Increased efficiency of analyses: cumulative logistic regression vs ordinary logistic regression

Abstract– The common practice of collapsing inherently continuous or ordinal variables into two categories causes information loss that may potentially weaken power to detect effects of explanatory variables and result in Type II errors in statistical inference. The purpose of this investigation was to illustrate, using a substantive example, the potential increase in power gained from an ordinal in-stead of a dichotomous specification for an inherently continuous response. Ordinary (OLR) and cumulative logistic regression (CLR) modeling were used to test the hypothesis that the risk of alveolar bone loss over 2 years is greater for subjects with poorer control of non-insulin-dependent diabetes mellitus (NIDDM) than for those who do not have diabetes or have better controlled NIDDM. There were 359 subjects; 21 of whom had NIDDM. Analysis of main effects using OLR for the dichotomous outcome (no change in radiographic bone loss vs any change) produced parameter estimates for better control and poorer control that were not statistically significant. CLR analysis of main effects using a 4-category ordinal specification for radiographic bone loss also produced a parameter estimate for better control that was not statistically significant, but which estimated poorer control to have a significant effect. Thet of this CLR model was significantly better at P < 0.05 than that for the OLR. While an OLR model testing the interaction between age and control status did not converge after 100 iterations, the CLR interaction model converged without difficulty and estimated a significant effect for interaction between age and poorer control. Results from the CLR analysis, in contrast to the OLR model, would lead one to conclude that the risk for more severe bone loss progression after 2 years is greater in subjects with poorer controlled NIDDM and that subjects with better controlled NIDDM may not have greater risk of bone loss progression than those without diabetes. The use of an ordinal instead of a dichotomous specification for an inherently continuous response provided increased power, more precise parameter estimates, and a significantly better fitting model. In estimating parameter estimates for odds ratios or risks, it is important to consider using ordinal logistic regression where the response is inherently continuous or ordinal.     

口源性口臭检测指标的Logistic回归分析

目的本研究旨在寻找使用其他口臭检测指标替代鼻测法对口源性口臭进行检测的可能性。方法从169名受试者中筛选出56名作为研究对象,分析鼻测法与数个口臭检测指标间的Logistic回归关系。结果在健康青年人群中,口腔可挥发性含硫化合物稳定值和口腔卫生指数是口源性口臭的两大主要影响因素,利用这两项指标建立的方程对口臭进行诊断的预测符合率达到78.6%。结论综合使用数个检测指标来替代鼻测法对口源性口臭进行诊断是可行的。     

A logistic regression model for the decision to perform access surgery

Abstract Access surgery may be recommended to about 80% of patients who present with advanced forms of periodontal disease. In this report, a multivariate logistic regression analysis which incorporated several clinical parameters for each tooth examined, i.e., tooth type, furcation involvement, bleeding on probing, attachment level, probing depth, mobility and BANA test score, was conducted using generalized estimating equations (GEE). This approach identified parameters that were significantly associated at p < 0.05 level with the need for access surgery or extraction for periodontal purposes. The estimated probabilities derived from the GEE model were plotted over the complete spectrum of operating conditions to obtain a receiver-operator characteristic (ROC) curve. At a probability cutpoint of 0.8, the decision threshold for surgery/extraction at the pre-treatment examination would have a sensitivity of 76.1% and a specificity of 75.3%. We have taken this 0.8 cut point to look at specific clinical decisions made by our examiners after the patients had received scaling and root planing plus 2 weeks unsupervised usage of systemic antimicrobials. The clinicians' decision was taken as the primary reference standard. The model's estimated decision agreed with the clinicians' decision in 226 of the 284 teeth, for an accuracy of 80%. The specificity was 90% and the sensitivity was 43%.     

安氏Ⅲ类错畸形病因的logistic分析

目的研究引起安氏Ⅲ类错牙合畸形的病因。方法对50例安氏Ⅲ类错牙合患者和50例正常牙合人作病因问卷调查,将结果用logistic法分析,提取有效病因。结果共有慢性扁桃体炎、遗传因素、咬上唇3项病因进入方程。结论按贡献大小,长期慢性扁桃体炎、经常咬上唇和遗传因素是导致安氏Ⅲ类错牙合畸形的危险因素。     

安氏Ⅲ类错(牙合)畸形病因的logistic分析

目的 研究引起安氏Ⅲ类错（牙合）畸形的病因。方法 对50例安氏Ⅲ类错（牙合）患者和50例正常（牙合）人作病因问卷调查，将结果用logistic法分析，提取有效病因。结果 共有慢性扁桃体炎、遗传因素、咬上唇3项病因进入方程。结论 按贡献大小，长期慢性扁桃体炎、经常咬上唇和遗传因素是导致安氏Ⅲ类错（牙合）畸形的危险因素。     

婴幼儿猛性龋危险因素的Logistic分析

目的：探讨引起婴幼儿猛性龋的危险因素。方法：龋病危险因素调查和菌斑ｐＨ检测,应用Ｌｏｇｉｓｔｉｃ回归分析的方法研究婴幼儿猛性龋与危险因素之间的数量关系。结果：喂养时间、奶瓶内容、甜食习惯和菌斑ｐＨ值４项是婴儿猛性龋的危险因素。婴幼儿猛性龋的发病情况与用Ｌｏｇｉｓｔｉｃ回顾分析结果的判断情况一致率为９６．２％,结论：用此４项来测婴儿猛性龋比单纯用细菌学参数更为准确。     

正畸治疗后复发危险因素的非条件logistic回归分析

目的 探讨正畸治疗后复发的危险因素,为预防正畸治疗后复发提供依据。方法 回访2006年结束正畸治疗的患者128例,对性别、患者依从性、牙根位置颊舌向、矫治结束时年龄、牙槽骨高度和牙周膜炎症、矫治结束后医生对患者的管理、保持器类型、矫治结束时的咬合关系、矫治结束时的牙根情况,以及矫治时间10个因素进行单因素分析,再把分析结果有意义的因素进行多因素非条件logistic回归分析。结果 回归分析显示3个因素相关：矫治结束后医生对患者的管理（OR=7.77）、矫治结束时的咬合关系（OR=4.54）、患者依从性（OR=3.53）。结论 矫治结束后医生对患者合理管理、矫治结束时咬合关系稳定、患者依从性好有利于降低正畸治疗后复发的风险。     

Clinical evaluation of the temporomandibular joint following orthognathic surgery--multiple logistic regression analysis

This study compares temporomandibular joint dysfunction (TMD) symptoms before and after bilateral sagittal split ramus osteotomy, and identifies predictive factors for the postoperative TMD symptoms by assessing the adjusted odds ratio using multiple logistic regression analysis. A consecutive series of 37 cases treated only with bilateral sagittal split ramus osteotomy were evaluated. New postoperative TMD symptoms appeared in 9 cases, preoperative TMD symptoms disappeared in 6 cases, and TMD symptoms were unchanged in 5 cases. The median period until the interincisal opening range attained 40 mm was 5 months (range, from 2 to 15 months). Age was a positive factor in patients with postoperative TMD symptoms, with an odds ratio of 1.43 (95 percent confidence interval, from 1.05 to 1.93). In addition, the maximum value of the bilateral setback distance of more than 9 mm was a positive factor of 6.95 (95 percent confidence interval, from 1.06 to 45.42). We concluded that surgical correction in skeletal malocclusion may affect temporomandibular joint dysfunction symptoms.     

乳牙反相关危险因素的Logistic回归分析

目的通过对儿童乳牙反畸形各种可能致病因素的逐步回归分析,探讨引起乳牙反畸形的主要危险因素。方法对100例乳牙反畸形患儿和100例正常儿童做问卷调查,将结果用Logistic回归法分析,提取有效因素,研究乳牙反畸形与相关危险因素之间的关系。结果按其贡献大小,共有喂养方式、家族史及咬上唇三种危险因素进入方程。结论喂养方式、家族史及咬上唇是导致乳牙反畸形的主要危险因素。     

���������ڼ���ʹΣ������Logistic�ع����

目的探讨根管治疗期间疼痛的危险因素,为提高临床疗效提供依据。方法2006年1月至2009年3月于泸州医学院附属隆昌医院对426颗患牙行根管治疗,对可能是根管治疗期间疼痛的危险因素进行Logistic回归分析。结果根管治疗期间疼痛的发生率为19.25%,通过Logistic回归分析发现:操作不慎、根管超充、封药不当以及牙位等是其危险因素。结论针对上述危险因素,应尽早采取防治措施,提高根管治疗的疗效和安全性。     

Variation in torus mandibularis prevalence in Norway A statistical analysis using logistic regression

The prevalence of torus mandibularis was assessed in two groups of dental patients, altogether 2010 individuals over 10 yr of age: 1181 individuals native to the Lofoten Islands in North Norway, situated at 68 degrees latitude; and 829 patients indigenous to Gudbrandsdal, an inland district in the southeastern part of the country at 61 degrees latitude. Both groups were supposed to be of the same Caucasian stock and, therefore, to have similar genetic predisposition to torus on the average. The following observations were found: 1) the prevalence of torus mandibularis was much greater in Gudbrandsdal than in Lofoten (P much less than 0.001); 2) the prevalence decreased among persons above 50 yr of age as compared with those of the age classes 10-49 yr (P less than 0.01); and 3) it was smaller among women than men (P less than 0.05), mainly due to such a decrease in Lofoten. In a recent investigation of people living in Gudbrandsdal the fraction of the variation of torus that was attributable to genetic differences was estimated as about 30%, whereas approximately 70% of the causes seemed to be ascribable to environmental influences in terms of occlusal stress. It is suggested that dietary habits and number of existing teeth seemed to be environmental variables with an influence on the observed variation of torus prevalence between geographical regions, age classes, and sexes. The question about a possible sexual difference as to the genetic component of liability to torus mandibularis was outside the scope of the present study.     

Risk factors for temporomandibular disorder: Binary logistic regression analysis

Objectives: To analyze the influence of socioeconomic and demographic factors (gender, economic class, age and marital status) on the occurrence of temporomandibular disorder. Study Design: One hundred individuals from urban areas in the city of Recife (Brazil) registered at Family Health Units was examined using Axis I of the Research Diagnostic Criteria for Temporomandibular Disorders (RDC/TMD) which addresses myofascial pain and joint problems (disc displacement, arthralgia, osteoarthritis and oesteoarthrosis). The Brazilian Economic Classification Criteria (CCEB) was used for the collection of socioeconomic and demographic data. Then, it was categorized as Class A (high social class), Classes B/C (middle class) and Classes D/E (very poor social class). The results were analyzed using Pearson’s chi-square test for proportions, Fisher’s exact test, nonparametric Mann-Whitney test and Binary logistic regression analysis. Results: None of the participants belonged to Class A, 72% belonged to Classes B/C and 28% belonged to Classes D/E. The multivariate analysis revealed that participants from Classes D/E had a 4.35-fold greater chance of exhibiting myofascial pain and 11.3-fold greater chance of exhibiting joint problems. Conclusions: Poverty is a important condition to exhibit myofascial pain and joint problems. Key words:Temporomandibular joint disorders, risk factors, prevalence.     

Multiple logistic regression analysis of risk factors for carcinogenesis of oral submucous fibrosis in mainland China

Oral submucous fibrosis (OSF), a chronic progressive disorder, is regarded as the premalignant lesion of oral squamous cell carcinoma (OSCC). Its distribution is associated with chewing betel quid (BQ). The objective of the present study was to investigate risk factors for the carcinogenesis of OSF in mainland China. A case-control study was performed in 42 patients with OSCC that originated from OSF and 40 OSF controls. Epidemiological data and information related to risk factors were collected using a short structured questionnaire. Odd ratios (OR) and 95% confidence intervals (CI) were derived from logistic regression analysis. In multivariate analysis, only age, duration of BQ chewing, duration of cigarette smoking, and OSF accompanied by oral leukoplakia or oral lichen planus were associated with significantly increased risk for the malignant transformation of OSF. These findings contribute to current knowledge on the prevention of carcinogenesis of OSF in mainland China.     

Dental malocclusion is not related to temporomandibular joint clicking: a logistic regression analysis in a patient population

Objectives:To assess the association of several dental malocclusion features with temporomandibular joint (TMJ) click sounds in a population of temporomandibular disorder (TMD) patients.Materials and Methods:Four hundred forty-two TMD patients (72% female; 32.2 ± 5.7 years, range 25–44 years) were divided into a TMJ clicking and a no-TMJ clicking group, based on the Research Diagnostic Criteria for Temporomandibular Disorders (RDC/TMD) assessment. Seven occlusal features were recorded for each patient: (1) posterior crossbite, (2) overbite, (3) open bite, (4) overjet, (5) mediotrusive and (6) laterotrusive interferences and (7) retruded contact position to maximum intercuspation (RCP-MI) slide length. A logistic regression model was created to estimate the association of occlusal features with TMJ clicking.Results:The difference between the groups as for the prevalence of the various occlusal features was generally not statistically significant, with minor exceptions. Mediotrusive interferences (P  =  .015) and RCP-MI slide ≥2 mm (P  =  .001) were the two occlusal features that were associated with the probability of having TMJ clicking, even if the adjusted odds ratios for TMJ clicking were low for both variables (1.63 and 1.89, respectively). Moreover, the amount of variance in the prevalence of TMJ clicking that was predicted by the final model was as low as 4.5% (R²  =  0.045).Conclusions:Findings from the present investigation suggested that in a population of TMD patients, the contribution of dental malocclusion features to predict TMJ click sounds is minimal with no clinical relevance.     

A logistic regression model for analyzing the relation between dentists' attitudes,behavior, and knowledge in oral radiology

The aim was to study the relation between risk attitude and knowledge in technical, patient-oriented, and organizationally related behavior within oral radiology. A questionnaire was mailed to 2000 randomly selected dentists listed in the register of the Swedish Dental Society, with a response rate of 69.3%. Regression analysis was used for analyzing the effects of the independent variables knowledge, risk attitude, continuing education in oral radiology, counties with specialists in oral radiology, type of practice, work experience, and sex on three categories of dependent variables: 1) technical behaviors: type of film, type of collimator, dose level, frequency of change of chemicals; 2) patient-oriented behaviors: use of patient protection barriers, strict indications for performing full-mouth X-ray examinations and bitewing radiography on new patients and recall patients; and 3) organizationally related behaviors: delegation of X-ray examinations to dental auxiliaries, influence on choice of collimator, influence on choice of film. Knowledge and education had strong direct effects for most of the dependent variables. The technical behaviors were mainly influenced by knowledge, education, and risk attitude, while organizationally related behaviors were influenced by type of practice and sex. The patient-oriented behaviors were influenced by a number of independent variables, such as education, type of practice, work experience, and sex. The present results indicate that both knowledge and the organizational context of dentists influence work.     

Correlation of factors associated with postoperative infection in patients with malignant oral and maxillofacial tumours: a logistic regression analysis

Patients with oral and maxillofacial malignant tumours have a relatively high incidence of postoperative infections. We have analysed the risk factors and the distribution of pathogens in infected patients to try and find out how to prevent them. We recruited 312 patients over 60 years old with malignant oral and maxillofacial tumours and investigated their incidence of postoperative infection between January 2007 and December 2017. Some factors were included in one way and multifactorial logistic analyses to find out which risk factors were likely to be associated with postoperative infections. Thirty-nine of the 312 (12.5%) developed postoperative infections, mainly at the surgical site (n = 21) and in the lower respiratory tract (n = 14). Pathogens were isolated from all patients and comprised aerobic (n = 4) and anaerobic (n = 5) bacteria, together with mixed infections (n = 30). Univariate analysis showed that age, size and site of tumour, body-mass index (BMI), diabetes, and duration of operation were risk factors. Multivariate logistic analyses, however, showed that BMI, the presence of diabetes, duration of operation, and American Society of Anesthesiologists (ASA) score were the main independent risk factors for postoperative infection. We conclude that there are abundant risk factors for postoperative infection in these patients, and it is important that they are evaluated preoperatively so that suitable postoperative treatments can be given.