Knowledge and Confidence of a Convenience Sample of Australasian Emergency Doctors in Managing Dental Emergencies: Results of a Survey

Background. We aimed to determine Australasian Specialist Emergency Physicians' and Emergency Physicians in Training (Trainees') level of knowledge of common dental emergencies. We also explored confidence in managing dental emergencies; predictors of confidence and knowledge; and preferences for further dental education. Methods. A questionnaire was distributed electronically (September 2011) and directly (November 2011) to Fellows and Trainees of the Australasian College for Emergency Medicine. It explored demographics, confidence, knowledge of dental emergencies, and educational preferences. Results. Response rate was 13.6% (464/3405) and college members were proportionally represented by region. Fewer than half (186/446; 42%) had received dental training. Sixty-two percent (244/391, 95% CI 57.5–67.1) passed (>50%) a knowledge test. More than 60% incorrectly answered questions on dental fracture, periodontal abscess, tooth eruption dates, and ulcerative gingivitis. Forty percent (166/416) incorrectly answered a question about Ludwig's Angina. Eighty-three percent (360/433) were confident in the pharmacological management of toothache but only 26% (112/434) confident in recognizing periodontal disease. Knowledge was correlated with confidence (r = 0.488). Interactive workshops were preferred by most (386/415, 93%). Conclusions. The knowledge and confidence of Australasian Emergency Physicians and Trainees in managing dental emergencies are varied, yet correlated. Interactive training sessions in dental emergencies are warranted.


Introduction
Patients presenting to the emergency department (ED) with dental problems account for 0.3 to 4% of ED visits [1][2][3][4]. In one United Kingdom (UK) study [5], dental problems were the second most common reason (after drug reactions) patients telephoned the ED for advice. The most common dental conditions presenting to ED were dental pain, dental infections, dental and maxillofacial trauma, and postdental treatment-related complications such as haemorrhage and dry socket [2,[6][7][8].
Emergency department medical staff may have difficulty making a specific diagnosis for dental presentations, and often manage these patients symptomatically [9]. This may be because they have had little training in dental emergencies and little exposure to these conditions. A survey of newly graduated physicians working in hospitals in Bahrain, Ireland, Kuwait, and the UK reported that 97% had not received any dental education prior to graduation [10]. In a large study from the United States involving 1030 pediatricians and family physicians, most respondents had less than two hours of preventive dental education during their whole medical training [11]. A further study from England showed that only 6% of ED medical staff had received any dental training in medical school [12].
In a cross sectional, nationwide survey that included 118 doctors working at Welsh emergency departments, knowledge for the management dental trauma was found to be partial, with greater knowledge associated with experience and training [13].

Emergency Medicine International
In a small US study of 72 physicians and directors, knowledge was poor for dental fractures, and good for luxations and avulsions. Specialist qualification in paediatric emergencies was found to be associated with great knowledge in managing dental trauma [14].
Some studies suggest that emergency physicians with inadequate training and insufficient knowledge had difficulty with diagnosis, investigation, management, and appropriate referral of dental conditions [1,15].
An assessment of Australasian ED staff knowledge and confidence is lacking and to our knowledge there are no known studies of factors that predict knowledge and confidence in dental emergencies. An understanding of these factors may permit the targeting of relevant educational resources. In addition, awareness of emergency doctors' views about preferred activities which can improve these weaknesses may enable decision makers to consider suitable teaching methods to improve knowledge and confidence.
We aimed to determine specialist emergency physicians' (fellows') and emergency physicians in training (trainees') knowledge of common dental emergencies. Secondary aims were to assess respondent's self-rated confidence in managing dental emergencies, to determine predictors of knowledge and confidence in managing dental emergencies, to determine the degree of correlation between knowledge and confidence and to determine attitudes towards further education in dental emergencies.

Study Design and Ethics.
This cross sectional survey was approved by St. Vincent's Hospital Melbourne's Ethics Committee and by the Scientific Committee of the Australasian College for Emergency Medicine (ACEM).

Inclusion and Exclusion Criteria.
Participants eligible for inclusion in the study were doctors registered with ACEM including Emergency Physicians ("fellows" or "FACEM"), advanced trainees, and provisional trainees. Doctors not registered with ACEM were excluded.

Tool Development and Validation.
A questionnaire was designed based on dentistry exam multiple choice questions [16][17][18], emergency medicine textbooks [19][20][21], and internet teaching sites. Face validity was ensured using iterative feedback from six ED consultants, two dentists, a researcher, and an emergency medicine trainee after which minor changes were made to the original survey. To ensure content validity, a sample of six emergency physicians and five trainees rated relevance of survey items on a four-point scale. This enabled the identification of items not to be centrally relevant to the topic, items requiring revision in formatting or wording, and items requiring no alteration.
The final questionnaire (see the appendix) comprised 40 items including nine demographic items, 11 items about confidence in diagnosis and management of some of the most frequent dental presentations, 15 multiple choice questions assessing participants' knowledge of common dental emergencies, and five questions regarding desired further training in dental emergencies. The survey incorporated graded responses using Likert scales for attitudinal questions, binary response formats for demographics, and multicategory format (demographics, knowledge questions, and further training in dental emergencies). Some open-ended questions were also included.

Survey Distribution.
The electronic survey, formatted using online SurveyMonkey software, was distributed in September, 2011, by email by the ACEM. Electronic surveys were delivered via hyperlink to the online questionnaire embedded in the email. Due to the rules of the ACEM who distributed the survey on behalf of the authors, only one reminder was sent (October 2011). For ethical and privacy reasons the authors were restricted from making direct contact with potential respondents via electronic means. For these reasons, the authors were limited in maximizing the response rate. The electronic survey remained open until January, 2012.
Although not permitted to send more than one reminder, the authors were able to disseminate a paper based version of the survey to conference registrants at the ACEM Annual Scientific Meeting in Sydney, November, 2011. This conference would not have been attended by all those eligible for the study; however the use of a paper based survey was used both to maximize recruitment and to minimise bias that may have occurred due to limited internet access (which can occur in remote or rural areas) and preference for paper based surveys over electronic formats. Paper surveys were collected in a box to preserve anonymity.
All participants (via electronic recruitment and face-toface) received an invitation to participate and a comprehensive information form outlining participant rights, the length of the survey, and the purpose of the study which was to assess Australasian emergency doctors' knowledge and confidence for dental emergencies. Survey commencement was taken as implied consent and survey completion took approximately 15 minutes.

Primary
Outcome. The primary outcome was the proportion of respondents obtaining a pass mark on the knowledge test. A pass mark was defined a priori by a panel of emergency physicians and registrars as >50% total score (at least 8/15) and was calculated by summing correct answers on the knowledge test for those participants who answered all questions.
2.6. Sample Size. Based on an ACEM membership of 3405 (1392 FACEM, 2013 Trainees) at the time of survey distribution (personal communication, Jane Macaulay, 9/9/11), a sample size of 346 was required in order to estimate the proportion of respondents passing the knowledge test within a 5% margin of error (assuming a 50% response distribution) at a 95% confidence level. This equates to a 50% pass rate for the knowledge portion of the survey. This figure was chosen as it provides the most conservative estimate of sample size required (i.e., maximizing the number required). Further power analyses indicated that a sample size of 346 would be sufficient to detect a difference in two proportions of 15 Emergency Medicine International 3 percentage points (50% versus 65%) with power at 80% and criterion for significance set at 0.05. For multiple regression analyses, we adhered to the rule of thumb for sample size; that is, number of cases = 50 + (8 × number of predictors).

Data Analysis.
Quantitative data were analysed using IBM SPSS Statistics 20.0 (Chicago, IL). For each survey summary statistics (%, 95% confidence interval (CI)) were calculated for the entire sample and by the demographic variables. Arithmetic mean (95% CI) was used to summarize the total number of correct knowledge items across that sample. All data reported were adjusted for missing data on an item by item basis.
Representation of membership per state and membership type was calculated by comparing the confidence intervals (CI) of the percentages of all members of ACEM to the percentages of respondents to the survey. If the CI overlapped, we considered that they were adequately represented.
Negatively worded survey items from the confidence section were reverse scored. Exploratory inferential analyses on confidence and attitudinal items were undertaken after collapsing Likert Scales to binary scales (strongly disagree/disagree/neutral versus agree/strongly agree). Fisher's exact test was used for 2 × 2 contingency tables and Pearson's Chi Square (linear by linear association) was used to identify linear trends for ordinal data (4 × 2 contingency tables).
A total score for overall confidence was calculated by summing collapsed items for confidence. Similarly, a total score was calculated for each participant that completed all items in the knowledge test. Each item was identified as being either correct or incorrect ("incorrect/do not know") prior to summing. To ensure validity of derived total scores, internal consistency of these items was verified using Cronbach's alpha ( = 0.805).
Since the primary outcome (proportion of respondents obtaining a pass mark on the knowledge test) was based on a panel determining that a pass mark be defined as 50% or more, further sensitivity analyses were undertaken comparing mean total knowledge score by Fellow (yes/no), received formal training in dental emergencies (yes/no), attended workshops in dental emergencies (yes/no), ED type and number of patients seen per year (0-25, 26-50, 51+). Data were analysed using independent samples -test and ANOVA in situations where there were 3+ categories.
Multiple regression ("enter" method) was then used to identify (demographic) predictors of knowledge score and all predictors of total confidence. For total knowledge, 11 demographic predictors were assessed: FACEM (no/yes); ED type (adult/other); whether formal education in dental emergencies had been received (yes/no); whether the respondent had participated in a conference workshop on dental emergencies (yes/no), and the number of patients seen annually (continuous); access to specialized dental service coded as six dummy variables (none; onsite dentist 24/7; onsite dentist, limited; dentist on call; refer to dental hospital; refer to private dentist). For confidence, these same predictors were included in the model with the addition of total knowledge score.
For multiple regression, variance inflation factor (VIF) was used to indicate the presence or absence of multicollinearity with a criterion of VIF <5 set for retention of variables. Other assumptions (outliers, linearity, homoscedasticity, and independence of residuals) were assessed by inspecting the residuals scatterplot and the normal probability plot of regression standardised residuals. Pearson's correlation was used to explore an association between knowledge score and total confidence score.
For all inferential tests, alpha was set at 0.05 and two tailed tests of significance were used.

Participation.
Four hundred and ten respondents were recruited via the online survey during the study period and 56 via the paper-based survey. Two participants were not registered with ACEM and were therefore excluded resulting in 464 participants commencing the survey from an eligible pool of 3405 ACEM members (13.6%). Of these, 14 completed the demographics section of the survey only and were excluded from all further analyses. There were no significant differences according to demographics in respondents that commenced the survey ( = 464) and those that completed all elements of the survey ( = 391).

Demographics.
Respondents were from a total of 117 hospitals, and two respondents were fulltime locums. The distribution of respondents by region is comparable to the membership of ACEM (Table 1)  One hundred and eighty six respondents (41.7%) reported having received some formal training in dental emergencies, while 121/447 (27.1%) indicated participating in either a conference or workshop specifically on dental emergencies. Respondents indicated having limited access to a dentist in ED ( Figure 1).

Knowledge of Dental Emergencies.
In total, 391 respondents completed all knowledge test items. For those participants, the mean (95% CI) total score out of a possible 15 was 8 (7.8-8.3; range 1-14). A total of 244/391 (62.4%, 95% CI 57.5-67.1) participants completing all items achieved a pass (>50%) in the knowledge test. Knowledge of dental emergencies varied considerably across topics ( Sensitivity analyses indicated that those that had attended formal education and workshops in dental emergencies, and those that were fellows obtained significantly higher total confidence scores compared to their counterparts. There were no significant differences according to ED type and number of patients seen per year (data not shown).  Figure 2). Receipt of formal dental education was significantly associated with confidence for all tasks apart from pharmacological management of toothache and managing a dental abscess (Table 3). other: 188/340; 55.3%, 95% CI 50.0-60.5). Confidence also varied significantly by recency of graduation as a specialist, with more senior graduates more confident in many areas ( Figure 3). Total confidence score was positively correlated with total knowledge score ( = 0.488, < 0.001). Significant linear relationships existed between confidence for most tasks explored and year fellowship obtained, with a trend toward <2000 * * * * * * * * * * * * * * * Figure 3: Practitioner confidence (agreed/strongly agreed) in specific skills relating to dental emergencies according to year the Fellowship was obtained. * denotes < 0.05. * * * denotes < 0.001. a greater level of confidence and earlier attainment of the fellowship (Figure 3).

Predictors of Knowledge and Confidence.
A significant model emerged for predictors of knowledge ( (11,348) = 6.722, < 0.001; adjusted square = 0.149) and also for predictors of total confidence score ( (12,350) = 23.545, < 0.001; Adjusted square = 0.428). For knowledge, the strongest predictor was being a FACEM, whereas the strongest predictor of confidence was receipt of formal education. Other significant predictors of knowledge were having attended a conference workshop in dental emergencies and having received formal training in dental emergencies (Table 4). Other significant predictors for confidence were being a FACEM having attended a conference workshop in dental emergencies, having a dentist on call, total score on the knowledge test, and number of patients seen (Table 4).

Attitudes towards Knowledge and Training in Dental
Emergencies. The majority of respondents (393/405, 97.0%, 95% CI 94.8-98.4) agreed or strongly agreed that it was important for emergency trainees to have practical knowledge about dental emergencies. Less than one percent (3/408; 0.7%, 95% CI 0.2-2.2) agreed or strongly agreed that there was too much attention on dental emergencies in the current emergency medicine training program.
There were no significant attitudinal differences according to year of fellowship, receipt of formal education, or participation in workshops or conferences on dental emergencies. Other suggested CPD methods included a short rotation at an acute dental clinic or dental hospital ED, bedside teaching with a dentist or maxillofacial surgeon, a course, and on the job teaching.

Continuing Professional
When asked to indicate the preferred frequency of live education programs (e.g., lectures and workshops) on dental emergencies, the modal responses were annually (200/405; 49.4%, 95% CI 44.5-54.2) and every six months (115/405; 28.4%, 95% CI 24.2-33.0). The most commonly preferred duration of live education programs ranged between half

Discussion
Patients with dental problems present to the ED [3,22] mostly on weekends and outside normal working hours, when dentists might not be available. One third of Australian adults do not visit the dentist because of expenses [1,23]. An adequate understanding of dental disease and trauma is important for ED physicians to be able to diagnose, treat, and refer patients with dental emergency efficiently [24]. Poor knowledge can result in premature referral, where management is possible by the emergency doctor but is not provided, or late referral in situations when a tooth-or lifethreatening problem exists. To our knowledge, this is the first study to assess the knowledge and confidence of Australasian Emergency Physicians and Trainees' with respect to dental emergencies. We have demonstrated that knowledge and confidence vary considerably and that they are intimately related. Both are enhanced by seniority and experience working in emergency medicine and by participation in teaching programs or educational activities. Clinicians who scored well on the knowledge score were justifiably confident in their ability to manage dental emergencies.
We found knowledge to be better in some areas than others. More than 70% of respondents correctly answered questions relating to common dental conditions such as tooth avulsion, antibiotics treatment, and treatment of dry sockets. Less than half the surveyed doctors however correctly answered questions about dental abscesses, dental fractures, and ulcerative gingivitis. A previous study [10] of 30 new graduates reported that "no physician" showed a high level of knowledge with respect to managing tooth avulsion. This contrasts with more than 80% of all respondents in our study who correctly answered this knowledge question.
While confidence is not an indicator for competence, it may affect the decisiveness of management. In our study, the range of confidence varied from 27% (95% CI 22.7-30.8) in recognition of periodontal disease to 83% (95% CI 79.1-86.1) for pharmacological management of toothache. Almost 70% of FACEMs and 37% of trainees in our study were confident in managing tooth avulsion which conflicts with a recent UK study [4] among 120 ED physicians with only 20% of them confident in managing dental avulsion injuries. Surprisingly, doctors' self-rated confidence was high in managing dental fractures and abscesses despite their lack of knowledge in these areas. They did however recognize their lack of knowledge in recognizing periodontal disease.
Confidence and knowledge were moderately strongly correlated, and knowledge was a significant predictor of confidence suggesting that the two are intimately related. It is possible that being a FACEM may increase confidence generally, but clearly education plays an important part in both knowledge and confidence since being a FACEM, having received formal training in dental emergencies, and having attended a conference workshop in dental emergencies were significant predictors of both.
These findings suggest that specific education in dental emergencies may be beneficial for doctors working in EDs. Virtually all respondents agreed that it was important for emergency trainees to have practical knowledge of dental emergencies. Interactive workshops were the most highly valued educational format, supported by visual presentations, lectures, and case-based conferences. This is unsurprising since interactive workshop which was most preferred format of future CPD in our study, has significant effect on professional practice based on a Cochrane Database Systemic Review [25]. Most members of the ACEM feel that such training should be provided once or twice a year, for a two hour to half day period. The ACEM may consider offering more sessions on dental emergencies at their annual scientific meetings, potentially involving dental practitioners with particular background, and expertise in dental emergencies.
An interesting finding of the recent UK study [7] showed that only 3.9% of ED physicians would choose to be seen and treated by another ED doctor if they had a traumatic dentofacial injury. Seventy-two percent preferred to be treated by a maxillofacial surgeon and 23.5% by a dentist [4]. While this may be a reflection of attitudes towards emergency clinicians dealing with dental injuries, it is also likely that survey respondents are aware of the "system" and the likely referral to specific dental or faciomaxiallary clinicians for definitive care.
The data from this study reinforce anecdotal evidence that access to dental services in emergency situations is very limited. This lack of access makes it even more important that emergency clinicians have adequate knowledge to manage dental emergencies. Referral off-site to other providers may delay treatment, increase morbidity, and inconvenience patients. Emergency clinicians should be able to manage most of the common emergencies in the ED.
Many dental emergencies rely on the clinician's ability to provide adequate pain relief, including local anesthetic blockade. Although this skill is a procedure for which a high level of competence is expected among ACEM members, only one third of respondents reported that they could confidently use local anesthetics in toothache. Regional anaesthesia techniques, including dental anaesthesia, were rated as one of the most requested education procedures by ACEM fellows in a national learning need analysis [26]. This analysis suggested that high quality and well explained videos and regular workshops could help members improve this core competency, as suggested by respondents to our survey. Our finding that only 17% of members without formal education could perform local anesthetics compared with 46% with some kind of formal education, supports the value of such educational activities.

Limitations
This study is not without limitation. The overall response rate was low which is not uncommon for survey research and may be lower among this cohort due to the possibility of eligible participants being asked to complete multiple survey-based research projects via the ACEM (Dr Andrew Gosbell, personal communication), thereby resulting in survey fatigue. The distribution of respondents by region was however comparable to the eligible study population, except for a slight preponderance of Victorian respondents and fewer respondents from Queensland. This suggests the sample was reasonably representative of ACEM members. Further demographic information about the total ACEM population was not available to the authors and therefore restricted further assessment of representativeness. The use of ACEM as the organization distributing the email to potential participant was an ethical requirement for the study due to privacy laws but is likely to have resulted in the use of correct email addresses as ACEM is the professional body that manages specialist status in Australasia for emergency physicians.
We attempted to minimise selection bias by recruiting participants through the two methods described: an online survey and by direct distribution of the survey to eligible registrants at the ACEM Annual Scientific Meeting. The direct recruitment approach was used in order to enroll eligible doctors whose participation in the study through completion of the online questionnaire was unlikely due to internet access barriers or a bias against online survey completion. We cannot exclude the possibility of responder bias; those more interested in dental emergencies may have been more likely to respond to the survey. If such effect were present this may have produced an overestimate the actual knowledge and confidence of ACEM members.
We cannot exclude the possibility of responder bias; those more interested, competent, or confident in dental emergencies may have been more likely to respond to the survey, and those less comfortable with their knowledge may have been less inclined to respond. The possible survey fatigue among the target group may have biased participation to those who are more engaged and possibly therefore more likely to have been involved in educational and other college activities; this might also have resulted in higher test scores.
We attempted to minimise measurement bias by establishing face validity and content validity of the tool, given the absence of a previously validated tool on dental emergencies. Additionally, we undertook further validation of the internal consistency of the knowledge component and confidence component of the tool using Cronbach's alpha. Although the multiple choice questions that comprised the knowledge assessment in this survey were reviewed by several ED consultants and registrars, it was not an exhaustive examination of dental emergencies, assessing just a few key presentations and therefore may not extrapolate well to other dental emergencies not included in the survey. Further, although the pass mark of 50% was set a priori by a panel of emergency physicians, this cutoff may be deemed by some as somewhat arbitrary; however the findings of sensitivity analyses were in line with the expectation that seniority and exposure to education and training would result in higher total scores.
We did not instruct participants to avoid consulting educational materials to improve performance. The anonymous nature of the survey, however, may have minimised any such Hawthorne Effect.

Conclusions
The knowledge and confidence of Australasian ED doctors with respect to dental emergencies is varied, being good in some areas but in need of improvement in others. Seniority in the ED is associated with both confidence and knowledge. However, formal interactive education is also associated with confidence and knowledge, and importantly, is desired by Emergency Medicine International 9 the majority of specialist and trainee emergency physicians in Australasia. The use of designated training workshops on dental emergencies is therefore warranted.