Cardiovascular disease (CVD) is responsible for about one third of global mortality every year, with 18 million CVD deaths in 2015 worldwide, of which 15 million deaths were due to ischemic heart disease and stroke.1 CVD was responsible for 29% of the 107000 deaths in Portugal in 2016.2 Although mortality from CVD in Portugal has decreased by more than 40% over the last 10 years,3 the number of patients discharged from the hospital having survived an event is not decreasing and therefore there are more patients alive with a history of the disease.3,4 Thus, the burden of CVD morbidity is expected to grow further, weighing heavily not only on healthcare systems and patients’ quality of life but also on work productivity, as well as increasing the need for informal care of people with CVD.

To our knowledge, existing information on productivity losses after CVD in Portugal is very limited.5 In a multi-country study, the direct and indirect costs of cardiovascular disease in Portugal for 2003 were estimated at around €1762 million, of which €969 million was healthcare costs.6 A more recent study conducted in the European Union in 2017 estimated the overall cost of CVD in Portugal at €2.8 billion a year, with direct healthcare costs accounting for 42% of total costs, work productivity losses for 27% and informal care for the remaining 31%.7

The aim of the current study was to estimate productivity loss and indirect costs for patients and caregivers in the first year after an acute coronary syndrome (ACS) or stroke in Portugal.

In our study we observed substantial productivity losses and indirect costs in the first year following a CVE in Portugal, with 47 work days lost by patients and caregivers after an ACS and 76 work days after a stroke, and mean total indirect costs of €5403 and €8726, respectively.

About 80% of productive time in ACS patients was lost due to the patient's absence immediately following the CVE. This is much more than the average of 60% that was observed in the other six European countries that participated in this study.4 Once Portuguese patients returned to work they continued to lose productive time due to absenteeism (five work days for ACS and 11 work days for stroke patients) and presenteeism (two work days for ACS and seven work days for stroke patients). The loss of productive time observed in Portuguese patients was considerably lower than the average time lost in the other six countries in the project. After returning to work Portuguese ACS patients lost on average seven work days during the first year after the event, while stroke patients lost 18 work days. The corresponding losses in the other six European countries were 24 work days for ACS and 22 work days for stroke patients. A possible explanation for the lower productivity loss compared to France, Poland, Spain, Switzerland and the UK could be the proportion of blue-collar workers in the Portuguese sample (45% in Portugal vs. 26% in the other six countries). Blue-collar workers may be more concerned about losing their job for being on sick leave for too long or being unproductive. However, it should be noted that unemployment in Portugal, although higher than in Switzerland, UK, Poland and Belgium, is much lower than in France and Spain, where such low productivity losses were not seen after returning to work.13 ACS patients in Portugal tended to lose less productive time due to the initial hospitalization and sick leave compared to the other participating countries (31 vs. 37 days). Portuguese stroke patients, on the other hand, lost more time than did stoke patients in the other six countries that participated in the project (47 days vs. 31 days).

The considerable presenteeism that we observed in stroke patients could be explained by cognitive problems and fatigue after the event that are not considered in the mRS score. Although the Portuguese stroke population had the same proportion of stroke patients with mRS ≤ 1 as did the other countries in the project (77%), the distribution of patients between mRS = 0 and mRS = 1 was different. In the Portuguese population 55% had mRS=1 whereas in the overall study this population comprised a mean of 38%. This could explain the longer initial hospitalization and sick leave and longer period of absenteeism in Portuguese patients.

According to the 2016 literature review by Gordois et al., very little information is available in Portugal on indirect costs in general and on cardiovascular disease in particular.5 Gouveia Pinto reported that the estimated cost of MI for the first 12 weeks was €5450, which is very close to what we observed in ACS patients (€5403).14 For stroke our estimate of indirect costs seems to be double the direct cost as reported by Miguel et al. (€8726 vs. €4136).15

An important strength of our study is that, unlike most indirect costs studies, which focus on retrospective investigation of databases or administrative records, we collected information directly from CVE patients to capture information on absenteeism after return to work, presenteeism, and caregiver help. We used the most recent validated tool (iPCQ) specifically designed for assessing productivity losses over the previous four weeks – the longest period that can be assessed without recall bias having a significant effect.9

We recruited ACS and stroke patients who were still of working age, were employed prior to their event and were able to return to work afterward. This population is expected to be healthier and younger than the overall population of ACS and stroke patients. Therefore, our study provides a conservative estimate of productivity losses in ACS and stroke patients; the burden of productivity loss and indirect costs in the overall population of ACS and stroke patients is likely higher, as patients with more severe ACS and stroke may take more than 12 months to return to work (for example because they require longer rehabilitation) or in fact never return to work due to acquired disability.

The main study limitations are typical of all cross-sectional studies with patient-reported data. Our patients were recruited from four different centers. Although every attempt was made to select centers in different regions across the country, most of the centers served urban populations, leading to the possibility of rural populations being under-represented. These factors may limit the generalizability of our results to a broader ACS and stroke population in Portugal.

Prorating was based on individual absenteeism, presenteeism and caregiver loss patterns over the previous four weeks and assumed to remain constant between return to work and the end of the year. We analyzed the distribution of patients who were recruited within six months of the event and those who were recruited 6-12 months after the event, and found no difference in productivity loss patterns depending on the time elapsed since the event.

By focusing on employed patients only, we did not take into account caregiver time in patients who were unemployed or did not return to work. Since the overall population of MI and stroke patients does include these patients and their caregivers, the estimates for overall caregiver productivity losses are likely to be higher.

Concerning wages, salary-related differences between genders and age groups could have been considered to refine this value. However, given that this analysis was part of a multi-country project we opted to use the same source (Eurostat) for all the different countries, with the associated limitations.

For this analysis, a validated questionnaire (iPCQ) was used and instructions for users on how to translate the answers into costs were followed. For caregivers it is not known whether they were employed, and if so, whether they performed their caregiving tasks during working time.

Our findings suggest that ACS and stroke were associated with substantial productivity loss and indirect costs in Portugal. ACS patients lost a mean of 17% of their work days during the first year after the event and stroke patients lost 27% of work days. In addition, caregivers lost 4% and 5% of their annual productive time helping ACS and stroke patients, respectively. In Portugal, productivity loss in terms of work days lost associated with ACS was lower than in the other six European countries that participated in the study. Productivity loss in Portuguese stroke patients was similar to that observed in the other participating countries.

The associated indirect costs, €5403 for ACS and €8726 for stroke, are comparable to the direct medical costs of these conditions as reported in earlier studies, and therefore potentially double the total cost of ACS and stroke from a societal perspective. Similar relationships between direct and indirect costs of ACS and stroke were found in all seven European countries that participated in the study.

Qualified researchers may request data from Amgen clinical studies. Complete details are available at http://www.amgen.com/datasharing.

NM, HP, SR, AGF, ADA: contributed to patient recruitment, interpreted the results, critically reviewed the manuscript.

ES, LG, MR, RM: developed the study concept and design, coordinated the centers and patient recruitment, supervised the analyses, critically reviewed the manuscript.

LG: wrote the study protocol, programmed the analyses.

MR: drafted the manuscript.

This study was funded by Amgen (Europe) GmbH.

ES is a full-time employee at Amgen (Europe) GmbH and owns Amgen stock options. RM is a full-time employee at Amgen Biofarmacêutica Lda (Portugal). IQVIA, employer of LG and MR, received consulting fees from Amgen (Europe) GmbH.