If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
A Matched Comparison of Postoperative Complications Between Smokers and Nonsmokers Following Open Reduction Internal Fixation of Distal Radius Fractures
Corresponding author: Jonathan N. Grauer, MD, Department of Orthopaedics and Rehabilitation, Yale School of Medicine, 47 College St, New Haven, CT 06510.
The purpose of the present study was to identify differences in 30-day adverse events, reoperations, readmissions, and mortality for smokers and nonsmokers who undergo operative treatment for a distal radius fracture.
Methods
The American College of Surgeons National Surgical Quality Improvement Program (NSQIP) database was queried for patients who had operatively treated distal radius fractures between 2005 and 2017. Patient characteristics and surgical variables were assessed. Thirty-day outcome data were collected on serious (SAEs) and minor adverse events (MAEs), as well as on infection, return to the operating room, readmission, and mortality. Multivariable logistic analyses with and without propensity-score matching was used to compare outcome measures between the smoker and the nonsmoker cohorts.
Results
In total, 16,158 cases were identified, of whom 3,062 were smokers. After 1:1 propensity-score matching, the smoking and nonsmoking cohorts had similar demographic characteristics. Based on the multivariable propensity-matched logistic regression, cases in the smoking group had a significantly higher rate of any adverse event (AAE) (odds ratio [OR], 1.75; 95% confidence interval [95% CI], 1.28–2.38), serious adverse event (SAE) (OR, 1.75; 95% CI, 1.22–2.50), and minor adverse event (MAE) (OR, 1.84; 95% CI, 1.04–3.23). Smokers also had higher rates of infection (OR, 1.73; 95% CI, 1.26–2.39), reoperation (OR, 2.07; 95% CI, 1.13–3.78), and readmission (OR, 1.83; 95% CI, 1.20–2.79). There was no difference in 30-day mortality rate.
Conclusions
Smokers who undergo open reduction internal fixation of distal radius fractures had an increased risk of 30-day perioperative adverse events, even with matching and controlling for demographic characteristics and comorbidity status. This information can be used for patient counseling and may be helpful for treatment/management planning.
Dr. Graham makes his analysis of the lead article for January 2021. "A Matched Comparison of Postoperative Complications Between Smokers and Nonsmokers Following Open Reduction Internal Fixation of Distal Radius Fractures.
Although tobacco use has dropped significantly over the past 60 years, a briefing by the Center for Disease Control and Prevention in 2016 reported that 15.5% of adults in the United States continue to smoke.
Prior work investigating the impact of smoking on outcomes after ORIF of distal radius fractures has shown that smokers are not at increased risk of operative complications or worse functional outcomes.
However, this work was limited by a comparatively small sample size and may have been inadequately powered to detect differences in perioperative complications. Understanding the effect of cigarette smoking on operative complications from distal radius ORIF patients remains important.
Worse outcomes after fracture fixation in smokers may be related to systemic effects leading to cardiopulmonary complications, cerebrovascular accidents, wound-healing problems, anesthesia complications, and thromboembolic events. Smoking may also have local effects at the fracture site related to decreased blood flow, lower capillary oxygen gradients in local tissue, impaired collagen production, and reduced extracellular matrix turnover. In orthopedic patients specifically,
published a retrospective chart review of 417 distal radius fractures at a single institution (of which 24.6% were smokers). The overall complication rate in this cohort was relatively high at 9.8% in smokers and 5.6% in nonsmokers. However, the generalizability of the study may be limited by its single-center design.
The purpose of the current study was to use a large, multicenter, national database, the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) to compare the rates of 30-day adverse events, reoperations, readmissions, and mortalities between smoking and nonsmoking patients who underwent operative intervention for distal radius fracture.
Methods
Data source
Open reduction internal fixation of distal radius fracture cases were extracted from the 2005 to 2017 NSQIP database. The NSQIP collects over 150 variables from eligible cases performed at more than 600 participating hospitals. Cases are tracked for 30 days following the operation, regardless of discharge status.
Study cohort
Distal radius ORIF cases were captured from the NSQIP database using the following Current Procedure Terminology codes: 25607, 25608, and 25609. Fracture type was inferred from the Current Procedure Terminology code. Extra-articular fractures are coded as 25607. Two-part intra-articular fractures are coded as 25608. Intra-articular fractures involving 3 or more parts are coded as 25609.
Demographic factors including age, sex, height, weight, functional status prior to injury, and American Society of Anesthesiologists (ASA) classification were directly extracted from the NSQIP database. The ASA classification and preoperative functional status have been used in several past studies as a proxy for comorbidity burden.
Predicting adverse outcomes after total hip arthroplasty: a comparison of demographics, the American Society of Anesthesiologists class, the Modified Charlson Comorbidity Index, and the Modified Frailty Index.
The ASA classification takes into account several classes of comorbidities, including chronic obstructive pulmonary disease, heart failure, coronary artery disease, and end-stage renal disease.
Body mass index (BMI, kg/m2) was calculated from height (in meters) and weight (in kilograms) data available in the database. Preoperative smoking status was also directly extracted from the NSQIP database where this was defined as having smoked cigarettes in the 12 months preceding operation.
Several surgical variables were also assessed. Anesthesia type (general vs local) was determined using the NSQIP anesthesia-type variable. General anesthesia was coded as such in NSQIP. The NSQIP variables of local and regional anesthesia were aggregated into a single definition for this study. Outpatient cases were defined as having a length of stay (LOS) of 0 days and inpatient cases were defined as having an LOS of greater than 0 days. Patients with missing demographic or surgical variables were excluded.
Thirty-day outcomes
Thirty-day outcomes were extracted from the database. For the purposes of the current study, complications were grouped into 3 categories: any adverse events (AAEs), minor adverse events (MAEs), and serious adverse events (SAEs). The same definitions of these composite end points have been utilized extensively in the orthopedic surgery literature.
Thirty-day postoperative complications and mortality following total knee arthroplasty: incidence and risk factors among a national sample of 15,321 patients.
Impact of resident involvement on orthopaedic surgery outcomes: an analysis of 30,628 patients from the American College of Surgeons National Surgical Quality Improvement Program Database.
Trends in the incidence and in-hospital outcomes of elective major orthopaedic surgery in patients eighty years of age and older in the United States from 2000 to 2009.
The MAEs included the occurrence of any of the following: superficial SSI, wound dehiscence, pneumonia, urinary tract infection (UTI), or postoperative renal insufficiency. The SAEs included the occurrence of any of the following: deep SSI, sepsis, failure to wean from mechanical ventilation, unplanned reintubation, postoperative renal failure, deep vein thrombosis, pulmonary embolism, cardiac arrest, myocardial infarction, or stroke. The occurrence of an AAE was defined as the occurrence of an MAE or an SAE.
In addition to grouping infections in the aforementioned outcome variables, the occurrence of infection was investigated as a separate outcome variable, defined as the occurrence of any of the following: superficial SSI, deep SSI, sepsis, UTI, or pneumonia. Death, an uncommon occurrence, was also separately assessed.
For 30 days following the index procedure, NSQIP tracks both hospital readmissions as well as revision operations related to the principal operative procedure. Occurrences of readmission within 30 days were available for cases performed in 2011 and later. Occurrences of reoperation within 30 days were available for cases performed in 2012 and later. The readmission analysis in this study included 15,139 of 16,158 cases (93.7%) and the analysis of reoperation included 14,376 of 16,158 cases (89.0%).
Statistical analysis
Smoker and nonsmoker cohorts were first compared on the basis of patient demographic/comorbidity characteristics and fracture/operative characteristics. Chi-square tests were used to compare categorical variables. Student t tests and analyses of variance were used to compare continuous variables.
After tabulating the occurrence/rates of adverse outcomes, 2 sets of multivariable logistic analyses were performed. The first multivariable logistic analyses were of the entire nonsmoking and smoking cohorts to determine the odds of experiencing an adverse outcome in the smoking relative to nonsmoking cohorts. These multivariable logistic regression models included age, sex, BMI, functional status, ASA classification, anesthesia type, inpatient/outpatient status, and fracture type as additional independent predictors.
Second, propensity-score matching was used to infer the odds of experiencing adverse events in the smoker cohort. Prior studies have found that propensity-score matching reduces bias while increasing precision and robustness of effect size estimates compared with multivariable regression.
algorithm was used to define a new nonsmoker cohort that was similar in demographic covariates (age, sex, BMI, functional status, ASA classification) and perioperative factors (anesthesia type, inpatient/outpatient status, fracture type) to the smoker cohort. The algorithm further dropped 181 cases from the smoker cohort that did not have a similar case in the nonsmoker cohort. Logistic regression was then performed on the propensity-matched groups to calculate the odds of experiencing adverse events in the smoker cohort, using the nonsmoker cohort as a referent.
The level of significance for all tests was set at P less than .05.
Results
In total, 16,990 patients were included in the study. After patients with missing demographic and surgical data were excluded, 16,158 patients (95.1%) remained. Of this study sample, 3,062 (19.0%) were smokers (Fig. 1). Table 1 shows the demographics and comorbidity markers of the smokers and nonsmokers. Smokers in the study sample were more likely to be younger (mean age of 50.2 years compared with 58.7 years; P < .05), male (38.2% male compared with 24.0% male), have a lower BMI (mean BMI of 27.3 compared with 28.3; P < .05), and have a higher ASA classification (8.8% ASA 1 compared with 19.9% ASA 1; P < .05). Table 2 shows the fracture and surgical variables for the 2 cohorts. Smokers were more likely to undergo general anesthesia (85.3% compared with 80.5%; P < .05) and have outpatient status (76.4% compared with 74.4%; P < .05).
Figure 1Forest plot shows propensity-matched ORs for different classes of adverse events in smokers relative to nonsmokers. Controlled for age, sex, BMI, functional status, ASA class, anesthesia type, inpatient/outpatient status, and fracture type.
In the aggregate cohorts (prior to propensity-score matching), the rate of AAEs within 30 days of index procedure was 3.89% in smokers, compared to 2.98% in the non-smoking cohort. The aggregate incidence rate of AAEs was 3.15%. The rate of SAEs within 30 days of index procedure was 2.91% in smokers compared with 2.37% in the nonsmoking cohort. The aggregate incidence rate of SAEs was 2.48%. The rate of MAEs was 1.18% in the smoking cohort compared with 0.71% in the nonsmoking cohort. The aggregate rate of MAEs was 0.80%. The smoking cohort also had higher rates of infection (3.66% compared with 2.76% in the nonsmoking cohort; aggregate rate of 2.93%), reoperation (1.18% compared with 0.83% in the nonsmoking cohort; aggregate rate of 0.90%), and readmission (2.19% compared with 1.52% in the nonsmoking cohort; aggregate rate of 1.65%). The rate of mortality within 30 days of the index procedure was lower in the smoking cohort (0.07% compared with 0.18% in the nonsmoking cohort). The aggregate rate of mortality was 0.15% in this study sample. These findings are summarized in Table 3.
Table 3Number of Adverse Events, Return to Operating Room, Readmissions, and Mortality for Smokers and Nonsmokers
Complication
Aggregate Population
Nonsmokers
Smokers
Multivariable OR Controlled for Demographics Multivariate Propensity-Matched OR
As described in the Methods section, the statistical significance of these differences in outcomes was first assessed using multivariable logistic regression models based on the aggregate cohorts. In this set of analyses, smokers had an increased OR of AAE (OR, 1.63; 95% CI, 1.29–2.06; P < .05), SAE (OR, 1.53; 95% CI, 1.17–2.00; P < .05), MAE (OR, 1.84; 95% CI, 1.20–2.83; P < .05), infection (OR,1.62; 95% CI, 1.27–2.06; P < .05), and readmission to the hospital within 30 days of index procedure (OR, 1.59; 95% CI, 1.17–2.16; P < .05) (Table 3). There were no statistically significant differences in the OR of return to the operating room within 30 days of index procedure or mortality within 30 days of index procedure.
Propensity-score matching was then used to define new smoker and nonsmoker cohorts that did not have significant differences in patient demographics, comorbidities, or perioperative factors (P > .300 for all). These findings are shown in the right columns of Tables 1 and 2.
After propensity-score matching was implemented, similar findings were observed. Smokers continued to have increased OR of AAE (OR, 1.75; 95% CI, 1.28–2.38; P < .001), SAE (OR, 1.75; 95% CI, 1.22–2.50; P < .05), MAE (OR, 1.84; 95% CI, 1.04–3.23; P < .05), infection (OR, 1.73; 95% CI, 1.26–2.39; P < .05), and readmission within 30 days of index procedure (OR, 1.83; 95% CI, 1.20–2.79; P < .05) (Table 3). There was additionally a statistically significant difference in return to the operating room within 30 days of index procedure (OR, 2.07; 95% CI, 1.13–3.78; P < .05) between the smoker and the nonsmoker cohorts. There remained no statistically significant difference in the OR of mortality within 30 days of index procedure between smokers and nonsmokers. Aggregate and propensity-score matched ORs for other covariates included in the models are reported in Table E1, Table E2, Table E3, Table E4, Table E5, Table E6, Table E7, Table E8, Table E9 (available on the Journal’s Web site at www.jhandsurg.org).
Discussion
The purpose of the current study was to compare 30-day perioperative outcomes of ORIF of distal radius fractures in the NSQIP database. Compared with nonsmokers, smokers in this study were found to be younger, more likely to be male, have a lower BMI, and a higher ASA classification. However, the absolute differences in these demographic factors are likely not clinically significant. Smokers were identified to have increased risk of perioperative adverse events, readmission, and reoperation even after matching and controlling for demographics and comorbidity status.
Smokers were found to have higher rates of AAEs, MAEs, and SAEs. These results are consistent with previous findings in the literature for acute and elective orthopedic surgery.
compared the incidence of various classes of complications between smokers and nonsmokers. That study found an increased rate of complications in smokers.
Similarly, this study found an increased rate of adverse events in smokers of all fracture types.
Infections (SSIs, pneumonia, UTIs, and sepsis episodes) were also found to be more common in smokers. This is consistent with prior studies that have found smoking to have an effect on the immune system with alterations in the humoral and cell-mediated immunity, both locally and systemically.
At the surgical sites specifically, a growing body of literature highlights that smoking is associated with impaired wound healing owing to poor tissue oxygenation and impaired collagen formation.
Past studies of increased SSIs have been reported for orthopedic procedures including ankle fractures, calcaneus fractures, open tibia fractures, and lumbar fusions.
in total joint arthroplasty patients. The significantly higher rate of general infection, including SSIs, among smokers in this study may have driven their increased rate of reoperation. Whereas NSQIP specifically tracks procedure-related reoperations, the specific reason for reoperation was not stated in the database. Notably, the present study sample had a low rate of reoperations and mortality.
Smokers had a higher rate of readmission in the current study. This finding was similar to the overall outpatient elective findings for elective hand and elbow surgery of Noureldin et al.
These findings are pertinent to daily practice because the 30-day readmission rate is a benchmark used by Medicare and Medicaid for reimbursements and as an indicator of hospital quality.
It is conceivable that smoking may lead to increased readmission rates in the distal radius fracture ORIF patients owing to higher rates of the adverse events warranting hospital admission. Additional investigation using more granular and complete readmission data is required to understand the mechanism of increased readmission rates in smokers.
The present study finds associations of large effect sizes (ORs between 1.73 and 2.07) between smoking and several classes of adverse events. Owing to the multicenter and multiyear nature of the database, the association holds for the more than 700 hospitals reporting to NSQIP from 2005 to 2017. However, the observational design of the present study precludes formally testing assertions of causality.
The current study has several limitations within the scope of which the results should be interpreted. Foremost, the study is limited by data elements captured by the NSQIP database. Most notably, fracture and surgery-specific variables were limited. There is also a lack of data on pain, return to function, patient-reported outcomes, and radiographic outcomes. Complications analyzed were, therefore, limited to those tracked by the NSQIP, although the present study included all complications available in the database. The number of pack-years of smoking exposure was also not available in the database, which precludes the analysis of dose-response relationships. The NSQIP has a bias toward the inclusion of larger, level-1 trauma centers in the database. Large-volume centers are known to have different complication profiles from smaller-volume centers, as prior research in the total joint arthroplasty literature has demonstrated.
Furthermore, distal radius fractures are amenable to being treated in private surgery centers, which are not included in the NSQIP database. However, the NSQIP database does include surgical cases performed in outpatient surgery centers affiliated with member institutions.
In conclusion, the present study used a large national database and corroborated the findings of a previous smaller study that identified smoking status as a risk factor for perioperative complications after ORIF of distal radius fractures. This information should be used for patient counseling and may be helpful for treatment planning.
Appendix
Table E1Multivariate OR for Adverse Events, Return to Operating Room, Readmissions, and Mortality for Patients of Varying Age
Predicting adverse outcomes after total hip arthroplasty: a comparison of demographics, the American Society of Anesthesiologists class, the Modified Charlson Comorbidity Index, and the Modified Frailty Index.
Thirty-day postoperative complications and mortality following total knee arthroplasty: incidence and risk factors among a national sample of 15,321 patients.
Impact of resident involvement on orthopaedic surgery outcomes: an analysis of 30,628 patients from the American College of Surgeons National Surgical Quality Improvement Program Database.
Trends in the incidence and in-hospital outcomes of elective major orthopaedic surgery in patients eighty years of age and older in the United States from 2000 to 2009.
30571-2&id=gr1.jpg){kind=link}