Patients rehospitalized with complications after major surgery are 26% more likely to survive if they return to the hospital where they had their operation compared to those readmitted to a different hospital, according to a national study involving over 9 million Medicare patients in the USA, published in The Lancet.
The findings stand in contrast to current health policies that aim to regionalise major surgical procedures into high volume centres of excellence.
"With up to one in four patients rehospitalized following complex surgery, our results could potentially translate into thousands of lives saved every year in the USA alone if patients returned to the hospital where they had the procedure and received care from their original surgical team" , explains lead author Dr Benjamin Brooke from the University of Utah School of Medicine in Salt Lake City, USA.
Brooke and colleagues examined data from more than 9 million (9440503) Medicare beneficiaries in the USA between January, 2001 and November, 2011, who underwent 12 common high-risk operations . They used different statistical models to investigate the association between readmission destination (ie, the index hospital where the procedure took place vs a different hospital) and risk of death within 90 days of the procedure. This included instrumental variable analyses to account for potential unmeasured bias. This approach effectively allows researchers to simulate randomisation of patients who were readmitted following surgery and to generate unbiased estimates.
The number of patients rehospitalised with complications within 30 days of their operation ranged from 5.6% (154203 patients) of knee replacement patients, to 22% (3665) of oesophagectomy (surgical removal of all or part of the oesophagus) patients. Of patients rehospitalized, those readmitted to the same hospital ranged from two-thirds (186336) after coronary artery bypass grafting, to 83% (142142) following colectomy (surgical removal of all or part of the colon).
The researchers found that patients readmitted to the same hospital were 26% less likely to die within 90 days than those readmitted to a different hospital, even after taking into account measures of surgical quality that can affect mortality such as hospital size, teaching status, and volume of procedures. When confounding was controlled with instrumental variable analysis, patients returning to the index hospital were 8% less likely to die than those returning to a different hospital.
According to Dr Brooke, "Patients increasingly travel long distances to have their operations done at hospitals that are recognised as providing high-quality care or because of lower costs for health insurers. The assumption has been that if patients need readmission for complications they can seek care at local hospitals without compromising outcomes. However, our findings suggest that maintaining continuity of care when readmissions occur is a more important predictor of survival than other established surgical quality measures such as hospital procedure volume and needs to be considered in the trade-offs when choosing a hospital for surgery." 
Dr Justin Dimick and Dr David Miller from the University of Michigan, Ann Arbor, Michigan, USA, authors of a linked Comment, say, "After many years of evidence supporting the advantages of regionalizing complex surgery, Brooke and colleagues provide the first definitive empirical evidence that travelling to a remote hospital for surgery may be potentially life threatening. These findings have important implications for existing selective referral and centre of excellence programmes. If patients need to travel long distances to receive care, every effort should be made to ensure that the post-surgical patient is readmitted to the hospital where they had surgery." 
NOTES TO EDITORS:
 Quotes direct from authors and cannot be found in text of Article / Comment.
 Open abdominal aortic aneurysm repair, infrainguinal arterial bypass, aortobifemoral bypass, coronary artery bypass surgery, oesophagectomy (surgical removal of all or part of the oesophagus), colectomy (surgical removal of all or part of the colon), pancreatectomy (surgery to remove all or part of the pancreas), cholecystectomy (surgical removal of the gallbladder), ventral hernia repair, craniotomy, hip replacement, or knee replacement.