final logistic regression model
Recently Published Documents


TOTAL DOCUMENTS

4
(FIVE YEARS 3)

H-INDEX

1
(FIVE YEARS 0)

2022 ◽  
Vol 8 ◽  
Author(s):  
Emily M. Leishman ◽  
Nienke van Staaveren ◽  
Vern R. Osborne ◽  
Benjamin J. Wood ◽  
Christine F. Baes ◽  
...  

Injurious pecking can cause a wide range of damage and is an important welfare and economic issue in turkey production. Aggressive pecking typically targets the head/neck (HN) area, and feather pecking typically targets the back/tail (BT) area; injuries in these separate areas could be used as a proxy for the level of aggressive and feather pecking in a flock. The objective of this study was to identify risk factors for integument injuries in Canadian turkey flocks. A survey containing a questionnaire about housing and management practices and a scoring guide was distributed to 500 turkey farmers across Canada. The farmer scored pecking injuries in two different body areas (HN and BT) on a 0–2 scale on a subset of birds within each flock. Multivariable logistic regression modeling was used to identify factors associated with the presence of HN and BT injuries. The prevalence of birds with integument injuries ranged widely between the flock subsets (HN = 0–40%, BT = 0–97%), however the mean prevalence was low (HN = 6%, BT = 10%). The presence of injuries for logistic regression was defined as flocks with an injury prevalence greater than the median level of injury prevalence in the dataset (3.3% HN and 6.6% BT). The final logistic regression model for HN injuries contained five variables: flock sex, flock age, number of daily inspections, number of different people during inspections, and picking up birds during inspections (N = 62, pR2 = 0.23, α = 0.05). The final logistic regression model for BT injuries contained six variables: flock sex, flock age, litter depth, litter condition, inspection duration, and use of hospital pens for sick/injured birds (N = 59, pR2 = 0.29, α = 0.05). Flock age, and to a lesser extent, sex was associated with both types of injuries. From a management perspective, aggressive pecking injuries appear to be influenced by variables related to human interaction, namely during inspections. On the other hand, the presence of feather pecking injuries, was associated with litter condition and other management factors like separating sick birds. Future research on injurious pecking in turkeys should focus on these aspects of housing and management to better describe the relationship between the identified variables and the prevalence and severity of these conditions.


2020 ◽  
Vol 11 (2) ◽  
pp. 121-125
Author(s):  
Siti Maemun ◽  
Nina Mariana ◽  
Surya Otto Wijaya ◽  
Dina Oktavia ◽  
Vivi Lisdawati ◽  
...  

Latar belakang : Hipoalbuminemia pada pasien rawat inap berkaitan dengan prognosis buruk pasien. Penelitian ini, mengidentifikasi bahwa hipoalbuminemia berat pada awal pasien masuk rawat inap sebagai prediktor andalan penanda laboratorium dalam mortalitas. Metode : Sebuah studi cross sectional pada pasien dewasa dengan hipoalbuminemia (kadar albumin < 3,5 g / dL) pada pasien rawat inap (usia > 18 tahun) pada periode Januari 2013 - Maret 2018. Kami mengevaluasi penanda prediktor kematian. Multivariat dengan regresi logistik diterapkan dalam penelitian ini. Hasil : Dari 747 hipoalbuminemia pada pasien rawat inap dengan rata-rata kadar albumin  pada awal adalah 2,0 ± 0,6 g / dL. Sebagian besar pasien (83,4%) memiliki kadar albumin ≤ 2,5 g / dL (hipoalbuminemia berat), 16,6 persen memiliki > 2,5 g / dL (hipoalbuminemia ringan-sedang). Kondisi yang mendasari pasien adalah infeksi HIV / AIDS (26,9%) dan sepsis (26,6%). Proporsi multiple komorbiditas pada kelompok hipoalbuminemia berat adalah 55,1 persen Pada kelompok hipoalbuminemia berat terutama untuk kadar albumin 2,01 - 2,5 g / dL, angka mortalitas adalah 28,3 persen. Berdasarkan model regresi logistik akhir, faktor risiko kematian meliputi kadar albumin pada awal dan lama rawat pasien. Mortalitas lebih tinggi pada pasien dengan hipoalbuminemia berat (rasio odds yang disesuaikan 2,91, 95% CI 1,88-4,50) dibandingkan pasien dengan hipoalbuminemia ringan-sedang. Kesimpulan: Hipoalbuminemia berat pada awal pasien rawat inap sebagai prediktor penanda kematian di rumah sakit.  Kata kunci: hipoalbuminemia, pasien rawat inap, mortalitas   Abstract Background: Hypoalbuminemia in hospitalized patients has been associated with poor prognosis. In this study, we attempted to identify that severe hypoalbuminemia at baseline in hospitalized patients is a reliable predictor of laboratory marker for mortality. Methods: A cross sectional study on adults of hypoalbuminemia (albumin level < 3.5 g/dL) in hospitalized patients (aged > 18 years old) in period January 2013 - March 2018. We evaluated the predictor marker of mortality. Multivariate with the logistic regression was applied in this study. Results: Of the 747 hypoalbuminemia in hospitalized patients with the mean albumin level at baseline was 2.0 ± 0.6 g/dL. Most patients (83.4 %) had less than or equal to 2.5 g/dL albumin level (severe hypoalbuminemia), 16.6 percent had over 2.5 g/dL (mild-moderate hypoalbuminemia). The underlying condition of patients was HIV/AIDS infection (26.9%) and sepsis (26.6 %). The proportion of multiple comorbidities in the severe hypoalbuminemia group was 55.1percent. In the severe hypoalbuminemia group especially for 2.01 – 2.5 g/dL albumin level, the mortality rate was 28.3 percent. Based on the final logistic regression model, known risk factors of mortality include albumin level at baseline and length of stay. Mortality was higher among patients with severe hypoalbuminemia (adjusted odds ratio 2.91, 95 % CI 1.88-4.50) than patients with mild-moderate hypoalbuminemia. Conclusion: Severe hypoalbuminemia at baseline in the hospitalized patients was a predictor laboratorymarker of hospital mortality.  Keywords: hypoalbuminemia, hospitalized patients, mortality


2019 ◽  
Vol 47 (1) ◽  
pp. 67-73
Author(s):  
Lanying Huang ◽  
Shiang-Fan Chen

SummaryMany previous studies argue that harsher forestry legislation should be enforced to handle the problem of tree poaching. However, empirical studies on the behavioural analysis of poachers’ decision making is largely lacking. Drawing from conversations with 65 inmates imprisoned for Forestry Act offences in Taiwan, we discuss the reasons behind the intention whether or not to stop tree poaching. The majority (81.5%) of the sample expressed their intention to stop tree poaching. Among the 16 demographic, offence and punishment characteristics, we identified only four variables to be included in the final logistic regression model to predict the decision to stop. We found that (1) having no previous experience of stout camphor tree (Cinnamomum kanehirae) theft, taking a log from a stout camphor tree and selling it to buyers; or (2) higher level of education could predict a greater likelihood of intending to quit. Given the limitation of the existing control approach, we propose a restorative justice approach to the poaching problem. A restorative justice approach, instead of focusing solely on the violation of law, recognizes the harm done and forms collaborative work to repair the harm and prevent future wrongdoings. It also helps break the vicious cycle of a poaching subculture.


2017 ◽  
Vol 20 (10) ◽  
pp. 913-921 ◽  
Author(s):  
Younjung Kim ◽  
Hyungjoon Kim ◽  
Dirk Pfeiffer ◽  
Dave Brodbelt

Objectives The objective of this study was to investigate potential risk factors for the diagnosis of feline idiopathic cystitis (FIC) in cats living in a primarily indoor environment. Methods A case-control study focusing on a cohort of cats attending a first-opinion veterinary practice in Seoul, South Korea, from 2012–2016, was undertaken. Data were collected from cats’ owners by questionnaire and analysed using a multivariable logistic regression analysis. Results Fifty-eight cases of FIC and 281 randomly selected controls were surveyed. Over 90% of the cases and controls had no access to the outside, and 100% and 91% of the cases and controls, respectively, were neutered. The estimated prevalence of an FIC diagnosis was 1.77% (95% confidence interval [CI] 1.36–2.18). The final logistic regression model included five variables associated with an FIC diagnosis. Males had 2.34 times the odds of being diagnosed with FIC compared with females (95% CI 1.18–4.62; P = 0.015). Cats reported as not having vantage points had 4.64 times the odds of an FIC diagnosis compared with those reported as having vantage points (95% CI 2.05–10.49; P <0.001). Cats living in an apartment had 2.53 times the odds of an FIC diagnosis compared with those living in a house (95% CI 1.30–4.93; P = 0.006). Cats cohabiting with other cats were more likely to be diagnosed with FIC than those living alone (odds ratio 3.16, 95% CI 1.61–6.22; P = 0.001). Cats using non-clumping litter had 2.62 times the odds of an FIC diagnosis compared with those using clumping litter (95% CI 1.38–4.96; P = 0.003). Conclusions and relevance This study was conducted in a different epidemiological context from previous studies in that the overwhelming majority of the cats studied were housed entirely indoors. This study identified several significant associations related to a cat’s indoor environment. These findings suggest that the cat’s physical and social environment may play a role in the development of FIC.


Sign in / Sign up

Export Citation Format

Share Document