Mortality for dialysis patients is highest during the

Transcription

1 The Right of Passage: Surviving the First Year of Dialysis Rebecca L. Wingard, Kevin E. Chan, J. Michael Lazarus, and Raymond M. Hakim Fresenius Medical Care North America, Waltham, Massachusetts Mortality risk for dialysis patients is highest in the first year. We previously showed a 41% mortality benefit associated with a pilot case management program for incident hemodialysis patients (n 918). The RightStart Program (RSP) provided prompt medical management and self-management education and was recently expanded to more facilities. We conducted a matched cohort analysis to validate the expanded program s continued effectiveness. Death risk was reduced for RS patients (n 4308) versus matched controls (C; n 4308) by 34% (hazard ratio 0.66, P < ) at 120 d and 22% at 1 yr (hazard ratio 0.78, P < ). RS patients had lower hospitalization during the first year (RS 15.5 days per patient year versus C 16.9, P < 0.01). At 120 d, more RS patients achieved hemoglobin 11 to 12 g/dl (RS 22.4% versus C 19.7%, P < 0.01), ekt/v > 1.2 (RS 66% versus C 53.5%, P < 0.01), albumin > 4.0 g/dl (RS 26% versus C 22%, P < 0.01), and phosphorus 3.5 to 5.5 mg/dl (RS 52.4% versus C 45.4%). At 120 d, RS patients had a greater reduction in catheter use (RS 32% versus C 25%, P < 0.01) and more vitamin D orders (RS 60% versus C 55%, P < 0.01). Expansion of RS to a larger incident patient population results in significant reduction of morbidity and mortality associated with improvement of intermediate outcomes. Clin J Am Soc Nephrol 4: S114 S120, doi: /CJN Mortality for dialysis patients is highest during the first year of chronic renal replacement therapy (Figure 1) (1). According to The U.S. Renal Data System (1), the adjusted first year mortality rate was 24.5% in 2004, which is 17% higher than the mortality rate for dialysis patients in the second year and onward. During the first year, the risk of death rises to a peak at month 3 (Figure 2), when the mortality rate has been estimated to be 40 deaths per 100 patient years (2). This increased mortality soon after the initiation of dialysis seems to be fairly universal, in that death risk is higher within the first 90 to 120 d regardless of age, gender, race (3), or country (personal communication with DOPPS, April 24, 2009). Despite improved survival across all dialysis modalities and ESRD diagnoses, little progress has been made in reducing first-year mortality (4). Several studies have attempted to explain this excess mortality during the first year. Analysis of the Dialysis Outcomes and Practice Patterns cohort reported 15 demographic, laboratory, and comorbid parameters associated with first-year mortality (3). Wolf et al. (5) reported an early mortality association with vitamin D deficiency and higher death rates in those who did not receive vitamin D on the initiation of dialysis. Late nephrology referral was also reported to be associated with a higher risk of death at 1 yr after the initiation of dialysis compared with early referral (6). Additionally, less tangible factors such as mental health and self-care knowledge (7) may also play a role in mortality and hospitalization. Thus, the excess mortality in the first year is likely multifactorial and attributable to a combination of actionable, fixed, and unmeasureable patient factors. Correspondence: Rebecca Wingard, Medical Department, Fresenius Medical Care North America, 750 Old Hickory Blvd, Suite 230, Brentwood, TN Phone: ; Fax: ; rebecca.wingard@fmc-na.com The disproportionately high mortality encountered in the first year inspired the conception and implementation of RightStart, a quality improvement program to provide incident hemodialysis patients with intensive patient education, psychologic evaluation, encouragement, and early clinical intervention in the areas of anemia management, dialysis adequacy, nutrition, and catheter reduction, to achieve quality outcome measures consistent with the National Kidney Foundation Kidney Disease Outcomes Quality Initiative (8). Initially, the RightStart program was implemented as a pilot study in new patients admitted to specific facilities and with the approval of the medical director. Because of practical difficulties and ethical concerns, nursing home residents, transients, non-english speaking, and patients with significant cognitive impairment were excluded from participating in the pilot. Controls were chosen from incident patients in facilities that were geographically proximate but where the RightStart program was not being implemented (9). During this pilot phase, 918 incident hemodialysis patients were enrolled in RightStart from May 2002 to November 2005 and compared with 1020 controls. As previously reported, the incidence rate for mortality was significantly lower in the RightStart patients at 3 mo (P 0.001): 17 deaths per 100 patient years in the RightStart group versus 30 deaths per 100 patient years in the control group (9). In addition, the differences in mortality remained significant at 1 yr of follow-up and there were fewer hospital days for RightStart patients versus controls during the first 90 d (2.1 versus 3.1 cumulative d/90 d, P 0.01). Compared with controls at 90 d, RightStart (RS) patients had a significantly higher average hematocrit (RS 37.9% versus control 36.7%, P 0.001) and albumin (RS 3.70 g/dl versus control 3.66 g/dl, P 0.05). Patients also showed improved dialysis knowledge test scores (surrogate for self-management potential) Copyright 2009 by the American Society of Nephrology ISSN: /

2 Clin J Am Soc Nephrol 4: S114 S120, 2009 The RightStart Program S115 Figure 1. Mortality rates by dialysis vintage. Mortality rates are 17% higher in the first year of dialysis compared with patients who survived the first year of chronic renal replacement therapy. Reproduced from the U.S. Renal Data System (1): U.S. Renal Data System, USRDS 2008 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, from 62.1 to 79.3% (P 0.001) and significant increases in the mental component of the Kidney Disease Quality of Life score (increase from 48.6 to 51.4; P 0.05). No significant benefit in catheter reduction was noted between the two groups at the end of 3 mo. This may reflect the time from referral to the vascular surgeon and creation of a permanent access, to cannulation, which often exceeds the 90-d observation period of the analysis. Overall, the findings suggested that a structured program of prompt medical and educational strategies can effectively improve first-year morbidity and mortality in chronic incident dialysis patients. Given the survival benefit associated with participation in RightStart, the program was formally introduced in 2005 to the general ESRD population to include 200 facilities. The nursing home patient exclusion criteria was eliminated, and the subjective evaluation of mental status was replaced with the objective Mini-Cog assessment for dementia (10). In addition, non-english speaking patients were included, and a Spanish version of the RightStart Handbook was provided to the patient if needed. We conducted a retrospective analysis to determine whether the survival benefit seen in the RightStart pilot study could be effectively translated to a broader and more diverse incident patient population as a facility-based quality improvement program. Materials and Methods Program Description Patients enrolled in the RightStart Program were followed by a RightStart case manager for 120 d or until the completion of the program patient education curriculum. The case manager was a registered renal nurse or renal dietitian who was integrated into the regular staff and covered, on average, six geographically proximate dialysis clinics and an average caseload of 50 patients, which provided annual coverage of 150 patients per full-time case manager. The case manager provided one-on-one patient education and health counseling during 8 to 12 sessions with the patient. In addition, the case manager reviewed the patients medical record, laboratory parameters, and medications and communicated as needed with other team members to facilitate appropriate and prompt care. Specifically, case managers were trained to facilitate care such that the following strategies would be implemented. For anemia and bone disease, a laboratory panel draw was initiated on the first treatment with prompt initiation/adjustment of erythropoietic stimulating agent, iron therapy, vitamin D, and phosphorus binders. Dietary counseling focused on liberal protein and calorie intake (other restrictions added later as indicated) provided by the dietitian, with reinforcement from the case manager. Nutritional supplements were prescribed during dialysis for patients with albumin 3.5 g/dl. For patients with ekt/v 1.2, communication was initiated with the team regarding strategies to improve dialysis adequacy. Placement of permanent access for patients with catheters was facilitated by assisting with surgeon appointments and follow-up and patient education about the adverse events associated with catheters. All patients received monthly medication reviews. Patients with diabetes received monthly foot checks, facilitation of retinopathy screening, and review of glucose control and medication management. Throughout the program, consultation with other team members was initiated as indicated, particularly for the management of psycho-social issues with social worker intervention. All patients received a RightStart Handbook, which was written at a fifth- to eighth-grade level according to principles of low literacy education and served as the guide for teaching content. The RightStart case manager provided treatment options education and assessed patients regarding their understanding of these options. Teaching sessions and content were focused on increasing knowledge and selfmanagement behaviors. The RightStart program was designed to guard against duplication of patient services but instead provided an enhanced model of delivering services by adding an additional focus on the unique needs of incident patients through a comprehensive approach that included patient evaluation, team collaboration, and time devoted to teaching self-management strategies. In clinics where clinical care may have included some of the interventions that were part of the RightStart program, the case manager modified activities to avoid duplication of services (for example, in clinics with an access manager, less activity was needed by the case manager to meet access goals). Although all clinics included a method of patient education, the RightStart approach was unique in its focus on self-management behaviors and low literacy method of teaching to enhance learning for patients of all educational levels. In addition, the RightStart curriculum plan included assignment of education topics to the appropriate discipline (social worker, dietitian), such that the case manager primarily covered topics ideally discussed by a nurse. This particular design for curriculum delivery allowed for teaching of content by the team member with the appropriate expertise. Support and integration of the program was important for program success. The case manager was hired locally and reported to a local manager, with oversight of activities by the corporate Medical Department team, which provided a 3- to 4-d training program. Before initiating RightStart in the clinic, the case manager presented the RightStart program content to the medical director, clinical manager, dietitians, and social workers to gain program support. The case manager informed all staff about his/her role and worked as an additional mem-

3 S116 Clinical Journal of the American Society of Nephrology Clin J Am Soc Nephrol 4: S114 S120, 2009 Figure 2. Adjusted cause-specific mortality in the first months of therapy: incident dialysis patients. Adjusted mortality rises from month 1 to a maximum level in month 3. Reproduced from the U.S. Renal Data System (4): U.S. Renal Data System, USRDS 2007 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, ber of the interdisciplinary team. Communication with team members was used for facilitation of care and input into the patient s plan of care. The case managers also participated in monthly conference calls, an annual educational conference, and ongoing continuous quality improvement (CQI) processes. Case managers collected and evaluated outcomes data and participated in small teams with monthly conference calls to develop and carry out improvement strategies, which would involve collaboration with the clinic s interdisciplinary team. For example, to improve the percent of diabetes patients with monthly foot checks, specific causes of missed foot checks for each individual clinic were determined, and the RightStart case manager worked with the team to make improvements. The patient care goals for RightStart were accepted standards of care in accordance with Kidney Disease Outcomes Quality Initiative and Centers for Medicare & Medicaid Services standards and were present in both the RightStart and control group clinics. However, the core difference in care between the RightStart and control groups was the facilitation of prompt and comprehensive care and patient education directed toward self-management provided by a consistent team member throughout the first 4 mo of treatment. In the RightStart clinics, the population of incident patients was singled out as a population with special needs and separate outcomes evaluation. Specific differences in care are detailed in the Appendix. Patient Population From December 2005 to May 2008, all incident hemodialysis patients at RightStart-based clinics were evaluated for enrollment in the Right- Start Program by the case manager within 2 wk of starting outpatient dialysis. Exclusion criteria were (1) imminent transplant, transfer to a non-rightstart clinic, or transfer to peritoneal dialysis within 2 wk; (2) cognitive impairment as assessed by the Mini-Cog exam; and (3) 18 yr of age. Of these 5042 patients, we were able to match 4308 cases to a non-rightstart patient within the same facility who started dialysis Table 1. Baseline characteristics for patients enrolled in RightStart compared to non-rightstart patients matched by age, gender, cause of ESRD, and facility (control group) RightStart Group Control Group P Value N Age (yr) 62.8 (0.2) 63.1 (0.2) 0.38 Gender (% male) 56.1 (0.5) 56.5 (0.5) 0.71 Diabetic (%) 58.7 (0.5) 53.6 (0.5) Race Black (%) 34.4 (0.5) 34.2 (0.5) Other (%) 6.6 (0.3) 9.5 (0.3) White (%) 59.0 (0.5) 56.3 (0.5) Access at dialysis initiation AVF (%) 18.2 (0.4) 15.5 (0.3) Catheter (%) 72.2 (0.4) 60.5 (0.5) Graft (%) 6.1 (0.2) 8.8 (0.3) Unknown (%) 3.5 (0.1) 15.3 (0.3) Hemoglobin (g/dl) 10.3 (0.02) 10.4 (0.02) Transferrin saturation (%) 20 (0.1) 20 (0.1) 0.69 Albumin (g/dl) 3.5 (0.008) 3.4 (0.008) 0.62 Phosphorus (mg/dl) 4.8 (0.02) 4.9 (0.02) ekt/v 1.26 (0.006) 1.28 (0.006) 0.09 Values are mean (SE). Baseline laboratory values established on the first blood draw after the initiation of dialysis; baseline ekt/v was measured after the first three treatments.

4 Clin J Am Soc Nephrol 4: S114 S120, 2009 The RightStart Program S117 Figure 3. One-year survival of patients enrolled in RightStart compared with all incident hemodialysis patients at Fresenius Medical Care from December 2005 to May within the 365 d preceding the implementation of the RightStart program in the same clinic (control). Additional matching criteria included age, gender, and cause of ESRD. Matching was done such that controls were chosen if they were alive on the day of case enrollment (i.e., matching on vintage in that it was based on the number of days since the initiation of chronic dialysis) (11). All patient data used for the analysis was prospectively charted in the Fresenius Medical Care North America electronic health record at each patient encounter. Implementation of company-wide data collection processes maintained the uniform collection and entry of patient data across all facilities. Further description of the data records can be found in previously published studies (12 14). Figure 4. One-year survival of patients enrolled in RightStart compared with a matched intrafacility cohort of non-rightstart patients. The first year mortality rate was 18 deaths per 100 patient years in the RightStart group versus 24 deaths per 100 patient years in the control group.

5 S118 Clinical Journal of the American Society of Nephrology Clin J Am Soc Nephrol 4: S114 S120, 2009 Table 2. Quality outcome measures in RightStart-enrolled patients compared to non-rightstart patients at 120 d RightStart Group Control Group P Value Access AVF (%) 32.6 (0.5) 23.4 (0.4) Catheter (%) 51.8 (0.5) 47.4 (0.5) Graft (%) 12.0 (0.3) 13.9 (0.3) Unknown (%) 3.5 (0.1) 15.4 (0.3) Hemoglobin Concentration (g/dl) 12.3 (0.02) 12.2 (0.02) Percentage within g/dl 22.4 (0.4) 19.7 (0.4) Iron saturation Level (%) 25 (0.2) 24 (0.2) Percentage within 20 50% 54.9 (0.5) 47.1 (0.5) Albumin Concentration (g/dl) 3.64 (0.007) 3.57 (0.008) Percentage 4.0 g/dl 26.0 (0.4) 22.2 (0.4) Percentage 3.5 g/dl 71.5 (0.4) 63.4 (0.5) Phosphorus Concentration (mg/dl) 5.4 (0.02) 5.5 (0.02) Percentage within mg/dl 52.4 (0.5) 45.4 (0.5) Dialysis adequacy ekt/v 1.36 (0.006) 1.32 (0.006) Percentage ekt/v (0.5) 53.5 (0.5) EPO (% use) 83.0 (0.4) 81.6 (0.4) 0.08 Vitamin D (% use) 60.1 (0.5) 55.4 (0.5) Episodes of foot checks (in first 120 d) 1.9 (0.05) 0.5 (0.03) Days to first laboratory draw 3.23 (0.1) 7.19 (0.3) Days to first EPO order 8.42 (0.3) (0.4) Values are mean (SE). EPO, erythropoietin. Outcomes The outcomes of the study were all-cause mortality and hospitalization as charted in the Fresenius Medical Care North America electronic health record. Data Analysis Standard baseline demographic variables and 17 prespecified quality outcome parameters were assessed for the RightStart and control groups. Case-control matching (described above) was used to balance for baseline differences in the patient characteristics before unadjusted Cox and Poisson regression were used to compare the first-year mortality and hospitalization rates between the RightStart and control patients. Quality outcome parameters on day 120 were also compared using t tests (continuous variables) and the 2 statistic (categorical variables). Results At baseline, no clinically appreciable or statistically significant difference in patient characteristics was noted between the cases and controls (Table 1), with the exception of higher baseline catheter use in the RightStart group. Cox regression modeling showed RightStart patients had a clinically and statistically significant 29% 1-yr survival benefit (P ) compared with all incident hemodialysis patients at Fresenius Medical Care during the study period (Figure 3) and a 22% 1-yr survival benefit (P ) over a matched intrafacility cohort of incident dialysis patients who were not enrolled in the program (Figure 4). Further analysis suggested the survival benefit associated with RightStart was evident early after enrollment, with a significant survival advantage in the first 120 d of dialysis (hazard ratio 0.66, P ). First-year hospitalization was also decreased: 15.5 hospital days per patient year in RightStart patients versus 16.9 hospital days per patient year in the control group (P ). Compared with matched controls, RightStart patients achieved statistically significant improved performance on 16 of 17 quality outcome measures (Table 2) at 120 d after the initiation of chronic dialysis. More specifically, RightStart patients achieved statistically better results in catheter reduction, anemia management, dialysis adequacy, nutrition, and bone disease, and foot checks completed for patients with diabetes (Table 2). Although the difference in average albumin concentration was statistically significant, the difference was of minimal clinical significance. Pre-enrollment mean dialysis knowledge scores improved from 56 to 85% after RightStart educational intervention (P ). Similarly, the pre-enroll-

6 Clin J Am Soc Nephrol 4: S114 S120, 2009 The RightStart Program S119 Appendix. Comparison of clinical care for RightStart versus control groups Item RightStart Control Program length 4 mo NA Case manager Nurse or RD None Labs at first treatment Yes Variable Prompt laboratroy follow-up Yes Variable Anemia Per clinic protocol Per clinic protocol Collaborate with anemia manager Nutrition Liberal protein/calorie intake reinforced Per usual RD counseling Oral nutritional supplements for albumin Supplement programs variable 3.5 g/dl Active collaboration with RD Dialysis adequacy Facilitate changes to improve ekt/v Per usual clinic process Access Facilitate permanent access placement Per usual clinic process Medication reviews Additional monthly reviews Per usual clinic process Diabetes Foot checks Foot checks Glucose control Eye exams Medication management Psycho-social Additional assessment, support, referrals Per usual clinic process to social worker Clinic plan of care Contributes through interdisciplinary Per clinic process team RightStart handbook Yes Usual clinic materials One-to-one education toward Yes Per usual clinic process self-management Consistent health coach Yes Various staff NA, not applicable. ment mean diabetes knowledge score increased from 69 to 88% (P ). Discussion This observational analysis in a large and generalized incident hemodialysis population validates the mortality and morbidity benefits of the RightStart disease management program as shown in a previous pilot study. The analysis found a significant 22% decrease in mortality and an 8% decrease in hospitalization associated with RightStart enrollment compared with matched intrafacility controls. As expected, the effect sizes were reduced compared with the initial pilot findings, likely for three main reasons: this analysis targeted a wider patient population (i.e., inclusion of nursing home patients), the length of the program was increased from 3 (pilot) to 4 mo (this analysis), and the reduction of facility effects through intrafacility matching (as opposed to interfacility controls used in the pilot). The reduction of mortality in the first year of dialysis was recently adopted as a unified national policy by the Kidney Care Partners (15), a broad-based coalition of kidney patient advocacy group, health care professional organizations, dialysis service providers, and manufacturers who seek to improve the quality of care for all Americans with renal disease. The Performance Excellence and Accountability in Kidney Care campaign was launched in June 2009 with the mission of empowering health care providers and patients with many of the tools pioneered in the RightStart program to improve survival in the first year. The program pledges to reduce mortality among first-year dialysis patients by 20% by the end of 2012, which, if successful, could extend the lives of 10,000 ESRD patients. Two main limitations of the study are worth further discussion. There remains the possibility of unmeasured confounding factors that could not be adequately corrected through casecontrol covariate matching. Second, some data points extracted from the medical records might have been misclassified or were missing (i.e., access). In conclusion, implementation of care processes that specifically target accepted quality outcome parameters, increase patient knowledge, and promote behavioral change can effectively decrease first-year mortality in incident dialysis patients; therefore, disease management programs that combine both expert medical care and patient empowerment methods show great promise at promoting longevity for all patients starting dialysis. Acknowledgments We greatly acknowledge the dedication of the RightStart case managers, staff, and nephrologists at participating dialysis units. Financial support for the program was also provided by a grant from Abbott (Abbott Park, IL). We also thank Nancy Gill for helping with the

7 S120 Clinical Journal of the American Society of Nephrology Clin J Am Soc Nephrol 4: S114 S120, 2009 manuscript preparation and Billie Axley and Kathryn McDougall for management of the RightStart Program. Disclosures RightStart is a disease management program of Fresenius Medical Care North America. All authors are employees of Fresenius Medical Care North America and have no financial interests in the RightStart program. References 1. U.S. Renal Data System: 2008 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States, Bethesda, MD, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Khan IH, Catto GR, Edward N, MacLeod AM: Death during the first 90 days of dialysis: A case control study. Am J Kidney Dis 25: , Bradbury BD, Fissell RB, Albert JM, Anthony MS, Critchlow CW, Pisoni RL, Port FK, Gillespie BW: Predictors of early mortality among incident US hemodialysis patients in the Dialysis Outcomes and Practice Patterns Study (DOPPS). Clin J Am Soc Nephrol 2: 89 99, U.S. Renal Data Systems: 2007 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States, Bethesda, MD, National Institutes of Diabetes and Digestive and Kidney Disease, Wolf M, Shah A, Gutierrez O, Ankers E, Monroy M, Tamez H, Steele D, Chang Y, Camargo CA Jr, Tonelli M, Thadhani R: Vitamin D levels and early mortality among incident hemodialysis patients. Kidney Int 72: , Kazmi WH, Obrador GT, Khan SS, Pereira BJ, Kausz AT: Late nephrology referral and mortality among patients with end-stage renal disease: A propensity score analysis. Nephrol Dial Transplant 19: , Curtin RB, Sitter DC, Schatell D, Chewning BA: Selfmanagement, knowledge, and functioning and well-being of patients on hemodialysis. Nephrol Nurs J 31: , Eknoyan G, Levin NW, Steinberg E: The National Kidney Foundation Dialysis Outcomes Quality Initiative. Curr Opin Nephrol Hypertens 6: , Wingard RL, Pupim LB, Krishnan M, Shintani A, Ikizler TA, Hakim RM: Early intervention improves mortality and hospitalization rates in incident hemodialysis patients: RightStart program. Clin J Am Soc Nephrol 2: , Borson S, Scanlan J, Brush M, Vitaliano P, Dokmak A: The mini-cog: a cognitive vital signs measure for dementia screening in multi-lingual elderly. Int J Geriatr Psychiatry 15: , Li YP, Propert KJ, Rosenbaum P: Balanced risk set matching. J Am Stat Assoc 96: , Chan KE, Lazarus JM, Thadhani R, Hakim RM: Anticoagulant and antiplatelet usage associates with mortality among hemodialysis patients. J Am Soc Nephrol 20: , Chan KE, Lazarus JM, Wingard RL, Hakim RM: Association between repeat hospitalization and early intervention in dialysis patients following hospital discharge. Kidney Int 76: , Chan KE, Lazarus JM, Thadhani R, Hakim RM: Warfarin use associates with increased risk for stroke in hemodialysis patients with atrial fibrillation. J Am Soc Nephrol August 27, 2009 [epub ahead of print] 15. Kidney Care Partners: Kidney Care partners launches health care campaign to improve survival rates of first-year dialysis patients. Available online at: com/files/final_peak_launch_release_06_08.doc. Accessed June 8, 2009