RANDOMIZED CONTROLLED TRIAL

Assessment of the effect of combined multimedia and verbal information vs verbal information alone on anxiety levels before bond-up in adolescent orthodontic patients: A single-center randomized controlled trial
Jemnique Pal Kaur Srai,a Aviva Petrie,b Fiona S. Ryan,c and Susan J. Cunninghamd London, United Kingdom

Introduction: Despite the high prevalence of dental anxiety in children, there is little research examining anxiety before orthodontic procedures. This parallel-group randomized controlled trial assessed whether provision of additional multimedia information regarding the bond-up procedure affected anxiety in adolescent orthodontic patients. The effects of sex, ethnicity, and age were also investigated. Methods: Participants were recruited from the orthodontic department of the Eastman Dental Hospital, University College London Hospitals Foundation Trust, in the United Kingdom; all were 10 to 16 years of age, with no history of orthodontic treatment, and patient assent and parental consent were obtained. The participants were randomized into control (n 5 45) and intervention (n 5 45) groups using a random number table. Both groups were given verbal information regarding the bond-up procedure, and the intervention group was additionally given a DVD showing a bondup. Anxiety was assessed in the department immediately before the bond-up using the State-Trait Anxiety Inventory for Children, with state anxiety as the primary outcome measure. The researchers were unaware of group allocations while enrolling patients, scoring questionnaires, and analyzing data. Results: A statistically signicant difference was found between groups, with a difference in scores of 2 (95% condence interval for the difference 5 0.15 to 3.85). The median state anxiety was 32 in the control group (n 5 42) and 30 in the intervention group (n 5 43; P 5 0.012). Sex, ethnicity, and age did not signicantly affect anxiety. No harmful effects were noted. Conclusions: Additional information reduces anxiety levels, but other methods could be more costeffective than the DVD. Sex, ethnicity, and age did not statistically affect the anxiety levels. (Am J Orthod Dentofacial Orthop 2013;144:505-11)

D
a b c

ental anxiety is a complex subjective emotion with many inuencing factors: eg, age, sex, socioeconomic group, parental dental anxiety, and previous dental attendance and experiences. Studies show conicting evidence with regard to the effects of

Specialist Registrar, Eastman Dental Institute, University College London. Biostatistics, Eastman Dental Institute, University College London. Consultant, Eastman Dental Hospital, University College London Hospitals Foundation Trust. d Professor, Orthodontic Department, Eastman Dental Institute, University College London. All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conicts of Interest, and none were reported. Reprint requests to: Susan J. Cunningham, Orthodontic Department, University College London Eastman Dental Institute, 256 Gray's Inn Rd, London, WC1X 8LD, United Kingdom; e-mail, s.cunningham@ucl.ac.uk. Submitted, March 2013; revised and accepted, June 2013. 0889-5406/$36.00 Copyright 2013 by the American Association of Orthodontists. http://dx.doi.org/10.1016/j.ajodo.2013.06.013

these factors on anxiety.1-7 This is understandable when one considers the methodologic difculties for researchers and the fact that it is not possible to objectively assess a patient's own experience. Thus, patients must be relied upon to self-assess their anxiety; this introduces the possibility of reporting or recall bias. Recall bias can be reduced through the use of prospective trials; however, the assessment of an emotional response is complex. Several self-assessment tools have been developed for use in research into dental anxiety, but some are specic to general dental procedures (ie, restorative work or extractions); naturally, these do not apply to orthodontics. Dental anxiety in children and adolescents is common and well documented, with direct repercussions on the care sought and received by patients.1 Despite the acknowledgment of the high prevalence of dental anxiety, there is currently little research examining
505

506

Srai et al

anxiety levels before orthodontic procedures and methods for reducing it. Previous research has suggested that adolescent orthodontic patients are unclear about the process of placing xed orthodontic appliances and would like more information to reduce their anxiety.8 The majority (87%) of participants in this research project stated that they thought that information in an audiovisual format would be benecial because it would allow them to see what actually happens. The reason for undertaking this study was therefore to examine the effect of additional multimedia information provision on anxiety levels in adolescent orthodontic patients before the bond-up procedure. The aim of this study was to ascertain whether the additional preparatory information regarding the orthodontic bond-up procedure reduces anxiety in adolescent orthodontic patients. A secondary aim was to establish the inuence of patients' sex, ethnicity, and age on anxiety before the bond-up procedure. The null hypotheses for this study were the following: (1) the provision of additional multimedia information regarding the orthodontic bond-up procedure does not affect the median anxiety score in adolescent orthodontic patients before the procedure; and (2) sex, ethnicity, and age do not inuence the median anxiety score before the orthodontic bond-up procedure.
MATERIAL AND METHODS

Ethical approval was obtained from the National Research and Ethics Service (reference number 11/LO/ 0392), and the University College London Hospitals Foundation Trust Research and Development Department granted research and development approval. This was a prospective parallel-group randomized controlled study in which patients were allocated into either the control or the intervention group. Verbal information regarding the bond-up procedure was provided by the treating orthodontist to the participants in both groups, according to the department's usual practice. Patients in the intervention group were also given a DVD created specically for this study and containing additional information regarding the bond-up procedure. Anxiety was assessed on the day of the procedure, just before the bond-up, using the State-Trait Anxiety Inventory for Children.9 The patients for this study were recruited from the orthodontic department at the Eastman Dental Hospital, University College London Hospitals Foundation Trust, London, United Kingdom. They were 10 to 16 years old and were due to undergo xed appliance treatment and able to give assent; parental consent was also obtained. Patients were excluded if they had a history of previous orthodontic treatment or were receiving other orthodontic interventions (eg, headgear) concurrently.

The participants were block randomized in groups of 10 to either the control or the intervention group using a random number table, with allocation of those with odd numbers to the control group and those with even numbers to the intervention group. To reduce bias, sealed opaque envelopes containing information relating to the allocations were prepared before the study. Since these were only opened once written consent and assent had been obtained, the researchers were not aware of the patient's allocation while obtaining consent. The researchers remained blinded to group allocations when scoring the questionnaires and entering data into the results spreadsheet. An Internet search was undertaken to assess the amount and quality of audiovisual information already available to patients and to establish whether any preexisting videos were suitable for use in this study. A number of videos concerning placement of xed appliances were available; most were homemade, but some were commercial. However, several issues were noted, including the lack of verbal explanations, poor visibility of the actual procedure, and use of computer-generated models rather than a patient in some cases. A DVD was therefore created specically for this study, with an introductory section explaining the components of a xed appliance and the materials and instruments used during the procedure, followed by video footage of a patient having xed appliances placed. A trial DVD was created using footage of a colleague who was due to commence xed appliance treatment and had consented to the lming of the bond-up for this purpose. This footage was edited to remove substantial repetitive scenes of the bond-up procedure because it was agreed that a viewer's interest would not be maintained over a protracted time period. The footage was therefore edited until the DVD lasted 15 minutes. This draft was critiqued by the researchers, and changes were made to ensure that the explanation was more child friendly and the quality of the DVD footage was improved. It was also agreed that the patient shown in the video of the bond-up procedure should be in the same age group as those recruited for this study to allow effective modeling. Once a suitable patient had been identied and consented for the lming of the bondup procedure, a second draft of the DVD was created. Again, this was appraised and edited until the researchers were satised that the DVD was as informative, engaging, and succinct as possible. DVD labels were then created to inform viewers that it should play automatically on insertion and to provide a contact e-mail address if they had any problems viewing the content.

October 2013  Vol 144  Issue 4

American Journal of Orthodontics and Dentofacial Orthopedics

Srai et al

507

The DVD was provided only to patients in the intervention group once written consent had been obtained and the group allocation had been checked. Because this was done before the bond-up date (eg, at the orthodontic records or treatment planning appointment), the patients could take the DVD home to watch before the bond-up appointment. The outcome measure for this study was the patient's anxiety score as measured by the State-Trait Anxiety Inventory for Children.9 Unlike some dental anxiety measurement tools, no scenarios are provided in this questionnaire. This, along with the fact that it was created specically to measure anxiety in children and adolescents, makes it applicable for use in most medical and dental settings and ideal for this orthodontic study. This self-reported questionnaire comprises 2 scales to distinguish between state anxiety and trait anxiety. The state anxiety scale poses 20 questions, asking how respondents feel at that particular moment in time, thus assessing transitory anxiety states. The trait anxiety scale also asks 20 questions, but these relate to how the subject usually feels, thus assessing each patient's relatively stable tendency to anxiety. Because information provision regarding a procedure is likely to affect anticipatory anxiety rather than the stable tendency to anxiety, the primary outcome measure was the state anxiety score. The questionnaire was distributed to the patients by a researcher (J.P.K.S.) on the day of the bond-up procedure, immediately before their appointment. Each questionnaire had the patient's unique identication number on the front page. The researcher provided verbal instructions about how to complete the questionnaire and highlighted the written instructions at the top of the questionnaire. An envelope was provided, and the patient was asked to seal the completed questionnaire inside and return it to the orthodontist or the reception staff. The patients were also asked to record on the outside of the envelope if they were in the intervention group but had not watched the DVD for any reason. The envelope was then returned to the researcher after the bond-up appointment, and the questionnaire was removed and placed into a le to be scored later. The outside of the envelope was checked for any notes before it was discarded. It was hoped that this method of collecting the questionnaires would reduce responder bias in patients who might have felt compelled to complete the questionnaire by selecting answers they perceived to be correct in the researcher's eyes. Additionally, it was hoped that this assurance of anonymity would reduce any anxiety that the patients might feel if the completed questionnaire was visible to both researchers and clinicians whom they might expect to judge them on their responses.

Because there were no studies in the orthodontic literature with the same version of the State-Trait Anxiety Inventory for Children questionnaire, a sample-size calculation was carried out using data from the rst 10 questionnaires returned from each group. The standard deviation of the state anxiety scores was 6.0, and the clinically relevant difference in state anxiety scores was set at 4.0 points. Using nQuery Advisor (version 4.0; Statistical Solutions Ltd, Cork, Ireland) and based on an independent 2-sample t test, it was estimated that 37 participants were required in each group for signicance of P \0.05 and 80% power.10 Since the remaining results were unlikely to be normally distributed, it was probable that a Mann-Whitney U test would be used, so a correction factor was applied, resulting in a recommended sample size of 43 per group.11 To allow for dropouts from the study, it was decided that 45 participants should be recruited per group: ie, a total of 90. The data were not normally distributed. To investigate differences between the control and the intervention groups, the median state, trait, and total anxiety scores were compared with the Mann-Whitney U test. This test was also used to investigate the effect of sex and ethnicity on anxiety. Univariable linear regression analysis was used to establish the independent effect of age on patient anxiety. It was thought that there might be a relationship between a patient's underlying stable trait anxiety and how he or she reacts to stressful situations, and this might affect the analysis of the effect of information provision on anxiety. Therefore, a multivariable linear regression analysis was used to assess the group effect, with control for trait anxiety. The assumptions underlying the regression analyses were checked by a study of the residuals and found to be satised. A signicance level of 0.05 was used for all hypothesis tests. All analyses were performed using SPSS software (version 21.0; IBM SPSS Statistics for Windows, Armonk, NY). The mean state anxiety scores of both the male and female participants in this study were also compared with the normative values provided in the state-trait anxiety inventory for children manual, for the control and the intervention groups. A 1-sample t test was used for this comparison because the only normative values were mean values. Only state anxiety scores were compared because state anxiety scores had greater importance in this study.
RESULTS

The participants were recruited between November 2011 and May 2012; 85 questionnaires were returned by December 2012. A CONSORT ow diagram (Fig) demonstrates the progression of patients in the study.

American Journal of Orthodontics and Dentofacial Orthopedics

October 2013  Vol 144  Issue 4

508

Srai et al

Assessed for elegibility (n = 92)

Enrolment
Excluded (n = 2) Randomised (n= 90) (declined to participate)

Allocation

Control group (n = 45)

Intervention group (n = 45)

Lost to follow-up (n = 3)

Lost to follow-up (n = 2)
[change in treatment plan n=1, patient declined treatment n=1]

Follow-up

[change in treatment plan n=2, patient discharged n=1]

Analysis

(n = 42) Excluded from analysis n = 3, as lost to follow-up

(n = 43 ) Excluded from analysis n = 2, as lost to follow-up

Fig. CONSORT ow diagram of participants in the study.

Of the 90 patients recruited, 85 returned to have xed appliances placed and completed the questionnaire. Five patients were lost to follow-up, and the reasons included a change in treatment plan (n 5 3), poor oral hygiene resulting in discharge (n 5 1), and treatment declined (n 5 1). The baseline demographics for the patients who completed the questionnaire are shown in Table I. Analysis was carried out per protocol; this was deemed appropriate because there were only minor losses to follow-up. The effect of group allocation on anxiety scores before bond-up is shown in Table II. A statistically signicant difference of 2 points (95% condence interval [CI] for difference, 0.15-3.85) was found in the median state anxiety scores between the groups, with

participants in the intervention group showing lower state anxiety scores (P 5 0.012). No signicant difference was found for trait or total anxiety scores between the 2 groups. The multivariable linear regression to assess the group effect, having controlled for trait anxiety, indicated that both trait anxiety and group allocation had signicant effects on state anxiety scores (P \0.001 and P 5 0.020, respectively) (Table III). However, the adjusted R2 value for this analysis was 0.187, indicating that only 18.7% of the state anxiety score was affected by the trait anxiety score and the group allocation. There were no signicant sex or ethnicity effects and no signicant correlation between age and state anxiety as assessed with the Spearman rank correlation test.

October 2013  Vol 144  Issue 4

American Journal of Orthodontics and Dentofacial Orthopedics

Srai et al

509

Table I. Baseline demographics of the patients who

completed questionnaires
Group Control (no DVD) (n 5 42) 18 24 28 14 13 Intervention (DVD) (n 5 43) 16 27 29 14 13 Total population (n 5 85) 34 (40.0) 51 (60.0) 57 (67.1) 28 (32.9) 13

No harmful effects were found at any point while conducting this study.
DISCUSSION

Demographic Sex, n (%) Male Female Ethnicity, n (%) White Other races Median age (y)

Table II. Relationship between median anxiety score

and group
Control group Intervention group (n 5 42) (n 5 43) Median Median score Range score State anxiety subscale 32 38 30 Trait anxiety subscale 30 30 30 Total anxiety 61 59 58 P Range value 22 0.012 30 0.850 39 0.160

Table III. Multivariable linear regression analysis to

assess the effect of underlying trait anxiety and group allocation on state anxiety
95% CI for B Independent variable Trait anxiety Group Regression coefcient B 0.393 3.034 Lower 0.187 5.580 Upper 0.599 0.489 P value \0.001 0.020

Table IV. Analysis of mean state anxiety scores for the

research group compared with normative values from the state trait anxiety inventory for children manual
Sex Female Male Study group compared with normative values Control Intervention Control Intervention P value 0.08 0.51 0.31 0.07

A 1-sample t test was carried out to allow comparison of the state anxiety of the study population with the mean normative values for the State-Trait Anxiety Inventory for Children. These results are shown in Table IV. The state anxiety scores for the patients in this study did not differ signicantly from the normative values.

The DVD was specically created for this research; this ensured that the information provided encompassed what the researchers thought was important, but it also included information requested by patients in a previous study in our department.8 The DVD was viewed and critiqued by orthodontic experts, colleagues, the patient lmed in the DVD, and her parents; all of them thought that the nal version met the requirements for this study. A sample size calculation was carried out to ensure that the study had sufcient power. In view of the loss of 5 patients, a retrospective power calculation was undertaken on the data from the 85 questionnaires returned and showed that the 80% power required had still been obtained. The study results are therefore statistically robust and allow conclusions to be drawn. The questionnaire was distributed by the researchers; this could potentially introduce bias. However, because the researchers met the patients only once, the likelihood of remembering the patient's group allocation when distributing the questionnaire was low. Additionally, the researchers scored the questionnaires and entered these data into the spreadsheet with no knowledge of the patients' groups, and the group was entered just before data analysis. It was intended that these measures would minimize bias. The ndings showed that before the bond-up, both the control and the intervention groups had state anxiety scores similar to normative data (Table IV). The authors of the questionnaire collected mean normative values for state anxiety by asking the children to imagine that they were about to sit for an academic school examination. This comparison therefore suggests that attending a bond-up seems to have a similar effect on anxiety as does a school examination; the DVD does not appear to have reduced state anxiety below expected normative levels. Most clinicians would like their patients to be less anxious when they are due to undergo the bond-up procedure than they would be for a school examination, and we should therefore continue to aim to reduce patient anxiety levels in an orthodontic setting. The median state anxiety levels in the intervention group were 2 points lower than those in the control group, a difference that was statistically signicant (P 5 0.012; 95% CI for the difference, 0.15-3.85) (Table II). This indicates that viewing the DVD before bond-up reduced state anxiety. No signicant difference was found between the 2 groups for either trait or total

American Journal of Orthodontics and Dentofacial Orthopedics

October 2013  Vol 144  Issue 4

510

Srai et al

anxiety levels. This was to be expected because the DVD only provided information regarding the bond-up procedure, and it should therefore affect only a participant's anticipatory anxiety. However, although the difference was signicant, it was less than the clinically relevant difference of 4 points used for the sample size calculation. Given the limited reduction in anxiety and the costs involved, it might not be justiable to routinely distribute a DVD in a clinic or a practice to reduce anxiety. It might, however, be viable to give patients the information in a way that would cost less; the footage from the DVD could be put on social media sites (eg, YouTube) or allowed to play in a clinic waiting room. This might reduce patients' anxiety in a more cost-effective manner. The provision of this information on social media sites might have certain benets when compared with playing the video in a clinical waiting area because it would allow patients to watch the video as many times as they would like and in a more relaxed environment. Clinicians could then use information leaets to direct patients to reliable videos placed on such sites. This signicant reduction in anxiety contrasted with results from previous studies examining the effect of information in a printed leaet.12-14 It is possible that the audiovisual format was seen as more interactive; because it allowed patients to see a real-time video of the procedure they would undergo, there might have been a greater positive effect on anxiety reduction. The possibility that patients with a tendency to higher trait anxiety scores might experience higher state anxiety scores was also considered, and this was accounted for by undertaking a multivariable regression analysis (Table III). The regression analysis showed that for every 1 point of increase in trait anxiety, the state anxiety score increased on average by 0.393 points; this was signicant (P \0.001). Patients with a tendency toward higher levels of trait anxiety had higher levels of state anxiety, regardless of the intervention. It might be worth considering giving DVDs to those with the greatest likelihood of anxiety. Having accounted for the effect of the underlying trait anxiety, the regression analysis showed that the DVD led to a 3.034-point reduction, on average, in state anxiety. This nding was signicant; however, the average reduction of 3.034 points was still below the clinically relevant difference. The 95% CIs for state anxiety (0.489 to 5.580 points), however, illustrate the overall positive effect of the DVD, and a reduction in state anxiety of up to 5.580 points was feasible. This demonstrates that some patients will experience a clinically relevant reduction in anxiety and conrms the potential benets of providing this type of information.

The adjusted R2 value indicated that only 18.7% of the state anxiety score was affected by the trait anxiety score and the group allocation. This highlights the multidimensional nature of anxiety as previously noted by Newton and Buck.15 It is unsurprising then that Klingberg and Broberg16 concluded that research into anxiety is fraught with methodologic and conceptual problems, and further work is required. Anxiety was not signicantly affected by sex; this contrasts with a prospective cohort study by Tickle et al,1 who found that girls had twice the odds of being dentally anxious than boys. Unlike the self-reported questionnaire used in our study, Tickle et al used a parent-reported questionnaire. The results also indicated no statistically signicant difference between white patients and those of other races and no signicant correlation between anxiety and age. There are, however, few previous studies with which to compare these ndings. Even though the results showed no signicant association between anxiety and age, sex, or ethnicity, as with most clinical decisions, each patient's needs must be assessed individually, and any information provided should be tailored accordingly.
CONCLUSIONS

The results of this study showed that the provision of additional information regarding the bond-up procedure does signicantly reduce state anxiety levels on average, although perhaps not at clinically relevant levels. It might be benecial to provide additional audiovisual information more cost-effectively: eg, on social media sites or in clinical waiting rooms. Sex, ethnicity, and age did not statistically affect anxiety levels before bond-up. Therefore, the rst null hypothesis can be rejected, but there was no evidence to reject the second hypothesis based on our results.
REFERENCES 1. Tickle M, Jones C, Buchannan K, Milsom KM, Blinkhorn AS, Humphris GM. A prospective study of dental anxiety in a cohort of children followed from 5 to 9 years of age. Int J Paediatr Dent 2009;19:225-32. 2. Nuttall NM, Gilbert A, Morris J. Children's dental anxiety in the United Kingdom in 2003. J Dent 2008;36:857-60. 3. Enkling N, Marwinski G, J ohren P. Dental anxiety in a representative sample of residents of a large German city. Clin Oral Invest 2006;10:84-91. 4. Lader D, Chadwick B, Chestnutt I, Harker R, Morris J, Nuttal N, et al. Childrens Dental Health in the United Kingdom, 2003. Summary Report Ofce for National Statistics. London, United Kingdom: Ofce for National Statistics; 2005. Available at: http://www.dh.gov.uk/prod_consum_dh/groups/dh_digitalassets/ @dh/@en/documents/digitalasset/dh_4107310.pdf. Accessed October 8, 2011.

October 2013  Vol 144  Issue 4

American Journal of Orthodontics and Dentofacial Orthopedics

Srai et al

511

5. Bolin AK. Childrens dental health in Europe: an epidemiological investigation of 5- and 12-year old children in eight EU countries. Swed Dent J Suppl 1997;122:1-88. 6. Klingberg G, Berggren U, Carlsson SG, Nor en JG. Child dental fear: cause-related factors and clinical effects. Eur J Oral Sci 1995;103: 405-12. 7. Bedi R, Sutcliffe P, Donnan PT, McConnachie J. The prevalence of dental anxiety in a group of 13- and 14-year old Scottish children. Int J Paediatr Dent 1992;2:17-24. 8. Stephens RM, Ryan FS, Cunningham SJ. Information seeking behavior in adolescent orthodontic patients. Am J Orthod Dentofacial Orthop 2013;143:303-9. 9. Spielberger CD, Gorsuch RL, Lushene R, Vagg PR, Jacobs GA. Preliminary test manual for the state-trait anxiety inventory for children. Palo Alto, Calif: Consulting Psychologists Press; 1973. 10. nQueryAdvisor. Available at: http://www.statistical-solutionssoftware.com/products-page/nquery-advisor-sample-size-software/. Accessed September 23, 2011.

11. Childrens-Mercy.com. Available at: http://www.childrensmercy. org/stats/size/mann.aspx. Accessed September 23, 2011. 12. Kerrigan DD, Thevasagayam RS, Woods TO, McWelch I, Thomas WE, Shorthouse AJ, et al. Who's afraid of informed consent? Br Med J 1993;306:298-300. 13. Olumide F, Newton JT, Dunne S, Gilbert DB. Anticipatory anxiety in children visiting the dentist: lack of effect of preparatory information. Int J Paediatr Dent 2009;19:338-42. 14. Wright NS, Fleming PS, Sharma PK, Battagel J. Inuence of supplemental written information on adolescent anxiety, motivation and compliance in early orthodontic treatment. Angle Orthod 2010;80:329-35. 15. Newton JT, Buck DJ. Anxiety and pain measures in dentistry: a guide to their quality and application. J Am Dent Assoc 2000; 131:1449-57. 16. Klingberg G, Broberg AG. Dental fear/anxiety and dental behaviour management problems in children and adolescents: a review of prevalence and concomitant psychological factors. Int J Paediatr Dent 2007;17:391-406.

American Journal of Orthodontics and Dentofacial Orthopedics

October 2013  Vol 144  Issue 4