Evaluating an integrated respiratory service

What we did

We looked at the new respiratory service, treating people in the community, to see how well it worked. We compared what happens to patients in the area with the service to what happens to patients in other areas, looking at the number of patients admitted to hospital with COPD.

Using data from NHS Digital that is collected by hospitals, we compared admissions for the two years before the new service was set up with the year after the new service started, so from December 2014 to November 2017.

We also looked at data from 10 NHS regions that were similar to the region where the new service was set up. This was based on factors like income level, population density, age and ethnicity.

We focused on:

• Cases where COPD was the main reason for admission
• Emergency admissions

We also looked at how the hospital trust and community provider work together and support each other, which could be used as a blueprint for other healthcare organisations to follow. We used interviews to see how the services worked together from the point of view of staff and the managers who delivered the service. This part of the study ran from May 2018 to May 2019. We have kept the hospital trust and community provider anonymous to maintain the confidentiality of participants. We conducted interviews with:

• Doctors, nurses, physiotherapists and managers from each partner organisation
• Health care commissioners who shaped the partnership between the organisations

What we found and what this means

The number of hospital admissions was our main focus because we expected that the service would lead to a drop in admissions. We also wanted to find out if the new service could lead to lower readmissions and possibly shorter hospital stays.

However, we found little evidence that admission avoidance changed the rate of emergency hospital admissions, 30-day readmissions, or length of stay. Admissions had a seasonal pattern, varying between 180 and 330 per 10,000 COPD patients, and were higher during the winter months. The average length of stay was around five days and did not change much over the study period. Around one in 10 COPD patients were back in hospital within 30 days of their first stay.

Our interview study found that, generally, hospital and community staff liked the idea of working together but also identified areas where the partner organisations struggled to integrate, including:

• poor communication
• lack of shared IT infrastructure
• lack of trust

The interview data did suggest an association between lack of integration and the commissioning process (the way in which the contract between the two organisations was drawn up). We propose that future initiatives like this should establish a dialogue and understanding between the partner organisations before the contract is put in place. This would mean the role of each organisation would be clearer before they started working together, allowing relationships to develop and for both organisations to feel involved from the start. This should lead to higher levels of trust and confidence.

What next?

More work is needed to understand why we found little evidence of an improvement. It may be that a longer follow-up is needed to see if the new services reduce admissions and whether working across organisations in this way changes over time.