All consecutive patients admitted to the cardiology department of Santa Marta Hospital aged <66 years (official retirement age) between July 2016 and July 2017 with MI and who survived to discharge were included in the present study. Demographic characteristics, clinical variables, treatment and outcome were obtained for each patient. Employment status on admission was assessed in every patient. All those reporting any form of regular paid work were considered employed. In subsequent follow-up contacts, performed at 30 days and one year after admission by telephone interview with a dedicated nursing team, the total duration of absence from work after the MI episode and any change in employment status during this period were assessed. The number of days absent from work for cardiology appointments and exams was also assessed.

To obtain the national numbers of patients with MI, the national Diagnosis-Related Group database was searched and all patients were selected with episodes in which the primary diagnosis was classified as CM-410 in the International Classification of Diseases (ICD)-9 or CM-I21 in the ICD-10 during 2016 (latest available data).4 For the present analysis, only patients aged <66 years were included.

Estimation of indirect costs followed the human capital approach, whereby productivity losses can be proxied by the cost of labor to employers. Monthly wages were estimated from market rates from the Ministry of Labor database (grossed up by social security contributions) according to gender and age in 2016 (Supplementary Table).5 Additional regular wage values were included on top of the base wage. These figures were adjusted for Social Security contributions by multiplying by 1.2375, because the economic value of lost production is based on gross wages and includes employers’ contributions to Social Security (23.75%). Other irregular wage components were also included in the estimates but not grossed up by contributions. Wages estimated on the basis of 2016 data were updated to 2018 using an index published regularly by the Bank of Portugal covering changes in wages reported to Social Security.

A daily cost was calculated to assess the cost of temporary absence from work by multiplying the monthly wage by 14/365. These figures assume weekends and holidays were included in the reporting of sick leave. The total cost of temporary absence from work (sum of the costs for each working patient) was calculated in this sample and the results were applied for the total number of MIs in Portugal during 2016. Indirect costs also included absence from work for cardiology appointments and exams, assuming a half-day absence from work for each. A strict definition of indirect costs was adopted in accordance with the guidelines of the National Authority for Medicines and Health Products (INFARMED).6 Data from the Portuguese Registry on Acute Coronary Syndromes (ProACS) were used to adjust our results for national data.7

Continuous data are presented as means ± standard deviation if they presented a normal distribution and as medians for data without a normal distribution, as assessed by the Kolmogorov-Smirnov test. Categorical variables are described as frequencies and percentages.

The variable ‘number of days absent from work’ was highly skewed, and so a natural logarithmic transformation (Ln) was performed that was used in the subsequent regression analysis.

Linear regression analysis was performed to identify predictors of ‘Ln absent days from work’. All variables with p<0.10 were included in the multivariate model.

The statistical analysis was performed with IBM SPSS, version 20 and STATA, version 14. A p-value below 0.05 was considered statistically significant.

During the study period, 233 patients aged <66 years were admitted to our cardiology department with an acute MI. Fourteen patients who were foreigners passing through Portugal or Portuguese nationals working in other countries were excluded from the present analysis. The final sample included 219 patients, with a mean age of 54±7 years (25-65 years), 83% male, 66.2% actively working, 16.4% early retired, 11.9% unemployed and 5.5% in long-term absence from work due to noncardiac disease (Table 1). A significant proportion of patients had retired before the age of 60. During the one-year follow-up, there were no changes in employment status.

Mean labor costs are presented in Table 2. In our sample, the mean monthly labor cost of working individuals was 1802 euros (69 euros/day), higher in males. The median number of days absent from work was 34 days (31 days in men and 52 days in women) (Figure 1 and Table 3). In total, there were 10 679 days of absence from work in this population and 285 cardiology appointments or exams (median two half-days each), with a total cost of 760 521.55 euros (sum of the costs for each employed patient). The mean indirect annual cost was 5244.97 euros per patient employed. Adjusting for the proportion of working individuals, the mean annual indirect cost was 3472.17 euros per patient.

Figure 1.

Density plot of days absent from work in men and women.

(0.05MB).

The above estimate of the mean indirect cost per patient does not take into account the large differences between patients with STEMI and those with NSTE-ACS.7 Significantly higher costs were obtained in STEMI patients, despite similar wages in the two groups (Table 4). This results in a mean total cost of absence from work of 5884.94 euros per patient with STEMI and of 2448.11 euros for NSTE-ACS.

In 2016 in Portugal, 4133 patients aged <66 years were admitted with MI who survived to discharge (Table 5). Based on data from ProACS, the proportion of STEMI vs. NSTE-ACS in the previous five years was 41% vs. 59%.7 In 2016, according to these proportions, 1695 patients would have had STEMI and the other 2438 NSTE-ACS, which would correspond to 1097 working patients in the STEMI group and 1497 working patients in the NSTE-ACS group, according to the proportions obtained in our sample (Table 4). We estimate a total indirect cost of 6 455 779.18 euros for STEMI and 3 664 820.67 euros for NSTE-ACS (total indirect cost of 10 120 599.85 euros).

In univariate analysis of the working population, potential predictors of absence from work were age, previous MI, STEMI and gender. In this preliminary analysis, gender was a significant variable, but in a more detailed analysis, gender was only significant when associated with admission diagnosis, and so an interaction term of ‘gender’ × ‘main diagnosis’ was introduced in the final regression model. In this multivariate linear regression analysis, the most important clinical predictors of the number of days absent from work were age, previous MI, STEMI and male gender in STEMI patients (Table 6).

Unit values of production, relying on wages, were estimated through official statistics and were based on gender and age group. This methodology provides only a rough estimate of individual values. Nevertheless, of all available options, to our knowledge this approach is the most appropriate.

In the present study, it was not possible to assess whether return to work was complete or whether productivity on return was somewhat reduced, which could have increased indirect costs further.

Although the retirement age is 66 years, some individuals remain in the labor force for longer periods. Our analysis only studied patients aged <66 years. Had we included older age groups, more indirect costs would probably have to be considered.

In addition, no indirect costs were calculated resulting from premature death in individuals at active ages or from productivity losses in individuals working in jobs outside the formal market but with social value (such as housewives).

Due to the relatively short follow-up, our data were also not suitable to study the influence of MI on early retirement.

Indirect costs were estimated on the basis of a single urban center. However, this is a tertiary hospital, and patients from peripheral hospitals who were referred to our hospital for coronary angiography were also included in the analysis. For this reason, the data are to some extent representative of the national situation. It has also been shown that in high-income countries such as Portugal, rates of major cardiovascular events and death are similar in urban and rural areas, particularly in countries with universal coverage, ensuring good health services access and treatment in both urban and rural areas.14 This is expected to be the case in Portugal, with possible exceptions in a few regions.