Gholian K, Hajian-Tilaki K, Akbari R. Changes of Creatinine and Albumin over two years of follow-up and the associated factors in patients under hemodialysis using generalized estimating equations. , Caspian J Intern Med. 2026; 17 (4) :13
URL:
http://caspjim.com/article-1-4793-fa.html
Changes of Creatinine and Albumin over two years of follow-up and the associated factors in patients under hemodialysis using generalized estimating equations. ، . 1405; 17 (4) :13
URL: http://caspjim.com/article-1-4793-fa.html
چکیده: (41 مشاهده)
Background: This study aimed to investigate the factors that influence changes in creatinine and albumin among dialysis patients by employing the method of generalized estimating equations (GEE) and quadratic inference function (QIF).
Methods: In this historical cohort study, information from 153 dialysis patients, were collected at the Shahid Beheshti Dialysis Center of Babol between 2019 and 2020. Serum albumin and creatinine were measured longitudinal as response variables. The independent variables were gender, age, age at the start of dialysis, history of hypertension (HTN), history of diabetes, history of cardiovascular disease, history of high cholesterol, history of smoking, duration of session of dialysis, dialysis vintage, the number of dialysis per week, and weight. The longitudinal data were analyzed using GEE and QIF models.
Results: The findings revealed an increasing trend in average creatinine levels. The QIF model with unstructured correlation demonstrated the best fit in changes of creatinine. Gender (P=0.001), age (P=0.001), age of starting dialysis (P=0.041), HTN (P=0.001), history of diabetes (P=0.05), history of high cholesterol (P=0.001), dialysis vintage (P=0.001), duration of session of dialysis (p=0.001), the number of dialysis per week (P=0.001), and weight (P=0.002), exhibited a significant association with the serum creatinine level. The relative efficiency of QIF compared to GEE1 and GEE2 were 1.36 and 1.32, respectively. The decreasing trend of albumin over the time showed a significant inverse association (P=0.001) but no association with other clinical characteristics in QIF model. For albumin, the relative efficiency of QIF versus GEE1 and GEE2 was 1.18 and 1.34, respectively.
Conclusion: The clinical characteristics are indicators of changes in creatinine. All three models yielded rather similar coefficient estimates. However, the QIF consistently indicated higher relative efficiency than GEE1 and GEE2.
نوع مطالعه:
Original Article |
موضوع مقاله:
Nephrology دریافت: 1404/3/12 | پذیرش: 1405/1/30 | انتشار: 1405/6/18