diagnosis or a desire to shield the feelings of the decedent's family.
One of the most important aspects of the physician's obliga tions in connection with mortality reporting is the statement concerning the cause or causes of death. Two grave errors are possible which seriously interfere with the accuracy of mortality returns. The first of these possible errors relates to mistakes in the diagnosis of the condition actually responsible for death. Any physician who has followed a considerable series of cases to postmortem is familiar with the fact that post mortem examinations not infrequently fail to corroborate the antemortem diagnosis, even when the patient had been under the care of extremely competent physicians. Where the post mortem examination serves as a check these errors are usually revealed, but where postmortem examinations are not made, and relatively few bodies are examined postmortem, these errors remain concealed. Another source of error is the em ployment by physicians of vague or inaccurate terms in stating the cause of death, so that it is difficult to associate the stated cause with any definite pathological condition. The first of these possible errors can only be reduced by greater refinements in the art of diagnosis and of greater efficiency on the part of the medical profession in their application. The second can be reduced by care. The Census Bureau is endeavoring to secure an improvement in the last direction by familiarizing physicians with the standard International Lists of Causes of Death, and also the indefinite, objectionable terms whose use should be avoided. The Bureau publishes a very convenient
Docket list of the proper terms.
Death records serve the following purposes; (a) they aid in the detection of crime, making the concealment of murders very difficult; (b) assist in the transfer of inherited property; (c) facilitate the settlement of life insurance, and lastly (d) in dicate the extent and changes in population produced by death. From a public health standpoint they have some value in the Deaths among a given population are expressed in rates fundamentally similar to those employed in expressing births. The most common rate employed is the general or crude death rate. This may be defined as the ratio of the number of deaths within a given time to the number of people alive at the middle of the period, referred to some round number (roo°, io,000 roo, 000) as a basis for comparison. The period commonly chosen is one year. In estimating populations for this purpose it is necessary to estimate the mid-year population, making suitable corrections with the census figures for the dates to which they have been referred. Their precision depends upon the accuracy of the population and death figures.
They are computed as follows: A population of 6o,000 has in one year goo deaths. What is the crude death rate per l000? Method A.
goo X I000 = 15 6o,000 Method B.
6o,000 goo : = 15 I000 Their most common use is in a comparison of the mortality in the same community in different years, or between two corn munities having similar populations. If the crude rate is em ployed in making comparisons between two communities having similar populations, falacious conclusions will almost certainly be drawn. For this purpose special rates should be employed of which we will shortly speak (Fig. 13o.) It is sometimes desirable to compute the death rates for short periods, such as weeks, months or quarters. When so com puted these are expressed in terms of annual rates, i.e. as what the annual rate would be providing the deaths occurred through out the year with the same frequency as during the week or month under consideration.