Don't miss a digital issue! Renew/subscribe for FREE today.
June 2018
Volume 39, Issue 6

Tranexamic Acid as a Hemostatic Adjunct in Dentistry

Riccardo I. Ambrogio, DMD; and Marci H. Levine, DMD, MD


Clinicians often perform dental procedures and dentoalveolar surgery on patients who take prescribed anticoagulant medications. Such medications can present clinical challenges with regard to modifying or temporarily discontinuing a patient’s anticoagulant regimen, which may not be medically indicated. In many instances discontinuation or altering of anticoagulation can be avoided through use of local hemostatic factors while patients are therapeutically continued on their prescribed warfarin doses. This article discusses a useful adjunct to dental treatment for the warfarin-anticoagulated patient—tranexamic acid (TA). The purpose of TA in these dental patients is to promote safe surgical site hemostasis for improved intraoperative visibility and postoperative hemostasis.

As patient care continues to progress, dentists often are treating an older patient population that is more medically complex than previously seen. Many practitioners perform dental procedures and dentoalveolar surgery on patients who take prescribed anticoagulant/antiplatelet medications (Table 1). These medications may be used to prevent emboli in patients with chronic atrial fibrillation, prevent thrombus formation in cardiac stents and valves, or prevent recurrent thrombus in patients with deep venous thrombosis. Anticoagulants provide clinical challenges to the dental community because holding them may increase thromboembolic events, while treating patients without discontinuing them may present problems of uncontrollable bleeding during and after dental procedures.

In private and clinical academic practices, medical/dental practitioners continue to advise their anticoagulated patients to modify or temporarily discontinue their anticoagulant regimens prior to dental procedures, especially dentoalveolar surgery.1,2 Lim et al1 found in a survey that warfarin was routinely discontinued by 70% of hematologists and 37% of oral surgeons for dental extractions and surgical dental procedures. This regimen is based on risk of procedure-related, perioperative and postoperative bleeding versus the risk of thromboembolic phenomenon occurrence or recurrence.3 A 1996 survey by Wahl noted that 56% of physicians indicated that they routinely discontinued anticoagulants in patients undergoing a single dental extraction.4

The aim of this article is to remind dental practitioners that in many instances these alterations/interruptions in anticoagulation are not medically indicated and, in fact, may add additional health risk to already medically compromised patients.5 Such risks often include heart attack, stroke, deep venous thrombosis, cardiovascular graft occlusion, and increased morbidity to pre-existing medical conditions.

Wahl examined patients who underwent dental extractions while anticoagulation therapy was temporarily discontinued and noted that although the incidence of complications was low (1%), those who had an adverse event tended to have more severe complications, such as stroke, heart attack, and death.6 Yasaka prospectively evaluated 412 ischemic stroke patients and concluded that among patients treated with warfarin who presented with stroke, those whose anticoagulation had been interrupted had a higher likelihood of having a cardioembolic stroke and being disabled by their stroke.7

Bacci et al suggested that discontinuing specific anticoagulants, eg, warfarin, is not recommended,8 which is consistent with the opinion of others such as Febbo et al.9 Morimoto evaluated 270 patients on anticoagulants who underwent extractions without adjusting their anticoagulation medications.10 Patients were treated with oxidized cellulose and suturing. Postoperative hemorrhage occurred in 3.6% of patients, and the study concluded that sufficient hemostasis is obtained in most cases of tooth extraction (warfarin related, INR <3). Bacci noted similar results comparing dental extraction patients on warfarin anticoagulation compared to control (non-anticoagulated patients).8 INR values ranged from 1.8 to <4 in 451 patients on warfarin, with seven bleeding complications occurring. The control group (not taking warfarin) of 449 subjects noted four bleeding complications.

In many instances discontinuation or altering of anticoagulation can be avoided through use of local hemostatic factors while patients are therapeutically continued on their prescribed warfarin doses.

This article aims to discuss a valuable adjunct to dental treatment for the warfarin-anticoagulated patient, namely tranexamic acid (TA). TA is a synthetic derivative of the amino acid lysine. It has antifibrinolytic effects by acting on lysine bonding sites on plasminogen molecules, and the binding site blockade is reversible. TA is soluble in water and has a pH ranging from 6.5 to 8. It is also available in pill form as a 500-mg tablet, and for injection. As a dental adjunct specifically for the anticoagulated patient, TA is most often used in oral solution and as a topical agent. TA solution has varying expiration. As a medical hemostatic agent it is available in several forms throughout a given procedure, including pill form taken as a preoperative medicament, an intraoperative irrigant solution during surgery, and an intravenous hemostatic agent for use during surgery.

Various modalities of TA often are used for the management of dental patients who are not on anticoagulation therapy. One example is its use intravenously prior to treatment of exodontia patients with hemophilia.11 Another example of using intravenous TA is during orthognathic surgery12 in patients with normal coagulation in an attempt to decrease intraoperative blood losses.

As a dental adjunct for the anticoagulated patient, 5% TA solution can be used in a variety of forms, such as a preoperative mouth rinse (prior to surgery),13 an intraoperative irrigant solution,13,14 a TA pre-packaged cellulose matrix packing or as a soaking/wetting solution for dental packing material and gauze,14,15 and, lastly, a postoperative mouth rinse.13,14

Use of Tranexamic Acid as a Dental Adjunct

TA use (either intravenously or orally by pill form) has been documented in patients with bleeding disorders. Note that the authors do not recommend that either the intravenous or pill form of TA be used for dentoalveolar patients who are on prescribed anticoagulation therapy. The purpose of TA in the dental anticoagulated patient is to promote safe local surgical site hemostasis for improved intraoperative visibility and postoperative hemostasis, not generalized systemic coagulation. Derived resorbable cellulose packing materials such as Surgicel® (Ethicon, or absorbable gelatin compressed sponges such as Surgifoam® (Ethicon) or Gelfoam® (Pfizer, can be used as vehicles for the local delivery of TA in solution.

Therapeutic patients on warfarin therapy with INR values of <4 have been used as a parameter for dental intervention (along with local measures as described below). INR levels >4 are considered above the range of therapeutic anticoagulation and should be managed by the patient's physician prior to any dental interventions.9 It is important to remember that there are medical patients who require INR values to be approximately 3.5 (such as patients with prosthetic heart valves and those who take anticoagulant drugs to prevent recurrent myocardial infarction).16,17

Locally infiltrated local anesthetic with a vasoconstrictor (such as 2% lidocaine with 1:100,000 epinephrine) should be considered when dealing with anticoagulated patients. This may necessitate consultation with medical providers during the treatment-planning phase.

Patients with INR values of <4 are advised not to alter their warfarin regimen, and preoperative medical consultation may be necessary. Key elements in the medical consultation may include medical clearance for proposed dental treatment and cardiac clearance for the judicious use of epinephrine in the local anesthetic. One could appreciate that a patient who is on prescribed anticoagulation therapy for deep venous thrombosis therapy may tolerate, in a safer manner, a local anesthetic with epinephrine, whereas a patient with crippling cardiac disease may be highly susceptible to it. When the amount of exogenous catecholamine is a medical issue (such as when the managing cardiologist requests minimal amounts of epinephrine in the local anesthetic to be used), the authors' recommendation is for the safe use of block local anesthesia (such as a mandibular nerve block or posterior alveolar superior nerve block) performed with 3% mepivacaine (without epinephrine or "plain") and followed by local infiltration in the area with a combination of 3% mepivacaine and 0.5% bupivacaine with a lower amount of epinephrine such as 1:200,000 epinephrine rather than 1:100,000. This regimen limits the amount of exogenous catecholamine absorbed while still providing for the hemostatic benefits of the local anesthesia both intraoperatively and postoperatively.

Increasing INR values does not necessarily correlate with an expected increase in postoperative bleeding. In a study by Morimoto, patients on warfarin monotherapy undergoing exodontia experienced more postoperative bleeding when in the INR range of 2 to 2.49 than in the INR group that ranged from 2.5 to 2.99.10 Blinder et al published similar findings.18 These results suggest that increasing INR values did not directly correlate with increased rates of postoperative bleeding. Soares et al also reported that postoperative bleeding was not directly dependent on patients' INR values.19

As a point of interest, it is worth noting that some US dental schools do not allow students to give bilateral inferior alveolar nerve blocks to patients who are on anticoagulant medications. The concern is for the potential for bilateral mandibular hematoma formation and potential airway issues. The authors have not found any reports in the literature to support this concern.

Suturing wounds in conjunction with the use of topical TA is the technique frequently presented in the scientific literature.8,10 Suturing infected mucosa and tissue is to be avoided, and often times the wound attains adequate hemostasis with direct pressure on gauze soaked with TA solution. It may take longer to attain hemostasis, but suturing infected tissue may interfere with wound healing and resolution of the primary infection.

Forms of Tranexamic Acid

TA as a Mouthwash

According to the literature, 5% TA solution can be used preoperatively as a mouth rinse prior to the patient undergoing a dental procedure. Some studies have suggested that patients rinse for 2 minutes with 10 ml of 5% TA solution before dental intervention, and then rinse post procedure up to four times a day.13 Earlier studies had suggested rinsing for longer periods, such as between 5 and 7 days postoperatively, but more recent studies have concluded that rinsing for two consecutive postoperative days (and still rinsing four times per day) produced similar outcomes and was sufficient for obtaining adequate hemostasis.9,13,14,20

TA as a Topical Agent

5% TA solution can be used topically on a surgical site. During dentoalveolar surgery, the surgical site can be irrigated with the solution just prior to suture closure. In cases of dental surgery, after the site is cleansed, debrided, and irrigated with saline, the wound is then irrigated with TA prior to coaptation of the mucosa. It has been recommended that surgical sites be irrigated with 1 or 2 milliliters of TA solution. If resorbable cellulose packing were to be used, it would be placed at this time. Surgicel has been used in this manner.20 If amenable, the packing material should be pre-soaked with TA before placement in the surgical site. The patient would then apply pressure to the area for 30 minutes with gauze moistened with 5% TA to assist with further hemostasis. This would be used in the same manner as a dental surgery patient would use saline-moistened gauze by gently applying steady pressure on the gauze to assist in achieving hemostasis. Surgicel or an absorbable gelatin compressed sponge, eg, Gelfoam or Surgifoam, also can be placed in/on the area, and then the patient should be instructed to use the TA-moistened gauze.

The authors do not advocate the use of TA solution where there is an exposed nerve, eg, inferior alveolar or mental nerve, as they do not have experience with its effect on exposed nerve tissue. After dental intervention has been completed, postoperative instructions are reviewed with the patient verbally and in writing. If indicated, eg, after tooth extraction, the patient should be appropriately instructed to moisten gauze with TA solution for use at home in the event of further oozing or bleeding. Most patients tend to discontinue its use after a few gauze changes (and most patients complete this step while they are still in the office after surgery).

TA as an Intraoperative Irrigant

When used as an intraoperative irrigant, TA solution has clearly been shown to significantly reduce blood loss. It can be used in patients who are anticoagulated on warfarin therapy. It has also been shown to be useful for patients who are not anticoagulated and are undergoing major surgery, such as jaw repositioning or orthognathic surgery.21 A recent article by Eftekharian demonstrated the use of tranexamic solution (1 mg/ml in 0.9% saline solution) for orthognathic surgery.22 The paper reported a statistically significant reduction in overall blood loss and improved visibility of the surgical field, because there was less bleeding despite a prolonged surgical time.

TA as a Pill

TA is readily available in a 500-mg pill formulation. Patients are advised to take 1 to 1.5 grams two to three times per day for local fibrinolysis.23 In patients with hemophilia, this regimen can also be used in conjunction with 5% tranexamic solution when employed as a postoperative adjunct. TA pills are not recommended for use in cases of dentoalveolar surgery on patients who are medically anticoagulated.

TA as an Intravenous Agent

For local fibrinolysis, eg, patients with hemophilia, the recommended dose of TA is 500 mg to 1 g delivered by slow intravenous injection three times per day. Pediatric dosages should be calculated and adjusted for body weight (with a reference dose as 25 mg/kg/dose). This regimen may best be used in conjunction with input from the patient's treating physician, and patients should be properly counseled preoperatively in the presence of risk factors for future thromboembolic events, eg, obesity, use of birth control, issues with immobility, etc.

Advantages and Disadvantages of Tranexamic Acid

Bleeding is common with any surgical procedure. Experienced surgeons who employ careful and atraumatic techniques in the healthiest of patients will still have some degree of bleeding with surgery. Good technique such as the use of properly placed incisions will aid in minimizing overall blood loss. The same principles hold true for soft-tissue surgery.

The use of TA as an adjunct to dental treatment is slowly gaining popularity among dentists and dental specialists who perform invasive procedures. However, the initial hurdle for practitioners is often related to obtaining the medication. It is not routinely stocked in most local pharmacies. Practitioners who use TA often rely on a compounding pharmacy or local hospital pharmacy to prepare the solution. Once the availability barrier has been resolved, the authors believe that dental practitioners will encounter many applications for TA. With a prescription, a local compounding company can prepare a 5% TA solution by properly combining sterile water, 1% benzalkoniun chloride, and TA in advance of surgery. One of the authors (RIA) frequently prescribes 30 ml to 50 ml of 5% TA solution for use after exodontia limited to several teeth or other minor preprosthetic surgical procedures. The patient is instructed to bring the solution to his or her dental appointment. The shelf life of the solution (depending on how the solution was manufactured) is variable. When ordering it, practitioners should request that the compounding pharmacist label the prescription bottle with an appropriate expiration date.

Patients who have underlying renal insufficiency and/or thromboembolic disease may have adverse side effects with TA. The use of TA in these patients is subject to a comprehensive medical assessment that includes the evaluation of factors relating to its risks and benefits. Topical oral use of TA does not produce high plasma levels. Rinsing with 5% aqueous solutions of TA results in plasma drug concentrations below 2 mg/L.23 There is a report in the literature of a specific instance with an unexpected outcome related to the use of TA; a patient with hemophilia experienced a myocardial infarction after taking prophylactic TA (20 mg/kg/day, oral administration) for 10 days before a dental extraction.24 This patient also had medical comorbidities that may have contributed to the cardiac event. It is unclear if the TA had a direct effect.

Allergy to TA would also be another contraindication for its use.

TA does cross the placenta. In the nomenclature, which is currently outgoing, TA is categorized as an FDA pregnancy category B drug. FDA category B medications are those in which animal reproduction studies have failed to demonstrate a risk to the fetus and for which there are no adequate and well-controlled studies in pregnant women. This pregnancy category is the same as lidocaine with epinephrine, acetaminophen, and penicillin V. (As of June 29, 2018, all FDA drugs will no longer use the pregnancy category rating, as it is being replaced with the Pregnancy and Lactation Labeling Rule [PLLR]). Consult the manufacturer, as some forms of TA are contraindicated for use in pregnancy. Caution also should be used in women who are lactating, and medical consultation with a patient's treating physician is strongly recommended prior to TA use in these settings.

Case Presentation

The patient was a 67-year-old white man who was referred for extraction of tooth No. 15 (maxillary left second molar). The tooth was mastication painful, and the patient was not interested in further dental therapy for the tooth. The patient was informed of treatment options and requested extraction.

His medical history was remarkable for coronary artery disease, peripheral vascular disease, and hypertension. His surgical history noted a mechanical aortic valve replacement, right and left carotid endarterectomy, coronary artery bypass graft of three vessels, femoro-popliteal bypass, and a recent pacemaker. The patient's medications were as follows:Coumadin (warfarin, anticoagulant), 81-mg aspirin, Crestor (rosuvastatin, to reduce cholesterol), metoprolol, Protonix (pantoprazole, proton pump inhibitor), finasteride (for benign prostrate hyperplasia), lisinopril (for hypertension), and multivitamin. He was allergic to Ceclor (cefaclor, a cephalosporin antibiotic).

Clinical Examination

The partially edentulous patient was well-developed. Teeth Nos. 14 and 16 were missing. Tooth No. 15 was percussion painful and had a large amalgam restoration making up the majority of the clinical crown, as noted on periapical radiography. The tooth had normal anatomic bone height. The patient was not interested in endodontic evaluation, and after discussion with his general dentist, the tooth was planned for extraction.

Consultation with the patient's cardiologist noted that the patient was doing well since his pacemaker placement. Ejection fraction was normal, and he was stable and cleared for the use of local anesthesia with judicious epinephrine for local infiltration. His INR taken 3 days preoperatively was 2.9.

The patient was instructed to continue all medicines in routine fashion, and he presented in the morning for his extraction.

Procedure and Findings

Mechanical heart valve prophylaxis was taken 1 hour preoperatively (two grams amoxicillin by mouth). (The authors note that penicillin/beta lactam antibiotics may have showed related allergy with cephalosporin antibiotics. The patient was using amoxicillin heart valve prophylaxis routinely prior to being their patient, and this was his prophylactic recommendation advised by his cardiologist.) Preoperative blood pressure was 140/80, pulse 68. Local anesthesia was administered with a 27-guage long needle. Mepivacaine 108 mg (two dental carpules of 3% mepivacaine) was administered via posterior superior alveolar block and buccal and palatal infiltration. Marcaine 9 mg with 0.009 mg epinephrine (one dental carpule of 0.5% marcaine with 1:200,000 epinephrine) was infiltrated locally, in the buccal and palatal mucosa.

After waiting approximately 5 minutes for the local anesthetic to take effect, the crown of tooth No. 15 was sectioned with a medium fissure bur at the buccal aspect, separating the mesial-buccal and distal-buccal roots from the crown. This sectioning was performed in a horizontal fashion at the dentoenamel junction of the tooth, going from mesial to distal, with the fissure bur attaining a depth of approximately 4 mm. The palatal root was not encroached, leaving it still intact with the clinical crown of the tooth. A 150 universal forcep was used to grasp the tooth crown and rotate the crown (and attached palatal root) in a "corkscrew" motion. This allowed the crown and palatal root to be delivered atraumatically. Next, a thin fissure bur was used to separate the mesio-buccal root from the distal-buccal root at the furca. The 150 forcep and thin straight elevator were then used to atraumatically remove the two remaining roots.

Normal saline was used to irrigate the socket. Visual inspection noted no granulation tissue for debridement. One ml of 5% TA was used to irrigate the socket, followed by three small pieces of Surgifoam placed in the socket. One piece, approximately 1 cm by 5 mm, was compressed into the shape of a wick and placed in each root socket. A 1-inch thick mat of 4x4 sterile gauze was then moistened with approximately 10 ml of TA, and this gauze was used, one piece at a time, for direct compression on the wound. The patient was given one piece of this gauze, which was folded into the shape of a tea bag. It was directly placed on the wound, and the patient was instructed to gently bite on the gauze. After 10 minutes, the gauze was replaced a second time.

Adequate hemostasis was attained after 10 minutes, and the use of gauze was discontinued. Total blood loss was minimal. Postoperative blood pressure was 138/70, pulse 64. The patient was given sterile gauze to take home and instructed to moisten it with the remaining TA if bleeding recurred. The patient had an unremarkable postoperative experience and was seen for follow-up 1 week later, with the wound healing well and the patient without complaints.


TA solution is an adjunct for hemostasis in dental surgery, especially for patients on anticoagulation therapy such as warfarin or dicoumarol. Once the initial challenge of finding a pharmacy or compounding company to prepare the medicament is met, use of TA in patient care is a relatively easy practice standard.

About the Authors

Riccardo I. Ambrogio, DMD

Clinical Assistant Professor, Departments of Oral and Maxillofacial Surgery,
New York University College of Dentistry, New York, New York, University of Connecticut School of Dental Medicine, Farmington, Connecticut; Private Practice, Wethersfield, Connecticut

Marci H. Levine, DMD, MD

Clinical Assistant Professor, Oral and Maxillofacial Surgery, New York University College of Dentistry, New York, New York; Private Practice, NYU Dental Faculty Practice, New York, New York


1. Lim W, Wang M, Crowther M, Douketis J. The management of anticoagulated patients requiring dental extraction: a cross-sectional survey of oral and maxillofacial surgeons and hematologists. J Thromb Haemost. 2007;5(10):2157-2159.

2. Pototski M, Amenábar JM. Dental management of patients receiving anticoagulation or antiplatelet treatment. J Oral Sci. 2007;49(4):253-258.

3. Levine MN, Raskob G, Landefeld S, Hirsh J. Hemorrhagic complications of anticoagulant treatment. Chest. 1995;108(4 suppl):276S-290S.

4. Wahl MJ, Howell J. Altering anticoagulation therapy: a survey of physicians. J Am Dent Assoc. 1996;127(5):625-634.

5. Wahl MJ, Pinto A, Kilham J, Lalla RV. Dental surgery in anticoagulated patients-stop the interruption. Oral Surg Oral Med Oral Pathol Oral Radiol. 2015;119(2):136-157.

6. Wahl MJ. Dental surgery in anticoagulated patients. Arch Intern Med. 1998;158(15):1610-1616.

7. Yasaka M, Naritomi H, Minematsu K. Ischemic stroke associated with brief cessation of warfarin. Thromb Res. 2006;118(2):290-293.

8. Bacci C, Maglione M, Favero L, et al. Management of dental extraction in patients undergoing anticoagulant treatment. Results from a large, multicentre, prospective, case-control study. Thromb Haemost. 2010;104(5):972-975.

9. Febbo A, Cheng A, Stein B, et al. Postoperative bleeding following dental extractions in patients anticoagulated with warfarin. J Oral Maxillofac Surg. 2016;74(8):1518-1523.

10. Morimoto Y, Niwa H, Minematsu K. Hemostatic management of tooth extractions in patients on oral antithrombotic therapy. J Oral Maxillofac Surg. 2008;66(1):51-57.

11. Zanon E, Martinelli F, Bacci C, et al. Proposal of a standard approach to dental extraction in hemophilia patients. A case-control study with good results. Haemophilia. 2000;6(5):533-536.

12. Song G, Yang P, Hu J, et al. The effect of tranexamic acid on blood loss in orthognathic surgery: a meta-analysis of randomized controlled trials. Oral Surg Oral Med Oral Pathol Oral Radiol.

13. Borea G, Montebagnoli L, Capuzzi P, Magelli C. Tranexamic acid as a mouthwash in anticoagulated-treated patients undergoing oral surgery. An alternative method to discontinuing anticoagulant therapy. Oral Surg Oral Med Oral Pathol. 1993;75(1):29-31.

14. Carter G, Goss A. Tranexamic acid mouthwash-a prospective randomized study of a 2-day regimen vs. 5-day regimen to prevent postoperative bleeding in anticoagulated patients requiring dental extractions. Int J Oral Maxillofac Surg. 2003;32(5):504-507.

15. Sammartino G, Marenzi G, Miro A, et al. Local delivery of the hemostatic agent tranexamic acid in chronically anticoagulated patients. J Craniofac Surg. 2012;23(6):e648-e652.

16. Aframian D, Lalla RV, Peterson DE. Management of dental patients taking common hemostasis-altering medications. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2007;103(suppl):S45.e1-e11.

17. Hirsh J, Dalen JE, Deykin D, et al. Oral anticoagulants: mechanism of action, clinical effectiveness, and optimal therapeutic range. Chest. 1995;108(4 suppl):231S-246S.

18. Blinder D, Manor Y, Martinowitz U, Taicher S. Dental extractions in patients maintained on oral anticoagulant therapy: comparison of INR value with occurrence of postoperative bleeding. Int J Oral Maxillofac Surg. 2001;30(6):518-521.

19. Soares EC, Costa FW, Bezerra TP, et al. Postoperative hemostatic efficacy of gauze soaked in tranexamic acid, fibrin sponge and dry gauze compression following dental extractions in anticoagulated patients with cardiovascular disease: a prospective, randomized study. J Oral Macillofac Surg. 2015;19(2):209-216.

20. Hewson I, Makhmalbaf P, Street A, et al. Dental surgery with minimal factor support in the inherited bleeding disorder population at the Alfred Hospital. Haemophilia. 2011;17(1):e185-e188.

21. Kaewpradub P, Apipan B, Rummasak D. Does tranexamic acid in an irrigating fluid reduce intraoperative blood loss in orthognathic surgery? A double-blind, randomized clinical trial. J Oral Maxillofac Surg. 2011;69(6):e186-e189.

22. Eftekharian H, Vahedi R, Karagab T, Tabrizi R. Effect of tranexamic acid irrigation on perioperative blood loss during orthognathic surgery: a double-blind, randomized controlled clinical trial. J Oral Maxillofac Surg. 2015;73(1):129-133.

23. Dunn CJ, Goa KL. Tranexamic acid: a review of its use in surgery and other indications. Drugs. 1999;57(6):1005-1032.

24. Günaldi M, Helvaci A, Yildirim ND, et al. Acute myocardial infarction in a patient with hemophilia A and factor V Leiden mutation. Cardiol J. 2009;16(5):458-461.

© 2024 BroadcastMed LLC | Privacy Policy