No association found between sickness absence and duration of pregnancy benefit

Abstract

Objective. Issuing sickness absence certifications is a time-consuming part of GPs’ and obstetricians’ work in antenatal care. The aim was to study whether a social benefit, the Pregnancy benefit, introduced in Sweden in 1980 and especially targeted at pregnant women working in arduous occupations, had any demonstrable effect on sickness absence among employed pregnant women.Design. Data from delivered women's medical and social security records were collected for 7459 consecutively delivered women in 1978, 1986, 1992, and 1997.Setting. Two delivery wards in southern Sweden.Main outcomes measures. The mean number of days of leave provided by the Pregnancy benefit increased by 8 days up to 1986, and by an additional 6 days between 1986 and 1997, making the total increase 14 days. The mean number of days of sickness absence increased by 19 days between 1978 and 1986, but fell by 24 days between 1986 and 1997. All occupational groups studied behaved in the same way.Results. No direct correlation was found between sickness absence and the number of days of the Pregnancy benefit that were used. Instead, changes in sickness absence among pregnant women were similar to changes in sickness absence among all persons insured, both men and women.Conclusions. Increasing costs of social security insurance among pregnant women are more probably coupled to levels in the social security insurance, as pregnant women seem more inclined to apply for a sickness absence certificate during periods when the economic compensation practically equals their salary.

• Pregnancy
• sickness absence
• social benefits
• women

Women in Sweden are employed almost to the same degree as men [1]. However, women are sick-listed far more frequently, especially pregnant women [2–5]. Comparisons are, however, not straightforward, as pregnant women have special social benefits that are provided in connection with work. These special benefits have been added over the past two decades, and each was intended to provide relief for pregnant women who were employed. Due to an economic recession, cuts in social benefits were made between 1992 and 1997. Compensation for the first days (1–3) of a sick-leave period were substantially reduced from almost 100% of the salary to 75–80% of that amount [6], [7]. The taking of sickness absence in Sweden peaked in 1989, and since then has decreased [3] both in the general population and among pregnant women [7].

In 1980, a new benefit, the Pregnancy benefit, was introduced for pregnant women working in occupations considered to be extra heavy and tiresome. The Pregnancy benefit was a complement to an already existing general benefit, the Parental benefit, available since 1974. The intention behind the Pregnancy benefit was to ensure that pregnant women working in heavy occupations would not have to make use of their Parental benefit days before delivery or apply for sick leave. It was understood that the special character of their work might entail a need for time off from these types of jobs [8]. The percentage of pregnant women granted the Pregnancy benefit by the social authorities increased gradually from its introduction in 1980 to 28% in 1997 [7].

It seemed therefore reasonable to assume that the introduction of the Pregnancy benefit would diminish the need for working pregnant women to apply for sick leave, as the pregnant women with the most strenuous working conditions had been supported. Working pregnant women seem to be most sensitive to strain and fatigue at the very end of the pregnancy and thus the work of the physicians employed in antenatal care would be facilitated [9–11]. Knowledge of the societal and public health effects of a new benefit ought to be of great general interest from a scientific as well as from a public health point of view, especially if a society is planning to extend its social benefits [12], [13].

For this reason, we collected information on 7459 delivered women for the years 1978, 1986, 1992, and 1997 with regard to occupation, age, parity, sickness absence, Parental benefit, and Pregnancy benefit.

The aim of the study was to investigate whether a more frequent and generous granting of the Pregnancy benefit contributed to any changes in the frequency with which working pregnant women took sickness absence.

Material and methods

In order to obtain a representative population of pregnant women, we combined samples of delivered women from two hospitals (the University Hospital in Linköping and the County Hospital in Värnamo) for the years 1978, 1986, 1992, and 1997. In Värnamo, all women were included, but as Linköping had far more deliveries, we included every second consecutively delivered woman on the assumption that they were randomly collected.

When we started the study in 1988, we wanted to go back 10 years for comparison, but finally chose the years 1978 and 1986, as the diagnoses were set according to the same ICD standard [5]. However, as the social benefits were cut in 1992, due to an economic recession, the year 1992 was added and included as well as 1997, because compensation was partly restored in 1998.

Information on occupation

Information concerning type of occupation and degree of employment was based on the pregnant woman's statement at her first visit to the Antenatal Care Unit [5]. As, by definition, only employed pregnant women could apply for the Pregnancy benefit, only women with stated occupations could be included. Students, housewives, the unemployed, and women with unknown occupation were therefore included in a group called “no occupation stated”.

A total of 25 occupations were registered and were merged into four main groups to facilitate comparisons:

1. Administrative work: Bank teller, postal clerk, secretary, etc.

2. Industrial work: Assembler, farming, weaver, truck driver, etc.

3. Service: Waitress, food counter, hairdresser etc.

4. Health and care: Registered nurse, auxiliary nurse, childcare worker, etc.

Social benefits for pregnant women 1978–1997

All information regarding sickness absence and the use of Parental and the Pregnancy benefits was collected manually from the women's files at the local social insurance offices by the authors and presented as whole days to facilitate comparisons. All days of sickness absence were included whether a GP or an obstetrician had issued the certificate.

Since 1974 the Parental benefit has been available to all pregnant women and was gradually expanded to 450 days in 1989. Sixty days of the Parental benefit could be used before delivery, but the number of days used is then subtracted from the total number of days available after delivery.

The Pregnancy benefit offers 50 days of paid leave during the last 2 months of the pregnancy but, in contrast to the Parental benefit, no days are subtracted after delivery [5], [8].

Statistics

Besides the chi-square and Student's t-test, the 95% confidence interval (95% CI) for the mean was used to test whether observed differences between means were different from zero at the 5% level of statistical significance [15].

Ethics

The Ethics Committee, Linköping University, approved the study.

Results

The Pregnancy benefit was granted to almost half of the women employed in all but the administrative types of occupations in 1997. The percentages of women in each group granted this benefit are presented in Table I.

Table I.  The Pregnancy benefit: Changes in four occupational groups among pregnant women employed 1986–97.

Type of occupation	1986		1992		1997
	%	95% CI	%	95% CI	%	95% CI
Administrative	2	1–3	6	4–9	6	4–8
Industrial	18	14–23	28	22–33	38	31–44
Service	19	14–25	37	30–43	44	37–51
Health and Care	29	25–33	37	33–41	47	42–52
Total	17	15–19	25	23–27	28	26–31

The average number of days that a pregnant woman was supported by sickness absence compensation, the Parental benefit, and the Pregnancy benefit are presented in Table II and in Figure 1. The number of days of Pregnancy benefit support provided in 1986 was 8.4. The number of days of sickness absence was 21.6 in 1978 and 40.1 in 1986, an increase of 18.5 days. As the number of days used under the Pregnancy benefit programme continued to increase, there was a concomitant decrease in the number of days of sick leave between 1986 and 1997. However, the increase in the number of days (5.6 days) of leave taken under the Pregnancy benefit was not matched by the decrease in the number of days of sickness absence. Between 1986 and 1997, there was a far greater decrease in the average number of days of sickness absence. In 1986, the average number of days of sick leave was 40.1, and in 1997 it was 16.1, a decrease of 24.0 days.

Figure 1. Relations of mean days of sickness absence, Parental and Pregnancy benefit 1978–97.

Table II.  Mean number of days provided by benefits in connection with pregnancy (sickness absence, Parental benefit, and Pregnancy benefit) among pregnant women employed 1978–97.

Type of occupation	1978		1986		1992		1997
and type of benefit	Days	95% CI	Days	95% CI	Days	95% CI	Days	95% CI
Administrative
Pregnancy benefit	(*)^1	(*)^1	1.0	0.4–1.7	3.2	2.2–4.3	3.2	2.2–4.2
Sickness absence	13.8	10.8–16.8	21.5	18.4–24.5	14.5	11.6–17.4	10.0	7.9–12.2
Parental benefit	10.7	9.6–11.9	6.7	5.9–7.5	5.6	4.7–6.4	5.6	4.8–6.4
Industrial
Pregnancy benefit	(*)^1	(*)^1	9.2	7.1–11.3	13.8	11.2–16.3	18.8	15.6–22.1
Sickness absence	38.6	31.8–45.5	61.9	55.2–68.7	30.5	24.9–36.2	20.5	15.7–25.2
Parental benefit	16.9	15.1–18.8	6.1	5.2–7.0	8.0	6.7–9.3	9.9	8.2–11.5
Service
Pregnancy benefit	(*)^1	(*)^1	9.6	6.9–12.3	18.3	14.9–21.6	22.0	18.5–25.6
Sickness absence	23.2	17.0–29.4	49.7	42.0–57.5	26.1	19.2–32.9	23.9	16.9–31.0
Parental benefit	15.6	13.4–17.7	6.1	5.0–7.2	5.6	4.2–7.1	6.5	5.0–8.0
Health and Care
Pregnancy benefit	(*)^1	(*)^1	14.6	12.4–16.7	18.6	16.4–20.7	23.5	21.1–25.9
Sickness absence	18.6	14.8–22.3	38.0	33.4–42.6	22.6	18.4–26.7	19.0	15.2–22.7
Parental benefit	15.7	14.4–17.0	6.0	5.3–6.7	5.9	5.1–6.8	7.1	6.0–8.1
Total
Pregnancy benefit	(*)^1	(*)^1	8.4	7.4–9.4	12.4	11.3–13.5	14.1	12.9–15.3
Sickness absence	21.6	19.3–23.9	40.1	37.4–42.8	21.9	19.7–24.2	16.1	14.2–18.0
Parental benefit	14.2	13.4–14.9	6.2	5.8–6.7	6.2	5.7–6.7	6.8	6.2–7.3

¹The Pregnancy benefit was introduced in 1980.

The changes were most clearly expressed in the industrial types of occupations, where the increase in the days supported by the Pregnancy benefit was 9.6 between 1986 and 1997. Sickness absence, however, decreased by 41.4 days from 1986 to 1997.

The average number of days of Parental benefit used per employed pregnant women prior to delivery was 14.2 days in 1978. The number of days then fell significantly to 6.2 in 1986 (p < 0.001) and remained low, 6.2 days in 1992 and 6.8 days in 1997 (ns).

The main change in the studied populations of delivered women was an increase in the percentage of women who could not be included in any of the four occupational groups. In 1978, this was 20% and in 1997 36% (p < 0.001). These women could not be placed in any of the 4 occupational groups, because no information on occupation was stated in their files and because there was an increase in the percentage of more women registered as students or as unemployed, 2% in 1978 and 18% in 1997 (p < 0.001).

No differences in the proportions of nulliparous/parous women were found, however. The mean age of the women increased from 27.1 years in 1978 to 29.8 in 1997 (p < 0.001).

In order to illuminate the concordance between sickness absence taken by pregnant women in this study and all insured persons (men and women aged 16–64 years in Sweden for the years 1975 to 2000), the mean number of days of sickness absence is plotted in Figure 2. A result from a survey in Stockholm County for the year 1988 is also included to illustrate the high level of sickness absence among pregnant women for that period [8].

Figure 2. Mean number of days of sickness absence for all insured persons in Sweden 1975–99 aged 16–64 years. The results for employed pregnant women in this study are plotted for the years 1978, 1986, 1992, and 1997. The mean number of days is multiplied by 9/12 in order to facilitate comparisons for one whole year. For 1988, a result from the urban area (the county of Stockholm) is included [8].

Discussion

Several studies on work and pregnancy in Western society conclude that society ought to provide improved social benefits or ease the working conditions for pregnant women instead of forcing them to apply for sick leave for such a normal condition as pregnancy [9], [12], [16–18], [20–22].

The results in this study, however, do not support such conclusions. We found no clear interaction between the number of days of Pregnancy benefits granted and the number of days of sick leave taken by working pregnant women. Paradoxically and contrary to what was expected, the average number of days of sickness absence increased from 1978 to 1986 in spite of the introduction of the Pregnancy benefit in 1980. Between 1986 and 1997, the use of the Pregnancy benefit programme increased significantly from 17% in 1986 to 28% in 1997 among pregnant women in this study. As there are no indications that the working conditions for women deteriorated in Sweden during the period 1987 to 1997, the increase in the percentage of women granted the Pregnancy benefit is most probably due to a more generous attitude among the social authorities that grant this benefit. With the exception of the administrative type of occupations, almost every second pregnant woman employed was granted the Pregnancy benefit in 1997. The number of days of sickness absence, however, fell five times more than the increase in the number of days granted by the Pregnancy benefit programme during the period 1986 to 1997. The observed changes in sickness absence were similar in all four occupational groups, irrespective of the change in the level of use of the Pregnancy benefit.

Studies on sickness absence are few and the general conclusion is that sickness absence among all persons insured and especially during pregnancy has increased in Western societies, such as for instance Sweden, Norway, Denmark, and France [5], [14], [19], [20]. So far, no clear medical explanations have been advanced to explain the high rates of sickness absence observed among pregnant women in Sweden or for instance other Nordic countries – countries with ample social benefits for pregnant women and a well-organized antenatal care [5], [11], [19], [20]. Instead, increasing levels of sickness absence have been attributed to possible changes in attitudes towards work during pregnancy and a broadening of the criteria for sickness absence to also include common complaints normally associated with pregnancy, complaints that otherwise might not be regarded as “illnesses” and that at the same time have been adapted to the prevailing social and economic benefits system [5], [7], [17–19]. However, if the compensation provided by the social benefits is cut as reported in this study, most pregnant women choose to go to work for economic reasons and do not try to apply for a sickness absence certificate.

Healthcare for normal pregnant women in the Nordic countries is generally organized within a primary healthcare setting and three main professions, general practitioners, midwives, and gynaecologists/obstetricians, provide care. In Sweden antenatal care is usually organized in separate maternity units staffed mainly by midwives [23]. However, in spite of different organizations, we believe that the problems encountered by physicians working in antenatal care are similar.

Our opinion in this study, that levels of sickness absence are determined by the patients’ attitudes and wishes, is in accordance with findings by Englund, who stated that a wish for a sickness absence certificate as a rule was put forward by the patient, even if the doctor could not in all cases find a suitable diagnosis [24], [25]. Studies on the practice of sick listing by physicians indicate no change in spite of an administrative reform intended to narrow sick-listing benefits [26]. The changes in our study may thus be a result of a diminishing demand for a sickness absence certificate owing to economic considerations among the pregnant women themselves. The results suggest that there may be a certain overuse of the sickness absence benefit in periods when the coverage practically equals the sick-listed person's salary. The observed decrease in sick leave taken by all insured women and men in official statistics supports such an explanation. The mean number of days of sick leave per insured person fell by 60% among pregnant women in this study and by 56% among men (from 24 days in 1986 to13.6 days in 1997) between 1986 and 1997 [3]. It is hardly likely that any medical conditions could cause practically identical changes in two such disparate groups. Furthermore, the recently observed pattern that sickness absence has increased by 62% (from 11.2 days in 1998 to 18 days in 2004) among all insured persons (both men and women) since 1998, when levels in the social benefits were again substantially increased, also supports such an assumption [3], [27–29].

The use of the general Parental benefit among working pregnant women in this study, from 1978 to 1998, also suggests that a pregnant woman's use of social benefits is rather more influenced by economic considerations than by recommendations from her doctor. The society encourages general practitioners and obstetricians to restrict the use of sickness absence and instead to make use of the general Parental benefit, which is available to all pregnant women [18]. This was so especially among pregnant women who had not been granted the Pregnancy benefit, as their occupations were not considered as heavy by the social authorities. One could therefore assume that if a woman could not fulfil the criteria for a sickness absence certificate, she would instead draw on more free days from her Parental benefit. However, the mean number of days used from the Parental benefit has been unchanged since 1986. This is most probably due to the fact that pregnant women prefer to save all the days provided by the Parental benefit for the period after delivery and practically all pregnant women who are refused the Pregnancy benefit are instead sick-listed [8], [30].

Our study is based on consecutively delivered women and supports the importance of longitudinal studies. So far, we have not found any other study focusing on changes in sickness absence and social benefits during a time period in which sickness absence both increased and decreased in concordance with economic changes in social benefits.

Conclusion

The introduction of the Pregnancy benefit in 1980 and the subsequent granting of 50 days free from work had no demonstrable influence on the number of days of sick leave taken among employed pregnant women. It is more probable that pregnant women behave in the same way with regard to sickness absence and personal economy as all other insured persons in society. The accordance in the levels of sickness absence between pregnant women and all insured women and all insured men suggests that economic considerations may play a vital role in explaining changing levels of sickness absence, as all the groups mentioned ought to have very disparate causes for sickness absence.

Key Points

It is generally assumed that improved conditions in the form of extended or new social benefits, for instance the Pregnancy benefit, introduced in Sweden in 1980, will have an impact on the use of sickness absence among pregnant women gainfully employed.

• A steady rise in the number of days of a new benefit, the Pregnancy benefit, aimed at pregnant women working in arduous or stressful occupations, had no correlation to sickness absence.

• Sickness absence among pregnant women instead correlated to changes among all insured persons, both men and women.

• Efforts by physicians working in antenatal care to increase the number of days of the Parental benefit before delivery, as wished by the authorities, had no visible effect.

The Health Research Council in the south-east of Sweden financially supported the study.