Key points

  • This briefing presents the findings of an evaluation into the early effects of introducing integrated care teams (ICTs) in North East Hampshire and Farnham (NEHF), as one part of the Happy, Healthy, at Home primary and acute care system vanguard. Through this evaluation, the Improvement Analytics Unit sought to provide the vanguard with evidence to help inform the development of its services as part of its commitment to learning and continuous improvement.
  • In NEHF, ICTs are multidisciplinary teams that meet weekly to develop a care plan for each of their patients and provide more coordinated care. During the study period, the main objectives of the ICTs were to reach patients with the highest need and at highest risk of going into crisis, and – by providing more coordinated care – to improve patients’ health, health confidence, experience and wellbeing and reduce A&E attendances and emergency admissions. Patients in NEHF were referred to ICTs by their GP and other health care workers, who selected patients they considered to have highest need and be at highest risk of going into crisis and who would most benefit from a multi-disciplinary approach.
  • The Improvement Analytics Unit examined the impact of ICTs on the hospital use of 774 patients referred to an ICT between July 2015 (when the ICT was first introduced) and May 2017. We could not examine the impact of the ICT on other outcomes, such as health and wellbeing, due to the limitations of NHS data sets. The evaluation assessed the impact of ICTs over and above the effect of other services available in the area, including those developed as part of the vanguard programme. We examined the hospital use of patients for an average of 7 months following referral.
  • Compared with other adult patients registered with a GP in the North East Hampshire and Farnham area who were not referred to an ICT, the ICT patients were on average older (81 vs 54 years), had more emergency admissions in the year prior to referral (1.93 vs 0.32) and had more health conditions, particularly those linked to frailty (1.3 vs 0.2). We therefore selected a matched subgroup of non-ICT patients to use as the basis of our comparisons.
  • Patients in the matched subgroup were aged 79 (vs 81 for ICT patients), had 1.67 emergency admissions in the prior year (vs 1.93) and 1.2 health conditions on average (vs 1.3). Looking at individual health conditions, the profile was similar to the ICT group, although there were some slight differences. We used multivariable regression analysis when comparing the hospital use of the two groups, with the aim of eliminating (“adjusting for”) the impact of these differences in baseline characteristics.
  • After statistically adjusting for differences between the two groups such as age, prior admissions and health conditions, ICT patients experienced emergency admissions 43% more often than matched control patients (95% confidence interval: 23-67% more often), which is equivalent to an average of 0.53 more emergency admissions per person per year after referral (95% confidence interval: 0.28-0.82 more).
  • ICT patients also attended A&E 33% more often than the matched control patients in the period following referral to the ICT (95% confidence interval: 16-54% more often), which is equivalent to an average of 0.54 more A&E attendances per person per year (95% confidence interval: 0.26-0.89 more). In contrast, patients referred to an ICT were admitted electively 24% less often than the matched control patients (95% confidence interval: 2-41% lower), equivalent to 0.15 less elective admissions per person per year (95% confidence interval: 0.01-0.25 less).
  • When interpreting these findings, it is important to remember that the ICT and the matched control group might have differed in unobserved ways (for example, in their degree of family support, social isolation or severity of or ability to manage their health conditions) and we could not adjust for these statistically as we did for age, prior admissions and health conditions.
  • In the absence of a randomised controlled trial, we cannot be sure whether the higher rates of emergency hospital use could be explained by unobserved differences in the characteristics of the two groups. However, it seems unlikely that unobserved differences could explain the much higher emergency admission rates amongst the ICT patients. Furthermore it is very unlikely that any such differences could hide a decrease in hospital use. Therefore, we interpret the findings to show that the ICTs did not reduce A&E attendances and emergency admissions in the early stages (first 23 months) of its implementation and may even have led to increases. Other evaluations of ICTs and similar interventions have reached similar conclusions.,,,
  • It is plausible that the greater emergency hospital use of ICT patients was a result of ICTs identifying urgent needs for health care that might otherwise have remained unmet or only been identified later. Another explanation is that the ICTs led to patients being more aware of their health needs, which in turn led to patients attending A&E and being admitted. The lower levels of elective admission among ICT patients might have arisen if some patients were referred to ICTs at a time when they and their doctors had decided to stop curative care such as cancer treatment.
  • It is important to remember that this evaluation was conducted less than two years after implementation of the ICTs and does not evaluate the effect of the ICTs after June 2017. Research has highlighted that implementing complex interventions needs time to take effect. This is so that teams introducing these changes can be supported to work through implementation challenges and learn from experience and evaluation. It may be that more time is needed for the benefits of ICTs on hospital use to be seen. It is also important to remember that this evaluation did not examine the impact of the ICTs in improving the coordination of care or improving patients’ health, health confidence, experience of care and quality of life. However, the value of integrated care teams might relate to improving these other areas rather than reducing emergency hospital admissions. There is also a need to monitor the ongoing impact of the ICTs in NEHF, especially since the ICTs have continued to evolve, and have already been adapted in response to learning.
Next