Portal de Periódicos da CAPES

British Committee for Standards in Haematology Guidelines on the Identification and Management of Pre-Operative Anaemia

2015; Wiley; Volume: 171; Issue: 3 Linguagem: Inglês

10.1111/bjh.13623

1365-2141

Alwyn Kotzé, Andrea Harris, C E Baker, Tariq Iqbal, N. G. Lavies, Toby Richards, Kate Ryan, Craig J. Taylor, Dafydd Thomas,

Cardiac, Anesthesia and Surgical Outcomes

Anaemia is most often defined in terms of the criteria established by the World Health Organization (WHO) in 1968 (WHO 2011), namely haemoglobin (Hb) concentration of <130 g/l for men and 100 g/l) increased relative risk by over 30%, with there also being a relationship between anaemia severity and outcome, i.e. a 'severity-response curve'. Anaemia also strongly predicted the need for transfusion, which is again predictive of poor outcome. Ferraris and colleagues (Ferraris et al, 2012) analysed data and conducted propensity-matched comparisons between 15 186 patients who received one unit of donated red cells and 893 205 patients who received no blood. Transfused patients suffered more morbidity and mortality. Patients transfused >1 unit had further increased rates of morbidity and mortality, in a dose-dependent fashion. Other authors (Pedersen et al, 2009; Glance et al, 2011) made similar conclusions. Detection of pre-operative anaemia thus helps to identify patients at risk of poor post-operative outcomes, including mortality. This is consistent with the surgical risk prediction literature. Pre-operative Hb influences risk in well-validated risk prediction models (Copeland et al, 1991; Whiteley et al, 1996; Prytherch et al, 2003; Richards et al, 2010; Smith & Tekkis, 2015). Furthermore, both iron status and Hb concentration are closely related to functional capacity (Anker et al, 2009), which in turn influences peri-operative risk (Hennis et al, 2011). Successive reports from the Royal College of Surgeons and Department of Health 2011) and National Confidential enquiry into Patient Outcome and Death (NCEPOD) (2011) recommend that the care and consent of surgical patients explicitly consider quantitative risk estimates made using predictive scores, e.g. decisions on critical care. The effect of anaemia on mortality is thus pertinent in routine practice and will impact on Trusts' critical care resource. Whilst anaemia is predictive of surgical risk, we could find no randomized evidence that correction of anaemia alters that risk. The exception is transfusion, where randomized evidence indicates that anaemia correction decreases transfusion requirements (Australian National Blood Authority 2011a). Given that allogeneic blood is, by definition, a limited resource, there is public health benefit to be gained from limiting its use where alternatives exist. Because previously unsuspected anaemia detected during surgical work-up may, of itself, be significant, anaemia screening also carries a health benefit beyond the planned surgical episode. The argument for screening and correction is thus strong in principle, but individualized decisions must still be made that take into account the magnitude of the proposed surgery as well as the patient's history and preferences. If patients are only screened close to the planned operation date, there will be insufficient time to correct anaemia (Blest et al, 2007). Missing the above opportunities for timely diagnosis should be viewed as a failure of care processes. An agreed referral pathway that incorporates anaemia screening as part of the referral for surgery from General Practice is ideal, as this maximizes the time available for investigation and treatment. Anaemia screening, by laboratory or bedside test, is easily carried out and places minimal burden on patients. Where surgery is a likely outcome from referral, such screening is strongly advocated. In an observational study, an approach including early screening and collaboration between primary and secondary care delivered outcome improvements for patients undergoing lower limb arthroplasty (Kotze et al, 2012), as well as significant cost savings (Spahn et al, 2012). Many surgical procedures may safely be scheduled to allow diagnosis and treatment of anaemia, and reduce risk. To do so constitutes best practice and provider organizations and commissioners may agree that referral-to-treatment time targets take account of this ('stopping the clock'). In procedures where delay presents a risk or burden to the patient, an individualized decision should be made that balances the risk of anaemia and/or transfusion against that of postponement. However, in non-urgent surgery this should only rarely be necessary if surgical pathways are properly set up. The magnitude of risk increase is proportional to anaemia severity (Musallam et al, 2011). Speed of response to treatment is furthermore greater in more severe anaemia, particularly in iron deficiency. Using whatever time is available before surgery to treat anaemia thus makes logical sense when surgery should be expedited. Broadly speaking, in the surgical context anaemia may be defined as that caused by disturbances of iron metabolism (and hence potentially correctable with iron alone) and anaemia of other causes (which may require other treatments). Erythropoiesis is iron-restricted if the body's absolute iron stores are insufficient (Iron Deficiency Anaemia, IDA) or if insufficient iron is available at bone marrow level in the presence of iron in the reticulo-endothelial system (Functional Iron Deficiency, FID) (Goddard et al, 2011; Thomas et al, 2013). Iron depletion may also occur without anaemia. Regenerating 10 g/l of lost blood takes approximately 165 g of stored iron in a 70 kg adult. If the serum ferritin is <100 μg/l, loss of 30 g/l Hb (1200 ml blood for a 70 kg adult) will deplete total body iron stores and precipitate IDA (Australian National Blood Authority 2011a). The serum ferritin concentration is the most powerful test to elucidate iron status in the absence of inflammation (Goddard et al, 2011). In anaemia, serum Ferritin <30 μg/l is a sensitive marker of iron deficiency (Mast et al, 1998), with <15 μg/l being pathognomonic of IDA (Goddard et al, 2011). The British Society for Gastroenterology recommends specialist referral for IDA, except in pre-menopausal women, because the prevalence of cancer in unexplained IDA approaches 15% (Goddard et al, 2011). Given that pre-operative workup often includes a battery of tests, hypoferritinaemia without anaemia may be detected. Hypoferritinaemia (<15 μg/l) without anaemia is not sinister in young women, but is associated with a cancer prevalence of 0·9% in men and postmenopausal women and referral is indicated (Goddard et al, 2011). Specialist referral may also be indicated according to the severity of unexplained anaemia (e.g. men with Hb <120 g/l and women with Hb <100 g/l, or according to any locally agreed criteria) (UK Renal Association 2010, Goddard et al, 2011). Ferritin is an acute phase reactant and may be elevated by inflammation or concurrent disease (Anker et al, 2009). Anaemia with ferritin levels below 50 μg/l or 100 μg/l is still strongly suggestive of iron deficiency in the presence of inflammation (Australian National Blood Authority 2011a, Goddard et al, 2011). In such circumstances, exclusion of iron deficiency may require the use of further tests (e.g. serum transferrin saturation <20%) or a therapeutic trial of parenteral iron (Australian National Blood Authority 2011a). Anaemia not related to iron disturbance may be due to other nutritional deficiencies (Vitamin B12 and folate), secondary to renal failure, or of other causes. Deficiency in vitamin B12 and folate is easily and cheaply tested for, and is recommended (UK Renal Association 2010). If other common causes are excluded and the estimated Glomerular Filtration Rate (eGFR) is <30 ml/min/1·73 m2 or 20 g/l below lower limit of normal or those with symptoms or splenomegaly should be referred. Carriers do not need haematology follow-up but their General Practitioner (GP) should be informed because the information may be important for genetic counselling. Women who have had pregnancies in recent years or partners of women with significant carrier states may have been screened as part of the national antenatal screening programme and may be aware of their haemoglobinopathy results. It is also useful to refer back to historical results, for haemoglobin, MCV and MCH, if available, as these remain relatively constant throughout adult life for a given individual; any significant deviation indicates an additional cause for anaemia. The Australian National Blood Authority and Network for Advancement of Transfusion Alternatives (NATA) review groups each published an investigation algorithm (Australian National Blood Authority 2011a, Goodnough et al, 2011). The UK context is different, particularly regarding referral-to-treatment time targets and the role of GPs in commissioning care. It is thus important that patients' GPs be involved from the outset in the management of anaemia discovered during operative work-up. GPs should include information about known anaemia in the surgical referral, so that patients are not inappropriately postponed or investigated. It may be practicable to submit anaemic patients to a standardized battery of tests (including repeat Hb, serum ferritin, vitamin B12 and Folate), rather than engage in sequential testing. The above diagnostic considerations and initial test results may be integrated into an investigations and referral algorithm. This writing group considers that such algorithms should be locally designed, to be implementable without disrupting surgical pathways. It should involve both primary and secondary care, and funding arrangements should be clear. The potential for gain is great, with reductions in spend on transfusion, potential reductions in length of stay, and a potential decreased burden of follow-up and treatment in primary care after discharge. Both primary and secondary care systems stand to gain from systematic pre-operative anaemia management. Example algorithms are given in Appendix S2. These are shown with the intent of supporting organizations in creating their own systems, rather than as recommendations per se. We therefore intend that clinicians use this guideline and the results of local pathway analysis to design management algorithms for the different procedures (or procedure categories) that they perform. Integrating anaemia management with pre-existing surgical pathways is ideal. Based on 13 studies of fair to good methodological quality, the Australian PBM guidelines (Australian National Blood Authority 2011a) recommend that iron deficiency be treated pre-operatively (Grade 1B). We identified no RCTs that should alter this. There are two options for IDA treatment: oral or intravenous iron supplementation. Oral formulations are widely available, cheap, and safe. However, oral iron is poorly bio-available (Finch, 1994). The absorption of oral iron is variably inhibited by dietary iron chelaters (e.g. phytates and tannins) and common medications including proton pump inhibitors. Iron tablets often cause dose-related gastrointestinal side effects. Furthermore, enteral iron is poorly absorbed in the context of chronic inflammation (Weiss, 2009) or chronic renal failure, making oral iron unlikely to be effective in FID. Even with optimal absorption, the rate of haemoglobin correction is slow, typically about 10 g/l per week (Weiss, 2009). After correction of anaemia, oral iron supplementation for up to 3 months is required before iron stores are replete (Goddard et al, 2011). The co-administration of vitamin C supplements is recommended to enhance duodenal iron uptake (Brise & Hallberg, 1962). Although oral iron can be as effective as intravenous iron in the medium term (Evstatiev et al, 2011), the intravenous route is increasingly used. Where rapid correction is desirable (e.g. cancer surgery), the parenteral route is more attractive (Beris et al, 2008). We identified one RCT (Kim et al, 2009) that compared pre-operative oral iron with parenteral iron sucrose. Target Hb attainment was more likely with parenteral iron. A single total dose infusion (ferric carboxymaltose) was furthermore found to correct anaemia quicker than iron sucrose in a small observational study (Bisbe et al, 2011). Recently developed intravenous iron formulations are safer than older preparations, because the iron is encased in a carbohydrate shell (Auerbach & Macdougall, 2014). Based on all prospective reports, where high molecular weight dextrans are avoided, the serious adverse event rate is estimated at 6 weeks, and for >12 weeks in 71/91 studies, making the applicability to pre-operative patients unclear. A Cochrane review also found strong evidence that peri-operative transfusion increases the likelihood of cancer recurrence after potentially curative surgery (Amato & Pescatori, 2006), so blanket advice to avoid ESAs in cancer surgery is not appropriate. A multicentre RCT powered to evaluate the safety of ESA therapy in elective spinal surgery without pharmacological thromboprophylaxis found higher rates of ultrasound-diagnosed deep vein thrombosis (Stowell et al, 2009). ESA therapy is expensive – USD 7300 per avoided transfusion in a recent RCT (So-Osman et al, 2014) – and the authors concluded that the cost was not justifiable. The clinical and cost-effectiveness of ESA therapy also falls within the scope of forthcoming guidance from the UK National Institute for Health and Clinical Excellence (2014). ESAs thus appear effective in reducing transfusion risk, but are expensive and have potential side effects and serious complications. Individual risk-benefit decisions are necessary and further research is urgently required that evaluate the safety of ESA therapy when used peri-operatively. Consequently, we make no recommendation for the use of ESA therapy, other than where transfusion avoidance of itself is clearly beneficial to the individual, e.g. in patients who refuse blood, or those with complex alloimmunization. The UK Renal Association (2010) recommends that if used, ESA therapy be given to achieve a target Hb of no higher than 120 g/l. The Australian PBM group (Australian National Blood Authority 2011a), Australian NATA (Goodnough et al, 2011) and the UK Renal Association) all recommend that iron be co-administered with an ESA in order to maximize its efficacy. We could find no evidence on the best route of iron administration to maximize ESA efficacy. Given that compliance with oral iron is often difficult and its absorption not certain (Finch, 1994; Weiss, 2009) and because ESAs are expensive, it may be pragmatic to consider the intravenous route so that therapeutic plasma iron levels are assured. Research is required before any recommendation can be made. There is emerging randomized evidence regarding very short-term management of anaemia before cardiac (Weltert et al, 2010; Yoo et al, 2011) and arthroplasty (Na et al, 2011) surgery. In these trials, an ESA and parenteral iron were started <1 week before surgery, without attempt to diagnose the cause of the anaemia. All three trials showed significant decreases in transfusion rate but did not report on other patient outcomes and the absolute numbers of patients involved are small. Short-term oral or parenteral iron alone proved ineffective in RCTs (Serrano-Trenas et al, 2011; Garrido-Martin et al, 2012), although there is observational evidence to the contrary (Munoz et al, 2014). In one small RCT (n = 79) in patients with intertrochanteric fracture (Kateros et al, 2010), ten daily ESA doses reduced transfusion requirements and accelerated Hb mass recovery. No other patient outcomes were reported. Short-term combination therapy that includes an ESA is thus potentially efficacious, but safety and cost-effectiveness data is lacking. This GDG considers that, until better safety data is available, combination therapy should only be used after patient counselling that is explicit about the lack of safety data. We therefore make no recommendation. 'Top-up' transfusions have traditionally been used to prepare anaemic patients for surgery, but no randomized evidence exists that this is of benefit (Shander et al, 2011). Traditional practice includes pre-operative transfusion at trigger values as high was 120g/l (Karkouti et al, 2012), i.e. transfusion to correct pre-operative anaemia, in anticipation of future blood loss. Trials of restrictive versus liberal transfusion (for haemodynamically stable patients) showed restrictive strategies to be superior in critical illness (Hebert et al, 1999) and gastrointestinal bleeding (Villanueva et al, 2013), and safe in elderly patients after hip fracture repair (Carson et al, 2011). Single studies of perioperative transfusion in major cancer surgery (de Almeida et al, 2015) and cardiac surgery (Murphy et al, 2015) suggested a trigger Hb of 90 g/l may be superior to 70 g/l or 75 g/l, respectively. However, even the 'liberal' groups in the above trials were managed restrictively when compared to traditional practice (Karkouti et al, 2012). Whilst the optimal transfusion threshold remains to be defined, and whilst it is probably different for different patient groups, we found no literature to support transfusion being beneficial where the target is a normal or near-normal value (i.e. transfusion given to correct anaemia). We further found no good evidence in support of pre-operative transfusion to improve surgical outcomes, and in the absence of other PBM measures, pre-emptive transfusion appears not to reduce total transfusion requirements (Karkouti et al, 2012). There is thus a lack of evidence for transfusion as an effective treatment to protect against the deleterious effects of pre-operative anaemia. This should be taken in combination with the dose-dependent relationship between transfusion and complications (Ferraris et al, 2012), with systematic reviews showing liberal transfusion to be either inferior (Carson et al, 2012; Rohde et al, 2014) or non-beneficial (Carson & Strair, 2014; Holst et al, 2015) compared to restrictive strategies, and evidence of both short-term (Bolton-Maggs et al, 2013) and longer-term (Amato & Pescatori, 2006) risks of transfusion. Elective transfusion to normal or near-normal Hb in anticipation of operative blood loss is therefore not recommended; particularly as evidence-based anaemia treatments are available. In situations where transfusion is likely to be unavoidable despite appropriate transfusion practice and intra-operative patient blood management (e.g. severe refractory anaemia or urgent major surgery), the question of whether pre-operative transfusion is superior to intra-operative transfusion is as yet unanswered. However, it should be emphasized that these situations are not common in non-emergent surgery if surgical pathways are properly set up. Appendix S4 gives an overview of the management options available for pre-operative anaemia. While the advice and information in these guidelines is believed to be true and accurate at the time of going to press, neither the authors, the British Society for Haematology nor the publishers accept any legal responsibility for the content of these guidelines. With thanks to: Dr Alan Nye, General Practitioner, Oldham, for his expert review. Appendix S1. Literature review. Appendix S2. Example algorithms. Appendix S3. Relative properties of intravenous iron preparations commercially available in the UK (2014). Appendix S4. Overview of treatment options for pre-operative anaemia. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.