Polytomous logistic regression analysis could be applied more often in diagnostic research

OBJECTIVE: Physicians commonly consider the presence of all differential diagnoses simultaneously. Polytomous logistic regression modeling allows for simultaneous estimation of the probability of multiple diagnoses. We discuss and (empirically) illustrate the value of this method for diagnostic research. STUDY DESIGN AND SETTING: We used data from a study on the diagnosis of residual retroperitoneal mass histology in patients presenting with nonseminomatous testicular germ cell tumor. The differential diagnoses include benign tissue, mature teratoma, and viable cancer. Probabilities of each diagnosis were estimated with a polytomous logistic regression model and compared with the probabilities estimated from two consecutive dichotomous logistic regression models. RESULTS: We provide interpretations of the odds ratios derived from the polytomous regression model and present a simple score chart to facilitate calculation of predicted probabilities from the polytomous model. For both modeling methods, we show the calibration plots and receiver operating characteristics curve (ROC) areas comparing each diagnostic outcome category with the other two. The ROC areas for benign tissue, mature teratoma, and viable cancer were similar for both modeling methods, 0.83 (95% confidence interval [CI]=0.80-0.85) vs. 0.83 (95% CI=0.80-0.85), 0.78 (95% CI=0.75-0.81) vs. 0.78 (95% CI=0.75-0.81), and 0.66 (95% CI=0.61-0.71) vs. 0.64 (95% CI=0.59-0.69), for polytomous and dichotomous regression models, respectively. CONCLUSION: Polytomous logistic regression is a useful technique to simultaneously model predicted probabilities of multiple diagnostic outcome categories. The performance of a polytomous prediction model can be assessed similarly to a dichotomous logistic regression model, and predictions by a polytomous model can be made with a user-friendly method. Because the simultaneous consideration of the presence of multiple (differential) conditions serves clinical practice better than consideration of the presence of only one target condition, polytomous logistic regression could be applied more often in diagnostic research.     

Genetic programming outperformed multivariable logistic regression in diagnosing pulmonary embolism

OBJECTIVE: Genetic programming is a search method that can be used to solve complex associations between large numbers of variables. It has been used, for example, for myoelectrical signal recognition, but its value for medical prediction as in diagnostic and prognostic settings, has not been documented. STUDY DESIGN AND SETTING: We compared genetic programming and the commonly used logistic regression technique in the development of a prediction model using empirical data from a study on diagnosis of pulmonary embolism. Using part (67%) of the data, we developed and internally validated (using bootstrapping techniques) a diagnostic prediction model by genetic programming and by logistic regression, and compared both on their predictive ability in the remaining data (validation set). RESULTS: In the validation set, the area under the ROC curve of the genetic programming model was significantly larger (0.73; 95%CI: 0.64-0.82) than that of the logistic regression model (0.68; 0.59-0.77). The calibration of both models was similar, indicating a similar amount of overoptimism. CONCLUSION: Although the interpretation of a genetic programming model is less intuitive and this is the first empirical study quantifying its value for medical prediction, genetic programming seems a promising technique to develop prediction rules for diagnostic and prognostic purposes.     

A review of two journals found that articles using multivariable logistic regression frequently did not report commonly recommended assumptions

BACKGROUND AND OBJECTIVE: To examine if commonly recommended assumptions for multivariable logistic regression are addressed in two major epidemiological journals. METHODS: Ninety-nine articles from the Journal of Clinical Epidemiology and the American Journal of Epidemiology were surveyed for 10 criteria: six dealing with computation and four with reporting multivariable logistic regression results. RESULTS: Three of the 10 criteria were addressed in 50% or more of the articles. Statistical significance testing or confidence intervals were reported in all articles. Methods for selecting independent variables were described in 82%, and specific procedures used to generate the models were discussed in 65%. Fewer than 50% of the articles indicated if interactions were tested or met the recommended events per independent variable ratio of 10:1. Fewer than 20% of the articles described conformity to a linear gradient, examined collinearity, reported information on validation procedures, goodness-of-fit, discrimination statistics, or provided complete information on variable coding. There was no significant difference (P>.05) in the proportion of articles meeting the criteria across the two journals. CONCLUSION: Articles reviewed frequently did not report commonly recommended assumptions for using multivariable logistic regression.     

Vital signs should be maintained as continuous variables when predicting bacterial infections in febrile children

ObjectiveTo determine how vital signs such as heart and respiratory rates should be included in prediction models for serious bacterial infections (SBIs) in febrile children.Study Design and SettingProspective observational study of 1,750 febrile children aged <16 years, visiting the emergency department of a university hospital; of them 13% (n = 222) had SBI. Common age-specific thresholds of heart and respiratory rates were used to define tachycardia and tachypnea. We compared seven strategies to handle vital signs as predictors of SBI (dichotomized or continuously in various ways).ResultsThe dichotomous predictors, namely tachycardia and tachypnea, containing information on the vital sign and age showed limited value to predict the presence of SBI (area under the receiver operating characteristic curve [AUC (ROC)]: 0.53 for heart rate and 0.55 for respiratory rate). In comparison, a model with age as a single continuous predictor resulted in an AUC of 0.58. Models with age and one of the vital signs included continuously showed the highest AUC (heart rate: 0.60 and respiratory rate: 0.63).ConclusionHeart and respiratory rates should be maintained as continuous variables in model development to predict SBI in febrile children, as dichotomization results in information loss and lower predictive ability.     

某三级医院医院感染死亡患者多因素logistic回归分析

目的探讨医院感染直接相关死亡患者的危险因素,为控制医院感染,降低医院感染病死率提供依据。方法对2010-2015年3月医院感染病例进行回顾性调查分析,采用χ2检验和logistic回归分析医院感染直接相关死亡的危险因素。结果 2010-2015年3月医院感染死亡患者304例,其中与医院感染直接相关死亡患者37例,占12.17%;感染部位以下呼吸道为主占68.29%,其次为血液占12.20%;单因素分析:ICU、使用呼吸机、年龄>75岁、长期住院、透析、免疫功能低下对医院感染相关死亡的影响因素,差异均有统计学意义;多因素分析:高龄、长期住院、免疫功能低、使用呼吸机和透析是发生医院感染相关死亡的独立危险因素。结论医院感染相关死亡患者感染部位以下呼吸道为主,预防下呼吸道感染,有助于降低感染死亡风险;应重点关注有高风险因素的患者,防范医院感染的发生,从而降低医院感染病死率。     

脑梗死患者肺部感染特点与多因素logistic回归分析

目的探讨脑梗死患者伴肺部感染特点,并对相关因素加以分析,为临床防止脑梗死患者发生肺部感染提供参考。方法选取2014年1月-2016年8月医院住院治疗的500例脑梗死患者,将未发生感染的患者作为对照组,发生肺部感染的患者为观察组;对患者临床病例资料、诊治情况等加以分析,对脑梗死患者伴肺部感染发生因素加以分析,并对有统计学意义的单因素行logistic多因素回归分析,同时对感染患者进行痰液培养,了解其病原菌分布。结果 500例脑梗死患者中发生肺部感染32例,感染率为6.4%;经痰液检测发现,32例脑梗死伴肺部感染患者共检出病原菌40株,其中革兰阴性菌28株占70.0%;革兰阳性菌9株占22.5%;真菌3株占7.5%;主要病原菌为铜绿假单胞菌、大肠埃希菌、金黄色葡萄球菌;观察组患者PH值为(7.05±0.02)明显低于对照组(7.51±0.07);PaO_2为(7.98±2.01)KPa明显低于对照组(10.36±0.71)KPa;PaCO_2为(6.19±0.57)KPa明显高于对照组(5.27±0.31)KPa;TNF-α为(0.27±0.02)μg/L明显高于对照组(0.07±0.02)μg/L;IL-6为(0.27±0.14)μg/L明显高于对照组(0.08±0.04)μg/L,以上指标两组比较,差异均有统计学意义(P<0.05);单因素分析发现,吸烟、使用呼吸机、气管插管、卧床时间、留置鼻饲管、慢性心功能不全、侵入性操作、意识障碍、年龄、糖尿病、肌力Ⅱ级以下、COPD、球麻痹为脑梗死患者伴肺部感染的影响因素;经多因素logistic回归分析发现,年龄、侵入性操作、气管插管、卧床时间、意识障碍、留置鼻饲管、使用呼吸机、糖尿病、肌力Ⅱ级以下、COPD、球麻痹为脑梗死患者伴肺部感染的高危因素。结论脑梗死患者伴肺部感染的影响因素较多,临床医师需加强对脑梗死患者病情的观察,减少侵入性操作等,制定合理有效防治对策,同时根据病原菌特点积极选择合理抗菌药物,以更好的改善患者预后。     

腹腔肿瘤患者围术期输血细菌感染的多因素logistic分析

目的 探讨分析腹腔肿瘤患者围手术期输血与术后细菌感染之间所存在的关系,为临床该类患者预防感染发生提供可靠有效依据.方法 选取医院收入治疗的112例腹腔肿瘤患者,根据是否感染分为感染组(18例)和未感染组(94例),通过单因素分析和多因素logistic分析法,分析围术期输血患者术后细菌感染的影响因素.结果 112例腹腔肿瘤患者患者中,术后细菌感染的有18例,感染率为16.1％;单因素分析显示,围术期输血患者术后感染16例,无输血患者仅感染2例;感染患者平均手术时间(243.2±82.7)min,无感染患者平均手术时间(175.6±56.3)min,两者比较差异有统计学意义(P＜0.05);其他性别、年龄、病程、术前血红蛋白、白细胞、血清白蛋白、术中失血量、手术部位、肿瘤分期等指标在有无感染患者间差异均无统计学意义;经多因素logistic分析校正后发现,输血与未输血患者发生术后细菌感染的OR值为1.246(95％CI:1.147～1.354,P＜0.05).结论 围术期输血是术后细菌感染的独立危险因素,在腹腔肿瘤患者术中应减少术中出血,以降低患者感染概率.     

113838例住院患者医院感染及危险因素的logistic回归分析

目的探讨医院感染流行趋势,确定相关危险因素及其影响程度。方法调查2002-2009年113 838例住院患者医院感染情况,对可能危险因素进行logistic回归分析。结果医院感染率为2.53%,以上呼吸道感染最高,占36.77%,其余依次为下呼吸道占24.09%、泌尿道占17.49%、胃肠道占7.00%、皮肤软组织占5.22%等;性别、年龄、住院天数、免疫抑制剂、化疗I、CU、血液病、糖尿病、肝病、精神病、抗菌药物、骨腰穿、尿路插管及动静脉插管、呼吸机、气管切开、急诊手术、其他慢性病和侵入性操作等19项因素是医院感染的独立危险因素。结论医院感染控制不平衡,但总体好转,应根据危险因素制定控制措施。     

Serotype changes in adult invasive pneumococcal infections in Portugal did not reduce the high fraction of potentially vaccine preventable infections

We determined the serotype and antimicrobial susceptibility of 1100 isolates responsible for adult invasive pneumococcal infections (IPD) in Portugal between 2006 and 2008. Serotypes 3 (13%), 1 (12%), 7F (11%), 19A (10%) and 14 (7%) were the most frequent causes of IPD and the two later serotypes accounted for the majority of erythromycin and penicillin nonsusceptible isolates. Serotype 1 was associated with younger adults whereas serotype 3 was associated with older adults. Despite the availability of the 23-valent polysaccharide vaccine (PPV23) in Portugal since 1996, the proportion of PPV23 preventable IPD remained stable and above 80%. Comparing with previous data from Portugal, we showed a continued decline of the serotypes included in the 7-valent conjugate vaccine (PCV7) in adult IPD and a rise of serotypes included in the 13-valent conjugate vaccine, increasing its potential coverage of adult IPD to 70% in 2008. Penicillin non-susceptibility remained stable (17%) whereas erythromycin resistance (18%) has continued to rise in the post-PCV7 years.     

小儿热性惊厥复发高危因素的Logistic回归分析

目的:对小儿热性惊厥复发的临床高危因素进行研究,为临床预防热性惊厥复发的发生提供可能的支持.方法:随机选取2002～ 2010年苏州大学附属儿童医院确诊为上呼吸道感染及热性惊厥复发患儿30例作为观察组,有上呼吸道感染及高热但未发生热性惊厥的患儿30例分为对照组,两组年龄及性别无统计学差异.观察两组患儿16个指标(如:性别、年龄、日托儿童等),再将阳性结果放入Logistic回归方程进行多因素分析.结果:发现观察组患儿中:日托儿童,一、二级亲属中热性惊厥史、围产期异常史、发育迟缓史、代谢性酸中毒、低血钠、低血锌、低钙、低血镁、缺铁性贫血、脑电图异常、头颅CT异常与对照组比较有统计学差异(P＜ 0.05).再将上述高危因素放入Logistic回归方程进行多因素分析,发现日托儿童,一、二级亲属中热性惊厥史、围产期异常史、低血钠、低血锌、缺铁性贫血与热性惊厥复发的发生有显著相关性.结论:日托儿童、家族史、围产期异常史、低血钠、低血锌,缺铁性贫血与热性惊厥复发有显著相关性,以上结论可进一步指导临床,预防热性惊厥复发的发生.     

Inpatient admission for febrile seizure and subsequent outcomes do not differ in children with vaccine-associated versus non-vaccine associated febrile seizures

Introduction

Recent data suggest that the risk factors for febrile seizure (FS) can differ depending on whether the FS was vaccine-associated (VA) or not. As such, there also may be differences in the risk of inpatient admission and/or the incidence of FS-related subsequent outcomes following the index FS depending on whether it was VA or non-vaccine associated (NVA). This could have useful clinical implications including caregiver education and planning for follow-up care.

Methods

This cohort study consisted of 3348 children who experienced an index FS between 6 months up to 3 years of age from July 1, 2003 through December 31, 2011. The index FS was determined to be VA-FS or NVA-FS; inpatient admission for FS, recurrent FS, and diagnosis of epilepsy were compared between exposure groups. Hazard ratios and relative risk estimates comparing between VA-FS and NVA-FS were estimated by Cox proportional models and Robust Poisson regression models, adjusted for race, sex, age at first FS, birth weight, gestational age, maternal age, and 1- and 5-min Apgar scores.

Results

The mean age at index FS was 1.5 years; the mean length of follow-up was 2.3 years. Of all index FS, 383 (11.4%) were VA and 2965 were NVA. Among index FS, 264 (7.9%) were admitted as inpatients. Subsequently, 703 (21.0%) children developed at least one recurrent FS, where the number of recurrences ranged from 0 to 9 events. Overall, 144 (4.3%) children were diagnosed with epilepsy during the follow-up period. In adjusted analyses, VA-FS did not differ in the risk for any of the outcomes of interest compared with NVA-FS.

Discussion

The risk of hospitalization for index FS or select subsequent FS outcomes did not differ between VA or NVA-FS. This suggests that the follow-up care of children with VA-FS does not warrant attention beyond that for NVA-FS.     

Increasing hospitalizations for serious skin infections in New Zealand children, 1990-2007

The incidence of serious skin infections in New Zealand children is significantly higher than in comparative countries. This study aimed to describe the epidemiology of these infections and identify changes in disease distribution over time. Discharge data were analysed for all children admitted to a New Zealand public hospital with a serious skin infection during the period 1990-2007. Patient and admission variables were compared between 1990-1999 and 2000-2007. The incidence of serious skin infections almost doubled from 298·0/100,000 in 1990 to 547·3/100,000 in 2007. The highest rates were observed in boys, preschool-aged children, Māori and Pacific children, those living in deprived neighbourhoods, urban areas and northern regions. Over time there were disproportionate increases in infection rates in Māori and Pacific children and children from highly deprived areas. Serious skin infections are an increasing problem for New Zealand children. Worsening ethnic and socioeconomic health inequalities may be contributing to increasing rates.     

肿瘤患儿化疗后细菌感染的临床分析

目的探讨肿瘤患儿化疗后细菌感染的危险因素,制定有针对性的感染干预措施。方法选取2010年7月-2011年7月医院诊治的恶性肿瘤患儿400例,观察肿瘤患儿化疗后细菌感染率及感染部位分布,分析患儿发生感染的危险因素,采用SPSS17.0软件进行统计分析。结果 400例患儿中105例发生感染,感染率为26.3%;其中呼吸道感染比例最高,占48.6%,其次为口腔感染,占17.1%;共检出病原菌105株,前3位病原菌依次为凝固酶阴性葡萄球菌、大肠埃希菌、铜绿假单胞菌,分别占35.2%、27.6%和21.0%;经单因素分析,年龄<3岁、接受侵入性操作、白细胞计数<2×109/L、中性粒细胞<0.5×109/L、住院时间>20d患儿发生感染率显著高于其他患儿,两者差异有统计学意义(P<0.05);经logistic回归分析,住院时间>20d、白细胞计数<2×109/L、中性粒细胞≤0.5×109/L及接受侵入性操作是肿瘤患儿化疗后细菌感染的危险因素。结论医院患儿化疗后细菌感染率较高,住院时间、接受侵入性操作等是导致感染的危险因素,针对上述危险因素进行有效干预,可显著降低肿瘤患儿化疗后细菌感染率。     

顺义区学龄前儿童超重/肥胖的危险因素Logistic分析

目的 了解北京市顺义区学龄前儿童超重/肥胖的流行特征及其危险因素,为儿童肥胖的预防和临床干预提供科学依据。方法 采取整群分层随机抽样的方式,于2017年9—11月抽取顺义区5所幼儿园共1 238名儿童作为研究对象。采用自行设计的调查问卷方式研究儿童肥胖的危险因素,对相关危险因素进行多因素Logistic回归分析。结果 顺义区学龄前儿童超重/肥胖发生率分别为13.3%、15.2%。学龄前儿童超重/肥胖的危险因素包括:剖宫产(OR=1.333,95%CI:1.032~1.722)、父亲超重/肥胖(OR=1.723,95%CI:1.293~2.295)、母亲超重/肥胖(OR=1.617,95%CI:1.234~2.119)、进食速度快(OR=1.403,95%CI:1.125~1.748)。而平均户外活动时间>2 h(OR=0.649,95%CI:0.549~0.771)是儿童超重/肥胖发生的保护因素。结论 学龄前儿童超重/肥胖是诸多因素共同作用的结果,应制定父母参与的综合防治措施,以减少儿童超重/肥胖的发生。     

Paradoxical health care utilization patterns among children in Korea who did not receive mandatory pneumococcal vaccination

BackgroundUnvaccinated children in the National Immunization Program (NIP) are a public health concern. We used Korean national population data to analyze health care utilization patterns of NIP-eligible children and identify the unvaccinated group.MethodsPneumococcal vaccination (PCV) records were reviewed to determine the vaccination status of children born between 2013 and 2015. Children who received three doses or more from a 3 + 1 schedule were defined as vaccinated, while those who had not received any pneumococcal vaccinations were defined as unvaccinated. Corresponding health care utilization records were retrieved from the National Health Insurance Review and Assessment Service. The incidence of combined pneumococcal infections and health care utilization rates were estimated and the proportion of complementary and alternative medicine (CAM) utilization among the total health care utilization records was measured.ResultsIn total, 26,893 (2.1%) of 1,272,685 children remained unvaccinated. The incidence of pneumococcal infection was lower in unvaccinated children, at 10.1 cases (9.8–10.3) per 1000 person-months. However, their health care utilization was significantly lower than that noted for vaccinated children (hospital visit rate: 26.6 (95% confidence interval [CI] 26.5–26.7) vs. 3.2 (3.2–3.3) visits annually), indicating underdetection. CAM treatment was sought at least three times often more in unvaccinated children than in vaccinated children (3.5% vs. 1.1%).ConclusionUnvaccinated children showed significantly lower utilization of overall health care than the vaccinated children; however, a higher preference for CAM was noted among unvaccinated children than among vaccinated children. These differences in care-seeking patterns should be considered when identifying unvaccinated children and providing protection through vaccination programs.     

多重聚合酶链反应系统用于儿童常见细菌及真菌血流感染的价值研究

目的探讨多重聚合酶链反应系统用于儿童常见细菌及真菌血流感染的价值。方法选取2014年1月-2016年1月医院住院疑诊血流感染患儿135例,采集患儿的血标本,分别采用血培养法、16SrDNA-PCR和多重聚合酶链反应系统进行检测,对比3种方法检测儿童血流感染的阳性率。结果血培养法检测阳性标本12份,阳性率为8.89%,16SrDNA-PCR检测阳性标本26份,阳性率为19.26%,多重聚合酶链反应系统检测阳性标本24份,阳性率为17.78%;16SrDNA-PCR和多重聚合酶链反应系统的阳性率均显著高于血培养法(P<0.05);16SrDNA-PCR和多重聚合酶链反应系统阳性率比较,差异无统计学意义。结论多重聚合酶链反应系统具有操作简单、检测时间短等优势,可在较短时间内鉴定儿童血流感染的常见病原菌,对临床诊疗有较好的指导意义。     

Clinical items not helpful in differentiating viral from bacterial lower respiratory tract infections in general practice

OBJECTIVE: Incorrect and unnecessary antibiotic prescribing enhancing bacterial resistance rates might be reduced if viral and bacterial lower respiratory tract infections (LRTI) could be differentiated clinically. Whether this is possible is often doubted but has rarely been studied in general practice. STUDY DESIGN AND SETTING: This was an observational cohort study in 15 general practice surgeries in the Netherlands. RESULTS: Etiologic diagnoses were obtained in 112 of 234 patients with complete data (48%). Viral pathogens were found as often as bacterial pathogens. Haemophilus (para-) influenzae was most frequently found. None of the symptoms and signs correlated statistically significantly with viral or bacterial LRTI. Erythrocyte sedimentation rate >50 (odds ratio [OR] 2.3-3.3) and C-reactive protein (CRP) >20 (OR 2.1-4.6) were independent predictors for viral LRTI and bacterial LRTI when compared with microbiologically unexplained LRTI. CONCLUSION: Extensive history-taking and physical examination did not provide items that predict viral or bacterial LRTI in adult patients in daily general practice. We could not confirm CRP to differentiate between viral and bacterial LRTI.