The Prevalence of Anemia of Chronic Disease in Patients With Spinal Cord Injuries and Pressure Ulcers and the Impact of Erythropoietin Supplementation on Wound Healing: A Descriptive Pilot Study
Anemia of chronic disease (ACD) is thought to impair the responsiveness of erythroid progenitor cells, but research has shown treatment with recombinant human erythropoietin (rhuEPO) can improve patient hemoglobin levels and, subsequently, overall patient health status and quality of life. A prospective pilot study was designed to estimate the prevalence of ACD in outpatients with spinal cord injury (SCI) and chronic pressure ulcers (PUs) and examine the impact of rhuEPO on PU healing in this population. The charts of 49 SCI patients with PUs were reviewed; of those, 17 had anemia (hemoglobin <110 g/L). The prevalence of anemia in SCI patients with PUs was found to be approximately 35%.
From these 17 potential participants, 5 had improved hemoglobin levels during the screening period (rendering them ineligible), 1 withdrew due to illness, and 7 died, leaving 4 participants to complete the study. Four patients (2 men, 2 women, average age 57 ± 16.5 years) ultimately were enrolled. Wound area and depth and cytokines were measured before, during, and after 6 weeks of treatment with rhuEPO, with a 3-month follow-up. Laboratory tests measuring hemoglobin, C-reactive protein, and prealbumin were used to monitor nutritional status and treatment response. No statistically significant changes were observed with treatment. Wound surface area and depth had mean decreases of 1.35 cm² and 0.58 cm, respectively, immediately post-treatment. Participants’ elevated C-reactive protein levels (91.1–14.2 mg/L) decreased with rhuEPO treatment, but returned to baseline levels post-treatment (83.2–14.3 mg/L). Prealbumin levels were consistently low (range of 0.1–0.21 g/L). This research indicates rhuEPO treatment may improve some outcomes for ACD-SCI PU patients, but larger randomized controlled trials are required. The results of this study suggest the prevalence of ACD in the SCI outpatient population with PUs is at least 35%, and increased vigilance of patient nutrition is recommended.
Potential Conflicts of Interest: This study was funded by a grant from the Ontario Neurotrauma Foundation, Toronto, Ontario, Canada.
Anemia of chronic disease (ACD), also known as anemia of inflammatory response, often is seen in chronic illness. High circulating levels of inflammatory cytokines in the body affect the signaling for the storage and release of iron1; they also are thought to impair the responsiveness of erythroid progenitor cells.2 The development of ACD is complex, but a series of case studies3-5 has shown treatment with recombinant human erythropoietin (rhuEPO) can improve patient hemoglobin levels, leading to improvements in the overall health status and quality of life in these patients. Additionally, results of animal studies6,7 suggest supplementation with erythropoietin may improve wound healing outcomes in a variety of clinical situations; research in animal models8 also indicates an increase in growth factors such as vascular endothelial growth factor (VEGF) and a decrease of pro-inflammatory cytokines such as Interleukin-6 (IL-6) and tumor necrosis factor alpha (TNF-α). These are markers of inflammation in chronic wounds, and previous in vitro and animal studies8,9 have indicated supplementation with rhuEPO may modify the inflammatory marker levels.
An observational study of 30 patients by Scivoletto et al4 first associated low hemoglobin and decreased healability in patients with pressure ulcers (PUs) and spinal cord injury (SCI), and a previous case series5 of 4 patients conducted by the current authors found the use of rhuEPO for treatment of ACD in persons with SCI and chronic PUs led to encouraging results in wound healing5: following a 6-week course of rhuEPO treatment in 4 patients, mean hemoglobin increased from 88 g/L to 110 g/L; mean ulcer surface area decreased from 42.3 cm2 to 37.3 cm2 (a 12% decrease) despite extensive deroofing of 1 ulcer; and mean ulcer depth decreased from 2.3 cm to 1.2 cm (a 48% decrease). In addition, patients reported greater energy levels and were better able to participate in their rehabilitation programs. Post-study, an additional 9 inpatients and 3 outpatients were treated with rhuEPO with positive results: hemoglobin levels were raised above 100 g/L and chronic PUs closed in 11 of 12 patients (unpublished chart review data). It was determined these results merited further study using an outpatient population.
The purpose of this prospective, descriptive pilot study was to: 1) obtain an estimate of the prevalence ACD in SCI patients with chronic PUs, 2) investigate wound healing outcomes, and 3) investigate whether rhuEPO administration reduces the level of chronic inflammatory mediators IL-1, IL-6, and TNF-α and/or enhances the levels of growth factors such as VEGF in the wound environment as measured by their presence in the wound fluid of SCI patients with ACD and chronic PUs.
This study was conducted under the supervision and approval of the Health Science Ethics Review Board at Western University, London, Ontario, Canada.
Inclusion and exclusion criteria. Study participants were recruited from the Spinal Cord Injury and Chronic Wound Management Clinics in the outpatient department of a Rehabilitation Hospital. The medical records of individuals with nonhealing ulcers (present for >3 months) as a result of SCI were screened for ACD, defined as hemoglobin <110 g/L with normal red cell indices and normal or elevated ferritin. Persons with 1 or more nonhealing ulcers and low hemoglobin levels were contacted by a member of that patient’s direct medical care team and informed about the study. Potential participants had to be able to provide informed consent. Exclusion criteria included current use of rhuEPO or a medical condition that contraindicates the use of rhuEPO, including uncontrolled hypertension, a history of developing pure red cell aplasia (PRCA) following treatment with any erythropoiesis-regulating hormone, known hypersensitivity to mammalian cell-derived products such as human albumin, or inability for any reason to receive adequate antithrombotic treatment. Some versions of rhuEPO are formulated with human albumin, and cases of erythropoietin-induced hypertension have been reported,10 presenting an increased risk factor even though the type of rhuEPO used in this study did not contain any human albumin. PRCA complications as a result of rhuEPO have been known to occur. These complications are considered to be the result of using prefilled syringes with uncoated rubber stoppers that produced leachates and potentiated the immunogenicity of the product; modern reformulations of rhuEPO have correlated with a steady decrease in PRCA incidence.11 As rhuEPO stimulates red blood cell production, risk of clotting is potentially increased, necessitating treatment with an antithrombotic treatment.
Because this was a prospective, descriptive pilot case series, no calculation or expectation was placed on population size.
Study design. Patients recruited into the study underwent an initial baseline assessment of hematological status, wound characteristics, and level of inflammatory mediators. Prevalence data were estimated from the screening data. Following baseline assessment, a 2-week standardization period was initiated to optimize wound dressings, improve nutritional status, maximize pressure redistribution strategies, and stabilize comorbid medical conditions. All patients involved in this study were seen in the Regional Spinal Cord Injury Program and provided occupational therapy and physiotherapy assessments, which included a seating clinic. All recommendations made by the authors in this study were done in accordance with the Canadian Association of Wound Care Best Practice Recommendations, which are based on the Registered Nurses Association of Ontario Clinical Practice Guidelines.12
Each participant received a dietary assessment by the study personnel following the aforementioned guidelines. In addition to providing iron supplementation, determined based on laboratory results, the study staff consulted with the Regional Spinal Cord Injury Program dietitian who worked with them to maximize protein and calorie status consistent with participant wishes. Upon completion of the standardization period, blood values and wound status were assessed again to establish pre-rhuEPO treatment levels of ACD, inflammatory markers, and wound status. If hemoglobin rose above 110 g/L, the participant was considered a screen failure. Wound status was assessed using the Photographic Wound Assessment Tool13 (PWAT) and by measuring size and depth. These measurements were taken at 2 weeks and 5 weeks after commencement of rhuEPO treatments and at 3 months after rhuEPO treatments were completed. Tissue swabs were taken from the wound bed using a swab-collection technique described by Wyffels et al14 at each data point for analysis of VEGF and inflammatory mediator levels of IL-1, IL-6, and TNF-α. A minimum of 4 swabs was taken for each time point, covering all areas of the wound bed post-dressing removal and before wound bed irrigation and debridement.
Treatment. All patients received regular care (physiotherapy, occupational therapy, and seating assessments) in addition to the rhuEPO treatments. All ulcers were debrided and bandaged to usual treatment standards.12 Patients were treated by their regular nurses in a home care setting. These nurses provided regular wound care and dressing changes and administered the rhuEPO injections. Patients enrolled into treatment were given a course of rhuEPO injections of 75 IU/kg, 3 times per week, for 6 weeks. Study participants were required to visit the clinic 6 times: once for the initial screening, once for enrollment into the study, twice during the treatment phase, and twice post-treatment (1 week and 10 weeks post-treatment). Depending on the requirements of the patient, an iron supplement may have been prescribed. Because this study was designed to investigate the possible effects of rhuEPO on the healing of chronic wounds and these patients already were involved in a program to help support their regular health care needs, additional comorbid conditions were not investigated in this study.
Outcome measures. Outcomes were measured in 3 different domains: laboratory testing, changes in wound size and appearance, and protein analysis of wound fluid.
Laboratory testing. In order to determine anemia, complete blood count, serum ferritin, serum iron, total iron binding capacity, serum vitamin B12, and serum folate levels of circulating blood all were measured and observed. C-reactive protein (CRP), an acute phase reactant that is elevated in the presence of inflammation or infection, also was measured as an additional assessment of the possible change in the patient’s inflammatory status. Prealbumin, creatinine, and electrolytes also were assessed to monitor nutritional status.
Wound size and appearance. Changes in wound surface area were obtained using acetate tracings and calculated with a digitizing tablet (Visitrak™, Smith and Nephew Canada), which calculates wound area using an acetate tracing of the wound bed. Wound depth was determined by measuring the maximum depth a sterile cotton-tipped swab could be inserted into the wound and make contact with the deepest point. Wound bed appearance was assessed using the PWAT.14
Protein analysis. In order to monitor possible changes in cytokine and growth factors in the wound bed, IL-1, IL-6, TNF-α, and VEGF present in the wound fluid at the time of the assessments were measured using swabs taken each visit at the base and edges of the wound.
Data collection and analysis. Participants enrolled in the study were under observation for 20 weeks; data were collected at week 0 (baseline) and at weeks 2, 5, 8, and 20 throughout the study. All study data were anonymized and patients were given study numbers upon enrollment. Laboratory data and wound measurement data (including the PWAT) were collected and stored onsite in a secure location. PWAT measurements were taken onsite using previously established guidelines,13 and wound measurements were taken using the aforementioned measurement system. Analyses of the wound measurements, including averages and standard deviations, were performed using Microsoft Excel™. Laboratory samples were evaluated and assessed by the London Laboratory Services (London, Ontario). Wound fluid swabs for cytokine data were collected and stored at -80˚ C in buffered solution according to previously established guidelines.13
Cytokines were assessed in a multiplex protein; analyte assay was performed by an offsite facility for these specific targets. Concentrations of each protein were standardized using total sample protein concentration. Analysis was performed using the R statistical software package, version 3.0.2 (R Foundation for Statistical Computing, Vienna, Austria).
Comparisons of analyte concentrations are presented as relative quantities (ratios) for each time point (weeks 2, 5, 8, and 20) relative to week 0. Ratio calculations and graphs were developed using Microsoft Excel™.
Participants. Of the 49 patients initially screened, 30 had relatively normal hemoglobin levels, 17 had hemoglobin levels >110 g/L at the time of screening, and 2 did not have hemoglobin levels included in their chart data. Of the 17 patients that met the study criteria for ACD and initially considered as potential patients and followed-up, 3 (18%) improved by the screening visit, 7 (41%) remained anemic (and therefore eligible), 1 (5%) withdrew due to illness, and 6 (35%) died before they could be enrolled. From the 7 participants who were enrolled in the study, 2 had improved health during the screening period (28%) and another individual died (14%). This left 4 patients (2 men, 2 women, average age 57 ± 16.5 years) able to complete the full rhuEPO treatment. All (100%) of the patients involved in the rhuEPO treatment were diagnosed with ACD, but the prevalence of ACD within the population of SCI patients originally examined was 35%. Three of the four patients had Stage IV PUs: 1 was located on the left greater trochanter (which was 4.1 cm2 and 12 months old at day 0); 1 on the left ischium (0.9 cm2 and 9 months old); and 1 on the coccyx (6.9 cm2 and 25 months old). The fourth patient had a Stage III PU PI on the heel, which was 16.7 cm2 and 18 months old at day 0 of treatment.
Laboratory and wound healing results. At baseline, the laboratory results revealed all participants had elevated CRP levels as well as low prealbumin and absolute lymphocyte counts, indicating possible inflammation and nutritional deficiencies.15 After week 20, no overall sustained biological effects from the rhuEPO treatment were noted. The laboratory results and wound measurements and evaluation are summarized in Table 1.
Remarkable biological variation was noted among participants in the response to erythropoietin supplementation, from patient 1 who had almost no improvement in hemoglobin levels and a standard deviation of 1.5 g/L to patient 3 who had a sustained improvement in hemoglobin levels that was maintained to week 20, 3 months post-treatment. The participants’ elevated CRP levels (reference range <5 mg/L), ranging from 91.1–14.2 mg/L, reflected the variation was too great to determine a statistically significant trend, but in 50% of the participants, CRP levels tended to decrease with the rhuEPO treatment and re-elevate in the 12-week follow-up period. Participant prealbumin also was low throughout the study (0.1–0.26 g/L; reference range 0.18–0.45 g/L), most notably at week 0 (0.1–0.21 g/L). Absolute lymphocyte counts, a marker for micronutrient deficiency, were low, ranging from 0.9 x 109 cells/L – 2.1 x 109 cells/L (reference range >1.5 x 109 cells/L).
The growth factor and cytokine assessment from the wound fluid samples yielded similarly inconsistent results, although some trends were observed, indicating the presence of IL-6 decreased during rhuEPO treatment and returned to pretreatment levels at week 20 (12 weeks post-treatment completion). The relative changes in quantities of cytokines (compared to baseline) are provided in Figures 1 through 4. The precise measurement of cytokines and these kinds of proteins combined with the limited sample size of participants made it difficult to determine statistically significant results. The quantities of cytokine ranged from 6.6 pg/mg total protein to 6332.8 pg/mg, and no statistically significant differences were found for standardized analyte concentrations or total concentrations between time points.
Wound healing results were similarly variable. Wound improvement was noted from pretreatment to post-treatment measurements. The average wound surface area from baseline to the 8 week treatment period was 7.5 cm2 to 5.8 cm2, and the average change in wound surface area for each patient was 1.35 cm2 (SD 1.42 cm2). Mean change in wound depth from baseline to week 8 was 1.85 cm to 1.275 cm, a decrease of 0.575 cm (SD 0.55 cm). The week 20 post-treatment follow up visit showed the average surface area increased an average of 7 cm2 (SD 8.1 cm2), which indicated a final average decrease in wound area of only 0.15 cm2 (SD 3.01 cm2). The mean wound depth at week 20 (post-treatment) was 1.58 cm (SD 1.8 cm), an average change of 0.275 cm (SD 0.311) and a sustained improvement from baseline than the other measures.
A paired, 2-tailed t-test of the changes in area did not reveal any statistically significant differences between baseline and either the week 8 or week 20 time points (P = 0.197 and P = 0.936, respectively). A paired, 2-tailed t-test analysis of the change in depth between baseline and weeks 8 and 20 revealed similar results (P = 0.17 and P = 0.223, respectively). No adverse events occurred as a result of the rhuEPO supplementation, and none of the participants reported any ill effects either before or after rhuEPO treatment.
The primary objectives of this study were to obtain an estimate of the prevalence of ACD and to investigate wound healing outcomes in SCI patients with chronic PUs. Among the study population, 35% had hemoglobin levels <110 g/L; of these, 1 died before he/she could be approached for the study and 14% were diagnosed with ACD, with consistently low hemoglobin levels.
Before potential participants could be approached for study participation, 6 died. The causes of death were not investigated in this study because these participants had agreed to be screened, but they were not enrolled and therefore had not provided consent for detailed medical information to be investigated. One potential participant declined to participate due to illness, which left 10 potential participants. Of the 10 potential participants, 3 became ineligible due to increased hemoglobin levels and 6 consented. Of these 6, 2 became ineligible during the run-in period due to increased hemoglobin levels. Thus, 4 participants completed the study.
This study indicates the prevalence of anemia in SCI patients with PUs is at least 35% and could be greater due to the high variability of hemoglobin levels. The authors could find no previously published prevalence information for ACD in SCI patients with PUs with which they could compare these findings.
An unfortunate observation in this pilot study was the high mortality rate in the population of interest. Of the 49 pre-screened patients, 7 died and 2 withdrew due to illness. The hemoglobin data are not consistent for the subset of patients that either died or withdrew, so they also may have had ACD but died or withdrew before it was confirmed.
All SCI patients approached and followed for this study had a PU at the time of pre-screening. It has been previously documented that PUs in SCI patients increase the risk for metabolic changes, particularly for anemia.4 Although this is a small-scale examination of persons with SCI and PUs, it is notable that 35% of patients with consistently low hemoglobin died during the prescreening or screening process.
The secondary objectives were to investigate whether rhuEPO administration reduces the level of chronic inflammatory mediators and/or enhances the levels of growth factors in the wound environment. The data gathered from this pilot study were insufficient to draw any conclusions about these cytokines and growth factors. Unfortunately, the analysis of wound fluid found no statistically significant differences between any of the markers, and individual variability in analyte concentrations in combination with the relatively low number of individuals in the data set prevented detection. No statistically significant findings could be determined.
From both the population and laboratory data gathered from this study, it is clear concomitant factors affect the ACD such that simple dosage of rhuEPO may not be enough to improve patient prognosis. A previous case series5 involving hospitalized inpatients with PUs demonstrated consistent improvement in hemoglobin levels. The mean hemoglobin for 4 patients increased from 88 (±7.4) g/L to 110 (±3.7) g/L. The primary consideration for inconsistency of results in this pilot study (given there were 4 patients in each observation) is the comparison is between an inpatient group that received consistent nutritional and medical support and a more realistic outpatient population, where factors such as nutrition and activity were not as closely regulated and lifestyle choices had a greater impact. As the results demonstrate, minimal gains were made with regard to wound healing, and 2 of the 3 patients had a minimal response to rhuEPO treatment in terms of their ACD.
The lack of significant improvements does not necessarily suggest rhuEPO is inconsequential to wound healing: 1 out of the 3 patients had significant improvement in hemoglobin levels (96 to 124 g/L), but only 2 improved (2–5 g/L). As previously mentioned, the lack of response to the rhuEPO in general could be a reflection of the general poor health of the patients and the impact of environmental factors, such as nutrition.
The low lymphocyte counts and prealbumin levels could indicate the participants were not meeting nutritional benchmarks sufficiently before entering the study.15 The participants in this study were asked questions about their health and nutritional status at each visit, but diet (eg, calorie counts or intake records) was not formally monitored by the study team. Previous research5 by the authors indicates when patients are given rhuEPO in hospital as part of a care regimen for their PUs, wound outcomes and general health status improve.
In order to determine the possible effects of treatment with rhuEPO on PU outcomes, a much larger participant population would be required. Ideally, a randomized controlled trial could provide information regarding the effects of rhuEPO treatment; however, given the authors’ experience with this specific patient population, this would either require several research sites to enlarge the catchment area significantly or change the upper limit on hemoglobin values to include SCI patients without ACD.
The main limiting factor in this study is the sample size, partially the result of the fragile nature of the health status of potential participants. Other clinical studies4,5-9 have demonstrated links between PUs, anemia, and SCI, but this study lacked the participant numbers to support the hypothesis rhuEPO may modulate growth factors and inflammatory cytokine levels. A more indepth analysis of participant comorbidities also might have provided further information on possible underlying factors related to illness and response to treatment, although a larger patient population would be needed to properly analyze those influences.
A prospective, descriptive pilot study was conducted to determine the prevalence of ACD in SCI patients with PUs and to examine the effects of rhuEPO on the wound bed and healability of ACD patients with SCI and PUs. This case series demonstrated the overall health status of ACD-SCI patients is suboptimal, including poor nutritional status and increased chronic inflammation. Although data were insufficient to provide conclusive evidence with regards to rhuEPO supplementation and chronic wound outcomes, this study has successfully provided an overview of the prevalence of ACD in a patient population with SCI and PUs of approximately 35% and supports the importance of close observation of this group. This study also clearly indicated the management of PUs in SCI patients is a complex process. Comparing this study to the previous findings of the authors, it appears closely monitored inpatients have better wound healing outcomes. Although the participant population was too small to determine the possible impact of rhuEPO on ACD-SCI patients with PUs, it is possible to consider, as previously found in the literature, supplementation with rhuEPO could have a positive impact on the blood and serum metabolism of patients and therefore their healability, but this would require a larger population of patients to draw from, with better monitored nutrition. n
The authors gratefully acknowledge the financial support of the Ontario Neurotrauma Foundation, as well as the contributions of Drs. Jennifer Wyffels and Laura Edsberg at Daemen College, Amherst NY for their technical support and expertise in the collection and analysis of the wound fluid; and of Dr. Patrick Potter, Dr. Keith C. Hayes, and Chris Fraser for their expertise and guidance throughout this project. They also thank Diny Warren and Angela Woolner for assistance with patient recruitment.
At the time of manuscript submission, Ms. Vair was a research associate and clinical study coordinator, Lawson Health Research Institute, Aging, Rehabilitation and Geriatric Care Research Centre, London, Ontario, Canada. Dr. Keast is a Family Physician, Wound Care Specialist, and Associate Scientist, Lawson Health Research Institute; and Medical Director, Outpatient Wound Management Clinic, Parkwood Institute, London, Ontario. Ms. LeMesurier is a Nurse Clinician for Wound and Skin Care at Parkwood Hospital, St. Joseph’s Healthcare, London, Ontario. Please address correspondence to: David Keast, MD, Aging, Rehabilitation and Geriatric Care Research Centre, St. Joseph’s Parkwood Institute, 550 Wellington Road London, Ontario N6C 0A7 Canada; email: David.Keast@sjhc.london.on.ca .
1. Iron Disorders Institute. Iron Disorders Institute. Anemia of Chronic Disease. Available at: www.irondisorders.org/anemia-of-chronic-disease. Accessed November 2, 2011.
2. Spivak JL. Iron and anemia of chronic disease. Oncology (Williston Park). 2002;16(9 suppl 10):25–33.
3. Arndt U, Kaltwasser JP, Gottschalk R, Hoelzer D, Moller B. Correction of iron-deficient erythropoiesis in the treatment of anemia of chronic disease with recombinant human erythropoietin. Ann Hematol. 2005;84(3):159–166.
4. Scivoletto G, Fuoco U, Morganti B, Cosentino E, Molinari M. Pressure sores and blood and serum dysmetabolism in spinal cord injury patients. Spinal Cord. 2004;42(8):473–476.
5. Keast DH, Fraser C. Treatment of chronic skin ulcers in individuals with anemia of chronic disease using recombinant human erythropoietin (EPO): a review of four cases. Ostomy Wound Manage. 2004;50(10):64–70.
6. Hamed S, Ullmann Y, Masoud M, Hellou E, Khamaysi Z, Teot L. Topical erythropoietin promotes wound repair in diabetic rats. J Invest Dermatol. 2010;130(1):287–294.
7. Buemi M, Galeano M, Sturiale A, Lentile R, Crisafulli C, Parisi A et al. Recombinant human erythropoietin stimulates angiogenesis and healing of ischemic skin wounds. Shock. 2004;22(2):169–173.
8. Hamed S, Bennett CL, Demiot C, Ullmann Y, Teot L, Desmouliere A. Erythropoietin, a novel repurposed drug: an innovative treatment for wound healing in patients with diabetes mellitus. Wound Repair Regen. 2014;22(1):23–33.
9. Bader A, Lorenz K, Richter A, Scheffler K, Kern L, Ebert S, et al. Interactive role of trauma cytokines and erythropoietin and their therapeutic potential for acute and chronic wounds. Rejuvenation Res. 2011;14(1):57–66.
10. Vaziri ND. Mechanism of erythropoietin-induced hypertension. Am J Kidney Dis. 1999;33(5):821–828.
11. Keast DH. Unexpected benefits of a formulation: case study with erythropoietin. In: Gad SC, ed. Handbook of Pharmaceutical Technology, Hoboken, NJ: John Wiley and Sons;2007:463–468.
12. Keast DH, Parslow N, Houghton PE, Norton L, Fraser C. Best practice recommendations for the prevention and treatment of pressure ulcers: update 2006. Wound Care Canada. 2008;4(1):447–462.
13. Houghton PE, Kincaid CB, Campbell KE, Woodbury MG, Keast DH. Photographic assessment of the appearance of chronic pressure and leg ulcers. Ostomy Wound Manage. 2000;46(4): 20–30.
14. Wyffels JT, Fries KM, Randall JS, Ha DS, Lodwig CA, Brogan MS, et al. Analysis of pressure ulcer wound fluid using two-dimensional electrophoresis. Int Wound J. 2010;7(4):236–248.
15. Beck FK, Rosenthal TC. Prealbumin: a marker for nutritional evaluation. Am Fam Physician. 2002;65(8):1575–1578.