Long acting benzodiazepine and pregabalin/gabapentin use was associated
with increased odds for postoperative delirium in patients undergoing hip
and knee replacements.
Age stratified analysis confirmed these findings, while indicating that
particularly among those aged ≥65 higher opioid use was associated with
decreased odds for postoperative delirium.
Given that postoperative delirium is linked to morbidity and increased
resource utilization, identification and subsequent intervention to address
modifiable risk factors is crucial.
Study design: retrospective population based analysis.
Population: 564,226 cases of total hip and 1,130,569
cases of total knee arthroplasties in >600 hospitals;
Premier Perspective Healthcare Database 2006-2016
(national claims data), IRB approval
Main outcome: postoperative delirium
Risk Factors: perioperative opioid dose prescription
(high: >75th percentile, medium: 25-75th percentile and
low: <25th percentile, based on quartiles of oral
morphine equivalents), type of anesthesia (general,
neuraxial, both, other/missing), benzodiazepine use
(long acting, short acting, both), meperidine, hypnotics,
ketamine, corticosteroids, and pregabalin/gabapentin.
Analysis: Multilevel models measured associations
between risk factors and postoperative delirium.
In addition to assessing effects in the full study cohort
we assessed effects of any of the potential risk factors
separately for those aged <65 and ≥65 by introducing
interaction terms in the model. Odds ratios (OR) and
95% confidence intervals (CI) are reported.
With the current aging US population and an ever
increasing surgical volume the burden of postoperative
delirium is likely to increase. This appears particularly
true for hip and knee replacements which are among the
most commonly performed inpatient surgical
procedures. In the absence of large scale real-world data
we therefore sought to study potential modifiable
perioperative risk factors for postoperative delirium
among patients undergoing hip and knee replacements.
Postoperative delirium prevalence: 0.9% (n=5,119)/ 1.0% (n=11,862) among
patients with hip/ knee replacements, respectively.
Use of benzodiazepines and pregabalin/gabapentin was associated with
altered odds for postoperative delirium, after adjustment for relevant
covariates. Knee cohort: long acting OR 1.79 CI 1.51-2.14, short/long acting
combined OR 1.44 CI 1.27-1.63 (both compared to no benzodiazepine use);
pregabalin/gabapentin (compared to no use) OR 1.46 CI 1.34-1.59; all
P<0.0001 after multiplicity adjustment.
Other risk factors were not associated with altered odds for postoperative
Age-stratified analysis: 42.3% (n=478,231) of knee replacement patients
were aged <65 while 57.7% (n=652,338) were aged ≥65.
Increased odds for postoperative delirium persisted for long acting
benzodiazepine use and pregabalin/gabapentin. The use of opioids was
associated with decreased odds for delirium particularly in patients ≥65.
Knee patients ≥65 : high dose OR 0.73 CI 0.64-0.82, medium dose OR 0.78 CI
0.71-0.86 (both versus low dose; P<0.0001). Knee patients <65: high dose OR
0.74 CI 0.57-0.97, medium dose OR 0.88 CI 0.68-1.13 ( both versus low dose
p=0.0112, p>0.999 respectively). Patterns were similar in hip arthroplasty.
Postoperative delirium after total hip and knee replacement: a population-based study on modifiable risk
Crispiana Cozowicz MD1,2, Lukas Pichler MD1,2, Jashvant Poeran MD PhD3, Nicole Zubizarreta MPH3, Sarah M. Weinstein BA1, Madhu Mazumdar PhD3,
Stavros G. Memtsoudis MD PhD1,2
1Hospital for Special Surgery, Department of Anesthesiology, Critical Care & Pain Management, New York, NY
2Paracelsus Medical University, Department of Anesthesiology, Perioperative Medicine, and Intensive Care Medicine, Salzburg, Austria
3Icahn School of Medicine at Mount Sinai, Department of Population Health Science and Policy, New York, NY
** Opioid prescription on day of surgery and day after; reference group: low
Risk factors for delirium in THA/TKA
Opioid prescription high**
Opioid prescription medium**
Anesthesia general + neuraxial
Benzodiazepines long acting
Benzodiazepines short acting
OR (95% CI)
0.7 0.85 1 1.15 1.3 1.45 1.6 1.75 1.9 2.05 2.2 2.35
Associations between risk factors and postoperative delirium