| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Department of Anesthesiology, Joint Appointments in Medicine and Health Policy & Management, Johns Hopkins Medical Institutions, Baltimore, Maryland
Address correspondence to: Lee A. Fleisher, MD, The Johns Hopkins Hospital, 600 N. Wolfe Street, Carnegie 280, Baltimore, MD 21287. Address e-mail to lfleishe{at}jhmi.edu
Over the past decade, there has been a clear and conscious effort to define the indications for laboratory testing before surgery. Up through much of the 1980s, there was a standard battery of tests that was considered routine and required for all but the simplest of procedures (1). The last decade has witnessed a marked change in this approach, in part because of economic pressures but also because there is a better understanding of the meaning and predictive value of abnormal tests (2–5). Clearly, there is an imperative today to limit preoperative testing to those that are "indicated." One such frequently cited indication has been advanced age. The study by Dzankic et al. (6) in this issue of Anesthesia & Analgesia clearly adds to the data regarding indications for preoperative diagnostic laboratory testing.
To understand how the indications for preoperative testing have evolved, it is important to review the history of preoperative testing research. Of note, the American Society of Anesthesiologists (ASA) had attempted to develop Practice Parameters on preoperative testing, but after an extensive review of the literature, they settled instead on producing a Practice Advisory. Why was this route taken? The simplest answer is the lack of strong evidence. Until recently, there have been no randomized trials designed to determine the utility of preoperative laboratory data. Virtually all of the literature in this field has reported on prospective cohort studies and case series that have evaluated outcomes (morbidity, mortality, and perioperative management changes) in patients who have and/or have not had specific laboratory tests (7).
Using a standard for evidence of the randomized clinical trial, the initial Task Force on Preoperative Testing did not find sufficient evidence to advocate any specific testing. Importantly, the lack of evidence regarding specific tests is not evidence that there is no value to testing (8). That is to say, there is insufficient evidence to either advocate routinely performing any given test or to clearly define preoperative tests that are without value. Therefore, more data are needed.
Of note, strong evidence regarding the value of preoperative testing has recently been published in the form of a large-scale randomized trial. Schein et al. (9) performed a study of almost 20,000 patients undergoing cataract surgery who were randomized to either routine laboratory testing or no-routine testing. They reported no difference in perioperative morbidity and mortality between those who did versus those who did not receive routine testing. In interpreting the study, it is important to take into account some of the comments made by Roizen (10) in his accompanying editorial. Specifically, all of these patients were evaluated by their primary medical doctor and therefore deemed to be in optimal medical condition to undergo elective cataract surgery. Patients in their no-routine testing group underwent tests "when the history or a finding on physical examination would have indicated the need for a test even if surgery had not been planned." Schein et al. (9) excluded patients who required medical care with laboratory testing within 28 days before enrollment. Second, the actual rates of morbidity and mortality are exceedingly low after cataract surgery and therefore few if any interventions could be shown to statistically affect outcome. Therefore, the ability to generalize these results beyond this setting (low-risk surgery performed under conscious sedation) is limited.
In the absence of more generalizable randomized clinical trial data, what is the best method to define the value of a test? Several studies have focused on the lack of morbidity associated with the elimination of tests in healthy subjects, yet few focus on patients with known risk factors other than cardiopulmonary disease (1–4). In virtually all risk indices, including the ASA physical status classification, advanced age (>70 yr) is defined as a high risk or high classification group (11,12). Therefore, age is frequently used as an indication for routine testing. In this issue of Anesthesia & Anlagesia, Dzankic et al. (6) attempt to address the implications of advanced age using a prospective cohort study of a consecutive group of patients >70 yr of age undergoing noncardiac surgery requiring general or regional anesthesia. They had two goals: to define the incidence of preoperative laboratory test abnormalities and to determine the predictive value of an abnormal test for an adverse event. With respect to the first goal, the authors demonstrated that platelet count abnormalities were rare although the frequency of creatinine, hemoglobin, and glucose abnormalities was in the 7%–12% range. With respect to the second goal, the authors used an aggregate end point of all-cause adverse events. Although some abnormal laboratory values predicted adverse outcomes, only ASA physical status (>2) and surgical risk were independent predictors of adverse outcome using multivariate modeling. These findings suggest that the prevalence of abnormal laboratory values is small in the elderly and may identify a high risk subset with more comorbidities, hence higher ASA physical status scores, which were independent predictors of poor outcome. On the basis of these results, the authors suggest that "routine preoperative laboratory testing for hemoglobin, creatinine, glucose, and electrolytes based on age alone may not be indicated in geriatric patients."
In principle, this recommendation fits within the currently evolving concept of testing based on patient comorbidity and surgical risk. However, before we totally abandon routine testing in the elderly, it is important to determine how the information about an abnormal test is used because a basic tenet of preoperative testing is that a test should only be performed if it will actually be used to modify care. One use of preoperative testing is that laboratory data only becomes important as baseline information if problems develop. This is particularly relevant for hemoglobin level, for which the risk of surgical bleeding should be incorporated into the indications. The other use is that the presence of an abnormal test will lead to the identification or confirmation of an abnormal disease process that affects management. For platelet count, the ability to confirm an abnormality in the patient with a history of a bleeding diathesis has been well described, and the small prevalence of abnormalities in the current study supports this approach. However, because the current study was not randomized, it is impossible to determine if cases were canceled before surgery based on the presence of an abnormal test or treatment of other abnormalities (e.g., high glucose, low hematocrit) during surgery that led to an improved outcome. Additionally, the preoperative test may in fact modify the assignment of ASA physical status, and therefore become less predictive in the multivariate model. For example, in the univariate analysis, creatinine was predictive of poor outcome, and the presence of chronic renal insufficiency may not be detected by history alone. Importantly, the perioperative period is not a black box, and presumably interventions by an anesthesiologist (e.g., change in fluid management or pharmacologic interventions in a patient with renal insufficiency) will lead to improved outcome compared with those situations in which management was not changed. Although the actual rate of abnormalities is small, it is not as small as in the healthy population (7). Therefore, total abandonment of routine testing based on age alone must be tempered against the probability that disease will be detected by the laboratory screening and that the extent of surgery is sufficient that management will be modified.
In summary, the current study extends the body of knowledge that defines the prevalence of laboratory abnormalities found on preoperative screening to an elderly population and identifies predictors of adverse events. As in a younger population, a thorough clinical history and determination of surgical risk remains critical as a means of identifying individuals in whom baseline testing should be performed. However, in the absence of randomized controlled trials in the noncataract surgery population, obtaining information that may change perioperative management in the elderly should continue to guide physicians in their decision to order preoperative laboratory tests.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
