| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
J Am Dent Assoc, Vol 139, No 6, 697-703.
© 2008 American Dental Association |
CLINICAL PRACTICE |
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Methods. This prospective cohort study involved use of an INR testing device to obtain INR test values in the dental office for patients thought to be at risk of experiencing bleeding complications after undergoing invasive dental procedures. The authors recorded demographic, social and medical history data, as well as clinical signs and symptoms of liver disease. The authors considered an INR out of range if it was greater than or equal to 1.4 for patients with potential liver disease and greater than 3.5 for patients receiving warfarin.
Results. The authors completed an in-office INR test for 66 patients receiving warfarin whose INR had not been tested within the preceding 48 hours and 34 patients suspected of having liver disease. Eleven (17 percent) patients receiving warfarin and seven (21 percent) patients suspected of having liver disease had INR values considered out of range. Dental treatment was deferred for eight of 11 patients in the warfarin group who had INR values in the range of 3.6 to 7.4, while three others had dental procedures without bleeding complications. Six of seven patients who had documented or suspected liver disease and an out-of-range INR (range 1.5–2.5) underwent their dental procedures without experiencing bleeding complications.
Conclusions. Use of an in-office INR test indicated a high incidence of elevated INR values. The results of this study point to the importance of obtaining current INR values before performing invasive dental procedures for patients receiving warfarin therapy whose INR values have not been tested recently, and for patients thought to be at risk of developing or having liver disease.
Key Words: Anticoagulation therapy; dental care for chronically ill patients; ethanol; hemorrhage; oral hemorrhage; oral surgery; risk assessment
Abbreviations: CLIA: Clinical Laboratory Improvement Amendment. • ED: Emergency department. • FDA: U.S. Food and Drug Administration. • INR: International normalized ratio. • POC: Point of care.
Patients with congenital or acquired coagulopathies are common in the general population. Dentists should have access to the patients’ appropriate and current laboratory values to prevent bleeding complications during and after invasive dental procedures, in particular for patients receiving warfarin and those who have liver disease.
Various protocols have been proposed for the dental care of patients receiving antithrombotic medications, but the lack of a definitive standard of care suggests a lack of necessary evidence. Bleeding after a dental procedure is rare and almost always easily controlled with local measures. However, some authors recommend withholding medication,1,2 despite the potentially devastating consequences, which can include life-threatening thromboembolic events such as stroke1–4 —the risk of which is three to five times greater than the risk of postoperative bleeding that is uncontrollable by local measures.5
Dental care of patients with liver disease requires a familiarity with the pathophysiology, signs and symptoms associated with liver damage. Liver impairment—whether resulting from infectious disease, alcohol abuse or vascular or biliary congestion—can significantly alter a patient’s ability to tolerate surgery and general anesthesia and poses a significant risk of perioperative and postoperative complications. Therefore, it is important to determine preoperatively the patient’s degree of hepatic compromise; risk of experiencing oral bleeding as a result of impairment of coagulation factors synthesized in the liver; and risk of experiencing thrombocytopenia, which may develop from splenomegaly secondary to portal hypertension.6–9
The risk of bleeding associated with warfarin use or with coagulopathies resulting from liver disease can be assessed preoperatively with a prothrombin time reflected as an international normalized ratio (INR) laboratory test result. This laboratory value reflects the extrinsic coagulation pathway, which is affected by both liver disease and warfarin therapy.
The major purpose of this study was to evaluate the utility of an in-office INR testing device in identifying the incidence of patients who are receiving warfarin therapy or have liver disease and have INR measurements considered too high (out of range) to receive dental treatment. A secondary objective was to determine if demographics, patient-reported social and medical history and clinical signs and symptoms were associated with an elevated INR. Finally, we determined the incidence of bleeding complications after invasive dental treatment among patients who received treatment while having elevated but clinically acceptable INR values.
We considered an INR out of range if it was 1.4 or higher for patients with potential liver disease, and higher than 3.5 for patients receiving warfarin.11–13 For patients receiving warfarin, we considered minor oral surgical procedures safe at an INR up to 3.5.13
Bleeding complications follow-up.
We evaluated bleeding complications after invasive dental treatment in all patients. We did so by reviewing whether patients returned to our dental clinic or emergency department (ED) with a complaint of bleeding. We conducted an electronic search for ED patient visits that occurred within two weeks of the dental procedure. We reviewed the medical records of patients who returned to the ED to determine the reason for their visits.
Statistical analysis.
We determined descriptive statistics, including mean, standard deviation, number of patients and percentages. We used t test or the Wilcoxon rank sum test for continuous variables and
METHODS
TOP
ABSTRACT
METHODS
RESULTS
DISCUSSION
CONCLUSIONS
REFERENCES
In this prospective cohort study, we used the CoaguChek system (Roche Diagnostics, Indianapolis) to obtain in-office INR values for patients thought to be at risk of experiencing bleeding complications after a planned dental procedure. Patients were scheduled for dental care in the dental clinic at a large academic medical center (Carolinas Medical Center, Charlotte, N.C.). Members of the clinical staff—including dental assistants, nurses or dental hygienists—performed the INR test if a patient had one of the following conditions: receiving warfarin without having completed an INR test in the 48 hours preceding the dental visit; history of viral hepatitis with a self-report of abnormal liver function test values; history of excessive bleeding after an invasive procedure; current or history of heavy alcohol use (more than 20 drinks per week for more than two years)10; clinical signs of jaundice, ascites or encephalopathy; or experiencing significant bleeding after undergoing an invasive dental procedure. We collected demographic data, social and medical history, information on concomitantly used medications reported to interfere with coagulation (such as aspirin, clopidogrel, cilostazol, nonsteroidal anti-inflammatory medications and selective serotonin reuptake inhibitors), and clinical signs and/or symptoms of liver disease. This study was approved by the medical center’s institutional review board. 
2 or the Fisher exact test for dichotomous variables. We made comparisons between out-of-range (that is, higher than 3.5 for warfarin and 1.4 or higher for liver disease) versus within-range INR values. We considered significance to be P < .05. We completed analyses by using statistical software (SAS, Version 9.1, SAS, Cary, N.C.).
RESULTS
TOP
ABSTRACT
METHODS
RESULTS
DISCUSSION
CONCLUSIONS
REFERENCES
We completed an in-office INR test for 100 consecutive patients who met the inclusion criteria from June 2004 to October 2005. These included 66 patients (mean age 55.6 ± 13.4 years, 27 [41 percent] male) receiving warfarin without having completed an INR test in the 48 hours preceding the dental visit and 34 patients (mean age 49.5 ± 8.7 years, 23 [68 percent] male) determined to be at risk of experiencing coagulopathy associated with liver disease (Table 1
).
|
). We also found out-of-range INR values in seven of 34 (21 percent) patients suspected of having liver disease (Table 1
). Several factors were significantly more common in patients with an out-of-range INR compared with safe INR, including older age; self-report of abnormal liver laboratory values; and clinical evidence of jaundice, ascites and encephalopathy (Table 3). More patients in the normal INR group reported current or past alcohol abuse.
|
|
, page 701). The other three patients receiving warfarin (range of INR values 3.6–4.0) underwent a dental procedure without experiencing bleeding complications. Dental treatment was deferred for one of seven patients with liver disease (out-of-range INR values 1.5–2.5); the remaining six were treated with extractions, scaling or a dental restoration (Table 4). Two patients with liver disease required alterations in dental care; one underwent fewer extractions than originally planned and the other received a small composite restoration instead of a dental cleaning. The hygienist discontinued a scaling procedure on another patient because of heavy bleeding.
|
Of interest and of clinical significance was our identification of 30 (45 percent) of 66 patients receiving warfarin whose INR values were below their therapeutic target level (that is, lower than 2.0). Three of the patients with a subtherapeutic INR had been asked by their physicians to stop taking warfarin before undergoing the dental procedure, and one patient had stopped taking warfarin of his or her own accord. Therefore, only 25 (38 percent) of 66 patients being treated with warfarin maintained an INR value within the therapeutic range of 2.0 to 3.5 at the time of dental treatment.
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
We found a high incidence of elevated INRs in this cohort of dental patients who visited our hospital-based dental clinic. Our routine clinical practice is to obtain an INR test for patients receiving warfarin whose INR values are outdated (and therefore unreliable), and it is based on the potential for INR fluctuations resulting from numerous factors such as changes in warfarin dose, drug and diet interactions with warfarin and poor patient compliance. There were no differences in patient reports of changes in warfarin dose or new medications since the last INR testing between the group receiving warfarin who had out-of-range INR values and the group with normal INR values. In addition, there was no difference in the previous INR values between these two groups. These findings may point to the limitations in these parameters in predicting whether a patient may have an out-of-range INR. Because of the high percentage of missing data regarding changes in warfarin dose and new medications since the last INR test, more data are required to confirm these findings. Data were missing primarily because of the patient’s uncertainty as to whether there had been changes in warfarin dose or if new medications had been added since the last INR testing.
Other investigators have identified a high prevalence of out-of-range INR values in a large cohort of patients seeking treatment in the ED. In a prospective, observational trial of patients receiving warfarin therapy with a self-report of compliance, 21 percent of patients had an INR value higher than 3.5.14 This finding is similar to ours of 17 percent, and it points to the importance of obtaining updated INR values, even for patients who report compliance with warfarin therapy.
Several variables were associated with an elevated INR in patients whose medical history suggested the potential for liver disease. For example, patients who said they had previous abnormal liver laboratory values were more likely to have an out-of-range INR. Additionally, the presence of clinical ascites, jaundice or encephalopathy was associated with an elevated INR. Similar findings of an association with clinical parameters of liver disease and oral bleeding as a result of dental procedures has been reported.15 Therefore, a patient’s self-report of abnormal laboratory test results or the presence of clinical symptoms of liver disease may be helpful in identifying patients with a risk of experiencing bleeding and who therefore require further laboratory assessment before receiving dental care. Although the INR may be the most useful sole laboratory measure of the risk for bleeding from liver disease, alone it may not provide a complete assessment of risk, and there are many drugs and diseases that also can predispose a patient to a coagulopathy.
Discussion with the patient’s physician may be useful, as may performing a more limited dental procedure initially. For example, scaling only one sextant or quadrant of teeth, or removing only one of several teeth planned for extraction, often will give a good indication of patients’ coagulation status without the risk of more widespread bleeding. Further laboratory assessment of hepatic compromise for a patient with an elevated INR without use of warfarin may include liver function tests such as alanine transferase and aspartate transferase, partial thromboplastin, platelet count, and albumin and bilirubin levels.6
The CoaguChek device used in our study has the advantage of providing point-of-care (POC) testing. The Clinical Laboratory Improvement Amendment (CLIA) of 1988 established minimum requirements for POC quality standards.16 The CoaguChek device has received a CLIA waiver as a POC test and, as established by CLIA, requires additional quality assurance requirements such as maintaining sufficient documentation of external controls, which may add to the challenges of implementing such a POC device in a hospital clinic setting. In addition to the upfront costs of the device (roughly $1,500–$1,800), the individual test strips cost between $3 and $5 (S. Cannon, Roche Diagnostics, oral communication, April 29, 2008). Recently, our clinic switched to the CoaguChek XS device, which runs an internal control with every test strip, thus eliminating the need for external controls. If the control does not pass, the meter will not provide a result for that strip.
A surprising and clinically significant finding in this study was that almost one-half of the patients studied were at risk of experiencing a thromboembolic event.
A surprising and clinically significant finding in this study was the high percentage of patients (45 percent) who had subtherapeutic INR values (that is, less than 2.0). Therefore, almost one-half of the patients studied were at risk of experiencing a thromboembolic event.12 A similar high prevalence (43 percent) was reported in a cohort of patients receiving warfarin who sought care in an ED.14
Because of the small cohort in our study, multivariate regression analysis was not possible. Future studies that examine potential risk factors evident from the medical and social history, and from the clinical examination, would be of value in identifying patients with the greatest risk of significant liver disease and those who should undergo an INR test before undergoing an invasive dental procedure. One of our criteria for ordering an INR test was a past or current consumption of more than 20 drinks of alcohol per week for more than two years. There were few patients with a history of alcohol abuse in the out-of-range INR group, so we found that this question was not helpful in suggesting the presence of an elevated INR.
The ability to generalize our findings may be limited, given that the patients had varying degrees of medical complexity and that they sought treatment at a hospital-based outpatient dental clinic at a large tertiary hospital that provides care for patients before and after liver transplantation. This likely accounts for a higher prevalence of liver disease in the patient population in this study compared with the patient population in a nonhospital–based dental practice. However, the population of patients receiving warfarin in our study is likely to be similar to that of similar patients seen in community-based dental practices. Larger studies of different patient cohorts would answer this question. Another limitation of our study was an October 2006 recall of CoaguChek test strips by the U.S. Food and Drug Administration (FDA). This recall took place during the course of our study; the FDA recommended duplicate testing for abnormal results, which we already had been doing routinely for INR values considered out of range.
The fact that none of the patients in the warfarin group in our study had bleeding episodes after dental treatment suggests that discontinuation of warfarin therapy for a patient with an INR below 4.0 is not recommended for minor surgical procedures.3,13,17,18 Finally, our data suggest that dentists should not rely on a patient’s previous INR test result as a reliable predictor of his or her current INR and that a new INR be obtained 24 to 48 hours before the patient undergoes an invasive dental procedure.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. A. Gallis Quality Improvement and Continuing Education in Dental Practice J Am Dent Assoc, June 1, 2008; 139(6): 656 - 659. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
