Health service use by patients with heart failure living in a community setting: a cross-sectional analysis in North West London

Summary

Overall health service utilisation was high. Almost everyone in the present study population used some kind of health service during the study period: outpatients (7 in 10), primary care (6 in 10 saw a GP), community services, especially nursing (2 in 10), and unplanned A&E visits (4 in 10). Community care use related to cardiology was low. Few patients used only primary care and ‘other’ services, which may reflect modest needs or a lack of community and primary care provision suitable for complex needs.

Patterns of health service utilisation depended on age and comorbidity but were highly heterogeneous. Younger patients with fewer comorbidities (clusters 1 and 2) had the lowest usage, which may partly indicate underutilisation and/or lack of access. For instance, those with infrequent GP consultations (clusters 1 and 3) were also more likely to be of mixed ethnicity and living in areas of higher deprivation and demographics known to be associated with poorer primary care access.^17,18 These patients also showed the highest levels of unknown values for patient variables (derived from GP data) and lower than average GP consultation rates, which could reflect both poor health management and low engagement of patients in their own health (they were also more likely to be smokers). The oldest and most likely to live in care homes (cluster 5) had the highest usage of emergency inpatient, A&E, and ‘other’ services, and had high levels of comorbidity, especially renal disease, and the highest mortality. Higher usage of care is expected in older patients with comorbidities,^19–21 but some use might be excessive and avoidable.^19,20 Health service use was high in the present population of adults with HF living in a community setting. However, overall, relatively low GP service use, which decreased over the study period, and high use of emergency and other unscheduled care in these vulnerable patients is of significant concern and may suggest challenges in access to primary care services. These findings warrant further investigation to ensure equity of access and appropriate integrated care provision for patients with HF.

Strengths and limitations

The authors used a linked dataset with near-complete population coverage for the region and employed both descriptive analysis and clustering algorithms to describe health service use by this highly heterogeneous population. The dataset is large and reflective of current medical practice, but the study has several limitations.

Electronic health records are not specifically intended for research, and coding is highly variable.²² Coding in some of the WSIC tables required additional cleaning and processing, which could introduce bias; however, a transparent coding methodology to mitigate this was provided. Moreover, coded data rely on recorded information, meaning that certain diseases or service components may have been underestimated, for example, cardiac rehabilitation, or that certain primary care data coding may have been affected by pay-for-performance schemes. The authors were unable to ascertain the reasons for the community care or mental health consultations as diagnosis coding was irregular. It is also likely that the fall in GP consultations during the study period was offset by more practice nurse contacts, which were not included in the dataset.

Lastly, though the dataset was based on adults living in a community setting from a large and ethnically diverse area in England, the findings may not be generalisable to the wider population of people with HF.

Comparison with existing literature

Few studies have attempted to quantify individual patterns of care in real-world settings beyond the hospital. Robertson et al described the burden of HF on the Australian healthcare system, but were only able to assess hospitalisation data,²³ as was the case for the present authors’ previous work.^24,25 Similarly, other studies have described a single dimensional aspect of health service use by the population of people with HF.^26,27 The present findings are consistent with these, showing that secondary care use is high,²³ participation in cardiac rehabilitation in the community is low,^26,28 and that requirement for personal care, such as nursing and homecare services, is relatively common.²⁷

An increase over time in most healthcare services use was observed in the present study, especially outpatient visits, but a surprising decrease of 11% in GP consultations. Furthermore, only 60% of patients had GP appointments during the study period, which contrasts with the national GP Patient Survey of 2019,²⁹ where 85% of patients reported having had a GP appointment in the past year. Potential explanations include the increasing workload and workforce pressures on primary care, changes in primary care practice with more frequent contacts with practice nurses and allied health professionals, a significant problem of access to care, and/or differences in case mix.

Another surprising finding is the apparently limited uptake of community cardiac rehabilitation. The National Audit of Cardiac Rehabilitation 2018 report²⁸ suggests that around half of eligible patients take up cardiac rehabilitation. The report did indicate significant regional variation. However, it seems likely that differences in coding of data are responsible for the very low uptake in this analysis; ‘rehabilitation’ events may be recorded elsewhere and currently unavailable in WSIC, and ‘community cardiology’ may also include HF nurse domiciliary care.

The authors further report low use of mental health and social care services by patients with HF, but whether this observed level is appropriate is unclear without further assessment.

Implications for research and practice

The present finding of increased secondary and urgent care service use, low GP appointment use, and apparently limited cardiac rehabilitation is of concern and suggests a lack of multidisciplinary HF care. National Institute for Health and Care Excellence guidelines recommend an MDT approach, but there is no standard definition besides who should be involved and what should be achieved.¹ Each local area has unique challenges and requires tailored solutions; research is needed to establish the nature, location, timing, and intensity of the support needed by patients with HF. In an ethnically diverse area with a relatively young population like NWL, where deprivation level and ethnicity may affect a person’s access to health care, creating a strong MDT embedded in primary care may be very pertinent. For example, practice nurses may target recently diagnosed patients in primary care, that is, younger patients with fewer comorbidities, on early education and management, which may include additional telephone and/or specialist community support for those with lower socioeconomic status. This, in conjunction with hospital-based solutions, like early supported discharge plans for older patients, who are the highest users of secondary care, may provide significant and long-term benefits for the NWL area. Local solutions like these have been shown not only to reduce utilisation of health services but also to improve patient wellbeing and result in large cost savings for the NHS.³⁰

Though the present data could not establish whether an MDT approach was implemented in the NWL area, it may well be that MDTs exist but their solutions are not translating into reduced secondary care use. Successful MDTs will require cooperation, coordination, and communication across health services. Reasons for ineffective multidisciplinary care could be posited through the following questions: is there an overarching coordinating unit for multidisciplinary care? Are the IT systems compatible for such care? Is information exchange readily available and safe? Is communication across settings both smooth and frequent? Is the approach sustainable? These questions illustrate how successful solutions will require sustained financial investments and the solid backing of all relevant stakeholders, and the sheer challenge of this may explain why many MDTs have had only neutral effects.²