Association between Post-Transplant Anemia and Mortality is Strengthened

Anemia, defined by the American Society of Transplantation as a hemoglobin concentration under 130 g/L in males and 120 g/L in females, is associated with an increased risk of mortality in affected patient populations, although no such link has been demonstrated in kidney-transplant patients. However, a study at the Department of Transplantation and Surgery at the Semmelweis University, Budapest is the first study to show a significant and independent association between post-transplant anemia (PTA) and mortality in this patient group. In the study, 938 kidney-transplant patients (18 years of age or older) were regularly followed at a single outpatient transplant center for a 4-year period. Data on mortality, graft failure (return to dialysis), or a combination of both were recorded. Patients were initially given a baseline assessment that included demographic data and medical history details such as age, gender, etiology of CKD, the presence or absence of diabetes, hypertension, and other comorbidities.

After the 4-year period, 118 patients had died and 79 patients returned to dialysis. A univariate analysis showed that anemic patients showed significantly higher mortality rates than nonanemic patients (18% vs. 10%; p<.001), as well as return to dialysis (17% vs. 6%; p<.001) and combined outcome (32% vs. 15%; p<.001). Furthermore, a multivariate analysis, including covariables such as gender, age, serum albumin, pretransplant time on dialysis, and transplantation "vintage" revealed similar trends, indicating that anemia is significantly and independently associated with mortality, return to dialysis, and combined outcome. When adjusting for these covariables, the analyses revealed that patients with anemia had a 1.69 times higher chance of mortality and a 2.46 times higher chance of graft failure within 4 years. Along with a large population size, utilizing covariable analysis has given this study a great advantage over previous studies.

These results strengthen the case for an association between anemia and mortality. There are multiple conditions seen in transplant patients, including hyperkinetic circulation, congestive-heart failure, thicker left ventricle walls, and malnutrition-inflammation complex syndrome. Anemia leads to many of these cardiovascular conditions, which are strong, independent predictors of mortality in transplant patients. Future randomized controlled interventional studies will be needed to determine if the relationship between anemia and outcomes in this population is causal and whether treating PTA will reduce mortality and graft failure.

Please reference the source article:
Anemia Is Associated with Mortality in Kidney-Transplanted Patients - A Prospective Cohort Study. Molnar, Czira et al. Am J Transplant. 2007; (4): 818-824.

NAAC Expert Commentary:
Anemia is associated with adverse outcomes (mortality, morbidity) in a variety of patient populations, including those with chronic kidney disease and those on dialysis. Kidney transplant patients commonly have mild or more severe chronic kidney disease and would thus be anticipated to have anemia on this basis. In addition, ongoing chronic inflammation and the effects of immunosuppressive medications also contribute to the significant prevalence of anemia in this population. This paper extends previous research by documenting the high prevalence of anemia in patients following kidney transplantation and the association of anemia with undesirable outcomes.

Like other such studies, however, one must be cautious in the interpretation of the results. In the non-transplant kidney patient the results of such observational studies have not been confirmed in randomized clinical trials. It appears that anemia may be one indicator of poor health and thus portend a poor outcome when present, rather than being the cause of the poor outcome. The authors are appropriately cautious in their conclusions and their call for additional prospective controlled trials to determine if anemia and its correction are primarily important in patient and graft survival following kidney transplantation, and to determine the most beneficial target hemoglobin in this population.

Preoperative Anemia is Associated with Increased Risk of Postoperative Mortality and Cardiac Events

While healthy adults show strong physiological tolerance towards preoperative anemia and polycythemia in the context of postoperative mortality rates, little is known about how these conditions affect elderly populations. A recent study of elderly male veterans undergoing major noncardiac surgery revealed that even mild degrees of preoperative anemia or polycythemia were associated with significant increased risk of 30-day postoperative mortality and cardiac events (Q-wave infarction or cardiac arrest), independent of the type of surgery.

The study included 310,311 records of veterans aged 65 years or older who underwent major noncardiac surgery between 1997 and 2004, and grouped the subjects into white and non-white categories to account for the possible influence of race or ethnicity. Coronary disease-related variables such as hospitalization for angina, myocardial infarction, and coronary bypass graft surgery prior to an operation were also collected from the National Surgical Quality Improvement Program (NSQIP), as well as the Veterans Health Administration hospital records.

Guidelines form the World Health Organization and Harrison's Principles of Internal Medicine were used to define anemia (hematocrit <39%), polycythemia (hematocrit >54%), and normal hematocrit levels (hematocrit 39.0%-53.9%). In addition, the study defined preoperative hematocrit value as the last hematocrit value measure prior to the surgery (99% of hematocrit values were taken 3 months prior to the surgery). The study found a rise (1.6%) in 30-day postoperative mortality and cardiac events for every percentage-point increase or decrease from normal hematocrit range. The further subjects deviated from the normal range, the greater the risk became. For example, subjects with a hematocrit level of 18% or less showed a 35.4% mortality rate, while subjects in the 45.0% to 47.9% hematocrit range showed a 1.5% mortality rate.

Future studies will be needed to determine the effect of treatment on morbidity and mortality and to account for several study limitations, including a small sample size in polycythemia patients and the inability to determine the causal relationship between varying hematocrit levels and postoperative adverse events.

Please reference the source article:
Preoperative Hematocrit Levels and Postoperative Outcomes in Older Patients Undergoing Noncardiac Surgery Wu, Schifftner et al. JAMA, June 13, 2007. 297. 2481-2488.

NAAC Expert Commentary:
The authors have given us evidence about the relationship that most of us have believed intuitively for some time: that there is a strong association between preoperative anemia and increased risk of death and/or cardiac event in (elderly) surgical patients. Interestingly, polycythemia also increases risk. However, the study raises some questions about what wasn't analyzed. NSQIP data tracks only preoperative transfusion of greater than 4 units of RBCs, a measure that will miss many transfusions, thus underestimating the impact of blood transfusion on mortality. In fact, the majority of patients get less than 4 units of blood and mortality in the low hematocrit group could be a function of more exposure to blood transfusions. Hematocrit is as much a measure of volume status as it is a measure of red cell mass, especially in the elderly who are often volume depleted. Hematocrit often drops in these patients when they receive non-sanguineous intravenous fluids, so preoperative measures underestimate the level of anemia.

Perhaps most worrisome shortcoming in this study is the long list of "also rans" as potential contributors to poor outcome. This includes "Do Not Resuscitate" orders (OR=3.1), ASA class >4 (OR=3.62), MI within 6 mos (OR=2.84), ascites (OR=2.24), and disseminated cancer (OR=3.62). Variables with odds ratios (OR) such as these often show up as independent variables in regression models, but we don't have enough information to know if this is the case here. Nonetheless, this report does have the strengths of large numbers and a proven data collection process. My hope is that physicians will interpret this report as compelling evidence that correctable anemia should be a contraindication to elective surgery in the elderly. The appropriate therapy still needs to be assessed. In light of other data from large observational studies, allogeneic blood as treatment for anemia may also worsen outcome.

Restrictive Hemoglobin Strategy May Safely Reduce Transfusion Requirements

A randomized controlled study enrolling 637 stable, critically ill children has shown that a hemoglobin threshold of 7 g/dL (restrictive-strategy) with reduced leukocytes can reduce transfusion requirements without negative effects such as death or multiple-organ dysfunction syndrome (MODS). The study, carried out in 19 tertiary-care ICUs in 4 countries, examined children aged 3 days to 14 years who, within the first 7 days after admission, had at least one hemoglobin concentration of 9.5 g/dL (liberal-strategy). The study's hypothesis was formulated from previous adult transfusion studies where a restrictive-strategy was shown to be superior to a liberal-strategy.

Patients were randomly distributed into three age groups (<29 days, 29 to 364 days, and >364 days), and baseline assessments of patients including recording of hemoglobin concentrations, the number of red-cell transfusions, types of medications, the use of mechanical ventilation and dialysis, and surgical interventions were recorded. Research staff was kept blind to the randomization strategy. Primary outcome was measured as the number of patients who died during 28 days after randomization, experienced new or progressive MODS, or experienced adverse reactions.

Interestingly, each strategy reported the same number of deaths (14) as well as similar numbers of patients experiencing new or progressive MODS (0.4% absolute reduction in risk; 95% confidence interval -4.6 to 5.5 with restrictive-strategy). Furthermore, there were no significant differences in adverse reactions between the restrictive-strategy and liberal-strategy groups (221 and 203 reactions, respectively). Most importantly, the restrictive-strategy greatly reduced the number of patients (96%) having any transfusion exposure, as well as decreasing the number of transfusions (44%) administered.

The study showed that a restrictive-strategy may be safe for stable, critically ill children, yet outcomes from trials in critically ill adults differ.

Please reference the source article:
Transfusion Strategies for Patients in Pediatric Intensive Care Units Lacroix, Hebert et al. NEJM. April 19, 2007. 356;16. 1609-1619

NAAC Expert Commentary:
In a formidable collaborative effort, the TRIPICU(1) trial was completed and published in Transfusion (2006 Nov). This was a prospective randomized trial of a restrictive vs. liberal threshold of transfusion (hemoglobin of 7.0 g/dL vs. 9.5 g/dL respectively) in pediatric patients admitted to intensive care units with stays of greater than 3 days. The age group spanned neonates to children of age 14. The study was conducted from Nov 2001 to Aug 2005, screening 5,399 pediatric patients of whom 81% met exclusion criteria, leaving 637 patients for final randomization. All red cell transfusions were pre-storage leukodepleted cells. Primary end points were 28-day mortality and 2 or more organ dysfunction. Results of the study showed a 96% reduction of exposure to allogeneic blood and a 44% reduction in the amount given to the restrictive group. These finding led the investigators to conclude that a restrictive transfusion practice was safe in the stable intensive care pediatric population but was not superior to the liberal arm. The difference in survival outcome compared to the adult study, TRICC(2), was thought to be the prestorage leukodepleted blood in this trial.

The study design included a predetermined 'suspension' clause where subjects were taken off protocol for a variety of predetermined condition such as surgery and active bleeding. Patients with high acuity illnesses were also excluded from the trial making the results narrowly applicable. As with the adult population, few units per patients are given to a large cumulative number of patients who fit these restrictive criteria. Because of the small number of neonates, these results may not reflect behavior in this population that has a different cardiovascular physiology than their older counterparts.

References in the commentary:
(1) Acute transfusion reactions in the pediatric intensive care unit. Gauvin F, Lacroix J, Robillard P, Lapointe H, Hume H. Transfusion. 2006 Nov; 46(11): 1899-908.

(2) A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. Hébert PC, Wells G, Blajchman MA, Marshall J, Martin C, Paqliarello G, Tweeddale M, Schweitzer I, Yetisir E. N Engl J Med. 1999 Feb 11; 340(6): 409-17.

Study Shows For-Profit Facilities Give Higher Doses of Epoetin

A study of dialysis facility ownership has revealed an association between the type of facility (profit, chain, and affiliation status) and epoetin dosing in anemic patients. The study showed that, regardless of anemia status, for-profit facilities consistently administered higher doses of epoetin to patients compared to non-profit facilities. In November and December of 2004, data from 159,522 adult Medicare-eligible end-stage renal disease (ESRD) patients receiving in-center hemodialysis were collected from the US Renal Data System (USRDS). The USRDS provides comprehensive data on demographic and clinical data for ESRD patients, as well as information on facilities.

The study's primary outcome was the average weekly dose of epoetin administered to each patient in December of 2004. Statistical regression models then estimated the mean epoetin dose and dose adjustment according to profit status or chain affiliation. The study found that the average hematocrit levels for patients treated in for-profit facilities was 36.2% and 35.6% in non-profit facilities (The Centers for Medicare and Medical Services (CMS) recommends a target range of 33% to 36%). Patients treated at non-profit facilities within this target range received reduced epoetin levels by an average of 634 units per week (U/wk), while patients treated at for-profit facilities received a 1,214 U/wk average increase.

The differences in epoetin dosing between for-profit and non-profit facilities are not explained by patient characteristics or by factors related to dialysis responsiveness. The study, however, does offer two explanations for the facilities ownership effect on epoetin dosing. First, CMS strives for a goal of 70% of treated patients to have a hematocrit level above 33%. As long as average hematocrit levels do not exceed 37.5%, facilities meeting this goal would be eligible for epoetin reimbursement through CMS, which provides an incentive to use large doses even though improved patient care did not necessarily result. Second, facilities using large amounts of epoetin are eligible for volume discounts - a further incentive to administer large doses to patients.

Some studies have failed to detect a cardiovascular or survival benefit in patients receiving epoetin dosing above the 36% hematocrit threshold. Moreover, adverse consequences may be associated with epoetin dosing beyond this threshold. However, these studies are controversial and future efforts must determine if existing guidelines need reevaluation to optimize epoetin-dosing practices.

Please reference the source article:
Dialysis Facility Ownership and Epoetin Dosing in Patients Receiving Hemodialysis. Thamer, Zhang et al. JAMA. April 18, 2007. 297. 1667-1674.

NAAC Expert Commentary:
The authors note that since 1991, reimbursement for epoetin has been based on the administered dose rather than a fixed or capitated fee. When payment by Medicare, was based on a capitated fee, a previous study(1) had concluded that for-profit dialysis centers administered smaller does of epoetin compared to non-profit centers, thus increasing their profit margins.

This recent analysis of epoetin use found that for-profit facilities used 20,838 U/wk on average, which was 3306 U/wk higher than the non-profit facilities. In patients with a hematocrit level <30%, the disparity in epoetin use was even greater between the two types of facilities. Most notably, dosing for patients within the hematocrit therapeutic target range of 33-36% decreased by an average of 634 U/wk at non-profit facilities, while increasing at for-profit facilities by 1,214 U/wk. In fact the most prominent differences in dosing were observed in the largest for-profit dialysis chains.

In for-profit dialysis centers, 23% of patients had a monthly hematocrit of 39% or higher (a statistic the non-profit facilities did not provide). Although facilities exceeded the upper hematocrit target for many patients, CMS did not suspend payment until 3-month average hematocrit levels rose above 37.5%. Thus, despite differences in the use of epoetin as described in this study, it appears that most facilities, for-profit and non-profit, met then applicable CMS guidelines.

Thus, even though dialysis facilities generally adhere to dosing requirements to ensure repayment, some have exhibited treatment practices which regularly bypass target hematocrit levels in the name of profit. This study has successfully demonstrated how difficult synchronizing payment and dosage has become and affirms guideline revisions must create appropriate reimbursement schedules to ensure the best clinical outcomes.

References in the commentary:
(1) The relationship of provider organizational status and erythropoietin dosing in end-stage renal disease patients. De Lissovoy G, Powe NR, Griffiths RI, et al. Med Care. 1994;32:130-140.