Advertisement

Sarcoidosis diagnosis and treatment in Sweden: A register-based assessment of variations by region and calendar period

  • Marios Rossides
    Correspondence
    Corresponding author. Karolinska Institutet, Department of Medicine Solna, Clinical Epidemiology Division, T2, Karolinska University Hospital, 171 76, Stockholm, Sweden.
    Affiliations
    Clinical Epidemiology Division, Department of Medicine Solna, Karolinska Institutet, Stockholm, Sweden
    Search for articles by this author
  • Susanna Kullberg
    Affiliations
    Respiratory Medicine Division, Department of Medicine Solna, Karolinska Institutet, Stockholm, Sweden

    Center for Molecular Medicine, Karolinska Institutet and Karolinska University Hospital, Stockholm, Sweden

    Respiratory Medicine, Theme Inflammation and Infection, Karolinska University Hospital, Stockholm, Sweden
    Search for articles by this author
  • Anders Eklund
    Affiliations
    Respiratory Medicine Division, Department of Medicine Solna, Karolinska Institutet, Stockholm, Sweden

    Center for Molecular Medicine, Karolinska Institutet and Karolinska University Hospital, Stockholm, Sweden

    Respiratory Medicine, Theme Inflammation and Infection, Karolinska University Hospital, Stockholm, Sweden
    Search for articles by this author
  • Johan Grunewald
    Affiliations
    Respiratory Medicine Division, Department of Medicine Solna, Karolinska Institutet, Stockholm, Sweden

    Center for Molecular Medicine, Karolinska Institutet and Karolinska University Hospital, Stockholm, Sweden

    Respiratory Medicine, Theme Inflammation and Infection, Karolinska University Hospital, Stockholm, Sweden
    Search for articles by this author
  • Elizabeth V. Arkema
    Affiliations
    Clinical Epidemiology Division, Department of Medicine Solna, Karolinska Institutet, Stockholm, Sweden
    Search for articles by this author
Open ArchivePublished:November 25, 2019DOI:https://doi.org/10.1016/j.rmed.2019.105846

      Highlights

      • Increase in healthcare use indicates sarcoidosis diagnosis takes about 6 months.
      • Sarcoidosis may run undetected for more than 2 years in some patients.
      • There are differences in sarcoidosis diagnosis by Swedish healthcare region.
      • More patients are treated in the South and Southeast than in Stockholm and the North indicating phenotypic differences or lack of treatment standardization across regions.

      Abstract

      Background

      In Sweden, sarcoidosis prevalence varies geographically, but it is unclear whether diagnosis and treatment patterns vary by geographical area and calendar period. We sought to investigate differences in sarcoidosis diagnosis and treatment by healthcare region and calendar period using nationwide register data.

      Methods

      We included 4777 adults who had at least two ICD-coded visits for sarcoidosis in the National Patient Register (2007–2012). We compared patterns of healthcare use (visits and medication dispensations), and data on sarcoidosis diagnosis and treatment spanning two years before to two years after diagnosis stratified by healthcare region and calendar period at diagnosis.

      Results

      Compared to other regions, individuals diagnosed in Stockholm were younger, more likely female, and had a higher education level. In all regions, there was an increase in healthcare use at least six months before sarcoidosis diagnosis with small variation among regions. Most patients were diagnosed in pulmonary and internal medicine outpatient clinics, but compared to the national average more patients were diagnosed in rheumatology in the West and ophthalmology and cardiology in the South. Corticosteroid dispensations at diagnosis varied widely by region (48% in the South/Southeast vs. 30% in Stockholm/North). Demographic factors could not explain these differences. We found no differences by calendar period.

      Conclusion

      Our findings suggest a six-month delay in sarcoidosis diagnosis irrespective of region. The observed regional variation likely reflects differences in diagnosis and treatment patterns. Stakeholders should ensure diagnosis and treatment recommendations are closely followed.

      Keywords

      1. Introduction

      Sarcoidosis, a granulomatous disease of unknown etiology, may present in multiple forms and affect several organs [
      Statement on sarcoidosis.
      ,
      • Grunewald J.
      • Grutters J.C.
      • Arkema E.V.
      • Saketkoo L.A.
      • Moller D.R.
      • Müller-Quernheim J.
      Sarcoidosis.
      ]. As a result and in the absence of a pathognomonic test, diagnosis is often a long and complex process involving invasive examinations and exclusion of alternative diagnoses [
      • Grunewald J.
      • Grutters J.C.
      • Arkema E.V.
      • Saketkoo L.A.
      • Moller D.R.
      • Müller-Quernheim J.
      Sarcoidosis.
      ,
      • Arkema E.V.
      • Eklund A.
      • Grunewald J.
      • Bruze G.
      Work ability before and after sarcoidosis diagnosis in Sweden.
      ,
      • Ma S.P.
      • Rogers S.L.
      • Hall A.J.
      • Hodgson L.
      • Brennan J.
      • Stawell R.J.
      • Lim L.L.
      Sarcoidosis-related uveitis: clinical presentation, disease course, and rates of systemic disease progression after uveitis diagnosis.
      ,
      • Gerke A.K.
      • Tang F.
      • Pendergast J.
      • Cavanaugh J.E.
      • Polgreen P.M.
      The high frequency of healthcare use in patients one year prior to a sarcoidosis diagnosis.
      ,
      • Gade Sikjær M.
      • Hilberg O.
      • Ibsen R.
      • Bock K.
      • Løkke A.
      Direct and indirect economic and health consequences related to sarcoidosis in Denmark: a national register-based study.
      ]. According to international and Swedish recommendations, pharmacologic treatment, initially systemic corticosteroids that are replaced thereafter by steroid-sparing agents, is reserved for patients with debilitating symptoms or progressive disease [
      • Selroos O.
      • Eklund A.
      ,
      • Svensk Lungmedicinsk Förening
      ,
      • Spagnolo P.
      • Rossi G.
      • Trisolini R.
      • Sverzellati N.
      • Baughman R.P.
      • Wells A.U.
      Pulmonary sarcoidosis.
      ].
      In Sweden where sarcoidosis is more common than other parts of the world, geographic variation of disease prevalence has been observed [
      • Arkema E.V.
      • Cozier Y.C.
      Epidemiology of sarcoidosis: current findings and future directions.
      ]. Individuals with sarcoidosis were more likely to live in northwestern than in southeastern areas of the country [
      • Arkema E.V.
      • Grunewald J.
      • Kullberg S.
      • Eklund A.
      • Askling J.
      Sarcoidosis incidence and prevalence: a nationwide register-based assessment in Sweden.
      ]. Similar observations have been made in other countries such as the United States and Italy [
      • Arkema E.V.
      • Cozier Y.C.
      Epidemiology of sarcoidosis: current findings and future directions.
      ]. Clustering of genetic, environmental, or social exposures in a geographical area is perceived to cause some of geographical variation in disease occurrence. However, some of the observed geographical clustering of disease could be due to differences in diagnosis and treatment. Therefore, by mapping healthcare use, diagnosis, and treatment patterns in different geographical areas, we can better understand how much of the differences in disease occurrence are due to variation in diagnosis and treatment.
      The standardization of sarcoidosis diagnosis and treatment procedures and the application of best practices is critical for achieving favorable outcomes for all sarcoidosis patients. Despite the fact that there are diagnosis and treatment recommendations for sarcoidosis since the early 2000s [
      • Selroos O.
      • Eklund A.
      ,
      • Svensk Lungmedicinsk Förening
      ], it is unknown to what extent these are followed across regions and calendar time. It is also unclear how long it takes between symptom onset and sarcoidosis diagnosis, which is critical to account for in studies of potential causes of sarcoidosis. For these reasons, we used national registers to describe healthcare use, diagnosis and treatment patterns across geographical areas in Sweden and investigate their evolvement over time.

      2. Methods

      2.1 Study setting and data sources

      In Sweden, sarcoidosis care is provided almost exclusively by a tax-funded system of public hospitals and specialized clinics. Healthcare is primarily administered at the county level, although counties collaborate for efficient allocation of resources and provision of specialized care, forming six healthcare regions.
      Information on healthcare contact, i.e. visits to specialists and dispensation of prescriptions is recorded in national registers. Records can be retrieved and linked among registers using a unique personal identification number allocated to individuals upon birth or immigration. Hospitalizations and visits to outpatient (non-primary) care are captured by the National Patient Register (NPR; nationwide coverage of inpatient and outpatient care since 1987 and 2001, respectively). Discharge diagnoses in the NPR are coded using the International Classification of Diseases (ICD) system. Data for somatic diseases are of very high quality [
      Socialstyrelsen, Kvalitet och innehåll i patientregistret
      Utskrivningar från slutenvården 1964–2007 och besök i specialiserad öppenvård (exklusive primärvårdsbesök) 1997–2007.
      ]. Since July 2005, Swedish pharmacies report all dispensed prescriptions in the Prescribed Drug Register. Over-the-counter purchases and in-hospital treatments (e.g. infusions) are not recorded.

      2.2 Study population

      Using the NPR, we identified all individuals aged 18 years and older with at least two inpatient or outpatient visits listing an ICD code for sarcoidosis (ICD-8/9 135, ICD-10 D86 including all sub-codes). Aiming to capture newly-diagnosed cases and allow for data to accumulate for a few years before and after diagnosis, we restricted inclusion to individuals diagnosed between 2007 and 2012. We regarded the date of first visit as the date of sarcoidosis diagnosis. To reduce the potential for sarcoidosis misdiagnosis, we excluded individuals with a hematopoietic or lung malignancy in the Cancer Register (ICD-7 162, 163, 200–205) six months before or after sarcoidosis diagnosis. Ethical permission for the study was granted by the Regional Ethics Review Board in Stockholm (decision no. 2014/230–31).

      2.3 Healthcare region and calendar period of diagnosis

      To investigate regional patterns in diagnosis and treatment, we identified the hospitals in the NPR where patients received their sarcoidosis diagnosis. Based on geographical location, hospitals were classified into six healthcare regions: Stockholm, Uppsala-Örebro, West, South, Southeast, and North (Fig. 1 and Table S1 in the Supplement). To study calendar period effects, we created three groups based on calendar year of sarcoidosis diagnosis: 2007–2008, 2009–2010, and 2011–2012.
      Fig. 1
      Fig. 1Map of Swedish healthcare regions (Stockholm [Stockholm and Gotland counties], Uppsala-Örebro [Uppsala, Södermanland, Värmland, Örebro, Västmanland, Dalarna, and Gävleborg], West [Västra Götaland and Halland], South [Skåne, Kronoberg, and Blekinge], Southeast [Östergötland, Jönköping, and Kalmar], and North [Västernorrland, Jämtland, Västerbotten, and Norrbotten]).

      2.4 Demographic information, healthcare use, and sarcoidosis diagnosis and treatment

      Information from national registers was obtained to describe demographic, healthcare contacts and sarcoidosis diagnosis and treatment patterns across regions and calendar periods. From the Total Population Register and the Longitudinal Integration Database for Health Insurance and Labor Market Studies, we received data on birth year (to calculate age), sex, birth country (Nordic, non-Nordic, missing), years of attained education at diagnosis (≤9, 10–12, ≥13 years, missing), and annual gross salary earned during the year before diagnosis and adjusted to 2014 inflation rates [].
      To approximate healthcare use, we calculated the proportion of individuals with at least one inpatient or outpatient visit in the NPR or at least one dispensation in the Prescribed Drug Register, as well as the median number of visits and dispensations of unique medications. This was done for each six-month interval 24 months before to 24 months after sarcoidosis diagnosis. In addition, to investigate the reasons for healthcare use over time, we identified the three most common (ICD-coded) main discharge diagnoses and the three most commonly dispensed medications before and after sarcoidosis diagnosis.
      The clinic where the sarcoidosis diagnosis was given was retrieved from the NPR and categorized into pulmonary, internal medicine, rheumatology, cardiology, ophthalmology, dermatology, and other. For 20 individuals (0.4% of all patients) who visited more than one clinic on the day of diagnosis, one clinic was chosen at random.
      We obtained data on sarcoidosis treatments including the date of dispensation, amount, and the daily defined dose (DDD) from the Prescribed Drug Register. We calculated the proportion of patients who were dispensed at least one prescription of the following sarcoidosis-related medications during a six-month-long interval as described above: (a) systemic corticosteroids (prednisolone, prednisone, and methylprednisolone [Anatomical Therapeutic Chemical classification system codes H02AB04/6/7]), (b) second line treatments (methotrexate, leflunomide, azathioprine, and hydroxychloroquine [L01BA01, L04AX01/3, L04AA13, P01BA02]), (c) inhaled corticosteroids (R03BA), and (d) prescribed non-steroidal anti-inflammatory drugs (M01A). For systemic corticosteroids, we also calculated the median dispensed DDDs per individual during the six-month interval. We did not investigate third line (biologic) treatments because they are not captured by the Prescribed Drug Register. They are nevertheless infrequently used.

      2.5 Statistical analysis

      All analyses were stratified by healthcare region and calendar period of diagnosis. We used descriptive statistics, i.e. proportions, means with standard deviations, and medians with interquartile ranges to compare data among regions and calendar periods. We plotted patterns over time (before and after sarcoidosis diagnosis) for healthcare use and sarcoidosis medication dispensation by region and calendar period. We compared differences by region and calendar period in each six-month interval using chi-square or Kruskal-Wallis tests as appropriate reporting two-sided p-values. To examine whether regional variation could be explained by demographic differences across regions, we used logistic regression models (multivariate or simple depending on the outcome) adjusting for age at diagnosis, sex, birth country, and education to estimate odds ratios and corresponding 95% confidence intervals.
      Data management and analysis were performed using SAS software version 9.4 (SAS Institute Inc., Cary, NC, USA) and R version 3.6.1 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria).

      3. Results

      We included 4777 individuals with at least two visits for sarcoidosis in the NPR from 2007 to 2012. Approximately 20% of patients were diagnosed in clinics in each of Stockholm, Uppsala-Örebro, the West, and South healthcare regions and 11% in each of the Southeast and North regions (Table 1). Individuals diagnosed in Stockholm were more likely to be younger, female, to be born outside a Nordic country, have higher education and salaries compared to other regions. No demographic differences were identified among the three calendar periods (Table 1).
      Table 1Demographic characteristics of individuals with sarcoidosis at diagnosis stratified by healthcare region and calendar period of diagnosis.
      Healthcare regionYear of diagnosisOverall
      StockholmUppsala-ÖrebroWestSouthSoutheastNorth2007–20082009–20102011–2012
      Patients, n (% of overall)987 (21)1043 (22)860 (18)818 (17)539 (11)530 (11)1617 (34)1739 (36)1421 (30)4777
      Mean age (SD), years48 (14.5)51 (15.2)51 (15.1)50 (15.2)52 (15.7)51 (15.5)50 (15.1)51 (15.4)51 (15.0)50 (15.2)
      Female, %47.044.144.244.444.044.044.846.142.944.7
      Country of birth*, %
      Nordic83.194.088.488.190.095.390.489.288.589.4
      Non-Nordic16.45.711.411.99.84.39.210.611.210.3
      Missing0.50.40.2<0.10.20.40.40.20.30.3
      Attained education, %
      ≤9 years14.821.419.721.923.719.820.020.019.719.9
      10–12 years43.953.850.346.354.952.650.350.448.649.8
      ≥13 years39.523.828.330.921.026.028.328.530.529.0
      Missing1.81.11.70.90.41.51.41.21.21.3
      Annual gross salary†
      Median (IQR), 1000 SEK295 (181, 414)284 (165, 361)268 (139, 365)262 (156, 348)266 (144, 349)252 (115, 350)266 (147, 358)278 (160, 360)279 (158, 374)275 (155, 364)
      Zero salary, %28.532.532.333.129.929.430.132.131.131.1
      Missing, %1.01.60.80.71.91.51.41.30.81.2
      SD = standard deviation; IQR = interquartile range; SEK = Swedish krona. *Nordic countries include Sweden, Denmark, Norway, Finland, and Iceland. †Adjusted to 2014 inflation rates. 1.00 SEK is approximately equal to 0.10 USD, 0.09 EUR, or 0.08 GBP.

      3.1 Healthcare use and sarcoidosis diagnosis by region and calendar period

      In all regions, about nine out of 10 individuals were diagnosed in specialized outpatient care and were most likely to have their second sarcoidosis visit within about two months from the first (Table 2). Most patients received their sarcoidosis diagnosis in pulmonary or internal medicine clinics (range 67–80% across regions; average 74%). Compared to the national average, a higher proportion of patients were diagnosed in rheumatology in the West (9% vs. 5%) and in cardiology and ophthalmology in the South (6% vs. 2% and 11% vs. 5%, respectively). These findings were independent of demographic variation across regions (Table S2).
      Table 2Information on sarcoidosis diagnosis stratified by healthcare region and year of diagnosis.
      Healthcare regionYear of diagnosisOverall
      StockholmUppsala-ÖrebroWestSouthSoutheastNorth2007–20082009–20102011–2012
      Patients, n (% of overall)987 (21)1043 (22)860 (18)818 (17)539 (11)530 (11)1617 (34)1739 (36)1421 (30)4777
      Diagnosed in outpatient care, %93.889.791.791.288.189.891.790.990.491.0
      Median time from first to second visit for sarcoidosis (IQR), months1.4 (0.6, 3.7)1.8 (0.6, 4.6)2.0 (0.9, 4.4)1.9 (0.9, 4.2)2.3 (0.9, 5.8)2.0 (0.7, 5.1)2.1 (0.9, 4.7)2.0 (0.7, 5.2)1.5 (0.7, 3.4)1.9 (0.7, 4.4)
      Clinic of diagnosis, %
      Pulmonary or internal medicine77.677.867.363.778.879.874.274.172.973.8
      Pulmonary medicine58.742.753.821.930.447.043.244.342.943.5
      Internal medicine18.935.113.541.848.432.831.029.830.030.3
      Rheumatology3.02.89.05.13.35.34.74.45.14.7
      Cardiology0.70.91.06.01.11.33.11.31.11.8
      Ophthalmology2.52.67.010.63.93.44.65.64.65.0
      Dermatology4.74.44.34.53.21.73.93.84.44.0
      Other11.411.611.410.09.68.59.510.812.010.7
      An increasing trend of healthcare use was apparent years before diagnosis but a surge was identified six months before diagnosis (Fig. 2). More than 60% of patients from all regions had at least one contact with specialist care at least six months before diagnosis, an increase from 40% in the previous six-month period (Fig. 3). On average, patients in the South and Stockholm had about three visits to the hospital during the periods before and after diagnosis (excluding the diagnostic visits) compared to about two visits in the other regions. Except in Stockholm, the most common discharge diagnoses were related to sarcoidosis (e.g. uveitis or abnormal imaging of the lung) for visits 18 to 12 months before diagnosis (Table S3 in the Supplement). Closer to diagnosis, abnormal findings on lung imaging and/or enlarged lymph nodes became the most prevalent discharge diagnoses in all regions. In all regions and up to six months before diagnosis, 60% of patients were dispensed at least one prescribed medication (a median of six unique medications per patient), increasing to 80% and a median of eight different medications thereafter (Fig. S1 and Table S5).
      Fig. 2
      Fig. 2Proportion of individuals with at least one visit in the National Patient Register (solid line) and proportion of individuals with at least one medication dispensation in the Prescribed Drug Register (dotted line) evaluated in monthly intervals before sarcoidosis diagnosis (time “0”).
      Fig. 3
      Fig. 3Proportion of individuals with at least one visit in the National Patient Register excluding first and second visits for sarcoidosis (upper panel) and proportion of individuals with at least one medication dispensation in the Prescribed Drug Register (lower panel) before and after sarcoidosis diagnosis (“Dx”) by healthcare region. P-value from a chi-square test for the comparison among healthcare regions.
      We did not identify any considerable variation by calendar period in the type of care, clinic of sarcoidosis diagnosis, history of discharge diagnoses for visits before sarcoidosis, or receipt of healthcare before or after sarcoidosis diagnosis (Table 1, Table 2, S4, S6, and Figs. S2 and S3).

      3.2 Sarcoidosis treatment by region and calendar period

      In all regions, dispensation of systemic and inhaled corticosteroids as well as non-steroidal anti-inflammatory drugs increased markedly starting six months before diagnosis (Fig. 4, Fig. 5). For systemic corticosteroids, there was considerable variation by region despite taking into consideration demographic variation across regions (Table S2). In Stockholm and the North, about 30% dispensed at least one prescription of systemic corticosteroids during the first six months after diagnosis. Almost half of the patients in South and Southeast and 40% in Uppsala-Örebro and the West dispensed at least one systemic corticosteroid prescription during the same period. No marked differences in the dispensed daily defined doses were identified (approximately 1.5 daily defined doses/day/individual). The proportion of individuals who were dispensed systemic corticosteroids decreased gradually thereafter with less than 20% of the patients in Stockholm and the North and 30% in the other regions dispensed a corticosteroid two years after diagnosis.
      Fig. 4
      Fig. 4Dispensation of systemic corticosteroids in the Prescribed Drug Register by healthcare region. Upper panel: proportion of individuals with at least one dispensation of systemic corticosteroids; lower panel: median daily defined doses of systemic corticosteroids per individual dispensed during the six-month interval. P-value from a chi-square or Kruskal-Wallis test as appropriate for the comparison among healthcare regions.
      Fig. 5
      Fig. 5Dispensation of sarcoidosis-related treatments in the Prescribed Drug Register by healthcare region. Upper panel: any second line treatment (methotrexate, leflunomide, azathioprine, or hydroxychloroquine); middle panel: inhaled corticosteroids; lower panel: non-steroidal anti-inflammatory drugs (NSAIDs). P-value from a chi-square test for the comparison among healthcare regions.
      Less than 10% of individuals were dispensed a second line (steroid-sparing) treatment for sarcoidosis after diagnosis with small variation among regions (Fig. 5). Differences among regions were more noticeable for inhaled corticosteroids and non-steroidal anti-inflammatory drugs. The dispensation of systemic corticosteroids and other sarcoidosis-related retreatments did not vary by calendar period (Figs. S4 and S5).

      4. Discussion

      In this large register-based study, we identified variation in the diagnosis and treatment of individuals with sarcoidosis by geographical area in Sweden. No variation by calendar time (2007–2012) was found. For most individuals, there was an increase in healthcare use indicating that at least six months were required for diagnosis to be established although a diagnostic delay of at least two years was identified in some cases with small differences across regions. The proportion of individuals treated with systemic corticosteroids around the time of diagnosis was higher in South and Southeast compared to Stockholm and the North.
      There is distinct demographic variation across Sweden. Compared to the rest of the country, patients diagnosed with sarcoidosis in the capital region of Stockholm are slightly younger, more likely to be female, to be born outside a Nordic country, better educated, and having a higher salary. The variation in demographics observed across regions is not unique to the sarcoidosis population, but likely reflects the underlying differences in the general population []. Despite the fact that access to healthcare is universal in Sweden, social and geographical disparities are well documented [
      • jämförelser Öppna
      folkhälsa [Open comparisons 2014: public health], Stockholm.
      ,
      Hälso- och sjukvårdsrapporten
      Health and health care report 2018.
      ] and are thus likely to have partly driven but could certainly not explain the differences we identified.
      While nine in 10 patients in all regions received their diagnosis in specialist outpatient care and had a second visit for sarcoidosis within two months from the first, there were noticeable differences in the clinics where patients were diagnosed. We should note that a clinic may include more specialties in a small local hospital compared to a large regional hospital and hence some variation is expected. In all regions except the West and South, up to 80% of patients were diagnosed in pulmonary and internal medicine clinics. Compared to the national average, a higher proportion of patients were diagnosed in rheumatology in the West, which we expected because of the local tradition of referring individuals to either respiratory medicine or rheumatology when sarcoidosis is suspected. In the South, sarcoidosis diagnoses from ophthalmology and cardiology clinics were also slightly more common compared to the national average. These differences may represent differences in referral patterns or sarcoidosis phenotype across regions. It remains to be determined which explanation is most probable and what the outcomes are for individuals diagnosed in clinics of variable experience with disease diagnosis and treatment.
      We observed that up to 80% of individuals in all regions were in contact with a specialist at least six months before their diagnosis or had been dispensed a prescription. A steep increase in the prescription of non-steroidal anti-inflammatory drugs during the same period is indicative of deteriorating general health. These findings suggest that six months are needed on average for individuals to be diagnosed with sarcoidosis. Our data also indicated that there was some diagnostic delay. Estimating how extensive this issue is with high precision is challenging due to difficulties accurately attributing ICD-coded diagnoses to preclinical sarcoidosis. Nonetheless, a small proportion of patients (<10%) received diagnoses that could be perceived related to sarcoidosis (e.g. uveitis/iridocyclitis), some of which occurred 24 months before sarcoidosis diagnosis (the maximum lookback period in this study). Differences in the proportion of, for example, uveitis among regions may be indicative of some regional variation in terms of delay in diagnosis.
      Long-lasting preclinical sarcoidosis was described in several other studies. Comparing sick leave absence, earnings, and health expenses as indicators of general health status between sarcoidosis and non-sarcoidosis comparators, large studies from Sweden [
      • Arkema E.V.
      • Eklund A.
      • Grunewald J.
      • Bruze G.
      Work ability before and after sarcoidosis diagnosis in Sweden.
      ] and Denmark [
      • Gade Sikjær M.
      • Hilberg O.
      • Ibsen R.
      • Bock K.
      • Løkke A.
      Direct and indirect economic and health consequences related to sarcoidosis in Denmark: a national register-based study.
      ] indicate that preclinical disease might last for at least a year. Using employer health insurance data from the United States, an investigation found more outpatient visits during the year before diagnosis in sarcoidosis compared to other pulmonary diseases [
      • Gerke A.K.
      • Tang F.
      • Pendergast J.
      • Cavanaugh J.E.
      • Polgreen P.M.
      The high frequency of healthcare use in patients one year prior to a sarcoidosis diagnosis.
      ]. Uveitis has been suggested as a predictor of delayed diagnosis in sarcoidosis [
      • Ma S.P.
      • Rogers S.L.
      • Hall A.J.
      • Hodgson L.
      • Brennan J.
      • Stawell R.J.
      • Lim L.L.
      Sarcoidosis-related uveitis: clinical presentation, disease course, and rates of systemic disease progression after uveitis diagnosis.
      ]. Similar to sarcoidosis, long preclinical disease was described in other chronic inflammatory diseases. Inflammatory blood markers can be identified up to 10 years before the diagnosis of rheumatoid arthritis is established [
      • V Arkema E.
      • Goldstein B.L.
      • Robinson W.
      • Sokolove J.
      • Wagner C.A.
      • Malspeis S.
      • Rosner B.
      • Grodstein F.
      • Karlson E.W.
      • Costenbader K.H.
      Anti-citrullinated peptide autoantibodies, human leukocyte antigen shared epitope and risk of future rheumatoid arthritis: a nested case–control study.
      ] and a delay in diagnosis of up to five years was reported in idiopathic pulmonary fibrosis [
      • Mooney J.
      • Chang E.
      • Lalla D.
      • Papoyan E.
      • Raimundo K.
      • Reddy S.R.
      • Stauffer J.
      • Yan T.
      • Broder M.S.
      Potential delays in diagnosis of idiopathic pulmonary fibrosis in Medicare beneficiaries.
      ].
      In line with both local and international recommendations [
      • Selroos O.
      • Eklund A.
      ,
      • Svensk Lungmedicinsk Förening
      ,
      • Spagnolo P.
      • Rossi G.
      • Trisolini R.
      • Sverzellati N.
      • Baughman R.P.
      • Wells A.U.
      Pulmonary sarcoidosis.
      ], systemic corticosteroids are the backbone of sarcoidosis treatment in Sweden. Four in ten patients with sarcoidosis received at least one prescription of systemic corticosteroids around the time of diagnosis, whereas second line (steroid-sparing) treatments were rarely prescribed at diagnosis. These observations are consistent with reports from the United States, the United Kingdom, Spain, and Poland [
      • Baughman R.P.
      • Judson M.A.
      • Teirstein A.
      • Yeager H.
      • Rossman M.
      • Knatterud G.L.
      • Thompson B.
      Presenting characteristics as predictors of duration of treatment in sarcoidosis.
      ,
      • Gibson G.J.
      • Prescott R.J.
      • Muers M.F.
      • Middleton W.G.
      • Mitchell D.N.
      • Connolly C.K.
      • Harrison B.D.
      British Thoracic Society Sarcoidosis study: effects of long term corticosteroid treatment.
      ,
      • Judson M.A.
      • Baughman R.P.
      • Thompson B.W.
      • Teirstein A.S.
      • Terrin M.L.
      • Rossman M.D.
      • Yeager Jr., H.
      • McLennan G.
      • Bresnitz E.A.
      • DePalo L.
      • Hunninghake G.
      • Iannuzzi M.C.
      • Johns C.J.
      • Moller D.R.
      • Newman L.S.
      • Rabin D.L.
      • Rose C.
      • Rybicki B.A.
      • Weinberger S.E.
      • Knatterud G.L.
      • Cherniak R.
      • Group A.R.
      • Access Research Group
      Two year prognosis of sarcoidosis: the ACCESS experience.
      ,
      • Mañá J.
      • Rubio-Rivas M.
      • Villalba N.
      • Marcoval J.
      • Iriarte A.
      • Molina-Molina M.
      • Llatjos R.
      • García O.
      • Martínez-Yélamos S.
      • Vicens-Zygmunt V.
      • Gámez C.
      • Pujol R.
      • Corbella X.
      Multidisciplinary approach and long-term follow-up in a series of 640 consecutive patients with sarcoidosis.
      ,
      • Martusewicz-Boros M.M.
      • Boros P.W.
      • Wiatr E.
      • Fijołek J.
      • Roszkowski-Śliż K.
      Systemic treatment for sarcoidosis was needed for 16% of 1810 Caucasian patients.
      ,
      • Thillai M.
      • Chang W.
      • Chaudhuri N.
      • Forrest I.
      • Ho L.-P.
      • Lines S.
      • Maher T.M.
      • Spencer L.G.
      • Spiteri M.
      • Coker R.
      Sarcoidosis in the UK: insights from British Thoracic Society registry data.
      ].
      Although the type of treatment did not vary by healthcare region, indicating that treatment guidelines are followed to some extent, there were large regional differences in the proportion of individuals who were dispensed systemic corticosteroids. Interestingly, regional variation in treatment does not directly map the geographical variation in disease occurrence [
      • Arkema E.V.
      • Grunewald J.
      • Kullberg S.
      • Eklund A.
      • Askling J.
      Sarcoidosis incidence and prevalence: a nationwide register-based assessment in Sweden.
      ]. We hypothesize that several factors might be behind those differences. First, there were (and to this day there are) no clear-cut treatment guidelines that are based on objective disease severity markers. Second, across regions, patients might be treated by physicians of different specializations who are likely to follow treatment strategies familiar with their specialty and may favor one treatment strategy over another. Last, real phenotypic differences in sarcoidosis occurrence and progression might be behind the observed regional differences, but there are no published observations to support this notion. It is worth noting that regional differences in treatment are not unique to sarcoidosis in Sweden; they have been well-documented in rheumatoid arthritis for which detailed guidelines and disease activity scores exist [
      • Neovius M.
      • Sundström A.
      • Simard J.
      • Wettermark B.
      • Cars T.
      • Feltelius N.
      • Askling J.
      • Klareskog L.
      Small-area variations in sales of TNF inhibitors in Sweden between 2000 and 2009.
      ].
      Our study has some limitations. Despite the fact that we could obtain almost complete longitudinal information from the nationwide registers, the amount of clinical information available was limited. Because our data relied on ICD codes, we could not investigate whether patients received a biopsy or underwent certain imaging examinations, nor could we analyze the results of those diagnostic examinations. In addition, primary care data, which could have provided some insight on diagnostic examinations, referral patterns across regions, and possibly the date of symptoms onset were not available.
      Furthermore, we had no data on hospital-administered treatments (e.g. anti-tumor necrosis factor α infusions) or on over-the-counter medication purchases. Some underestimation of the dispensation of non-steroidal anti-inflammatory drugs is therefore expected. We investigated dispensation of medications, not prescription (physician's intention-to-treat) or actual use of the medications, which may have led to some misclassification of medication use. Misclassification is unlikely to have varied much by region or calendar period so as to alter the robustness of our conclusions. Last, our analysis assumed that there was no attrition due to emigration or death after sarcoidosis diagnosis so that all patients were at risk of healthcare interaction after inclusion (at the time of diagnosis). Previous analyses suggest that attrition after the first two years from inclusion is negligible [
      • Rossides M.
      • Kullberg S.
      • Askling J.
      • Eklund A.
      • Grunewald J.
      • Arkema E.V.
      Sarcoidosis mortality in Sweden: a population-based cohort study.
      ].
      The strengths of our study should also be emphasized. This is a large investigation of patterns of diagnosis and treatment that used nationwide prospectively-collected data. Our sarcoidosis population has not been selected based on ethnic background, employment, or income. Despite the fact that this study describes the Swedish setting, we believe that several of our general observations and conclusions could be extrapolated to other high-income countries, especially those with similar publicly-funded healthcare systems.
      In summary, we found that for the average individual with sarcoidosis diagnosis takes about six months to be established and receipt of treatment around diagnosis largely depends on which part of the country they are seen. We call for active collaboration between pulmonologists and other specialists as well as research directed towards early diagnosis and standardization of treatment procedures. We believe this could effectively diminish any disparities in the diagnosis and management of individuals with sarcoidosis.

      Author contributions

      Marios Rossides: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing - original draft, Writing - review & editing. Susanna Kullberg: Writing - review & editing. Anders Eklund: Writing - review & editing. Johan Grunewald: Writing - review & editing. Elizabeth V. Arkema: Funding acquisition, Supervision, Writing - review & editing.

      Funding/support statement

      This study was supported by funding from the Swedish Heart-Lung Foundation (Hjärt-Lungfonden). The register linkage used in this study was funded by grants from the Swedish Society of Medicine (Svenska Läkaresällskapet).

      Declaration of competing interest

      The authors declare the following financial interests/personal relationships which may be considered as potential competing interests. EVA received research grants from the Swedish Heart-Lung Foundation (Hjärt-Lungfonden) and the Swedish Society of Medicine (Svenska Läkaresällskapet). MR, SK, AE, and JG have no conflicts of interest.

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

      1. Statement on sarcoidosis.
        Am. J. Respir. Crit. Care Med. 1999; 160: 736-755https://doi.org/10.1164/ajrccm.160.2.ats4-99
        • Grunewald J.
        • Grutters J.C.
        • Arkema E.V.
        • Saketkoo L.A.
        • Moller D.R.
        • Müller-Quernheim J.
        Sarcoidosis.
        Nat. Rev. Dis. Prim. 2019; 5https://doi.org/10.1038/s41572-019-0096-x
        • Arkema E.V.
        • Eklund A.
        • Grunewald J.
        • Bruze G.
        Work ability before and after sarcoidosis diagnosis in Sweden.
        Respir. Med. 2018; 144: S7-S12https://doi.org/10.1016/j.rmed.2018.09.016
        • Ma S.P.
        • Rogers S.L.
        • Hall A.J.
        • Hodgson L.
        • Brennan J.
        • Stawell R.J.
        • Lim L.L.
        Sarcoidosis-related uveitis: clinical presentation, disease course, and rates of systemic disease progression after uveitis diagnosis.
        Am. J. Ophthalmol. 2019; 198: 30-36https://doi.org/10.1016/j.ajo.2018.09.013
        • Gerke A.K.
        • Tang F.
        • Pendergast J.
        • Cavanaugh J.E.
        • Polgreen P.M.
        The high frequency of healthcare use in patients one year prior to a sarcoidosis diagnosis.
        Sarcoidosis Vasc. Diffuse Lung Dis. 2014; 31: 256-261
        • Gade Sikjær M.
        • Hilberg O.
        • Ibsen R.
        • Bock K.
        • Løkke A.
        Direct and indirect economic and health consequences related to sarcoidosis in Denmark: a national register-based study.
        Respir. Med. 2019; 152: 7-13https://doi.org/10.1016/j.rmed.2019.04.004
        • Selroos O.
        • Eklund A.
        Scandinavian Sarcoidosis Association. Student litterature, Lund2005 (Sarkoidos [Sarcoidosis])
        • Svensk Lungmedicinsk Förening
        Vårdprogram För Sarkoidos [Sarcoidosis Care Recommendations]. Mediahuset AB, Göteborg2014
        • Spagnolo P.
        • Rossi G.
        • Trisolini R.
        • Sverzellati N.
        • Baughman R.P.
        • Wells A.U.
        Pulmonary sarcoidosis.
        Lancet Respir. Med. 2018; 6: 389-402https://doi.org/10.1016/S2213-2600(18)30064-X
        • Arkema E.V.
        • Cozier Y.C.
        Epidemiology of sarcoidosis: current findings and future directions.
        Ther. Adv. Chron. Dis. 2018; 9: 227-240https://doi.org/10.1177/2040622318790197
        • Arkema E.V.
        • Grunewald J.
        • Kullberg S.
        • Eklund A.
        • Askling J.
        Sarcoidosis incidence and prevalence: a nationwide register-based assessment in Sweden.
        Eur. Respir. J. 2016; 48: 1690-1699https://doi.org/10.1183/13993003.00477-2016
        • Socialstyrelsen, Kvalitet och innehåll i patientregistret
        Utskrivningar från slutenvården 1964–2007 och besök i specialiserad öppenvård (exklusive primärvårdsbesök) 1997–2007.
        ([Quality and contents of the National Patient Register], Stockholm)
        • Centralbyrån Statistiska
        Konsumentprisindex.
        (Consumer price index)
        • Centralbyrån Statistiska
        Befolkningsstatistik.
        (Population statistics)
        • jämförelser Öppna
        folkhälsa [Open comparisons 2014: public health], Stockholm.
        2014 (2014)
        • Hälso- och sjukvårdsrapporten
        Health and health care report 2018.
        2018
        • V Arkema E.
        • Goldstein B.L.
        • Robinson W.
        • Sokolove J.
        • Wagner C.A.
        • Malspeis S.
        • Rosner B.
        • Grodstein F.
        • Karlson E.W.
        • Costenbader K.H.
        Anti-citrullinated peptide autoantibodies, human leukocyte antigen shared epitope and risk of future rheumatoid arthritis: a nested case–control study.
        Arthritis Res. Ther. 2013; 15: R159https://doi.org/10.1186/ar4342
        • Mooney J.
        • Chang E.
        • Lalla D.
        • Papoyan E.
        • Raimundo K.
        • Reddy S.R.
        • Stauffer J.
        • Yan T.
        • Broder M.S.
        Potential delays in diagnosis of idiopathic pulmonary fibrosis in Medicare beneficiaries.
        Ann. Am. Thorac. Soc. 2019; 16: 393-396https://doi.org/10.1513/AnnalsATS.201806-376RL
        • Baughman R.P.
        • Judson M.A.
        • Teirstein A.
        • Yeager H.
        • Rossman M.
        • Knatterud G.L.
        • Thompson B.
        Presenting characteristics as predictors of duration of treatment in sarcoidosis.
        QJM. 2006; 99: 307-315https://doi.org/10.1093/qjmed/hcl038
        • Gibson G.J.
        • Prescott R.J.
        • Muers M.F.
        • Middleton W.G.
        • Mitchell D.N.
        • Connolly C.K.
        • Harrison B.D.
        British Thoracic Society Sarcoidosis study: effects of long term corticosteroid treatment.
        Thorax. 1996; 51: 238-247https://doi.org/10.1136/thx.51.3.238
        • Judson M.A.
        • Baughman R.P.
        • Thompson B.W.
        • Teirstein A.S.
        • Terrin M.L.
        • Rossman M.D.
        • Yeager Jr., H.
        • McLennan G.
        • Bresnitz E.A.
        • DePalo L.
        • Hunninghake G.
        • Iannuzzi M.C.
        • Johns C.J.
        • Moller D.R.
        • Newman L.S.
        • Rabin D.L.
        • Rose C.
        • Rybicki B.A.
        • Weinberger S.E.
        • Knatterud G.L.
        • Cherniak R.
        • Group A.R.
        • Access Research Group
        Two year prognosis of sarcoidosis: the ACCESS experience.
        Sarcoidosis Vasc. Diffuse Lung Dis. 2003; 20: 204-211
        • Mañá J.
        • Rubio-Rivas M.
        • Villalba N.
        • Marcoval J.
        • Iriarte A.
        • Molina-Molina M.
        • Llatjos R.
        • García O.
        • Martínez-Yélamos S.
        • Vicens-Zygmunt V.
        • Gámez C.
        • Pujol R.
        • Corbella X.
        Multidisciplinary approach and long-term follow-up in a series of 640 consecutive patients with sarcoidosis.
        Medicine (Baltim.). 2017; 96e7595https://doi.org/10.1097/MD.0000000000007595
        • Martusewicz-Boros M.M.
        • Boros P.W.
        • Wiatr E.
        • Fijołek J.
        • Roszkowski-Śliż K.
        Systemic treatment for sarcoidosis was needed for 16% of 1810 Caucasian patients.
        Clin. Res. J. 2018; 12: 1367-1371https://doi.org/10.1111/crj.12664
        • Thillai M.
        • Chang W.
        • Chaudhuri N.
        • Forrest I.
        • Ho L.-P.
        • Lines S.
        • Maher T.M.
        • Spencer L.G.
        • Spiteri M.
        • Coker R.
        Sarcoidosis in the UK: insights from British Thoracic Society registry data.
        BMJ Open Respir. Res. 2019; 6e000357https://doi.org/10.1136/bmjresp-2018-000357
        • Neovius M.
        • Sundström A.
        • Simard J.
        • Wettermark B.
        • Cars T.
        • Feltelius N.
        • Askling J.
        • Klareskog L.
        Small-area variations in sales of TNF inhibitors in Sweden between 2000 and 2009.
        Scand. J. Rheumatol. 2011; 40: 8-15https://doi.org/10.3109/03009742.2010.493895
        • Rossides M.
        • Kullberg S.
        • Askling J.
        • Eklund A.
        • Grunewald J.
        • Arkema E.V.
        Sarcoidosis mortality in Sweden: a population-based cohort study.
        Eur. Respir. J. 2018; 51 (1701815)https://doi.org/10.1183/13993003.01815-2017