A study to assess management of patients on warfarin by general dental practitioners (GDPS) in the west midlands

Materials and methods

A questionnaire (Figure 1) was sent to 638 GDPs in the West Midlands in February 2010. Details of the GDPs in the West Midlands area were manually searched using the online General Dental Council database of registered dental practitioners, using regional postcodes as the search criteria. All GDPs from the search were sent the questionnaires to eliminate any potential bias. Clinicians whose practice was limited to orthodontics and hospital dental services were excluded. No incentives to complete the questionnaire were offered other than a stamped addressed envelope to return the questionnaire. Dental practitioners were asked to reply anonymously.

The questionnaire itself was based on one used in a previous study with some additions.20 Additional questions included:

Results

Replies to questions

GDP's position in practice

One hundred and ninety-eight (42%) of the respondents were principal GDPs, 199 (42%) were associates and 24 (5%) were Vocational Dental Practitioners (VDPs). Of the remaining 50 respondents, 39 (8%) were either locums or not known and 11 (2%) were assistants.

Numbers of years since qualification?

Out of 471 respondents, the majority 260 (55%) had been qualified for 5–10 years, 97 (20%) GDPs had been qualified for 10–15 years, 48 (10%) had been qualified for 1–5 years, 41 (8%) were less than a year qualified and the remaining 25 respondents (5%) had been qualified for more than 15 years.

What is your main income source?

A total of 224 GDPs (47.5%) were working mainly in NHS practice, 172 (36.5%) of them worked in a mixed NHS and private practice and the remaining 75 (16%) of the respondents worked mainly in the private sector.

Do you treat patients on warfarin?

A total of 438 GDPs (93%) treated patients on warfarin and the remaining 33 (7%) did not. Reasons provided for not treating patients on warfarin included: ‘time consuming’ (2.5%), ‘difficulty in co-ordinating INR prior to treatment’ (2%), ‘not sure about the regulations’ (1.5%), ‘potential medico-legal problems’ (0.5%) and ‘no support from GP/cardiologist’ (0.5%).

Opinion on procedures and perceived risk of bleeding (Table 1)

Table 1. GDPs' responses to ‘risk of bleeding’ assessment for clinical procedures.

Procedure	High Risk No. of GDPs (%)	Low risk No. of GDPs (%)	Not noted No. of GDPs (%)
Sub-gingival debridement	49%	43%	8%
Supra-gingival scaling	4%	86%	10%
Periodontal examination	4.8%	93.6%	1.4%
Inferior dental block (local anaesthetic)	39.7%	48.4%	9.7%
Infiltration (local anaesthetic)	5%	91.7%	3%
Pulp extirpation and RCT	8%	82.3%	9.5%
preparations RCT obturation	1.4%	87.8%	10.6%
Implant 1 stage procedure (surgical)	74.7%	4%	21.2%
Implant abutment connection	14%	67%	18.8%
Extraction	86%	5.3%	8.4%

When asked their opinion on the potential risk of bleeding associated with various dental procedures, the majority of respondents (93.6%) considered periodontal examination as low. Seventy four percent considered surgical implant placement as high risk and 86% considered dental extraction as high risk.

Management of anticoagulation

Advice to patients when carrying out a procedure considered high risk for bleeding (Figure 2). When planning a procedure considered high risk for bleeding on a patient taking warfarin, a total of 181 (38%) GDPs said that they would liaise with the GMP or cardiologist before treating the patient. Only 9 (2%) respondents indicated that they would reduce the dose of warfarin before performing a high-risk procedure and 6 (1%) said that they would ask the patient to cease their warfarin prior to a high risk procedure without seeking a medical opinion.

INR assessment prior to treatment (Figure 3) In response to a question asking if or how the INR was assessed prior to dental treatment, the majority (88%) of GDPs would ensure that the patient had an INR test performed prior to any dental treatment. Out of this 88%, 52% ensured that the patient had an INR test performed within 24 hours prior to the procedure. Only 5% of GDPs would not check the INR prior to a procedure considered high risk for bleeding.

Safe upper limit of INR to carry out ‘high risk of bleeding’ procedures (Figure 4) With regard to the safe upper limit for the INR only 117 (25%) GDPs suggested that an INR of 4.0 was the safe upper limit when undertaking a procedure associated with a high risk of bleeding. With respect to an INR of 3 or 3.5, 138(29%) and 98(21%) respondents, respectively, considered these levels as safe upper limits.

Factors influencing the INR (Figure 5) A total of 93% (n = 438) and 68% (n = 320) of GDPs highlighted that drug interaction and alcohol, respectively, would influence the patient's INR. Twenty-five percent of GDPs indicated stress and anxiety affected the INR and 22% thought dietary habits of patients could influence the INR.

Discussion

This study set out to ascertain how well the management of anticoagulated patients by GDPs in the West Midlands is currently practised and to evaluate how it correlates with the standard guidelines. A similar study in South West Wales identified that differences between published guidelines for the management of anticoagulated patients was one of the reasons that GDPs were poorly compliant with national standards. Since the Welsh study was undertaken, the British National Formulary22 has updated its guidance in line with that produced by the North West Medicines Information Centre (NWMIC).15

Since 2001, the North West Medicines Information Centre (NMWIC) has published guidance on the surgical management of patients on warfarin in primary care. UK Medicines Information is a pharmacist-led service that supports the safe, effective and efficient use of medicines by the provision of information and advice.20 The guidance is supported by various medical bodies. A working group of medical professionals and representatives from the BCSH (British Committee for Standards in Haematology), NPSA (National Patient Safety Agency), British Society for Haematology (BSH) and British Dental Association (BDA) produced a set of guidelines and the essential recommendations have been summarized by Perry et al.18 The BNF is a joint publication of the British Medical Association and the Royal Pharmaceutical Society of Great Britain and includes guidance in relation to the treatment of patients taking warfarin.22

Various sources of information which discuss the current evidence about this topic are available to dentists. An overwhelming amount of the evidence supports the conclusion that most dental or minor oral surgery procedures can be carried out safely on patients with a stable INR of below 4.0.11^,12^,13^,14^,15^,17^,23^,24^,25

In comparison with the Welsh study, a similar number of GDPs treated patients on warfarin in primary care (93% in this survey and 92% Wales). The practitioners who did not treat patients on warfarin provided reasons which included: ‘time consuming’, ‘difficulty in co-ordinating INR prior to treatment’, ‘unsure about the regulations’ and ‘potential medico-legal problems’. The lack of difference between this and the previous study suggests that there still appears to be a group of practitioners who are reluctant to treat this group of patients. This may suggest a lack of awareness of current recommendations. However, it may suggest that, despite the evidence supporting the guidelines, some practitioners lack confidence in managing patients on warfarin when carrying out procedures considered high risk for bleeding, even though it has shown that bleeding can be clinically controllable with simple local measures in the majority of cases.3^,7^,13^,15^,21

Even in patients not receiving anticoagulant therapy, the incidence of post-operative bleeding is cited as being between 0% and 3.8%.15 In an anticoagulated patient, the risk of post-operative bleeding requiring intervention is greater but most cases can be managed successfully by using local measures.15^,21 This has also been shown in analysis of the literature done by Wahl.13^,26 Stopping warfarin provides no guarantee that the risk of bleeding will be eliminated because intra-oral bleeding could be excessive even in non-anticoagulated patients as the tooth support structures are highly vascular and saliva contains constituents with fibrinolytic activity.15^,27

In a review of the literature which identified 950 anticoagulated patients undergoing dental surgery, the incidence of uncontrollable clinically significant bleeding was found to be low.26 Only 12 patients (1.3%) experienced bleeding that was uncontrollable by local measures. None of these patients was reported to have experienced serious harm. Of these patients, 7 had an INR greater than 4.0 and, of these 7 patients, 3 had been given a course of post-operative antibiotics that may have interacted with the warfarin. In a randomized controlled trial involving 57 patients, there was no increase in clinically important bleeding amongst anticoagulated patients.3 Similar incidences (1.2%) of bleeding were reported in two studies involving 260 patients who had never taken an oral anticoagulant.28^,29

Clinically significant post-operative bleeding has been defined as that which:15^,27

There is limited documented guidance on procedures which are considered to have either a high or low risk of bleeding, except the recommendations by the consensus committee consisting of the British Committee for Standards in Haematology (BCSH), British Society of Haematology (BSH), British Dental Association (BDA) and the National Patient Safety Agency (NPSA).18^,30^,31^,32^,33 The NWMIC15 recommendations focus mainly on the surgical and periodontal aspects of dental care, as these procedures are most likely to cause bleeding. The evidence suggests that minor oral surgery procedures can be carried out safely without altering warfarin dose. Up to 3 extractions in the same quadrant is associated with a low risk of bleeding. When multiple extractions of more than 3 teeth are planned, then multiple visits should be arranged. In general, the guidance makes a broad assumption that the procedures performed in primary care are unlikely to involve a high risk of haemorrhage.

There was a notable (10%) decrease in respondents who did not consider subgingival debridement a high bleeding risk procedure compared to the Welsh study respondents. This was surprising as the NWMIC guidance considered subgingival debridement as higher-risk if the tissues are inflamed.15 This discrepancy may be due to the fact that the BNF does not include risks associated with gingival manipulation as a variable in their guidelines. The BNF states that scaling (not specified whether they are referring to supra- or subgingival scaling) and root planing should be restricted to limited areas to assess for bleeding. The risk of bleeding following subgingival debridement depends on a number of factors of which the degree of inflammation is an important consideration.20^,34 Good plaque control is important in anticoagulated patients and scaling and root surface debridement should be restricted initially to limited areas (eg one quadrant) to assess if the bleeding is problematic.15

In response to the safe upper limit for the INR, only 25% of GDPs said that they considered the safe upper limit for the INR to be 4.0 for procedures associated with risk of bleeding. Twenty one percent of respondents said that they consider 3.5 to be the safe upper limit and the majority (30%) regarded 3.0 as a safe upper limit for the high risk bleeding procedures.

Awareness of 4.0 as the recommended safe upper limit for the INR was 25% in our study, a great improvement compared to the 10% reported in the Welsh study. When comparing opinions on those respondents who considered that the safe upper limit for the INR was lower than 4.0; 29% and 21% of GDPs in our study considered INRs of 3.0 and 3.5, respectively, as safe compared to 30% and 10% reported in the Welsh study. The overall trends indicate that respondents are considering higher limits of INR to be safe. However, there is still a large proportion of GDPs who are still not aware of, or not adopting, the current guidelines. Interestingly, two of the respondents commented that, whilst they are aware that the current guidelines suggest a safe upper limit of 4.0, they still would not be comfortable or confident in performing some procedures in primary care.

When should the INR be measured before a dental procedure?

The current NWMIC guidance suggests that, whilst measuring the INR within 24 hours is ideal (as it provides the most accurate assessment of the INR status), this may be difficult to achieve in primary care and the evidence suggests that an INR measure within 72 hours is acceptable.15^,22 In this study, the majority (52%) of respondents would have the INR checked within 24 hours of the procedure being carried out and 17% of the respondents would check the patient's INR within 72 hours. Checking the INR within 72 hours is clearly mentioned in the guidelines and provides an opportunity to change the warfarin dosage by consulting with the patient's physician if necessary.

Conclusion

This study provides an insight into the current views and practices adopted by GDPs in the West Midlands with regards to the management of patients taking warfarin.

Evidenced-based guidelines suggest that patients with a stable INR within the therapeutic range of between 2 and 4 ie < 4 (and with no concomitant conditions such as liver disease) should not have their warfarin interrupted prior to dental procedures, nor should they routinely be referred to secondary care for management. Unnecessary referral creates greater stress on resources and incurs a cost to secondary care but, more importantly, it creates a delay in treatment. Patients would certainly benefit if their treatment could be carried out safely, closer to their home and in a timely manner. The evidence supports that patients taking warfarin can be safely managed in primary care by following well-defined guidance, such as BNF, that provides good and very accessible guidance for GDPs and includes the use of local haemostatic measures applied to control bleeding.

This study has identified that current practice deviates significantly from evidence-based guidelines. The authors recommend that GDPs follow the guidelines (Table 2) published with regard to management of patients on warfarin in primary dental care.