Automatic Interaction Detection Decision Tree Research Articles

Clinical prediction rule for bacterial arthritis: Chi-squared automatic interaction detector decision tree analysis model.

Differences in demographic factors, symptoms, and laboratory data between bacterial and non-bacterial arthritis have not been defined. We aimed to identify predictors of bacterial arthritis, excluding synovial testing. This retrospective cross-sectional survey was performed at a university hospital. All patients included received arthrocentesis from January 1, 2010, to December 31, 2020. Clinical information was gathered from medical charts from the time of synovial fluid sample collection. Factors potentially predictive of bacterial arthritis were analyzed using the Student's t-test or chi-squared test, and the chi-squared automatic interaction detector decision tree analysis. The resulting subgroups were divided into three groups according to the risk of bacterial arthritis: low-risk, intermediate-risk, or high-risk groups. A total of 460 patients (male/female = 229/231; mean ± standard deviation age, 70.26 ± 17.66 years) were included, of whom 68 patients (14.8%) had bacterial arthritis. The chi-squared automatic interaction detector decision tree analysis revealed that patients with C-reactive protein > 21.09 mg/dL (incidence of septic arthritis: 48.7%) and C-reactive protein ⩽ 21.09 mg/dL plus 27.70 < platelet count ⩽ 30.70 × 104/μL (incidence: 36.1%) were high-risk groups. Our results emphasize that patients categorized as high risk of bacterial arthritis, and appropriate treatment could be initiated as soon as possible.

Radiomics analysis of amide proton transfer-weighted and structural MR images for treatment response assessment in malignant gliomas.

The purpose of this study was to evaluate the value of amide proton transfer-weighted (APTw) MRI radiomic features for the differentiation of tumor recurrence from treatment effect in malignant gliomas. Eighty-six patients who had suspected tumor recurrence after completion of chemoradiation or radiotherapy, and who had APTw-MRI data acquired at 3 T, were retrospectively analyzed. Using a fluid-attenuated inversion recovery (FLAIR) image-based mask, radiomics analysis was applied to the processed APTw and structural MR images. A chi-square automatic interaction detector decision tree was used for classification analysis. Models with and without APTw features were built using the same strategy. Tenfold cross-validation was applied to obtain the overall classification performance of each model. Sixty patients were confirmed as having tumor recurrence, and the remainder were confirmed as having treatment effect, at median time points of 190 and 171 days after therapy, respectively. There were 525 radiomic features extracted from each of the processed APTw and structural MR images. Based on these, the APTw-based model yielded the highest accuracy (86.0%) for the differentiation of tumor recurrence from treatment effect, compared with 74.4%, 76.7%, 83.7%, and 76.7% for T1 w, T2 w, FLAIR, and Gd-T1 w, respectively. Model classification accuracy was 82.6% when using the combined structural MR images (T1 w, T2 w, FLAIR, Gd-T1 w), and increased to 89.5% when using these structural plus APTw images. The corresponding sensitivity and specificity were 85.0% and 76.9% for the combination of structural MR images, and 85.0% and 100% after adding APTw image features. Adding APTw-based radiomic features increased MRI accuracy in the assessment of the treatment response in post-treatment malignant gliomas.

Chi-square automatic interaction detector decision tree analysis model: Predicting cefmetazole response in intra-abdominal infection

BackgroundCefmetazole is used as the first-line treatment for intra-abdominal infections. However, only a few studies have investigated the risk factors for cefmetazole treatment failure. AimsThis study aimed to develop a decision tree-based predictive model to assess the effectiveness of cefmetazole in initial intra-abdominal infection treatment to improve the clinical treatment strategies. MethodsThis retrospective cohort study included adult patients who were unexpectedly hospitalized due to intra-abdominal infections between 2003 and 2020 and initially treated with cefmetazole. The primary outcome was clinical intra-abdominal infection improvement. The chi-square automatic interaction detector decision tree analysis was used to create a predictive model for clinical improvement after cefmetazole treatment. ResultsAmong 2,194 patients, 1,807 (82.4%) showed clinical improvement post-treatment; their mean age was 48.7 (standard deviation: 18.8) years, and 1,213 (55.3%) patients were men. The intra-abdomせinal infections were appendicitis (n = 1,186, 54.1%), diverticulitis (n = 334, 15.2%), and pancreatitis (n = 285, 13.0%). The chi-square automatic interaction detector decision tree analysis identified the intra-abdominal infection type, C-reactive protein level, heart rate, and body temperature as predictive factors by categorizing patients into seven groups. The area under the receiver operating characteristic curve was 0.71 (95% confidence interval: 0.68–0.73). ConclusionThis predictive model is easily understandable visually and may be applied in clinical practice.

Predictors of Quadriceps Strength Asymmetry after Anterior Cruciate Ligament Reconstruction: A Chi-Squared Automatic Interaction Detection Decision Tree Analysis.

The influence of graft type on the restoration of quadriceps strength symmetry after ACL reconstruction (ACLR) has been widely studied. However, an important consideration when evaluating quadriceps symmetry is the fact that this measure can be influenced by numerous factors beyond graft type. This study sought to determine if graft type is predictive of quadriceps strength asymmetry during the first 12 months post-ACLR taking into consideration potentially influential factors (i.e., age, sex, body mass index, time post-ACLR). We retrospectively reviewed quadriceps strength data from 434 patients (303 female patients and 131 male patients) who had previously undergone ACLR with an autograft (hamstring tendon, quadriceps tendon [QT], patellar tendon [PT]) or allograft. Chi-Squared Automatic Interaction Detection decision tree analysis was used to evaluate if graft type is predictive of quadriceps strength asymmetry during the first 12 months post-ACLR taking into consideration age, sex, body mass index, and time post-ACLR. The best predictor of quadriceps strength asymmetry was graft type. Specifically, three graft categories were identified: 1) allograft and hamstring tendon autograft, 2) PT autograft, and 3) QT autograft. The average quadriceps strength asymmetry for each of the three identified categories was 0.91, 0.87, and 0.81, respectively, and differed statistically from each other ( P < 0.001). The second-best predictor of quadriceps strength asymmetry was sex, albeit only in the PT and QT groups (with female patients having increased asymmetry). Female patients post-ACLR with a QT autograft were at highest risk for quadriceps strength asymmetry. Graft type and sex are important predictors of quadriceps strength asymmetry after ACLR. Clinicians should take these factors into consideration when designing rehabilitation protocols to restore quadriceps strength symmetry during the postoperative period.

Subjective assessment underestimates surgical risk: On the potential benefits of cardiopulmonary exercise testing for open thoracoabdominal repair.

BackgroundInitial clinical evaluation (ICE) is traditionally considered a useful screening tool to identify frail patients during the preoperative assessment. However, emerging evidence supports the more objective assessment of cardiorespiratory fitness (CRF) via cardiopulmonary exercise testing (CPET) to improve surgical risk stratification. Herein, we compared both subjective and objective assessment approaches to highlight the interpretive idiosyncrasies.MethodsAs part of routine preoperative patient contact, patients scheduled for major surgery were prospectively “eyeballed” (ICE) by two experienced clinicians before more detailed history taking that also included the American Society of Anesthesiologists score classification. Each patient was subjectively judged to be either “frail” or “not frail” by ICE and “fit” or “unfit” from a thorough review of the medical notes. Subjective data were compared against the more objective validated assessment of postoperative outcomes using established CPET “cut‐off” metrics incorporating peak pulmonary oxygen uptake, V̇O2PEAK at the anaerobic threshold (V̇O2‐AT), and ventilatory equivalent for carbon dioxide that collectively informed risk stratification. These data were retrospectively extracted from a single‐center prospective National Health Service database. Data were analyzed using the Chi‐square automatic interaction detection decision tree method.ResultsA total of 127 patients were examined that comprised 58% male and 42% female patients aged 69 ± 10 years with a body mass index of 29 ± 7 kg/m2. Patients were poorly conditioned with a V̇O2PEAK almost 20% lower than predicted for age, sex‐matched healthy controls with 35% exhibiting a V̇O2‐AT < 11 ml/kg/min. Disagreement existed between the subjective assessments of risk with ∼34% of patients classified as not frail on ICE were considered unfit by notes review (p < .0001). Furthermore, ∼35% of patients considered not frail on ICE and ∼31% of patients considered fit by notes review exhibited a V̇O2‐AT < 11 ml/kg/min, and of these, ∼28% and ∼19% were classified as intermediate to high risk.ConclusionsThese findings highlight the interpretive limitations associated with the subjective assessment of patient frailty with surgical risk classification underestimated in up to a third of patients compared to the validated assessment of CRF. They reinforce the benefits of a more objective and integrated approach offered by CPET that may help us to improve perioperative risk assessment and better direct critical care provision in patients scheduled for “high‐stakes” surgery including open thoracoabdominal aortic aneurysm repair.

Developing a Process for Deploying Systemic Countermeasures at Unsignalized Intersections: Case Study of Virginia

This paper discusses the process used to develop a safety improvement plan for unsignalized intersections using systemic low-cost countermeasures. The scope of this project focused on unsignalized intersections with stop sign control on the minor approaches. The first objective was to perform an assessment of Virginia’s unsignalized intersection crashes over a five-year period to determine predominant crash trends and collision types to target for treatment. The four focus collision types with the highest frequency of crashes and the greatest potential reduction in crashes were 3-leg angle, 3-leg fixed object off the road, 4-leg angle and 4-leg rear end. Chi-square automatic interaction detection decision tree analysis was used to perform a systemic analysis to identify a group of intersections associated with potential risk factors related to the focus collision types. A tiered list of systemic countermeasures to deploy was developed. The countermeasures were intended to warn of the stop ahead, make the stop sign and stop location more visible on a minor street, and to warn of the intersection ahead on a major street. The potential for safety improvement measure was used to prioritize the candidate treatment intersections. Before deployment, a study of the intersection by district traffic engineering staff was planned to finalize the plan. The output from the research was a safety improvement plan to systemically deploy treatments to unsignalized intersections as part of the safety program.

Ski injury predictive analytics from massive ski lift transportation data

Ski injury research is traditionally studied on small-scale observational studies where risk factors from univariate and multivariate statistical models are extracted. In this article, a large-scale ski injury observational study was conducted by analyzing skier transportation data from six consecutive seasons. Logistic regression and chi-square automatic interaction detection decision tree models for ski injury predictions are proposed. While logistic regression assumes a linearly weighted dependency between the predictors and the response variable, chi-square automatic interaction detection assumes a non-linear and hierarchical dependency. Logistic regression also assumes a monotonic relationship between each predictor variable and the response variable, while chi-square automatic interaction detection does not require such an assumption. In this research, the chi-square automatic interaction detection decision tree model achieved a higher odds ratio and area under the receiver operating characteristic curve in predicting ski injury. Both logistic regression and chi-square automatic interaction detection identified the daily time spent in the ski lift transportation system as the most important feature for ski injury prediction which provides solid evidence that ski injuries are early-failure events. Skiers who are at the highest risk of injury also exhibit higher lift switching behavior while performing faster runs and preferring ski slopes with higher vertical descents. The lowest injury risk is observed for skiers who spend more time in the ski lift transportation system and ski faster than the average population.

Patterns of attendance to health checks in a municipality setting: the Danish 'Check Your Health Preventive Program'.

We aimed to investigate the determinants of attendance to a preventive health check program and to explore the homogeneity of the attenders.4853 eligible persons living in the municipality of Randers, Denmark, from 2012 to 2013, aged 30–49 years, received an invitation to attend the ‘Check Your Health Preventive Program’. Data was obtained from the Danish National Registers. Socio-demographic factors, use of preventive services, morbidity were examined as determinants of attendance by Poisson regression analyses. A chi-squared automatic interaction detection decision tree analysis was used to identify mutually exclusive groups.In total, 55% of the invited population attended (49% men). Attenders were more likely to be: of higher age; immigrants; cohabiting; have: higher socio-economic status; higher use of preventive services and lower morbidity. Decision tree analysis revealed six groups, with the most important variable being income: 1) low income, low education (A = attendance rate: 38%; P = population size: 11%); 2) low income, education higher than 10 years, living alone (A: 41%; P: 5%); 3) low income, education higher than 10 years, cohabiting (A: 56%; P: 16%); 4) middle income (A: 60%; P: 34%); 5) high income, living alone (A: 56%; P: 4%); 6) high income, cohabiting (A: 69%; P: 30%).More than half of a general population voluntarily attended a general health check, despite a resource intensive offer. People with low resources had lower attendance rates. This study adds a detailed description of mutually exclusive groups of attenders, for use in future planning and implementation of preventive actions.

Chi-squared Automatic Interaction Detection Decision Tree Analysis of Risk Factors for Infant Anemia in Beijing, China.

Background:In the past decades, studies on infant anemia have mainly focused on rural areas of China. With the increasing heterogeneity of population in recent years, available information on infant anemia is inconclusive in large cities of China, especially with comparison between native residents and floating population. This population-based cross-sectional study was implemented to determine the anemic status of infants as well as the risk factors in a representative downtown area of Beijing.Methods:As useful methods to build a predictive model, Chi-squared automatic interaction detection (CHAID) decision tree analysis and logistic regression analysis were introduced to explore risk factors of infant anemia. A total of 1091 infants aged 6–12 months together with their parents/caregivers living at Heping Avenue Subdistrict of Beijing were surveyed from January 1, 2013 to December 31, 2014.Results:The prevalence of anemia was 12.60% with a range of 3.47%–40.00% in different subgroup characteristics. The CHAID decision tree model has demonstrated multilevel interaction among risk factors through stepwise pathways to detect anemia. Besides the three predictors identified by logistic regression model including maternal anemia during pregnancy, exclusive breastfeeding in the first 6 months, and floating population, CHAID decision tree analysis also identified the fourth risk factor, the maternal educational level, with higher overall classification accuracy and larger area below the receiver operating characteristic curve.Conclusions:The infant anemic status in metropolis is complex and should be carefully considered by the basic health care practitioners. CHAID decision tree analysis has demonstrated a better performance in hierarchical analysis of population with great heterogeneity. Risk factors identified by this study might be meaningful in the early detection and prompt treatment of infant anemia in large cities.

ALicante sUrgical Community Emergencies New Tool for the enUmeration of Morbidities: a simplified auditing tool for community-acquired gastrointestinal surgical emergencies

BackgroundIn a previous study, we found that Physiological and Operative Severity Score for the enUmeration of Mortality and Morbidity (POSSUM) overpredicts morbidity risk in emergency gastrointestinal surgery. Our aim was to find a POSSUM equation adjustment. MethodsA prospective observational study was performed on 2,361 patients presenting with a community-acquired gastrointestinal surgical emergency. The first 1,000 surgeries constituted the development cohort, the second 1,000 events were the first validation intramural cohort, and the remaining 361 cases belonged to a second validation extramural cohort. Results(1) A modified POSSUM equation was obtained. (2) Logistic regression was used to yield a statistically significant equation that included age, hemoglobin, white cell count, sodium and operative severity. (3) A chi-square automatic interaction detector decision tree analysis yielded a statistically significant equation with 4 variables, namely cardiac failure, sodium, operative severity, and peritoneal soiling. ConclusionsA modified POSSUM equation and a simplified scoring system (aLicante sUrgical Community Emergencies New Tool for the enUmeration of Morbidities [LUCENTUM]) are described. Both tools significantly improve prediction of surgical morbidity in community-acquired gastrointestinal surgical emergencies.

Correlation of Driver Gender with Injury Severity in Large Truck Crashes in Missouri

Large trucks are vital for shipping freight in the United States but are also a dangerous mode of transportation. From 2002 to 2012, 91,145 crashes involving large trucks were recorded in Missouri, resulting in 1,156 fatalities and 18,457 injuries. Many factors contribute to the severity of a large truck crash; these factors and their effects on severity may vary as a function of gender. Missouri crash data from 2002 to 2012 were analyzed to identify circumstances that increased the probability of injury severity outcomes in the event of a large truck crash. Chi-square automatic interaction detector decision tree models were developed for each gender to understand predictor importance better and to uncover interactions between contributing factors. Results suggest that the major correlations with levels of injury severity are following too closely, physical impairment, and improper passing for female Missouri drivers with a commercial driver’s license (CDL) and driving too fast for conditions, failing to yield, and physical impairment for male Missouri CDL drivers. Results suggest that environmental factors are not significant predictors of severity outcome for female CDL drivers; however, road and lighting conditions play a tertiary role for male CDL drivers. Recommendations based on study results include encouraging truck driver training programs to focus educational efforts on gender-specific behaviors that affect crash injury severity to enhance road safety.

The Validity of the 16-Item Version of the Prodromal Questionnaire (PQ-16) to Screen for Ultra High Risk of Developing Psychosis in the General Help-Seeking Population

In order to bring about implementation of routine screening for psychosis risk, a brief version of the Prodromal Questionnaire (PQ; Loewy et al., 2005) was developed and tested in a general help-seeking population. We assessed a consecutive patient sample of 3533 young adults who were help-seeking for nonpsychotic disorders at the secondary mental health services in The Hague with the PQ. We performed logistic regression analyses and CHi-squared Automatic Interaction Detector decision tree analysis to shorten the original 92 items. Receiver operating characteristic curves were used to examine the psychometric properties of the PQ-16. In the general help-seeking population, a cutoff score of 6 or more positively answered items on the 16-item version of the PQ produced correct classification of Comprehensive Assessment of At-Risk Mental State (Yung et al., 2005) psychosis risk/clinical psychosis in 44% of the cases, distinguishing Comprehensive Assessment of At-Risk Mental States (CAARMS) diagnosis from no CAARMS diagnosis with high sensitivity (87%) and specificity (87%). These results were comparable to the PQ-92. The PQ-16 is a good self-report screen for use in secondary mental health care services to select subjects for interviewing for psychosis risk. The low number of items makes it quite appropriate for screening large help-seeking populations, thus enhancing the feasibility of detection and treatment of ultra high-risk patients in routine mental health services.

Using decision trees to identify tourism stakeholders

PurposeThe purpose of this paper is to investigate the use of decision tree analysis in the identification of stakeholders based on their participation in tourism and political activities in a community.Design/methodology/approachA survey was sent to tourism stakeholders in two rural counties. Responses were collected and analyzed using the exhaustive chi‐square automatic interaction detection decision tree analysis.FindingsBased on the results of the decision tree analysis four tourism stakeholder groups were identified based on their participation in tourism and political activities in a community: high participants, high‐moderate participants, low‐moderate participants, and low participants.Research limitations/implicationsOwing to a low response rate, an issue of non‐response bias could exist, but the information from the respondents can give insight on stakeholders in these communities. Also, the specific results of this study can only be applied to eastern North Carolina and are not generalizable to other areas.Practical implicationsResults from this study demonstrate the use of decision tree analysis in identifying community stakeholders. Using decision tree analysis tourism planners can identify stakeholder groups that will participate in tourism and political activities. With this knowledge, tourism planners can identify which stakeholder groups will be the most influential and vocal in a community with regard to tourism development.Originality/valueDecision tree analysis is a tool for partitioning a data set based on the relationships between a set of independent variables and a dependent variable. The research reported here tests the application of decision tree analysis, an analytical technique that is not traditionally used to segment stakeholders in tourism.

Using Decision Trees to Identify Tourism Stakeholders: The Case of Two Eastern North Carolina Counties

This paper explores stakeholder involvement in tourism planning, development, and management. For tourism planners to include stakeholders in the tourism planning process those stakeholders and their interests need to be identified. The research reported in this paper describes and applies an analytical technique that is not traditionally used to identify stakeholders. A questionnaire was developed and mailed to stakeholders in two rural communities in North Carolina. The data were analysed with an Exhaustive Chi-square Automatic Interaction Detection decision tree. From the results of the decision tree, stakeholder groups were identified in relation to their support for sustainable tourism development in their community.