Transforming health care in nursing homes: An evaluation of a dedicated primary care service in outer east London

The study indicates that providing residents of care homes with increased GP access and continuity of primary care more generally may help to reduce hospital admissions.

Background

People residing in care homes generally have more complex health care needs than the average older adult, yet their experiences of health care services are of variable quality (British Geriatrics Society, 2012). In particular, the provision of primary care to care homes is too often reactive, with little in the way of continuity (Robbins and others, 2013). As a result, residents often have a poor experience, receive sub-optimal medication and have frequent unplanned admissions to hospital. In fact, care home residents have between 40% and 50% more emergency admissions than the general population aged over 75 (Smith and others, 2015).

Because care homes often provide services that fall under both the NHS and social care systems, the commissioning of those services is too often considered in isolation and not as part of overall joined-up service provision in a locality (Humphries and others, 2016).

Where residents within a single nursing home are registered with several different general practitioners (GPs), it can be difficult for staff to establish relationships with GPs in order to gain timely access to advice and to proactively manage risks. This can lead to greater use of emergency services and inefficient use of medicines.

The Nuffield Trust was commissioned by the Barking and Dagenham, Havering and Redbridge Clinical Commissioning Groups (BHR CCGs) to evaluate a new primary care service (Health 1000) that was being piloted in four nursing homes in the London Borough of Havering who previously had difficulty accessing GP services. Among the features of this service were:

  • the assignment of a single GP practice to all residents in a home
  • access to healthcare professionals with expertise in caring for older people with complex needs
  • extended access beyond normal GP hours; care guidance to nursing home staff
  • improved medicines management
  • new approaches for managing people who are at the end of life.

The aims of the evaluation were to assess the impact of the new service on staff and the use of hospital services.

Methods

We employed a mixed methods approach, incorporating retrospective cohort analysis and semi-structured interviews conducted with staff.

For the cohort analysis, we analysed changes in use of hospital services by 431 residents of the four nursing homes selected for the intervention, before and after registration with the new service. These were compared with outcomes over the same period for 1,495 residents of a set of similar nursing homes that were chosen as controls. Hospital services included inpatient spells and bed days, attendance at accident and emergency (A&E) departments, and outpatient appointments. These were further evaluated at the end of life.

For the qualitative element of the evaluation, we undertook interviews with 14 nursing home staff, managers and GPs, which we coded and thematically analysed. These interviews offered an insight into the impact of the new service on staff members’ working lives and how they manage risk.

Experiences before Health 1000

The majority of the nursing home staff we spoke to described a poor experience of working with primary care before the introduction of the Health 1000 service, both in terms of their ability to access primary care services on behalf of residents and the decision-making support they themselves received. This appears to have been driving risk-averse behaviours within the homes, such as immediately sending residents to A&E.

Staff suggested it had also led to behaviours which could be construed as an inefficient use of resources: for instance, residents being sent to hospital unnecessarily and nursing home staff spending large amounts of time on administrative tasks because they had to liaise with multiple GPs. This could also impact on quality of care, for example in relation to how medications were managed. Despite this, nursing home staff recognised that local GPs were under pressure and sympathised with their situation.

Impact of the new service on staff

The most frequent improvements since the new services mentioned by nursing home staff related to access: to the GPs themselves and to clinical advice, both for triaging and ongoing learning support. There was a view that GPs spending more time in the homes face-to-face with residents had improved care quality. Also, several staff indicated that their own approach to risk sharing had improved as a result of having better access to GPs.

There were also noted improvements in medicines management, enabling quicker access to medicines. Reductions in polypharmacy (concurrent use of multiple medications by a patient) and waste were achieved through medicines reviews and deleting unnecessary repeat prescriptions. However, staff mentioned risks associated with potential discrepancies in notes, where Health 1000 and the nursing homes operate different systems. There were also difficulties in implementing Health 1000 policies where these conflicted with nursing homes’ own policies, for example, in terms of the paperwork and checklists that staff are required to complete. As nursing homes are usually privately run and are often sub-units of large chains, there appears to be a potential for conflict where parent companies’ protocols diverge from the Health 1000 operating approach. For example, one GP expressed frustration that they were required to complete forms for the nursing home company that were not directly relevant to the patient’s care.

Impact on use of hospital services

The improvements in the service have coincided with a reduction in emergency admission of 36%, compared to a 4% reduction observed among the comparator homes over the same period. After accounting for differences in case mix, the marginal reduction associated with the Health 1000 service was 35% (95% confidence interval, 6% to 55%). Also, total bed days following emergency admission fell by 53% compared to no reduction among the comparator homes. After accounting for differences in case mix, the marginal creduction in bed days was 50% (95% confidence interval, 10% to 72%). Reductions in emergency admissions were most notable during the last three months of a person’s life.

There were also reductions in A&E attendance associated with Health 1000, but these were not statistically significant. Outpatient attendance increased by 45%, although this was strongly influenced by a very small number of individuals from one of the nursing homes attending multiple times for anti-coagulation.

Implications

This study suggests that Health 1000 has had a positive impact on the working experience of staff, has reduced emergency admissions and has potential to improve the quality of care for nursing home residents.

However, the four nursing homes selected for the intervention were chosen because they had difficulties with primary care access. This may mean that there has been greater potential for change within these homes, and an equivalent scale of change may not be seen in nursing homes where there are fewer problems.

Our findings from this evaluation are broadly consistent with findings from the limited number of other studies in this area and with emerging findings from NHS England’s care home vanguard sites, most of which are demonstrating mixed impacts; highlighting the complexities of implementation.

Whether the results of this scheme can be replicated elsewhere is likely to depend on the context in which it is being implemented. Success will depend on the quality and continuity of relationships between the GPs and nursing homes, and it will take time to establish such relationships where they do not already exist. This will be more difficult in areas where there are staffing shortages or high numbers of temporary staff, and may be further exacerbated by provider instability in the nursing homes market. It is also important to note that Health 1000 was driven forward by a group of committed individuals who were instrumental in designing and implementing the service.

The ability of an initiative to overcome the challenges that arise from operating across both the NHS and social care will be a crucial determinant of success. Successful implementation may be easier where there is a system-wide commitment to integrated working, in which nursing homes are seen to be an integral part.

Key points

  • The Health 1000 service provides proactive primary care, operated from a single practice, to four nursing homes in the London Borough of Havering who previously had difficulty accessing GP services.
  • The service includes expertise in caring for older people with complex health needs, and features include enabling access to primary care for extended hours, providing advice and support to nursing home staff, medicine reviews, and new approaches for managing people at the end of life.
  • Since registration with the Health 1000 service, emergency inpatient admissions fell by 36%, and emergency bed days by 53%. After adjusting for case mix, these are all significantly larger reductions than observed among a control group of similar nursing homes in the area, although there is a reasonably large margin of error. The biggest reductions occurred towards the end of a person’s life.
  • There were significant increases in outpatient appointments since registration with Health 1000, but this seemed to be influenced by a very small number of residents in one of the homes attending multiple times for anti-coagulation.
  • Previous lack of timely access to primary care appeared to have been driving risk-averse behaviour at the homes, which meant that staff were more likely to send residents to A&E than to call the GP.
  • Staff reported feeling more supported and more confident in managing risk as a result of having quick access to GP advice via phone and the certainty of a regular weekly visit from a GP who knew the home, staff and residents.
  • Staff, managers and GPs said it was important to develop a trusting relationship between professionals, as that provided confidence, support and a sense of shared responsibility.
  • Staff and managers observed improvements in the proactive management of medications and in end-of-life planning, but said there was still room for improvement in both these areas.

References cited

British Geriatrics Society (2012) Failing the frail: A chaotic approach to commissioning healthcare services for care homes: Analysis of data collected by CQC about PCT support for the healthcare of older people living within nursing and residential care homes

Humphries R, Thorlby R, Holder H, Hall P and Charles A (2016) Social care for older people: Home truths. The King’s Fund

Robbins I, Gordon A, Dyas J, Logan P and Gladman J (2013) Explaining the barriers to and tensions in delivering effective healthcare in UK care homes: a qualitative study. BMJ Open 3, e003178. https://doi.org/10.1136/bmjopen-2013-003178

Smith P, Sherlaw-Johnson C, Ariti C and Bardsley M (2015) Focus on: hospital admissions from care homes. QualityWatch: Nuffield Trust and The Health Foundation 

Suggested citation

Sherlaw-Johnson C, Crump H, Curry N, Paddison C and Meaker R (2018) Transforming health care in nursing homes: An evaluation of a dedicated primary care service in outer east London. Research report, Nuffield Trust.

Appears in

Dedicated GP support for nursing homes sees significant drop in emergency admissions 04/04/2018

Press release

A new scheme piloted in four nursing homes has resulted in a 36% reduction in emergency admissions to hospital, with the biggest reductions...

Read more