OBJECTIVE: To evaluate the role of delirium in the natural history of funct
ional recovery after hip fracture surgery, independent of prefracture statu
s.
DESIGN: Prospective cohort study.
SETTING: Orthopedic surgery service at a large academic tertiary hospital,
with follow-up extending into rehabilitation hospitals, nursing homes, and
the community.
PARTICIPANTS: One hundred twenty-six consenting subjects older than 65 year
s (mean age 79 +/- 8 years, 79% women) admitted emergently for surgical rep
air of hip fracture.
MEASUREMENTS: Detailed assessment at enrollment to ascertain prefracture st
atus through interviews with the patient and designated proxy and review of
the medical record. Interviews included administration of standardized ins
truments (Activities of Daily Living (ADL) Scale, Blessed Dementia Rating S
cale, Delirium Symptom Interview) and assessment of ambulation, and prefrac
ture living situation. Medical comorbidity, the nature of the hip fracture,
and the surgical repair were obtained from the medical record. All subject
s underwent daily interviews for the duration of the hospitalization, inclu
ding the Mini-Mental State Examination and Delirium Symptom Interview, and
delirium was diagnosed using the Confusion Assessment Methods algorithm. Pa
tients and proxies were recontacted 1 and 6 months after fracture, and unde
rwent interviews similar to those at enrollment to determine death, persist
ent delirium, decline in ADL function, decline in ambulation, or new nursin
g home placement.
RESULTS: Delirium occurred in 52/126 (41%) of patients, and persisted in 20
/52 (39%) at hospital discharge, 15/52 (32%) at 1 month, and 3/52 (6%) at 6
months. Patients aged 80 years or older, acid those with prefracture cogni
tive impairment, ADL functional impairment, and high medical comorbidity we
re more likely to develop delirium. However, after adjusting for these fact
ors, delirium was still significantly associated with outcomes indicative o
f poor functional recovery 1 month after hip Fracture: ADL decline (odds ra
tio (OR) = 2.6; 95% confidence interval (95% CI), 1.1- 6.1), decline in amb
ulation (OR = 2.6; 95% CI, 1.03-6.5), and death or new nursing home placeme
nt (OR = 3.0; 95% CI, 1.1-8.4). Patients whose delirium persisted at 1 mont
h had worse outcomes than those whose delirium had resolved.
CONCLUSIONS: Delirium is common, persistent, and independently associated w
ith poor functional recovery 1 month after hip fracture even after adjustin
g for prefracture frailty. Further research is necessary to identify the me
chanisms by which delirium contributes to poor functional recovery, and to
determine whether interventions designed to prevent or reduce delirium can
improve recovery after hip fracture.