Valued and received forms of support among Swedish working carers of older people: A descriptive study with focus on ICT-mediated support

1.1Work and care

About 70% (900 000) of Swedish carers are in work [4]. The largest group of working carers are middle-aged adults that regularly provide care for a family member, about 80% of whom are an older family member, most often a parent [5]. Both women and men provide care, while women provide the most demanding care and experience more difficulties to focus on their job [6].

It is acknowledged that working carers can gain satisfaction from providing care, which can sustain them in their role [7]. Satisfaction can arise from the relationship with the care recipient, from feeling appreciated and supported as a carer, and from a sense of accomplishment [8, 9]. However, most research has focused on the negative impact of the challenges arising from providing care [10], such as high levels of dependency and behavioural disturbance in the care recipient that can cause work interruptions, and restrictions on the carer’s social life, leading to a reduction in his or her psychological wellbeing [8, 9]. A negative impact can also be related to the strain arising from the combination of care and work obligations. For example, providing care can impact on work performance, and work-related pressures can interfere with the provision of care commensurate with a spillover-crossover effect [10, 11]. However, working carers may equally value their work as it provides social contact and support, a form of respite from caring and opportunity for self-realisation and independence [10]. The complexity of work-care reconciliation with fluctuating care needs and varying forms and levels of support can make achieving a work-care balance more difficult. This can potentially be detrimental to working carers’ mental and physical health [12] and may lead to financial insecurity and loss of pension benefits [13]. While giving up work to provide care is not an option for many Swedish carers, 10% reduce their working hours to care for an older family member while some 5% give up work altogether [4, 14, 15], with no difference between males and females [6].

1.2Support for working carers

In Sweden, the need to support working carers is receiving attention from policy and carer organisations [16, 13]. Family care support may consist of several forms in four areas addressing carer needs: 1) Health and social care management, such as to learn and access information about caring issues of relevance to their own situation (e.g., about the care recipient’s needs and available support); have practical assistance in caring (e.g., via information and training in care provision) and in planning and coordination of care/help/support; 2) Social, peer-to-peer, and emotional support, such as to talk about their situation and share their experiences with others (e.g., to cope with caregiving) via social support, counselling, and carer support groups; 3) Carer relief and respite care, and care recipient independence, such as to have access to relief and respite care (temporary supervision of the care recipient by a formal care provider to relieve family carers) and day care for the care recipient; and 4) Financial support [17, 18, 19]. Anecdotal reports from Swedish municipal family support units imply that existing services such as information, counselling, and carer support groups are underused by working carers [16, 15]. Earlier empirical evidence validates such claims, with data indicating that Swedish working carers, especially non-cohabiting/working daughters of older care recipients, report low frequency of use and limited benefit of support [17]. Be that as it may, in these studies the perspectives of working spousal carers seem underrepresented, and results are based on a fairly narrow definition of support and venues for implementation.

Other forms of support may also be added to the list above, such as various information and communication technologies (ICTs). These are increasingly recognized as innovative means to increase accessibility of, and flexibility in, support [20]. ICT-mediated support, such as support available via the Internet, has been proposed as a means to provide Swedish working carers with flexible access to information, psychosocial support, and learning opportunities, which would otherwise be difficult to obtain due to limited time while working and providing care [21, 22]. Such forms of ICT-mediated support and social, peer-to-peer, and emotional support have been reported to reduce carer burden [23, 24] and stress [25, 26, 27]. Telecare and assistive technologies for remote monitoring, alerts and distance communication, used both outside and at the workplace, may help the care situation and promote independence of the care recipient [19]. In various countries (e.g., the United States and the United Kingdom), these forms of support for working carers are reported to reduce burden, provide relief and peace of mind from being relieved from caregiving duties and from knowing that the care recipient is safe [ct. [24, 26, 28]]. However, such forms of support may also produce an unintended burden arising from the additional responsibility of responding to alerts and from the daily maintenance of the technology [23]. ICT-mediated support can also have a positive effect on secondary stressors arising from work–care conflict [cf. [10, 11]], connected to reduced work-care interference, and increased work productivity and work participation [26, 27]. However, to our knowledge studies on support for particularly Swedish working carers that focuses on ICT-mediated support are still sparse.

1.3The present study

The main aim of this study was to describe the perceived value of different forms of support among Swedish working carers of older people, with a focus on ICT-mediated support.

The primary research questions were: (1) What forms of support are valued? (2) How are ICT-mediated forms of support valued relative to non-ICT forms of support? and (3) To what extent are valued forms of support received? The secondary research questions were: (1) What aspects of working life are affected by providing care? and (2) are caring characteristics and work-care conflict associated with the perceived value of forms of support?

2.1Design

A cross-sectional questionnaire-based survey design was used as we wished to access a broad range of carers and measure their perceived value of support at a set point in time.

2.2Sampling frame and participants

A convenience sampling strategy was used to recruit participants via two channels, the Swedish Dementia Association (Svenska Demensförbundet) and Carers Sweden (Anhörigas Riksförbund) as a way to access a sample of working carers from a large, estimated population of working carers. Inclusion criteria for participants were: being over 18 years of age; being self-employed or in paid employment/work; and presently caring for a person aged 65 years or older. A total of 129 participants were recruited. Due to the nature of the recruitment process (see Procedure, below), a response rate cannot be determined.

2.3Materials

The questionnaire contained a selection of standard items for demographic information, items developed by the research team, and validated items and/or scales. The questionnaire consisted of four main sections: demographic details; the family care situation; work-care conflict; and valued and received forms of support.

2.4Procedure

An advertisement for the study was placed in a member bulletin of the Swedish Dementia Association in February 2016, and also sent in May 2016 via an email registry of delegates that had previously attended the annual National Carers’ Day conference (Anhörigriksdagen) organized by Carers Sweden. The advert directed potential participants to a website containing an online questionnaire. For the Swedish Dementia Association potential participants were also offered a paper-based format, which none of the participants requested. The questionnaire was accessible from February to July 2016. One advert update was placed in the Swedish Dementia Association member bulletin and two email reminders were sent out to Carers Sweden’s email registry.

2.5Data analysis

SPSS v. 23.0 was used for all statistical analyses. Descriptive (univariate) statistics were used to describe the sample characteristics and study variables. Spearman’s rho (rs) was calculated for associations between the perceived value of support and work–care conflict, while point biserial correlation (rp⁢b) was calculated for associations between the perceived value of support and carer characteristics (being a spouse/non-spouse and dementia/non-dementia carer). Reliability analysis was performed to calculate Cronbach’s alpha (α) as a measure of the internal consistency of the work conflict scale. Alpha was set at p< 0.05 for each test, with no adjustment for multiple testing; note should therefore be taken of the potential for an inflated type I error rate, and consideration given to the effect size of significant associations.

The response options “very often,” “often,” and “sometimes,” for the questions on the influence of work on using support were collapsed into one category. The response options for the perceived value of support were collapsed from four into two, with “very” and “quite” combined and “a little” and “not” combined. The 24 forms of support were categorized, via a triangulation process involving study authors (SA, LM, EH, KM) until consensus, into four support areas (as listed in the Introduction), and another category, i.e., support that didn’t fit in to a clear category.

2.6Ethics

This study was conducted in compliance with ethical principles of the Helsinki Declaration [30] and was approved by the Swedish regional ethics committee (dnr 2016/8-31).

3.1Demographic characteristics

Descriptive data on the sample is presented in Table 1. The majority of the sample was female (n= 98; 95.1%), married/cohabiting (n= 87; 84.5%), college or university-educated (n= 78; 75.7%), and aged 45–65 (n= 86; 83.5%). The majority of participants worked full-time (n= 70; 64.8%), with most employed in administrative work (n= 38; 40.9%).

3.2Family care situation

The majority of participants had been caring for their care recipient for more than 4 years (n= 78; 60.4%) and was caring for a parent/parent-in-law (n= 88; 68.2%), in a large number of cases a care recipient with dementia (n= 71; 55%). Just over half of the sample (n= 66; 51.2%) provided 6 hours or more of care per week. The majority (n= 100; 77.5%) cared for someone living in her/his own dwelling, or for someone with whom they were cohabiting. Further data on the care situation is presented in Table 1.

3.3Work-care conflict

Relatively few participants indicated that their caring role did not influence their work (n= 17; 14.8%). Data on how the remaining (n= 98) participants’ caring commitments affected their working situation is presented in Table 2. The majority of participants indicated that their work was affected by psychological and physical strain arising from their caring role, i.e., by being worried about the older person while at work (n= 72; 81.8%) and having less energy for work (n= 70; 77.8%). Fewer participants reported that their caring role affected logistical or practical aspects of their work, although in some cases these effects were reported by a substantial number (e.g., being disturbed by telephone calls about, and from, the care recipient during work, (n= 47; 53.4%); using vacation days (n= 42; 42.9%) and adapting their work (n= 32; 32.7%) to be able to provide care). The mean score for participants’ responses to the five items of the work-care conflict scale was 2.30 (SD = 0.85), close to the mid-point of the scale (note: additional analysis found no significant difference in work-care conflict between carers of home-based care-recipients, and carers of care-recipients in residential care). Table 2 also shows that about a third of participants was unable to access support groups for information, education, and carer networking during their day due to work, while two-fifths did not have time to use carer support due to work and care commitments.

3.4Valued forms of support and support areas

Table 3 shows, for each form of support, the proportion of participants who indicated the support was very/quite valuable, and the proportion who indicated the support was a little/not valuable. Nearly two-thirds (n= 15, 62.5%) of the forms of support was valued by 50% or more of the sample. The most valued form of support was ICT-mediated (web access to information, n= 86; 76.8%), but otherwise eight out of the ten ICT-mediated forms of support were valued by less than half the participants. By comparison, only one of the non-ICT forms of support was valued by under half the participants.

In terms of the most valued support areas, when considering ICT-mediated support Health and social care management was the most valued support area (63.6% of participants, as an average of two forms of support within this area), followed by Social, peer-to-peer, and emotional support (47.7% of participants, one form of support), then Carer relief and respite care, and care recipient independence (38.1% of participants, averaged across five forms of support combined). For non-ICT support, Health and social care management was the most valued support area (63.6% of participants, an average of four forms of support), followed by Social, peer-to-peer, and emotional support (58.7% of participants, averaged across six forms of support), then Financial support (55.2%, one form of support) and finally Carer relief and respite care, and care recipient independence (52.7%, an average of three forms of support). Thus, the ranking of support area by value was comparable between the two delivery modes, but the level of value was generally higher when the mode of delivery was non-ICT. It should also be acknowledged that the level of value for forms of support could range substantially within a given support area. Thus, within the Carer relief and respite care and care recipient independence support area (non-ICT) whereas nearly two-thirds of the sample valued relief from caring through the help of family and friends (n= 67; 63.9%), only just over a third of the sample (n= 37; 35.3%) valued help to rest and have time for themselves e.g., through respite care. However, the majority of participants (n= 66; 58.9%) valued having information about where and how to get time for self and rest.

3.5Received forms of support

Of the fourteen possible forms of support specified (including ‘other’), the median number received by participants was 2.0 (n= 110; IQR 1.0–3.0). For those forms of support for which data was gathered on both value and receipt (n= 11), we examined the proportion of participants who received a valued form of support. Thus, Fig. 1 shows, for those participants who responded that a form of support was valued, the proportion that was and the proportion that was not in receipt of that support. As would be anticipated given the low median value for the number of forms of support received, in general the proportion of participants in receipt of valued support was low. Averaged across the 11 forms of support, the proportion of participants in receipt of valued support was 21.7%.

Figure 1.

Proportion of participants receiving valued forms of support.

The proportion of participants in receipt of ICT-mediated support averaged at 16.9% across three forms of support. However, this average hides very low levels of receipt for two forms of web-based support, support for becoming informed and educated in caregiving (n= 6; 7.0%) and support for meeting peer carers (n= 4; 7.2%). By comparison, the receipt of technologies in the home of the care recipient (e.g., alarms and sensors) was considerably higher (n= 19; 36.5%).

The proportion of participants in receipt of non-ICT support averaged at 23.4% across eight forms of support. The lowest level of receipt was for support enabling rest or time for self, respite care, which was received by none of the 33 participants who valued that support, while the highest level of receipt was for relief received from family and friends (n= 32, 47.8%).

3.6Associations between value of support, caregiving characteristics, and work-care conflict

The bivariate associations between caregiving characteristics, work-care conflict, and value of support are presented in Table 4, separately for ICT-mediated and non-ICT based support.

4.1Perceived value of ICT-mediated and non-ICT forms of support

Forms of support in the area of health and social care management were the most valued by participants, regardless of whether they were ICT-mediated or not. This finding is given credence by research acknowledging that carers are particularly stressed when lacking information, and, with regard to working carers, when being obliged to organise, plan, and coordinate health and social care [31, 6]. The relatively lower level of perceived value for social, peer-to-peer, and emotional forms of support found in our study are reflected in research reporting low frequency of use and limited benefit of municipally organised support groups among Swedish working carers [13, 14, 15, 16, 17]. Further, the relatively lower perceived value of support for relief and respite, e.g., through respite care in our study, has been reflected in research on Swedish working daughters low service use when caring of older care recipients refusing such services when allowed to decide her/his own care arrangements [17].

When comparing the perceived value of ICT-mediated forms of support to non-ICT forms of support, we would argue that our results suggest that ICT-mediated support is complementary to non-ICT support. Firstly, having information (e.g., about support for the carer and the care recipient) and getting assistance with planning and coordination for health and social care management were highly valued. Using the web to access such information stands out in our results as the most valued support. Further, the ability to participate in care planning or doctor’s appointments from a distance via ICT may suggest an alternative means for working carers to stay informed, and may be valued when planning health and social care for their care recipients.

Secondly, a comparison of different modes for social and peer-to-peer support shows that web-based support for carers, allowing them to meet their peers, was only slightly less valued than an onsite, physical, meeting. This difference should be considered in the context of other findings of this study, i.e. the one-third of carers who indicated that they were unable to access support groups during the day due to work, the 40% who had no time to use carer support due to their work and care commitments, and the 25% that indicated a preference for anonymity when meeting other carers. ICT-mediated forms of support would clearly have application for such carers, as also described by Andersson et al. [21]. A preference for physical meetings with other carers rather than ICT-mediated meetings may be related to experiencing fewer digital skills or the view that communication is too impersonal when not involving synchronous interaction or physical contact [21].

Thirdly, between one-third and two-fifths of participants valued information on safety and prevention via remote monitoring technologies. Such technologies may assist working carers in providing everyday care while maintaining their employment (e.g., alarms, sensors as a form of in-house respite care), as shown in Mahoney et al. [26] and Jarrold and Yeandle [28], and offer a viable alternative to traditional institutional respite care or day care at a time when the withdrawal of the welfare state and the increase in numbers of older people is putting pressure on such services in many countries, including Sweden [15]. Nevertheless, we found a higher perceived value for being introduced to telecare and assistive technologies than for actually using these forms of support for safety and preventive monitoring. This discrepancy may be due to participants seeing such technologies as more for supporting the care-recipient’s independence and safety than being a support for themselves personally, i.e., as carers [19]. Alternatively, the discrepancy may indicate a general need for having such support available when required, as opposed to using such support when available, which is suggested as a key characteristic of support need among European family carers [17]. Finally, nearly one-quarter of our participants cared for someone living in a residential care facility, which may have reduced the perceived value of all forms of respite care in our sample.

4.2Receipt of valued support

The median number of different forms of support received by participants in our sample was two, which, when combined with the general low level of receipt of valued forms of support, suggests a poor level of received support for working carers. Though it cannot be stated for certain that something that has a high perceived value is in fact needed, still there is an argument to be made that our results suggest a high degree of unmet need for support in our participants. The forms of support that were considered in our study reflect those domains of support for which high levels of need have been found among carers in other research, namely, the need to be able to: (1) learn and access information about caring issues of relevance to their own situation; (2) talk about their situation and share their experiences with others; (3) have practical assistance in caring; and (4) have access to relief and respite care [32]. The low level of receipt of valued forms of support may indicate an underdeveloped municipal family support system, which fails to acknowledge the difficult situation of working carers and/or fails to provide information on the availability of support and how to access it [16, 13, 33].

4.3The influence of caring on working life

Our results confirm those from other studies, namely that caring influences working life [6, 34]. Our results show that the great majority of participants had been affected in their work situation by their caring role. Psychological and physical strain arising from their caring role (i.e., worry about the older person while at work and having less energy for work) was the most commonly reported effect; however, a substantial number of participants were also affected logistically or practically (e.g., adapting their work for caring, losing too many workdays, or finding new work).

Our analyses also demonstrated a strong link between higher levels of work-care conflict and a higher perceived value of support, for both ICT-mediated and non-ICT forms of support. The most common contributions to work-care conflict in this study, as measured via the items of the work-care conflict scale, were worry about the care recipient, having less energy for work, and being disturbed at work by phone calls about, and from, the care recipient. Interestingly, the strongest associations between work-care conflict and value of support were found for those forms of support that may address such contributing factors: namely, being introduced to available technologies such as alarms, and using such technologies for locating the care recipient (not for emergency alerts), especially during work hours. Our findings reflect those of other studies, where working carers experience peace of mind and emotional and physical respite from using such technologies [24, 26, 28, 35]. Also, using the described technologies during work hours in worksite programs may allow working carers to better balance care and work, thus reducing the negative impacts of caring [21], although a different research design than that of the present study would be required to test this claim. While the associations between work-care conflict and perceived value of support cannot be viewed as causal due to the cross-sectional nature of our study, it is logically easier to argue that the direction of influence is that conflict between care and work increases the perceived value of support rather than that an increased value for support might cause an increase in work-care conflict.

It is acknowledged that caring for someone with dementia, and for a spouse, can often strongly influence carers’ support needs [36]. It is rather surprising, therefore, that these characteristics of the caring situation had relatively few associations with the perceived value of different forms of support. One reason for this may be that the categories of dementia care vs. non-dementia care, and spouse vs. non-spouse, while defining characteristics of the caring role, are fairly unsophisticated descriptions and statistically insensitive measures of what is a highly complex care situation.

4.4Study strengths and weaknesses

The study’s main weakness relates to its convenience sample, which was obtained through two recruitment channels for which no response rate can be accurately determined. It is likely that our sample is therefore biased in unquantifiable ways, and may not be a good representation of our study population, which in itself offers only a limited representation of our target population. Our sample is representative in age distribution, marital status, weekly caring hours, relationship to the care recipient, and living and caregiving situation compared with Swedish national based surveys [4, 6]. We do however acknowledge that our convenience sample has a majority of females. While most carers are female, the combination of a relatively small sample size with a large majority of females limited the extent to which we could carry out analyses by gender. Women are known to experience more mental and physical strain than men and experience more difficulties in finding time for leisure activities and reduced ability to focus on their job [6]. This may explain why a higher proportion of our sample,18.3%, had reduced their working hours compared to 10% reported in a survey conducted by the Swedish National Board of Health and Welfare [4].

We acknowledge that not knowing if participants have experienced or were aware of the existence of certain forms of support limits what conclusions can be drawn. Nevertheless, due to the questionnaire items identifying forms of ICT and non-ICT support that are relatively generic, we believe that a lack of awareness of such forms of support is likely to be low and bias in the analysis minimal.

We acknowledge that the web-based survey format required that our participants had some familiarity with accessing and using the internet and this may have biased our sample in favour of those carers with such experience. However, it should be noted that our other recruitment channel offered a paper-based survey format making it possible for carers with less experience of the internet to participate.

We also acknowledge 11.6 –18.6% missing data in our questions on perceived value of support, received support, and work–care influence, which may have introduced bias into our analyses. Missing value analysis revealed patterns of not completing the whole questionnaire, i.e., giving up halfway through (n= 26), which may have to do with fatigue at finding the questionnaire too extensive, or having technical issues when completing the web-based questionnaire. However, an MCAR test [37] was non-significant, so the null hypothesis that our data are missing at random cannot be rejected. In our analyses, multiple bivariate tests were carried out, inflating the chance of type I error. However, we chose not to correct for this due to our relatively small sample for specific analyses also increasing the potential for type II error. Taking all the above information together, generalization of our findings to the wider population of working carers of older people must therefore be considered with due caution.

It might also be considered to what extent findings from a study concerning the perceived value of ICT-mediated and non-ICT forms of support among Swedish working carers might have implications for understanding the situation of working carers in other countries. After all, there are substantial differences between countries relating to welfare systems, working and caring cultures, and ICT use. Nevertheless, we would argue that our findings should be of interest to academic, policy, and practitioner communities internationally, as working carers in many different countries experience the dual pressure to work longer and care more. Where there is a lack of services and support for working carers, the price for many is the loss of paid work, impaired health, and an uncertain financial future, as a study comparing the United Kingdom and Sweden has shown [33]. Understanding how work-care conflict relates to the value of different forms of support among working carers, particularly ICT-mediated support, is critical from an international policy perspective given the reduction of welfare provision in many countries and the need to retain people of working age in the labour market [15].

4.5Conclusions and implications for policy and practice

Our study found high levels of work-care conflict among working carers, with conflict associated with higher levels of perceived value for support, while receipt of valued forms of support was low. Taken together our findings suggest a worrying level of unmet support needs in working carers which is inflated by the work-care conflict they experience. Our findings are perhaps indicative of underdeveloped carer support in municipalities. These findings are relevant for health and social care professionals, municipalities and formal care organisations, and for employers wanting to develop strategies to meet the support needs of working carers of older people. While on one level we found non-ICT forms of support to be more highly valued by working carers than ICT-mediated forms, reflecting on different elements of our results suggests that ICT-mediated and non-ICT forms of support are highly complementary. We suggest that future research should explore the effects of ICT on relief and respite care, and the use of ICT at the worksite and in the planning and coordination of care.