Composition of the group of service-users
Ninety Four participants (64%female) , age range 65 to 99 years (mean age 82 years) from nine centres were recruited to the study and completed baseline measures. Table 1 provides a description of each day care centre. All those who the day care service manager / leader believed were eligible and wished to contact the researcher agreed to participate. The number of LTCs ranged from two to nine (mean of 4.3 LTCs). The most commonly reported LTCs were arthritis, heart disease, early to moderate dementia, stroke and mental health issues. Thirty two percent were married, 56% widowed, and 12% either separated, divorced or never married and 52% of participants lived alone. Over a third (37%) identified a carer who was a family member living with them; 27% had a carer who was a family member living elsewhere. More than a third (36%) of the sample lived in one of the 20% most deprived local authorities in England and Wales. On average older people travelled 3 miles to attend the day services, (range 0.1 mile to 20 miles in a rural area). 73 participants (78%) completed follow-up at 6 week and 12 weeks. ( Table 2 lists reasons for attrition by service type).
Baseline characteristics of service-users, and type and location of service
Table 3 reports the demographic profile of the older people attending day care services delivered by paid staff, paid and voluntary staff (‘blended’) and voluntary staff only.
Blended service participants were significantly older (mean age 84.7 vs 80.6, P=0.04). All participants described their ethnicity as white. A significantly higher proportion of the paid service group lived in one of the most socially deprived neighbourhoods 56% compared to 27% of participants attending other services, (P=0.02). Those attending voluntary services had a significantly greater distance to travel (mean 5.8 versus 2.2 miles) (P=0.001). A significantly lower proportion of those attending blended day care services responded positively to the De Jong item relating to having “plenty of people to rely on when having problems” (13% compared to 37% of all other participants, P=0.04). The number of LTCs reported at baseline was comparable across all service types (paid staff mean 4.4, blended 4, voluntary 4.7, P=0.39).
Service type, location and change in EQ-5D-3L self-reported health status
The proportions of older people of the three service-types (paid, blended, voluntary) reporting individual health problems on the EQ-5D-3L, along with mean number of problems and VAS score, are reported in Table 4. A significantly higher proportion of participants attending paid day care services reported a self-care problem at 6-week follow-up (46%, compared to 32% of blended service clients and only 10% of voluntary service users, P=0.02). However, there were no other statistically significant differences between the clients attending different services, at any time-point or in change in proportions/scores over time in terms of self-reported health status.
When comparing services, the mean number of problems reported by older people attending voluntary day care services declined between baseline and 12 weeks, whilst those attending blended and paid services increased. For the two domains of mobility and self-care, the proportion of participants reporting problems on these domains declined in those attending voluntary services but increased at blended and paid services between baseline and twelve weeks.
Anxiety and depression domains revealed a decline in both paid and voluntary services between baseline and 12 weeks. There was a small increase in anxiety and depression levels at blended services between baseline and 12 weeks. Pain was reported as increased between baseline and 12 weeks in paid and voluntary services. The mean VAS score for all services reported positive change in health and well-being from baseline to twelve weeks.
Service type, location and change in reported loneliness
Statistical significant differences were not reported in loneliness (Table 5). However, the change in mean total loneliness between baseline and 12 weeks reduced in Blended services and Voluntary services but increased in those attending Paid Staff services. In order to examine this further, the mean scores for emotional loneliness and social loneliness were compared by service group. The results for total loneliness showed the mean score for those attending Voluntary and Blended services reduced over 12 weeks, whilst the mean score for those attending Paid Staff services increased. However, when the social loneliness group means across the three services were analysed from baseline to 12 weeks, it could be seen that the group mean score reduced across all services. Therefore, the lack of reduced loneliness for those attending the Paid staff services appeared to be connected to levels of emotional loneliness rather than social loneliness
Likelihood of ‘any improvement’ in outcome
Table 6 illustrates the likelihood of improved outcomes for people attending blended services or voluntary services when compared with paid staff services. Older people attending a voluntary service were over twice as likely to experience a reduction in De Jong loneliness score between baseline and their final follow-up. Those receiving a service delivered by voluntary staff also had an increased likelihood of reporting reduced loneliness. Older people attending a ‘blended’ service had a raised likelihood of experiencing a reduction in the number of reported EQ5 health problems. The voluntary service group had a statistically significant increase in the likelihood of reporting fewer health problems over follow-up. In terms of reporting an improvement in the global health rating (VAS) from baseline, those attending voluntary services had a reduced likelihood however, users of blended services had raised odds of reporting a higher VAS rating