| Summary: | 碩士 === 長庚大學 === 醫務管理學系 === 104 === Background
The prevalence rate of Type 2 Diabetes (T2DM) is dramatically increasing in Taiwan which demonstrated 7.1% prevalence according the statistic report in 2012. Among those T2DM patients, one-third subjects will develop as chronic kidney disease (CKD) in the life time, which is highly associated with diabetes control. The burden of diabetes leads the complications to higher medical expenditure, especially for the chronic kidney disease and end-stage of renal disease (ESRD). Based on the primary clinic care with appropriate education and surveillance for disease controlling, the incidence rates and severity of complications would be reduced. However, regarding the risk of CKD prediction, there was no substantial study to investigate the risk factors on CKD development based on Taiwanese diabetes patients. Furthermore, the optimal period of surveillance for CKD is still controversial, but it is essential issue for healthcare policy making.
Aim
There were two parts in our study, first, we aimed to estimate the mean sojourn time (MST), also call dwelling time, of CKD using the prevalence and incidence rate of CKD (P/I ratio) approach based on the database of shared care for diabetes. Second, we conducted the prospective cohort design based on shared care of diabetes in Keelung Chang Gung Memorial Hospital to investigate the predictive risk factors for CKD incidence.
Materials and Methods
Based on the database of Shared Care for Diabetes in Keelung Chang Gung Memorial Hospital, that Type 1 diabetes cases were excluded from our study. Those who CKD history by self-report or baseline value of eGFR less than 60 were defined as prevalent cases of CKD that contributed for prevalence calculation. Excluding those prevalent CKD cases, others with 2 or more clinic visits were followed up to last clinic visit or the end of 2014. For the incidence cases, the follow-up time were calculated till the date of first eGFR <60, otherwise, we treated those as censored. The MST was estimated by P/I ratio. Using the prospective cohort, those baseline factors were conducted for risk prediction model estimation. The stepwise method was used for significant variable selection with p-value<0.1 criterion. Both Poisson regression and Cox proportional hazards regression model were employed for risk prediction of CKD incidence, which were reported by adjusted relative risk (aRR) and hazards ratio (aHR) for both respectively after adjustment.
Results
The overall prevalence rate of CKD complication was 19.62% (18.55%~20.69%) among diabetes patients. The prevalence rate was higher in Men compared with female. For those who have coronary heart disease (CAD), hypertension, abnormal HbA1c level, higher triglyceride (TG) or lower high density lipoprotein (HDL), the prevalence rate were higher than others. The overall incidence rate of CKD was 4.50(2.98~6.02)/per 100. For those who were male, CAD, hypertension, higher TG level, the incidence was higher compared with normal subjects. The overall MST was 5.42 years, which were also 6.41 and 4.52 years for male and female respectively. About the risk factors for CKD prediction, the risk of those patients were being male (aRR=1.35, 95%CI:1.16, 1.56), elders(aRR=2.25, 95%CI:2.17, 2.93), hypertension (aRR=1.62, 95%CI:1.35, 1.94), abnormal HbA1c level (aRR=1.23, 95%CI:1.06, 1.45), lower HDL (aRR=1.36, 95%CI:1.16, 1.61), and higher TG (aRR=1.30, 95%CI:1.12, 1.52)were significantly increased. Both Poisson and Cox regression showed the quite close results.
Conclusion
According to our result in MST, the appropriate interval for CKD surveillance would be suggested by 2-3 years (half of MST)。Our study also revealed those diabetes patients with male, elder, hypertension, abnormal HbA1c, lower HDL, and higher TG made significantly effect on CKD incidence.
|
|---|
