HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Riley, M. Articles by Bearn, D. R. Search for Related Content PubMed PubMed Citation Articles by Riley, M. Articles by Bearn, D. R.

A systematic review of clinical trials of aligning archwires

University of Manchester, UK

David R. Bearn

University of Dundee, UK

Address for correspondence: Professor David R Bearn, University of Dundee, UK., Email: d.bearn{at}dundee.ac.uk

Received 25 August 2008; accepted 23 November 2008

	Abstract

Top Abstract Background Objectives Methods Results Discussion Conclusions References

 
Objectives: This review aimed to identify the evidence for the efficacy of archwires used in the alignment stage of orthodontic treatment by undertaking a systematic review of the literature.

Data Sources: MEDLINE, the Cochrane Central Register of Controlled trials (CENTRAL), EMBASE, and the meta Register of Controlled Trials were searched up to July 2008. Reference lists of identified articles and relevant review articles were checked for further possible studies.

Review Methods: Controlled clinical trials and randomised clinical trials that compared aligning archwires and reported objective measures of alignment were selected for inclusion. Validity and quality assessment were undertaken to identify studies with a low risk of bias. Details of the study methodology and the reported results were then abstracted.

Results: 100 studies were identified by the searches and 7 of these were identified as meeting the selection criteria. Four studies were deemed, after quality assessment, to have a low risk of bias and data was extracted from these. No two studies shared a common methodology or common reporting of outcome. Meta-analysis was therefore not possible.

Conclusions: There is insufficient data in these studies to make clear recommendations regarding the most effective archwire for alignment. Recommendations on future study design have been made.

Key words: Orthodontics, archwires, systematic review

	Background

Top Abstract Background Objectives Methods Results Discussion Conclusions References

 
The aligning efficacy and effectiveness of many archwire materials and dimensions used on a daily basis by orthodontists have not been scrutinized by clinical trial. Proffit¹ described the ideal aligning archwire as one which has excellent strength, excellent springiness and a long range of action with force values of about 50 g (the optimum force for tipping the teeth into alignment). In addition to this, Kusy² suggested that such wires should be aesthetic and biocompatible. The archwire materials used to achieve levelling and alignment at the present time meet most of these ideals except for aesthetics and occasional biocompatibility problems (Nickel allergy). The commonly used materials currently used fall into two broad categories; stainless steel and nickel titanium. Newer materials which add Kusy’s demand for aesthetics have also been developed.

The range of archwires available have been subjected to in vitro laboratory testing to ascertain their physical properties by too many investigators to list here. However, in vitro tests, no matter how complex will never be able to fully reproduce or predict the performance of an aligning archwire in clinical practice. To understand the clinical trials are required. However in 1996 Evans commented on this paucity of clinical evidence and added that manufacturers were in such a headlong rush to produce the ultimate aligning archwire that very little attention had been paid to the in vivo behaviour of these materials.³ This review aimed to see if more than a decade later the situation had improved.

	Objectives

Top Abstract Background Objectives Methods Results Discussion Conclusions References

 
To systematically evaluate all clinical trials which investigate the effectiveness of archwires for alignment and levelling, and if possible identify the most effective archwire through the use of meta-analysis. This review is reported according to the QUORUM statement recommendations.

	Methods

Top Abstract Background Objectives Methods Results Discussion Conclusions References

 
Searching
A simple search of, for example, Medline is generally not considered adequate⁴ and so a number of sources were searched to retrieve the relevant literature:

Electronic databases. All searches were initially conducted in January 2005 and no language restrictions were applied.

Two basic sets of terms were applied in the search. Firstly those terms used to identify records related to the health condition of interest (orthodontic alignment), and secondly those terms used to identify records related to the intervention being evaluated (archwire therapy). It is generally standard procedure to include a third basic set of terms which identifies those records related to the type of study designs to be included. This was not included in our final search because in pilot runs of the strategy it reduced the number of papers found by all databases to zero. Details of the searches of electronic databases are shown in Table 1.

Other sources (reference lists). The bibliographies of papers and review articles identified were checked for studies published outside the electronically searched journals which may have been otherwise missed.

Selection
Study design. It was decided to include both randomized clinical trials and controlled clinical trials to have the best chance of finding all evidence of an acceptable level.

Participants. Children and adults who had aligning and/or levelling archwires used as part of orthodontic treatment.

Intervention. Fixed orthodontic appliances consisting of brackets and archwires to achieve aligning and levelling. The type of archwires investigated would be used to put studies into homogenous groups, where applicable, for meta-analysis.

Outcome measures. Studies needed to report an objective measurement of alignment/irregularity to be included.

Validity assessment
Decisions on validity and quality assessment were made independently by two assessors. The factors which were considered when assessing the quality of the studies were:

• Was the sample size reported?
  
• Was the sample size based on a power calculation?
  
• Were the eligibility criteria described?
  
• Was assignment to groups random?
  
• Was treatment allocation concealed?
  
• Were groups similar at baseline in terms of prognostic factors?
  
• Was the care provider blinded?
  
• Were the patients blinded?
  
• Were outcome assessors blinded to treatment allocation?
  
• Were the point estimates and measure of variability presented for the primary outcome measure?
  
• Were the statistical methods used to compare the groups appropriate?
  
• Did the analysis include an intention to treat analysis?
  

On the basis of these questions, a score out of 12 was calculated for each study. A consensus meeting was then arranged to discuss which studies showed low risk of bias and should therefore be included for data extraction.

Data abstraction
Two assessors independently extracted data from the eligible trials and this was entered into customized data abstraction forms. The following details were recorded:

General details

• Name of paper (Author and year)
  
• Name of assessor
  

Care setting

• Location of care setting
  

Participant details

• Age
  
• Gender
  
• Number of subjects in each group
  

Methodological quality

• Study design (RCT or CCT)
  

Interventions

• Types of archwire investigated
  

Outcome measures

• Which teeth were measured?
  
• How the teeth were were measured at baseline?
  
• What was measured after the intervention?
  
• What was used to take the measurements?
  

Results

• How much alignment was achieved with each archwire?
  
• Data was extracted according to the authors’ method of assessment of alignment. This meant there were three broadly different possible types of data to extract for each trial:
  1. What was the change in total irregularity for the given period of time?
     
  2. What was the change in contact point movement for the given period of time?
     
  3. What was the duration taken to achieve 2 mm irregularity index?
     
  
  
• How many drop outs were there and what was the duration of the trial?
  

Meta-analysis. Once data extraction was completed, it was planned to enter this into Revman to undertake a meta-analysis and produce forest plots showing the overall effect of the archwire interventions.

	Results

Top Abstract Background Objectives Methods Results Discussion Conclusions References

 
The study flow is shown in Figure 1 as described in the QUORUM statement.

View larger version (23K): [in this window] [in a new window]   Figure 1 Study flow, as described in QUORUM statement

	Discussion

Top Abstract Background Objectives Methods Results Discussion Conclusions References

 
There have been only six controlled clinical trials of aligning archwires up to the time of the final search.^5–9,11 These trials aimed to test whether there was a difference in time taken to achieve alignment between archwires and compared either a multistranded stainless steel wire and nickel titanium archwire, or different types of NiTi archwire. Prior to these trials (pre-1990), various archwires were being used for levelling and aligning without having been scrutinized by clinical trial.

A summary of all of the controlled clinical trials of aligning archwires can be seen in Table 1.

Care setting
Samples seem to have been drawn exclusively from university teaching hospitals but it was only Cobb et al.⁸ that actually made this explicit. When trials are undertaken in such a setting, it has been suggested that they might not be relevant to the real world as they measure the efficacy of the intervention rather than its effectiveness.¹² If patients are seen in less ‘ideal’ conditions in orthodontic practices than those in teaching hospitals with respect to archwire placement then this suggestion may be true. Conversely as many university departments are training establishments, it is likely that the operators in this setting, if trainees, were less experienced. The number of weeks between each adjustment of the appliance could well be longer in government funded practice than university teaching hospitals and this may have yielded different results between the archwires had the trials been carried out there instead.

Sample size
Sample sizes have ranged from 15–123 patients or 15–158 arches. Only two of the six trials reported power calculations, West et al.⁷ and Evans et al.¹¹ Despite a power calculation, Evans et al. had a sample size below the threshold which had been set after drop outs. It is not known whether West et al. met their power calculation threshold as drop outs were not published. It is possible that samples in the remaining studies lacked the power to detect differences between the archwires.

Age and gender of subjects
Where reported, the ages of subjects in the trials were broadly similar.^7,8 The gender distribution was not widely reported – in fact, only West et al.⁷ reported it, where females outnumbered males by 2 : 1. There is, however, no reason to suspect that there would be a difference in archwire performance between the genders so the absence of this data is not critical.

Quality assessment
The purpose of the quality assessment was to identify possible sources of bias, and make a judgment based on this as to which studies to include. The cut off score of 6 which was used is to some degree arbitrary, but in order to make the rationale behind this clear the possible biases identified are listed and discussed here.

Design of studies. This systematic review sought to find only controlled clinical trials and randomized controlled clinical trials. Given the number of different types of archwires currently available, it was somewhat surprising that only six of these types of trials were found (five randomized clinical trial and one controlled clinical trial).

Sample size and eligibility criteria. All six trials reported their sample size but only two of these samples had been based on power calculations. Eligibility criteria were described in all trials except Dalstra et al.⁹

Was assignment to groups random? Random allocation to treatment group is important because it gives groups that are likely to be balanced for known as well as unknown confounding variables. Jones et al.’s study⁶ was not a randomized controlled trial and was therefore susceptible to selection bias. Dalstra et al. used a spit mouth design with random allocation to quadrants within each subject.

Was treatment allocation concealed? Without concealment of the randomly generated allocation sequence, the operator may choose to manipulate it so that subjects are allocated to the group that the operator chooses. This is called selection bias and is one of the most important factors which may distort treatment comparisons.¹³ None of the six trials had their treatment allocation reported as concealed.

Were the groups similar at baseline in terms of prognostic factors? It would seem important that the groups were comparable at baseline with respect to alignment as otherwise differences between groups after intervention may be attributed (wrongly) to the archwire rather than the fact that the groups were not comparable at baseline. Only O’Brien et al.⁵ and Cobb et al.⁸ confirmed that both groups were similar in terms of alignment at baseline.

Was the care provider blinded? Blinding of the care provider is necessary as otherwise performance bias (a systematic difference in care provided apart from the intervention being evaluated) is possible. The operators were not blind in any of the trials leading to the possibility of performance bias. However, it would have been very difficult in most trials to blind the operator as the wires either had a different number of strands^6–8,11 or were a different colour.⁹ In O’Brien et al.’s study⁵ blinding would have been possible, although the operator may have realized which wire they were using from its handling characteristics.

Were the patients blinded? Blinding of patients to their group allocation helps to stop reporting bias. None of the trials had patient blinding but reporting bias is not a risk for these studies as the patients are not reporting the primary outcome.

Were the outcome assessors blinded to treatment allocation? Detection bias is a systematic difference between comparison groups in how outcomes are ascertained and can be minimized by blinding the outcome assessors.

The assessment of irregularity was not blind in three of the studies (Cobb et al.,⁸ West et al.⁷ and Dalstra et al.⁹). Cobb et al. measured Little’s irregularity index directly on the patients allowing the operator to identify the archwire in use. West et al. digitized contact point displacement on study models which included the archwire – the multistranded could have been differentiated from the single stranded archwires. Dalstra et al. took an occlusal photograph on which the type of wire could be identified. In all of these trials there was the possibility of detection bias.

Were the point estimates and measure of variability presented for the primary outcome measure? Without the mean and a measure of variability for the primary outcome measure, data cannot be extracted and a meta-analysis performed. All studies included such data and from this point of view were admissible to the data abstraction stage.

Were the statistical methods used to compare the groups appropriate? Broadly speaking, the statistical methods used to compare the groups were appropriate. However, only Cobb et al.⁸ and Dalstra et al.⁹ analysed the distribution of the data. On both occasions this showed that alignment time/extent was not normally distributed. Consequently non parametric statistics were used. In the remaining four trials parametric statistics were used without confirming their suitability with distribution analysis.

Did the analysis include an intention to treat analysis? An intention to treat analysis is one where participants are analysed according to the group to which they were initially allocated. It is important to analyse data on an intention to treat basis because otherwise attrition bias (systematic differences between comparison groups in terms of withdrawals or exclusions of participants) is possible. O’Brien et al.⁵ Jones et al.⁶ and Dalstra et al.⁹ analysed data on an intention to treat basis but this was only by virtue of the fact that there were no drop outs. In the remaining trials, West et al.⁷ did not report on whether there were any drop outs and Evans et al.¹¹ had drop outs but omitted them from analysis.

Intervention. A summary of the different interventions in each of the four trials finally included can be seen in Table 5. Investigation of the same intervention by more than one author would potentially make a meta-analysis of this intervention possible. 0.016 NiTi was investigated by both O’Brien et al. and Cobb et al. whilst 0.0155 stainless steel was investigated by both West et al. and Evans et al. However, a meta-analysis is only possible if the outcomes measured and reported are the same.

Assessment of archwire performance. The extent of alignment achieved by each archwire was assessed in three broadly different ways:

1. 3D contact point position with respect to palatal rugae: O’Brien et al.⁵ assessed archwire performance by measuring the 3D movement of each anatomical contact point. This was achieved by digitizing each anatomical contact point with respect to the medial palatal rugae at the beginning of the trial and then repeating these measurements at the end. The mean contact point movement was then calculated for each archwire.
   Whilst the internal validity of this form of assessment appears to be good, the external validity is perhaps not as good. As a purchaser of archwires, an operator might want to know whether alignment with one archwire is faster than another. The form of assessment used by O’Brien et al. will tell us whether contact points move faster with one archwire than another but it won’t tell us how many weeks faster this translates to.
   
2. Sum of contact point displacements: Jones et al.,⁶ West et al.⁷ and Cobb et al.⁸ all summated anatomical contact point displacement to give a sum of contact point displacement both before and after archwire therapy.
   Evans et al.¹¹ measured the displacement of incisal edges anteriorly whilst effectively measuring inter-bracket span posteriorly. This irregularity was then summated to give a sum of ‘inter-tooth distances’ before and after archwire therapy. The reason given for measuring irregularity in this way was that it reduced the measurement error posteriorly.
   The internal validity of the assessment of alignment used by Jones et al., West et al. and Cobb et al. is good; it quantifies the amount of irregularity before and after archwire therapy. However, the internal validity of Evans et al.’s assessment of alignment is questionable. Whilst measuring the inter-bracket span of posterior teeth may reduce measurement error, this is at the cost of not actually measuring the irregularity of the teeth themselves. For example, when considering correction of an individual tooth rotation during alignment, it is conceivable that the sum of ‘inter-tooth distances’ would remain the same despite the fact that considerable alignment has been achieved. A common example where this may occur is when two premolars are rotated towards each other prior to alignment, producing a very short inter-bracket span.
   The external validity of assessing archwire performance by the degree of change in the sum of irregularity over time is reasonable but it is open to the same criticism as O’Brien et al.’s methodology – it does not tell us how many weeks will be saved by using one archwire over another. Cobb et al.⁸ used the time taken to achieve 2 mm total irregularity as their endpoint. This approach has more external validity in that one can then tell how much time might be saved between archwires to reduce total initial irregularity to 2 mm.
   
3. Computer aided photographic analysis: Dalstra et al. assessed archwire performance in their split mouth study by measuring irregularity on each side of the mouth before and after archwire therapy. This was achieved by taking intra-oral photographs of the occlusal aspect of the upper arch at the beginning and end of alignment and analysing them. The photographs were taken with a custom made plexiglass plate fitted over the stable medial palatal rugae which had a grid of perpendicular lines drawn on it. The x and y coordinates of the bracket wings relative to the grid were measured pre and post alignment and this movement was compared between sides. There was no investigation of the validity or reproducibility of this technique.
   

Data abstraction
Duration of trial. The period over which alignment was assessed varied considerably from 21 to 56 days. Only Cobb et al. used a clinical end point for initial alignment (2 mm Little’s irregularity index) rather than an arbitrary period of time. They found the median time to initial alignment was 51 days. Only Evans et al.¹¹ had a follow up of more than this and so the remaining trials^5–7 may not have fully reflected the alignment phase.

Meta-analysis. A meta-analysis was not possible because even though the same archwires were assessed in multiple studies, there was no homogeneity in the way that their performance was reported; no one methodology was used more than once. In addition to this, data could not be extracted from either West et al. or Evans et al.’s papers. In the case of West et al., data were presented as geometric mean ratios of total irregularity of NiTi/Multistranded stainless steel with no data on the actual amount of irregularity for each archwire at each interval. In the case of Cobb et al., the data was only presented in graphical form with insufficient detail for data to be extracted.

Only one of the four trials showed a significant difference between the wires studied and on both occasions the clinical significance was questionable; West et al.⁷ showed that NiTi was significantly quicker at aligning teeth in the lower labial segment than multistranded stainless steel although the clinical significance of this difference appears to be small.

Evans et al. suggested that in the case of multistranded stainless steel and NiTi, differences may not have been found because the NiTi wires were not clinically deformed enough to take advantage of their superelastic properties.

In summary, it would appear that there have been little clinically significant differences found between the archwires studied.

	Conclusions

Top Abstract Background Objectives Methods Results Discussion Conclusions References

 
Seven clinical trials have been undertaken investigating aligning archwires. After quality assessment, four were selected for data extraction but due to a lack of homogeneity, a meta-analysis was not possible. There is insufficient data in these studies to make clear recommendations regarding the most effective archwire for alignment.

Of the seven clinical trials evaluating aligning arch-wires, none have been ideal. It would seem that the best trial design to evaluate the clinical performance of aligning archwires would:

• be a randomized controlled trial;
  
• standardized appliance variables other than the archwires;
  
• have a power calculation to establish the sample size;
  
• have adequate randomization, allocation concealment and blinding where possible to protect against bias;
  
• use a valid, reproducible, quick and well recognized assessment of alignment such as Little’s irregularity index;
  
• have a long enough follow up to accurately reflect the alignment phase. This could be a clinically defined endpoint e.g. 2 mm irregularity rather than an empirical time point;
  
• use an intention to treat analysis to protect against attrition bias;
  
• include an analysis of the distribution of data so that the most suitable statistical tests are used;
  
• report data in its raw form to make it more amenable to meta-analysis.
  

	References

Top Abstract Background Objectives Methods Results Discussion Conclusions References

 
1 Proffit WR. Contemporary Orthodontics, 3rd Edn. London: Mosby, 2000.

2 Kusy RP. The future of orthodontic materials: the long-term view. Am J Orthod Dentofacial Orthop 1998; 113(1): 91–95.[CrossRef][Medline]

3 Evans TJ, Durning P. Aligning archwires, the shape of things to come? – a fourth and fifth phase of force delivery. Br J Orthod 1996; 23(3): 269–75.[Abstract]

4 Higgins JPT, Green S, editors. Cochrane Handbook for Systematic Reviews of Interventions 4.2.6. Chichester: John Wiley and Sons Ltd, 2006.

5 O’Brien K, Lewis D, Shaw W, Combe E. A clinical trial of aligning archwires. Eur J Orthod 1990; 12(4): 380–84.[Abstract/Free Full Text]

6 Jones ML, Staniford H, Chan C. Comparison of superelastic NiTi and multistranded stainless steel wires in initial alignment. J Clin Orthod 1990; 24(10): 611–13.[Medline]

7 West AE, Jones ML, Newcombe RG. Multiflex versus superelastic: a randomized clinical trial of the tooth alignment ability of initial arch wires. Am J Orthod Dentofacial Orthop 1995; 108(5): 464–71.[CrossRef][Medline]

8 Cobb NW, 3rd, Kula KS, Phillips C, Proffit WR. Efficiency of multi-strand steel, superelastic Ni-Ti and ion-implanted Ni-Ti archwires for initial alignment. Clin Orthod Res 1998; 1(1): 12–9.[Medline]

9 Dalstra M, Melsen B. Does the transition temperature of Cu-NiTi archwires affect the amount of tooth movement during alignment? Orthod Craniofac Res 2004; 7(1): 21–25.[CrossRef][Medline]

10 Pandis N, Polychronopoulou A, Eliades T. Self-ligating vs conventional brackets in the treatment of mandibular crowding: a prospective clinical trial of treatment duration and dental effects. Am J Orthod Dentofacial Orthop 2007; 132(2): 208–15.[CrossRef][Medline]

11 Evans TJ, Jones ML, Newcombe RG. Clinical comparison and performance perspective of three aligning arch wires. Am J Orthod Dentofacial Orthop 1998; 114(1): 32–39.[CrossRef][Medline]

12 O’Brien K, Wright J, Conboy F, et al. Effectiveness of early orthodontic treatment with the Twin-block appliance: a multicenter, randomized, controlled trial. Part 1: Dental and skeletal effects. Am J Orthod Dentofacial Orthop 2003; 124(3): 234–43; quiz 339.[CrossRef][Medline]

13 Kunz R, Oxman AD. The unpredictability paradox: review of empirical comparisons of randomised and non-randomised clinical trials. BMJ 1998; 317(7167): 1185–90.[Abstract/Free Full Text]

14 Richmond S. Feasibility of categorising treatment difficulty – the use of three dimensional plotting. Cardiff: University of Wales, 1984.

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Riley, M. Articles by Bearn, D. R. Search for Related Content PubMed PubMed Citation Articles by Riley, M. Articles by Bearn, D. R.