| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Department of Anesthesiology, University of Ottawa, Ottawa Hospital, Ottawa, Ontario, Canada.
Address correspondence to: Dr. Gregory L. Bryson, Department of Anesthesiology, The Ottawa Hospital – Civic Campus, 1053 Carling Avenue, Ottawa, Ontario K1Y 4E9, Canada. Phone: 613-761-4169; Fax: 613-761-5209; E-mail: glbryson{at}ohri.ca
 |
Abstract |
|---|
 TOP Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Methods: Following Research Ethics approval all patients undergoing elective surgery at TOH-Civic Campus in the first three months of 2004 were identified. One hundred charts from each month were randomly selected and analyzed retrospectively by a single reviewer. The ordering and results of four preoperative tests were characterized. The indication for preoperative testing was identified from preoperative notes and the source of non-compliant orders was identified. Compliance with the CAS, OPTF, and TOH guidelines was documented. Abnormal test results were analyzed for evidence of a subsequent change in clinical management.
Results: The charts of 294 of the 2,116 patients who underwent elective surgery at TOH-Civic Campus in the first three months of 2004 were reviewed. A total of 534 tests were ordered on 198 patients (67%). Non-compliance rates varied significantly (5–98%) depending on test and guideline analyzed. Results of 329 tests (61.6%) were normal. Management was changed by 14 of 534 tests ordered (2.6%). Surgery proceeded as scheduled in all cases.
Conclusion: The majority of elective surgical patients undergo preoperative testing. Non-compliance with guidelines ranged from 5 to 98%. The results of most tests were normal and influenced management in only 2.6% of cases.
|
Introduction |
|---|
|
TOPAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In an attempt to control the costs associated with perioperative care, a variety of agencies have published guidelines to rationalize preoperative testing practice. As a result, preoperative testing at The Ottawa Hospital (TOH) is subject to influences at the national, provincial, and local levels. At the national level the Canadian Anesthesiologists’ Society (CAS) publishes recommendations for preoperative testing in the Guidelines to the Practice of Anesthesia.4 The CAS advocates that investigations should not be ordered on a routine basis but rather based on the patient’s health status, drug therapy, and with consideration to the proposed surgical intervention. At the provincial level the Ontario Preoperative Task Force (OPTF), in collaboration with the Guidelines Advisory Committee of the Ontario Medical Association, published the Ontario Preoperative Testing Grid in 2003.5 The OPTF grid describes indications for a variety of preoperative tests in use in hospitals across Ontario and provides consensus recommendations for appropriate testing practice. At the local level, the pre-admission unit at TOH implemented a policy defining the indications for preoperative testing in our institution. The TOH policy was approved by the departments of Anesthesiology, Surgery and Nursing and was implemented in June 2003.6 The ordering of tests was to be individualized for each patient, and nurses in the clinic were responsible for ensuring that required tests were completed.
The primary objective of this study was to determine if testing practices at TOH-Civic Campus were compliant with any of the three applicable guidelines. Secondary objectives included: identification of the source of testing orders non-compliant with guidelines; documentation of the rates of normal and abnormal test results, and, an evaluation of the impact of abnormal test results on perioperative care.
|
Methods |
|---|
|
TOPAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Orders and results for the four most common preoperative tests were considered: electrocardiogram (ECG), chest x-ray (CXR), complete blood count (CBC), and serum electrolytes (LYT). The source of the order for the preoperative test was identified and compliance with the CAS, OPTF, and TOH guidelines was documented. Indications for testing were found in the surgeon’s history or the preadmission assessment. The status of a given laboratory test in each patient, indicated vs non-indicated, was made for each test for all three guidelines. The indications for the individual laboratory tests and guidelines are summarized in the Appendix (available as Additional Material at www.cja-jca.org). Clinical agreement regarding the indication for each test ordered was assessed for all possible pairs of guidelines (CAS-OPTF, CAS-TOH, OPTF-TOH).
Results of the tests were classified as either normal or abnormal using hospital laboratory reference standards. The charts of patients with abnormal laboratory values were then reviewed to identify if a change in patient management could be identified. A change in patient management was broadly defined and included, but was not restricted to, request for old records, orders for repeat testing, requests for consultation, and new medication orders.
Categorical data were described as proportions and compared with Chi-square statistic. Continuous data were described using mean and standard deviation and compared using unpaired Student’s t tests. Agreement was characterized using the Cohen’s kappa statistic. A kappa of 0.5 defined agreement between guideline pairs no greater than that expected by chance alone while a kappa of 1.0 defined complete agreement. In all analyses a P value of < 0.05 was considered statistically significant.
|
Results |
|---|
|
TOPAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
.
|
. Non-compliance rates were lowest when the OPTF guidelines were considered (64 of 534 tests; 12%) and greatest with CAS guidelines (166 of 534 tests; 31.4%). Surgeons’ orders were responsible for 425 of 534 tests (80.0%) ordered preoperatively; it is therefore of little surprise that surgeons were the source of most orders non-compliant with guidelines. The attending surgeon ordered 67.6 to 90% of those tests in this review that were non-compliant with guidelines.
|
The level of agreement regarding the indication for preoperative laboratory testing is shown in Table III. Indications for LYT were remarkably similar across guidelines and led to high levels of agreement among all guidelines. In contrast, indications for CXR varied significantly between guidelines and were responsible for significant disagreement. When CBC and ECG guidelines were evaluated kappa scores > 0.5, indicating agreement greater than that expected by chance, were found in approximately half of guideline pairs. Agreement between guideline pairs for most tests was greatest when TOH and CAS guidelines were compared.
|
. The 329 of 534 (61.6%) tests ordered returned with normal results. Abnormal results were noted in 189 tests yet evidence of action prompted by these results could be documented in only 14 of 534 cases (2.6%).
|
provides an example of the influence of the CAS indications for testing on the proportion of abnormal test results. In this example, only LYT was statistically more likely to be abnormal in those tests indicated by the guideline. When other guidelines (not shown) were considered, relationships between indication for LYT and ECG and abnormal results were inconsistent. In all cases there was no significant relationship between the indication for testing and abnormal results of CXR and CBC.
|
|
Discussion |
|---|
|
TOPAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Agreement of testing indications between the guidelines studied was highly variable. Indications for preoperative evaluation were most similar for LYT across all guidelines. All three guidelines clearly defined and prescribed indications for preoperative assessment of LYT. Indications common to all guidelines included patients with renal disease, hypertension, diabetes mellitus, and conditions affecting the adrenal hypopituitary axis as well as for patients taking diuretics and digoxin. The consistency of the guidelines, if not the ordering physicians, was reflected in similar non-compliance rates ranging from 25 to 36%. For all other tests clinical agreement with the recommendations of the OPTF were most commonly found in conflict with the other guidelines. The OPTF guidelines are relatively lenient and recommend a CBC for all women undergoing surgery and an ECG on all patients aged 45 or greater. These more inclusive standards led to non-compliance rates for both tests of < 6%; compare this to the more restrictive CAS standards where non-compliance was > 20%. It is clear that there is potential to rationalize and harmonize the indications for testing across all three levels of regulation. A consistent and evidence-based approach to preoperative testing should reduce the numbers of tests performed preoperatively, but whether these tests provide clinically useful information is an altogether different question.
If testing is to be decreased, what criteria can one use to reform testing practice? Gender in the present study was poorly predictive of abnormal results and a source of significant disagreement between the guidelines evaluated. Dzankic et al. demonstrated that abnormal laboratory results were common in patients > 70 yr of age, but only surgical risk and ASA classification were associated with adverse outcome.2 We also found age > 65 yr was more commonly associated with abnormal results in LYT and CBC but again noted that age based inclusion criteria were inconsistent between guidelines. Elimination of gender and age based indications for testing in favour of indications based on comorbidity would therefore seem logical first steps in rationalizing preoperative testing practice. In the wake of Schein’s findings regarding the utility of preoperative testing in cataract surgery,3 many clinicians question the need for testing in other ambulatory procedures. Abnormal results were equally common in patients having ambulatory and inpatient surgical procedures in this series suggesting that a reevaluation of testing guidelines need not be restricted to outpatient procedures. Would more careful application of existing guidelines be the answer? Tests indicated by CAS testing guidelines were no more likely to be abnormal than those ordered without indication, LYT being the notable exception. These findings suggest that the guidelines are probably overinclusive and could be further refined. If testing policy is refined to reduce unnecessary testing what benefits might be expected?
The cost savings associated with rationalizing preoperative testing practice can be substantial. Estimating the cost of testing can be challenging in centres, such as TOH, which receives institutional funding for all laboratory services. Using the Ontario Schedule of Benefits and TOH laboratory charges the cost of the tests in this study can be estimated as follows: ECG = $18.24 CDN, CXR = $38.54 CDN; CBC $9.60 CDN; and LYT $8.22 CDN. The cost incurred by tests ordered in our review of 300 patients that were non-compliant with TOH guidelines was estimated at $3,300.00 CDN. At an average cost of just over $10 CDN per patient we calculate that non-compliant testing in elective surgical patients costs our institution over $87,000 per year. Given the absence of adverse postoperative adverse events in this study, and the further difficulty of attributing such events to an absent test, we cannot estimate the costs tests indicated but not completed. Regardless of the economic issues at hand one wonders if preoperative laboratory testing offers significant clinical value.
Abnormal results were noted in approximately one third of all tests ordered. The frequency of abnormal results reported in this series was similar to that reported in a systematic review of preoperative testing conducted for the National Health Service (NHS) by the NHS R&D Health Technology Assessment Programme.1 That action was taken in < 3% of test results is also consistent with the NHS review. The lack of correlation between abnormal results and clinician response suggests that the abnormalities reported were of a minor nature, or that clinicians were willing to proceed with anesthesia and surgery in the face of significant abnormalities. In either event it is clear that the elimination of a significant proportion of preoperative laboratory investigation could be accomplished without changing clinical care.
This study has several limitations. Foremost among them is the inability to generalize these results to other centres. That said, manufacturing change is challenge for all institutions, large and small. The TOH experience suggests that implementation of our own policy, agreed upon by the departments of Anesthesiology, Surgery, and Nursing, still resulted in a large number of unnecessary tests. Each institution will have its own unique problems and solutions in the rationalization of preoperative testing practice. Old habits die hard and departments wishing to change their testing habits will have to identify their unique problems and propose their own solutions. Another significant limitation was the retrospective determination of testing indication and outcome assessment. It is quite possible that our test indications may have been incorrectly assigned based on written patient records or that clinical action on test results was underestimated. That no case was delayed, cancelled, or rescheduled lends support to the relatively low impact of testing on clinical decision making.
In conclusion, the results of this review suggest that unnecessary laboratory testing is common in our institution despite the presence of an approved policy dictating preoperative testing practice. Age- and gender-related indications for testing in the OPTF guidelines significantly broaden indications for testing and resulted in poor agreement with other guidelines. Regardless of indication, preoperative tests were frequently normal and rarely changed perioperative management. There is ample opportunity to rationalize testing practice and decrease testing-related costs without altering patient outcome.
|
Footnotes |
|---|
Accepted for publication July 7, 2005. Revision accepted September 18, 2005.
|
References |
|---|
|
TOPAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
2 Dzankic S, Pastor D, Gonzalez C, Leung JM. The prevalence and predictive value of abnormal preoperative laboratory tests in elderly surgical patients. Anesth Analg 2001; 93: 301–8.
3 Schein OD, Katz J, Bass EB, et al. The value of routine preoperative medical testing before cataract surgery. Study of Medical Testing for Cataract Surgery. N Engl J Med 2000; 342: 168–75.
4 Canadian Anesthesiologists‘ Society. Guidelines to the practice of anesthesia. The pre-anesthetic period. Available from URL; http://www.cas.ca/members/sign_in/guidelines/practice_of_anesthesia/default.asp?load=preanesthetic. 2004 (accessed 01.08.2005).
5 Badner N, Bryson G, Kashin B, et al. Ontario Preoperative testing grid. Endorsed by the Ontario Guidelines Advisory Committee. Available from URL; http://gacguidelines.ca/pdfs/tools/Ontario%20Preoperative%20Testing%20Grid.pdf. 2004 (accessed 01.08.2005).
6 PAU Committee, The Ottawa Hospital. Guidelines for laboratory testing prior to elective surgery. Available from URL; http://www.anesthesia.org/civic/Policies/PAU-Preoperative%20Testing%20Policy.pdf. 2003 (accessed 01.08.2005).
7 Davis DA, Taylor-Vaisey A. Translating guidelines into practice. A systematic review of theoretic concepts, practical experience and research evidence in the adoption of clinical practice guidelines. CMAJ 1997; 157: 408–16.[Abstract]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
