Does type of anesthesia influence risk of postoperative delirium in total knee arthroplasty patients?
Sarah M. Weinstein BA, Eva E. Mörwald MD, Crispiana Cozowicz MD, Lila R. Baaklini MD, Lazaros Poultsides MD PhD, Jashvant Poeran MD PhD, Nigel Sharrock MD, Stavros G. Memtsoudis MD PhD FCCP
Postoperative delirium has drawn considerable attention as its association with other adverse outcomes, including readmission, longer hospital stay, higher cost of care, and mortality have been highlighted in recent publications. While previous research has suggested that older age and psychiatric comorbidities increase risk for delirium, the impact of modifiable risk factors, such as anesthesia type, is still debated.[1-2]
We hypothesized that patients who underwent total knee arthroplasties (TKA) and received general anesthesia would have higher odds of developing postoperative delirium than those who received neuraxial anesthesia.
MATERIALS & METHODS
For this retrospective study, with approval from the Hospital for Special Surgery (HSS) Institutional Review Board, we identified patients who underwent unilateral TKA at HSS from 2005-2014. Using the statistical software R version 3.3.1, we merged various administrative datasets to combine anesthesia billing data, ICD-9 codes, and demographic information. The resulting dataset consisted of 22,179 patients. Patients whose initial anesthesia billing transaction did not occur until after their date of admission (n=73) and those whose primary type of anesthesia was missing (n=901) were excluded.
For the remaining 21,205 patients, complications and comorbidities were identified using ICD-9 code definitions. A multivariable logistic regression model was used to compute odds ratios (OR) and corresponding 95% confidence intervals (CI) and p-values. Covariates included in the model were age (in years), gender (female or male), type of anesthesia (general, combined spinal epidural – CSE, epidural, or spinal), and the following comorbidities: hypothyroidism, renal, diabetes, obesity, sleep apnea, valvular disease, hypertension, congestive heart failure, tobacco abuse, pulmonary, depression/anxiety, and psychoses.
454 of the 21,205 patients (2.14%) in this cohort had an ICD-9 diagnosis code indicating postoperative delirium. These patients were, on average, older than those without postoperative delirium and had higher prevalence of psychoses and depression/anxiety (Table 1).
Controlling for age, gender, and comorbidities, type of anesthesia was a significant predictor of a patient’s odds of developing postoperative delirium (Table 2). Compared to patients who had general anesthesia, the OR of postoperative delirium for CSE patients was 0.55 (CI 0.34-0.89, p=0.0155), for epidural patients was 0.45 (CI 0.20-0.99, p=0.0483), and for spinal patients was 0.52 (CI 0.30-0.91, p=0.0211).
An AUC of 0.7844 suggested the regression model’s effectiveness at discriminating between patients who had and did not have postoperative delirium.
In this cohort of unilateral TKA patients, general anesthesia was associated with higher odds of postoperative delirium. Given that other risk factors for this complication are difficult or impossible to alter, it is especially important to consider anesthesia type as a modifiable risk factor for postoperative delirium (as well as for other complications).
Further research will be needed to validate our results and demonstrate the extent to which these findings are generalizable to other patient cohorts.
These results are a portion of an ongoing analysis, and we are in the process of expanding this cohort to include other orthopedic patients as well as covariates from pharmacologic and laboratory datasets.
1. Razak, Hamid Rahmatullah Bin Abd, Yung WYA. Postoperative delirium in patients undergoing total joint arthroplasty: A systematic review. J Arthroplasty. 2015;30(8):1414-1417.
2. Scott J, Mathias J, Kneebone A. Incidence of delirium following total joint replacement in older adults: A meta-analysis. Gen Hosp Psychiatry. 2015;37(3):223-229.