Article Text

Download PDFPDF

Prognosis and causes of death of patients with acute exacerbation of fibrosing interstitial lung diseases
  1. Johanna Salonen1,2,
  2. Minna Purokivi3,
  3. Risto Bloigu4 and
  4. Riitta Kaarteenaho1,2
  1. 1Respiratory Medicine, Research Unit of Internal Medicine, University of Oulu, Oulu, Finland
  2. 2Medical Research Center, Oulu University Hospital, Oulu, Finland
  3. 3The Center of Medicine and Clinical Research, Division of Respiratory Medicine, Kuopio University Hospital, Kuopio, Finland
  4. 4Medical Informatics and Statistics Research Group, University of Oulu, Oulu, Finland
  1. Correspondence to Dr Johanna Salonen; johanna.salonen{at}


Background The aim of this study was to compare the clinical characteristics, causes of death and factors impacting on the prognosis of patients with idiopathic pulmonary fibrosis (IPF) and other fibrosing interstitial lung disease (FILD) with a history of acute exacerbation (AE) of IPF or FILD.

Methods Retrospective data of hospital treatment periods caused by AE-IPF and AE-FILD were collected from medical records. Clinical features and survival data of IPF and non-IPF cases were evaluated and compared. The underlying and immediate causes of death were gathered from death certificates.

Results A total of 128 patients fulfilled the criteria for inclusion. IPF (n=79/62%), rheumatoid arthritis-associated interstitial lung disease (RA-ILD; n=17/14%) and asbestosis (n=11/8.6%) were the most common FILD subgroups in the study. The median survival after hospitalisation in AE-IPF was 2.6 months compared with 21 months in other AE-FILDs (p<0.001). The survival difference was not explained by age, gender or pulmonary function test results at the time of hospitalisation. Patients with non-specific interstitial pneumonia and RA-ILD had the most favourable prognosis. ILD was the most common underlying cause of death in both patients with IPF and with other FILD accounting for 87% and 78% of deaths, respectively.

Conclusions We detected a significantly longer survival in AE of patients with non-IPF compared with that of AE-IPFs. The prognosis of patients was affected by the underlying lung disease since pulmonary fibrosis was the underlying cause of death in the majority of all patients with FILD having experienced an AE.

  • idiopathic pulmonary fibrosis
  • interstitial lung disease
  • acute exacerbation
  • asbestosis
  • survival

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Key messages

Why read on?

  • This is the first European study investigating comprehensive data on patients with acute exacerbation of fibrosing interstitial lung disease.

What is the key question?

  • What are the prognoses and causes of death in patients with acute exacerbation of fibrosing interstitial lung disease?

What is the bottom line?

  • The survival after acute exacerbation was dependent on the pulmonary fibrosis type of the patient.


Idiopathic pulmonary fibrosis (IPF) and other fibrosing interstitial lung diseases (FILD) are serious, often fatal disorders in individual patients.1–4 The phenomenon of acute exacerbation (AE) has been associated with high mortality in IPF, and AEs also seem to cause significant mortality in other FILDs.5–14 According to previous studies, every year about 5%–15% of patients with IPF will experience an AE.5 13 Recently, it was shown that the incidence of AE or death in a study consisting of various types of patients with FILD was 7.8% per year.15 The treatment of AEs is based on clinical practice in each hospital, although it mainly involves the administration of corticosteroids and antibiotics, while there is a lack of randomised, controlled studies investigating this subject.5 16

The first guidelines for diagnostic criteria of AE-IPF were issued in 2007 with the definition being revised in 2016.5 17 It has been proposed that it would be beneficial to apply the definition of AE-IPF also to non-IPF FILDs.16 Currently, AE-IPFs are divided into triggered and idiopathic exacerbations, and the exclusion of infection by bronchoalveolar lavage or endotracheal aspirate is no longer a requirement as was the case in the older recommendations. Thus, a significant number of cases that would fulfil the present criteria for triggered AEs have been excluded from earlier research materials.

AE-IPF mortality rates have tended to be lower in more recent studies when compared with values published prior to 2007. Recent data collected from several previous studies by Kondoh et al17 including 995 patients indicates 1-month survival of 66% and a 3-month survival of 44 %, whereas in older studies, the mortality rates have tended to be much higher, 80%–100%.17 18 Much less is known about survival after AEs in patients with non-IPF FILDs. Most studies having investigated both patients with IPF and non-IPF with AE have revealed equally high mortality rates in both groups,8 10–12 14 although a poorer survival of patients with AE-IPF has also been reported.6 7 13

The aim of this study was to investigate the prognosis and factors impacting on the outcome of patients with different types of FILDs, including both IPF and non-IPF patients treated non-electively due to AE in the respiratory wards or intensive care units (ICUs) in the hospitals of the Northern Ostrobothnia Hospital District, that is, Oulu University Hospital (OUH) and Oulaskangas Hospital (OH) during 2008–2017 in northern Finland. Both AE and the types of interstitial lung disease (ILD) were redefined by the current international criteria. Clinical features, including age, gender, radiological findings, pulmonary function test (PFT) results, smoking status, pharmacological therapy, survival data and causes of death, were evaluated and compared between IPF and non-IPF cases.


Patient and public involvement

It was neither appropriate nor possible to involve patients or the public in the design, or conduct, or reporting, or dissemination plans of our research.

Patient and data collection

The study cohort consists of patients hospitalised due to an AE of IPF or non-IPF FILD in OUH and OH between 1 January 2008 and 31 December 2017. A flow chart of the study is presented in figure 1. The patients and their treatment periods were collected from the hospital’s medical records according to the International Classification of Diseases version 10 (ICD-10) codes J84.1, J84.8 and J84.9. To find the hospitalisations related to asbestosis and rheumatoid arthritis-associated ILD (RA-ILD), another search was performed with ICD-10 codes J61, J99, J99.0* and J99*M05.1. The type of ILD was re-evaluated using the current international criteria.19 20 A more detailed description of the diagnostic evaluation of the patients is presented in online supplementary material.

Figure 1

Flow chart of the study. FILD, fibrosing interstitial lung diseases; ICD-10, International Classification of Diseases version 10; ILD, interstitial lung disease; IPF, idiopathic pulmonary fibrosis; UIP, usual interstitial pneumonia.

The data included date of birth, age at diagnosis and at hospitalisation, gender, smoking status at hospitalisation, date of FILD diagnosis, date of hospital admission, duration of hospital treatment, PFT, high-resolution CT (HRCT) reports and images, use of long-term oxygen treatment (LTOT) prior to hospital admission, pharmacological therapy for FILD prior to admission and during hospital treatment and the possible trigger for AE. The diagnosis date recorded was the first visit to the pulmonary clinic or the hospital admission date for those diagnosed with ILD during their hospital treatment. The date and causes of death for each study subject were collected from death certificates housed in the national registry of Statistics Finland. The histological diagnoses from surgical biopsies and autopsies were reviewed. Patients with a pattern of probable or consistent with usual interstitial pneumonia (UIP) in HRCT according to the present classification or histopathological finding of UIP were classified as UIP in statistical analysis. Patients with ILD with less than five pack years of smoking history were classified as non-smokers. Recorded PFT values at hospitalisation were measured maximally 6 months prior to or after the hospital admission date. Survival time was calculated from the hospital admission date to death, lung transplantation or last follow-up date (31 August 2019).

Readmissions after AE within 3 months were not regarded as new, separate AEs, if the clinical presentation of FILD had not been stabilised in that time frame and a new episode meeting the criteria of AE could not be ensured. In one patient without HRCT data, the diagnosis of AE was based on chest X-ray, typical clinical features and histological findings (diffuse alveolar damage (DAD) in addition to the UIP pattern in lung tissue sample). The causes for hospitalisations were classified into AEs and other reasons applying the novel criteria of AE-IPF to all FILDs.5 The clinical information was collected systematically from electronic patient records dating back about 20 years in OUH and OH in a form specially designed for the present study. Complete case analysis was used to deal with variables with missing data.

Statistical analysis

IBM SPSS Statistics for Windows, V.25.0 (IBM Corp.) was used to perform statistical analysis and Origin(Pro), Version 2019b (OriginLab Corporation, Northampton, Massachusetts, USA) was utilised for graphs. Means and SD were calculated for continuing variables, which were normally distributed. Medians and IQR were used for continuing variables which were not normally distributed. Group differences of continuous variables were tested by using independent samples t-test and variance analyses or Mann-Whitney U test. Differences in the categorised variables were calculated with χ2 or Fisher’s exact test. Survival was estimated by using Kaplan-Meier curves and the groups were compared with each other by using log rank tests. Log rank test with a linear trend for factor levels was used when the survival differences of five different FILD subgroups were evaluated. The effects of gender, age, PFT results, smoking status and corticosteroid treatment preceding hospitalisation on survival were assessed with Cox regression analysis. As this was a register-based retrospective study and the majority of study subjects were deceased, no patient consent forms were gathered in accordance with Finnish legislation.


Most patients with AE were men with either IPF or CTD-ILD

We investigated a total of 128 patients (79 with IPF and 49 with other FILD) with 142 non-elective hospital treatment periods due to AE (online supplementary E-Table 1 and E-Table 2). Table 1 shows each hospital admission analysed as a unique event. The majority of patients had IPF (62 %), while patients with CTD-ILD were the most common subgroup (16 %). Eleven patients (8.6 %) had an asbestosis-related AE accounting for 13 treatment periods. A slight majority of treatment periods (56%) took place in a respiratory ward, not in the ICU or intermediate care unit, without the need for non-invasive or invasive mechanical ventilation (IMV) support. Sixteen patients (13 %) required treatment with IMV; six of them survived the hospital treatment. In 86% of all treatment periods, there were no identifiable triggers for AE. The medical treatment of the patients is presented in online supplementary E-Table 3.

Table 1

Data of the hospital treatment periods of the patients with acute exacerbation (AE) of idiopathic pulmonary fibrosis (IPF) and other fibrosing interstitial lung diseases (FILD)

IPF was associated with shorter survival time and pulmonary fibrosis was the most common underlying and immediate cause of death

Median survival after hospitalisation in AE-IPF was 2.6 months, whereas patients with some other FILD had a much longer median survival, 21 months (p<0.001; figure 2(A)). The difference was clear even after adjustment for gender, age, smoking, forced vital capacity (FVC) and diffusion capacity for carbon monoxide (DLCO) determined at the time of hospitalisation (HR 2.39 (95% CI 1.41 to 4.05), p=0.001). When corticosteroid treatment preceding hospitalisation was included into the multivariate model, HR was even higher 2.86 (95% CI 1.65 to 4.96), with a p value<0.001. The effect of corticosteroid treatment on survival time was significant in the above-mentioned adjustment model (p=0.006). When patients with FILD were divided into five subgroups, survival time was dependent on the FILD subtype (p<0.001; figure 2(B)). Patients with IPF or asbestosis seemed to have the shortest survival compared with other FILDs, and non-specific interstitial pneumonia (NSIP) was the subgroup with the most favourable prognosis. When possible risk factors for death were analysed separately for IPF and other FILD groups (online supplementary E-Table 4), it was revealed that age more than 80 years, UIP pattern and LTOT preceding hospitalisation seemed to elevate the risk for death in the non-IPF group, while this was not the case in patients with IPF. In contrast, treatment with corticosteroids preceding hospitalisation, treatment in an intermediate care unit or treatment with steroids (more than 150 mg/day) during AE were risk factors for death in IPF, but not in other FILDs. Pulmonary fibrosis was the most common underlying and immediate cause of death accounting for 84% and 45% of deaths, without statistically significant difference between patients with IPF and non-IPF (table 2).

Table 2

Underlying and immediate causes for death in patients with idiopathic pulmonary fibrosis (IPF) and other fibrosing interstitial lung disease (FILD)

Figure 2

Survival time from the first hospitalisation caused by an acute exacerbation of a fibrosing interstitial lung disease (FILD). (A) Patients with idiopathic pulmonary fibrosis (IPF) versus non-IPF and (B) five different pulmonary fibrosis subgroups. NSIP, non-specific interstitial pneumonia; RA-ILD, rheumatoid arthritis-associated interstitial lung disease.


We studied a cohort of 128 patients from northern Finland with AE-FILD, assessing their clinical characteristics and prognosis. As reported previously, the patients with AE-IPF had a poor prognosis.5 17 18 We detected a longer posthospitalisation survival in patients with non-IPF in comparison with patients with IPF, a finding which has not been so explicit in previous studies restricted to AE of patients with IPF and non-IPF. Furthermore, we could discern survival differences between five subclasses of FILD, namely IPF, NSIP, RA-ILD, asbestosis and other FILD, which has not been shown in earlier studies. In addition, we could demonstrate that most patients with a history of AE-FILD also died due to their underlying lung disease, a finding that has not been identified earlier, because there are no previous studies on causes of death in study populations consisting of merely patients with AE-FILD.

AE-IPF and AE of CTD-ILD formed the largest groups in our cohort, a finding consistent with the previous studies on AE in patients with IPF and non-IPF.6–8 10 21 22 In contrast to these publications, our study revealed that the patients with an AE of asbestosis were more common than those in whom the AE was related to NSIP, chronic hypersensitivity pneumonitis or scleroderma-associated ILD. To our knowledge, there are no previous publications examining patients with asbestosis with clinically confirmed AE, although a previous Japanese survey stated that histological DAD was observed in 35% of 40 autopsy cases of asbestosis,23 which suggests that an AE might evoke significant mortality among patients with asbestosis, thus supporting the results of our study.

Most of the AE-IPFs found here occurred in winter or spring, which is consistent with previous investigations.24 25 Patients with IPF experiencing an AE had a poor prognosis with a median survival of only 2.6 months, a result in line with earlier reports of median survival times ranging between 1 and 4 months after AE-IPF.9 13 24 26 27 Unlike most of the published reports, we detected a clear difference in survival between patients with IPF and non-IPF after the AE. As presented in table 3, three studies from USA investigating cohorts with both patients with IPF and other FILD treated for acute respiratory failure, reported 1-year mortality rates of more than 50% for patients with non-IPF, that is, much higher than in our study where patients with non-IPF had a median survival of 21 months.6–8 The studies mentioned above, however, did not use the current definition of AE-IPF in their inclusion criteria, and in addition, one study included patients treated in the ICU,8 which is known to be a risk factor for death, at least in patients with IPF.28 The other published studies concerning AE-FILD have been conducted in Japan,9–14 of which only one demonstrated longer survival of patients with non-IPF compared with that of IPF13; in another study, the prognosis of AE-IPF compared with other FILDs was even better.9 Altogether, it can be stated that the earlier publications focusing on acute respiratory worsening or AE of IPF or non-IPF FILDs have consisted of small and heterogenic study populations with variable inclusion criteria, which complicate making comparisons between the investigations. Our results showing a remarkable difference in survival time between patients with IPF and non-IPF after AE suggested that AE-FILD and AE-IPF cannot be considered as completely identical disease entities, and furthermore, even subtypes of non-IPF FILDs may have a variable prognosis after an AE.

Table 3

Previous publications which have examined an acute respiratory worsening of idiopathic pulmonary fibrosis (IPF) and fibrosing interstitial lung diseases (FILD)

It is noteworthy that all previously published studies using the current criteria of AE-IPF for inclusion and involving both patients with IPF and non-IPF have been implemented in Japan. Differences in genetic and environmental factors may, however, have some impact on the disease course after AE-FILD. For instance, it has been proposed that African Americans with IPF, who are less likely than Caucasians to develop IPF, nonetheless die younger than non-Hispanic Caucasians with pulmonary hypertension.29 Environmental factors, such as air pollution, have been reported to be associated with AE-IPF.30 In a previous study, the mortality of IPF was associated with an increased cumulative exposure to particulate matter.31 However, further data are needed to establish convincing evidence of the effect of different genetic and environmental backgrounds on prognosis of AE-FILDs.

We observed no differences in age, gender, smoking history, FVC or DLCO between the IPF and non-IPF groups and furthermore, these factors did not explain the survival difference after AE between patients with IPF and other FILD. Clinical factors like gender, age, PFT and Gender-Age-Physiology (GAP) index have been demonstrated to be relevant in the prognosis in patients with IPF and non-IPF with or without a history of an AE.28 32–35 There are also studies designating these clinical factors as risk factors for AE,9–11 13 36–38 but their relevance in prognostic assessment of the patients who have already received an AE-FILD is less clear. A recent study described a risk assessment model for patients with AE-FILD including male sex, interstitial pulmonary fibrosis diagnosis, body mass index and some other parameters relevant to the circumstances behind their treatment in the ICU.8 However, in our study, less than 20% of patients were treated in ICU with no statistically significant differences between IPF and non-IPF, that is, the above-mentioned risk score might not be applicable for the majority of patients with AE-FILD. Furthermore, since the prognostic significance of age, LTOT, treatment unit and corticosteroid treatment seemed to be different in patients with AE-IPF and non-IPF, it might prove difficult to plan a functional risk-predicting model for all types of patients with AE-FILD.

In our study, corticosteroid treatment preceding the hospitalisation was a risk factor for shortened survival in patients with IPF, a result which supported the previous guidelines according to which corticosteroids or other immunosuppressive medication is not recommended for IPF.39 Since most hospital admissions were accomplished when no antifibrotic medication was available in Finland, rather few patients were administered nintedanib or pirfenidone in our study, even though current evidence supports using antifibrotic treatment for patients with IPF.40 However, we were not able to show any survival benefit of antifibrotic drugs in seven patients receiving pirfenidone or nintedanib.

The pharmacological therapy for AE was similar in both FILD-groups, being mostly corticosteroids and antibiotics. The mortality rates of our study were equal or even lower than in the investigations where pulse steroid treatment has been more common,7 9 12 14 with the exception of a study which estimated a 3 months’ mortality rate of only about 35% for 17 patients with AE-IPF treated with pulse steroids.11 We observed that pulse steroid treatment or treatment with a steroid dose more than 150 mg/day was risk factors for death in patients with IPF; however, this finding cannot be the explanation for higher mortality of patients with IPF, because only 17 patients (22 %) had been treated with high (more than 150 mg) corticosteroid dose. Possibly those patients suffering from more severe respiratory failure were selected to receive higher doses of steroids, and therefore it is difficult to draw any conclusions of the utility of this treatment.

As far as is known, this is the first time that the causes of death have been determined in a study population involving AE-IPF and AE of patients with other FILD. Pulmonary fibrosis was the underlying cause of death in 84% of patients, with no statistically significant difference between patients with IPF and other FILD. According to data from the USA, pulmonary fibrosis was the underlying cause of death in 60% of patients with IPF with a similar result (67.5 %) presented in a Finnish cohort of 117 patients with IPF .1 3 Less is known about the causes of death in patients with other FILD, even though high mortality, though usually somewhat lower than in IPF, is known to be related to systemic sclerosis-associated ILD, RA-ILD, chronic hypersensitivity pneumonia, fibrotic NSIP and unclassifiable ILD.2 41–47 It seems that an AE of FILD is a characteristic of a more progressive disease course, which is reflected in the relatively high proportion of pulmonary fibrosis deaths observed here. The immediate causes of death were mainly respiratory in both patients with IPF and non-IPF, with no statistically significant differences between the groups. The higher mortality rate of patients with AE-IPF compared with other AE-FILDs was not reflected in these rates, which is probably related to the fact that there is no specific ICD-10 code for an AE. The immediate causes of death in patients who died of AE were mostly recorded as ‘pulmonary fibrosis’ or ‘pneumonia’ diagnoses. It is possible that some AE-FILD deaths remain unrecognised, for example, because HRCT imaging is not expected to provide any benefits for an individual patient in the terminal phase of his/her pulmonary disease.

The retrospective study design, the study implementation being limited to two hospitals and the heterogeneity of the non-IPF study population can be considered as limitations of our study. Moreover, some AE-FILDs may have been excluded due to insufficient data for evaluation (eg, lack of HRCT) or faulty diagnosis codes in the medical records. Nevertheless, all non-elective hospitalisations of each study subject were assessed systematically, and we succeeded in gathering comprehensive data despite the retrospective study design.


To conclude, we have investigated a diverse group of patients with AE-FILD in a real-life setting; their prognosis varied depending on the underlying ILD. Pulmonary fibrosis represents a fatal condition and proved to be the underlying cause of death in the majority of all patients with FILD experiencing an AE. Further research is needed on this highly challenging disease group with such a poor prognosis.


The authors would like to thank Eerika Keskitalo (BM) for help with the SPSS program and Ewen MacDonald for language and editorial assistance.



  • Contributors JS collected the study material, planned the data collection form, interpreted and analysed the data and prepared the draft of the manuscript. RB planned and participated in the statistical analyses. MP and RK participated in planning the data collection form, study design and the interpretation of the data. RK managed the study and contributed substantially to data interpretation by re-evaluating study patients and data from medical records. All authors participated in the preparation of the manuscript, read and approved the final manuscript.

  • Funding JS has received personal grants for scientific work from Foundation of the Finnish Anti-Tuberculosis Association and the Research Foundation of the Pulmonary Diseases HES. RK has received grants for the study group from Foundation of the Finnish Anti-Tuberculosis Association, the Research Foundation of the Pulmonary Diseases, Jalmari and Rauha Ahokas Foundation and the Research Foundation of North Finland.

  • Competing interests JS reports congress fees and travel costs from Boehringer-Ingelheim, Novartis, Orion Pharma, Ratiopharm and Roche, and lecture fees from Boehringer-Ingelheim, Chiesi, GlaxoSmithKline, Mundipharma, Orion Pharma and Roche outside the submitted work. MP reports lecture fee from Boehringer-Ingelheim Finland and congress fee and travel costs from Roche, outside the submitted work. RB has nothing to disclose. RK reports consultant fees from GlaxoSmithKline and Boehringer-Ingelheim, lecture fees from Roche and Boehringer-Ingelheim, and a congress travel costs from Orion outside the submitted work.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Patient consent for publication Not required.

  • Ethics approval The study protocol was approved by the Ethical Committee of the Northern Ostrobothnia Hospital District (statement 2/2015). Permission to use death certificates was given by Statistics Finland (Dnro: TK-53-515-15). The study was conducted in compliance with the Declaration of Helsinki.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability statement Data are available upon reasonable request. The datasets generated and analysed during the current study are not publicly available due to the relatively small population of Northern Finland since we could not guarantee individuals’ anonymity as the data were collected in a detailed manner, but it is available from the corresponding author on reasonable request.