Abstract

ORIGINAL ARTICLE

Mortality caused by idiopathic pulmonary fibrosis in the State of Rio Grande do Sul (Brazil)^* * Work carried out in Pereira Filho Pavillion – ISCMPA, Fundação Faculdade Federal de Ciências Mëdicas de Porto Alegre (FFFCMPA).

Fabrício Piccoli Fortuna (te sbpt)^I; Christiano Perin^II; Leticia Cunha^II; José da Silva Moreira^III; Adalberto Sperb Rubin^IV

^IPneumologist of the General Hospital of Caxias do Sul – Fundação Universitária of Caxias do Sul (FUCS). Specialist degree from SBPT.

^IIMedical undergraduate student (FFFCMPA).

^IIIAssociate Professor of the Department of Internal Medicine of UFRGS and FFFCMPA.

^IVDoctor in Pneumology from UFRGS. Pneumologist (ISCMPA). Professor of FEEVALE (RS).

^{Correspondence} Correspondence to Adalberto Sperb Rubin Rua Almirante Abreu, 246/402 90420-010 – Porto Alegre, RS E-mail: arubin@terra.com.br

ABSTRACT

Epidemiologic data on idiopathic pulmonary fibrosis are relatively scarce, and its real incidence and prevalence are unknown. Recent studies suggest that mortality due to idiopathic pulmonary fibrosis is rising in developed countries.

OBJECTIVE: To describe mortality caused by idiopathic pulmonary fibrosis in the State of Rio Grande do Sul (RS), Brazil, from 1970 to 2000, analyzing its trend and comparing it with that from other countries.

METHOD: Prevalence study, using data from the Brazilian Institute of Geography and Statistics (IBGE), analyzing death certificates in which idiopathic pulmonary fibrosis was stated as the ultimate cause of death.

RESULTS: Annual mortality rate adjusted to the population was 0.22/100,000 people in the 1970’s, 0.3/100,000 people in the 1980’s, and 0.48/100,000 people in the 1990’s. Total mortality raised 36% from 1970 to 1980, and 73% from 1980 to 1990. The mortality rate adjusted to the population raised 36% and 60% during the same periods. The rise in both total and adjusted mortality from IPF was statistically significant (p < 0.05). The mean mortality rate per 100,000 inhabitants between 1996 and 1998, however, was 0.683, corresponding to a 70% increase when compared to the previous three-year period, which was 0.4 (p = 0.0002), probably reflecting coding practices.

CONCLUSION: There was a significant increase in IPF mortality in RS from 1970 to 2000, partly due to changes in coding practices. This increase is in conformity with observations in other countries, although mortality rates in RS are considerably lower.

Key words: Pulmonary fibrosis. Mortality.

Abbreviations used in this paper

ICD – International Classification of Diseases

IPF – Idiopathic Pulmonary Fibrosis

BIGS – Brazilian Institute of Geography and Statistics

Introduction

Idiopathic pulmonary fibrosis (IPF) is one of the idiopathic interstitial pneumonias, being histologically characterized by the usual pattern of interstitial impairment. Its natural history shows a progressive evolution of the fibrotic process with structural disruption and haphazard therapeutic response ^(1-3). As a group, pulmonary interstitial diseases are not rare, but there are few epidemiological data about IPF occurrence and prevalence in the general population ^(2,4,5).

Recent studies in Europe ^(6,7) and in the United States ⁽⁸⁾ detected increased mortality due to IPF since 1979. There are no available data on mortality from IPF in the Brazilian population. The objective of the present study is to describe the mortality from IPF in Rio Grande do Sul (RS) from 1970 to 2000, assessing the trends of mortality in this population, and comparing them to similar studies on other populations.

Methods

Data on mortality from IPF were obtained between 1970 and 2000 from the Annual Statistics on Mortality published by the Regional Center of Health Information of Rio Grande do Sul. Only the cases in which pulmonary fibrosis was mentioned as the main cause of death were considered. The term pulmonary fibrosis was researched (code S17), referring to the International Classification of Disease (ICD) 8, from 1970 to 1978. From 1979 to 1995, the researched the terms cryptogenic fibrosing alveolites (code 516.3/2) and post-inflammatory fibrosis (code 515.9/4), referring to ICD-9. From 1996 to 2000, the terms researched referring to ICD-10 were fibrosing alveolites and pulmonary fibrosis (J.84.1),

Populational data were obtained by reports from the census of the Brazilian Institute of Geography and Statistics (BIGS), using linear interpolation to estimate the population in the years between the censuses. Comparisons between the interval means were made by the Student’s-t-test.

Results

Throughout the 30-year period, there were 777 deaths whose main cause was IPF. In 1970, the State population was 6,664,891 inhabitants, and in 1998 it was 9,810,471, representing a net increase of 3,145,580 inhabitants, corresponding to 47.1%. Figure 1 shows the evolution in absolute numbers of deaths from IPF, whereas Figure 2 shows the evolution of mortality rate by 100,000 inhabitants.

In the 70’s, 145 deaths from IPF were recorded. In the 80’s and 90’s, 220 and 412 deaths were respectively recorded, representing a 51% and 184% increase in total mortality from IPF in these periods. Mean annual mortality adjusted to the population was 0.22/100,000 in the 70’s, 0.3/100,000 in the 80’s and 0.48/100,000 in the 90’s. Total mortality increased 36.3% from 1970 to 1980, and 73% from 1980 to 1990. The rate adjusted to the population increased 36.3% from the 70’s to the 80’s and 60%, from the 80’s to the 90’s. The increase of both total mortality from IPF and mortality adjusted to the population were statistically significant (p < 0.05). Figure 3 compares the mortality rates in the State of RS (Brazil) with that of seven other countries.

In order to investigate a possible effect of coding change on mortality rate from IPF, the mean rates in the 3-year period before or after each change in the ICD were compared. The mean mortality rate per 100,000 inhabitants in the period between 1976 and 1978 was 0.247, identical to that of the period from 1979 to 1981. The mean mortality rate per 100,000 inhabitants between 1996 and 1998, however, was 0.683, corresponding to an increase of 70% in comparison to the three previous years, which was 0.4 (p = 0.0002), suggesting that this increase was influenced by the change of ICD, which took place in 1996. This fact also occurred when comparing total mortality rates in the same period: 37.6 in 1993-1995, and 66.3 in 1996-1998, an increase of 76% (p = 0.004).

Discussion

Our data show an increase in mortality due to IPF from 1979 to 2000, both total mortality and by 100,000 inhabitants. This significant increase is partly due to the coding pattern alteration that occurred in 1996.

It is difficult to obtain accurate epidemiological data regarding the actual incidence and prevalence of IPF ⁽⁴⁾ for several reasons: alterations of diagnostic criteria throughout the years, from their classification by Liebow ⁽⁹⁾ until a recent modification proposed by Katzenstein and Myers ⁽¹⁰⁾, added to the uneven application of such criteria to clinical practice, and dependence on the available technology, generate doubts concerning the accuracy of the cases coded as such. Moreover, deaths from IPF may be attributed to other frequently co-existing causes, such as pneumonia or heart failure, thus reducing the real estimate. Finally, the same histological pattern of IPF can be found in a variety of other conditions, such as occupational exposure and collagenosis ⁽¹¹⁾. Therefore, the exact incidence, prevalence and mortality of IPF are unknown. Based on available data, the incidence is estimated at approximately 7-11/100,000 inhabitants, and the prevalence is close to 3-20/100,000 inhabitants ⁽⁵⁾.

Johnson et al. ⁽⁶⁾ analyzed the mortality due to IPF in England from 1979 to 1988, and found a two-fold increase of total mortality and a proportional increase of the adjusted mortality. The analysis of the hospital records of some patients known as having IPF suggests that the simple checking of the death certificates may cut the number of cases in half. Mannino et al ⁽⁸⁾ separated the death certificates that contained the code of IPF from those that listed it as the direct cause of death. Whereas the adjusted rate for age per 100,000 inhabitants increased 14% from 1979 to 1991 for IPF patients, this increase was of 63% in the same period, for patients whose determining cause of death was IPF. This study also revealed a trend of increase when ICD refers to post-inflammatory pulmonary fibrosis (516.3) and an ICD decrease when it refers to IPF (515) in the USA.

Hubbard et al. ⁽⁷⁾ compared the mortality rates due to IPF in seven different countries and found significant differences, being the highest rates in England and the lowest in the USA and Germany. In Germany and New Zealand there was no increase of mortality, and in the USA there was even a reduction of mortality, likewise attributed to a transferring of the code to post-inflammatory pulmonary fibrosis.

When comparing the rates of the State of RS (Brazil) with the other above-mentioned studies, we realize that the first ones are much lower. We can only hypothesize that such difference can be explained by the control of death recordings, as well as by the medical practice per se, including a higher availability of centers with diagnostic resources necessary for the correct identification of IPF cases and its follow-up in industrialized countries. It is important to emphasize that we analyzed only the cases in which IPF was mentioned in the certificate as the direct cause of death, which, as previously explained, tends to considerably reduce the number of cases. Nonetheless, we must also consider the factors which can falsely increase the mortality due to IPF, such as the inclusion of patients with fibrotic consequence of infectious diseases, either mycobacterial or mycotic, both highly prevalent in our environment, of the migration of IPF patients to the State of RS, secondary to the implementation of a pulmonary transplant program, recently disclosed in the media.

We could also demonstrate that the increase of total and adjusted mortality from IPF in the State of RS is at least partially due to coding changes which occurred in 1996, by comparing the means in the three previous years with the means of the three years following the new ICD. The difference was highly significant, as opposed to 1978. Such difference would hardly be justified by a real mortality increase in such a short time, considering the average survival of patients with IPF. However, the increase found throughout the three decades is not justified only by this fact.

In summary, we observed an increase of mortality rate from IPF in RS from 1979 to 2000, and this increase is partially due to coding changes which occurred in 1996. Mortality from IPF in the State of RS seems to be lower than that observed in other industrialized countries. Other studies are necessary to clear out if these findings are actually due to the increased prevalence of IPF or due to factors related to the notification and/or diagnosis.

Received for publication on 7/29/02.

Approved, after revision, on 3/10/03.

Publication Dates

History