Impact of Smoking and Drinking on Complications After Revision Total Joint Arthroplasty

A Matched Cohort Analysis

Venkatsaiakhil Tirumala, MS; Georges Bounajem, MD; Christian Klemt, PhD; Ameen Barghi, MD; Young-Min Kwon, MD, PhD

Disclosures

J Am Acad Orthop Surg. 2021;29(15):e769-e781. 

In This Article

Results

Propensity Cohort Matching

After exclusions, a total of 3,336 revision TJAs remained for categorization into four substance usage cohorts. These cohorts presented notable differences in (1) age, (2) sex, (3) body mass index, (4) ASA, (5) Charlson comorbidity index, (6) comorbidities, and (7) indications for revision TJA. Race, laterality, revised joint type (hips versus knees), and anesthesia type (general versus spinal) remained statistically similar across all cohorts. After propensity score matching, a total of 1,924 patients remained, which included (1) 808 nonusers (42.0%), (2) 421 smokers (21.9%), (3) 483 drinkers (25.0%), and (4) 212 concurrent users (11.1%). The demographics and procedural details of the four matched cohorts are summarized in Table 1, which shows statistically similar baseline characteristics.

Nonusers Versus Users

Compared with nonusers, substance users (as a whole) were at higher odds for experiencing two or more inhospital complications between surgery and discharge (OR, 1.63, P < 0.05) (Table 2 and Table 3). The most prevalent inhospital complication among users was blood transfusion (6.2%), which they had higher odds for (OR, 1.87, P < 0.01). In addition, users had a longer total LOS (3.8 versus 3.4 days, P < 0.001) and higher odds for an extended LOS (OR, 1.56, P < 0.001). Users had a higher rate of 60-day readmissions (OR, 1.35, P = 0.02); however, 30- and 90-day readmissions did not significantly differ. Users had overall higher odds for developing major complications (OR, 1.72, P < 0.01) within 90 days postdischarge—specifically for developing wound dehiscence (OR, 2.79, P = 0.03) and undergoing revision surgery (OR, 1.84, P = 0.01). In addition, users had a higher rate of aseptic (OR, 1.37, P = 0.03) and septic failures (OR, 1.84, P = 0.01).

Nonusers Versus Smokers

Compared with nonusers, smokers were at higher odds for experiencing two or more inhospital complications between surgery and discharge (OR, 2.40, P < 0.01) (Table 2 and Table 3). The most prevalent inhospital complications among smokers were blood transfusions (7.0%) and acute renal failures (3.1%). When compared with nonusers, smokers had 56% higher odds for requiring a blood transfusion (P < 0.01), a longer total LOS (3.9 versus 3.4 days, P < 0.001), and higher odds for an extended LOS (OR, 1.54, P = 0.02). Furthermore, smokers were more frequently discharged to inpatient rehabilitation facilities (4.1% versus 6.9%, OR, 1.73, P = 0.04). The rates of 30-, 60-, and 90-day readmissions were similar for nonusers and smokers (all P > 0.05) (Table 4 and Table 5). However, smokers had overall higher odds for developing major complications (OR, 1.56, P < 0.01) within 90 days postdischarge—specifically for developing wound dehiscence (OR, 1.86, P < 0.01), deep surgical site infections (OR, 1.46, P = 0.04), and undergoing revision surgery (OR, 1.76, P = 0.046). Although smokers had a higher prevalence of re-revision TJAs for aseptic and septic failures, these outcomes were not statistically significant.

Nonusers Versus Drinkers

Compared with nonusers, drinkers were at higher odds for experiencing an inhospital complication (OR, 1.55, P < 0.001) (Table 2 and Table 3) with blood transfusions (7.7%) and urinary tract infections (3.7%) being the most common. Drinkers had higher odds for requiring perioperative blood transfusions (OR, 1.72, P < 0.01) and a similar total LOS and odds of extended LOS as nonusers. Furthermore, drinkers had similar discharge dispositions as nonusers. Compared with nonusers, drinkers had statistically similar 30-, 60-, and 90-day readmissions (all P > 0.05) (Table 4 and Table 5). However, drinkers had higher overall odds for developing major complications within 90 days of discharge (OR, 1.57, P < 0.01). Specifically, drinkers had higher odds for requiring blood transfusions (OR, 1.72, P < 0.01), developing wound dehiscence (OR, 1.98, P < 0.01) and deep surgical site infection (OR, 2.07, P < 0.01), and undergoing revision surgery (OR, 1.61, P = 0.049). Moreover, drinkers also had higher odds for undergoing re-revision surgery for periprosthetic joint infection (PJI) (OR, 2.09, P = 0.01).

Nonusers Versus Concurrent Users

Compared with nonusers, concurrent users were at higher odds for developing two or more inhospital complications (OR, 4.13, P < 0.01) (Table 2 and Table 3). The most prevalent inhospital complications among concurrent users were blood transfusions (11.7%) and acute renal failure (3.8%). Concurrent users had 174% increased odds for requiring blood transfusions compared with nonusers (P < 0.01). In addition, concurrent users had a significantly longer LOS (4.3 versus 3.4 days, P < 0.001) and higher odds for an extended LOS (OR, 2.02, P < 0.01). Concurrent users were also more frequently discharged to inpatient rehabilitation (4.1% versus 9.0%, OR, 2.31, P < 0.01). Compared with nonusers, concurrent users had higher rates of 60-day (13.2% versus 19.3%, OR, 1.60, P = 0.02) and 90-day readmissions (13.2% versus 19.3%, OR, 1.57, P = 0.01) (Table 4 and Table 5). Concurrent users had higher overall odds for developing major complications (OR, 2.74, P < 0.01) within 90 days postdischarge. In particular, concurrent users had higher odds for superficial surgical site infection (OR, 4.30, P = 0.01), deep surgical site infection (OR, 2.81, P = 0.04), wound dehiscence (OR, 5.73, P < 0.001), and revision surgery (OR, 2.26, P < 0.01). In addition, concurrent users underwent re-revision surgery for PJI more frequently (OR, 2.44, P = 0.01).

Drinkers Versus Smokers

Compared with drinkers, smokers had higher odds for developing two or more inhospital complications (OR, 1.93, P = 0.03), having a longer LOS (3.5 versus 3.9 days, P < 0.01) and being discharged to inpatient rehabilitation care (OR, 1.73, P = 0.04) (Table 2 and Table 6). Differences in readmissions, complications, and re-revisions were statistically similar between drinkers and smokers (Table 4 and Table 7).

Concurrent Users

Compared with smokers, concurrent users had higher odds for developing two or more inhospital complications (OR, 2.14, P < 0.01) and having longer LOS (3.9 versus 4.3 days, P < 0.01) (Table 2 and Table 6). In particular, concurrent users had frequently more blood transfusions (OR, 1.76, P = 0.049) during their stay. Concurrent users also had higher overall odds for developing major complications (OR, 1.62, P = 0.03) within 90 days of discharge; however, no individual complication rate was statistically different between both cohorts (Table 4 and Table 7).

Compared with drinkers, concurrent users had higher odds for developing two or more inhospital complications (OR, 4.13, P < 0.01), having longer LOS (3.5 versus 4.3 days, P < 0.01), and being discharged to inpatient rehabilitation care (OR, 3.09, P < 0.01) (Table 2 and Table 6). For complications, compared with drinkers, concurrent users had higher overall odds for developing major complications (OR, 1.64, P = 0.02) within 90 days of discharge. Wound dehiscence was the only individual complication concurrent users had specifically higher odds for (OR, 2.89, P = 0.049) (Table 4 and Table 7).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....