Incidence and prevalence of interstitial lung diseases in Norway: analysis of the national Norwegian patient registry

Abstract

Background

Interstitial lung diseases (ILDs) are progressive chronic lung diseases associated with poor prognosis despite available disease-modifying therapy. Reported incidence and prevalence of ILD varies, and few studies have described epidemiology across the range of ILD subtypes in complete population-based cohorts. This study aimed to estimate incidence and prevalence of ILD, overall and by subtype, in a complete nationwide cohort from Norway.

Methods

The Norwegian Patient Registry and Norwegian Prescribed Drug Registry were used to identify all patients with diagnosis codes identifying an ILD between 2011 and 2023. Using diagnosis code-based algorithms adapted from previous studies, the incidence and prevalence of ILD subtypes was estimated overall by year, and by subgroups defined by age and sex.

Results

Using a broad definition, the overall prevalence of ILD in Norway has increased to over 450 cases per 100,000 from 2012 to 2023. The prevalence of idiopathic pulmonary fibrosis (IPF) has increased to 46 per 100,000 and prevalence of connective tissue disease related ILD (CTD-ILD) has increased to nearly 42 per 100,000. Current estimates of incidence ranged from about 1 per 100,000 for hypersensitivity pneumonitis, to about 5 per 100,000 for CTD-ILD, to 8–10 per 100,000 for IPF.

Conclusions

Incidence and prevalence of ILDs in Norway are comparable to published reports from other regions. The prevalence of ILDs in Norway appears to be increasing substantially.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12890-026-04515-5.

Background

Interstitial lung diseases (ILDs) comprise a diverse group of chronic lung diseases characterized by inflammation and/or fibrosis affecting interstitial or alveolar lung tissue. While prognosis and disease course vary substantially by ILD subtype, many patients will experience progressive loss of lung function and are at high risk of mortality [1, 2]. Accordingly, ILDs have recently been the focus of significant drug development efforts, leading to the approval of several antifibrotic drugs that slow, but do not prevent, disease progression [3–7]. Despite broad interest in ILDs, defining their epidemiology has been challenging, likely due to several factors. Heterogeneity in research methodology and study cohorts, differences in disease detection related to practice patterns or settings, changes in diagnostic criteria over time, and perhaps true changes over time in incidence may all contribute [8–11]. Few studies have evaluated incidence and prevalence across ILD subtypes in truly population-based cohorts.

Because ILDs are relatively uncommon diseases, larger administrative datasets have often been used to estimate their incidence and prevalence, and previous studies have developed algorithms based on International Classification of Diseases (ICD) codes for identifying and classifying patients with ILD [12]. Such algorithms have been refined over time to improve specificity with respect to idiopathic pulmonary fibrosis (IPF) in particular [13]. In recent years, similar algorithms have been established for classification of non-IPF subtypes [14]. Despite these advantages, many administrative databases are limited to specific insurance providers and thus may not be representative of the overall population. By contrast, countries with single-payer healthcare systems may offer opportunities to advance understanding of the true epidemiology of ILD.

This study used the Norwegian Patient Registry, which contain administrative records of all encounters within the national healthcare system and private practice specialists accepting referrals from the national system, to estimate the incidence and prevalence of ILD in Norway, overall and by ILD subtype.

Methods

Data sources

The Norwegian Patient Registry (NPR) and the Norwegian Prescribed Drug Registry were used to identify patients with ILD. The NPR contains records for all patient encounters within the nationwide publicly funded specialist healthcare system, including all hospital encounters, ambulatory specialist encounters, and private practice specialists contracting with the national system. Records include basic demographic information, encounter date and location, and all associated international classification of disease-10 (ICD-10) diagnosis codes, diagnostic procedure codes and month and year of death. The Norwegian Prescribed Drug Registry includes records of all prescriptions dispensed by Norwegian pharmacies, including date dispensed, the Anatomical Therapeutic Chemical (ATC) codes, and the International Classification of Diseases (ICD-10) or International Classification of Primary Care (ICPC) codes corresponding to the indication given by the prescriber. Data from NPR and the Norwegian Prescribed Drug Registry were linked using the unique personal identification number of every Norwegian resident, which made it possible to follow individual patients over time.

Publicly available official census data from Statistics Norway [15] was used to determine the number of potential adult individuals at risk, overall and by age and sex categories, to serve as the denominator for estimates of incidence and prevalence.

Identification of patients with ILD

The overall dataset contained all patients with at least one encounter in the NPR associated with a diagnosis code for ILD (Supplemental Table 1), during the period 2008 to 2023. The first ILD diagnosis code established the index date, for estimation of incidence. Patients under 18 years of age at the time of first ILD diagnosis code were excluded. To distinguish incident from prevalent cases, data from 2008 to 2010 was used as a “lookback” period. Consequently, patients whose first ILD diagnosis appeared in this period were excluded from incidence analyses, and incidence is estimated from 2011 onward.

Classification of ILD subtype

Patients with ILD were then classified by subtype. In keeping with prior studies [14], some ICD10 codes were sufficient to establish both the presence of ILD and the subtype, whereas other less specific ILD codes required additional ICD10 codes corresponding to an underlying condition or cause to establish the subtype (Supplemental Table 1). Each patient was assigned to one of the following subtypes: connective tissue disease-associated ILD (CTD-ILD), hypersensitivity pneumonitis (HP), other exposure-related ILD, sarcoidosis-ILD, idiopathic pulmonary fibrosis (IPF), multiple, or unclassifiable.

Codes for underlying conditions appearing up to 730 days (2 years) before or after the first ILD code were considered in the subtype classification process. Patients with multiple (“competing”) underlying condition codes during the ± 730-day window were classified based on the last underlying condition appearing during the window where possible, a strategy chosen to reflect that classification of ILD often evolves over time as additional clinical information becomes available. When multiple underlying conditions were coded during the last encounter within the window, the ILD subtype was classified as “multiple.” Once classified, a given patient’s ILD subtype applied to the entire observation period, i.e. retroactively to the index date for ILD. Codes for underlying conditions occurring > 730 days before or after the first ILD code were not used for classification. Classification as IPF required the absence of codes corresponding to any underlying condition associated with another ILD subtype during the classification window, in keeping with previous studies [12, 13].

Analyses

For descriptive purposes, age, sex, comorbidities, diagnostic procedures and ILD-related treatments were summarized by ILD subtype and expressed as median (interquartile range) or n (%). Comorbidities were defined as any ICD10 code for the given condition within one year before or after the ILD index date. Treatments were defined as any use from ILD index date to the end of available follow-up for a given patient.

Incident diagnoses were defined by the year of the encounter with the first ILD diagnosis code and are expressed as cases per 100,000 person-years. Patients were considered prevalent patients during the index year and in every subsequent year up to, but not including, the year of death. Prevalence is expressed as cases per 100,000 individuals. Because of the large proportion of cases in which ILD subtype classification was not possible, overall ILD prevalence is reported both with and without unclassifiable patients included, i.e. for “all ILD” and “classifiable ILD.”

Ethical approvals

The study was approved by the Regional Ethics Committee for Southeastern Norway (Case #519,297), with a waiver of informed consent.

Results

Patient characteristics and treatment

In all, 44,988 unique patients had at least one ICD10 code corresponding to an ILD between 2011 and 2023 and were included in the analysis (Table 1). Of these, 70.6% were unclassifiable based on ICD-10 codes. Among those who were classifiable, the most common subtype was IPF (12.1% of the overall cohort), followed by CTD-ILD (6.8%). Patients with IPF were generally older, with a median age of 75.0 years, and 65.3% were male. Patients with CTD-ILD were comparatively younger, with a median age of 66.0 years, and 36.2% were male. Patients with HP and sarcoidosis-ILD were generally the youngest, with a median age of 57.0 years for both groups. The most common comorbidities were heart failure, chronic obstructive pulmonary disease (COPD), and atrial fibrillation.

Table 1.

Demographic and clinical characteristics

	IPF N  = 5442	CTD-ILD N  = 3062	Other exposure N  = 3020	Sarcoidosis-ILD N  = 831	HP N  = 625	Multiple N  = 221	Unclassified ILD N  = 31,787	All N  = 44,988
Age	75 (67–82)	66 (55–75)	71 (62–78)	57 (47–70)	57 (45–69)	72 (64–77)	71 (56–81)	71 (58–80)
Male	3555 (65.3%)	1109 (36.2%)	1891 (62.6%)	527 (63.4%)	307 (49.1%)	117 (52.9%)	19,981 (62.9%)	27,487 (61.1%)
Comorbidities
Ischemic heart disease	696 (12.8%)	233 (7.6%)	262 (8.7%)	48 (5.8%)	34 (5.4%)	26 (11.8%)	3111 (9.8%)	4410 (9.8%)
Heart failure	1202 (22.1%)	394 (12.9%)	489 (16.2%)	85 (10.2%)	49 (7.8%)	42 (19.0%)	5983 (18.8%)	8244 (18.3%)
COPD	1240 (22.8%)	396 (12.9%)	546 (18.1%)	80 (9.6%)	51 (8.2%)	31 (14.0%)	3783 (11.9%)	6127 (13.6%)
Connective tissue disease	41 (0.8%)	2638 (86.2%)	24 (0.8%)	49 (5.9%)	9 (1.4%)	172 (77.8%)	952 (3.0%)	3885 (8.6%)
Chronic liver disease	83 (1.5%)	64 (2.1%)	67 (2.2%)	11 (1.3%)	3 (0.5%)	6 (2.7%)	916 (2.9%)	1150 (2.6%)
Lung cancer	171 (3.1%)	66 (2.2%)	789 (26.1%)	11 (1.3%)	8 (1.3%)	11 (5.0%)	762 (2.4%)	1818 (4.0%)
Pulmonary hypertension	310 (5.7%)	148 (4.8%)	73 (2.4%)	21 (2.5%)	10 (1.6%)	12 (5.4%)	455 (1.4%)	1029 (2.3%)
Atrial fibrillation	835 (15.3%)	303 (9.9%)	536 (17.7%)	54 (6.5%)	33 (5.3%)	32 (14.5%)	5228 (16.4%)	7021 (15.6%)
Pulmonary embolism	145 (2.7%)	98 (3.2%)	201 (6.7%)	23 (2.8%)	13 (2.1%)	13 (5.9%)	942 (3.0%)	1435 (3.2%)
Venous thromboembolism	65 (1.2%)	46 (1.5%)	89 (2.9%)	5 (0.6%)	3 (0.5%)	1 (0.5%)	434 (1.4%)	643 (1.4%)
Diagnostic procedures*
Bronchoscopy	689 (12.7%)	397 (13.0%)	334 (11.1%)	207 (24.9%)	137 (21.9%)	33 (14.9%)	2978 (9.4%)	4775 (10.6%)
Surgical lung biopsy	113 (2.1)%	28 (0.9%)	7 (0.2%)	8 (1.0%)	17 (2.7%)	4 (1.8%)	71 (0.2%)	248 (0.6%)

*within one year following first ILD diagnosis

Medication use across the entire follow-up period is summarized in Table 2. Nintedanib and pirfenidone were used by 612 (11%) and 517 (10%) patients with IPF during follow-up, respectively, and in 0–2% of those with other ILD subtypes. Treatment with prednisolone was most frequent among those with CTD-ILD, used by 2376 (78%) during follow-up, but also by over 50% of patients in all other ILD subtype groups. Among those with CTD-ILD, mycophenolate, methotrexate, and tumor necrosis factor-alpha inhibitors were used by 338 (11%), 757 (25%) and 368 (12%), respectively. Methotrexate was also used by 136 (16%) of those with sarcoidosis. Non-steroidal immunosuppressive agent use was infrequent in other subgroups.

Table 2.

ILD-related treatments at any time during follow-up

	IPF N = 5442	CTD-ILD N = 3062	Other exposure N = 3020	Sarcoidosis-ILD N = 831	HP N = 625	Multiple N = 221	Unclassified ILD N = 31,787	All N = 44,988
Nintedanib	612 (11.2%)	64 (2.1%)	9 (0.3%)	13 (1.6%)	11 (1.8%)	7 (3.2%)	162 (0.5%)	878 (2.0%)
Pirfenidone	517 (9.5%)	10 (0.3%)	2 (0.1%)	10 (1.2%)	2 (0.3%)	2 (0.9%)	108 (0.3%)	651 (1.4%)
Prednisolone	2818 (51.8%)	2376 (77.6%)	1884 (62.4%)	592 (71.2%)	383 (61.3%)	152 (68.8%)	8428 (26.5%)	16,633 (37.0%)
Mycophenolate	94 (1.7%)	338 (11.0%)	28 (0.9%)	20 (2.4%)	18 (2.9%)	10 (4.5%)	264 (0.8%)	772 (1.7%)
Rituximab	1 (0.0%)	0 (0.0%)	0 (0.0%)	0 (0.0%)	0 (0.0%)	0 (0.0%)	0 (0.0%)	1 (0.0%)
Azathioprine	55 (1.0%)	256 (8.4%)	13 (0.4%)	18 (2.2%)	24 (3.8%)	11 (5.0%)	219 (0.7%)	596 (1.3%)
Methotrexate	70 (1.3%)	757 (24.7%)	38 (1.3%)	136 (16.4%)	10 (1.6%)	40 (18.1%)	390 (1.2%)	1441 (3.2%)
Cyclophosphamide	7 (0.1%)	8 (0.3%)	8 (0.3%)	1 (0.1%)	0 (0.0%)	0 (0.0%)	21 (0.1%)	45 (0.1%)
Tumor necrosis alpha inhibitors	45 (0.8%)	368 (12.0%)	30 (1.0%)	25 (3.0%)	11 (1.8%)	17 (7.7%)	240 (0.8%)	736 (1.6%)
Tocilizumab	2 (0.0%)	77 (2.5%)	2 (0.1%)	1 (0.1%)	0 (0.0%)	4 (1.8%)	7 (0.0%)	93 (0.2%)

Incidence

The overall incidence of ILD increased from approximately 57 cases per 100,000 in 2011 to 73 cases per 100,000 in 2023 (Fig. 1). Most of this increase represented patients whose ILD subtype could not be classified based on ICD-10 codes. Incidence of IPF ranged from 7 to 9.5 cases per 100,000 during the study period, with an apparent small increase in the period 2018–2019 that was sustained thereafter. Incidence of CTD-ILD varied from approximately 3.5 to 5.5 cases per 100,000, with a slight apparent general increase over time.

Fig. 1.

Incidence overall and by ILD subtype

Prevalence

The overall prevalence of ILD increased steadily from 164 to 465 patients per 100,000 individuals during the study period (Fig. 2), with the large majority (327 per 100,000 in 2023) being unclassifiable. The prevalence of “classifiable ILD” increased from 17 to 137 patients per 100,000 individuals during the same period (Fig. 2). Prevalence generally increased over time across all classifiable ILD subtypes (Fig. 3), and increased from approximately 7 to 46 patients per 100,000 for IPF. Overall ILD prevalence varied by age group, ranging from 115 per 100,000 among individuals under 40 years of age to 1702 per 100,000 among those ≥ 80 years of age in 2023 (Fig. 4). ILD was more prevalent among men across the entire study period, but increased steadily among both men and women (Supplemental Fig. 1).

Fig. 2.

Overall Prevalence of ILD

Fig. 3.

Prevalence by ILD subtype

Fig. 4.

Prevalence by age group

Antifibrotic treatment use

To further characterize the large subgroup whose ILD subtype could not be classified based on diagnosis codes, we examined the frequency of pirfenidone and nintedanib use among prevalent patients with IPF and those who were unclassifiable, on the basis that antifibrotic treatment was only approved and reimbursed for IPF (i.e. not for progressive pulmonary fibrosis [PPF] of other causes) in Norway during the study period. Antifibrotic treatment use increased steadily from 2012 to 2023 among those with IPF, but remained minimal over time among patients who were unclassifiable in our study (Fig. 5).

Fig. 5.

Frequency of antifibrotic use among prevalent patients with IPF and unclassifiable ILD. Represents any use of nintedanib and/or pirfenidone during the given year

Discussion

In this analysis of an administrative dataset inclusive of essentially all specialist healthcare encounters within the Norwegian health system, the incidence and especially prevalence of ILD have increased substantially in the past decade. Prevalence was consistently higher among older individuals. While ILD subtype could not be classified in a majority of patients due to unspecific diagnosis coding, IPF and CTD-ILD were consistently the most commonly diagnosed subtypes.

Among ILD subtypes, the epidemiology of IPF is the best characterized in prior studies, allowing for several illustrative comparisons. Estimated incidence of IPF in our Norwegian cohort was similar to other reports in general, especially other large European cohorts based on administrative databases [9, 16] and a recent meta-analysis [8]. Prevalence of IPF was similar to that reported in Finland [17] and Sweden [18], both of which are based on similar national healthcare systems, but higher than both a recent meta-analysis, which likely incorporated more selective cohorts [8], and a recent population-based study from Tuscany [19]. Hyldgaard et al. have previously reported a lower incidence of IPF in a study from a single tertiary referral center, for which all diagnoses were clinically reviewed [20]. By comparison, our dataset did not allow for clinician confirmation of IPF cases, but was broadly inclusive and not limited to tertiary care centers.

By comparison to IPF, the epidemiology of non-IPF ILD subtypes is more sparsely reported and heterogeneous. As a recent systematic review has highlighted, reported prevalence of CTD-ILD subtypes varies widely, and has often been estimated using smaller, more highly selected cohorts [10]. Several authors have pointed out the geographic variability in prevalence estimates, among international regions [21] and even among individual U.S. states [22]. As such, our estimates from a complete nationwide population-based cohort represent a meaningful contribution to the understanding of ILD epidemiology.

The observed increase in ILD incidence in our study may be related to several causes in addition to a potential “true” increase related to risk factors or other aspects of disease biology, including increased awareness and perceived benefit of diagnosing ILDs given the advent of disease-modifying therapies. A generally recognized increase in use of computed tomography (CT) imaging [23, 24] over time is also likely to play a role, although published data on CT utilization are not available specifically for Norway. Similarly, the use of multidisciplinary meetings has anecdotally become more standardized and widespread across Norwegian centers during the study period. While our study cannot differentiate among these potential drivers, the increase in IPF incidence following the availability of antifibrotic therapies suggests that perceived therapeutic consequence may have played a role. Given that incidence is higher among older adults, increased longevity in the general population is also a likely contributor.

The large proportion of patients for whom ILD subtype was unclassifiable represents the primary limitation of this analysis and highlights the inherent challenge in defining disease epidemiology based on administrative data sources. While 10–20% of patients with ILD may be truly unclassifiable following appropriate clinical evaluation [20, 25, 26], unspecific ICD coding practices almost certainly contributed to the larger unclassifiable subgroup, both in our study and in previous studies based on large administrative datasets [14]. In other words, many patients who were unclassifiable in this study are likely to have a definable ILD subtype given access to complete clinical data; this may have led to underestimating the prevalence of specific ILD subtypes. On the other hand, the specificity and positive predictive value of our overall inclusion criteria for clinically confirmed ILD are unknown, but in any case less than 100%; as such, the overall ILD prevalence in our study is likely overestimated to some degree, and “false positive” patients who do not in fact have ILD may be overrepresented among unclassifiable cases. It is reasonable to speculate that the true overall prevalence of ILD in Norway is higher than the “classifiable ILD” prevalence but lower than the overall prevalence in our study, such that these estimates serve to suggest broad upper and lower bounds for consideration.

Within the Norwegian health system, reimbursement for outpatient physician services does not depend on the specificity of the diagnosis code used, perhaps leading to less incentive for specific coding compared with other geographic regions; by contrast, antifibrotic therapies were approved and reimbursed only for IPF in Norway during the study period. Accordingly, our data demonstrate that most patients with treated IPF were coded as such. The extent to which patients clinically diagnosed with IPF but not treated with antifibrotics are being coded less specifically, is unknown but likely to be higher. Nintedanib has subsequently been approved for PPF in Norway, a development that will undoubtedly be reflected in increased use across a broader range of ILD subtypes, but also potentially influence diagnosis coding practices, given the documentation required for reimbursement.

Regardless of underlying pathology behind the large group of patients with J84 codes not possible to sub-classify, the large number represents a reminder of the need to diagnose and treat these patients as specifically as possible. Here, our data may suggest opportunities to improve and standardize ILD care in Norway, such as the relatively low frequency of bronchoscopy early in the disease course among patients with HP or sarcoidosis, where such procedures have an established role. In any case, it is likely that many unclassified patients in our cohort in fact have either IPF or progressive pulmonary fibrosis of other causes. Thus, they represent a large cohort of patients for whom systematic follow-up and perhaps specific treatments are warranted as their disease progresses.

The lack of clinical and radiological data to confirm the presence and subtype of ILD represents an additional and related limitation of this analysis. Some previous studies based on administrative data have imposed additional constraints, such as requiring an ILD diagnosis to recur across multiple episodes and/or persist following chest CT imaging, to improve the specificity of an ICD-based diagnosis [13]. We were unable to incorporate CT imaging in our study as these data were not available. However, the observed age and sex distributions and frequency of immunosuppressive and antifibrotic treatments aligned with clinically based expectations for each classifiable ILD subtype. These observations, in the context of a complete nationwide cohort not limited by payor factors, support these data as a meaningful contribution to our understanding of ILD epidemiology.

Conclusion

In a complete nationwide cohort, incidence and prevalence of ILD in Norway are comparable to other reported regions. The prevalence of ILD is increasing substantially in Norway, both overall and across identifiable subtypes.

Abbreviations

ATC

Anatomical Therapeutic Chemical

CT

Computed Tomography

CTD-ILD

Connective Tissue Disease-associated interstitial lung disease

HP

Hypersensitivity Pneumonitis

ICD

International Classification of Diseases

ICPC

International Classification of Primary Care

ILD

Interstitial Lung Disease

IPF

Idiopathic Pulmonary Fibrosis

NPR

Norwegian Patient Registry

Authors’ contributions

MTD, PPD, SSL and TMLE designed the study and interpreted the data. SSL performed the analysis. MTD drafted the manuscript. All authors provided critical feedback and reviewed the manuscript.

Funding

Open access funding provided by University of Oslo (incl Oslo University Hospital). The study was funded by a grant from the Norwegian Respiratory Society (PP Diep) and an unrestricted grant from Boehringer Ingelheim (MT Durheim). Neither funding organization played any role in the data acquisition, analysis, interpretation, manuscript preparation or decision to publish.

Data availability

Patient-level data from the NPR and Norwegian Prescribed Drug Registry cannot be made publicly available due to restrictions included in the data use agreements for this study and the approved waiver of informed consent. The data are available upon request and with appropriate permissions according to the following instructions: https://helsedata.no/en/apply-for-data/

Declarations

Ethics approval and consent to participate

The study was approved by the Regional Ethics Committee for Southeastern Norway (Case #519297), with a waiver of informed consent. The study was performed in accordance with the Declaration of Helsinki.

Competing interests

MTD reports research funding (to his institution) from Boehringer Ingelheim and Roche unrelated to the current study, participating in advisory boards for Boehringer Ingelheim, and a consultancy relationship with Calluna Pharma. PPD has received speaker fees from Boehringer Ingelheim, attended advisory boards for NordicInfu Care AB and Boehringer Ingelheim, and received independent grants from the Norwegian Respiratory Society. Her institution has received an unrestricted grant from Boehringer Ingelheim. TMLE reports research funding (to his institution) from GSK unrelated to the current study, reimbursement for travel and lectures from AstraZeneca and Boehringer Ingelheim, and advisory board fees from Boehringer Ingelheim. SSL has no competing interests to declare.