Economic burden of bronchiectasis in Germany

Discussion

The objective of this study was to provide the first comprehensive estimate of the economic burden of incident bronchiectasis in any European country. This was accomplished by a retrospective observational design comparing medical and productivity-related expenditure for 231 bronchiectasis patients to a 1:3 matched comparison group of 685 control patients without bronchiectasis over a period of 3 years in Germany. The economic burden attributed to bronchiectasis is expected to be great, because it is a chronic disease that may require frequent medical consultations, long-term treatment with multimodal regimens and hospitalisations for pulmonary exacerbations in order to minimise the risk of further progression. However, to date, published annual cost data for the treatment of bronchiectasis are sparse.

The studies of Sanchez-Munoz et al. [17] and de la Rosa et al. [18], which retrospectively determined the hospital costs of bronchiectasis in Spain (cost year 2013), reported mean costs of EUR4672±6281 per patient [17] or EUR3515 for patients with bronchiectasis as a primary diagnosis and EUR4559 for patients with a secondary diagnosis [18]. Three North American studies calculated the annual costs of prevalent bronchiectasis patients ranging from USD13 244 (cost year 2001) [19] to over USD37 030 (in patients with exacerbation, 2008–2011) [20] and up to USD67 764 (in patients with P. aeruginosa, 2007–2013) [21].

With respect to incident bronchiectasis patients, only the study of Joish et al. [20] examined the increase in resource use and costs for patients with bronchiectasis before and after the first year of diagnosis. US MarketScan Research data were analysed for the period from 2005 to 2009 for patients diagnosed with bronchiectasis (n=9146) and controls without bronchiectasis (n=27 438) who were matched to each bronchiectasis patient based on age, sex, geographical region and type of health plan enrolled in a 3:1 ratio. This resulted in an increase of USD2319 per bronchiectasis patient in the first year after diagnosis, but of only USD1607 for control patients. However, in contrast to our study, where COPD as one of the most important associated conditions of bronchiectasis was explicitly allowed in both bronchiectasis cases and matched control patients, individuals with COPD were excluded if the claim was either 12 months prior (baseline) or during the subsequent 12-month follow-up period.

To the best of our knowledge, our pilot study represents the first investigation of healthcare resource use and costs brought about by incident bronchiectasis cases worldwide who were matched according to type and severity of their comorbidities, and for which accumulated post-diagnosis costs have been followed for several years. Furthermore, it includes (for the first time) both mortality and costs due to lost patient productivity.

When considering only respiratory-related resource utilisation, bronchiectasis outpatients sought specialised respiratory care 80% more often and were referred to radiologists 30% more often (table 4). Based on records for the 25 most prescribed drugs for both groups, we found no notable difference between the two groups in prescription patterns for antibiotics addressing exacerbations (fluoroquinolones, aminopenicillins and macrolides) or for bronchodilators (predominantly salbutamol and long-acting β₂-agonists) (table 5). In total, however, bronchiectasis patients received prescriptions for bronchodilators and antibiotics 27% and 29% more frequently, respectively, than was the case with matched controls (all p<0.001).

Beyond the expected economic burden on the healthcare system, mortality in our bronchiectasis group after 3 years of follow-up was surprisingly high at 26.4%. Although in Cox's regression the presence of COPD had no significant influence on mortality when comparing the bronchiectasis group and the matched control group in which a considerable proportion of patients (35.9%) also had COPD, mortality in the bronchiectasis group was significantly higher in patients with COPD than in those without COPD. This is in line with recently published evidence that mortality may be higher among those who also have COPD than in patients with bronchiectasis alone [22]. Because mortality in the bronchiectasis group was highest at 8.23% (95% CI 5.33–12.49%) within the first year following diagnosis, immediate and appropriate medical treatment should be offered as complications may arise.

Apart from the contribution of COPD, we were not able to investigate the reasons for the high mortality found among bronchiectasis patients. It may, however, be speculated that high-resolution computed tomography, the radiological investigation of choice (regular chest radiography being generally insensitive to the changes caused by bronchiectasis), was performed too late. Thus, a considerable fraction of bronchiectasis patients may have been chronically infected with bacterial pathogens well before the diagnosis of bronchiectasis was made. Unfortunately, we could not evaluate P. aeruginosa infection in our study sample as ICD-10 coding of bacterial pathogens is voluntary in Germany and therefore not routinely performed by physicians in the outpatient setting where >90% of bronchiectasis patients are cared for [9]. According to data from the representative German Bronchiectasis Registry PROGNOSIS (The PROspective German NOn-CF bronchiectaSIS patient registry), which is part of the European Bronchiectasis Initiative EMBARC (www.bronchiectasis.eu) and which currently has recruited more than 1000 patients, P. aeruginosa is the most prevalent pathogen during stable disease and during pulmonary exacerbations, with detection rates of 30% and 36%, respectively [23]. A high proportion of bronchiectasis patients (40.3% (93 out of 231)) were hospitalised at least once in the first year following diagnosis. This observation suggests that the sicker patients whose disease was already fairly advanced at diagnosis may have started treatment with medication that proved at least partly ineffective with regard to the prevention of severe exacerbations requiring hospitalisations. This may have contributed to the poor prognosis of these patients within the follow-up period.

We also took into account reduced productivity caused by ongoing symptoms and exacerbations of bronchiectasis, which amounted to EUR1916.43 per bronchiectasis patient in the first year of follow-up and decreased slightly to EUR1273.61 in the third year. As the total cost of EUR4230.39 per bronchiectasis patient due to absenteeism from work that occurred within the 3 years of observation accounted for 22.7% of the burden caused by direct total costs, indirect costs may be considered to have a substantial economic impact on healthcare. Because short-term disability with sick leave of <3 days is not captured by the statutory company health insurance, these costs are most certainly underestimated in our calculations. The importance of indirect costs in bronchiectasis patients is not diminished by the fact that these did not differ significantly from the indirect costs in the matched control group, which was characterised not only by an equally high percentage of COPD (35.9%), but also by a comparable prevalence of cardiac comorbidities and asthma.

When extrapolating our age- and sex-adjusted incidence rate for bronchiectasis of 6.1 (95% CI 6.0–6.3) per 100 000 insured German inhabitants and keeping this incidence rate constant, on average a total of 5045 (range 4962–5210) newly diagnosed bronchiectasis patients per year can be expected in the 2017 German population of 82.7 million [24], whereby males aged >65 years, probably due to a high rate of concomitant COPD, are the most strongly represented bronchiectasis subpopulation. The annual direct and indirect expenditures attributable to the disease can be expected to amount to more than EUR38.45 million per year (range EUR37.82–39.71 million per year).

Our analysis was limited by several factors. First, we only followed bronchiectasis patients for 3 years after the quarter when the diagnosis was established and thus did not capture the long-term burden of the disease. Accordingly, future studies of the long-term costs of bronchiectasis that also take the burden arising from prevalent cases into account are warranted. Second, other costly comorbid conditions that were not accounted for via the CCI in the matching process may have influenced the cost burden estimates. Third, beyond determining the difference of costs with respect to the presence or absence of bronchiectasis, matching on the basis of an equivalent CCI score level does not allow the costs of COPD or other diseases included in the CCI to be determined separately.

Fourth, although the number of insured persons from which bronchiectasis patients and their matched controls were taken was large at more than 4.8 million, our patients were not formally selected as part of a representative sample. Furthermore, our higher proportion of males with bronchiectasis, a disease considered more common in females than in males in Western countries [25, 26], may reflect the generally higher proportion of insured males in all age groups of our study population, which in 2012 included in total 614 744 more males than females in German company health insurance funds [27]. Thus, it is not certain whether our results may be generalised to the entire German population. In addition, patients were only included in our analysis if they had no diagnosis of bronchiectasis within the 3 years prior to the coding date in 2012 and if they were continuously eligible for comprehensive health benefits in the 3 years following the diagnosis. Consequently, the figure of 6.1 newly diagnosed bronchiectasis patients per 100 000 insured persons in our sample is lower than those estimated in the recently published studies of Quint et al. [10] and Weycker et al. [11], with 21.2 and 29 incident cases per 100 000 person-years in the UK and USA, respectively.

Nevertheless, our results not only demonstrate that per-patient spending on bronchiectasis is high, but also suggest that efforts at providing earlier and more effective treatment are warranted with a view towards lowering both mortality and costs. Indeed, although the present findings have to be seen in the context of the German healthcare system, we expect that studies subsequently performed in other Western countries would arrive at similar economic conclusions. However, although European guidelines have recently been published [25], they have not yet been implemented in Germany. National guidelines are desperately needed given the drastic risk of refund claims for costly off-label prescriptions from statutory health insurance companies, as indicated by the infrequent use of inhaled antibiotics in our “real-life” study sample.