Does adding an aortic root replacement or sinus repair during arch repair increase postoperative mortality? Evidence from the Canadian Thoracic Aortic Collaborative

Abstract

 

Open in new tabDownload slide

OBJECTIVES

The aim of this study was to examine the effect of the addition of an aortic root replacement or sinus repair on mortality and morbidity during aortic arch repair.

METHODS

A total of 2472 patients underwent proximal or total aortic arch repair with hypothermic circulatory arrest between 2002 and 2018 at 12 centres. Multivariable logistic regressions (MV) and propensity score (PS) with inverse probability of treatment weighting (IPTW) analyses were performed.

RESULTS

A total of 1099 (44.5%) patients had additional aortic root replacement (n = 934) or sinus repair (n = 165). Those with aortic root interventions were younger (61 ± 13 vs 64 ± 13 years, P < 0.001) and had less females (23% vs 35%, P < 0.001), less dissection (31% vs 36%, P = 0.004), less urgent cases (35% vs 39%, P = 0.047), more connective tissue disease (7% vs 3%, P < 0.001) and less total arch replacements (14% vs 22%, P < 0.001). On adjusted analyses, the addition of aortic root procedure was associated with increased mortality [MV: odds ratio (OR) 1.41, 95% confidence interval (CI) 1.03–1.92; PS-IPTW: risk increased by 3.7%, 95% CI 1.2–6.3%, P = 0.004]. Reoperation for bleeding was also increased with the addition of aortic root intervention (MV: OR 1.48, 95% 1.10–1.99; PS-IPTW: risk increased by 3.2%, 95% CI 0.8–5.6%, P = 0.009). The risks of stroke and dialysis-dependent renal failure were similar. When looking only at non-elective cases, the increased risk of mortality was more pronounced (MV: OR 1.60, 95% CI 1.11–2.32, P = 0.013; PS-IPTW: risk increased by 6.8%, 95 CI 1.7–11.8%, P = 0.008, and a number need to harm of 15 patients to cause 1 additional death).

CONCLUSIONS

The addition of aortic root replacement or sinus repair during proximal or total aortic arch repair seems to increase postoperative mortality only in non-elective cases.

INTRODUCTION

Continuous debate still exists regarding extent of proximal aortic repair in patients presenting with an index aortic arch procedure. Aortic arch repairs are known to be associated with high mortality and morbidity, and it could be argued that the addition of an aortic root procedure (with increased surgical complexity and cardiopulmonary and cross-clamp times) could result in worsened outcomes. A previous multicentre study of 1169 cases reported no increased postoperative morbidity or mortality when adding an aortic root operation to an aortic arch procedure [1]; however, this study was only limited to elective cases and did not report outcomes of patients presenting with aortic dissection. To address this question, we set out to evaluate the effect of the addition of an aortic root replacement or sinus repair on mortality and morbidity during proximal or total aortic arch repair. We hypothesized that the addition of aortic root procedure to an indexed proximal or total aortic arch procedure would increase mortality.

MATERIALS AND METHODS

Study population

The Canadian Thoracic Aortic Collaborative is a network of aortic surgeons across Canada who retrospectively compiled a comprehensive multicentre registry of consecutive patients undergoing proximal arch or total arch surgery with circulatory arrest. An aortic root procedure was defined as any of the following: sinus root repair (most frequently repair of the non-coronary sinus, and could include 2 sinus repairs but not all 3), modified Bentall procedure, or valve-sparing root replacement (either the reimplantation or the remodelling technique). The major indications for aortic root procedures were a combination of annulo-aoritc ectasia, aneurysmal disease, aortic root dissections, and organic aortic valve disease. Total arch replacement (versus hemiarch replacement) was performed in the presence of any of the following: intimal tear in the aortic arch; circumferential involvement/dissection of the aortic arch; aortic arch aneurysm ≥4.5 cm; or cerebral malperfusion with complex arch tear. We included both elective and emergent cases. Cases involving thoraco-abdominal aortic repair were excluded.

Ethical statement

Each centre obtained local ethics approval from their respective institutional review boards, and individual informed consent was waived at all centres.

Outcomes

We evaluated 4 primary outcomes: in-hospital mortality, in-hospital stroke, reoperation for bleeding and dialysis-dependent renal failure. Secondary outcomes included major postoperative complications such as delirium, transient neurologic deficit, prolonged mechanical ventilation (>48 h), sepsis, in addition to blood transfusion and intensive care unit (ICU) length of stay (LOS) and hospital LOS.

Statistical methods

Continuous variables were expressed as mean ± standard deviation or median (interquartile range) and were compared using Student’s t-test or Wilcoxon rank-sum test, as appropriate. Categorical variables were expressed as frequencies (%) and were compared using Pearson’s chi-squared test or Student’s t-tests/ANOVA, as appropriate.

We built a separate multivariable logistic model for each of the 4 primary outcomes in which we forced the exposure of interest (addition of aortic root procedure versus no addition of aortic root procedure) to be included in all of the models. The logistic regression models included age, sex, aortic dissection, coronary artery disease, peripheral vascular disease, redo-cardiac surgery, connective tissue disease, renal failure, concomitant procedures, hemiarch/total arch replacement, hypothermic circulatory arrest temperature, cerebral perfusion strategy and emergency status. The effect of the individual centres was accounted for by using a mixed effects logistic regression models with a random effect (random intercept) for the centres.

In an additional effort to isolate the effect of the addition versus no addition of aortic root procedure on the primary outcomes, we sought to address confounding by performing a propensity score (PS) analysis with inverse probability of treatment weighting (IPTW) based on the PS of adding an aortic root procedure or not. The PS was calculated with a priori logistic regression model that was built using covariates that were judged to strongly influence the addition of an aortic root procedure and to impact the primary outcomes, such as patients’ age, sex, aortic dissection, coronary artery disease, peripheral vascular disease, redo, chronic obstructive pulmonary disease, connective tissue disease, renal failure, concomitant procedures, hemiarch/total arch replacement, hypothermic circulatory arrest temperature, cerebral perfusion strategy and emergency status. We computed standardized differences to compare patients’ covariates before and after IPTW adjustment with an absolute value above 10% indicating significance. We used the teffects ipw package in Stata version 16.0 (StataCorp LP, College Station, TX, USA) and the average treatment effect to perform the IPTW-adjusted analysis of our outcomes.

P-values <0.05 were considered statistically significant. Statistical analysis was conducted using Stata version 16.0.

RESULTS

Baseline characteristics

A total of 2472 patients were included in the analysis: 1099 patients had an additional aortic root procedure (‘additional root’: Bentall procedure: 711; valve-sparing root replacement: 223; sinus root repair: 165) and 1373 patients did not have an additional aortic root procedure (‘non-additional root’). Out of the 2472 patients, there were 2013 hemiarch replacements (of whom 188 patients had a valve-sparing root procedure and 612 patients had a Bentall procedure) and 459 total arch replacements (of whom 35 patients had a valve-sparing root procedure and 99 patients had a Bentall procedure). Table 1 summarizes the baseline characteristics for both groups. Several baseline differences were noted between both groups, such as the rates of connective tissue disorder (7% for additional root vs 3% for non-additional root, P < 0.001) and aortic dissection (31% for additional root vs 36% for non-additional root, P = 0.004).

Operative characteristics

Table 2 summarizes the intraoperative details for patients with the additional root versus non-additional root. Several intraoperative differences were noted between both groups, such as the rates of total arch replacement (14% for additional root vs 22% for non-additional root, P < 0.001) and frozen elephant trunk (8% for additional root vs 11% for non-additional root, P = 0.002). The additional root group had longer cardiopulmonary bypass time [median (interquartile range): 197 (157–249) vs 159 min (121–210), P < 0.001] and aortic cross-clamp time [142 (110–185) vs 85 min (60–113), P < 0.001].

Univariate analysis of in-hospital outcomes (aortic root procedure defined as either a Bentall procedure, a valve-sparing root replacement or a sinus root repair)

Effect on mortality

In-hospital mortality rate was 8.9% for the entire cohort and did not differ between the additional root and non-additional root groups (9.0% for additional root vs 8.8% for non-additional root, P = 0.87) (Table 3). For elective cases (total of 1555 cases), the in-hospital mortality rate was 3.7% (n = 58) and differed between the additional root and non-additional root groups (2.7% for additional root vs 4.6% for non-additional root, P = 0.039). For non-elective cases (total of 917 cases), the in-hospital mortality rate was 17.7% (n = 162) and differed between the additional root and non-additional root groups (20.8% for additional root vs 15.4% for non-additional root, P = 0.033).

Effect on stroke

Postoperative stroke occurred in 8.3% among the entire study cohort. There was no difference between both groups (7.6% vs 8.8%, P = 0.30) (Table 3).

Effect on reoperation for bleeding

The rate of postoperative reoperation for bleeding was 8.4% for the entire cohort, and there was a numerically higher rate in the additional root group (9.6% for additional root vs 7.4% for non-additional root, P = 0.058) (Table 3).

Effect on dialysis-dependent renal failure

The rate of dialysis-dependent renal failure was 5.2% for the entire cohort, and there was no difference between both groups (4.9% for additional root vs 5.5% for non-additional root, P = 0.49) (Table 3).

Multivariable analysis (Table 4) (aortic root procedure defined as either a Bentall procedure, a valve-sparing root replacement or a sinus root repair)

Effect on the primary outcomes

On multivariable analysis, when compared to the non-addition of aortic root procedure, the addition of aortic root procedure was found to be associated with increased odds of in-hospital mortality [odds ratio (OR) 1.41, 95% confidence interval (CI) 1.03–1.92, P = 0.031], increased reoperation for bleeding (OR 1.48, 95% CI 1.10–1.99, P = 0.010), but resulted in similar odds of stroke (OR 1.1, 95% CI 0.81–1.51, P = 0.53) and dialysis-dependent renal failure (OR 1.16, 95% CI 0.78–1.72, P = 0.48).

Other factors that impacted the in-hospital mortality were age (per 1-year difference, OR 1.03, 95% 1.02–1.05, P < 0.001), prior sternotomy (OR 1.57, 95% CI 1.05–2.35, P = 0.028), pre-operative renal failure (OR 2.48, 95% CI 1.63–3.75, P < 0.001), non-elective status (OR 3.37, 95% CI 2.23–5.10, P < 0.001), total arch replacement (OR 1.79, 95% CI 1.24–2.60, P = 0.002), hypothermic circulatory arrest temperature (per 1-degree difference, OR 0.93, 95% CI 0.89–0.97, P = 0.002) and the use of cerebral perfusion (OR 0.56, 95% CI 0.40–0.79, P = 0.001).

Other factors that impacted reoperation for bleeding were hypothermic circulatory arrest temperature (per 1-degree difference, OR 0.94, 95% CI 0.91–0.98, P = 0.001) and the use of cerebral perfusion (OR 0.70, 95% CI 0.52–0.95, P = 0.023).

Factors that impacted the odds of stroke were female sex (OR 1.42, 95% CI 1.03–1.95, P = 0.033), non-elective status (OR 1.60, 95% CI 1.04–2.50, P = 0.031), total arch replacement (OR 1.89, 95% CI 1.30–2.76, P = 0.001) and the use of cerebral perfusion (OR 0.59, 95% CI 0.41–0.85, P = 0.004).

Factors that impacted the odds of dialysis-dependent renal failure were pre-operative renal failure (OR 4.42, 95% CI 2.64–7.39, P < 0.001) and non-elective status (OR 3.91, 95% CI 2.23–6.87, P < 0.001).

Effect on the secondary outcomes

On multivariable analysis, when compared to the non-addition of aortic root procedure, the addition of aortic root procedure was associated with similar odds of delirium (OR 0.95, 95% CI 0.76–1.20, P = 0.69), transient neurologic deficit (OR 0.88, 95% CI 0.47–1.64, P = 0.68), prolonged mechanical ventilation (> 48 h) (OR 1.24, 95% CI 0.96–1.59, P = 0.10), sepsis (OR 1.27, 95% CI 0.70–2.33, P = 0.43) and similar ICU (mean difference between both groups: 0.54 days, P = 0.13) and hospital LOS (mean difference between both groups: 0.50 days, P = 0.47). The additional of aortic root procedure was associated with increased red blood cells transfusion (mean difference between both groups: 0.42 U, P = 0.003), fresh frozen plasma transfusion (mean difference between both groups: 0.68 units, P < 0.001) and platelet transfusion (mean difference between both groups: 0.51 units, P = 0.042).

PS analysis with IPTW (aortic root procedure defined as either a Bentall procedure, a valve-sparing root replacement or a sinus root repair)

Table 5 shows the baseline characteristics for both groups after adjustment with inverse probability of treatment weighting. After IPTW adjustment, all of the variables were balanced between both groups with absolute values of standardized differences below 10%.

The group with additional aortic root procedure had a PS-adjusted risk difference for mortality of 3.4% (95% CI 0.9–5.9%, P = 0.008), which yields a number needed to harm of 30 patients to cause 1 additional death (Table 6) and a PS-adjusted risk difference for reoperation for bleeding of 2.8% (95% CI 0.4–5.2%, P = 0.021). There was no difference between both groups in the adjusted risk difference for stroke (0.2%, 95% CI −2.1% to 2.6%, P = 0.86) or dialysis-dependent renal failure (−0.1%, 95% CI −2.0% to 1.7%, P = 0.89).

Sub-group analyses

To investigate whether the urgency status (elective versus non-elective, as defined by the Society of Thoracic Surgery Adult Cardiac Surgery Database Data Specification Version 2.9) [2] modified the effect of the addition of aortic root procedure on the risk of the primary outcomes, we divided the entire cohort into 2 subsets, elective and non-elective cases, and we performed the multivariable analysis and PS-IPTW analysis on each of these 2 subsets. Acute aortic dissection constituted the majority of non-elective cases, with 76.2% of non-elective patients presenting with an acute aortic dissection.

Looking at the elective cases, the addition of aortic root procedure was found to be associated with similar risks of in-hospital mortality (multivariable analysis: OR 0.95, 95% CI 0.52–1.76, P = 0.88; PS-IPTW-adjusted risk difference: 0.3%, 95% CI −2.0% to 2.6%, P = 0.81), stroke (multivariable analysis: OR 1.20, 95% CI 0.71–2.00, P = 0.50; PS-IPTW-adjusted risk difference: 1.5%, 95% CI −1.7% to 4.6%, P = 0.36) and dialysis-dependent renal failure (multivariable analysis: OR 0.67, 95% CI 0.28–1.62, P = 0.38; PS-IPTW-adjusted risk difference: −0.62%, 95% CI −2.0% to 0.72%, P = 0.37). However, the addition of aortic root procedure was associated with increased risk of reoperation for bleeding (multivariable analysis: OR 2.06, 95% CI 1.35–3.15, P = 0.001; PS-IPTW-adjusted risk difference: 5.0%, 95% CI 2.0–8.0%, P = 0.001).

When looking at non-elective cases, the addition of aortic root procedure was found to be associated with increased in-hospital mortality (multivariable analysis: OR 1.60, 95% CI 1.11–2.32, P = 0.013; PS-IPTW-adjusted risk difference: 6.8%, 95% CI 1.7–11.8%, P = 0.008), but similar risks of stroke (multivariable analysis: OR 1.02, 95% CI 0.68–1.54, P = 0.92; PS-IPTW-adjusted risk difference: 0.3%, 95% CI −4.2% to 4.9%, P = 0.89), dialysis-dependent renal failure (multivariable analysis: OR 1.28, 95% CI 0.81–2.03, P = 0.29; PS-IPTW-adjusted risk difference: 1.2%, 95% CI −2.9% to 5.4%, P = 0.56) and reoperation for bleeding (multivariable analysis: OR 1.12, 95% CI 0.72–1.76, P = 0.62; PS-IPTW-adjusted risk difference: 0.9%, 95% CI −3.2% to 4.9%, P = 0.66).

We performed another sub-group analysis where we investigated the effect of the addition of aortic root procedure on the risk of mortality only in patients who underwent hemiarch replacement. We found that in elective cases, the additional of aortic root procedures was not associated with increased mortality (multivariable regression OR 0.74, 95% CI 0.32–1.70, P = 0.48; PS-adjusted risk difference for mortality −0.5%, 95% CI −2.6% to 1.5%, P = 0.60). However, in non-elective cases, the addition of aortic root procedures was associated with increased mortality (multivariable regression OR 1.70, 95% CI 1.11–2.60, P = 0.014; PS-adjusted risk difference for mortality 8.0%, 95% CI 2.4–13.5%, P = 0.005).

We performed an additional sensitivity analysis where we excluded patients who had sinus root repair and defined ‘aortic root procedure’ as either modified Bentall procedure, or valve-sparing root replacement. We then investigated the effect of the addition of aortic root procedure on the risk of mortality and found that in elective cases, the additional of aortic root procedures was not associated with increased mortality (multivariable regression OR 1.10, 95% CI 0.58–2.08, P = 0.77; PS-adjusted risk difference for mortality 0.7%, 95% CI −2.0% to 3.3%, P = 0.62). However, in non-elective cases, the addition of aortic root procedures was associated with increased mortality (multivariable regression OR 1.80, 95% CI 1.20–2.70, P = 0.004; PS-adjusted risk difference for mortality 8.2%, 95% CI 2.4–14.0%, P = 0.006).

DISCUSSION

This collaborative study examined the postoperative outcomes of patients undergoing concomitant aortic root procedures (aortic root replacement or sinus repair) during an indexed proximal or total aortic arch repair. Our findings indicate that the addition of aortic root procedures is associated with increased postoperative mortality and increased reoperation for bleeding (Fig. 1). The effect on mortality is only present in non-elective cases and is not observed in patients presenting for elective proximal or total arch repair.

Aortic root replacement aims at addressing the diseased aortic root and possibly aortic valve to minimize future risk and restore aortic valve function [3–5]. If left untreated, residual aortic root aneurysms remain at risk for further aneurysmal growth, acute aortic dissection and need for reoperation. Reoperation on the aortic root following previous aortic surgery represents a technically challenging sub-group of patients that carries a higher risk of postoperative mortality and morbidity [6, 7].

Given the added complexity of an additional aortic root procedure, there has been a debate regarding the appropriate proximal extent of aortic reconstruction that should be undertaken in patients undergoing aortic arch repair. This issue becomes paramount when little aortic valve dysfunction exists. On the one hand, the optimal procedure aims to fix the diseased and at risk aortic segments to successfully treat the patient and prevent late complications. On the other hand, the operative mortality associated with contemporary open aortic arch repair can range between 2.5% for elective and 25% for emergency cases, therefore suggesting that limiting the complexity of the overall operation might help avoid additional perioperative mortality and morbidities associated with this already high-risk surgery [8, 9]. Importantly, adding an aortic valve repair or replacement should be carefully considered as it is associated with significant early and long-term implications, particularly in patients with normal aortic valve function. Previous studies have examined the outcomes of the addition of hemiarch repair to aortic root surgery and found that the combination of these 2 procedures did not increase operative mortality and complications when compared to isolated aortic root or ascending aortic replacement [10, 11]. However, few studies have looked at the outcomes of aortic arch surgery with and without the addition of an aortic root procedure. A recent study by the International Aortic Arch Surgery Study Group examined the effect of adding an aortic root procedure to elective arch surgery and showed that the addition of an aortic root procedure prolongs cardiopulmonary bypass and aortic cross-clamp times but does not increase postoperative morbidity or mortality [1]. This report only included elective arch procedures and did not report outcomes of patients presenting with aortic dissection. In our study, we performed a sub-group analysis to look at both elective and non-elective subsets of patients. Our finding in elective arch repair is consistent with that from the International Aortic Arch Surgery Study Group where the addition of aortic root procedure was found to be associated with the similar risk of in-hospital mortality. However, when looking at non-elective cases, the addition of aortic root procedure was found to be associated with the accentuated risk of in-hospital mortality. As expected, aortic dissection constituted the majority of non-elective cases. The differential survival for the addition of aortic root replacement between elective and non-elective proximal or total arch surgery supports the need to apply different treatment strategies in non-elective patients where focus could be to seek the most expeditious and simplified concomitant procedures required.

It is plausible that the decision for surgeon to add a root procedure was influenced by the extent of the proximal disease, where a surgeon might preferentially decide to add a root procedure when performing a hemiarch rather than a total arch repair, or inversely to limit the extent of the arch surgery when a root procedure is absolutely indicated. In fact, in our study, root replacement patients were more likely to receive a hemiarch rather than a total arch repair. Other factors that could have influenced the decision for the addition of root replacement include the presence of aortic dissection, aortic valve function, patient’s comorbidities and the acuity of the presentation. To account for these potential confounders, all baseline characteristics collected in our study were included in our inverse probability of treatment weighting analysis.

Limitations

Our study has certain limitations. Given the observational nature of our study, the non-elective group was perhaps somewhat heterogenous, most importantly with regard to the extent of the aortic dissection. The specific information regarding the extent of aortic root involvement with acute aortic dissection was unavailable and could have confounded the extent of root surgery and our findings. Specific information regarding the extent of aortic sinus repair or the technique of valve-sparing root replacement (reimplantation versus remodelling) was not available. We also did not have information on the exact prior cardiac operations of the patients who had a redo sternotomy. We were also unable to delineate whether the primary indexed operation was a root surgery (to which a hemiarch/total arch replacement was added) or a hemiarch/total arch replacement (to which a root procedure was added). The Canadian Thoracic Aortic Collaborative database lacked time-to-event data and all of our in-hospital outcomes were retrospectively collected. Therefore, by definition, the effect of the addition of aortic root procedure on morbidity (such as stroke, dialysis-dependent renal failure and reoperation for bleeding) was only examined among survivors.

CONCLUSIONS

Taken together, our findings show that after adjusting for multiple baseline characteristics and potential factors affecting surgical decision-making, the addition of an aortic root replacement or sinus repair during proximal or total aortic arch repair was associated with increased postoperative mortality only in non-elective cases.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

Conflict of interest: Michael W. A. Chu has received speaker’s honoraria from Medtronic, Edwards Lifesciences, Boston Scientific, Terumo Aortic and Abbott Vascular. There are no other disclosures.

Presented at the 34th Annual Meeting of the European Association for Cardio-Thoracic Surgery, Barcelona, Spain, 8–10 October 2020.

Author contributions

Fadi Hage: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Resources; Software; Validation; Visualization; Writing—original draft; Writing—review & editing. Ali Hage: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Project administration; Software; Validation; Visualization; Writing—original draft; Writing—review & editing. Francois Dagenais: Conceptualization; Data curation; Investigation; Project administration; Resources; Supervision; Validation; Writing—original draft. Andreanne Cartier: Data curation; Methodology; Writing—original draft. Maral Ouzounian: Conceptualization; Investigation; Methodology; Project administration; Supervision; Validation; Writing—original draft. Jennifer Chung: Conceptualization; Data curation; Methodology; Validation; Writing—original draft. Ismail El-Hamamsy: Conceptualization; Investigation; Methodology; Project administration; Supervision; Validation; Writing—original draft. Vincent Chauvette: Data curation; Methodology; Writing—original draft. Mark D. Peterson: Conceptualization; Investigation; Methodology; Project administration; Supervision; Validation; Writing—original draft. Kevin Lachapelle: Conceptualization; Investigation; Methodology; Project administration; Supervision; Validation; Writing—original draft. Khalid Ridwan: Data curation; Methodology; Writing—original draft. Munir Boodhwani: Conceptualization; Investigation; Methodology; Project administration; Supervision; Validation; Writing—original draft. Ming Guo: Data curation; Methodology; Writing—original draft. John Bozinovski: Conceptualization; Data curation; Investigation; Methodology; Project administration; Supervision; Validation; Writing—original draft. Michael C. Moon: Conceptualization; Investigation; Methodology; Project administration; Supervision; Validation; Writing—original draft. Abigail White: Data curation; Methodology; Writing—original draft. Michael Yamashita: Conceptualization; Investigation; Methodology; Project administration; Supervision; Validation; Writing—original draft. Carly Lodewyks: Data curation; Methodology; Writing—original draft. Rony Atoui: Conceptualization; Data curation; Investigation; Methodology; Project administration; Supervision; Validation; Writing—original draft. Darrin Payne: Conceptualization; Data curation; Investigation; Methodology; Project administration; Supervision; Validation; Writing—original draft. Michael W. A. Chu: Conceptualization; Investigation; Methodology; Project administration; Resources; Supervision; Validation; Visualization; Writing—original draft.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Yutaka Okita and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

 
• CI

  Confidence interval

 
• IPTW

  Inverse probability of treatment weighting

 
• LOS

  Length of stay

 
• OR

  Odds ratio

 
• PS

  Propensity score

Author notes