Do paediatric patient-related factors affect the need for a dental general anaesthetic?

Aims The aims were to determine the profile of patients referred for treatment with dental general anaesthetic (DGA) and identify patient-related factors that contributed to a child requiring a DGA. Design setting In total, 42 patients were recruited from new patient assessment clinics at the Edinburgh Dental Institute, Scotland. Prospective questionnaires were given to the patients' caregivers with questions regarding their child's dental anxiety level, ethnicity, level of English spoken and languages spoken at home. Height and weight (BMI) measurements were taken and other information (age, sex, medical history status and postcode for level of social deprivation) were collected from their medical records. Results Compared to the local population, children referred for treatment with DGA had a significantly decreased level of English (p = 0.0001) with an increase in non-English languages at home (p = 0.0004). Patients from minority ethnic groups (p = 0.008) and children from socially deprived areas (p = 0.0001) were significantly overrepresented. Self-reported dental anxiety was high and more common in primary than secondary school-aged children (p = 0.039). Conclusions This study highlighted several patient-related variables that may increase the likelihood of DGA treatment. Further work is required to confirm these associations and develop interventions to reduce the number of children requiring DGAs. Children from minority ethnic groups and areas of increased socioeconomic deprivation were significantly overrepresented in requiring treatment under general anaesthetic. Children requiring treatment under general anaesthetic had significantly lower levels of self-reported English and significantly higher levels of non-English languages at home compared to the background population. Highlights several patient-related factors that may increase the likelihood of dental treatment under general anaesthetic. Children from minority ethnic groups and areas of increased socioeconomic deprivation were significantly overrepresented in requiring treatment under general anaesthetic. Children requiring treatment under general anaesthetic had significantly lower levels of self-reported English and significantly higher levels of non-English languages at home compared to the background population. Highlights several patient-related factors that may increase the likelihood of dental treatment under general anaesthetic.


Introduction
Dental general anaesthetic (DGA) provides dental treatment during a 'state of controlled unconsciousness' . 1 Although it is a treatment modality commonly used in paediatric dentistry, DGA comes with well-documented risks. 1,2,3 Additionally, there often is an extended waiting period for paediatric DGA procedures with associated impacts on the child's wellbeing. 4 In the UK, all DGAs are undertaken in hospital settings, 5 resulting in significant costs. In 2018-2019, NHS spending data estimated that £41.5 million was spent funding 44,685 hospital operations for extractions of teeth in children in England. 6 A misconception is that a DGA may prevent children becoming more anxious about dental treatment. 3 Research has shown that dental anxiety may increase following a DGA, 4,7 suggesting 'behavioural therapy and desensitisation may be required following a DGA' . 8 However, post-DGA dental anxiety is offset with an improved quality of life following DGAs. 8 Caries risk factors are well-established and national guidelines have been published on this topic. 9,10 These risk factors are previous or current clinical signs of caries, dietary habits, socioeconomic status, fluoride usage, level of oral hygiene, low salivary flow and medical history status. 10 However, this guidance recognises that there is no consensus for the best predictors of caries. A systematic review concluded that socioeconomic and demographic information and baseline caries were good predictors of caries risk. 11 Despite similarities with caries risk factors, the factors that increase the likelihood of treatment under DGA are not identical. One cross-sectional survey concluded that a patient's age and country of residence were associated with treatment under DGA, with children living in Wales being significantly more likely to receive DGA treatment than those in England. 12 There is a paucity of information in the literature on patient-related risk factors for children receiving a DGA. This study aims to determine the profile of patients needing treatment with DGA and to identify patientrelated factors that may contribute to a child requiring a DGA. The null hypothesis was there was no association between any patient-related factors and receiving treatment under a DGA.

Study design and population
This prospective, cross-sectional study was undertaken in the paediatric dentistry department at the Edinburgh Dental Institute. Ethical approval was gained from the Berkshire Children from minority ethnic groups and areas of increased socioeconomic deprivation were significantly overrepresented in requiring treatment under general anaesthetic.
Children requiring treatment under general anaesthetic had significantly lower levels of selfreported English and significantly higher levels of non-English languages at home compared to the background population.
Research Ethics Committee (19/SC/0494). For all study participants aged 16 and under, consent was given by their parent or legal guardian. Additionally, the study participants aged eight and above consented to participate in the study and to have their data used as part of the research by completing an assent form.

Data collection
Patients aged between four and the day before their sixteenth birthday, examined in new patient assessment clinics between 11 November 2019 and 1 April 2020, and where DGA had been chosen for treatment, were invited to take part in the study. Patients' caregivers with limited English attended with interpreters as is routine within the department. Children without an accompanying adult with parental rights were excluded from the study.
Patient data were transferred to a Microsoft Excel (Microsoft, USA V14.2.0, Santa Rosa, CA, ISA) spreadsheet. Data collection forms were populated from the completed questionnaires and the patient's dental records. Data collected included the patient's age, sex, any significant medical history, postcode, height, weight, ethnicity and languages spoken at home. These data were compared with Scottish data obtained from the most recent Scottish Census (2011), 13 Scottish Index of Multiple Deprivation (SIMD) 2020 14 and Scottish Information Services Division (2019). 15 The format of the questions regarding the patient's ethnicity, languages spoken at home and how well they spoke English were derived from the most recent Scottish Census. 13 Dental anxiety was measured by one of two questionnaires depending on the patient's age; if aged between 4-12 years, the Modified Child Dental Anxiety Scale with Faces (MCDASF) was used. The MCDASF questionnaire has been validated for this age group. 16,17,18 Patients above 13 used the Modified Dental Anxiety Scale (MDAS). There are no validated dental anxiety scales available for this cohort. However, the MDAS was validated from 18-60-years-of-age, hence provided the nearest age-match. 19 For children of age eight and under, the clinician would ask the children the questions from the MCDASF questionnaire. Children over nine years of age would self-complete the questionnaires. The clinicians were available to assist the child if this was required. We recommended against parents completing the questionnaire to minimise bias.

Statistical analysis
Data were transferred from Microsoft Excel spreadsheets to IBM SPSS v24 (IBM, Armonk, NY, USA) for processing and statistical analysis. T-tests compared self-reported dental anxiety scores against other variables. Chisquared tests compared our datasets against the most recent City of Edinburgh Scottish Census 13 (ethnicity, languages spoken at home, how well English is spoken) data. Chi-squared tests compared against the City of Edinburgh SIMD 2020 14 (socioeconomic status) and the National Scottish Government Primary 1 (P1) age statistics 15

(body mass index [BMI]).
Residence from relative deprivation was selected as the primary outcome measure. Based on the local population it would be expected that 20% of patients would be from SIMD 1 or 2 (most deprived). In order to detect a doubling of this deprivation rate to 40% among our DGA sample, the minimum number of children required was 36 (p = 0.05) which would show a significant difference in SIMD deciles 1 and 2 membership.

Patient sex and age
The demographics of the children in our study compared with the overall data from the Scottish Census (2011), 13 SIMD (2020) 14

Ethnicity
Study participants reported a non-white British ethnicity in 23.8% of cases, which contrasts with 12.6% in the census data, a 1.89 times increase (p = 0.008). Participants of Asian ethnicity were reported 3.8 times higher in the study group (14.2%) compared to the census (3.7%).

Level of spoken English
Six times more study participants reported speaking English either not well or not at all compared to census data (9.6% versus 1.6%; p = 0.0001).

Languages spoken at home
A language other than English was spoken at home 4.04 times more in the study group compared to the census data (19.0% versus 4.7%; p = 0.0004).

SIMD
Study participants resided in areas of high relative deprivation (SIMD 1 and 2) at a rate 3.20 times greater than the background population (38.1% versus 11.9%; p = 0.0001). Conversely, there was a 0.43 times reduction in study participants from areas of low relative deprivation (SIMD 9 and 10) compared to the background population (19.0% versus 44.1%; p = 0.0001).

BMI
In total, 19.5% of our patients were above a healthy weight, (BMI 91 st centile or greater). Two participants (4.9%) were above the 98 th centile, which placed them in the obese/ severely obese category. The remaining participants (80.5%) were all within the healthy weight categories (healthy weight BMI ranged from 0.4 th to <91 st centile). One participant was excluded from BMI measurements as the child was wheelchair bound.

Medical history
In total, 88.1% of participants were healthy (n = 38), with only four having significant medical issues (autism [n = 2], motor neuron disease [n = 1] and challenging behaviour with fine motor difficulties [n = 1]).

Dental anxiety level
Overall, 60.4% (n = 27) of participants had a self-reported dental anxiety (score of 19 or more). 16,17,18,20 Only one was aged 12 or older. Chi-squared tests comparing our findings with those reported by Haworth et al. (2016) 21 showed the differences in the numbers of dentally anxious patients that had DGAs were statistically significant (X 2 [1, N = 42] 44.935, p = 0.0001). Table 2 compares the patient-related variables between children with self-reported dental anxiety (MCDASF and MDAS scores ≥19) against those without self-reported dental anxiety (MCDASF and MDAS scores <19). Dental anxiety was higher in the primary school-aged children compared with secondary school-aged children. The mean dental anxiety score of primary school children was 20.3 (SD = 5.01) while that of secondary school children was 15.0 (SD = 4.94). This difference was statistically significant (p = 0.039). For all the other patient-related variables, there were no statistically significant relationships found with self-reported dental anxiety.

Discussion
Language would appear to impact on children requiring a DGA. In our study, whether English was spoken well/very well or another language beyond English was spoken at home were significant factors when compared with the background population. These findings are novel as we are unaware of previous studies which have reported any associations between spoken language and the risks of receiving a DGA.
The reason for the association between language and children requiring a DGA is likely to be multifactorial. Preventative advice and oral hygiene instructions (OHI) by dental professionals may be challenging if children and their caregivers have communication difficulties. Behaviour management techniques rely heavily on good communication between the clinician and the child. A study of a similar cohort of children in Lothian (Scotland) found that only 39 out of the 80 children (48.8%) had OHI and diet advice given by their general dental practitioner in the 12 months preceding their DGA. 22 This suggests that if less than half of all children needing a DGA had the recommended preventative advice 9,10 then those with difficulty communicating in English would be further disadvantaged, increasing the risk of requiring a DGA.
There were fewer children in our study from a White British background and an overrepresentation of other ethnic backgrounds compared with the Scottish Census 13 data, which was statistically significant. While no previous studies have shown a correlation between ethnicity and the likelihood of receiving a DGA, two studies concluded that children from 'non-British' or other ethnic backgrounds had significantly more teeth extracted under DGA compared to White British. 23,24 Ethnic diversity could lead to differences in oral health behaviours, perceived importance of oral health or barriers to oral healthcare access. Further research is required to investigate the differences between   Table 2 Comparison of patient-related variables who received treatment under DGA in dentally anxious and non-anxious children ethnicities and cultures so individualised oral health promotion can be delivered. We are mindful that, while it is the most up-to-date dataset currently available, the Scottish census data are over a decade old. The majority of the study sample resided in the most relatively deprived areas (SIMD 1 or 2). This is statistically significant when compared with the SIMD (2020) 14 datasets. Children from the least relatively deprived areas (SIMD 9 or 10) were underrepresented compared with the SIMD (2020) 14 datasets which were also significant. These findings indicate that a child's socioeconomic status is a potential determinant on the receipt of dental treatment under DGA. It is possible that children from the least deprived areas may have accessed private care due to longer waiting time for treatment within the NHS. However, there are no private general anaesthetic treatment centres within the Lothian area.
As previously mentioned, socioeconomic status is a known risk factor for caries. 9,10 Caries prevalence in children maps to socioeconomic status, so the prevalence of children in the study from the different SIMD groups may be a reflection on the disease burden in society. 25 This may be reflective of the access to oral healthcare service within these different socioeconomic groups. Families with greater resources may be able to access services earlier and support the child attending multiple appointments to avoid DGA, whereas more disadvantaged families may not attend services until disease reaches a 'crisis' point and may only be able to support the treatment modality that resolves the issues as quickly as possible. These findings correlate with the previously reported evidence linking DGA use and social deprivation. 26 Most of the children in our study had selfreported dental anxiety, which correlates with previous studies. 12,21,26 However, our study had a higher proportion of children that had dental anxiety when compared with previous studies which was statistically significant. 12,21,26 This provides further evidence that dental anxiety is a risk factor for children receiving dental treatment under DGA.
Our study found that primary school aged children (4-to-11-year-olds) had a significantly higher level of self-reported dental anxiety than secondary school aged children (12-to-16-year-olds). This contradicted a previous finding where dental anxiety in 12-to-16-year-olds was twice that of the 7-to-11-year-old children. 27 However, these comparisons are of limited benefit as study populations differ. It would seem logical that primary-school-aged children were pre-cooperative and unlikely to manage chairside comprehensive caries management plans. Secondary school-aged children in our study did not have self-reported dental anxiety. The absence of dental anxiety in the older cohort was likely due to differences in the nature of their treatment under DGA, with younger children more likely to need DGA for caries management while older children requiring DGA for provision of invasive procedures, such as surgical removal of ectopic teeth.
In this study, sex, BMI and a significant medical history status were not associated with an increased risk of treatment under DGA. Currently, there is no evidence that sex is associated with an increased risk of receiving dental treatment under the DGA. Studies investigating links between BMI and effects on caries incidence reported conflicting findings. 28,29 A systematic review, investigating a potential link between caries and obesity, only found three studies with children to provide high enough evidence on the topic which had conflicting results. 30 There is currently no clear evidence that the medical history status of paediatric patients affects the caries risk. 31,32 The majority of the patients were treated for caries. This was expected as caries management is the most common reason for referrals to our paediatric services. Surprisingly, only one child was treated for a traumatic dental injury. This is most likely due to children having treatment done chairside with or without inhalation sedation.
One limitation of this study is the crosssectional design, which does not allow analysis of patient-related factors over a period of time. Dental anxiety may be different on the day of surgery and following DGA. The study design does not allow for a control group or random sampling, potentially increasing the risk of selection bias.
In terms of future directions for clinical practice, we would consider which patientrelated risk factors have shown an increased risk of children undergoing DGAs. The lack of spoken English from the children and their caregivers may create barriers to routine healthcare and health education. Thus, these families may seek care once the disease had progressed to a crisis point necessitating a DGA. To gain a further understanding of this, qualitative studies with patients who do not speak English well or who speak a different primary language would be helpful.
Socioeconomic deprivation is recognised as a caries risk factor and forms the basis of the targeted elements of the Childsmile community oral health programme. 33 The study data demonstrate the health inequality of children from areas of socioeconomic deprivation being more likely to be seen for treatment under DGA. The basis for this inequality is likely to be multifactorial. However, a fuller understanding of what health experiences drive children from these backgrounds towards DGA may help inform interventions to reduce the need for DGA by this group. Again, qualitative studies would likely be helpful in investigating this.

Conclusion
In conclusion, within its limitations, our study showed that level of spoken English, languages spoken at home, ethnicity (other than White British), self-reported dental anxiety and socioeconomic deprivation may be factors that increase the likelihood of receiving dental treatment under DGA.