| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Departmentof Anesthesiology, Duke University Medical Center, Duke Health Technology Solutions, Duke University Medical Center, Durham, North Carolina, Department of Anesthesiology, University of Miami, Miami, Florida
Address correspondence and reprint requests to Ronald P. Olson, MD, Duke University Medical Center, Box 3094, Durham, NC 277l0, Telephone 919 684 2025 Fax 919 681 8484, e-mail olson012{at}mc.duke.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
There is increasing realization that the perioperative period is not an appropriate setting for screening and investigating asymptomatic anemia (10–12). This is largely because the prevalence is infrequent except in neonates, the elderly, smokers, patients using anticoagulants, or patients with specific disease (12,13) and the likelihood of perioperative management of anemia is small (2,4).
We hypothesize that the prevalence of anemia of clinical interest by current practice in generally healthy outpatient surgical candidates is too infrequent to warrant routine screening. There has been no large study of the prevalence of anemia in this increasingly large proportion of surgical candidates, especially anemia as defined in currently practical terms. This retrospective study assesses the prevalence of anemia defined as hemoglobin (hgb) <9 g/dL and also examines whether the discovery of this level of anemia in asymptomatic outpatient surgical candidates results in any change in perioperative management.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
The laboratory database was analyzed to find all patients who had hgb 
9 mg/dL. This threshold was selected as one at which even asymptomatic anemia might require treatment. The admission database was used to determine the ASA physical classification, age, and gender of all patients with a preoperative hgb 9 mg/dL. The prevalence of preoperative hgb levels 9 mg/dL in various subgroups of the 14,337 patients were calculated and compared. The electronic common data repository records of any of the above patients who were ASA class 1 or 2 with preoperative hgb 9 mg/dL were then examined to see if there was evidence of further anemia investigations or treatment. If there was insufficient information in the electronic record, the written hospital record was examined as well. Statistical comparisons of group proportions were made using Pearson’s 
2 test or Fisher’s exact test with small counts. Alpha <0.05 was considered significant.
The ASA classification was based only on information that would have been available preoperatively from the preoperative clinical assessment, paper medical record, and electronic medical record that included previous lab results and anesthetic records.
As an additional crosscheck, the transfusion service’s database was reviewed to see if any of these patients received blood products. As no therapy was apparently impacted by measuring the preoperative hgb in this subset of patients, zero numerator statistics give a 99% confidence interval that the chance of a hgb test result impacting management is at most 0.048% (14).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
9 g/dL, giving a prevalence of 0.8% (95% confidence interval [CI], 0.6%–1.0%). In the ASA III patients, 138 of 3499 patients had hgb 9 g/dL, giving a prevalence of 3.9% (95% CI, 3.0%–4.9%). In the ASA IV patients, 18 of 205 patients had hgb 9 g/dL, giving a prevalence of 8.8% (95% CI, 5.3%–13.5%). All 3 pairwise comparisons showed Fisher’s exact P < 0.0032.
The prevalence of hgb 9 g/dL was slightly higher in females (0.9%; 95% CI, 0.7%–1.2%) than males (0.5%; 95% CI, 0.3%–0.9%) (P = 0.0340). When analyzed by age, the prevalence in patients aged <13 yr was 4.6%, which was significantly more than in patients aged >13 yr, where it was 1.6%. In the 66 infants <1 yr of age, the rate was 6.1%, although this was not significantly different from the rate of 4.4% in the 1–12 yr group.
Of the 75 ASA I–II patients with hgb 9 mg/dL, the average preoperative hgb was 8.3 g/dL with a standard deviation of 0.7 g/dL. The lowest was 5.6 g/L. The 4 patients with results <7 g/dL had clear indicators of potential anemia (advanced human immunodeficiency virus, sickle cell disease, or history of bleeding). It is questionable whether these individuals should have been classified as ASA II. There was no evidence of further investigation or perioperative treatment of anemia in these 75 patients undergoing elective surgery, including those with hgb <7 g/dL.
Review of the transfusion records showed that 4 of the 9584 ASA I–II patients (0.05%) received red blood cells on the same day or the day after the hgb test. All had hgb >9 g/dL. All had clear pretest clinical indictors of potential anemia. Three had a history of anemia, one was receiving heparin, and one had advanced cancer. There was no evidence that the decision to transfuse or any other perioperative management had occurred as a result of the preoperative hgb. Of these 9584 patients, 11 had received transfusions in the 6 mo before the procedure. Review of the medical records showed that in no case had a screening or preoperative hgb result been part of the decision to transfuse.
One other patient, initially classified as ASA II, was transfused postoperatively. She had sickle cell disease and had presented to the emergency department with increased vaginal bleeding; her hgb was <9 g/dL. She was reclassified as an ASA IIE and removed from the ASA I–II study group.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
In previous times, surgery usually entailed significant risk of blood loss and physiological insult as well as a lengthy hospital admission. Most patients with a hgb <10 g/dL received a transfusion. Therefore, the prevalence of anemia that required transfusion was frequent and routine hgb testing was appropriate.
Today, many surgical procedures involve minimal risk. There has also been a change in transfusion thresholds. The National Institute of Health, American College of Physicians, and American Society of Anesthesiology consensuses state that transfusion is not necessarily indicated in a patient with a hgb as low as 7 g/dL if that patient is normovolemic, asymptomatic, and no further blood loss is anticipated (15–18).
Screening in low-prevalence groups yields so many false positives that a positive test is not useful information (10–12). Guidelines for anemia screening set by authoritative organizations state that it should only be done in high-risk infants once before the age of 9 months and in menstruating women every 5–10 years (13,19). If the prevalence of anemia in asymptomatic outpatient surgical candidates is similar to that in the general population, screening criteria should be similar to those recommended by these authorities. Unindicated hgb screening, especially without the clinical context and follow-up of primary care, exposes patients to the net negative effects of testing with little benefit.
It has been shown that actual follow-up of abnormal hgb results in the increasingly short perioperative period is quite rare (2,20). It has been stated that there is now more risk of litigation from not following up a test than from not ordering it (11,20–22). We agree.
In this study, we assessed the prevalence of hgb 9 mg/dL in healthy outpatient surgical candidates. This threshold was chosen as a balance between increasing evidence that the traditional transfusion threshold of hgb 10 mg/dL is no longer appropriate in asymptomatic patients, (15,16,18) and the concern of many clinicians to at least be monitoring hgb at levels slightly below the traditional one.
The prevalence of hgb 9 g/dL in ASA I–II patients in this study was only 0.8%. Generally, a prevalence of 1%–5% is needed to make screening beneficial (23–27). As an example, the prevalence of iron deficiency anemia (hgb <13.5 g/dL) in adult males is 2%, and screening is not recommended for this group (13).
This study of 9584 such patients showed that only 4 had transfusions, and all of these patients had been expected based on information other than the preoperative hgb. Unexpected anemia requiring management is so rare in these patients that routine hgb testing has no benefit.
Routine laboratory tests cannot substitute for a clinical examination (1–9). But if there are any clinical indicators of potential anemia, including high-risk surgery or high-risk patients, then hgb testing is indicated. Clinical indicators of anemia are any feature of the patient history or examination that might lead one to suspect anemia. One such list is given in a recent textbook (21).
There are shortcomings of this study. It is possible that patients with low hgb results had surgery delayed and so were excluded from the sample or only entered the study after correction of anemia. This possibility would exist even if hgb results for all 21,080 patients were recorded. Analysis of the events leading up to the referral to surgery would be needed. That information was not reliably available in this study.
Only 14,337 of 21,080 outpatients (68%) had hgb tests performed within the 30 days before surgery. There is a possibility of selection bias, which excluded some cases of low hgb. However, all of the unmeasured patients proceeded to surgery without further testing, which would have been unlikely if there had been anemia concerns. In addition, analysis of the preoperative transfusions in the studied group showed that all transfusion decisions were based on historical information independent of screening hgb tests. This suggests that these test results are rarely part of management decisions.
More likely the bias is in the other direction, with the clinicians choosing to not order the test because of low probability of anemia or knowledge of previous outside test results showing normal levels. Thus the population studied likely had a more frequent incidence of abnormal hgb than the general ASA I–II outpatient surgical population. This would effectively increase the prevalence of anemia in our tested population, making the conclusions of this study even more valid in a less selected population.
Because 90% of the study population was between the ages of 13 and 79 years, conclusions cannot be applied to children or the very old.
This large study supports previous ones showing that selective testing based on the clinical examination is safe (4,28), avoids the drawbacks of false positives (1,7,21), and reduces unnecessary expense (1,6,7,28,29). The probability that preoperative hgb will impact on management is estimated to be <0.05%.
|
Conclusion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  W.-C. Wu, T. L. Schifftner, W. G. Henderson, C. B. Eaton, R. M. Poses, G. Uttley, S. C. Sharma, M. Vezeridis, S. F. Khuri, and P. D. Friedmann Preoperative Hematocrit Levels and Postoperative Outcomes in Older Patients Undergoing Noncardiac Surgery JAMA, June 13, 2007; 297(22): 2481 - 2488. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. P. Olson and D. Lubarsky Anemia Screening in Elective Surgery: Definition, Significance and Patients' Interests Anesth. Analg., September 1, 2006; 103(3): 779 - 780. [Full Text] [PDF]
|
|
|
|
|
|
A. Shander, M. Javidroozi, and L. T. Goodnough Anemia screening in elective surgery: definition, significance and patients' interests. Anesth. Analg., September 1, 2006; 103(3): 778 - 779. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
