Abstract
For some years, there has been vast international interest in creating models for joint efforts between geriatricians and orthopedic surgeons. We present data from two such models.
For the first time in Denmark, the Department of Orthopedics Bispebjerg University Hospital (BUP) recruited two full-time geriatricians in September 2009. They were assigned an independent unit meant for severely ill orthopaedic patients with high comorbidity and polypharmacy. These two geriatricians had, during the previous two years, consulted another orthopaedic department at Gentofte University Hospital (GUH) in a neighbouring community three times a week. The aim of their intervention was then to optimize treatment for comorbidity, to clarify indication of acute fall-assessment, osteoporosis diagnosis and treatment, presence of delirium and dementia.
Methods
A total of 1344 hip fracture patients (age 70 years) divided into three populations were included in this study. Mortality data were collected from the Danish Civil Registry.
Population 1 (P1), n = 645 was included at GUH from January 1, 2006 to December 31, 2007. During the entire period, the patients had access to a senior consultant in geriatric medicine three times a week. In this population, the majority of patients were assessed for dementia (n = 636), delirium (n = 627) and Barthel Index (Barthel100) at admission (n = 394).
Population 2 (P2), n = 381 included at BUH from September 1, 2009 until July 8, 2010 with orthogeriatric access.
Population 3 (P3), n = 318 were included at BUH from September 28, 2008 until August 31, 2009 with no orthogeriatric access.
Age mean (SD): P1 84.7 (6.8), P2 85.5 (7.3), P3 85.3 (14.3) P = 0.1(ANOVA) Sex ratio: females/males: P1 0.73/0.27, P2 0.80/0.20, P3 0.75/0.25 P = 0.09 Chi square).
Results
In-hospital mortality rate: P1 4.8%, P2 6.3%, P3 9.1% P = 0.03 (Chi square).
Three month mortality: In P1 dementia, delirium and Barthel Index (below 50 versus above 50) were all strong predictors: No dementia: 53/383 (13.8%) versus dementia present 68/253 (26.9%) died, P = 0001 (log-rank test).
No delirium 69/456 (15.1%) versus delirium present 47/171 (27.5%) died F = 0.0004 (log-rank test) Barthel Index 50 38/372 (10.2%) versus Barthel<50 7/22 (31.8%) died P = 0.0004
Conclusion
This paper reports data from two different models with orthogeriatric service.
Our data present delirium, dementia and Barthel Index to be very strong predictors for three month mortality (P1). Despite the time needed to implement a new orthogeriatric unit and the fact that the geriatricians only assessed a proportion of hip fracture patients, in-hospital mortality was reduced significantly for the total hip-fracture population (P2) within the first 10 months.
