Use of phosphate-binding agents is associated with a lower risk of mortality
2013; Elsevier BV; Volume: 84; Issue: 5 Linguagem: Inglês
10.1038/ki.2013.185
ISSN1523-1755
AutoresJorge B. Cannata‐Andía, José Luis Fernández Martín, Francesco Locatelli, Gérard M. London, José Luis Górriz, Jürgen Floege, Markus Ketteler, Aníbal Ferreira, Adrian Covic, Leszek Rutkowski, Dimitrios Memmos, Willem Jan W. Bos, Vladimı́r Teplan, Judit Nagy, Christian Tielemans, Dierik Verbeelen, David Goldsmith, Reinhard Kramar, Pierre‐Yves Martin, Rudolf P. Wüthrich, Draško Pavlović, M Benedik, José Emilio Hernández Sánchez, Pablo Martínez‐Camblor, Manuel Naves‐Díaz, Juan Jesús Carrero, Carmine Zoccali,
Tópico(s)Potassium and Related Disorders
ResumoHyperphosphatemia has been associated with higher mortality risk in CKD 5 patients receiving dialysis. Here, we determined the association between the use of single and combined phosphate-binding agents and survival in 6797 patients of the COSMOS study: a 3-year follow-up, multicenter, open-cohort, observational prospective study carried out in 227 dialysis centers from 20 European countries. Patient phosphate-binding agent prescriptions (time-varying) and the case-mix–adjusted facility percentage of phosphate-binding agent prescriptions (instrumental variable) were used as predictors of the relative all-cause and cardiovascular mortality using Cox proportional hazard regression models. Three different multivariate models that included up to 24 variables were used for adjustments. After multivariate analysis, patients prescribed phosphate-binding agents showed a 29 and 22% lower all-cause and cardiovascular mortality risk, respectively. The survival advantage of phosphate-binding agent prescription remained statistically significant after propensity score matching analysis. A decrease of 8% in the relative risk of mortality was found for every 10% increase in the case-mix–adjusted facility prescription of phosphate-binding agents. All single and combined therapies with phosphate-binding agents, except aluminum salts, showed a beneficial association with survival. The findings made in the present association study need to be confirmed by randomized controlled trials to prove the observed beneficial effect of phosphate-binding agents on mortality. Hyperphosphatemia has been associated with higher mortality risk in CKD 5 patients receiving dialysis. Here, we determined the association between the use of single and combined phosphate-binding agents and survival in 6797 patients of the COSMOS study: a 3-year follow-up, multicenter, open-cohort, observational prospective study carried out in 227 dialysis centers from 20 European countries. Patient phosphate-binding agent prescriptions (time-varying) and the case-mix–adjusted facility percentage of phosphate-binding agent prescriptions (instrumental variable) were used as predictors of the relative all-cause and cardiovascular mortality using Cox proportional hazard regression models. Three different multivariate models that included up to 24 variables were used for adjustments. After multivariate analysis, patients prescribed phosphate-binding agents showed a 29 and 22% lower all-cause and cardiovascular mortality risk, respectively. The survival advantage of phosphate-binding agent prescription remained statistically significant after propensity score matching analysis. A decrease of 8% in the relative risk of mortality was found for every 10% increase in the case-mix–adjusted facility prescription of phosphate-binding agents. All single and combined therapies with phosphate-binding agents, except aluminum salts, showed a beneficial association with survival. The findings made in the present association study need to be confirmed by randomized controlled trials to prove the observed beneficial effect of phosphate-binding agents on mortality. During the past decade, knowledge on the pathogenesis and management of chronic kidney disease mineral bone disorders (CKD-MBD) has grown considerably, and the diagnosis, prognosis, and management of these disorders is now formally systematized in specific KDIGO guidelines.1.Slatopolsky E. Moe S. 50 years of research and discovery in chronic kidney disease and mineral & bone disorder: the central role of phosphate.Kidney Int Suppl. 2011; 121: S1-S2Abstract Full Text Full Text PDF PubMed Scopus (1) Google Scholar, 2.Goodman W.G. Quarles L.D. Development and progression of secondary hyperparathyroidism in chronic kidney disease: lessons from molecular genetics.Kidney Int. 2008; 74: 276-288Abstract Full Text Full Text PDF PubMed Scopus (136) Google Scholar, 3.Kidney Disease: Improving Global Outcomes (KDIGO) CKD-MBD Work Group KDIGO clinical practice guideline for the diagnosis, evaluation, prevention, and treatment of Chronic Kidney Disease-Mineral and Bone Disorder (CKD-MBD).Kidney Int Suppl. 2009; 76: S1-S130Google Scholar, 4.Locatelli F. Cannata-Andia J.B. Drueke T.B. et al.Management of disturbances of calcium and phosphate metabolism in chronic renal insufficiency, with emphasis on the control of hyperphosphataemia.Nephrol Dial Transplant. 2002; 17: 723-731Crossref PubMed Scopus (184) Google Scholar The control of serum phosphorus at all stages of CKD is considered key to improve clinical outcomes in CKD-MBD, including survival.5.Floege J. Kim J. Ireland E. et al.Serum iPTH, calcium and phosphate, and the risk of mortality in a European haemodialysis population.Nephrol Dial Transplant. 2011; 26: 1948-1955Crossref PubMed Scopus (362) Google Scholar, 6.Cannata-Andia J.B. Naves-Diaz M. Phosphorus and survival: key questions that need answers.J Am Soc Nephrol. 2009; 20: 234-236Crossref PubMed Scopus (13) Google Scholar, 7.Zoccali C. Ruggenenti P. Perna A. et al.Phosphate may promote CKD progression and attenuate renoprotective effect of ACE inhibition.J Am Soc Nephrol. 2011; 22: 1923-1930Crossref PubMed Scopus (185) Google Scholar, 8.Palmer S.C. Hayen A. Macaskill P. et al.Serum levels of phosphorus, parathyroid hormone, and calcium and risks of death and cardiovascular disease in individuals with chronic kidney disease: a systematic review and meta-analysis.JAMA. 2011; 305: 1119-1127Crossref PubMed Scopus (533) Google Scholar, 9.Naves-Diaz M. Passlick-Deetjen J. Guinsburg A. et al.Calcium, phosphorus, PTH and death rates in a large sample of dialysis patients from Latin America. The CORES Study.Nephrol Dial Transplant. 2011; 26: 1938-1947Crossref PubMed Scopus (120) Google Scholar, 10.Pelletier S. Roth H. Bouchet J.L. et al.Mineral and bone disease pattern in elderly haemodialysis patients.Nephrol Dial Transplant. 2010; 25: 3062-3070Crossref PubMed Scopus (35) Google Scholar In clinical and experimental studies, phosphorus accumulation has been shown to have a negative impact in several aspects of the CKD-MBD constellation, such as parathyroid hyperplasia, vascular calcification, cardiovascular disease, bone strength, bone mass, and bone fractures.2.Goodman W.G. Quarles L.D. Development and progression of secondary hyperparathyroidism in chronic kidney disease: lessons from molecular genetics.Kidney Int. 2008; 74: 276-288Abstract Full Text Full Text PDF PubMed Scopus (136) Google Scholar, 10.Pelletier S. Roth H. Bouchet J.L. et al.Mineral and bone disease pattern in elderly haemodialysis patients.Nephrol Dial Transplant. 2010; 25: 3062-3070Crossref PubMed Scopus (35) Google Scholar, 11.Goodman W.G. Vascular calcification in chronic renal failure.Lancet. 2001; 358: 1115-1116Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar, 12.Roman-Garcia P. Carrillo-Lopez N. Fernandez-Martin J.L. et al.High phosphorus diet induces vascular calcification, a related decrease in bone mass and changes in the aortic gene expression.Bone. 2010; 46: 121-128Abstract Full Text Full Text PDF PubMed Scopus (130) Google Scholar, 13.Hruska K. Mathew S. Lund R. et al.Cardiovascular risk factors in chronic kidney disease: does phosphate qualify?.Kidney Int Suppl. 2011; 79: S9-S13Abstract Full Text Full Text PDF Scopus (39) Google Scholar, 14.Wigner N.A. Luderer H.F. Cox M.K. et al.Acute phosphate restriction leads to impaired fracture healing and resistance to BMP-2.J Bone Miner Res. 2010; 25: 724-733PubMed Google Scholar The potential adverse effects of high serum phosphorus and/or phosphorus accumulation for human health are not limited to patients with end-stage kidney disease and extends to stages 2–4 CKD7.Zoccali C. Ruggenenti P. Perna A. et al.Phosphate may promote CKD progression and attenuate renoprotective effect of ACE inhibition.J Am Soc Nephrol. 2011; 22: 1923-1930Crossref PubMed Scopus (185) Google Scholar and to the general population.15.Foley R.N. Phosphorus comes of age as a cardiovascular risk factor.Arch Intern Med. 2007; 167: 873-874Crossref PubMed Scopus (15) Google Scholar Maintaining a serum phosphorus level as close to normal values as possible has become a challenge in the management of CKD-MBD. Several new phosphate-binding agents (PBAs) have been developed and reached the market just at turn of the last century.16.Chertow G.M. Burke S.K. Raggi P. Sevelamer attenuates the progression of coronary and aortic calcification in hemodialysis patients.Kidney Int. 2002; 62: 245-252Abstract Full Text Full Text PDF PubMed Scopus (1327) Google Scholar,17.Raggi P. Vukicevic S. Moyses R.M. et al.Ten-year experience with sevelamer and calcium salts as phosphate binders.Clin J Am Soc Nephrol. 2010; 5: S31-S40Crossref PubMed Scopus (44) Google Scholar So far, all available PBAs have proven to be effective in reducing serum phosphorus, but their effects on clinical outcomes remain unknown, and the need of large-scale trials based on clinical end points cannot be overemphasized. However, funding and organizing such trials remains a tantalizing undertaking. In this scenario, observational studies testing the comparative effectiveness of PBAs may provide important information to further explore the hypothesis that these medications may reduce mortality in stage 5D-CKD patients. In this regard, a large cohort study by Isakova et al.18.Isakova T. Gutierrez O.M. Chang Y. et al.Phosphorus binders and survival on hemodialysis.J Am Soc Nephrol. 2009; 20: 388-396Crossref PubMed Scopus (309) Google Scholar showed that treatment with phosphorus binders is independently associated with decreased mortality, whereas in other analyses based on an incident USRDS cohort that started dialysis in 1996–1997—at a time when only calcium-containing phosphate binders were used in the United States—no association was found between the use of these agents and mortality.19.Winkelmayer W.C. Liu J. Kestenbaum B. Comparative effectiveness of calcium-containing phosphate binders in incident US dialysis patients.Clin J Am Soc Nephrol. 2011; 6: 175-183Crossref PubMed Scopus (37) Google Scholar The use of instrumental variable analysis techniques may help answer the question.20.Stel V.S. Dekker F.W. Zoccali C. et al.Instrumental variable analysis.Nephrol Dial Transplant. 2012https://doi.org/10.1093/ndt/gfs310Google Scholar The death risk in European patients on chronic dialysis is lower than that of US patients, and risk factors in these two populations differ in part.21.Goodkin D.A. Bragg-Gresham J.L. Koenig K.G. et al.Association of comorbid conditions and mortality in hemodialysis patients in Europe, Japan, and the United States: the Dialysis Outcomes and Practice Patterns Study (DOPPS).J Am Soc Nephrol. 2003; 14: 3270-3277Crossref PubMed Scopus (628) Google Scholar For example, body mass index (BMI) is substantially higher in American patients22.Kramer H.J. Saranathan A. Luke A. et al.Increasing body mass index and obesity in the incident ESRD population.J Am Soc Nephrol. 2006; 17: 1453-1459Crossref PubMed Scopus (256) Google Scholar than in European patients.23.Hecking E. Bragg-Gresham J.L. Rayner H.C. et al.Haemodialysis prescription, adherence and nutritional indicators in five European countries: results from the Dialysis Outcomes and Practice Patterns Study (DOPPS).Nephrol Dial Transplant. 2004; 19: 100-107Crossref PubMed Scopus (123) Google Scholar These differences are of potential relevance because the prescription of phosphate binders is strongly associated with better nutritional status, i.e., higher BMI and other nutritional indicators.24.Lopes A.A. Tong L. Thumma J. et al.Phosphate binder use and mortality among hemodialysis patients in the Dialysis Outcomes and Practice Patterns Study (DOPPS): evaluation of possible confounding by nutritional status.Am J Kidney Dis. 2012; 60: 90-101Abstract Full Text Full Text PDF PubMed Scopus (139) Google Scholar Thus, exploring the link between the use of phosphate binders and major clinical outcomes in European patients may provide relevant information for advancing knowledge on this issue. One of the main aims of the Current management Of Secondary hyperparathyroidism: A Multicenter Observational Study (COSMOS), which included randomly selected patients in 227 dialysis centers from 20 European countries, was to investigate the association between PBA prescription and survival in European patients on dialysis. In the analyses presented herein, in order to limit bias by indication, we modeled the relationship between PBA prescription and clinical outcomes by time-varying, multivariable Cox regression analysis, propensity score matching, and instrumental variable analysis. A total of 6797 patients were recruited for COSMOS, 4500 of them randomly selected at baseline and 2297 to replace patients lost to follow-up. Patients who had only baseline data (no follow-up) or patients with lacking information on prescription of PBAs were excluded. After exclusions, 6297 patients (4313 (68.5%) randomly selected and 1984 (31.5%) replacements) were available for analysis. The main baseline characteristics of the patients included in the study are detailed in Table 1. Patients not prescribed PBAs represented 14.9% of the full cohort, whereas PBAs-prescribed patients made up the remaining 85.1%. The latter were younger, with higher BMI; there were more men and smokers, and fewer diabetics. They referred less events related to cardiovascular disease, they had been on HD for a longer period, and they received more hours of dialysis per week. In the group of patients prescribed PBAs, there were also more patients who were prescribed Vitamin D receptor activators (VDRAs), calcimimetics, and erythropoietin-stimulating agents, and they showed higher serum levels of phosphorus, parathyroid hormone (PTH), and albumin. The propensity score–matched subcohorts showed no differences in the characteristics of patients prescribed or not prescribed PBAs (Table 1).Table 1Main baseline characteristics of patients included in the studyFull cohortPropensity score–matched cohortAll patients (6297)Untreated (N=941)Treated (N=5356)P-valueUntreated (N=823)Treated (N=823)P-valueSex (% males)60.857.461.40.02156.756.10.8Age (years) (mean±s.d.)64.0±14.469.1±12.763.1±14.5<0.00169.0±12.669.0±12.60.9BMI (kg/m2) (mean±s.d.)25.4±6.324.7±5.125.6±6.5<0.00124.7±5.124.6±4.40.6Current smokers (%)13.910.614.50.00210.910.70.9Diabetics (%)30.636.329.6<0.00135.735.81.0CVD history (%)72.075.371.50.01674.874.71.0Parathyroidectomy (%)4.93.85.10.14.03.60.7Months on HD (mean±s.d.)38.9±49.532.3±50.140.1±49.2<0.00132.9±52.533.4±40.20.8Hours of dialysis per week (mean±s.d.)12.0±2.111.6±2.312.1±2.1<0.00111.7±2.411.6±2.00.6Dialysis technique0.40.9 HD conventional low flux (%)54.155.353.956.656.1 HD conventional high flux (%)37.135.337.532.833.8 Hemodiafiltration and others (%)8.89.58.710.610.1Calcium concentration in dialysate<0.0010.7 2.5mEq/l (%)30.021.631.520.920.8 3.0mEq/l (%)50.053.749.455.353.8 3.5mEq/l (%)19.924.719.123.825.4Patients treated with VDRAs (%)47.539.548.9<0.00140.241.40.7Patients treated with calcimimetics (%)6.23.76.60.0013.32.90.7Patients treated with ESAs (%)90.487.790.90.00287.888.10.9PTH (pg/ml) (median (IQR))201.0 (275.2)186.5 (232.9)206.0 (287.0)<0.001182.0 (230.0)162.4 (220.5)0.8Calcium (mg/dl) (mean±s.d.)9.1±0.99.0±0.89.1±0.90.19.1±0.89.1±0.90.2Phosphorus (mg/dl) (mean±s.d.)5.4±1.74.7±1.65.5±1.7<0.0014.7±1.64.8±1.50.4Albumin (g/dl) (mean±s.d.)3.8±0.53.7±0.63.8±0.5<0.0013.7±0.63.7±0.51.0Hemoglobin (g/dl) (mean±s.d.)11.5±1.511.5±1.511.5±1.50.911.5±1.511.5±1.51.0Abbreviations: BMI, body mass index; CVD, cardiovascular disease; ESAs, erythropoietin-stimulating agents; HD, hemodialysis; IQR, interquartile range; PTH, parathyroid hormone; VDRAs, vitamin D receptor activators. Open table in a new tab Abbreviations: BMI, body mass index; CVD, cardiovascular disease; ESAs, erythropoietin-stimulating agents; HD, hemodialysis; IQR, interquartile range; PTH, parathyroid hormone; VDRAs, vitamin D receptor activators. The comparison between random baseline and replacement patients is shown in the Supplementary Material online. The replacement patients were younger, with more men and higher BMIs. Diabetes as a cause of end-stage renal failure was more frequent, and consequently there were more diabetics on HD. Replacement patients showed a lower number of patients with cardiovascular disease history and parathyroidectomies; conventional high-flux dialysis was less used in this group. Download .pdf (.29 MB) Help with pdf files Supplementary Information During the 3-year follow-up, the overall COSMOS crude all-cause mortality rate was 13.3 deaths per 100 patient-years, 14.2 in baseline random patients, and 10.8 in replacement patients. The crude cardiovascular mortality rate was 5.9 cardiovascular deaths per 100 patient-years, 6.4 in baseline random patients, and 4.6 in replacement patients. During that period, 1642 patients died (26.1%). The mean time of follow-up was 23.5 months (median 24.0), in the whole study, 25.2 months (median 30.0) in the random baseline patient group, and 19.9 months (median 18) in the replacement patients group. During the 3-year follow-up, 4430 patients were always prescribed PBAs, 451 were never prescribed PBAs, and 1416 were required to either stop or initiate PBA prescription during follow-up. The percentages of patient-years prescribed PBAs either as monotherapy or combined therapy were as follows: monotherapy: calcium-containing PBAs, 36.7%; sevelamer, 14.9%; aluminum salts, 3.3%; lanthanum carbonate, 2.5%; and others, 2.6%. Combined therapy: calcium-containing+sevelamer, 11.1%; calcium-containing+aluminum salts, 3.3%; calcium-containing+lanthanum carbonate, 1.6%; sevelamer+aluminum salts, 1.4%; sevelamer+lanthanum carbonate, 0.8%; and other combinations, 4.4%. In univariate analyses, patients prescribed PBAs showed a 53% (95% confidence interval (CI): 48–58%) and a 42% (95% CI: 31–51%) lower all-cause and cardiovascular mortality risk, respectively, as compared with patients not prescribed PBAs (Table 2). After adjustments additively using the three different multivariate models, the PBA-prescribed patients still showed lower all-cause and cardiovascular mortality risks (model 3) with hazard ratios of 0.71 and 0.78, respectively. As Figure 1 shows, unadjusted relative risk of mortality was significantly lower in most patients prescribed PBAs. This was the case in the vast majority of sensitivity analyses by stratification.Table 2Relative all-cause and cardiovascular mortality in patients prescribed versus not prescribed PBAsAll-cause mortalityCardiovascular mortalityNo. of patientsHazard ratio (95% CI)P-valueNo. of patientsHazard ratio (95% CI)P-valueUnivariate62970.47 (0.42–0.52)<0.00162970.58 (0.49–0.69)<0.001Model 1 (general and demographic characteristics)59120.59 (0.52–0.67)<0.00151810.75 (0.61–0.91)0.004Model 2 (model 1+treatments)56660.63 (0.55–0.72)<0.00148850.76 (0.62–0.93)0.009Model 3 (models 1+2+biochemical parameters)52760.71 (0.61–0.82)<0.00145310.78 (0.62–0.97)0.029Abbreviations: BMI, body mass index; CI, confidence interval; CKD, chronic kidney disease; CVD, cardiovascular disease; ESAs, erythropoietin-stimulating agents; HD, hemodialysis; PBAs, phosphate-binding agents; VDRAs, vitamin D receptor activators.General and demographic characteristic variables: country, center funding type (public or private), age, sex, BMI*Variables included in the multivariate model as time-dependent covariates., smoking habit, etiology of CKD, time on HD, diabetes, cardiovascular disease, calcification (valvular+vascular+calciphylaxis), and parathyroidectomy*Variables included in the multivariate model as time-dependent covariates..Treatment variables*Variables included in the multivariate model as time-dependent covariates.: dialysis type, dialysate calcium, hours of HD per week, native vitamin D or calcidiol, VDRAs (calcitriol, alfacalcidol, or paricalcitol), calcimimetics, and ESAs.Biochemical parameter variables*Variables included in the multivariate model as time-dependent covariates.: serum calcium, phosphorus, parathyroid hormone, and albumin plus hemoglobin.All multivariate analyses were stratified by facility.* Variables included in the multivariate model as time-dependent covariates. Open table in a new tab Abbreviations: BMI, body mass index; CI, confidence interval; CKD, chronic kidney disease; CVD, cardiovascular disease; ESAs, erythropoietin-stimulating agents; HD, hemodialysis; PBAs, phosphate-binding agents; VDRAs, vitamin D receptor activators. General and demographic characteristic variables: country, center funding type (public or private), age, sex, BMI*Variables included in the multivariate model as time-dependent covariates., smoking habit, etiology of CKD, time on HD, diabetes, cardiovascular disease, calcification (valvular+vascular+calciphylaxis), and parathyroidectomy*Variables included in the multivariate model as time-dependent covariates.. Treatment variables*Variables included in the multivariate model as time-dependent covariates.: dialysis type, dialysate calcium, hours of HD per week, native vitamin D or calcidiol, VDRAs (calcitriol, alfacalcidol, or paricalcitol), calcimimetics, and ESAs. Biochemical parameter variables*Variables included in the multivariate model as time-dependent covariates.: serum calcium, phosphorus, parathyroid hormone, and albumin plus hemoglobin. All multivariate analyses were stratified by facility. Patients prescribed PBAs at least once since baseline had a lower relative risk of all-cause and cardiovascular mortality than patients never prescribed PBAs during the follow-up period (fully adjusted hazard ratio (aHR)=0.56 (95% CI:0.45–0.68) and aHR=0.59 (95% CI:0.43–0.81), respectively). After excluding the 1416 patients who required either stopping or initiation of PBA therapy, the analyses of the remaining 4881 patients revealed that patients prescribed PBAs had a lower relative risk of all-cause mortality (aHR=0.63 (95% CI: 0.51–0.79)). Similar results were observed in cardiovascular mortality, except that the fully adjusted model was not statistically significant (aHR=0.73 (95% CI: 0.52–1.04))—exposure variable was considered as a fixed covariate, not time-varying, in these last two analyses. The analyses performed by excluding the 4309 patients who used VDRAs at least once showed that patients prescribed PBAs, but had never used VDRAs, also exhibited a lower relative risk of all-cause mortality (aHR=0.73 (95% CI: 0.56–0.93)). Similar results were observed in cardiovascular mortality, but the fully adjusted hazard ratio was not statistically significant (aHR=0.72 (95% CI: 0.49–1.06)). The analysis of all-cause mortality in the propensity score–matched cohorts and in all propensity score quartiles showed a significantly lower relative all-cause mortality risk in patients prescribed PBAs. Similar results were observed in cardiovascular mortality, but no significant differences were found in the third and fourth quartiles (Table 3).Table 3Unadjusted relative all-cause and cardiovascular mortality in PBA-prescribed versus non-prescribed patients using a matched cohort by propensity score and different propensity score quartiles in the full cohortAll-cause mortalityCardiovascular mortalityHazard ratio (95% CI)P-valueHazard ratio (95% CI)P-valuePS-matched cohort (N=1646)0.48 (0.41–0.58)<0.0010.64 (0.49–0.83)0.001Full cohort (N=5597) PS quartile 1 (N=1399)0.60 (0.50–0.71)<0.0010.71 (0.54–0.94)0.018 PS quartile 2 (N=1399)0.47 (0.38–0.60)<0.0010.57 (0.41–0.81)0.001 PS quartile 3 (N=1400)0.64 (0.47–0.89)0.0060.73 (0.46–1.17)0.2 PS quartile 4 (N=1399)0.54 (0.36–0.83)0.0040.71 (0.36–1.41)0.3Abbreviations: CI, confidence interval; PBAs, phosphate-binding agents; PS, propensity score.Propensity score was calculated by using binary logistic regression. Baseline treatment with phosphate binders was introduced as a dependent variable, whereas all the baseline variables used for the full adjustment in Table 2 were included as independent variables. PS quartile 1 is the lowest quartile. Open table in a new tab Abbreviations: CI, confidence interval; PBAs, phosphate-binding agents; PS, propensity score. Propensity score was calculated by using binary logistic regression. Baseline treatment with phosphate binders was introduced as a dependent variable, whereas all the baseline variables used for the full adjustment in Table 2 were included as independent variables. PS quartile 1 is the lowest quartile. The median case-mix–adjusted facility percentage of PBA prescription was 88.5% (ranging from 0 to 100%). Small differences in patient baseline characteristics were observed among the different quartile categories of this variable (Table 4). Patients in the highest quartiles were slightly younger and there were a lower percentage of diabetics, but very similar values in the biochemical parameters were found across categories (including serum phosphorous). The instrumental variable analyses showed an 8 and 7% decrease in all-cause and cardiovascular mortality risk, respectively, per every 10% increase in the case-mix–adjusted facility percentage of PBA prescription in the fully adjusted model (Table 5).Table 4Baseline characteristics of patients by quartiles of adjusted facility percentage of PBA prescriptionPrescription of PBAs 96.0% (N=1580)Sex (% males)58.661.060.962.5Age (years) (mean±s.d.)66.3±14.164.7±14.262.8±14.362.1±14.7BMI (kg/m2) (mean±s.d.)25.4±5.325.9±9.124.9±4.925.5±4.8Current smokers (%)14.113.415.113.2Diabetics (%)34.932.429.725.6CVD history (%)73.473.068.273.6Calcification (%)40.036.739.633.8Parathyroidectomy (%)4.64.05.35.6Months on HD (mean±s.d.)39.3±50.836.9±48.040.3±51.639.0±47.6PTH (pg/ml) (median (IQR))200.0 (269.1)190.9 (260.0)198.0 (286.0)224.0 (292.7)Calcium (mg/dl) (mean±s.d.)9.0±0.99.2±0.89.1±0.99.1±0.9Phosphorus (mg/dl) (mean±s.d.)5.3±1.75.4±1.75.4±1.75.4±1.7Albumin (g/dl) (mean±s.d.)3.7±0.63.8±0.53.9±0.53.8±0.5Hemoglobin (g/dl) (mean±s.d.)11.6±1.511.6±1.411.5±1.511.3±1.7Abbreviations: BMI, body mass index; CVD, cardiovascular disease; HD, hemodialysis; IQR, interquartile range; PBAs, phosphate-binding agents; PTH, parathyroid hormone. Open table in a new tab Table 5Relative all-cause and cardiovascular mortality per every 10% increase in the case-mix–adjusted center percentage of PBA prescriptionAll-cause mortalityCardiovascular mortalityNo. of patientsHazard ratio (95% CI)P-valueNo. of patientsHazard ratio (95% CI)P-valueUnivariate62930.92 (0.88–0.95)<0.00162850.92 (0.86–0.97)0.003Model 1 (general and demographic characteristics)62560.94 (0.90–0.98)0.00362480.94 (0.88–0.99)0.033Model 2 (model 1+treatments)59920.94 (0.90–0.98)0.00259820.94 (0.88–1.00)0.052Model 3 (models 1+2+biochemical parameters)55820.92 (0.89–0.96)<0.00155690.93 (0.87–0.99)0.018Abbreviations: BMI, body mass index; CI, confidence interval; CKD, chronic kidney disease; CVD, cardiovascular disease; HD, hemodialysis; IQR, interquartile range; PBAs, phosphate-binding agents; PTH, parathyroid hormone; VDRAs, vitamin D receptor activators.General and demographic characteristics variables: center funding type (public or private), age, sex, BMI*Variables included in the multivariate model as time-dependent covariates., smoking habit, etiology of CKD, time on HD, diabetes, cardiovascular disease, calcification (valvular+vascular+calciphylaxis), and parathyroidectomy*Variables included in the multivariate model as time-dependent covariates..Treatment variables*Variables included in the multivariate model as time-dependent covariates.: dialysis type, dialysate calcium, hours of HD per week, native vitamin D or calcidiol, VDRAs (calcitriol, alfacalcidol, or paricalcitol), calcimimetics, and erythropoietin-stimulating agents.Biochemical parameter variables*Variables included in the multivariate model as time-dependent covariates.: serum calcium, phosphorus, PTH, albumin, and hemoglobin.All the regression models were stratified by country.* Variables included in the multivariate model as time-dependent covariates. Open table in a new tab Abbreviations: BMI, body mass index; CVD, cardiovascular disease; HD, hemodialysis; IQR, interquartile range; PBAs, phosphate-binding agents; PTH, parathyroid hormone. Abbreviations: BMI, body mass index; CI, confidence interval; CKD, chronic kidney disease; CVD, cardiovascular disease; HD, hemodialysis; IQR, interquartile range; PBAs, phosphate-binding agents; PTH, parathyroid hormone; VDRAs, vitamin D receptor activators. General and demographic characteristics variables: center funding type (public or private), age, sex, BMI*Variables included in the multivariate model as time-dependent covariates., smoking habit, etiology of CKD, time on HD, diabetes, cardiovascular disease, calcification (valvular+vascular+calciphylaxis), and parathyroidectomy*Variables included in the multivariate model as time-dependent covariates.. Treatment variables*Variables included in the multivariate model as time-dependent covariates.: dialysis type, dialysate calcium, hours of HD per week, native vitamin D or calcidiol, VDRAs (calcitriol, alfacalcidol, or paricalcitol), calcimimetics, and erythropoietin-stimulating agents. Biochemical parameter variables*Variables included in the multivariate model as time-dependent covariates.: serum calcium, phosphorus, P
Referência(s)