FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W1564157571> ?p ?o ?g. }

• W1564157571 endingPage "10" @default.
• W1564157571 startingPage "8" @default.
• W1564157571 abstract "Separating the effects of transfusion from the effects of the “need for transfusion” has bedevilled observational studies of blood transfusion. History shows that we are rarely clever enough to allow correctly for all the possible confounding factors. Numerous retrospective studies have suggested that transfusion increased the risk of recurrence after cancer surgery. Finally a prospective randomized study showed that there was no effect at all.1 It was the “need for transfusion” that caused the trouble, not the transfusion. Even randomized prospective studies are not without flaws. They are often highly focused in an attempt to reduce unwanted variability. The most quoted of all these, the TRICC trial, dealt with patients on intensive care and used a nonleukoreduced and plasma-rich red blood cell (RBC) product of indeterminate storage age.2 Whether the results can be safely extrapolated to the use of optimal additive leukoreduced fresh RBCs given after surgery for a fractured hip is arguable. It is therefore refreshing to go back to basics and to try to analyze the effects of anemia, or the need for transfusion, alone. By understanding the effects of anemia we may decide how important correction is and make rational judgments about use of transfusion with due allowance for the quality of the product we have available. Two articles in this edition of TRANSFUSION attempt to examine the effects of anemia, rather than blood transfusion, on outcome from hip surgery. It is unfortunate that in routine clinical practice (away from those who refuse blood), anemia and blood transfusion, a little like electron pairs in quantum mechanics, are inevitably entangled. Observation of one factor, hemoglobin (Hb), by a clinician determines the second factor, transfusion. Despite this connection, the authors manage to pick out some illuminating findings. The two studies examine opposite ends of the process. Mantilla and coworkers3 have analyzed 30-day mortality and/or myocardial infarction (MI) in patients having emergency or routine hip or knee replacement operations. They correlate this outcome with preoperative Hb. To do this they use a case-control study. Case-control studies are particularly valuable when the outcome is relatively uncommon and a prospective trial would be hopelessly unwieldy in size and cost. By selecting patients who died or had a confirmed MI over 20 years of surgery and matching these with controls they are able to study a critical outcome from a very large cohort of patients. They conclude that anemia is an association that is common both to a poor outcome and to underlying disease and that the underlying disease, not the preoperative anemia, is the driving cause of the poor outcome. Unfortunately the entanglement of anemia and transfusion does muddy the water. The more anemic a patient is preoperatively, the more likely he is to receive transfusion. More “cases” were anemic and, as would be expected, transfusion was more common in the “cases” compared to the controls (53% vs. 37%). The pessimistic view might then be that transfusion was the cause of the poor outcome, but the lack of an independent association between preoperative anemia with poor outcome and, in a secondary and post hoc analysis, between transfusion and outcome, make this conclusion most unlikely. A more optimistic view would be that transfusion negated the effect of anemia on the patient's outcome; if the patients had not been transfused, mortality and MI in the vulnerable early postoperative period would have been more common and anemia would then have been independently associated with poor outcome. This possibility is more difficult to refute but the authors reference a number of other studies from their institution showing no change in morbidity after surgery despite changing transfusion practice. In their view neither anemia nor transfusion was guilty of causing the poor outcome; that culprit was the underlying disease. This conclusion contrasts with that of some other retrospective studies, particularly in cardiac surgery where the blame for poor outcome is laid firmly at the door of transfusion, particularly of older blood.4, 5 It may be that any effect of anemia or transfusion is heavily dependent on the type of surgery or perhaps the degree of anemia. More likely it demonstrates the serious limitations of those retrospective studies taking transfusion as the prime variable to be studied. The report of the FOCUS trial presented at the American Heart Association in 2009 by Carson suggests that in patients with fractured hip repair, there was no difference in mortality or serious cardiac events between patients in a liberal target-driven transfusion group and those transfused symptomatically. These results if confirmed in full publication strongly suggest that Mantilla and colleagues are correct in that moderate anemia is not a direct cause of mortality or serious cardiac events in patients undergoing lower limb arthroplasty. We might then ask why transfuse any surgical patient above a reasonable threshold of say 8 g/dL? One answer is that marked postoperative anemia, even if not increasing cardiac and general mortality might impair rehabilitation, slow time to discharge and possibly lead to greater readmission rates, particularly in vulnerable patients with traumatic hip fracture. An additional article in this edition of TRANSFUSION examines a similar cohort of patients undergoing hip surgery, this time all elective, and examines quality of life (QoL) postoperatively.6 So-Osman and coworkers7 carried out a randomized trial of two different transfusion strategies at three hospitals to determine whether RBC usage was associated with a prolonged hospital stay. The results of this study turned out to be somewhat ambiguous because of prestudy differences in the transfusion policies of the three hospitals. Tagged on to this trial, and reported here, was the study of QoL and anemia. The question asked was did postoperative Hb correlate with QoL scores on Day 14 postoperatively. The answer was no. Although anemia and transfusion are again entangled, this is unlikely to be an issue a week or two after any blood transfusion. Does QoL over a brief postoperative period matter? What really matters is medical outcome in terms of death or debility and economic outcome in terms of time to discharge, time to living independently, and avoidance of readmission. QoL may, however, be a useful measure if it reflects speed of rehabilitation for which it is likely to be a good surrogate. There are relatively few studies of QoL or other functional scores in relation to anemia or transfusion and none that is directly comparable because of differences in patient groups and times of measurements of QoL or Hb. Conlon and colleagues8 in a much smaller study of 79 hip surgery patients noted a clear but small correlation between Hb on Day 8 postoperatively and change in QoL on Day 56. It has been previously shown that Hb levels are substantially recovered by Day 56 in patients such as these and it must be likely that Day 8 Hb in the study of Conlon and coworkers was a surrogate for some other factor affecting the change in QoL, which was determined 2 months postoperatively without a concurrent Hb.9 Halm and colleagues10 reported that patients having fractured hip repair, a frail and vulnerable group, showed a higher readmission rate (but not mortality) after discharge with a lower Hb though we have been unable to reproduce that finding.11 Lawrence and colleagues12 found that mobility at discharge as measured by straight line walking was significantly associated with Hb, although they admit some shortcomings in the quality of the data. Foss and colleagues13, 14 have also argued that rehabilitation is slowed by anemia and in an observational study found supporting evidence, but in a randomized prospective trial were unable to show a significant effect when comparing transfusion triggers of 8 and 9 g/dL. In all there is little good evidence for a clear or significant effect of postoperative Hb on QoL or speed of rehabilitation. In otherwise healthy patients, Hb recovers rapidly after blood loss and So-Osman and colleagues found a Day 14 mean Hb level of 11.4 g/L, already up 1 g from Day 4, and perhaps this and other factors such as increased 2,3-DPG ameliorating the effects of anemia explain the poor correlation with QoL.9 It seems from these two articles that there is no good evidence that either moderate preoperative or 14-day postoperative anemia substantially affects the measured outcomes in patients undergoing elective hip or knee replacement or emergency hip repair. We cannot extrapolate these findings without qualification to more severe degrees of anemia, to other groups of patients, or to transfusion and anemia in the early postoperative period, which are only studied indirectly here. Those planning randomized trials of blood transfusion might, however, note that they may struggle to get useful answers if there is, as suggested by these studies, little or no signal associated with moderate anemia. I have no conflicts of interest with regard to this document or the articles on which it comments." @default.
• W1564157571 created "2016-06-24" @default.
• W1564157571 creator A5042893907 @default.
• W1564157571 date "2011-01-01" @default.
• W1564157571 modified "2023-10-14" @default.
• W1564157571 title "Disentangling anemia and transfusion" @default.
• W1564157571 cites W127034668 @default.
• W1564157571 cites W1565676850 @default.
• W1564157571 cites W1576627395 @default.
• W1564157571 cites W1981950803 @default.
• W1564157571 cites W1986294961 @default.
• W1564157571 cites W2070506725 @default.
• W1564157571 cites W2096876696 @default.
• W1564157571 cites W2097517425 @default.
• W1564157571 cites W2097601835 @default.
• W1564157571 cites W2101005384 @default.
• W1564157571 cites W2123207734 @default.
• W1564157571 cites W2135405317 @default.
• W1564157571 cites W2164059065 @default.
• W1564157571 doi "https://doi.org/10.1111/j.1537-2995.2010.02976.x" @default.
• W1564157571 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/21219321" @default.
• W1564157571 hasPublicationYear "2011" @default.
• W1564157571 type Work @default.
• W1564157571 sameAs 1564157571 @default.
• W1564157571 citedByCount "13" @default.
• W1564157571 countsByYear W15641575712012 @default.
• W1564157571 countsByYear W15641575712013 @default.
• W1564157571 countsByYear W15641575712014 @default.
• W1564157571 countsByYear W15641575712015 @default.
• W1564157571 countsByYear W15641575712016 @default.
• W1564157571 countsByYear W15641575712019 @default.
• W1564157571 crossrefType "journal-article" @default.
• W1564157571 hasAuthorship W1564157571A5042893907 @default.
• W1564157571 hasBestOaLocation W15641575711 @default.
• W1564157571 hasConcept C126322002 @default.
• W1564157571 hasConcept C177713679 @default.
• W1564157571 hasConcept C187212893 @default.
• W1564157571 hasConcept C2778248108 @default.
• W1564157571 hasConcept C2780014101 @default.
• W1564157571 hasConcept C3019864176 @default.
• W1564157571 hasConcept C71924100 @default.
• W1564157571 hasConceptScore W1564157571C126322002 @default.
• W1564157571 hasConceptScore W1564157571C177713679 @default.
• W1564157571 hasConceptScore W1564157571C187212893 @default.
• W1564157571 hasConceptScore W1564157571C2778248108 @default.
• W1564157571 hasConceptScore W1564157571C2780014101 @default.
• W1564157571 hasConceptScore W1564157571C3019864176 @default.
• W1564157571 hasConceptScore W1564157571C71924100 @default.
• W1564157571 hasIssue "1" @default.
• W1564157571 hasLocation W15641575711 @default.
• W1564157571 hasLocation W15641575712 @default.
• W1564157571 hasOpenAccess W1564157571 @default.
• W1564157571 hasPrimaryLocation W15641575711 @default.
• W1564157571 hasRelatedWork W2012564050 @default.
• W1564157571 hasRelatedWork W2056034497 @default.
• W1564157571 hasRelatedWork W2103728012 @default.
• W1564157571 hasRelatedWork W2151621769 @default.
• W1564157571 hasRelatedWork W2268976010 @default.
• W1564157571 hasRelatedWork W2329126096 @default.
• W1564157571 hasRelatedWork W2970788022 @default.
• W1564157571 hasRelatedWork W3004855973 @default.
• W1564157571 hasRelatedWork W4237351857 @default.
• W1564157571 hasRelatedWork W4291005422 @default.
• W1564157571 hasVolume "51" @default.
• W1564157571 isParatext "false" @default.
• W1564157571 isRetracted "false" @default.
• W1564157571 magId "1564157571" @default.
• W1564157571 workType "article" @default.