It is now accepted that surgery is best carried out in centres that have sufficient throughput to maintain the skills and competence of surgical teams. Such ‘centres of excellence’ are shown to result in improved outcomes and reduced costs. Apart from inconvenience for patients in travelling longer distances to regional centres, there appear to be few downsides. However, concerns have been raised about continuity of care, especially when the patient travels far from home. Until recently this concern has not been supported by evidence.

In a recent study in the Lancet, Brooke et al¹ provide the first empirical evidence to suggest that surgery in a remote hospital could potentially be life threatening for patients. In a study of US Medicare patients who had surgery and were later readmitted to hospital due to complications, those who were readmitted to the same hospital where they had the procedure (ie. the index hospital) had a lower risk of 90-day mortality than did those readmitted to a different hospital. The investigators studied more than nine million patients undergoing one of 12 complex surgical procedures. Patients who were readmitted to their index hospital were 26% less likely to die after surgery (odds ratio 0·74, 95% CI 0·66 to 0·83) than were those readmitted to other hospitals.

There are some reservations about the design of the study, such as patients readmitted to index hospitals might have been different to those readmitted to other hospitals in some way that explains their improved outcomes (eg. they might have been healthier or more likely to seek care at the index hospital). The study did use a sophisticated instrumental variable approach to the analysis which attempts to account for unmeasured variables and generate unbiased results. This result suggests that some, but not all, of the difference in outcomes between readmission to the index hospital versus other hospitals is caused by unmeasured differences in patients’ characteristics.

There are several reasons why patients might fare better when admitted to the hospital that performed their operation. First, index hospitals might simply have better systems for effective management of complications than do non-index hospitals. Patients readmitted after surgery almost always have a postoperative complication, either medical or surgical. Procedure volume not only improves primary outcomes but also the ability to treat complications effectively. Second, the index-hospital staff might have more first-hand knowledge of the details of the specific patient, their surgery and postoperative course before discharge.

A final reason could be the lower barriers for access to care at the index hospital than at a non-index hospital. Patients who were readmitted to the index hospital were admitted faster than those readmitted to non-index hospitals. Early presentation might lead to more timely management of surgical complications and improved outcomes. Patients could have difficulty accessing care at non-index hospitals, where they do not have a pre-existing relationship with a surgeon, thereby delaying presentation and treatment. Moreover, the team at the index hospital might have a stronger sense of duty towards the patient. This last observation has important implications for groups of hospitals that are designed to provide a balance between convenience and scale. If patient care is to be shared between local and regional sites then the level of communication, commitment and primacy of the patient must also be spread across those sites. Common governance structures are vital; old institutional loyalties must be set aside. 

Reference

1. Brooke BS, Goodney PP, Kraiss LW et al. Readmission destination and risk of mortality after major surgery: an observational cohort study. Lancet 2015(Aug 29); 386(9996): 884-95