Background
Haemophilia A is the most common severe inherited bleeding disorder characterized by spontaneous or traumatic bleeding owing to an inability to produce the clotting factor VIII. It is a rare genetic disease with an X-linked recessive pattern and thus affects primarily males (one in 5000–10,000 male births [
1]) while females are carriers. Haemophilia is essentially an inherited bleeding disorder but one third of patients result from a de novo gene mutation [
1]. Haemophilia A is traditionally classified as mild (6–40 IU/dL), moderate (1–5 IU/dL), or severe (< 1 IU/dL) [
2,
3].
Bleeding episodes are the hallmark of haemophilia A and their extent depends on the severity of the disease and the presence/absence of neutralizing anti-factor VIII antibodies, known as inhibitors. Recurrent muscular bleeding and chronic joint disease is common in patients treated on demand, primarily as a result of bleeding into the joints, which can lead to long-term haemophilic arthropathies [
4,
5]. These complications will lead to severe loss of function due to loss of motion, pain, deformity and activity limitation [
4], affecting patients’ quality of life, psychological well-being and thus productivity [
6,
7]. Orthopaedic complications usually require surgical interventions, such as orthopaedic surgery (total joint replacement), arthroplasty and mechanical or chemical synovectomy (synoviorthesis).
The standard treatment of haemophilia A is based on clotting factor replacement therapy on-demand (treatment for a bleeding episode control) or prophylaxis (continuous long-term treatment to prevent bleeding). Prophylactic treatment can lead to better clinical outcomes with better health-related quality of life and is changing the prognosis scenario, leading to a life expectancy close-to-normal in high- and middle- income countries [
8,
9]. However, prophylactic treatment is more costly than on-demand therapy [
10].
About 30% of patients with severe haemophilia A who are exposed to factor VIII concentrates will develop neutralizing anti-factor VIII antibodies (inhibitors), making clotting factor replacement therapy ineffective [
11]. Development of inhibitors is the most serious complication of haemophilia and results in poor bleeding control and infeasible standard prophylaxis [
12,
13], which subsequently means higher levels of morbidity and mortality and reduced health-related quality of life [
10,
14,
15]. Inhibitor development risk is higher in patients with severe haemophilia A than in mild/moderate disease and mainly occurs in children [
12]. Inhibitors have a great impact on therapeutic strategy, since patients with inhibitors require treatment with bypassing agents, namely recombinant activated factor VII (rFVIIa) or activated prothrombin complex concentrate (aPCC) [
11]. Although both agents are considered effective, the efficacy of bypassing agents is suboptimal as compared with FVIII replacement therapy [
12] and both options involve frequent intravenous infusions that depend on adequate venous access. However, the use of bypassing agents results in a significant increase to overall treatment costs [
16,
17]. Immune tolerance induction (ITI) is the only strategy to eradicate high titre inhibitors (≥5 Bethesda units) and consists in the use of high doses of the clotting factor that triggered the development of inhibitors [
17,
18].
Although other studies have estimated the health and/or economic burden of haemophilia A in other countries [
14,
16,
18‐
23], differences in population, available resources, clinical practice and unitary costs vary considerably among countries, making this an important topic for nations to address individually.
To our knowledge, health and economic burden of haemophilia in Portugal have not yet been assessed. There was a Portuguese retrospective hospital-based study conducted using electronic patient records, but its aim was estimating only medical costs [
13]. Considering the current economic setting in Portugal, the growing concerns about National Health Service sustainability [
24], and the arrival of new treatment options for patients with haemophilia A, it is crucial to understand the current burden of this disease in Portugal, to help support priority setting in healthcare policies and resource allocation.
This study was conducted to pursue the following main objectives: (a) assess the economic burden of haemophilia A in Portugal from a societal perspective and (b) estimate the health burden of this disease by calculating the disability-adjusted life years (DALYs) or healthy life years lost, as a result of haemophilia A.
Discussion
This study was performed using detailed data about epidemiology, resource consumption and productivity loss related with haemophilia A. Our findings confirm that management of this disease is costly, namely when patients develop inhibitors. The study suggests that haemophilia A imposes a significant economic burden on Portuguese NHS and subsequently, on society.
To our best knowledge this is the first study that attempts to characterize the health and economic burden of haemophilia A in Portugal approaching direct/indirect costs and DALYs.
We have chosen the broader societal perspective, which includes costs supported by the NHS and Social Security for patients/families. We found an annualised mean cost of €56,875 per patient with haemophilia A in Portugal in the societal perspective and €53,948 in the perspective of the NHS. This estimate is about 34 times higher than the public health expenditure per capita in Portugal of approximately €1602/year/person [
48].
Direct comparison with studies performed in other countries is limited, as mentioned in the Background section, but it is however possible to analyse trends in costs. Considering the proportion between direct costs and indirect costs estimated in our study (95% vs. 5%) similar results had already been reported in previous studies [
19,
20]. This is not surprising, since (a) the disease does not greatly interfere with professional activity in mild/moderate cases (approximately half of the population with haemophilia A) and (b) productivity loss has a low value compared with treatment costs due to high intensity of treatment, especially in prophylactic regimens, and cost of drugs. Our estimations show that prophylactic procedures, ITI treatment and
minor bleeding treatment are the
major cost drivers, accounting for 89% (€50,712) of the overall annual cost/patient. This finding is in accordance with results found by O’Hara [
18], Henrard [
19] and Rocha [
13] et al., the latter of which estimated an average cost of €44,134 per patient-year in a Portuguese hospital setting. Considering direct costs alone, these treatment resources account for nearly 94% of costs, a proportion similar to the previously described for Spain, Germany, Hungary and the UK by Cavazza et al. [
22]
Considering a total of 750 haemophilia A patients in Portugal and the annual cost obtained per patient, it is expected that around 40.4 million euros are spent every year by the NHS in the treatment of this disease. This amount corresponds to around 0.4% of the national health budget estimated for 2017, which consisted of 9.9 million € with additional 2.2 million euros from lost productivity [
49]. Assuming that the true haemophilia population ranges from a minimum of 539 [
50] to a maximum of 800 patients as reported by the experts, the true economic burden of haemophilia A to society may range between €31.8 million to €45.2 million. The high impact of the population with inhibitors to the economic burden of haemophilia A should be highlighted, since about 35% of haemophilia A societal annual expenditure is related with these patients, who represent only 6.5% of haemophilia A population.
Comparing different age groups, the annual cost per patient is 39% higher in children than adults. This can be explained by: (a) development of inhibitors mainly occurring in childhood; (b) consequent ITI administration and; (c) children being mostly treated with rFVIIa which is a more expensive treatment than aPCC, while the majority of adults are treated with aPCC. Nerich et al. [
14] also found differences in costs between age cohorts, but this study only included severe cases. The author identified that children have a higher consumption of clotting factor/kg because, on average, they may experience more bleeding episodes than adults, with more extensive secondary disease prevention and a less favourable post-surgical recovery. In the study performed by Rocha et al. [
13] in a Portuguese hospital, the total annual cost/adult (€49,338) was very similar to our estimate but the cost/paediatric patient (€37,805) was about half of our result with no statistically difference between adults and children (
p = 0.47). Rocha provides three possible explanations for the similar costs obtained in adults and children: the vast majority of children were teenagers and their body weight would be comparable to adults, different pharmacokinetics of clotting factor in children leading to a high consumption of clotting factor/kg in children and prophylaxis was used in 100% of children/teenagers versus 40–70% in adults with severe disease.
Cost comparisons by severity showed that yearly costs increase exponentially with severity of disease ranging from €401 per mild patient to €302,189 per patient with inhibitors. Our annual direct cost estimates per mild, moderate and severe (+/− inhibitors) patients are different from those found by Rocha et al. [
13] (€793, €8,495 and €77,587, respectively). Nevertheless, increments between severity levels follow a similar trend.
The development of inhibitors results in a great impact to the economic burden of the disease. Our data suggests that a patient with inhibitors costs 7.6 times more than a patient without inhibitors. This can be explained by: (a) higher incidence of bleeding episodes requiring consumption of clotting factor; (b) bypassing agents rFVIIa and aPCC for patients with inhibitors are more costly than plasma-derived FVIII and rFVIII for patients without inhibitors and; (c) need of ITI. Surprisingly, our estimate for the cost per patient with inhibitors was about the double of that described by Rocha et al. (€134,032) [
13] but we obtained a similar cost per patient without inhibitors (€40,318).
Additionally, in our study, the cost of severe patients without inhibitors was 3.5 times lower than patients with inhibitors, which can also be explained by b) and c) above. Similarly, Nerich et al. [
14] found that direct drugs costs in severe patients without inhibitors were two times lower than the cost of patients with inhibitors.
DALY is the primary metric used by WHO to assess the global burden of disease. The years lived with disability have the highest contribution to haemophilia-related DALYs (99%) whereas the years of life lost due to premature mortality were considerably low (1%). This result was expected, since life expectancy of patients with haemophilia A is close to that of general population due to great advances in the development of safer FVIII replacement products. The large contribute of YLDs in total DALYs demonstrate that patients with haemophilia A have poorer quality of life compared with general Portuguese population. In our study, patients with inhibitors had the largest DALY estimate per patient. Improving the quality-of-life and avoiding premature mortality is critical in these patients.
The results show that DALY estimation is largely dependent on the assumptions made in the calculation, namely, discount and weighting of age. DALYs are strongly related with the prevalence of diseases and therefore, DALYs from haemophilia A are normally lower since this is a rare disease. Further research and quantification of burden of disease in DALYs using disease specific disability weights, when available, will make the estimated burden more applicable to people with haemophilia A.
The strengths of this study are the comprehensive inclusion and description of all relevant components of cost when performing the cost-of-illness study, the use of real-world data to estimate bleeding frequency and the inclusion of productivity loss in adult patients and informal caregiver’s absenteeism. Given that many bleeding episodes are self-controlled by the patient, the use of patients’ self-reported frequency of bleeding was preferred compared to the information collected from the panel, since the latter was based on physician perception [
27].
Nevertheless, there are some limitations of our study that should be pointed out.
This study was partially based on data provided by an expert panel. This is a valid method that is widely used in several fields including health sciences. Experts were invited to a face-to-face meeting to present individual data based on their own casuistic regarding management of haemophilia A. Discussion between experts was encouraged by an independent medical moderator, namely to understand differences found between population of patients or practices. Since experts were treating 90% of the entire haemophilia A population in Portugal, the data herein presented is considered consistent, representative and without selection bias.
The cross-sectional survey had some limitations, namely, sample size by severity group, a one-year recall period, potential selection bias due to voluntary participation and missing values. The survey showed a trend for an increase in productivity loss with the severity of the disease, except when we move from adults with severe disease to adults with inhibitors probably due to missing data in lost working days. Nevertheless, it should be stressed that this observational study included 16% of the Portuguese population with haemophilia A.
Several other resources were not covered by this study such as loss of earnings by other informal caregivers (e.g. partners, other relatives of adult patients), use of a professional caregiver, emergency transportation in case of bleedings, home changes, alternative therapies (such as massage, nutritionist or homeopathy), treatment of adverse events resulting from anti-haemophilic drugs, presenteeism, premature death or intangible costs related with the disease such as anxiety, pain, reduced quality of life and suffering of patient and family. Further studies are needed to estimate the costs mentioned above on a proper and robust manner.
Moreover, experts mentioned other haemophilia A complications in adults besides bleeding, such as foot ulcer or limb pseudotumor that may result in amputation prosthesis, wheelchairs use, technical help, and orthotics, which were not quantified in this study.
Costs related with HCV co-infections were not considered, since we assumed, as reported by the experts, that in 2017 there were no new cases and all co-infected were successfully treated in the past. Nevertheless, re-treatments may occur in HCV and therefore costs are probably underestimated but were not possible to be quantified by the experts. Regarding HIV infection, estimates used for the HIV annual treatment cost did not include indirect costs and we have assumed only haemophilia A productivity loss to avoid double counting.
We have assumed that patients are 100% compliant with treatments but this assumption may not be valid in the real-world, resulting in lower treatment costs.
We were not able to find data about unemployment of parents related with having children with haemophilia A.
Hospitalization, consultations and emergency costs correspond to hospital reimbursement costs and were used as a proxy of the real cost. Nevertheless, these costs are irrelevant compared with clotting factor. Prices used for clotting factor are also a limitation since hospital tenders may decrease prices.
Weight is an extremely important factor when calculating the cost of clotting factor. Since we have not found any national available source describing weight by year of age, it was not possible to have the desirable granularity over this parameter. The impact of using a constant weight within age intervals should be further investigated with sensitivity analysis.
Although we recognise several limitations in the estimation of the economic burden, given the proximity between our results and other reported in previous studies and considering the wide national representativeness of the experts that provided the data, we believe that estimates from our study are accurate and robust and may be considered as representative of the Portuguese population with haemophilia A.
For the DALY calculation, a limitation relies on the fact that disability weights are a crucial factor in YLD calculation as they allow for adjustment of the number of years lived with disability and for comparison with the number of life years lost due the premature mortality. Unfortunately, these weights are not available for Portugal and estimates may lack external validity due to country specificities.