Methods

General

A systematic literature search was carried out to identify published research on costs in Europe for each lung disease. For the most part only papers in English were included with occasional exceptions where English abstracts or translations to English were available. The literature identified for each medical condition was reviewed for data on estimates of disease-related costs per patient. Estimates were combined with prevalence or incidence rates as given in chapter 1 and elsewhere in this book, together with population data, to obtain costs per European Union (EU) country. There is a considerable volume of non-diagnosed respiratory disease, which is, by definition, difficult to cost and define, and so has been excluded from the calculations.

For COPD, asthma, OSAS, and cystic fibrosis, cost estimates have been combined with prevalence data to estimate national costs. The costs of lung cancer were estimated using an incidence approach due to the availability of incidence rather than prevalence data, the usually short survival of patients making this approach appropriate. Likewise for TB, incidence data were used; since TB treatment and other related costs are likely to occur within a given year in most cases, estimation from incidence data is justifiable. For pneumonia, calculations were based on annual hospital bed-days and an estimate of total pneumonia cases in the EU.

We estimated the economic burden of each disease for each of the EU 28 countries, although cost estimates were available from only a small set of countries, and we used adjusted cost data from some comparable non-EU European countries (Norway and Switzerland) where necessary. The costs of labour, hospital inpatient care, outpatient care and drugs vary between countries and we therefore used World Bank data on gross domestic product (GDP) per capita to convert costs, as GDP may be considered a major indicator of costs involving labour. The estimated costs per patient for each disease were adjusted proportionally to each country’s relative income level and then aggregated into total costs for each country using prevalence or incidence and national population data. No, or insufficient, data were available for sensible estimation of the costs of several diseases including bronchiectasis, pulmonary fibrosis, pulmonary vascular diseases and occupational lung diseases. Inevitably, therefore, the overall economic burden of respiratory disease is seriously underestimated.

The direct costs of healthcare relate to inpatient and outpatient costs of both hospital and primary care, together with the costs of drugs (including oxygen). Indirect costs include costs of lost production due to absence from work and early retirement and are valued according to the average daily salary including social benefit payments, using the human capital approach. All costs are calculated in year 2011 euros, expressed in sums equivalent to the purchasing power of Belgium, using purchasing power parities and national inflation rates.

In addition to these direct and indirect costs, respiratory disease is associated with a considerable cost due to disability and loss of life-years. To put a monetised value on this disability and premature mortality, a ‘willingness to pay’ methodological approach has been used, based on the value of a statistical life. Society is willing to pay a considerable amount to save life, as shown, for example, by expenditure on road and other safety precautions, healthcare and rescue costs. The values of disability and life-years lost are based on surveys and observation of the trade-offs which society is prepared to make between risk and monetary gain. A European Commission research study has collated such estimates and reported a typical range of €50 000–100 000 for the value of a life-year, with a median value of €52 000 in 2009 – equivalent to €55 000 in 2011 values. This estimate is applied here to the projected disability-adjusted life-years (DALYs) lost due to respiratory causes. Data on DALYs lost are available from WHO World Health Statistics 2011 and from the Global Burden of Disease study.

Chronic obstructive pulmonary disease

It has been reported that only 21–25% of persons identified at screening as having COPD, already had a prior diagnosis of COPD. Undiagnosed individuals with COPD may have indirect costs related to morbidity, but, since the large majority of these have mild disease, we have assumed that those without a diagnosis have no treatment or indirect costs attributable specifically to COPD. The lack of data on this point may, however, again result in significant underestimation of costs.

Cost estimates for COPD from seven European countries were identified in the literature, some of which gave direct cost estimates by disease severity, which is important as disability and costs vary widely according to severity of the disease. Cost estimates were modelled using linear regression analysis, adjusting for severity and for the setting from which the patients were recruited (see online methods section). Indirect costs, by degree of severity, were obtained from two studies, and combined with prevalence by severity and population size to give costs per country. The grading systems of severity used in the studies included the Global Initiative for Chronic Obstructive Lung Disase (GOLD) definition of severity of  COPD (I: forced expiratory volume in 1 second (FEV1) ≥80% of predicted; II: 50% ≤ FEV1 < 80%; III: 30% ≤ FEV1 < 50%; IV: FEV1 < 30% of predicted) and those of the French Pneumology Society (SPLF), the American Thoracic Society (ATS) and the Spanish Society of Pneumology and Chest Surgery (SEPAR). All the grading systems use the criterion FEV1/FVC <70%. We used the GOLD grading system, and reclassified SPLF, ATS and SEPAR grades to the nearest GOLD equivalent. Studies which presented national aggregated costs were used as such with no prevalence calculation required.

Asthma

The severity of asthma was graded according to the Global Initiative for Asthma (GINA) classification. As for COPD, estimates of direct costs presented by disease severity were modelled using linear regression analysis, adjusted for severity and the setting from which the patients were recruited (see online methods section). Indirect costs by severity of disease were obtained from two studies from Sweden and Germany (Jansson et al.; Schramm et al.). The studies used fairly similar standards of severity, based on the GINA guidelines of 1995 to 2003. The direct healthcare costs of a child with asthma were determined from the costs for an adult, by adjusting for the relation of costs between age-groups, using the demographic population structure of the European Union based on Eurostat data. Children incur indirect costs if, for example, a parent needs to be absent from work in order to care for the child. Costs by severity were combined with prevalence by severity; cost estimates representative of asthma patients overall, and not presented by degree of severity, were used together with overall prevalence to determine national costs.

Lung cancer

Two studies were identified giving estimates of direct costs, including costs of surgery, inpatient and outpatient care, chemotherapy and other drugs from onset of the disease to death. These were assumed to be representative of the mean and variation in direct costs between the EU countries. Indirect costs were estimated from Organisation for Economic Co-operation and Development (OECD) Health Data for Germany, adjusted by national GDP and extrapolated to each of the 28 countries.

Tuberculosis

Estimates of costs related to TB are based on the recent review by Diel et al. for 27 EU countries (excluding Croatia), quoted at 2012 values; this analysis includes estimation of the direct healthcare costs associated with multidrug-resistant (MDR) and extensively drug-resistant (XDR) TB.

Pneumonia/ALRIs

Cost estimates for hospital admissions for pneumonia were used together with data on the number of hospital bed days to make a best estimate of inpatient costs. However, this total seriously underestimates the total costs of pneumonia/ALRI due to the lack of usable data on patients treated as outpatients (including in primary care) as well as the indirect costs of pneumonia and the costs of other acute respiratory infections.

Obstructive sleep apnoea syndrome

A registry-based study of patients diagnosed with OSAS identified direct costs for outpatient and inpatient care and drugs, and indirect costs due to absence from work.

Cystic fibrosis

Three studies presented estimates of the total direct costs of cystic fibrosis. Costs were shown to increase with age. Prevalence data were similar in the three countries, so the average cost per patient was used. Again, no estimates of indirect costs were identified. See chapter 14 for further discussion of the costs of cystic fibrosis treatment.

  Drug cost Outpatient cost Inpatient cost Total direct# Total indirect
COPD 7.1 8.9 7.3 23.3 25.1
Asthma 8.0 6.7 4.8 19.5 14.4
Table 1 – Aggregated annual costs of inpatient stay, drugs and outpatient care including primary care attributable to chronic obstructive pulmonary disease (COPD) and asthma in the EU (billions of euro at 2011 values). #: Total of costs for drugs, outpatient including primary care, and inpatient care; : costs for absence from work and early retirement (adults).

See the entire Economic Burden of Lung Disease Chapter