before his death, despite the fact that he had been constantly visiting the doctor
during the last six years of his life, to deal with high blood pressure and heart
lThis may suggest that 'morbidity' exists only if the patient is aware of it and regards the condition as an illness.
deficiency. Not only was the patient unaware of the development of the disease which was found to have begun long before his death, but he was seeking treatment for only some symptoms which eventually led to the immediate cause of death (Verbrugge, 1983:225). The verbatim report is as follows:
My husband died 40 days ago from renal failure. He was 49 years old. His illness began with a high blood pressure, but this last disease (renal failure) was diagnosed only eight days before he died. When he first visited a doctor, six years ago, he was told he had very high blood pressure and that his heart was not functioning well either. He started visiting a doctor each one or two months, but in the last four or six months his situation worsened. The doctor then prescribed aldomet, which is a medicine to deal with heart problems. It did not help. He was not feeling well, he was constantly nervous and we decided to visit another doctor. This other doctor confirmed that his blood pressure was indeed very high but the real problem was other. The doctor told me 'Dona Efigenia, there is nothing wrong with the functioning o f your husband's heart. But he has a serious renal failure which seems to have been going on fo r a long time. He has to stay in hospital and go through an haemodialysis at least three times a week. But I warn you that he will not live long'. In fact, after eight days he died (Efigenia, 44 years).
The focus on the relationship between morbidity and mortality in this chapter is less an attempt directly to compare levels o f morbidity and mortality, than to try to understand whether the patterns o f age-specific rates and cause-age- specific rates o f mortality are equally reflected on the morbidity side. To envisage that, an overview o f the comparison between the age-specific morbidity rates and age-specific mortality rates is set out.
6.2 Data: Sources and Limitations
M ortality rates for both regions in 1981 are derived from the vital
registration statistics, while morbidity rates are estimated from information from
PNAD-81. The next analyses focus on the morbidity and mortality differentials by
income and education: the mortality estimates are based on PNAD-84, as shown
in Chapter 3, and the morbidity rates are derived from PNAD-81 (see Chapter 4).
For the analysis involving causes of death and illness, mortality rates for Belo
H orizonte in 1979 are estim ated using vital registration statistics, whereas
morbidity statistics are drawn from information collected during the in-depth
interviews also in 1979. Additionally, data available on multiple cause o f death for
Säo Paulo in 1985, derived from the vital registration system, are also considered.
Concerning the information on morbidity, it is important to observe that
the 'ideal' data set, when all morbidity conditions during the time reference period
are reported is, in practice, impossible to obtain, partly because o f their own
definition o f illness and morbidity conditions, either established by a particular
survey or im plicit in people's own perception and definition, as discussed in
section 4.4 o f Chapter 4. However, even such 'ideal' data are not directly
com parable to m ortality rates and some o f the problem s related to this are
mentioned in the preceding section.
The way in which the data provided by PNAD-81 differs from the 'ideal'
m orbidity data, and the way in which the PNAD-81 data differ from the
information collected during the in-depth interviews, with respect to definition and
coverage, are also discussed in section 4.4 o f Chapter 4. For example, it was
mentioned that the PNAD-81 data exclude persons who already had a permanent
limitation, unless they had an acute episode or other disease in the time reference
the in-depth interviews do include chronic illnesses. Both, however, do not
include particular illnesses considered 'normal' according to people's perception,
although in the in-depth interviews some of such conditions were clearly
identified. Ignorance o f their conditions, omitting to state conditions which they
considered unimportant, or of which they were unaware are other limitations of
the data. Some of the statements related to these issues and collected during the in-
depth interviews are transcribed in section 4.4 of Chapter 4.
A nother lim itation also pointed out elsew here is that the in-depth
interviews refer to a sample of 613 persons, and from the poorer parts of Belo
Horizonte, and therefore would not expected to be totally com parable with
mortality rates for all Belo Horizonte. Finally, PNAD-81 data separated morbidity
conditions into those associated with illness, teeth and an accident or injury. As
m entioned in subsection 4.4.1 o f Chapter 4, interviewers were instructed to
record more than one type, according to the answers given by the respondents.
That is, all questions in the morbidity block o f PN A D -81 were asked for each
type o f health problem. To avoid duplication or overestimation, only the subgroup
o f people who reported having case of illness is referred to in this chapter, the
same procedure as adopted in Chapter 4 (see Figure 4.1 in subsection 4.4.1 of
Chapter 4).
6.3 Multiple Causes of Death
The information on causes o f death provided by the vital statistics system
is based on a single cause of death, known as basic or underlying cause.
Nevertheless, in the international form of medical certificate of cause o f death, this
is not the only illness or morbidity condition recorded. This form, filled in by
accounts for the chain of events leading to death. It is divided in two parts: in the
first one, the underlying cause (item c) and its chain of events - terminal cause
(item a) and intermediate cause (item b) - are stated. Items a and b of part I
constitute the so-called consequential causes. In the second part, two contributory
causes are stated.
As the possibility exists of misreporting the causes of death, especially the
misplacement of causes in the respective items, a selection o f the underlying cause
o f death is made in accordance with the definition and rules set forth in the
International Classification o f Diseases. Underlying cause is defined as 'the
disease or injury which initiated the train of morbid events leading directly to
death'; while contributory cause is 'any other significant condition which
unfavorably influenced the course of the morbid process and thus contributed to
the fatal outcome, but which was not related to the disease or condition directly
causing death’ (Puffer and Serrano, 1973:74).
Because only the underlying cause of death has usually been made
available through the classification provided by the vital registration system,
mortality statistics do not use all the diagnostic information stated in the death
certificate. Nevertheless, the importance of considering multiple causes of death
has been well established by the study of interrelations of causes carried out in the
Inter-American Investigation of Mortality, which concluded that:
the implications o f this concept are exceedingly important from the preventive viewpoint because measures aimed at preventing the underlying cause are not sufficient if the contributory conditions remain. One example is the common association o f nutritional deficiency and infection, which necessitates prevention o f both com ponents o f the com plex, w ithout which the phenomenon o f 'substitution' may come into play - that is, a death prevented by suppressing one underlying cause (usually an infection) may later occur as the result o f another underlying cause (another infection) if the contributory condition (nutritional deficiency) remains (Puffer and Serrano, 1973:74).