Castlefields Health Centre was the first practice to demonstrate that social workers for the elderly and district nurses, working together, orientated around practice populations, can reduce the pressure on acute hospitals without increasing the pressure on primary, community and social services.

Castlefields is situated in an electoral ward which ranks in the 2% most deprived in the UK. The practice population was 11,900, of whom 9.8% were over 65. The practice had noticed that social interventions were at least equally as valuable as medical ones in the older population, yet social services and health operated in separate organisations without communicating well without each other, let alone integrating care. Social services were also aware of the difficulties and were willing to try something different.

The model of care

A full-time social worker was appointed to work with the practice for a 12 month pilot period in August 1999. Funding was provided equally from social services, the practice and the health authority. One of the existing full time district nurses was selected from the team to spend half her time working jointly with the social worker, to form a new service for the practice. To cover the time she spent doing this the practice offered £10,000 from their PMS budget to provide backfill.

Referrals to the social services older people's team and a substantial number of district nurse referrals were re-routed to the new service. The practice-based social worker and the district nurse carried out joint assessments and joint packages of care were put in place to meet identified needs of the older people. The social worker could provide resource up to £200 a week without going through a social services panel.

The local hospital faxed the practice daily with a list of patients over 65 currently admitted. The social worker and district nurse went to the hospital to work with the hospital staff in discharge planning. Another element was to identify high-risk people not already known to the team so that anticipatory assessment and care packages could be done. A list was generated of patients with high use of health or social services or impairment in activities of living. The intention was to avoid missing a needy individual.

Care outcomes

The team managed a total of 409 patients. Two hundred and fifty three were active referrals via the practice or social services; 108 were new cases; and 48 individuals were proactively assessed before referral or admission. One hundred and fifty one were visited in the hospital setting

Social work assessments were carried out on the same day of referral for 97% (245 out of 253). Previously, a ‘routine' social services assessment took six weeks or more. The model facilitated quick decisions and packages of care were put in place quickly. In year three, the length of time for social services assessment and input reverted to the baseline.

The nurse and social worker visited the local hospital wards to begin discharge planning within two working days in most cases. Previously, discharge planning began towards the end of the in-patient stay, if at all.

Of the forty-eight ‘high-risk' individuals, 41 were maintained at home, three entered residential or nursing care, two died and only two required an acute admission to hospital (each only once). Thus, only 4.2% of this group had an acute admission, compared to 18.1% of the practices' over-65 population. The 48 people identified as having high risk did do well. However, the ‘case management' of these patients was a relatively small part of a whole systems approach and represented only 12% of the workload.

During the project, the practice's social care for older people remained within budget, including nursing and residential home costs. In contrast, the whole of the borough council's social services budget for older people was overspent. The district nurse team were given resources to provide back-fill. However, they found that the burden on their workload fell and, in the end, no backfill was required.

There was no evidence that the GPs had to work harder as a result of patients occupying fewer hospital beds as their consultations remained steady and home visits fell by 17%. There was no evidence of an increased burden on district nursing, as their workload remained manageable without backfill for the nominated nurse.

Figure 1: Total number of acute medical admissions per 1,000 population for over-65s

(Click on the diagram to view)

Figure 1 shows the project practice began from a higher starting point (year 1), commensurate with its higher level of deprivation than the rest of the town. Admissions fell by 15%, whereas the rest of the town rose by 9%, during the project year (year 2). After the social worker left (year 3) some of the reduction was lost.

Figure 2: Over 65s average length of stay in hospital

(Click on the diagram to view)

Figure 2 shows the average length of stay started at a lower level for the project practice (year 1) and fell by 31% (compared with 6% for the rest of the town) during the project year. Once the social worker left the practice much of the reduction was lost (year 3).

Figure 3: Over 65s bed days per 1000 population

(Click on the diagram to view)

Figure three shows the combined effect of reducing the number of admissions by 15% and the average length of stay by 31% had a dramatic effect on the total bed days during the project year (year 2). The project practice's patients' bed occupancy fell by 41% compared to a 2.25% rise for the town. Almost all of this was lost in year 3 when the social worker was no longer with the practice.