1957 - United Nations Statistics Division

sion of the age attribute shown in Tables 10, II, 12,
and 13.
Period of gestation or gestational age is defined III
accordance with international recommendations 50 as
follows:
Period of gestation is the number of completed
weeks which have elapsed between the first day of the
last menstrual period and the date of delivery, irrespective
of whether the product of conception was
live-born or born without evidence of life.
Gestational age is classified in 8 categories, namely,
under 17 weeks, 17-19 weeks, 20-27 weeks, 28-31 weeks,
32-35 weeks, 36 weeks, 37-39 weeks, and 40 weeks and
over. This classification, which provides no "unknown
age", is in accord with, but more detailed than, that
recommended in the Principles for a Vital Statistics
System 51 and by the World Health Organization. 52 Foetal
deaths of unknown gestational age are included in total
foetal deaths.
Where "age" has been reported in months or days
rather than completed weeks, conversion has been made
according to the following scale:
I - 3 months under 17 weeks
4 months 17 - 19 weeks
5 months 20 - 23 weeks
6 months 24 - 27 weeks
7 months. . .28 - 31 weeks
8 months. . 32 - 35 weeks
9 months 36 - 39 weeks
10 months 40 weeks and over
Coverage: Only 28 geographic areas were able to provide
foetal deaths classified as required and, for some of these,
there are minor variations as shown in revised run-in
headings or footnotes.
Ratio computation: Ratios are the number of foetal deaths
which occurred at a gestational age during the year per
1,000 live births which occurred during the same year.
The "live-birth base" was adopted rather than the combined
live births and foetal deaths because it is assumed
to be more stable in time and more comparable from
country to country.
It should be noted that, although the use of total live
births as a ratio base at each gestational age should provide
a set of ratios which, summed, produce the total
foetal-death ratio, this is not always the case because of
the relatively large number of foetal deaths with gestational
age unstated. Although methods of allocating
unstated age are available, none has been used in Table
21.
Limitations: Foetal deaths are probably the least accurate
of all vital statistics. In addition, they are subject to all
other limitations of vital statistics as described on p. 23.
Principal of their defects is lack of comparability as
a result of variation in definition in terms of gestational
age. There are only 17 countries which provide "foetal
deaths" in the complete sense, and even these admit the
possibility of a high degree of underregistration, espe-
50 Principles for a Vital Statistics System, op. cit. Principle 309.
510p. cit.
52 World Health Organization. Technical Report Series, No. 25, p.
13, section (d).
38
cially at the younger "ages". Therefore, it may be said
categorically that no basis exists for comparability between
statistics of foetal death for various countries. The
only use of the statistics at this stage of development is to
show the relative distributions by gestational age.
Table 22
4. Marriage statistics
Statistics of marriages are given in Table 22 only. For
this reason, the explanatory notes for marriage statistics
will be given in connexion with this table, rather than as
a separate section.
Table 22, corresponding in format to the other trend
tables, presents the number of marriages performed in
1948-1956, together with the crude marriage rates for the
same period.
In accordance with the recommendation made in the
Principles for a Vital Statistics System,53 the statistics
refer, unless otherwise noted, to formal marriages, that is,
to the legal union of persons of opposite sex. The legality
of the union may be established by civil, religious, or
other means as recognized by the laws of each country.
'Wherever possible, the data presented in the table relate
to both civil and religious marriages (if both types of
marriage are recognized) and are the number of marriages
performed during the specified year, rather than
the number registered.
Data relate to marriages performed within the present
boundaries of the geographic areas indicated and, unless
otherwise noted, within the present-in-area population
of those areas. However, by virtue of the legal concept to
which the data are bound, marriages among tribal or
indigenous groups isolated from the dominant culture
are almost invariably excluded from the statistics. These
cases are generally footnoted, nevertheless, as well as all
other deviations from the de facto concept.
Coverage: Statistics are presented for 151 individual geographic
areas. As in similar tables on live births and
deaths, some geographic areas are omitted in the absence
of recent data. It should also be noted that the geographic
coverage, extensive as it appears, is still deficient for
much of Africa where data relate to small segments of
the total population only, but these serve to indicate the
availability of marriage statistics.
Rate computation: The rates in Table 22, like corresponding
crude rates in other tables, are the number of marriages
reported as having occurred within the boundaries
of the geographic area per 1,000 total persons present in
the same area at the midpoint of the year in question.
In only a few cases is a census figure or an estimate for some
other date within the year used in place of the midyear
population.
The populations, midyear or other, used in the computation
of the rates are, in general, the official figures
presented in Table 3. Therefore, in the absence of an
official population for any year, a rate will not be computed,
despite the fact that the frequencies are available.
In general, reference should be made to Table 3 for the
type and coverage of the populations used in computing
the rates.
53 op. cit. Principle 202.