Multiple Poisson Regression Models Research Articles

Periodontitis and cancer mortality: Register-based cohort study of 68,273 adults in 10-year follow-up.

Periodontitis, a multifactorial infection-induced low-grade chronic inflammation, can influence the process of carcinogenesis. We studied with 10 years follow-up of 68,273 adults-based cohort the involvement of periodontitis as a risk factor for cancer mortality. Periodontal status was defined based on procedure codes of periodontal treatment. Rate ratios and absolute differences of overall and cancer mortality rates were assessed with respect to periodontal status using multiplicative and additive Poisson regression models, respectively. We adjusted for effect of age, sex, calendar time, socio-economic status, oral health, dental treatments and diabetes. Data about smoking or alcohol consumption were not available. Altogether 797 cancer deaths occurred during 664,020 person-years accumulated over a mean 10.1-year follow-up. Crude cancer mortality rate per 10,000 person-years for participants without and with periodontitis was 11.36 (95% CI 10.47-12.31) and 14.45 (95% CI 12.51-16.61), respectively. Crude rate ratios for periodontitis indicated an increased risk of overall (RR 1.27, 95% CI 1.08-1.39) and pancreatic cancer (RR 1.69, 95% CI 1.04-2.76) mortality. After adjustment, the results showed even stronger associations of periodontitis with increased overall (RR 1.33, 95% CI 1.10-1.58) and pancreatic cancer (RR 2.32, 95% CI 1.31-3.98) mortality. A higher pancreatic cancer mortality among individuals with periodontitis contributed considerably to the difference in overall cancer mortality, but this difference was not due to pancreatic cancer deaths alone.

Child abuse and work stress in adulthood: Evidence from a population-based study

The objective of this study was to examine the association between child abuse and work stress in adulthood. We used data from the 2012 Canadian Community Health Survey (CCHS) Mental Health, a nationally representative cross-sectional survey of Canadians. This study included all participants aged 20years or older who reported being employed the past 12months (N=14,581). Child physical abuse, sexual abuse, and exposure to intimate partner violence were assessed in relation to several work stress-related indicators. Multiple linear and Poisson regression models adjusted for age, sex, education, household income, marital status, occupation group, and any lifetime mental disorder. Child abuse was significantly associated with greater odds of high work stress (IRR: 1.29; 95% CI: 1.16–1.43) in adulthood. More specifically, child abuse was associated with greater odds of job dissatisfaction (IRR: 1.69; 95% CI: 1.31–2.18), job insecurity (IRR: 1.56; 95% CI: 1.27–1.91), and self-perceived low support (IRR: 1.33; 95% CI: 1.22–1.46). It was also associated with high levels of psychological demand (b=0.348; 95% CI: 0.229–0.467) and job strain (b=0.031; 95% CI: 0.019–0.043). Examination of the Karasek's Demand-Control Model using multinomial logistic regression analyses indicated that child abuse was significantly associated with high strain (RRR:1.39; 95% CI: 1.14–1.72) and active (RRR: 1.56; 95% CI: 1.28–1.90) jobs. These findings suggest the negative influence of child abuse on work experience. Success in preventing child abuse may help reduce work-related stress in adulthood.

Maternal socioeconomic factors and adverse perinatal outcomes in two birth cohorts, 1997/98 and 2010, in São Luís, Brazil.

Estimate the LBW, PTB, and IUGR rates in São Luís, Maranhão, Brazil, in 2010, and check for associations between socioeconomic factors and these indicators. This study is based on a birth cohort performed in São Luís. It included 5,051 singleton hospital births in 2010. The chi-square test was used for proportion comparisons, while simple and multiple Poisson regression models with robust error variance were used to estimate relative risks. LBW, PTB and IUGR rates were 7.5, 12.2, and 10.3% respectively. LBW was higher in low-income families, while PTB and IUGR were not associated with socioeconomic factors. The absence or weak association of these indicators with social inequality point to improvements in health care and/or in social conditions in São Luís.

Effect of Alcohol Consumption on Clinical Attachment Loss Progression in an Urban Population From South Brazil: A 5-Year Longitudinal Study.

The aim of this study is to investigate the impact of alcohol consumption on clinical attachment loss (AL) progression over a period of 5 years. A multistage probability sampling strategy was used to draw a representative sample of the metropolitan area of Porto Alegre, Brazil. Five hundred thirty-two individuals (209 males and 293 females) aged 18 to 65 years at baseline with no medical history of diabetes and at least six teeth were included in this analysis. Full-mouth periodontal examinations with six sites per tooth were conducted at baseline and after 5 years. Alcohol consumption was assessed at baseline by asking participants about the usual number of drinks consumed in a week. Four categories of alcohol consumption were defined: 1) non-drinker; 2) ≤1 glass/week; 3) >1 glass/week and ≤1 glass/day; and 4) >1 glass/day. Individuals showing at least two teeth with proximal (clinical AL) progression ≥3 mm over 5 years were classified as having disease progression. Multiple Poisson regression models adjusted for age, sex, smoking, socioeconomic status, and body mass index were used to estimate relative risks (RRs) and 95% confidence intervals (CIs). Overall, individuals who consumed >1 glass/day had 30% higher risk for clinical AL progression (RR = 1.30; 95% CI: 1.07 to 1.58) than non-drinkers. Among males, risk of clinical AL progression for individuals drinking >1 glass/day was 34% higher than non-drinkers (RR = 1.34; 95% CI: 1.09 to 1.64). Never-smoker males drinking ≤1 glass/week had significantly lower risk for clinical AL progression than non-drinkers (RR = 0.52; 95% CI: 0.30 to 0.89), whereas those drinking >1 glass/day had significantly higher risk (RR = 1.50; 95% CI: 1.08 to 1.99). Among females, no association between alcohol consumption and clinical AL progression was observed. Alcohol consumption increased the risk of clinical AL progression, and this effect was more pronounced in males. Low dosages (≤1.37 g of alcohol/day) of alcohol consumption may be beneficial to prevent periodontal disease progression in males. The impact of alcohol cessation initiatives on periodontal health should be evaluated.

Influence of functional dentition on satisfaction with oral health and impacts on daily performance among Brazilian adults: a population-based cross-sectional study

BackgroundDental esthetics, chewing and speech should be preserved in a dentition denominated functional and are closely related to satisfaction with oral health (SOH), impacts caused by oral problems and have a possible association with Oral Health-Related Quality of Life. Thus, the purpose of the present study was to investigate the influence of different concepts of functional dentition (FD) on both SOH and impacts on daily performance (IDP) among Brazilian adults.MethodsA cross-sectional study was conducted with 9564 adults (35–44 years). SOH and IDP were evaluated using the Oral Impacts on Daily Performance (OIDP) questionnaire. FD was considered based on four different definitions: I-classification of the World Health Organization (FDWHO = ≥20 teeth); II-well-distributed teeth (WDT = ≥10 teeth in each arch); III-classified by esthetics and occlusion (FDClass5 = sequential presence of one tooth in each arch, ≥10 teeth in each arch, 12 anterior teeth, ≥three posterior occluding pairs [POPs] of premolars and ≥one POP molar bilaterally); and IV-classified by esthetics, occlusion and periodontal status (FDClass6 = FDClass5 plus all sextants with CPI ≤ 3 and/or CAL ≤ 1). The proportion of adults satisfied with oral health and without overall impact (OIDP = 0) was calculated for each definition of FD. Multiple Poisson regression models were adjusted by demographic-socioeconomic characteristics, self-reported oral problems and the use of dental services for each dependent variable.ResultsWhen FDClass5 and FDClass6 were considered a greater proportion of adults reported being satisfied (52.1 and 53.1%, respectively) and have OIDP = 0 (52.4 and 53.3, respectively). In the multiple models, SOH was associated with FDClass5 (RP = 1.21) and FDClass6 (RP = 1.24) and OIDP = 0 was associated with WDT (RP = 1.14) and FDClass6 (RP = 1.21).ConclusionsThe greater influence of WDT, FDClass5 and FDClass6 on aspects related to quality of life in comparison to FDWHO demonstrates the need for the establishment of a broader definition of FD that encompasses subjective aspects.

Multimorbidity as an important issue among women: results of a gender difference investigation in a large population-based cross-sectional study in West Asia

ObjectivesTo investigate the impact of gender on multimorbidity in northern Iran.DesignA cross-sectional analysis of the Golestan cohort data.SettingGolestan Province, Iran.Study population49 946 residents (age 40–75 years) of Golestan Province, Iran.Main...

Variations over time in the effects of age and sex on hospitalization rates before and after admission to a nursing home: A German cohort study

ObjectivesWe examined hospitalization rates for nursing home residents before and after their entry to the home, stratified by sex and age. Study designA cohort study was conducted using data from a large health insurance fund on 127,227 residents aged 65 years and over newly admitted to a nursing home between January 1, 2010, and December 31, 2014. Main outcome measurementsWe assessed hospitalization rates and proportions being hospitalized in 6-month intervals one year before nursing home placement and up to 5 years thereafter. Multiple Poisson regression models were fitted to calculate relative risks (RR). ResultsMean age was 84.0 years and 74.6% of the cohort were females. Hospitalization rates were 194.4 per 100 person-years (PY) in the 12 months before entry to the nursing home and 120.0 per 100 PY thereafter. Rates were highest immediately before entry in both sexes. The influence of age was most pronounced in the 12–7 months before entry (RR: 2.37 for 65–74 vs. 95+ years) and declined thereafter (1.29–1.38 up to month 24 after entry). In contrast, the influence of sex was greater after entry (RR: 1.13 for males vs. females in the 12–7 months before and 1.23–1.31 up to month 24 after entry). ConclusionsHospitalization rates of nursing home residents are much higher in Germany than in other Western countries. We have provided some insight into the influence of age and sex on hospitalization rates, which varied over the period (time before and after entry to the nursing home) analyzed. We urgently recommend that future studies on the hospitalization of residents stratify their analyses by sex, age and period.

Impact of Malocclusion on Oral Health-Related Quality of Life of 8-12 Years Old Schoolchildren in Southern Brazil.

The aim of this study was to estimate the impact of malocclusion on the oral health-related quality of life (OHRQoL) of schoolchildren aged 8-12 years old in Southern Brazil. A two-stage cluster procedure was used to select 1,199 children in 20 public and private schools in Pelotas/Brazil. Cross-sectional data was collected, consisting of a socioeconomic questionnaire to parents, children's interview and clinical oral examination. The clinical variables were obtained from clinical examination, and the Child Perceptions Questionnaire (CPQ) was assessed during children's interview. To measure malocclusion and orthodontic treatment need the Dental Aesthetic Index was used. For data analysis multiple Poisson regression models estimating the rate ratios (RR) and their respective confidence intervals (95%CI) were used. Among 1,206 participants, 789 were aged 8-10 years and 417 between 11-12 years. The orthodontic treatment need was higher among the younger children (44.6%) than in the older ones (35.0%) (p value ≤0.05). There was a significant association in the CPQ social and emotional domains with malocclusion in the older schoolchildren. In the adjusted analysis (for socioeconomics and clinical variables) the effect of very severe malocclusion on OHRQoL was confirmed in both 8-10 and 11-12 age groups (RR(95%CI) of 1.24(1.02;1.51) and 1.28(1.01;1.62), respectively). The findings demonstrated that children with very severe malocclusion experienced greater negative impact on OHRQoL compared to those with mild or no malocclusion. The results suggest that malocclusion impacts the quality of life. The higher impact occurs in the social and emotional well-being domains.

How many water molecules are detected in X-ray protein crystal structures?

The positions of several water molecules can be determined in protein crystallography, either buried in internal cavities or at the protein surface. It is important to be able to estimate the expected number of these water molecules to facilitate crystal structure determination. Here, a multiple Poisson regression model implemented on nearly 10 000 protein crystal structures shows that the number of detectable water molecules depends on eight variables: crystallographic resolution, R factor, percentage of solvent in the crystal, average B factor of the protein atoms, percentage of amino acid residues in loops, average solvent-accessible surface area of the amino acid residues, grand average of hydropathy of the protein(s) in the asymmetric unit and normalized number of heteroatoms that are not water molecules. Furthermore, a secondary analysis tested the effect of different software packages. Given the values of these eight variables, it is possible to compute the expected number of water molecules detectable in electron-density maps with reasonable accuracy (as suggested by an external validation of the model).

Relative efficacy of tranexamic acid and preoperative anemia treatment for reducing transfusions in total joint arthroplasty.

This study aimed to evaluate the efficacy of a perioperative blood management (PBM) protocol at a large, tertiary hospital at reducing blood transfusions after total hip or knee arthroplasty (THA or TKA). A retrospective review of the PBM for patients undergoing THA or TKA was performed. Adjusted multiple logistic and Poisson regression models examined the effect of patient characteristics and preoperative, intraoperative, and postoperative factors on the likelihood of transfusion and units transfused. Of 883 study patients, 330 (37.4%) had surgery before PBM protocol implementation and served as the control population while 553 (62.6%) were eligible for the protocol. Having a higher preoperative hemoglobin (Hb) was independently associated with a decreased odds of transfusion (odds ratio [OR], 0.480; p < 0.001). Preoperative treatment for anemia (88 [15.9%] patients) did result in a significant, yet modest, increase in preoperative Hb (11.92 g/dL to 12.35 g/dL; p < 0.001) but treatment was not a significant predictor of transfusion. Receiving intraoperative tranexamic acid (TXA; 204 [36.9%] patients) had the greatest effect in reducing the odds of transfusion (OR, 0.289; p < 0.001) and the number of units transfused (-0.6; p = 0.008). Having a decreased Hb was shown to be an independent risk factor both for requiring a perioperative blood transfusion and for the volume of transfusion. The very modest increase in Hb achieved by the costly and time-consuming preoperative anemia optimization program, however, may not be justified when the use of intraoperative TXA led to drastic reductions in both transfusions and transfusion volumes.

Trends in Mortality After Primary Cytoreductive Surgery for Ovarian Cancer: A Systematic Review and Metaregression of Randomized Clinical Trials and Observational Studies.

Primary cytoreductive surgery (PDS) followed by platinum-based chemotherapy is the cornerstone of treatment and the absence of residual tumor after PDS is universally considered the most important prognostic factor. The aim of the present analysis was to evaluate trend and predictors of 30-day mortality in patients undergoing primary cytoreduction for ovarian cancer. Literature was searched for records reporting 30-day mortality after PDS. All cohorts were rated for quality. Simple and multiple Poisson regression models were used to quantify the association between 30-day mortality and the following: overall or severe complications, proportion of patients with stage IV disease, median age, year of publication, and weighted surgical complexity index. Using the multiple regression model, we calculated the risk of perioperative mortality at different levels for statistically significant covariates of interest. Simple regression identified median age and proportion of patients with stage IV disease as statistically significant predictors of 30-day mortality. When included in the multiple Poisson regression model, both remained statistically significant, with an incidence rate ratio of 1.087 for median ageand 1.017 for stage IV disease. Disease stage was a strong predictor, with the risk estimated to increase from 2.8% (95% confidence interval 2.02-3.66) for stage III to 16.1% (95% confidence interval 6.18-25.93) for stage IV, for a cohort with a median age of 65 years. Metaregression demonstrated that increased age and advanced clinical stage were independently associated with an increased risk of mortality, and the combined effects of both factors greatly increased the risk.

Association between workplace psychosocial factors and mental health in Black, Hispanic, and White women: Cross-sectional findings from the National Health Interview Survey

ABSTRACTResearch evaluating the relation of workplace psychosocial factors to mental health among U.S. women of different racial/ethnic backgrounds is limited. This study investigated the relationship between work-related psychosocial factors and mental health among non-Hispanic Black, Hispanic, and non-Hispanic White women using data from the 2010 National Health Interview Survey. Independent variables of interest included job insecurity, workplace harassment, and work-family conflict (WFC). Multiple Poisson regression models were used to examine the associations between the outcome and independent variables. The prevalence of unfavorable mental health was highest among non-Hispanic Black women (36%) compared to Hispanic (34%) and non-Hispanic White (30%) women. A higher proportion of non-Hispanic Black women reported WFC compared to Hispanics and non-Hispanic Whites (χ2 = 15.50, p < .01), while more Hispanics reported job insecurity (χ2 = 116.81, p < .01). Prevalence of workplace harassment did not differ significantly by race/ethnicity. Odds of unfavorable mental health were significantly higher for women reporting psychosocial work factors. Unexpectedly, a greater association between psychosocial work factors and unfavorable mental health was observed among non-Hispanic White women compared to non-White women; however, caution should be taken in interpreting these cross-sectional results. Future studies should investigate temporal associations and additional psychosocial variables that were not available for use in the current study.

Association of hypertension-related distress with lack of self-care among hypertensive patients

Research on factors associated with poor adherence to self-care focuses primarily on psychiatric emotional disorders such as depression and anxiety, whereas non-psychiatric chronic-disease-related emotional distress has received little attention in hypertensive patients. The objective of this study was to evaluate the association of hypertension-related distress with the lack of self-care including low adherence to pharmacological treatment, lack of regular physical activity, low intake of fruits and vegetables and frequent intake of high-salt foods. A cross-sectional survey was conducted in 2014 in two family medical units affiliated with the Mexican Institute of Social Security. The study included 487 hypertensive patients >19 years of age. The conceptual framework for the study was based on the Health Promotion Model. The analysis included multiple Poisson regression models. We found that 21.1% of participants had hypertension-related distress. Low adherence to pharmacological treatment was identified in 45.8% of patients, whereas 46.8% lacked regular physical activity, 30.8% reported a low consumption of fruits and vegetables, and 54.6% frequently consumed foods high in salt content. Hypertension-related distress was associated with lack of regular physical activity and low intake of fruits and vegetables. These findings highlight the importance of addressing distress in order to improve self-care of hypertensive patients.

Child undernutrition in one of the cities with greater nutritional risk in Brazil: population-based study in the Western Brazilian Amazon.

To estimate the prevalence of child undernutrition and associated factors in a municipality with high nutritional risk in Brazil. This cross-sectional, population-based study was conducted with a sample of 478 children aged under 5 years in the city of Jordão, Acre, Brazil. The following indicators were calculated: weight for age (W/A), height for age (H/A), and weight for height (W/H), using the growth curves of the WHO as reference, which adopts a cutoff of -2 z scores for identification of malnourished children. Adjusted prevalence ratios (PRs) were obtained using multiple Poisson regression models with robust error estimate (p < 0.05). A high prevalence of stunting (35.8%) was observed. Children with indigenous ancestry living in rural areas showed the highest prevalence of malnutrition (59.4%). After controlling for age, gender, and indigenous ancestry, the factors associated with stunting risk were: living in rural area (PR = 1.6; 95%CI 1.2 - 2.1); lower tertile of household wealth index (PR = 1.6; 95%CI 1.1 - 2.3); living in houses made of walking palm (PR = 1.6; 95%CI 1.1 - 2.4); maternal height less than or equal to 146.4 cm (PR = 3.1; 95%CI 1.9 - 5.0); and history of introduction of cow's milk before 30 days of age (PR = 1.4; 95%CI 1.0 - 1.8). Children with updated vaccination cards were inversely associated with stunting risk (PR = 0.7; 95%CI 0.5 - 0.9). Child undernutrition remains a serious public health problem in the Amazon, indicating additional difficulties in facing the problem in this region of the country.

Anemia and micronutrient deficiencies in infants attending at Primary Health Care in Rio Branco, Acre, Brazil.

The present study investigated the factors associated with anemia and micronutrient deficiencies in a cross-sectional analysis conducted with 150 children aged 11 to 14 months attending at basic health centers in Rio Branco, Acre. Venous blood samples were obtained to assess the occurrence of anemia and deficiencies of iron (ID), vitamin A (VAD), and B12 (VB12D). Multiple Poisson regression models were used to identify factors associated with anemia. Anemia, ID, VAD and B12D were observed in 23%, 76%, 18% and 20% of children, respectively. The factors associated with anemia were: do not be only child, living in households without access to cable TV or internet, stunting, late introduction of the complementary feeding (more than 240 days), VAD, VB12D, and current evidence of infection (plasma CRP > 5 mg/L). There was a lower prevalence of anemia among children with birth weight > 3,500 g. Overall, 82% of the study children had at least one of the micronutrient deficiencies (ID, VAD and VB12D). Actions with emphasis on timely and healthy feeding practices, better management of morbidities and supplementation with other micronutrients should be focused on the improvement of child care services at primary health care in this county.

A exposição à vibração e a hipertensão arterial em trabalhadores do transporte coletivo metropolitano

Resumo Objetivo: esclarecer possíveis associações entre a exposição à vibração no trabalho em ônibus da região metropolitana de Belo Horizonte-MG e o autorrelato de hipertensão arterial (HA). Métodos: questionário eletrônico preenchido em entrevista com motoristas (n = 853) e cobradores (n = 754), em pontos de parada das empresas, em 2012. Amostra estimada considerando-se as quotas de trabalhadores (universo amostral = 17.470 motoristas e cobradores) distribuídos nas cidades de Belo Horizonte, Betim e Contagem. A força das associações foi medida pelas razões de prevalência e seus intervalos de confiança de 95%. Após análise bivariada, a associação independente entre vibração e HA foi avaliada por modelos múltiplos de regressão de Poisson com variâncias robustas, ajustada por potenciais fatores de confusão. Resultados: quanto à resposta sobre exposição à vibração: nunca/raramente = 39,7% (nunca = 33,5% e raramente = 6,2%); às vezes = 21%; quase sempre/sempre = 39,3% (quase sempre = 5,9%; sempre = 33,4%). HA foi autorrelatada por 14,3%, sendo 33% mais frequente no grupo que informou sentir o corpo vibrar quase sempre/sempre, quando comparado ao grupo que informou nunca ou raramente (RP: 1,33; IC 95%: 1,02-1,73). Conclusão: a associação entre vibração e hipertensão salienta a relevância de se considerar a exposição ocupacional na abordagem das morbidades cardiovasculares em adultos.

Epidemiology of Pelvic Fractures in Germany: Considerably High Incidence Rates among Older People.

Epidemiological data about pelvic fractures are limited. Until today, most studies only analyzed inpatient data. The purpose of this study was to estimate incidence rates of pelvic fractures in the German population aged 60 years or older, based on outpatient and inpatient data. We conducted a retrospective population-based observational study based on routine data from a large health insurance company in Germany. Age and sex-specific incidence rates of first fractures between 2008 and 2011 were calculated. We also standardized incidence rates with respect to age and sex in the German population. Multiple Poisson regression models were used to evaluate the association between the risk of first pelvic fracture as outcome and sex, age, calendar year and region as independent variables. The total number of patients with a first pelvic fracture corresponded to 8,041 and during the study period 5,978 insured persons needed inpatient treatment. Overall, the standardized incidence rate of all first pelvic fractures was 22.4 [95% CI 22.0–22.9] per 10,000 person-years, and the standardized incidence rate of inpatient treated fractures 16.5 [16.1–16.9]. Our adjusted regression analysis confirmed a significant sex (RR 2.38 [2.23–2.55], p < 0.001, men as reference) and age effect (higher risk with increasing age, p < 0.001) on first fracture risk. We found a slight association between calendar year (higher risk in later years compared to 2008, p = 0.0162) and first fracture risk and a further significant association with region (RR 0.92 [0.87–0.98], p = 0.006, Westfalen-Lippe as reference). The observed incidences are considerably higher than incidences described in the international literature, even if only inpatient treated pelvic fractures are regarded. Besides which, non-inclusion of outpatient data means that a relevant proportion of pelvic fractures are not taken into account. Prevention of low energy trauma among older people remains an important issue.

Predictors of hospitalization for lower respiratory tract infection in children aged

Young age, adverse environmental conditions and infectious agents are established risk factors of lower respiratory tract infection (LRTI), whereas pneumococcal conjugate vaccines may be protective. To explore their relative role as predictors of hospitalizations under the continental climate prevailing in the province of Quebec, Canada, an ecological study was performed. Records with a main diagnosis of LRTI in children born during 2007-2010 and observed up to their second-year anniversary were extracted from the provincial hospital administrative database. Respiratory virus surveillance data and statistics on ambient air temperature were obtained. Vaccine use in different birth cohorts was derived from the Quebec City Immunization Registry. Additive and multiplicative Poisson regression models were applied to estimate attributable fractions. Age, month of birth, ambient temperature, and respiratory syncytial virus (RSV), human metapneumovirus (hMPV) and influenza-positive test proportions were significant predictors of LRTI hospitalizations. No substantial differences were observed in cohorts exposed to the 7-valent or 10-valent pneumococcal conjugate vaccines. In the additive model, the fraction of hospitalizations explained by temperature variation was 37%, whereas RSV circulation explained 28%, hMPV 4% and influenza 1%. Complex interplay between biological, environmental and social mechanisms may explain the important role of ambient air temperature in predicting LRTI hospitalization risk in young children.

Incidence of hospitalization in patients with head and neck cancer treated with intensity-modulated radiation therapy.

Patients with primary head and neck cancer managed with radiation therapy (RT) +/- chemotherapy may experience significant treatment-related toxicities. We assessed hospitalization as a metric for severe treatment-related toxicities and evaluated patient and treatment factors for possible association. A retrospective review was performed on 147 patients with head and neck cancer treated with definitive or adjuvant intensity-modulated radiation therapy (IMRT) +/- chemotherapy. Multiple Poisson regression model was used to analyze relationships between patient or treatment factors and number of hospital stays during RT and within 8 weeks after RT. Multivariate analysis showed preexisting diabetes or pulmonary disease, primary carcinoma of oral cavity, and prescribed radiation dose (p < .05) were associated with increased number of patient hospital stays during or shortly after RT. We found that 34.7% of patients experienced a chemoradiation toxicity-related hospitalization during or shortly after treatment. Prior pulmonary disease, diabetes, and increasing prescribed radiation dose were associated with increased hospital stays.

Risk Factors for Prolonged Mechanical Ventilation for Children on Ventricular Assist Device Support

Patients with end-stage heart failure possess many attributes that place them at risk for prolonged mechanical ventilation (MV). However, there are only limited data on MV support among children after ventricular assist device (VAD) implantation. We report the duration of MV after VAD placement, indications for respiratory support in the postimplantation period, and associated patient factors. This single-center retrospective study included 43 consecutive children (aged <18 years) with end-stage heart failure who were supported with a VAD as a bridge to transplantation from January 2005 to December 2011. Multivariable analysis was performed using the multiple Poisson regression model for the duration of MV. Overall, 33% (n = 14) remained on MV until heart transplant or death. Of those requiring pre-VAD extracorporeal membrane oxygenation (ECMO) support, 63% (n = 12 of 19) remained on MV until heart transplant or death compared with 8% (n = 2 of 24) among those not on ECMO before VAD (p < 0.001). Patients with moderate or severe mitral regurgitation while on VAD support had 1.7-times more MV days compared with those with none or trivial on-VAD mitral regurgitation. In addition, previous support on ECMO, those with moderate or severe tricuspid regurgitation, and those with only left VAD implants had an increased risk of prolonged MV. Our results suggest that VAD recipients previously supported on ECMO, those with moderate or severe mitral regurgitation, moderate or severe tricuspid regurgitation, and those with only left VAD implants had an increased risk of prolonged MV. Future studies in larger cohorts are necessary to confirm the findings from this single-institutional experience.