BACKGROUND: Few studies linking hospital discharge records with the population register of chronic dialysis (CD) patients are available. This study aimed to evaluate the frequency and the determinants of hospitalization, taking into account the demographic, clinical and biochemical data. METHODS: We conducted a retrospective cohort study in 3411 patients starting dialysis from 1996-2000, reported to the Lazio Dialysis Registry (RDL) (Italy). These patients were linked with the hospital information system from 1996-2002. Hospital admission probability was calculated using the Kaplan-Meier method. To evaluate the determinants of hospitalization risk we used Cox's proportional hazards for the first admission and a marginal model considering competitive effect of mortality, the Wei-Lin-Weissfeld model, for any admission. RESULTS: We found 7530 hospital admissions, referring to 1711 patients (50.7%), with a rate of 63/100 person-years. The most prevalent diagnoses were "diseases of the genitourinary system", (37.4%), and "diseases of the circulatory system", among secondary diagnoses (46.6%). Hospitalization probability was 34.4% at 1 yr after starting dialysis. The risk of first and any hospital admission was higher (p<0.05) for patients having more than one comorbid disease, hematocrit (Hct) level <30%, serum albumin level <3.5 g/dL, and a low degree of self-sufficiency. CONCLUSIONS: Hospitalization frequency, mainly during the first months of dialysis, suggests the need to improve the early management of chronic renal failure and indicates the importance of preventing complications and maximizing functional status among the dialysis population.
BACKGROUND: Few studies linking hospital discharge records with the population register of chronic dialysis (CD) patients are available. This study aimed to evaluate the frequency and the determinants of hospitalization, taking into account the demographic, clinical and biochemical data. METHODS: We conducted a retrospective cohort study in 3411 patients starting dialysis from 1996-2000, reported to the Lazio Dialysis Registry (RDL) (Italy). These patients were linked with the hospital information system from 1996-2002. Hospital admission probability was calculated using the Kaplan-Meier method. To evaluate the determinants of hospitalization risk we used Cox's proportional hazards for the first admission and a marginal model considering competitive effect of mortality, the Wei-Lin-Weissfeld model, for any admission. RESULTS: We found 7530 hospital admissions, referring to 1711 patients (50.7%), with a rate of 63/100 person-years. The most prevalent diagnoses were "diseases of the genitourinary system", (37.4%), and "diseases of the circulatory system", among secondary diagnoses (46.6%). Hospitalization probability was 34.4% at 1 yr after starting dialysis. The risk of first and any hospital admission was higher (p<0.05) for patients having more than one comorbid disease, hematocrit (Hct) level <30%, serum albumin level <3.5 g/dL, and a low degree of self-sufficiency. CONCLUSIONS: Hospitalization frequency, mainly during the first months of dialysis, suggests the need to improve the early management of chronic renal failure and indicates the importance of preventing complications and maximizing functional status among the dialysis population.
Authors: Anna T Mathew; Lisa Rosen; Renee Pekmezaris; Andrzej Kozikowski; Daniel W Ross; Thomas McGinn; Kamyar Kalantar-Zadeh; Steven Fishbane Journal: Kidney Int Rep Date: 2017-11-03