Perbandingan Analisis Regresi Logistik dengan Analisis Propensity Score Matching pada Studi Kasus Imunisasi Bayi

Analisis multivariat konvensioanal tidak selalu merupakan metode ideal untuk memprediksi efek pajanan pada studi-studi observasional. Ketika distribusi kovariat antara kelompok pajanan berbeda besar, penyesuaan dengan teknik multivariat konvensioanl tidak cukup menyeimbangkan kelompok tersebut. Bias yang tersisa dapat menghambat penarikan kesimpulan yang valid. Tujuan penelitian ini adalah membandingkan hasil analisis multivariat konvensional dengananalisis metoda propensity score matching pada studi kasus data sekunder imunisasi bayi ASUH KAP2 2003. Penelitian ini menemukan nilai OR metoda regresi logistik (0,99) berbeda dengan metoda propensity score matching (0,96). Metoda propensity score matching berhasil menjodohkan 574 subjek(68,27%). Untuk evaluasi pengaruh faktor risiko disarankan menggunakan model PSM karena mengurangi bias seleksi, tetapi untuk analisis faktor determinan yang banyak variabel independent, gunakan matching kerena variabel tersebut mempunyai posisi yang sama.Kata kunci : Regresi logistik, propensity score matching.AbstractConventional multivariable analyses may not always be the ideal method for estimating exposure effects in observational studies. Where there are large differences in the distribution of covariates between expose groups, adjusting with conventional multivariable techniques may not adequately balance the groups, and the remaining bias may limit valid causal inference. The objective of this research is to compare the result of convensional multiariate analysis versus propensity score matching analysis in case study of infant immunization using secondary data of ASUH KAP2 2003. Model will be compared without interaction variable. The results show that the OR from logistic regression (0,99) differs to propensity score matching (0,96). Propensity score matching is successfulin matching 574 subjects (68,27%). It is recommended to evaluate risk factor effect using PSM model, but to use logistic regression analysis for determinat factor analysis with many independent variables because the variables have the same position.Keywords: Logistic regression, propensity score matching.

Download Full-text

More metalwork removals in patients with olecranon fracture treated by tension band wiring than plate fixation-a propensity score matching analysis

BMC Musculoskeletal Disorders ◽

10.1186/s12891-021-04559-0 ◽

2021 ◽

Vol 22 (1) ◽

Author(s):

Longhai Qiu ◽

Yi Li ◽

Hongbo Wu ◽

Ruixiong Chen ◽

Zhiwen Zhang ◽

...

Keyword(s):

Logistic Regression ◽

Propensity Score ◽

Propensity Score Matching ◽

Plate Fixation ◽

Treatment Method ◽

Tension Band ◽

Tension Band Wiring ◽

Olecranon Fracture ◽

Propensity Score Matching Analysis ◽

Olecranon Fractures

Abstract Background Traditional tension band wiring and plate fixation represent the commonest methods for treating olecranon fractures. However, there is no agreement on which method provides the best outcome. The aim of this retrospective study is to compare the outcomes of tension band wiring (TBW) and plate fixation (PF) for treating displaced olecranon fractures. This is the first study to use propensity score matching analysis to compare treatment methods for olecranon fracture. Method A total of 107 patients aged between 18 and 85 had acute isolated and displaced olecranon fractures. The patients were divided into either TBW (n = 49) or PF (n = 58) groups. To conduct propensity score matching for the treatment method (TBW versus PF), 58 patients were analyzed by logistic regression (29 patients in each group). Various demographic and treatment-related variables were examined and analyzed to determine their correlation. Results Functional effects between two groups are similar (in terms of Mayo Elbow Performance Score (MEPS), the patients’ range of elbow motion (ROM) and forearm rotation (RFR), the time return to work (RTW)). The total adverse events rate and metalwork removal events rate are higher in TBW than that in PF. After propensity score matching analysis, similar primary treatment efficacy (indicated by MEPS> 90) in 2 groups and more primary adverse events (indicated by metalwork removal) were perceived in TBW than that in PF. Logistic regression analysis revealed that fracture type was an independent factor that affected the efficacy of a treatment (regression coefficient = − 1.24 < 0, P = 0.03), indicating that fracture severity was inversely proportional to the efficacy of a treatment for olecranon fracture. Furthermore, logistic regression analysis demonstrated that the treatment method was an independent factor that affected metalwork removal of olecranon fracture (regression coefficient 2.38 > 0, OR = 10.77, P < 0.01), indicating that the risk of metalwork removal in the TBW Group was 10.77 times that in the PF Group. Conclusion When initially discussing the surgical approach with patients, physicians should fully weigh the possibility that TBW may lead to a second surgery due to the higher risk of internal fixation removal and that TBW won’t yield better functional outcomes than PF .

Download Full-text

More Metalwork Removals of Olecranon Fractures After Tension Banding Wiring than Plate Fixation-A Propensity Score Matching Analysis

10.21203/rs.3.rs-154380/v1 ◽

2021 ◽

Author(s):

Yi Li ◽

Longhai Qiu ◽

Hongbo Wu ◽

Ruixiong Chen ◽

Zhiwen Zhang ◽

...

Keyword(s):

Logistic Regression ◽

Propensity Score ◽

Propensity Score Matching ◽

Regression Coefficient ◽

Plate Fixation ◽

Treatment Method ◽

Olecranon Fracture ◽

Independent Factor ◽

Propensity Score Matching Analysis ◽

Olecranon Fractures

Abstract Background: Traditional tension band wiring and plate fixation represent the most common methods for treating olecranon fractures. However, there is no agreement as to which method provides the best outcome. The aim of this retrospective study was to compare the outcomes of tension band wiring (TBW) and plate fixation (PF) for treating displaced olecranon fractures. This is the first study to use propensity score matching analysis to compare treatment methods for olecranon fracture.Method: A total of 107 patients aged between 18 and 85 years old had acute isolated and displaced olecranon fractures. The patients were divided into either TBW (n=49) or PF (n=58) groups. To conduct propensity score matching for the treatment method (TBW versus PF), 58 patients were analyzed by logistic regression (29 patients in each group). Various demographic and treatment-related variables were examined and analyzed to determine their correlation. Results: Functional effects between two groups are similar( in terms of Mayo Elbow Performance Score (MEPS), the patients’ range of elbow motion (ROM) and forearm rotation (RFR), the time return to work (RTW)). The total adverse events rate and metalwork removal events rate are higher in TBW than that in PF. After propensity score matching analysis, the similar primary treatment efficacy (indicated by excellent MEPS) and higher primary adverse events rate ( indicated by metalwork removal) were perceived in TBW than that in PF. Logistic regression analysis revealed that fracture type was an independent factor that affected the efficacy of a treatment (regression coefficient=-1.24<0, P=0.03), indicating that fracture severity was inversely proportional to the efficacy of a treatment for olecranon fracture. Furthermore, logistic regression analysis demonstrated that the treatment method was an independent factor that affected metalwork removal of olecranon fracture (regression coefficient 2.38 > 0, OR = 10.77, P < 0.01), indicating that the risk of metalwork removal in the TBW Group was 10.77 times that in the PF Group.Conclusion: The results indicated that plate fixation should be the first choice of treatment method for olecranon fractures, especially in younger patients with higher quality of life requirements.Trial regislation: The study was not registered in the clinical trial registration center, because it was a retrospective case-control study, which was an observational study with no intervention measures.Level of evidence: Level III; Retrospective Cohort Comparison; Treatment Study

Download Full-text

Comparison of the survival and neurological outcomes in OHCA based on smoking status: investigation of the existence of the smoker's paradox

10.22514/sv.2021.231 ◽

2021 ◽

Keyword(s):

Logistic Regression ◽

Propensity Score ◽

Propensity Score Matching ◽

Smoking Status ◽

Secondary Outcome ◽

Neurological Outcomes ◽

Propensity Score Matching Analysis ◽

Multivariable Logistic Regression ◽

Stratification Method ◽

Never Smokers

The smoker’s paradox has been reported to reduce mortality following out-of-hospital cardiac arrest (OHCA). However, recent studies on this paradox have reported contradictory findings, with some indicating that it does not exist. Therefore, the purpose of this study was to evaluate the association between smoking status and OHCA outcomes. This retrospective observational study was conducted using multicenter registry data. The associations between smoking status and OHCA outcomes were assessed using multivariable logistic regression analyses and propensity score-adjusted methods. We compared outcomes among current, former, and never-smokers, as well as between current and non-smokers and between ever- and never-smokers. The primary outcome was survival to hospital discharge, and the secondary outcome was favourable neurological outcomes. Among 4443 patients with OHCA, 19.9% were current smokers, 15.2% were former smokers, and 64.9% were never-smokers. Current smokers had significantly better outcomes than former or never-smokers. However, the significant differences observed in univariable analysis or before propensity score matching were not observed after adjustments with multivariable logistic regression or after propensity score matching analysis in both current vs. non-smokers and ever- vs. never-smokers. Other propensity score adjusted models also did not show significant differences, except for the stratification method. This study suggests that smoking is not an independent prognostic factor for OHCA. The statistically significant better outcomes observed in current or ever-smokers were not maintained after adjusting for confounders. Therefore, the smoker’s paradox should be investigated in additional prospective studies.

Download Full-text

Abstract 13351: Association of Atrial Fibrillation With In-hospital Mortality Among Patients With Takotsubo Cardiomyopathy

Circulation ◽

10.1161/circ.142.suppl_3.13351 ◽

2020 ◽

Vol 142 (Suppl_3) ◽

Author(s):

Qiying Dai ◽

Abhishek Bose ◽

Pengyang Li ◽

PENG CAI ◽

Douglas Laidlaw

Keyword(s):

Atrial Fibrillation ◽

Logistic Regression ◽

Mortality Rate ◽

Propensity Score ◽

Hospital Mortality ◽

Propensity Score Matching ◽

Mortality Rates ◽

Primary Diagnosis ◽

Propensity Score Matching Analysis ◽

Significant Difference

Introduction: Takotsubo cardiomyopathy (TC), characterized by transient left ventricular dysfunction, is known to be precipitated by sudden physical or emotional stress. Atrial fibrillation (AF) is a common arrhythmia and its presence increases mortality in patients with chronic underlying diseases. Furthermore, multiple studies have suggested that in TC patients, underlying AF correlates with a higher in-hospital mortality. However, these studies were constrained by either a disproportionate percentage of AF patients or lack of risk factor matching. To systematically explore the association between TC and AF, we decided to perform a propensity score matching analysis. Methods: We performed a retrospective cohort analysis using the ICD-10 codes for a primary diagnosis of TC from the 2016 National Inpatient Sample. To adjust for underlying risk factors a multivariable logistic regression model was employed followed by a propensity score matching analysis for the AF group and the non-AF group. A similar analysis method was followed for the subset of patients with paroxysmal AF. In-hospital mortality rates were compared between the two groups. Results: Out of the total 3139 patients with a primary diagnosis of TC, 433 (13.7%) had AF and 243 (7.7%) had paroxysmal AF. In the unmatched group, patients with AF appeared to be older (74.7 vs. 65.3 years, P<0.001) and more likely to have comorbidities like diabetes mellitus (24.0% vs 19.2%, P=0.024), chronic kidney disease (15.7% vs 7.6%, P<0.001) and obstructive sleep apnea (8.8% vs 4.9%, P=0.002). On multivariable logistic regression, in-hospital mortality was higher in the AF group (3.9% vs 1.3%, P< 0.001). However, after propensity score matching, there was no significant difference in the mortality rate between the two groups (3.7% vs 1.9%, P=0.082). Similarly, on logistic regression followed by propensity matching for patients with paroxysmal AF, no significant difference was noted for in-hospital mortality rates (3.3% vs 1.3%, P=0.112). Conclusions: In patients with TC, the presence of underlying AF had no effect on the in-hospital mortality rate.

Download Full-text

The Relationship Between AKI in Patients With STEMI and Short-Term Mortality: A Propensity Score Matching Analysis

Angiology ◽

10.1177/0003319721998567 ◽

2021 ◽

pp. 000331972199856

Author(s):

Zhongyuan Meng ◽

Yaxin Zhao ◽

Xifeng Zheng ◽

Yan He

Keyword(s):

Logistic Regression ◽

Propensity Score ◽

Propensity Score Matching ◽

Regression Models ◽

Medical Information ◽

Short Term ◽

Long Term Outcome ◽

Logistic Regression Models ◽

Propensity Score Matching Analysis ◽

Short Term Mortality

Acute myocardial infarction (AMI) in patients with acute kidney injury (AKI) is associated with poor long-term outcome. However, the short-term prognosis of AKI in patients with ST-elevation AMI (STEMI) needs to be explored further. We assessed this relationship between these patients and short-term mortality in relation to AKI and chronic kidney disease (CKD). All data were extracted from the Medical Information Mart for Intensive Care III database. The primary outcome was 28-day mortality. Kaplan-Meier curves, logistic regression models, and propensity score matching analysis were used to evaluate the associations between AKI in patients with STEMI and outcomes. A total of 1031 patients with STEMI met the inclusion criteria. For 28-day mortality, in the multivariable logistic regression models, the odds ratio (95% CI) of group 2 (AKI but no CKD) and group 3 (AKI in the presence of CKD) were 3.24 (1.46-7.18) and 4.57 (1.83-11.37), respectively, compared with group 1 (no AKI and no CKD). Comorbid AKI increased the risk of short-term mortality among patients with STEMI, especially for those with AKI in the presence of CKD.

Download Full-text