Lung Pattern Classification for Interstitial Lung Diseases Using a Deep Convolutional Neural Network

Automated tissue characterization is one of the most crucial components of a computer aided diagnosis (CAD) system for interstitial lung diseases (ILDs). Although much research has been conducted in this field, the problem remains challenging. Deep learning techniques have recently achieved impressive results in a variety of computer vision problems, raising expectations that they might be applied in other domains, such as medical image analysis. In this paper, we propose and evaluate a convolutional neural network (CNN), designed for the classification of ILD patterns. The proposed network consists of 5 convolutional layers with 2×2 kernels and LeakyReLU activations, followed by average pooling with size equal to the size of the final feature maps and three dense layers. The last dense layer has 7 outputs, equivalent to the classes considered: healthy, ground glass opacity (GGO), micronodules, consolidation, reticulation, honeycombing and a combination of GGO/reticulation. To train and evaluate the CNN, we used a dataset of 14696 image patches, derived by 120 CT scans from different scanners and hospitals. To the best of our knowledge, this is the first deep CNN designed for the specific problem. A comparative analysis proved the effectiveness of the proposed CNN against previous methods in a challenging dataset. The classification performance (~85.5%) demonstrated the potential of CNNs in analyzing lung patterns. Future work includes, extending the CNN to three-dimensional data provided by CT volume scans and integrating the proposed method into a CAD system that aims to provide differential diagnosis for ILDs as a supportive tool for radiologists.

The independent role of prenatal and postnatal exposure to active and passive smoking on the development of early wheeze in children.

Maternal smoking during pregnancy increases childhood asthma risk, but health effects in children of nonsmoking mothers passively exposed to tobacco smoke during pregnancy are unclear. We examined the association of maternal passive smoking during pregnancy and wheeze in children aged ≤2 years.Individual data of 27 993 mother-child pairs from 15 European birth cohorts were combined in pooled analyses taking into consideration potential confounders.Children with maternal exposure to passive smoking during pregnancy and no other smoking exposure were more likely to develop wheeze up to the age of 2 years (OR 1.11, 95% CI 1.03-1.20) compared with unexposed children. Risk of wheeze was further increased by children's postnatal passive smoke exposure in addition to their mothers' passive exposure during pregnancy (OR 1.29, 95% CI 1.19-1.40) and highest in children with both sources of passive exposure and mothers who smoked actively during pregnancy (OR 1.73, 95% CI 1.59-1.88). Risk of wheeze associated with tobacco smoke exposure was higher in children with an allergic versus nonallergic family history.Maternal passive smoking exposure during pregnancy is an independent risk factor for wheeze in children up to the age of 2 years. Pregnant females should avoid active and passive exposure to tobacco smoke for the benefit of their children's health.

The winner takes it all? Characteristics of adolescents at-risk/problem gamblers in Switzerland

Background: Gambling has and still entertains people in almost all societies throughout the world. Western societies have faced considerable changes in the amount and accessibility of gambling possibilities during the last decades, and the rates of both adolescent gamblers and problem gamblers have increased significantly. Objective: To determine the characteristics of at-risk and problem adolescent gamblers in Switzerland. Subjects: The study population consisted of 3134 students (1669 females) attending post-mandatory education in 15 randomly chosen centres. Participants were divided into non- (n=2207), non-problematic (n=754) and at-risk/ problematic gamblers (n=176). Methods: Both gambling groups were compared to non-gamblers on socio-demographic measures, substance use and Internet use. Overall, 29.6% had gambled during the past year and 5.6% had gambled in a risky or problematic way. Compared to non-gamblers and after controlling for potential confounders, non-problem gamblers were significantly more likely to be male, apprentices and to misuse alcohol. At-risk/problematic gamblers were additionally more likely to smoke cannabis, to be problematic Internet users and to be non-Swiss than non-gamblers. Conclusion: At-risk and problematic adolescent gambling is associated with other health risk behaviours. Health practitioners should include gambling in the psycho-social screening and preventive counselling of adolescents.

Transmission of Mycobacterium chimaera from Heater-Cooler Units during Cardiac Surgery despite an Ultraclean Air Ventilation System.

Heater-cooler units (HCUs) were recently identified as a source of Mycobacterium chimaera causing surgical site infections. We investigated transmission of this bacterium from HCUs to the surgical field by using a thermic anemometer and particle counter, videotape of an operating room equipped with an ultraclean laminar airflow ventilation system, and bacterial culture sedimentation plates in a nonventilated room. Smoke from the HCU reached the surgical field in 23 s by merging with ultraclean air. The HCU produced on average 5.2, 139, and 14.8 particles/min in the surgical field at positions Off, On/oriented toward, and On/oriented away, respectively. Culture plates were positive for M. chimaera <5 m from the HCU in the test room. These experiments confirm airborne transmission of M. chimaera aerosols from a contaminated HCU to an open surgical field despite ultraclean air ventilation. Efforts to mitigate infectious risks during surgery should consider contamination from water sources and airflow-generating devices.

Primary lung cancer after breast cancer: The role of radiation therapy and cigarette smoking

The magnitude of the interaction between cigarette smoking, radiation therapy, and primary lung cancer after breast cancer remains unresolved. This case control study further examines the main and joint effects of cigarette smoking and radiation therapy (XRT) among breast cancer patients who subsequently developed primary lung cancer, at The University of Texas M. D. Anderson Cancer Center (MDACC) in Houston, Texas. Cases (n = 280) were women diagnosed with primary lung cancer between 1955 and 1970, between 30–89 years of age, who had a prior history of breast cancer, and were U.S. residents. Controls (n = 300) were randomly selected from 37,000 breast cancer patients at MDACC and frequency matched to cases on age at diagnosis (in 5-year strata), ethnicity, year of breast cancer diagnosis (in 5-year strata), and had survived at least as long as the time interval for lung cancer diagnosis in the cases. Stratified analysis and unconditional logistic regression modeling were used to calculate the main and joint effects of cigarette smoking and radiation treatment on lung cancer risk. Medical record review yielded smoking information on 93% of cases and 84% of controls, and among cases 45% received XRT versus 44% of controls. Smoking increased the odds of lung cancer in women who did not receive XRT (OR = 6.0, 95%CI, 3.5–10.1) whereas XRT was not associated with increased odds (OR = 0.5, 95%CI, 0.2–1.1) in women who did not smoke. Overall the odds ratio for both XRT and smoking together compared with neither exposure was 9.00 (9 5% CI, 5.1–15.9). Similarly, when stratifying on laterality of the lung cancer in relation to the breast cancer, and when the time interval between breast and lung cancers was >10 years, there was an increased odds for both smoking and XRT together for lung cancers on the same side as the breast cancer (ipsilateral) (OR = 11.5, 95% CI, 4.9–27.8) and lung cancers on the opposite side of the breast cancer (contralateral) (OR= 9.6, 95% CI, 2.9–0.9). After 20 years the odds for the ipsilateral lung were even more pronounced (OR = 19.2, 95% CI, 4.2–88.4) compared to the contralateral lung (OR = 2.6, 95% CI, 0.2–2.1). In conclusion, smoking was a significant independent risk factor for lung cancer after breast cancer. Moreover, a greater than multiplicative effect was observed with smoking and XRT combined being especially evident after 10 years for both the ipsilateral and contralateral lung and after 20 years for the ipsilateral lung. ^

The role of smoking status on postoperative complications and survival in lung cancer patients presenting to M.D. Anderson Cancer Center

It is estimated that 50% of all lung cancer patients continue to smoke after diagnosis. Many of these lung cancer patients who are current smokers often experience tremendous guilt and responsibility for their disease, and feel it might be too late for them to quit smoking. In addition, many oncologists may be heard to say that it is 'too late', 'it doesn't matter', 'it is too difficult', 'it is too stressful' for their patients to stop smoking, or they never identify the smoking status of the patient. Many oncologists feel unprepared to address smoking cessation as part of their clinical practice. In reality, physicians can have tremendous effects on motivating patients, particularly when patients are initially being diagnosed with cancer. More information is needed to convince patients to quit smoking and to encourage clinicians to assist patients with their smoking cessation. ^ In this current study, smoking status at time of lung cancer diagnosis was assessed to examine its impact on complications and survival, after exploring the reliability of smoking data that is self-reported. Logistic Regression was used to determine the risks of smoking prior to lung resection. In addition, survival analysis was performed to examine the impact of smoking on survival. ^ The reliability of how patients report their smoking status was high, but there was some discordance between current smokers and recent quitters. In addition, we found that cigarette pack-year history and duration of smoking cessation were directly related to the rate of a pulmonary complication. In regards to survival, we found that current smoking at time of lung cancer diagnosis was an independent predictor of early stage lung cancer. This evidence supports the idea that it is "never too late" for patients to quit smoking and health care providers should incorporate smoking status regularly into their clinical practice.^

Tobacco use and related psychosocial risk factors among youth in urban India: Assessment of Project MYTRI follow-up surveys

Project MYTRI (Mobilizing Youth for Tobacco-Related Initiatives in India) was a large 2-year randomized school-based trial with a goal to reduce and prevent tobacco use among students in 6th and 8th grades in Delhi and Chennai in India (n=32 schools). Baseline analyses in 2004 showed that 6th grade students reported more tobacco use than 8 th grade students, opposite of what is typically observed in developed countries like the US. The present study aims to study differences in tobacco use and psychosocial risk factors between the 6th grade cohort and 8th grade cohort, in a compliant sub-sample of control students that were present at all 3 surveys from 2004-06. Both in 2004 and 2005, 6th grade cohort reported significantly greater prevalence of ever use of all tobacco products (cigarettes, bidis, chewing tobacco, any tobacco). These significant differences in ever use of any tobacco between cohorts were maintained by gender, city and socioeconomic status. The 6th grade cohort also reported significantly greater prevalence of current use of tobacco products (cigarettes, chewing tobacco, any tobacco) in 2004. Similar findings were observed for psychosocial risk factors for tobacco use, where the 6th grade cohort scored higher risk than 8th grade cohort on scales for intentions to smoke or chew tobacco and susceptibility to smoke or chew tobacco in 2004 and 2005, and for knowledge of health effects of tobacco in all three years.^ The evidence of early initiation of tobacco use in our 6th grade cohort in India indicates the need to target prevention programs and other tobacco control measures from a younger age in this setting. With increasing proportions of total deaths and lost DALYs in India being attributable to chronic diseases, addressing tobacco use among younger cohorts is even more critical. Increase in tobacco use among youth is a cause for concern with respect to future burden of chronic disease and tobacco-related mortality in many developing countries. Similarly, epidemiological studies that aim to predict future death and disease burden due to tobacco should address the early age at initiation and increasing prevalence rates among younger populations. ^

Smokeless tobacco use in 2005 US Military population

U.S. Military personnel are more likely to use smokeless tobacco than civilians. The purpose of this study was to describe the relationship between smokeless tobacco use and sociodemographic, behavioral, and occupational variables, using data from the 2005 Department of Defense Survey of Health Related Behaviors among Active Duty Military Personnel. The DoD survey was comprised of representative active duty U.S. military members (N=16,146). In adjusted multivariate logistic regression models, this study found smokeless tobacco use to be more prevalent in younger age, males, whites, and enlisted-rank members. By service, higher rates were reported among members of the Army and Marine Corps than among the Air Force and Navy members. Smokeless tobacco use among those who also smoke or drink heavily was also much higher than among those who did not report smoking or heavy alcohol use. Results also showed increased prevalence of smokeless tobacco use among those who reported moderate or high impulsive behavior and among those who recently deployed. These findings contribute to improving the understanding of factors related to smokeless tobacco use in the military and may help design strategies to reduce the use of this potentially toxic substance and improve health for military members.^

Transcriptional regulation of lung tumor suppressor GPRC5A in malignant and normal cells

Chronic exposure of the airways to cigarette smoke induces inflammatory response and genomic instability that play important roles in lung cancer development. Nuclear factor kappa B (NF-κB), the major intracellular mediator of inflammatory signals, is frequently activated in preneoplastic and malignant lung lesions. ^ Previously, we had shown that a lung tumor suppressor GPRC5A is frequently repressed in human non-small cell lung cancers (NSCLC) cells and lung tumor specimens. Recently, other groups have shown that human GPRC5A transcript levels are higher in bronchial samples of former than of current smokers. These results suggested that smoking represses GPRC5A expression and thus promotes the occurrence of lung cancer. We hypothesized that cigarette smoking or associated inflammatory response repressed GPRC5A expression through NF-κB signaling. ^ To determine the effect of inflammation, we examined GPRC5A protein expression in several lung cell lines following by TNF-α treatment. TNF-α significantly suppressed GPRC5A expression in normal small airway epithelial cells (SAEC) as well as in Calu-1 cells. Real-time PCR analysis indicated that TNF-α inhibits GPRC5A expression at the transcriptional level. NF-κB, the major downstream effectors of TNF-α signaling, mediates TNF-α-induced repression of GPRC5A because over-expression of NF-κB suppressed GPRC5A. To determine the region in the GPRC5A promoter through which NF-κB acts, we examined the ability of TNF-α to inhibit a series of reporter constructs with different deletions of GPRC5A promoter. The luciferase assay showed that the potential NF-κB binding sites containing region are irresponsible for TNF-α-induced suppression. Further analysis using constructs with different deletions in p65 revealed that NF-κB-mediated repression of GPRC5A is transcription-independent. Co-immunoprecipitation assays revealed that NF-κB could form a complex with RAR/RXR heterodimer. Moreover, the inhibitory effect of NF-κB has been found to be proportional to NF-κB/RAR ratio in luciferase assay. Finally, Chromatin IP demonstrated that NF-κB/p65 bound to GPRC5A promoter as well as RAR/RXR and suppressed transcription. Taken together, we propose that inflammation-induced NF-κB activation disrupts the RA signaling and suppresses GPRC5A expression and thus contributes to the oncogenesis of lung cancer. Our studies shed new light on the pathogenesis of lung cancer and potentially provide novel interventions for preventing and treating this disease. ^

Why is asthma mortality higher in Puerto Ricans?

Asthma is a significant public health issue and the most common chronic disease in children. The disease burden of asthma is rising around the world and especially in certain populations. In the United States Puerto Rican Americans have the highest rates of mortality due to asthma, while Mexico Americans have the lowest asthma mortality in the U.S. The reasons for this have been the cause of much speculation in the past; however, no clear cause for these differences has been recognized. The present work reviews the literature bearing on this question to show that there are good reasons to believe that individuals with unusually responsive innate immune responses may be predisposed to the development of asthma. Also reviewed is the molecular basis for this connection. The evidence shows that the history and anthropology of the Puerto Rican people is quite different from that of any other surviving North American or Caribbean population, as it was a relatively isolated island population for 400 years with an environment that tended to eliminate individuals with weak innate immune systems. The Puerto Rican population successfully survived the Columbian exchange of microbes but may be poorly adapted to the modern pro-inflammatory diet coupled with exposure to cigarette smoke as well as cockroach and house dust mite feces.^

Hormone therapy and physical activity as predictors of lung cancer risk and survival: The California Teachers Study

Cigarette smoking is responsible for the majority of lung cancer cases worldwide; however, a proportion of never smokers still develop lung cancer over their lifetime, prompting investigation into additional factors that may modify lung cancer incidence, as well as mortality. Although hormone therapy (HT), physical activity (PA), and lung cancer have been previously examined, the associations remain unclear. This study investigated exposure to HT and PA that may modulate underlying mechanisms of lung cancer etiology and progression among women by using existing, de-identified data from the California Teachers Study (CTS).^ The CTS cohort, established in 1995–1996, has 133,479 active and retired female teachers and administrators, recruited through the California State Teachers Retirement System, and followed annually for cancer diagnosis, death, and change of address. Each woman enrolled in the CTS returned a questionnaire covering a wide variety of issues related to cancer risk and women's health, including recent and past HT use and physical activity, as well as active and environmental cigarette smoke exposure. Complete data to assess the associations between HT and lung cancer risk and survival were available for 60,592 postmenopausal women. Between 1995 and 2007, 727 of these women were diagnosed with invasive lung cancer; 441 of these died. Complete data to assess the associations between PA and lung cancer risk and survival were available for 118,513 women. Between 1995 and 2007, 853 of these women were diagnosed with invasive lung cancer; 516 of these died.^ After careful adjustment for smoking habits and other potential confounders, no measure of HT use was associated with lung cancer risk; however, any HT use (vs. no use) was associated with a decrease in lung-cancer-specific mortality. Specifically, among women who only used estrogen (E-only), decreases in lung cancer mortality were seen for recent use, but not for former use; no association was observed for estrogen plus progestin (E+P). Furthermore, among former users of HT, a statistically significant decrease in lung cancer mortality was observed for E-only use within 5 years prior to baseline, but not for E-only use >5 years prior to baseline. Neither long-term recreational PA nor recent recreational PA alone were associated with lung cancer risk; however, among women with a BMI<25 and ever smokers, high long-term moderate+strenuous PA was associated with a decrease in lung cancer risk. Women with non-local disease showed a decrease in lung cancer mortality associated with increasing duration of strenuous long-term activity, and 1.50-3.00 h/wk/y of recent moderate or recent strenuous PA. Long-term moderate PA was associated with decreased lung cancer mortality in never smokers, whereas recent moderate PA was associated with increased lung cancer mortality in current smokers. ^ Placing our findings in the context of the current literature, HT does not appear to be associated with lung cancer risk and previous studies reporting a protective effect of HT use on lung cancer risk may be subject to residual confounding by smoking. Looking at our findings regarding PA overall, the evidence still remains inconclusive regarding whether or not physical activity influence lung cancer risk or mortality. Our results suggest that recreational PA may associated with decreased lung cancer risk among women with BMI<25 and ever smoking-women; however, residual confounding by smoking should be strongly considered. To our knowledge, this is the first study to investigate lifetime recreational PA and lung cancer mortality among women. Our results contribute to the growing body of knowledge regarding non-smoking-related risk factors for lung cancer incidence and mortality among women. Given the potential clinical and interventional significance, further study and validation of these findings is warranted.^

ETIOLOGIC AND CLINICAL FACTORS WHICH DIFFERENTIATE PATIENTS WITH CHROMOSOMAL ABNORMALITIES FROM DIPLOID PATIENTS IN ACUTE NONLYMPHOCYTIC LEUKEMIA

Tumor-specific chromosomal abnormalities have been demonstrated in bone marrow of approximately 50% of newly diagnosed acute nonlymphocytic (ANLL) patients. This study examined two hypotheses: (1) Aneuploid (AA) patients are diagnosed later in the course of their disease than diploid (NN) patients; and (2) AA patients are more likely to have been exposed to environmental agents. Of 324 patients eligible for study, environmental exposure data were obtained for 236 (73%) of them. No evidence was found to suggest that AA patients had more advanced disease than NN patients. Aneuploid patients were more likely than NN patients to: (a) report treatment with cytotoxic drugs for a prior medical condition (odds ratio, adjusted for age, sex and other exposures (OR) = 4.25, 95% confidence intervals, 1.38 to 13.17); (b) smoke cigarettes, OR = 1.82 (1.02, 3.26) and (c) drink alcoholic beverages, OR = 1.91 (1.05, 3.48). No statistically significant associations between aneuploidy and occupational exposures were present, OR = 3.59 (0.76, 17.13). Problems in interpreting these ORs are discussed. ^

MARITAL STATUS AND CARDIOVASCULAR RISK (SERUM, CHOLESTEROL, CIGARETTE, SMOKING, BLOOD, PRESSURE, SYSTOLIC)

There were three purposes of this study. The first was to describe the association between stable marital status and serum cholesterol, systolic blood pressure and cigarette smoking. The second purpose was to determine whether individuals who were married at one point and became widowed or divorced/separated had higher serum cholesterol, higher systolic blood pressure or were more likely to smoke prior to the change in marital status compared with individuals who did not change marital status. The third purpose was to determine whether the changes in marital status described above were related to increases in serum cholesterol or in cigarette smoking behavior. The rationale for the study was to determine whether previously reported associations between marital status categories and cardiovascular mortality may be mediated through higher values of risk correlates for cardiovascular disease among unmarried individuals.^ The study group selected for this dissertation was a sample from the Hypertension Detection and Follow-up Program (HDFP) population. The HDFP population was aged 30-69 years at the initial visit and included blacks and whites, males and females. The population was followed five years after the initial visit and periodic measurements of serum cholesterol, blood pressure and cigarette smoking behavior were obtained.^ Serum cholesterol was not associated with stable marital status category or with marital status prior to change. Changes in serum cholesterol were associated with marital status categories after change but the serum cholesterol values deceased rather than increased. Married individuals were shown to have higher serum cholesterol values compared with unmarried. Selection of the HDFP population may have influenced an ability to detect a significant association between marital status and serum cholesterol but it is doubtful that use of a general population would alter the direction of the association.^ Systolic blood pressure was significantly higher at the initial visit among unmarried white males and females compared with their married counterparts. No association between systolic blood pressure was found among black males or females. Those individuals who were married at the initial visit who experienced a change in marital status were found to have higher systolic blood pressure prior to the change in marital status. . . . (Author's abstract exceeds stipulated maximum length. Discontinued here with permission of author.) UMI ^

THE OCCURRENCE OF GROUP G STREPTOCOCCUS IN THE DOMESTIC CAT: COMPARISON OF THE PHYSIOLOGICAL PROPERTIES OF GROUP G STREPTOCOCCUS ISOLATED FROM HUMANS AND CATS

The occurrence of group G streptococci in cats and evaluation of the recovered organisms as potential human pathogens was investigated. Throat swabs were obtained from 89 cats (47 males and 42 females) and vaginal swabs from 39 female cats. Eighty-three of the examined cats were housed in individual cages at a University Animal Care Facility. Six cats, 2 mature males, 2 mature females and 2 young females were family pets in a rural area. Beta-hemolytic streptococci were recovered from 33 (37%) of the 89 cat throats cultured, and 27 (30.3%) were identified as group G. More males (34%) than females (24%) had throat cultures positive for group G. From the 39 vaginal cultures examined, 24 (61.5%) contained beta-hemolytic streptococci and 23 (58.9%) were identified as group G streptococci. Streptococci were not recovered from the vaginal cultures of the 5 females under 6 months of age.^ Thirty one group G streptococci isolated from cats were compared with 37 isolates of group G obtained from humans (health status or site of origin unknown). More group G cat isolates (81%) produced deoxyribonuclease (DNase) than did the human isolates (36%). The proportion of cat throat and vaginal isolates producing DNase was the same. Production of nicotinamide adenine dinucleotide glycohydrolase (NADase) by group G isolates of human origin was 70%, cat throat isolates 53% and cat vaginal isolates 37%. The Serum Opacity Factor was present in 73% of the cat throat isolates of group G, 43.7% of the cat vaginal isolates and 58.6% of the human isolates. Possession of an anti-phagocytic factor (M protein like substance) demonstrated by the ability to multiply in fresh human blood was greater in the group G from cat throats (46.7%) than from cat vagina (37.5%) or from the human isolates (13.5%). Many of the biochemical characteristics of the group G streptococci of cat origin were more similar to the biochemical characteristics of group A streptococci, than to the characteristics of group G of human origin. The group G streptococci, found in a large number of cats, could be potential human pathogens, as their physiological and biological characteristics are very similar to those of group A, a known human pathogen. ^

Motivational interviewing and feedback for college drinkers: Handling missing values with complete case analysis and multiple imputation

Objective: In this secondary data analysis, three statistical methodologies were implemented to handle cases with missing data in a motivational interviewing and feedback study. The aim was to evaluate the impact that these methodologies have on the data analysis. ^ Methods: We first evaluated whether the assumption of missing completely at random held for this study. We then proceeded to conduct a secondary data analysis using a mixed linear model to handle missing data with three methodologies (a) complete case analysis, (b) multiple imputation with explicit model containing outcome variables, time, and the interaction of time and treatment, and (c) multiple imputation with explicit model containing outcome variables, time, the interaction of time and treatment, and additional covariates (e.g., age, gender, smoke, years in school, marital status, housing, race/ethnicity, and if participants play on athletic team). Several comparisons were conducted including the following ones: 1) the motivation interviewing with feedback group (MIF) vs. the assessment only group (AO), the motivation interviewing group (MIO) vs. AO, and the intervention of the feedback only group (FBO) vs. AO, 2) MIF vs. FBO, and 3) MIF vs. MIO.^ Results: We first evaluated the patterns of missingness in this study, which indicated that about 13% of participants showed monotone missing patterns, and about 3.5% showed non-monotone missing patterns. Then we evaluated the assumption of missing completely at random by Little's missing completely at random (MCAR) test, in which the Chi-Square test statistic was 167.8 with 125 degrees of freedom, and its associated p-value was p=0.006, which indicated that the data could not be assumed to be missing completely at random. After that, we compared if the three different strategies reached the same results. For the comparison between MIF and AO as well as the comparison between MIF and FBO, only the multiple imputation with additional covariates by uncongenial and congenial models reached different results. For the comparison between MIF and MIO, all the methodologies for handling missing values obtained different results. ^ Discussions: The study indicated that, first, missingness was crucial in this study. Second, to understand the assumptions of the model was important since we could not identify if the data were missing at random or missing not at random. Therefore, future researches should focus on exploring more sensitivity analyses under missing not at random assumption.^