This was a retrospective study with two historical cohorts. The first cohort served as control group. The patients in the second cohort constituted the intervention group.
The medical record of 495 patients age 65 years or older, with a primary diagnosis of hip fracture (femoral neck, intertrochanteric), who were consecutively admitted to the Orthopedic Department of Aarhus University Hospital, in two periods between July 1st to December 31st, in 2000 (control group), and July 1st to December 31st, in 2003 (intervention group), were reviewed. While the intervention group received extended treatment by a team of geriatricians and orthopedic surgeons, the control group only received traditional rehabilitation in the orthopedic ward. All patients in both groups were admitted from the same area of the city, and those who were transferred to other hospitals, after the surgery, were excluded from the study.
The two year planning, development, and implementation of the geriatric multidisciplinary team in Aarhus University Hospital, Orthopedic surgery ward, ended in 2003. This explains the gap in patient recruitment for control and the intervention group during the period between year 2000 and 2003.
A geriatric team (GO-team), consisting of a geriatrician, a physiotherapist and a nurse with geriatric expertise, provided full-time geriatric and orthopedic care during daytime on weekdays. On weekends or off days the GO-team usually was not available. The GO-team included the following components in the care of their patients: conducting initial physical and mental screening and evaluation; providing continuity of care including geriatric focused care, and early discharge planning. Geriatricians and orthopedic care providers shared responsibility for patients’ care throughout the hospitalization period. The GO-team conducted rounds with the staff in the orthopedic ward and provided written report in patients’ medical charts/records.
One of the main goals of the GO-team was to minimize time to surgery and to avoid risk of iatrogenic illness. Information on the physical and mental status of the patient before hospitalization was obtained from the homecare system, the general practitioner, and the relatives, in order to elucidate complex social and medical conditions. To prevent future “fall-episodes”, patients were evaluated for diagnosis of osteoporosis and subsequently applied preventive measures. Of measures that GO-team took to prevent delirium or iatrogenic illness were to facilitate early discharge from the orthopedic ward; i.e., one or two days after the surgery. Patients who lived at home or lived at handicap-friendly housing were either: 1) escorted directly from the orthopedic ward to their homes with follow-up home-visits by a physiotherapist, or 2) transferred to a geriatric rehabilitation unit with multidisciplinary expertise. The choice between the two options depended on required personal and medical assistance. After the surgery, patients in the intervention group received training in ways to increase their mobility in activities of daily living (ADL). The GO-team also supervised patients’ nutritional in-take and managed patients’ concurrent medical problems including pain management and fluid therapy, which are among the most important factors in management of orthopedic elderly patients.
Patients admitted from nursing homes were returned to the nursing home directly from the orthopedic ward, without receiving geriatric follow-up since these patients had regular access to medical care by general practitioners, and the 24-hour staff in the nursing homes was close to the residents.
Data collection and main outcome measurement
Data regarding gender, age, housing, cause of fall, osteoporosis prophylaxis and iron treatment before and after admission, time to surgery, number of blood transfusions, and re-operations were collected from patient records. B-hemoglobin measurements at the time of the admission and three to six months post admission were obtained from the hospital electronic laboratory information system. Variations in the levels of hemoglobin between the time of admission and three to six months post admission, changes in discharge destination, three-month readmission rate, readmission rate within three to six months, new fractures within two years, in-hospital mortality and three-month mortality were collected as patient outcomes.
Length of stay (LOS) was the main outcome variable measuring hospital efficiency. Measurements for osteoporosis prophylaxis were an additional process outcome on hospital efficiency (). Data on LOS, readmissions, reoperations, and deaths were obtained from the Danish Health Data Bank (“Sundhedsdatabanken”). Information regarding new fractures was obtained from the emergency department, out-patient clinic, and from the Patient Registration System (Figure 1
). All data were collected retrospectively by a nurse specialist.
Table 1:Baseline characteristics
The differences in clinical characteristics between the two groups were tested by χ2 tests for proportions of categorical variables and presented in percentages and p-values with a significance level of 5%. In calculating differences in the continuous baseline characteristics, unpaired t-tests were used after a test of equal standard deviation in the two groups. LOS was tested for normal distribution and a logarithmic transformation was made to perform a multiple linear regression. Logistic regression was used for variables with dichotomized outcomes and adjusted for relevant potential effect variables. Then the residual-assumption in the regression model was checked and accepted. In analyzing the incidence of new fractures within two years, “lost to follow-up” was taken into account.19
Data were analyzed in Intercooled Stata.9,1