Preventing Wrong Tooth Extraction

Abstract

Objective of Work

Wrong-site tooth extraction (WSTE) is the most common serious patient safety incident in dentistry. Safety checklists have significantly reduced wrong-site surgery, although their benefit is unproven in primary care dentistry. Our quality improvement project developed and implemented a checklist optimised for oral surgery procedures in primary care to reduce WSTE risk.

Material and Methods

Local best practice for tooth extraction record-keeping (LBP), using national guidelines and standards was devised. We then retrospectively audited tooth extraction record-keeping against LBP. Deficiencies in current record-keeping practice were identified and used to design a checklist aimed at improving compliance. We provided a computerised safety checklist compliant with LBP to eleven clinicians at three general dental clinics within our region. The checklist included a pre-operative safety check, a pause to re-confirm the surgical site and a post-operative record-keeping proforma. The checklist was linked to our record-keeping software for use during tooth extraction. We audited checklist completion and compliance with LBP fortnightly for ten weeks.

Results

The introduction of a safety checklist resulted in increased compliance with LBP for tooth extraction record keeping. At week ten, 67% of records contained the computerised safety checklist. This resulted in a 50% increase in overall compliance with LBP for tooth extraction compared to baseline.

Conclusions

A computerised safety checklist for tooth extraction in primary care has potential to improve patient safety by adopting measures to prevent WSTE and standardising communication between clinicians. Checklists in general practice should be encouraged.

Introduction

Although no longer a "never event", wrong-site tooth extraction (WSTE) is the most common serious patient safety incident in the National Health Service (NHS) (1). A 2016 survey of dentists and dental therapists working in Wales, UK found that 12% of clinicians had extracted the wrong tooth, demonstrating that WSTE is not a rare event. The study also found that only 25% of clinicians used a recognised checklist for tooth extraction (2).

The South Powys General Dental Service (GDS) is staffed by ten dentists and one dental therapist, with a wide range of competence in oral surgery. The majority of clinicians practice across multiple sites, with different clinicians providing a treatment plan and carrying out the required treatment, increasing the risk of WSTE (3).

The causes of WSTE are multifactorial and can be classified into active failures, such as human error, and latent failures which occur due to poor workplace and organisational conditions (3). A paper analysing malpractice claims for WSTE in Israel found most errors occurred due to miscommunication between clinicians (4). Furthermore, a Korean study identified tooth specific risk factors, including multiple teeth planned for extraction, partially erupted and grossly carious teeth (5). Ambiguous tooth notation also increases the risk of WSTE, particularly when patients are missing teeth (3). This correlates with a study of Welsh general dental practitioners which found that 50% of WSTEs were prior to orthodontics (2).

155 WSTEs were reported to NHS England between 2015-19 (1). A systematic review showed that surgical checklists are a simple strategy for improving patient safety culture and are associated with increased detection of hazards and improved communication amongst team members (6).

The vast majority of patient safety research in dentistry has taken place in secondary care. However, 95% of dental treatment in the United Kingdom is performed in primary care (7), and specialist Oral Surgeons are increasingly working from dental practices (8). Currently there is no evidence demonstrating the effectiveness of safety checklists in general dental practice (7). However, general dental practitioners are knowledgeable about patient safety and are keen to adopt evidence-based interventions to drive improvement (9).

We aimed to explore whether implementing a patient safety checklist reduced the risk of WSTE in general dental practice by defining LBP record-keeping for tooth extraction and undertaking a quality improvement project to evaluate the effect of checklist implementation.

Material and methods

Model of Improvement

Our quality improvement project utilised plan, do study, act (PDSA) cycles. The project's aim was for all clinical records for non-surgical tooth extraction in South Powys GDS to adhere to LBP by March 2021. The project's outcome measure was the percentage of records for non-surgical tooth extraction complying with LBP, with a corresponding process measure assessing the percentage of records for non-surgical tooth extraction containing the safety checklist. The NHS research ethics committee decision toolkit was used to determine that this project did not require ethical approval.

Development of Local Best Practice

LBP was designed and optimised for non-surgical tooth extraction under local anaesthetic, using the Faculty of General Dental Practice (UK) Clinical Examination and Record-Keeping: Good Practice Guidelines (10), Oral Surgery Local Safety Standards for Invasive Procedures (LocSSIP) (11) and a consensus discussion between clinicians. LBP consisted of 5 domains: LocSSIP, pre-operative details, local anaesthetic, intraoperative details and post-operative details (Table 1).

Table 1. Record keeping features assigned to each domain of local best practice record keeping for non-surgical tooth extraction.

Domain of Local Best Practice
LocSSIP	Pre-Operative	Local Anaesthetic	Intra-Operative	Post-Operative
Assistant	Medical History	LA Site	Instruments Used	Haemostasis Confirmed
Patient Details	Pre-Op Risks / Consent	LA Solution	Apices Extracted	Post-Op Instructions
Diagnosis	Patient Eaten	LA Dose
Treatment Plan		LA Batch Number
Radiographs Displayed		LA Expiry Date
Site Confirmed		Confirm Anaesthesia

Baseline Compliance

A pilot study retrospectively assessed compliance with LBP, generating baseline data and facilitating the creation of an intervention. We analysed 28 non-surgical tooth extraction records obtained from the Software of Excellence computerised record-keeping system. The patients were treated by two general dentists at one South Powys GDS clinic. A data collection sheet was designed to collect all data defined as local best practice for non-surgical tooth extraction. Clinicians were calibrated prior to collecting data.

Baseline Results

The initial study found that overall compliance with LBP was poor, with LBP achieved in only four categories. 12 of the 17 LBP features were present in over 50% of records. However, none of the records achieved complete compliance with LBP (Fig.1).

Figure 1.

Pareto chart demonstrating non-compliance in each record keeping feature associated with LBP record keeping for non-surgical tooth extraction.

Intervention

Baseline results highlighted the need for improvement in record-keeping standards for non-surgical tooth extraction. Potential interventions were mapped on an ease-benefit matrix (Fig. 2). A computerised safety checklist was deemed the most appropriate intervention due to its ease of implementation and high likelihood of improving LBP compliance.

Figure 2.

Ease-benefit matrix exploring potential interventions to increase compliance with LBP for non-surgical tooth extraction.

A computerised safety checklist corresponding with LBP was created on Software of Excellence (Fig. 3). The checklist consisted of three phases. The "sign-in" phase, completed before delivery of local anaesthetic. Sign-in is used to check the patient's identity and medical history, confirm the surgical site with the patient and dental team, ensure that relevant investigations are available and that procedural risks are discussed. The "time-out" phase is a definite pause before applying instruments to the tooth, re-confirming the surgical site before extraction. The "sign-out" phase is completed after the tooth has been extracted to identify any operative complications, ensure the patient has received post-operative instructions and that records comply with LBP.

Figure 3.

South Powys GDS record keeping checklist for non-surgical tooth extraction.

10 GDS dentists and one dental therapist consented to participating in the study.

The checklist was linked to all participants Software of Excellence accounts and instructions for its use were provided. When clinicians treatment planned an extraction, the checklist automatically became available. All non-surgical extractions completed by participating clinicians were audited against LBP using a data collection sheet.

Results

Ten dentists and one dental therapist performed seventy non-surgical tooth extractions at three GDS clinics over ten weeks.

The median percentage checklist completion over ten weeks was 75.5%. Initial uptake was low, with only 18% of records containing the safety checklist at two weeks. Checklist completion significantly increased, reaching 94% at week eight before falling to 67% at week 10 (Fig. 4). 56 of the 70 records assessed contained the safety checklist.

Figure 4.

Percentage template completion and overall compliance with LBP for non-surgical tooth extraction over ten weeks in Powys GDS.

We defined overall compliance with LBP as the presence of all record-keeping features in each of the five domains. Median compliance with LBP was 39.5%. Initial compliance with LBP was low, reaching 9% at two weeks. Compliance peaked at 77% in week four before falling to 29% at week eight and rising to 50% at week 10 (Fig. 4). When clinicians used the safety checklist, we noted weaknesses in naming the assistant for the procedure and recording local anaesthetic details. Compliance with LBP for each domain is explored below.

The LocSSIP domain of LBP consisted of six record-keeping features: Name of assistant, patient details checked, diagnosis recorded, treatment plan checked, radiographs displayed, and site confirmed with the patient and dental team. Compliance with the LocSSIP domain was 9% at two weeks. LocSSIP compliance peaked at 82% in week six and fell to 50% at week ten. Weaknesses in compliance with the LocSSIP domain were predominantly due to failures in recording patient details and displaying radiographs.

We defined compliance with the pre-operative details domain of LBP as the presence of three record-keeping features: Relevant medical history and allergies detailed, pre-operative risks and discussion of consent detailed and patient eaten. Compliance with the pre-operative details domain dropped from a baseline of 46% to 18% at week two, peaking in week six at 91% before falling to 50% at week 10. The recording of medical history (93%) and pre-operative risks/consent (90%) was strong. However, recording of whether the patient had eaten was poor, being present in 67% of records.

Compliance with the local anaesthetic (LA) domain of LBP was defined as the presence of six record-keeping features: LA site, solution, dose, batch number, expiry date and confirmation of successful anaesthesia. After introducing the safety checklist, compliance with LA local best practice was 36% at two weeks. LA compliance peaked at 85% in week four before falling to 61% at week ten. Weak LA compliance resulted from the poor recording of anaesthetic batch number and expiry dates, present in 71% and 70% of records, respectively.

We defined compliance with the intra-operative details domain of LBP as the presence of two record-keeping features: instruments used for tooth extraction and confirmation of complete tooth extraction. Compliance with intra-operative details was strong, increasing from 54% at baseline to 100% at week two. We also achieved 100% compliance with recording intra-operative details in weeks six and eight, with the lowest compliance recorded at 92% in week 4.

Compliance with the post-operative details domain was defined as the presence of two record-keeping features: haemostasis and delivery of post-operative instructions. Compliance in this domain was strong, falling from 100% at baseline to 91% at week two before maintaining 100% compliance in weeks four to ten (Fig. 5).

Figure 5.

Percentage compliance with local best practice for non-surgical tooth extraction in South Powys GDS by Domain.

Discussion

We carried out a quality improvement project to improve record-keeping standards for non-surgical tooth extractions in primary care, to protect patients and the dental team against WSTE. Since implementing a computerised safety checklist in our service, there have been no serious patient safety events or WSTEs.

In the absence of a serious patient safety event, a key measure of success was whether using the computerised safety checklist resulted in increased compliance with LBP record-keeping for non-surgical tooth extraction. This is a strong indicator of patient safety as implementing a surgical safety checklist has been linked with a reduction in harm to patients undergoing tooth extraction (3, 6, 12).

After ten weeks we found that checklist completion led to increased compliance with LBP record-keeping in a primary care setting. Informal conversations with clinicians revealed that they found the computerised safety checklist accessible and efficient, increasing the likelihood of usage and potentially reducing the high levels of workplace stress experienced by dental professionals (13, 14). However, at week 8, 94% of records contained a safety checklist, but only 29% complied with LBP.

Weaknesses in compliance with LBP in the presence of a safety checklist were generally the result of failure to record the assistant's name and the LA batch number and expiry date. There was no dedicated area in the checklist for the clinician to name their assistant. This error in checklist design relied on the clinician to record the assistant's name in a free text box. There is no legal or obligatory requirement to record LA batch number and expiry dates (15). However, we included this as part of LBP because in our department we do not operate a robust system to track LA. We postulated that clinicians did not record LA details as this required the clinician to refer back to the cartridges after surgery. The cartridges' information is displayed in small font, and cartridges are often discarded before the end of surgery. To rectify this weakness in compliance, we aim to introduce a whiteboard system where LA details are updated before each session.

Our study demonstrated the benefit of implementing a safety checklist for non-surgical tooth extraction in primary care to protect patients by ensuring compliance with LBP record-keeping and standardising communication between clinicians. However, data was collected over a short timeframe from three general dental clinics with similar demographics, limiting generalisability.

Furthermore, implementation of a safety checklist is one small part of four strategic areas identified by Pemberton (16) when developing a patient safety culture in dentistry:

1. Incident reporting; 2. Evaluating incidents to identify best practice; 3. Communication and education about patient safety; 4. Building a safety culture

Although safety checklists have shown promise in reducing harm to patients undergoing dental procedures, Saksena (3) reported multiple WSTEs, despite introducing a safety checklist in a large teaching hospital. There is also some concern surrounding the notion that implementing a safety checklist improves patient care, with the nuances of checklist culture requiring further exploration (7).

Our experience was that the safety checklist was often used as a "box-ticking exercise", with clinicians failing to record the "sign-in" process before the procedure and neglecting to undertake a formal pause for site verification. This view is supported by Saksena (3), who identified deficiencies in teamwork and communication leading to poor compliance with patient safety measures in the presence of strong checklist compliance.

Safety checklists are a powerful tool in the patient safety armamentarium for reducing the risk of WSTE. They act as an aide-mémoire and standardise record-keeping procedures. However, for checklists to be effective, they must be used correctly and contribute to the broader departmental safety culture. We must take a pro-active approach to patient safety and avoid becoming comfortable in the absence of adverse patient safety incidents. All team members must understand their role, communicate effectively, have accountability for their actions and feel empowered to voice any ideas or concerns (3, 7). Empowering team members and creating a blame-free environment stems from effective leadership. Patient Safety First recommends introducing patient safety champions at all seniority levels within an organisation (3). Workforce training is also key to ensuring patient safety and should start at undergraduate level for clinicians, continuing through their career (17). A systematic review demonstrated that involving junior staff in a training session on wrong-site surgery and clinical guidelines resulted in a reduction in the incidence of WSTE (18).

To further investigate the role of safety checklists in preventing WSTE, we must refine our checklist for use in primary care and ensure that it complies with LBP. Firstly we will formalise initial discussions held during this study by undertaking a series of structured interviews with participating clinicians to investigate their views on patient safety culture and gather suggestions for checklist optimisation. We will then undertake a second PDSA cycle to assess changes in compliance with LBP and checklist completion. We will also build upon our departmental patient safety culture by appointing staff champions and ensuring that all staff receive patient safety training.

Patient safety is a relatively new concept within primary care dentistry, although it is strongly supported by general dental practitioners (7, 9). We intend to expand our quality improvement project to a wider cohort of general dental practices to benefit the greatest number of patients and dental teams.

Conclusion

A computerised safety checklist for non-surgical tooth extraction in primary care demonstrates the potential to improve patient safety, by adopting measures to prevent WSTE and standardising communication between clinicians. However, we must ensure that safety checklists are used correctly and contribute to a broader patient safety culture. Further research amongst general dental practitioners is needed to optimise the use of safety templates in primary dental care.