Abstract
Summary Statement
Pulmonary embolism (PE) after total knee arthroplasty can have a significant impact on patient outcomes and healthcare costs. Efforts to prevent or minimise PE over the last 10 years have not had a significant impact on its occurrence at the national level.
Introduction
Pulmonary embolism (PE) is a rare but known potentially devastating complication of total knee arthroplasty (TKA). Significant healthcare resources and pharmaceutical research has been recently focused on preventing this complication but limited data exists regarding the early results of this great effort. The purpose of this study was to assess recent national trends in PE occurrence after TKA and evaluate patient outcomes related to this adverse event.
Methods
International Classification of Disease - 9th Revision (ICD-9) procedure codes were used to search the National Hospital Discharge Survey (NHDS) for all patients admitted to US hospitals after primary TKA for each year between 2001 and 2010. ICD-9 diagnosis codes were then used to identify patients from this population who developed an acute PE during the same admission. Data regarding patient demographics, hospitalization length, discharge disposition, deep vein thrombosis, mortality, and hospital size/location were gathered from the NHDS. Trends were evaluated by linear regression with Pearson's correlation coefficient (r) and statistical comparisons were made using Student's t-test, z-test for proportions, and chi-square analysis with a significance level of 0.05.
Results
35,220 patients admitted for a primary TKA were identified. 159 (0.045%) of these patients developed an acute PE during the same admission. After adjusting for fluctuations in annual TKA performed, the development of PE after TKA demonstrated a weak negative correlation with time (r=0.17), insignificantly decreasing from an average rate of 0.049% between 2001–2005 to 0.041% between 2006–2010 (p=0.26). The size of the hospital was found to significantly impact the incidence of PE and primary TKA, with the lowest rate seen in hospitals under 100 beds (0.23%) and the highest rate seen in those with over 500 beds (0.65%, p=0.01). No significant differences in PE incidence were noted based on US region (p=0.38). The mean age of patients with PE was 67.7 years. This group included 54 men and 105 women. The non-PE group had a mean patient age that was insignificantly lower at 66.6 years (p=0.21) and included 12,450 men and 22,611 women. Gender was also not significantly different (p=0.68) between those with PE and those without PE. The number of medical co-morbidities was significantly higher in those with PE (mean 6.42 diagnoses) than those without PE (mean 4.89 diagnoses, p<0.01). Average hospitalization length also varied based on PE status, with significantly longer stays for those with PE (8.2 days, range 2–53) compared to those without PE (3.7 days, range 1–95, p<0.01). The rate of deep vein thrombosis was higher in the PE group (12.7%) versus the non-PE group (0.48%, p<0.01). Mortality was also significantly higher for the PE group (3.9%) compared to the non-PE group (0.09%, p<0.01). Discharge disposition did not significantly vary based on PE status, with 61.5% of PE and 64.0% of non-PE patients able to go directly home (p=0.59) after their inpatient stay.
Discussion/Conclusion
This study demonstrates that PE can have a significant impact on patient outcomes and healthcare costs, with an associated 43-fold increase in mortality and a doubling of the inpatient admission duration. Additionally, although the risk of PE after primary TKA remains rare, it still persists. Efforts to prevent or minimise this complication over the last 10 years have not had a significant impact on its occurrence at the national level. This risk of PE appears to be greatest in patients with multiple medical co-morbidities and established DVTs. Interestingly, the PE rate also demonstrated variability based on hospital size. The reasons for this are not clear, but we suspect larger hospitals are more likely to be tertiary-care centers and thus care for more medically-complex patients.
