We report the hospitalization risk among participants and non-participants in a Fresenius Medical Care North America program offering monitored intradialytic oral nutritional supplements (ONS) for chronic hemodialysis patients with low serum albumin. All ONS program-eligible patients with albumin <= 3.5 g/dL in Q4-2009 without ONS in the prior 90-days and did not receive any ONS external to the program were eligible. Participants initiated ONS within Q4-2009 while controls did not receive ONS at anytime during the follow-up period. In addition to the full study cohort, a 1: 1 geographic region and propensity score matched study population (utilizing case-mix, labs, access type, 30-day prior hospitalization, incident patient status) was followed for hospitalization until 12/31/2010. The sensitivity analysis restricted eligible control patients to 30+ day survivors after qualifying for participation, providing sufficient "lag" period for the opportunity to be offered ONS. Hospitalization was less in the ONS cohort at 2.5 vs. 2.7 episodes/patient-year and corresponding 19.2 vs. 20.4 hospital days/patient-year (both p<0.001) as well as a lower observed death rate of 29.4% vs. 36.6%. Hazard ratios for 1st hospitalization with ONS were 0.93 (0.90, 0.96) for the entire cohort, 0.97 (0.92, 1.02) in the matched cohort, and 0.90 (0.86, 0.95) in the matched "lag" cohort - all adjusted for case-mix, prior hospitalization within 30 days of study entry, SMR and laboratory results. Chronic hemodialysis patients with albumin <= 3.5 g/dL who received monitored intradialytic ONS exhibited significantly lower hospitalization risk than similar, matched patient controls. These results require confirmation in prospective clinical trials.

Use of intradialytic oral nutritional supplements (ONS) improve nutritional biomarkers but an association with survival has not been proven. We evaluated mortality, comparing patients with serum albumin ≤3.5 g/dL who received monitored ONS (at no patient cost) during chronic outpatient hemodialysis (HD) as part of a standardized national program in Fresenius Medical Care, North America facilities, with eligible patients who did not receive ONS (controls). Participation was promoted but not mandatory. Study enrolment covered Q4‐2009 and follow‐up was until 12/31/10. Patients who received ONS outside of the program or started during 2010 were excluded. Two protein bars and two liquid formulations of ONS options were offered, which were available until albumin was ≥ 4.0 g/dL. Crude mortality in the ONS group (N= 7,264) was 29.4% vs. 36.6% for controls (N= 13,853), p<0.001. Compared to controls, the unadjusted mortality hazard ratio for ONS was 0.70 (0.67, 0.74) and after adjustment for baseline case‐mix and 5 quality indicators was 0.68 (0.64, 0.71). Although limited by the observational design, these results indicate favorable survival associated with ONS use in malnourished chronic HD patients with albumin ≤3.5 g/dL.

BACKGROUNDInsufficient clinical data exist to determine whether provision of oral nutritional supplements during dialysis can improve survival in hypoalbuminemic maintenance hemodialysis patients.STUDY DESIGNRetrospective matched-cohort study.SETTING & PARTICIPANTSAll oral nutritional supplement program-eligible in-center maintenance hemodialysis patients with albumin level ≤3.5 g/dL in quarter 4 of 2009 without oral nutritional supplements in the prior 90 days at Fresenius Medical Care, North America facilities.QUALITY IMPROVEMENT PLANMonitored intradialytic oral nutritional supplements were provided to eligible maintenance hemodialysis patients upon physician order, to continue for a year or until serum albumin level was ≥4.0 g/dL.OUTCOMEMortality (including deaths and withdrawals), followed up until December 31, 2010.MEASUREMENTSBoth an intention-to-treat (ITT) and an as-treated analysis was performed using a 1:1 geographic region and propensity score-matched study population (using case-mix, laboratory test, access type, 30-day prior hospitalization, and incident patient status) comparing patients treated with intradialytic oral nutritional supplements with usual-care patients. Cox models were constructed, unadjusted and adjusted for facility standardized mortality ratio and case-mix and laboratory variables.RESULTSThe ITT and as-treated analyses both showed lower mortality in the oral nutritional supplement group. The conservative ITT models with 5,227 matched pairs had 40% of controls subsequently receiving oral nutritional supplements after January 1, 2010 (because many physicians delayed participation), with comparative death rates of 30.1% versus 30.4%. The corresponding as-treated (excluding crossovers) death rates for 4,289 matched pairs were 30.9% versus 37.3%. The unadjusted ITT mortality HR for oral nutritional supplement use was 0.95 (95% CI, 0.88-1.01), and the adjusted HR was 0.91 (95% CI, 0.85-0.98); the corresponding as-treated HRs were 0.71 (95% CI, 0.66-0.76) and 0.66 (95% CI, 0.61-0.71) before and after adjustment, respectively.LIMITATIONSLimited capture of oral nutritional supplement intake outside the facility and potential residual confounding from unmeasured variables, such as dietary intake.CONCLUSIONSMaintenance hemodialysis patients with albumin levels ≤3.5 g/dL who received monitored intradialytic oral nutritional supplements showed survival significantly better than similar matched patient controls, with the as-treated analysis highlighting the potentially large effect of this strategy in clinical practice.

Use of intradialytic oral nutritional supplements (ONS) improve nutritional biomarkers but an association with hospitalization risk is unexplored. We evaluated hospitalization events from all causes (HOS), comparing patients with serum albumin ≤3.5 g/dL who received monitored ONS (at no patient cost) during chronic outpatient hemodialysis (HD) as part of a standardized national program in Fresenius Medical Care, North America facilities, with eligible patients who did not receive ONS (controls). Study enrolment covered Q4‐2009 and follow‐up was until 12/31/10. Patients who received ONS outside of the program or started during 2010 were excluded. Two protein bars and two liquid formulations of ONS options were offered – – which were available until albumin was ≥ 4.0 g/dL. Overall, HOS was lower in the ONS group (N= 7,264) at 2.5 vs. 2.7 episodes/pt‐year for controls (N= 13,853), p<0.001. The unadjusted time to 1st HOS hazard ratio for ONS was 0.92 (0.88, 0.95) and after adjustment for baseline case‐mix, HOS 30‐days prior to study entry, and 5 HD quality indicators was 0.93 (0.90, 0.96). Although limited by the observational design, these results indicate lower HOS associated with ONS use in malnourished chronic HD patients with albumin ≤3.5 g/dL.