Intervenciones para la interrupción de los hábitos de succión no nutritivos en niños
Resumen
Antecedentes
Los comportamientos reconfortantes, como el uso de chupetes, mantas y la succión del dedo o el pulgar, son habituales en los recién nacidos y en los niños pequeños. Estos hábitos reconfortantes, a los que se puede hacer referencia como “hábitos de succión no nutritivos” (HSNN), tienden a finalizar a medida que los niños crecen, bajo su propio impulso o con el apoyo de los padres y los cuidadores. Sin embargo, si el hábito continúa mientras se está estableciendo la dentición permanente, puede contribuir a, o causar, el desarrollo de una maloclusión (mordida anormal). Se ha utilizado una variedad diversa de estrategias para ayudar a los niños a interrumpir los HSNN. Las mismas incluyen asesoramiento, eliminación del objeto reconfortante, uso de aparatos ortodónticos para interferir con el hábito, aplicación de un sabor aversivo en el dedo o técnicas de modificación de la conducta. Algunas de estas intervenciones son más fáciles de aplicar que otras y menos molestas para el niño y el padre; algunas son aplicables a un tipo de hábito concreto.
Objetivos
El objetivo principal de la revisión fue evaluar los efectos de diferentes intervenciones para la interrupción de los HSNN en los niños. Los objetivos secundarios incluyeron la posibilidad de determinar qué intervenciones funcionan más rápidamente y son las más efectivas en cuanto a los desenlaces centrados en los niños y los padres y cuidadores de menor malestar y angustia psicológica con la intervención, así como las medidas dentales de la maloclusión (reducción de la mordida abierta anterior, resalte y corrección de la mordida cruzada posterior) y la relación entre coste y efectividad.
Métodos de búsqueda
Se realizaron búsquedas en las siguientes bases de datos: Registro de ensayos del Grupo Cochrane de Salud Oral (Cochrane Oral Health Group) (hasta 8 de octubre de 2014), Registro Cochrane central de ensayos controlados (Cochrane Central Register of Controlled Trials; CENTRAL) (The Cochrane Library 2014, número 9), MEDLINE vía OVID (1946 hasta 8 octubre 2014), EMBASE vía OVID (1980 hasta 8 octubre 2014), PsycINFO vía OVID (1980 hasta 8 de octubre de 2014) y en CINAHL vía EBSCO (1937 hasta 8 de octubre de 2014), el US National Institutes of Health Trials Register (Clinical Trials.gov) (hasta 8 de octubre de 2014) y en la Plataforma de registros internacionales de ensayos clínicos de la OMS (hasta 8 de octubre de 2014). En las búsquedas en las bases de datos electrónicas, no hubo restricciones de idioma ni de fecha de publicación. Se revisaron las listas de referencias de los artículos pertinentes y se contactó con los autores de los estudios elegibles para obtener información adicional cuando fue necesario.
Criterios de selección
Ensayos controlados aleatorizados o cuasialeatorizados en niños con hábitos de succión no nutritivos que compararan una intervención con otra o un grupo de control sin intervención. El desenlace principal de interés fue el abandono del hábito.
Obtención y análisis de los datos
Se utilizaron los procedimientos metodológicos estándar previstos por la Colaboración Cochrane. Tres autores de la revisión participaron en el cribaje de entradas identificadas; dos realizaron la extracción de datos, dos evaluaron el riesgo de sesgo y dos evaluaron la calidad general de la base de evidencia. La mayor parte de los datos no se pudieron combinar y sólo se pudo realizar un metanálisis.
Resultados principales
Se incluyeron seis ensayos con 252 niños (de 2 años y medio a 18 años de edad), aunque presentaron datos de seguimiento de sólo 246 niños. La succión del dedo fue el único HSNN evaluado en los estudios. Cinco estudios compararon intervenciones únicas o múltiples con ninguna intervención o un grupo de control en lista de espera y un estudio realizó una comparación directa. Todos los estudios se encontraban en alto riesgo de sesgo debido a las importantes limitaciones en la metodología y la presentación de informes. Había un pequeño número de participantes en los estudios (20 a 38 participantes por estudio) y los tiempos de seguimiento variaron de uno a estudio 36 meses. Los desenlaces a corto plazo se observaron antes del año tras la intervención, y los desenlaces a largo plazo se observaron al cabo de un año o más de la intervención.
Aparatos ortodónticos (con o sin intervención psicológica) versus ningún tratamiento
Dos ensayos que evaluaron esta comparación evaluaron el desenlace principal del abandono del hábito. Uno de los ensayos evaluó la rejilla palatina y uno utilizó una combinación de rejillas y arcos palatinos. Ambos ensayos presentaron alto riesgo de sesgo. El aparato ortodóntico presentó una probabilidad mayor de interrumpir la succión del dedo que ningún tratamiento, ya sea con el uso a corto plazo (razón de riesgos [RR] 6,53; intervalo de confianza [IC] del 95%: 1,67 a 25,53; dos ensayos, 70 participantes) o a largo plazo (RR 5,81; IC del 95%: 1,49 a 22,66; un ensayo, 37 participantes) o utilizado en combinación con una intervención psicológica (RR 6,36; IC del 95%: 0,97 a 41,96; un ensayo, 32 participantes).
Intervención psicológica versus ningún tratamiento
Dos ensayos (78 participantes) en riesgo alto de sesgo evaluaron el refuerzo positivo (solo o en combinación con el logro de la cooperación del niño) o el refuerzo negativo en comparación con ningún tratamiento. El agrupamiento de los datos mostró una diferencia estadísticamente significativa a favor de las intervenciones psicológicas a corto plazo (RR 6,16; IC del 95%: 1,18 a 32,10; I2 = 0%). Un estudio, con datos de 57 participantes, informó sobre el efecto a largo plazo del refuerzo positivo y negativo sobre la interrupción de la succión del dedo y encontró una diferencia estadísticamente significativa a favor de las intervenciones psicológicas (RR 6,25; IC del 95%: 1,65 a 23,65).
Comparaciones directas
Sólo un ensayo demostró una diferencia clara en la efectividad entre las diferentes intervenciones activas. Este ensayo, que tuvo sólo 22 participantes, encontró una probabilidad mayor de interrupción del hábito con la rejilla palatina que con el arco palatino (RR 0,13; IC del 95%: 0,03 a 0,59).
Conclusiones de los autores
Esta revisión halló evidencia de calidad baja de que los aparatos ortodónticos (arco palatino y rejilla palatina) y las intervenciones psicológicas (incluido el refuerzo positivo y negativo) son efectivos para mejorar la interrupción de la succión en los niños. Hay evidencia de calidad muy baja de que la rejilla palatina es más efectiva que el arco palatino. Esta revisión ha destacado la necesidad de realizar ensayos de alta calidad que evalúen las intervenciones para detener los hábitos de succión no nutritivos y la necesidad de un método consolidado y estandarizado de informe de los desenlaces en estos ensayos.
PICO
Resumen en términos sencillos
Intervenciones para la interrupción de los hábitos de succión del dedo o pulgar o chupete en los niños
Pregunta de la revisión
Esta revisión se propuso evaluar los efectos de las intervenciones para detener los hábitos de succión, no vinculados a los alimentos, en los niños. Algunas consideraciones importantes son: qué tratamiento o combinación de tratamientos funciona de forma más efectiva, cuándo debe comenzar el tratamiento, cuál es el período de tiempo óptimo para la intervención y qué causa menos molestias a los niños y los padres.
Antecedentes
A menudo los recién nacidos y los niños desarrollan el hábito de succión de objetos lo cual los reconforta y los calma. Con frecuencia succionan chupetes, los dedos, los pulgares u otros elementos como las mantas. Con el tiempo, la mayoría de los niños superan el hábito, o lo abandonan debido al estímulo de los padres. Algunos niños, sin embargo, continúan succionando como un hábito. Si continúan haciéndolo cuando comienzan a aparecer los dientes permanentes (alrededor de la edad de seis años), existe el riesgo de que dichos dientes permanentes crezcan en la posición incorrecta, lo cual causa que sobresalgan demasiado o que no se junten de forma adecuada al morder. Como resultado, estos niños a menudo necesitan tratamiento dental para resolver los problemas causados por el hábito de succión.
Los tratamientos posibles para ayudar a los niños a interrumpir sus hábitos de succión examinados en los estudios de esta revisión incluyen el uso de dos aparatos ortopédicos diferentes en la boca; proporcionar asesoramiento e incentivos para cambiar su conducta (conocido como orientación/tratameinto psicológico); aplicar una sustancia con sabor amargo y desagradable a pulgar o los dedos de los niños o estos tratamientos combinados. Ninguno de los estudios incluidos consideró los métodos de barrera, por ejemplo el uso de guantes o vendas o el retiro de los chupetes.
Características de los estudios
Los autores de la revisión del Grupo Cochrane de Salud Oral realizaron esta revisión de los estudios existentes y la evidencia está actualizada hasta el 8 de octubre de 2014. La revisión incluye seis estudios publicados desde 1967 hasta 1997, que involucraron a 252 niños como participantes (aunque solo se proporcionaron datos de 246 de los niños). Tres de los estudios se realizaron en los EE.UU., uno en Canadá, uno en Suecia y uno en Australia.
No todos los estudios proporcionaron las edades de los niños involucrados; en cuatro de los estudios los niños tenían de dos años y medio a 18 años de edad, en un estudio tenían a partir de cuatro años de edad y en otros a partir de nueve años de edad.
Resultados
El uso de un aparato ortodóntico (como una rejilla o arco palatino) o una intervención psicológica (como el uso de refuerzo positivo o negativo), o ambos, tuvo más probabilidades de dar lugar al abandono del hábito que ningún tratamiento. La mayoría de los ensayos que compararon dos intervenciones diferentes no fueron concluyentes, aunque un estudio indicó que, de dos tipos diferentes de aparato ortopédico, la rejilla palatina es más efectiva que un diseño de arco palatino.
Calidad de la evidencia
La evidencia presentada es de baja calidad debido al pequeño número de participantes en los pocos estudios disponibles y a los problemas en la forma de realizar los estudios. Hubo un alto riesgo de sesgo en los estudios.
Conclusión
Los aparatos ortodónticos o la intervención psicológica parecen ser efectivos para ayudar a los niños a abandonar la succión que no tiene una finalidad de alimentación, aunque la evidencia es de baja calidad. Se necesitan más ensayos clínicos de alta calidad para guiar la toma de decisiones sobre lo que constituye un problema común que puede requerir un tratamiento dental prolongado y costoso para su corrección.
Authors' conclusions
Summary of findings
| Orthodontic appliances compared with no treatment for the cessation of pacifier or digit sucking habits in children | ||||||
|---|---|---|---|---|---|---|
| Patient or population: children with pacifier or digit sucking habits Settings: home Intervention: orthodontic appliances (palatal arch/palatal crib) Comparison: no treatment | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | Number of participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| No treatment | Orthodontic appliance | |||||
| Sucking cessation (short term) | 7 per 100 | 46 per 100 | RR 6.53 (1.67 to 25.53) | 70 (2) | Lowa,b | Long‐term data also show orthodontic appliances to be beneficial when compared with no treatment |
| Occlusion (changes in overbite) | MD 4.10 (2.93 to 5.27) | 24 (1) | Very lowa,b,c | Difference in mean net change in overbite between no treatment (‐0.4 mm) and orthodontic appliance (3.7 mm) = 4.1 mm | ||
| Adverse events | Insufficient information reported | |||||
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| aDowngraded due to risk of bias. | ||||||
| Psychological interventions compared with no treatment for the cessation of pacifier or digit sucking habits in children | ||||||
|---|---|---|---|---|---|---|
| Patient or population: children with pacifier or digit sucking habits Settings: home Intervention: psychological interventions Comparison: no treatment | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | Number of participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| No treatment | Pyschological interventions | |||||
| Sucking cessation (short term) | 3 per 100 | 19 per 100 | RR 6.16 (1.18 to 32.10) | 87 (2) | Lowa,b | Long‐term data also show psychological interventions to be beneficial when compared with no treatment |
| Occlusion | Not reported | |||||
| Adverse events | Insufficient information reported | |||||
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| aDowngraded due to risk of bias. | ||||||
Background
Description of the condition
The term 'non‐nutritive sucking habit' (NNSH) encompasses the use of pacifiers ('dummies', 'soothers'), blankets and digit sucking. Although the incidence of sucking habits varies considerably between different countries, these comforting habits are common in children in many populations. A Swedish study looked at 60 consecutive births and found the incidence of NNSH to be 82% during the first five months of life (Larsson 2001) and a USA‐based study reported the incidence as 73% for a group of 130 children between two and five years of age (Adair 1992). The incidence of NNSH reduces with age. Available data has shown that around 48% of four‐year‐olds maintain a digit or pacifier sucking habit (Modeer 1982), 12.1% of children past the age of seven years (Patel 2008), reducing to 1.9% of children by 12 years of age (Baalack 1971).
Sucking is one of the earliest reflexes exhibited and is a very strong urge in young babies. Sucking behaviours are very common in babies and young children as they give a feeling of security and many parents introduce the use of pacifiers to babies to help them settle. There are other positive effects of pacifiers in young children, which may also contribute to their widespread use. The use of pacifiers has been shown to reduce crying in children during painful experiences such as venepuncture, and has been advocated for use in paediatric emergency departments (Blass 1999; Curtis 2007). Pacifier placement for babies going to sleep has also been identified as a factor in the reduction of sudden infant death syndrome (SIDS) (Hauck 2005; Li 2006), although the mechanism behind this is currently unknown. However, the incidence of SIDS is highest at two to four months and declines towards zero at one year of age and therefore is not a reason for encouraging continuation of pacifier use in children over the age of two and a half years, when primary teeth are all present in the mouth.
Children with a history of a NNSH are more likely to develop a malocclusion compared to children with no NNSH history (Bowden 1966; Farsi 1997; Fukuta 1996; Mistry 2010; Svedmyr 1979; Vazquez‐Nava 2006). In addition, there is evidence that the more prolonged the duration of the habit, the more severe the developing malocclusion tends to be (Baalack 1971; Singh 2008; Warren 2002). However, rather than there being a direct cause‐effect relationship between NNSH and development of a malocclusion, the effects of a habit seem to be superimposed on genetic predispositions to a malocclusion. Therefore, the NNSH might worsen or, conversely, counteract an underlying malocclusion and lead to an improvement. For example, in a child who has a Class III incisor relationship, a NNSH may push the upper anterior teeth forwards and the lower ones backwards, resulting in a less severe malocclusion.
Although NNSHs do not inevitably lead to a predictable malocclusion, different sucking habits generally have different effects on the position of the teeth. A malocclusion can develop through use of a NNSH, through application of pressure by the object or digit on the teeth, interfering with their normal path of eruption. Prolonged pacifier habits are associated with the development of posterior crossbites and prolonged digit habits with increased overjet (Bishara 2006; Ogaard 1994; Warren 2002), and both are associated with an increased prevalence of reduced overbite and anterior open bite (Warren 2002). Children with an increased overjet and incompetent lips (often associated with an anterior open bite) are at greater risk of dental trauma due to the prominence of the upper teeth and lack of protection from the lips (Burden 1995; Glendor 2009; Norton 2012). Incompetent lips and prominent upper anterior teeth are both associated with poor facial aesthetics. Speech can also be affected by tooth position. Laine 1987 found a significant relationship between increased overjet and distortions of the “s” sound and Bernstein 1954 noted that speech is commonly defective where there is an anterior open bite, often presenting with a lisp. There have also been reports of digit deformities developing as a result of prolonged digit sucking requiring surgical correction (Reid 1984), although these are uncommon.
If a NNSH continues while the permanent dentition is establishing, it may be associated with a malocclusion that will require fixed orthodontic appliances (Figure 1), resulting in time consuming, complex and costly treatment required to be carried out by a specialist orthodontist (Greenlee 2011; Petren 2008; Sandler 2011).
Orthodontic deterrent appliance
A number of different interventions have been described in the literature to assist the child who wishes to stop the habit and to support parents who seek advice on this. However, it is not known which is most effective, or even if they are effective, or which are favoured by children and parents.
Description of the intervention
There is no standard intervention for cessation of NNSHs. A wide variety of different approaches and interventions have been described, which range from removal of the comforting object, through fitting an orthodontic appliance to directly interfere with the habit, application of an aversive tasting substance to the digit, to behaviour modification techniques (Al‐Jobair 2004; Friman 1986). Some of the interventions are easier to apply than others, less disturbing for the child and their parent or carer and certain ones are likely to be more applicable to a particular type of habit.
The interventions are likely to differ with respect to their:
-
effectiveness in habit cessation;
-
ease for children to cope with and ease of implementation from a parent/carer perspective;
-
time to stop the NNSH; and,
-
reduction in severity of the malocclusion.
How the intervention might work
The different ways in which the interventions might work depend on the habit and the type of intervention. Where the habit involves an object (blanket, pacifier etc), its removal will stop the habit (or lead to it being replaced by another). For habits that involve digit sucking, there are a number of different types of intra‐oral appliances to prevent placement of the digit in the habit position. Other appliances prevent the sense of gratification that the child feels through carrying out the habit although the digit can still be sucked. Other approaches involve replacing the feeling of comfort with an unpleasant stimulus such as an aversive taste. Behavioural modification techniques such as cognitive behavioural therapy, reward‐based strategies or use of positive reinforcement can also be employed.
Why it is important to do this review
NNSHs are common and this is a topic of significant interest to parents. There is a need to determine the most effective and timely management option(s) for cessation of NNSHs and it is important that consideration is also given to those associated with the least distress for children and their parents or carers. There is a wide variety of treatment strategies available to help children and parents with stopping sucking habits, but a lack of clarity about relative effectiveness and side effects. The aim of this review is to draw together the evidence and identify which interventions are the most successful.
Objectives
Primary objective
To evaluate the effects of different interventions for cessation of non‐nutritive sucking habits in children.
Secondary objectives
To determine which interventions work most quickly and are the most effective in terms of child and parent‐ or carer‐centred outcomes of least discomfort and psychological distress from the intervention, as well as the dental measures of malocclusion (reduction in anterior open bite, overjet and correction of posterior crossbite) and cost‐effectiveness.
Methods
Criteria for considering studies for this review
Types of studies
Randomised controlled clinical trials (RCTs) and quasi‐randomised controlled clinical trials comparing an intervention for cessation of non‐nutritive sucking habits with either a different intervention(s) or no treatment or control.
Types of participants
Children (up to age 18 years of age) who have a digit sucking habit or any other NNSH, including a pacifier habit (dummy).
Types of interventions
For the intervention group we looked at:
-
orthodontic appliances;
-
barrier techniques ‐ gloves/plasters etc.;
-
chemical techniques ‐ topical substances applied to pacifier or digit;
-
behaviour modification techniques;
-
non‐treated control; and
-
any combination of the above.
For the control group we looked at:
-
any combination of the above or pacifier withdrawal.
Types of outcome measures
Primary outcomes
The primary outcome was cessation of the habit.
Secondary outcomes
1. Time taken for intervention to be effective.
2. Child and parent‐ or carer‐centred outcomes of discomfort from the intervention, psychological effects of teasing associated with the intervention, and distress caused by removal of the comfort/habit.
3. Reduction in malocclusion as measured by:
-
reduction in anterior open bite (mm);
-
reduction in overjet (mm);
-
correction of posterior crossbite.
4. Costs of interventions.
Search methods for identification of studies
To identify studies to be included in this review, we developed detailed search strategies for each database to be searched based on the search strategy developed for MEDLINE (OVID) (see Appendix 1). This search strategy was revised appropriately for each database to take account of differences in controlled vocabulary and syntax rules. The MEDLINE search strategy combined the subject search with the Cochrane Highly Sensitive Search Strategy (CHSSS) for identifying randomised trials in MEDLINE: sensitivity maximising version (2008 revision) as referenced in Chapter 6.4.11.1 and detailed in box 6.4.c of the Cochrane Handbook for Systematic Reviews of Interventions, version 5.1.0 (updated March 2011) (Higgins 2011). The searches of EMBASE, CINAHL and PsycINFO were linked to search strategies for identifying RCTs developed by the Cochrane Oral Health Group.
We contacted authors of included studies for information on unpublished studies but no further attempt was made to identify unpublished literature.
Electronic searches
The following databases were searched:
-
MEDLINE via OVID (1946 to 8 October 2014) (Appendix 1);
-
Cochrane Oral Health Group's Trials Register (to 8 October 2014) (Appendix 2);
-
Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library 2014, Issue 9) (Appendix 3);
-
EMBASE via OVID (1980 to 8 October 2014) (Appendix 4);
-
PsychINFO via OVID (1980 to 8 October 2014) (Appendix 5);
-
CINAHL via EBSCO (1937 to 8 October 2014) (Appendix 6).
Searching other resources
Trials Registries
We searched the following databases for ongoing trials (see Appendix 7):
-
US National Institutes of Health Trials Register (http://clinicaltrials.gov) (to 8 October 2014);
-
The WHO Clinical Trials Registry Platform (http://apps.who.int/trialsearch/default.aspx) (to 8 October 2014).
Handsearching
The following relevant journals have been handsearched as part of the Cochrane Worldwide Handsearching Programme (see the Cochrane Masterlist for further information).
-
American Journal of Orthodontics and Dentofacial Orthopedics, 1970‐2004
-
Angle Orthodontist, 1979‐2006
-
ASDC Journal of Dentistry for Children, 1948‐2003
-
British Dental Journal, 1958‐2007
-
European Journal of Orthodontics, 1979‐2005
-
International Journal of Paediatric Dentistry, 1991‐2007
-
Journal of Orthodontics, 1973‐2007
Trials found as a result of this handsearching have been entered into the Cochrane Oral Health Group Trials Register and relevant studies were retrieved during the electronic searches.
No additional handsearching was undertaken as part of this review.
Language
Databases were searched with no language restrictions; however, all articles found were in English or provided an English abstract.
Correspondence
We contacted the first named authors or corresponding authors of studies included in the review in an attempt to identify unpublished studies and to obtain any further information about the trials. There was a response from only one author (Dr Larsson) who was not aware of any other studies.
Reference lists
The reference lists of the 14 full text articles were checked for eligibility and were scrutinised for further relevant studies.
Data collection and analysis
Selection of studies
Two review authors (Felicity Borrie (FB) and Nicola Innes (NI)) independently assessed the titles and abstracts of all reports identified by the search strategy for relevance to the review. We obtained full copies of all relevant and potentially relevant studies that appeared to meet the inclusion criteria, or for which there was insufficient data in the title and abstract to make a clear decision. A third review author (David Bearn (DB)) assisted with study selection where there was doubt about the inclusion of a trial. Studies rejected at this or subsequent stages were recorded in the Characteristics of excluded studies tables and the reasons for exclusion recorded.
Data extraction and management
Two review authors (FB and DB) independently extracted data for all reports on a specially designed data extraction form. Consensus was reached for all data included and any disagreements were resolved by the third author (NI). For each trial, the year of publication, country of origin and source of study funding were recorded as well as the following information.
1. Trial methods
-
Method of allocation
-
Number of losses to follow‐up, and reasons by study group
2. Participants
-
Age
-
Gender
-
Sample size
3. Intervention
-
Type
-
Duration, and duration of follow‐up
4. Control
-
Type of control
-
Duration, and duration of follow‐up
5. Outcomes
-
Primary and secondary outcomes as described in the outcome measures section of this protocol.
Assessment of risk of bias in included studies
For the studies included in this review, two review authors (FB and NI) undertook assessment of risk of bias independently for all studies (Higgins 2011). An overall 'Risk of bias' judgment was obtained for each study by addressing six specific domains: sequence generation, allocation concealment, blinding (because of the nature of the interventions this was only potentially possible for the outcome assessors), completeness of outcome data, risk of selective outcome reporting and other potential sources of bias. For each entry within these domains, what the study reported was stated in the 'Risk of bias' table, and a judgment made of the risk of bias for that entry (see Figure 2). The summary assessments of the risk of bias for individual studies was guided by Table 8.7a. in Cochrane Handbook for Systematic Reviews of Interventions, version 5.1.0 (Higgins 2011).
'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies
Measures of treatment effect
The data were analysed by FB and DB using Review Manager (RevMan) software and reported as suggested in Chapter 9 of the Cochrane Handbook for Systematic Reviews of Interventions 5.1.0 (Higgins 2011).
For ordinal data, including patient‐centred outcomes, discomfort and psychological effects, these were, as appropriate, dichotomised and then risk ratios (RRs) were calculated.
For dichotomous data, including cessation of habit and correction of crossbite, RRs and their 95% confidence intervals and number needed to treat for an additional beneficial outcome (NNTB) would have been calculated if data had been available.
For continuous data, including reduction in habit (measured in hours per day), time to cease habit (measured in days) and overjet and overbite (measured in millimetres (mm)), we planned to calculate the mean difference (MD) and 95% confidence intervals if data were available. However, only one study (Villa 1997) had continuous outcomes; overbite, overjet and arch length (mm) and these all presented standard deviations but it was not possible to calculate the MD from the data presented (Table 1).
| Villa 1997 study (n = 24) | Control | Palatal Crib | Statistically significant |
| Change in mandibular arch length (mm) | 0.03 SD 0.19 | ‐1.2 SD 0.8 | Yes, P value < 0.01 |
| Change in maxillary arch length (mm) | 0.01 SD 0.33 | ‐1.4 SD 1.4 | Yes, P value < 0.05 |
| Net change in overjet (mm) | 0.02 | ‐0.2 |
SD: standard deviation
P probability
For counts, including appliance breakages, we planned to calculate the rate ratio for each appliance type but there were no data available so this was not calculated.
Unit of analysis issues
We included RCTs and quasi‐RCTs.Although we had planned to analyse data presented at six months and 12 months after commencement of the intervention, this was not possible due to the lack of data and studies. For multi‐arm trials we extracted data to allow comparison between treatments as individual pairs.
Dealing with missing data
We contacted all authors of included studies by letter to obtain further information. However, none of the authors replied. We reported the proportions of participants for whom no outcome data were obtained in the 'Risk of bias' table. We used an available case analysis approach and included data only on those whose results were known, using the total number of individuals who had data recorded for each particular outcome as the denominator as in section 16.2.2 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).
Assessment of heterogeneity
We assessed clinical heterogeneity by examining the characteristics of the studies, the similarity between the types of participants, the interventions and the outcomes, as specified in the criteria for included studies.
We had planned to assess statistical heterogeneity using a Chi2 test and the I2 statistic where I2 values over 50% indicate substantial to considerable heterogeneity. Heterogeneity would have been considered to be significant when the P value was less than 0.10 (Higgins 2011). However, there were insufficient data for this calculation to be performed.
Assessment of reporting biases
We planned to assess publication bias according to the recommendations on testing for funnel plot asymmetry (Egger 1997) as described in section 10.4.3.1 of the Cochrane Handbook for Systematic Reviewsof Interventions 5.1.0 (Higgins 2011), but there were insufficient studies to allow this to be carried out.
Data synthesis
We followed The Cochrane Collaboration statistical guidelines and analysed the data using Review Manager software and reported according to Cochrane Collaboration criteria. If significant heterogeneity had been detected, we planned to assess the significance of treatment effects using the random‐effects model, providing that there were more than three studies. If not, we planned to use the fixed‐effect model. However, we only found sufficient data to allow one meta‐analysis with two studies with 59 patients, comparing palatal cribs versus no treatment (Haryett 1967; Larsson 1972).
Subgroup analysis and investigation of heterogeneity
We had planned subgroup analyses for age and gender of participants. However, these were not possible as there was only a single meta‐analysis involving two small studies and 59 participants. In addition, the lack of information provided in the studies with respect to gender and age distribution precluded these analyses.
Sensitivity analysis
There were insufficient data to allow the effect on the overall estimates of random sequence generation, allocation concealment or blinded outcome assessment to be investigated.
Results
Description of studies
Results of the search
We identified 195 publications from the search strategy, after removal of duplicates. From reading titles and abstracts, we rejected 181 as not being relevant to the review. No further potentially eligible studies were identified from the references checked. We obtained full text articles for the remaining 14, all of which were in English language. Of these 14 studies, we excluded eight as they were not RCTs or quasi‐RCTs (RCTs). The remaining six studies all of which were RCTs (Azrin 1980; Christensen 1987; Friman 1990; Haryett 1967; Larsson 1972; Villa 1997) met the inclusion criteria and have been included in the review (Figure 3).
Study flow diagram
Included studies
We included six studies with 252 enrolled participants (see Characteristics of included studies). The studies dated from 1967 to 1997.
Characteristics of the trial setting and investigators
Three trials were conducted in the USA (Azrin 1980; Friman 1990; Villa 1997), one in Canada (Haryett 1967), one in Sweden (Larsson 1972) and one in Australia (Christensen 1987). The sample sizes ranged from 22 (Friman 1990) to 76 participants (Larsson 1972); however, none of the studies gave information on power calculations.
Characteristics of the participants
There were a total of 252 enrolled participants in the six trials. The NNSH assessed in all the studies was digit sucking. None of the studies reported on the use of a pacifier.
Two studies did not give clear inclusion criteria for the participants (Azrin 1980; Villa 1997).
Details of the participants’ age range were given for four studies (two and a half years to 18 years). The mean ages of the participants were:
-
8.3 years (range two and a half to 14 years) (Azrin 1980);
-
6.3 years (range four to nine years) (Christensen 1987);
-
6.4 years in intervention group, 6.8 years in the control group (range four to 11.6 years) (Friman 1990);
-
12.1 years in intervention group, 13.5 years in the control group (range eight to 18 years) (Villa 1997).
One study stated the participants were four years old and over (Haryett 1967) and another that the participants were nine years old (Larsson 1972).
Characteristics of the interventions
We expected to find interventions under the following groupings: orthodontic appliances; barrier techniques ‐ gloves/plasters etc; chemical techniques ‐ topical substances applied to pacifier or digit; behaviour modification techniques; non‐treated control; and pacifier withdrawal. There were no studies that looked at barrier techniques or pacifier withdrawal. There were a variety of intervention techniques applied and some were combined within the one intervention group.
There were variations in both the control and intervention comparisons in the trials. Two of the studies investigated a single intervention versus a control group (Friman 1990; Villa 1997); one compared two intervention groups with a control group (Christensen 1987); another had three intervention groups and a control group (Larsson 1972); and in one study there were five intervention groups and a control group (Haryett 1967). The sixth study (Azrin 1980) made a head‐to‐head comparison of two intervention groups.
In five of the studies, where the interventions were psychological in nature or an aversive tasting substance was applied, parents administered the interventions at the participants’ home (Azrin 1980; Christensen 1987; Friman 1990; Haryett 1967; Larsson 1972). For the three studies involving the use of orthodontic appliances as the sole intervention or co‐intervention, the orthodontic appliances were provided in an orthodontic clinic (Haryett 1967; Larsson 1972; Villa 1997).
The included studies are described below under headings relating to the type of intervention employed in the studies.
The interventions were grouped into psychological interventions, aversive tasting substance application to digits and orthodontic appliances. In some of the studies, combinations of interventions were applied to the same individual and the details of the interventions in the studies are detailed below as 'combination treatment'.
Orthodontic appliances
Three of the studies (Haryett 1967; Larsson 1972; Villa 1997) included in this review used two types of orthodontic appliances (palatal cribs and palatal arches) as interventions.
i. Palatal crib: Three studies used palatal cribs, the design of which differed between the studies. These minor differences were unlikely to have an effect on the way they work (Figure 4; Figure 5).
Palatal crib
Palatal crib with spurs
Haryett 1967 defined a palatal crib as an appliance that has bands on either the maxillary second deciduous molars or first permanent molars, with pictures showing a stainless steel wire fitted behind the upper incisors, over the palatal rugae with “vertical fencelike projections extended as deep as the lateral excursions of the mandible will allow.”
Larsson 1972 used a palatal crib with a modified design. The authors describe it as having spurs welded to bands cemented to the maxillary first molar teeth. “The crib lay a millimetre or so from the mucosa and extended just behind the maxillary incisors. The spurs were rounded in front and so adjusted that they did not disturb the occlusion.”
Villa 1997 used a “palatal crib” but did not specify the design of this appliance. The authors mention in their study that they feel that the appliance they fitted would have made sucking difficult, “if not impossible.” If this appliance had been a palatal arch some degree of sucking would likely have been possible. As there was no response from the authors to clarify the design of the palatal crib used, it has been assumed that this crib was similar to that used by Haryett 1967.
ii. Palatal arch: The palatal arch placed in Haryett 1967 had bands on the molars and a wire sitting on the gingival margins of the palatal side of the upper incisors, but had no projections. Although called a palatal arch in this study, this design is different from a standard Goshgarian palatal arch used in orthodontics, where the wire connecting the bands is situated across the middle of the palate and has an omega loop. For the purpose of this review, when a palatal arch is referred to, it is the design used by Haryett 1967.
Psychological Interventions
Five types of psychological interventions were assessed in four studies (Azrin 1980; Christensen 1987; Haryett 1967; Larsson 1972).
i. Habit reversal (HR) (Azrin 1980; Christensen 1987): The children were taught competing behaviours such as making a fist or grasping a convenient object for one to three minutes (measured by counting to 100). In Azrin 1980, parents were instructed to praise the child when sucking was absent, provide pleasant treats and surprises when sucking was absent for an extended period and stop television or bedtime stories when sucking occurred.
ii.Differential reinforcement of other behaviour (DRO): In Christensen 1987, DRO involved an increasing schedule of reinforcement using tokens as rewards when thumb sucking was avoided.
iii. A two‐part strategy; gaining child’s co‐operation to break the habit and parental reward for periods of no sucking (Haryett 1967): Co‐operation was gained by creating a desire in the child to avoid negative aesthetic effects. This was done by showing the child that digit sucking could alter the position of the teeth using both their own teeth (with mirrors) and pictures or models of other teeth with undesirable aesthetics. The second part of the strategy involved the parent rewarding the child for periods of “no sucking” by giving them their full attention and ignoring them if the habit occurred.
iv. Positive reinforcement (Larsson 1972): Participants’ mothers were given specific instructions about different forms of encouragement and reinforcement was also given by a psychologist.
v. Negative reinforcement (Larsson 1972): Children and their parents were given information about the consequences and risks of prolonged finger sucking. They were given models of the children’s teeth home with them.
Aversive tasting substance application
One study (Azrin 1980) made a head‐to‐head comparison involving aversive tasting substance (ATS) application and habit reversal. In the ATS group, the parents of the children received a single phone call informing them to apply a bitter tasting substance to the digit, morning and evening.
Combination treatment
There were three combination treatments reported in two studies (Friman 1990; Haryett 1967).
i. Psychological intervention and palatal arch (Haryett 1967): The design of the appliance is described above and the psychological component involved a two‐part strategy; gaining child’s co‐operation to break the habit and parental reward for periods of no sucking.
ii. Psychological intervention and palatal crib (Haryett 1967): The design of the appliance is described above, and the psychological component involved a two‐part strategy; gaining child’s co‐operation to break the habit and parental reward for periods of no sucking.
iii. Application of an aversive tasting substance and psychological intervention (Friman 1990): This involved both application of an aversive tasting substance to the thumb and a psychological component, a treat chosen at random from a grab bag.
Characteristics of the controls
One study compared two interventions and did not have a no‐treatment control group (Azrin 1980). In the remaining five trials (Christensen 1987; Friman 1990; Haryett 1967; Larsson 1972; Villa 1997), the control groups all consisted of no treatment, with Christensen 1987 and Friman 1990 using a waiting list control group.
Characteristics of the outcome measures
We were able to extract data for the primary outcome of habit cessation, and two of the secondary outcomes, child and parent‐ or carer‐centred outcomes detailed in Table 2 (% of children with oppositional behaviour before the intervention and at follow‐up; number upset by treatment; number reporting eating difficulty; development of mannerisms) and reduction in malocclusion. Results for more than one time point were available for two studies. These were percentage time intervals with thumb sucking immediately post treatment, and at three‐month follow‐up (Friman 1990), and cessation of habit at one month, one year, two years and three years follow‐up (Haryett 1967).
| Study | Outcome | Intervention A | Intervention B | Intervention C | Intervention D | Intervention E | Control |
| Christensen 1987 (n = 30) | Habit reversal | Differential reinforcement of other behaviour |
|
|
| Waiting list (no treatment) | |
| % of children with oppositional behaviour before intervention and at follow‐up | 3.1% to 0.2% | 2.5% to 0.6% | 2.7% to 2.7% | ||||
| Haryett 1967 (n = 66) | Psychology | Palatal arch | Palatal arch & psychology | Palatal crib | Palatal crib & psychology | No treatment | |
| Number upset by treatment | 1 | 0 | 2 | 6 | 2 | 0 | |
| No. reporting eating difficulty | 0 | 0 | 0 | 3 | 6 | 0 | |
| Development of mannerisms | 6 | 1 | 1 | 4 | 0 | 1 |
Primary outcome
Sucking cessation was measured in five studies (Azrin 1980; Christensen 1987; Friman 1990; Haryett 1967; Larsson 1972) at timepoints between five days (Friman 1990) and three years (Haryett 1967) but in a number of different ways:
-
four of the studies measured cessation of the habit by proportion of participants who had stopped their NNSH in each group (Azrin 1980; Christensen 1987; Haryett 1967; Larsson 1972);
-
one of these studies (Christensen 1987) additionally measured the proportion of time spent digit sucking before and after the intervention; and
-
one study (Friman 1990) only measured the percentage intervals of time with observed thumb sucking immediately after the intervention.
Sucking cessation was measured and reported at <12 months (short term) or ≥12 months (long term).
Secondary outcomes
Time taken for sucking habit to cease was not reported by any of the included studies.
No study reported adverse events. However, adverse outcomes were reported in one study (Haryett 1967); reported upset, speech and eating difficulties and development of mannerisms (Table 3).
| Psychological treatment | Orthodontic treatment | Both Psychological and Orthodontic treatment | Control | |
| Reported upset | 1/11 (9%) | 6/22 (27%) | 4/22 (18%) | 0/50 (0%) |
| Speech difficulty | 0/11 (0%) | 9/22 (41%) | 6/22 (27%) | 0/50 (0%) |
| Eating difficulty | 0/11 (0%) | 3/22 (14%) | 6/22 (27%) | 0/45 (0%) |
| Developing mannerisms | 6/11 (55%) | 5/22 (23%) | 1/22 (5%) | 3/50 (6%) |
Reduction in malocclusion was reported by only one study (Villa 1997).
Costs were not measured by any of the studies.
Excluded studies
We excluded eight studies (see Characteristics of excluded studies). Two were observational, non‐intervention trials (Adair 1992; Friman 1986), three had no control group or inadequate controls (Al‐Emran 2005; Haryett 1970; Woods 1999), two were longitudinal studies (Cozza 2006; Cozza 2007), and one did not have outcomes relevant to this review (Degan 2005).
Risk of bias in included studies
A 'Risk of bias' graph (Figure 2) and summary diagram (Figure 6) were completed for the included studies (see Characteristics of included studies). All of the studies were assessed as being at overall high risk of bias although for one of the studies, this was purely on the basis of lack of blinding (Villa 1997).
'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study
Allocation
Random sequence generation (selection bias)
Four of the studies had adequate sequence generation: coin flip (Azrin 1980; Friman 1990), sampling without replacement procedure (Christensen 1987) and random sample tables (Larsson 1972). For the other two studies, sequence generation was unclear (Haryett 1967; Villa 1997).
Allocation concealment
Allocation concealment was unclear in all studies.
Blinding
There was no blinding of participants or operators in any of the studies so we assessed all studies as at high risk of performance bias.
For the outcomes assessors, there was adequate blinding in one study (Villa 1997) where study models were assessed by an independent assessor. Blinding was unclear in two studies (Christensen 1987; Larsson 1972) and there was no blinding in three of the studies (Azrin 1980; Friman 1990; Haryett 1967) where parents acted as outcome assessors.
Incomplete outcome data
There were two studies that provided clear information about incomplete data and dropouts (Haryett 1967; Larsson 1972). They both had low dropout rates, with Haryett 1967 having one participant out of 66 lost to follow‐up at 10 months and Larsson 1972 having one participant out of 76 decline post‐intervention follow‐up. Incomplete data and follow‐up were unclear in two studies: Christensen 1987 implied that there were no dropouts in the control group but there were no figures to confirm this and for Villa 1997, dropout rates were not reported, although it was implied that all children were followed up. For two of the studies there was high risk of bias as one study (Azrin 1980) only provided information on the 50% loss to follow‐up for the intervention group with loss to follow‐up not reported for the control group. In Friman 1990, dropout rates were not reported, although it was implied that all children were followed up "posttest" and there was no detail on whether follow‐up participants had been allocated to the intervention or control group. No intention‐to‐treat analysis were performed to account for missing data.
Selective reporting
Two studies showed no signs of selective reporting (Christensen 1987; Villa 1997) in terms of outcome measures. However, data presented in Villa 1997 were insufficient to calculate mean difference. In one study (Haryett 1967), it was unclear whether individual outcomes had been prespecified and in three studies (Azrin 1980; Friman 1990; Larsson 1972), there was no prespecification of the outcomes.
Other potential sources of bias
Only two of the studies were assessed as being at low risk of other potential sources of bias (Friman 1990; Haryett 1967). One study (Villa 1997), was assessed as unclear as it was uncertain whether the sample was representative of the general population having been were recruited from an orthodontic department patient population. Three studies were assessed as being at high risk of bias for different reasons. In one study (Azrin 1980), bias may have been introduced into the sample as recruitment was through a newspaper advertisement and two children, who had been allocated to the control arm, withdrew as they had used that procedure previously without success. Christensen 1987 recruited participants through self‐referral following a newspaper article and Larsson 1972 limited recruitment to children who were judged by their parents "to be intense suckers ‐ children who sucked their fingers at least every evening...".
Effects of interventions
See: Summary of findings 1 Summary of findings: orthodontic appliances versus no treatment; Summary of findings 2 Summary of findings: psychological interventions versus no treatment
The included studies reported on effects of the different interventions and combination of interventions on sucking cessation and occlusion. The comparisons have been summarised as follows.
Interventions versus no treatment
-
Orthodontic appliance versus no treatment (Analysis 1.1; Analysis 1.2; Analysis 1.3)
-
Psychological intervention versus no treatment (Analysis 2.1; Analysis 2.2)
-
Psychological intervention and orthodontic treatment versus no treatment (Analysis 3.1)
Head‐to‐head comparisons of active interventions
-
Aversive taste versus psychological intervention (Analysis 4.1)
-
Orthodotic appliances versus alternative orthodontic appliances (Analysis 5.1)
-
Psychological intervention versus alternative psychological intervention (Analysis 6.1; Analysis 6.2)
-
Psychological intervention plus orthodontic appliance versus psychological intervention plus alternative orthodontic appliance (Analysis 7.1)
We have consolidated other aspects of the studies into Additional tables to present a picture of the data more fully and show explicitly why data could not be combined for most of the interventions and outcomes.
Table 1 presents all data related to reduction in malocclusion which could only be reproduced with means and standard deviations.
Table 2 gives an overview of the child/parent/carer‐centred outcomes.
Table 3 where the adverse outcomes are detailed.
Table 4 shows the primary outcome of cessation of digit sucking after intervention for each different intervention, detailing the variety of interventions, controls, outcome measures.
| Study | Intervention A | Intervention B | Intervention C | Intervention D | Intervention E | Control | How cessation was measured | Cessation of digit sucking units | |
| Azrin 1980 (n = 30) | Habit reversal |
|
|
|
| Bitter substance | % of children with cessation of the habit | 47% HR group vs 10% control at 3 months | |
| Christensen 1987 (n = 30)
| Habit reversal | Differential reinforcement of other behaviour |
|
|
| Waiting list (no treatment) | number of children with cessation of the habit | 2/10 HR group vs 1/10 DRO group vs 0/10 in WL control group | |
| Friman 1990 (n = 34) | Aversive taste treatment and reward system |
|
|
|
| Waiting list (no treatment) | % intervals of time with observed thumb sucking before and after intervention | 44% to 4% aversive taste vs 44% to 51% control (% intervals with thumb sucking) | |
| Haryett 1967 (n = 66) | Psychology | Palatal arch | Palatal arch & psychology | Palatal crib | Palatal crib & psychology | No treatment | % of children with cessation of the habit | 9.1% A vs 9.1% B vs 27.3% C vs 100% D vs 100% E vs 10% no treatment | |
| Larsson 1972 (n = 76) | Positive reinforcement (pos) | Negative reinforcement (neg) | Palatal crib (crib) |
|
| No treatment | % of children with cessation of the habit | 26% pos, 53% neg, 42% crib, 5% control |
DRO: differential reinforcement of other behaviour
HR: habit reversal
WL: waiting list
Interventions versus no treatment
Orthodontic appliances versus no treatment
Three studies at high risk of bias reported on the effect of orthodontic appliances compared with no treatment. Outcomes reported were sucking cessation (short and long term) and occlusion.
Sucking cessation (short term)
Two studies, with data from 70 participants, compared palatal arch or palatal crib with no treatment) and reported data at one month (Haryett 1967) and two and a half months (Larsson 1972). Both studies were at high risk of bias. Pooling of data showed a statistically significant benefit in favour of the orthodontic appliances (risk ratio (RR) 6.53, 95% confidence interval (CI) 1.67 to 25.53). There was no significant heterogeneity (I2 = 0%) (Analysis 1.1).
Sucking cessation (long term)
One study comparing palatal crib with no treatment in 37 participants measured sucking cessation at 12 months (Larsson 1972). The study was at high risk of bias. Palatal crib was more likely to stop digit sucking compared to no treatment (RR 5.81, 95% CI 1.49 to 22.66) (Analysis 1.2).
Occlusion
One study, comparing palatal crib with no treatment in 24 participants, assessed changes in malocclusions at three months (Villa 1997). Although changes in mandibular and maxillary arch lengths were reported as statistically significant, the authors did not mention the clinical significance of the findings. The actual changes were less than 1.5 mm for the palatal crib intervention group, and would generally not be considered clinically important. However, the statistically significant reduction in anterior open bite of 3.7 mm in the palatal crib group (P < 0.05) is clinically important (Analysis 1.3). There was no statistically significant net change in overjet between the groups. Correction of posterior crossbite was not measured in this study despite being a well recognised feature in the malocclusion of a child with a thumb‐sucking habit.
Psychological interventions versus no treatment
Two studies at high risk of bias reported on the effect of various psychological interventions compared with no treatment on sucking cessation.
Sucking cessation (short term)
Two studies, with data from a total of 78 participants, evaluated positive reinforcement (alone or in combination with gaining the child's co‐operation) or negative reinforcement compared with no treatment (Haryett 1967; Larsson 1972). Pooling of data showed a statistically significant difference in favour of the psychological interventions (RR 6.16, 95% CI 1.18 to 32.10; I2 = 0%) (Analysis 2.1).
Sucking cessation (long term)
One study, with data from 57 participants,reported on the long‐term effect of positive and negative reinforcement on sucking cessation. A statistically significant difference was shown in favour of the psychological interventions (RR 6.25, 95% CI 1.65 to 23.65) (Analysis 2.2).
Psychological intervention plus orthodontic treatment versus no treatment
A combination of psychological intervention and orthodontic treatment was compared with no treatment in on study. The study reported on short‐term sucking cessation only.
Sucking cessation (short term)
One study, analysing 32 participants, compared psychological interventions plus orthodontic treatment (either palatal crib or palatal arch) with no treatment (Haryett 1967). The psychological intervention was a combination of positive reinforcement and cooperation. The study, at high risk of bias, found a statistically significant difference in favour of the combined intervention (RR 6.36, 95% CI 0.97 to 41.96) (Analysis 3.1).
Head‐to‐head comparisons of active interventions
Aversive taste versus psychological intervention
Sucking cessation (short term)
One study, which analysed 29 participants, compared habit reversal with aversive taste application and found that aversive taste was less likely to stop digit sucking than psychological intervention but the difference was not statistically significant (RR 0.18 (95% CI 0.03 to 1.24) (Analysis 4.1). The study was at high risk of bias.
Orthodontic appliances versus alternative orthodontic appliances
Sucking cessation (short term)
One study, at high risk of bias, compared palatal arch with palatal crib (Haryett 1967). The study evaluated 22 participants for this comparison. A statistically significant difference was shown in favour of palatal crib (RR 0.13, 95% CI 0.03 to 0.59)
Psychological interventions versus alternative psychological interventions
Sucking cessation (short term)
One study, analysing 20 participants, compared habit reversal (HR) with differential reinforcement of other behaviour (DRO) (Christensen 1987). No statistically significant difference was shown between treatment groups.
A second study compared positive reinforcement with negative reinforcement, with data from 38 participants (Larsson 1972). Again, no statistically significant difference between groups were shown.
Sucking cessation (long term)
Larsson 1972, comparing positive reinforcement with negative reinforcement, also provided long‐term data. Again, no statistically significant between group differences were shown.
Psychological interventions plus orthodontic appliance versus psychological interventions plus alternative orthodontic appliance
Sucking cessation (short term)
One study compared a combination of psychological intervention with palatal arch to psychological intervention with palatal crib (Haryett 1967). The trial evaluated 22 participants for this comparison and showed a statistically significant difference in sucking cessation (short term) in favour of the psychological intervention and palatal crib combination (RR 0.30, 95% CI [0.13 to 0.74). The trial was at a high risk of bias.
Other comparisons
These are detailed under interventions, as well as being grouped by outcomes under Table 4 for the primary outcome of cessation of the habit, Table 2 for the secondary outcome relating to child and parent/carer measures, Table 3 for adverse outcomes after one month and Table 1 for the secondary outcomes related to reduction in malocclusion.
Due to poor reporting, differences in interventions and a lack of standardisation in outcomes, no useful interpretation can be drawn from these results.
Discussion
Summary of main results
summary of findings Table 1; summary of findings Table 2.
The objective of this review was twofold: to assess the effectiveness of different interventions for stopping non‐nutritive sucking habits (NNSHs), and to identify acceptability of interventions. Six trials, with 252 enrolled children, aged between two and a half and 18 years, were included. Four studies (Azrin 1980; Christensen 1987; Haryett 1967; Larsson 1972) measured the primary outcome of cessation of NNSH while two studies reported secondary outcome data related to behaviour of child and parent‐ or carer‐centred measures.
There was a range of clinical interventions, of differing duration and follow‐up, evaluated (behaviour modification, application of an aversive tasting substance to digits and use of intra‐oral orthodontic appliances). In addition, the studies were at high risk of bias. The paucity of studies for each intervention type and their high risk of bias means that the body of evidence to support clinical decision making for cessation of NNSH is low.
Orthodontic appliances (palatal arch and palatal crib) were shown to be beneficial at increasing the number of children stopping sucking in both the short and long term, in comparison to no treatment. Palatal crib was also shown to beneficial in term of effect on occlusion (short term) compared with no treatment. When palatal arch and palatal crib were compared directly in a single study, a statistically significant increase in the number of children stopping sucking was seen in favour of palatal crib. This was the same whether the palatal arch or palatal crib was used alone or in combination with a psychological intervention.
Psychological interventions, such as positive or negative reinforcement were also shown to significantly increase the number of children stopping sucking in both the short and long term, in comparison to no treatment. There was insufficient information to determine whether one psychological intervention was more effective than another.
There was insufficient information to determine whether aversive taste was more effective than psychological interventions.
Secondary outcome data for child and parent‐ or carer‐centred measure reported were ‘oppositional behaviour’ (Christensen 1987), upset during treatment, eating difficulties and development of mannerisms (Haryett 1967). However, the numbers of patients for which these were reported were small and the findings inconsistent. Given the conflicting nature of the data and the small numbers, it is not possible to draw clear conclusions.
Overall completeness and applicability of evidence
This review has highlighted that the body of evidence for this subject is weak. There are very few trials that met the inclusion criteria with small numbers of children included, a wide age range, and they are published over a 30‐year period. This review identified a wide range of interventions, with differing durations, and interventions that were aimed at children or parents or both. A range of outcomes were found along with different durations of follow‐up for participants following the interventions.
There were no studies included in the review that used removable orthodontic appliances to stop NNSHs. This method is still commonly used in the UK but is not considered good practice in the USA (Proffit 2013). None of the included studies looked at interventions for pacifier habits, but this may be due to the fact that it is easy to withdraw pacifiers and consequently eliminate the problem.
There was very little data in the studies about the duration of treatment and long‐term follow‐up regarding cessation of habit and it was not possible to determine the age at which treatment would be most effective. Furthermore, there may be a period of time when the occlusion is establishing where absence of a NNSH may be more critical than at another stage. For example, it may be that re‐uptake of a NNSH might have very little effect in a 12‐year‐old, where the occlusion is more established than in, for example, a seven‐year‐old. No conclusions could be drawn about the time taken for different interventions to be effective, as they were in place for different lengths of time and follow‐up occurred at differing time points. There was also a lack of precision in measuring outcomes regarding time taken for effective treatments. For example, in Haryett 1967 where the palatal crib was in place for 10 months and 100% success was achieved with cessation of habit, there was no information about which time point this occurred or even whether the same result would have been achieved in a shorter timeframe (Table 4),
Orthodontic appliances were used in three studies (Haryett 1967; Larsson 1972; Villa 1997). However, no information was given on ease of fitting or removing the appliances. All were fixed appliances although a removable deterrent appliance is sometimes used in the UK for managing persistent thumb sucking habits, particularly if it is a nighttime‐only occurrence and the patient is motivated to stop.
There was no information in any of the studies relating to costs of the interventions.
Quality of the evidence
The quality of the evidence ranged from low to very low due to imprecision and high risk of bias in the studies. Imprecision was as a result of very few events from the small number of participants recruited as well as wide confidence intervals that include both appreciable benefit and appreciable harm. Methodological limitations were mostly due to lack of blinding of participants and personnel involved, attrition, selective reporting and sampling bias from participant recruitment (other bias). There is little consideration of adequate controls in the studies as some individuals could potentially cease a habit without intervention. The lack of standardised measures and the low quality of these studies, may, at least in part, be a result of their age.
Potential biases in the review process
We tried to limit bias in the review process by following standard methodological procedures expected by The Cochrane Collaboration. A sensitive search strategy was used in this review with every effort made to identify all relevant studies. No studies were excluded due to language. Data collection and analyses were carried out independently with any disagreement resolved by discussion amongst the review authors or with the assistance of the Cochrane Oral Health Group to minimise/exclude bias in this review.
The post‐hoc decision to regard outcomes observed less than 12 months post‐intervention as short‐term outcomes and 12 months or more as long‐term outcomes could also be considered a potential bias.
Agreements and disagreements with other studies or reviews
There are no other studies or reviews to compare our findings with.
Orthodontic deterrent appliance
'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies
Study flow diagram
Palatal crib with spurs
'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study
Comparison 1: Orthodontic appliances versus no treatment, Outcome 1: Sucking cessation (short term)
Comparison 1: Orthodontic appliances versus no treatment, Outcome 2: Sucking cessation (long term)
Comparison 1: Orthodontic appliances versus no treatment, Outcome 3: Occlusion (short term)
Comparison 2: Psychological intervention versus no treatment, Outcome 1: Sucking cessation (short term)
Comparison 2: Psychological intervention versus no treatment, Outcome 2: Sucking cessation (long term)
Comparison 3: Psychological intervention + orthodontic treatment versus no treatment, Outcome 1: Sucking cessation
Comparison 4: Aversive taste versus psychological intervention, Outcome 1: Sucking cessation (short term)
Comparison 5: Orthodontic appliances versus alternative orthodontic appliance, Outcome 1: Sucking cessation (short term)
Comparison 6: Psychological intervention versus alternative psychological intervention, Outcome 1: Sucking cessation (short term)
Comparison 6: Psychological intervention versus alternative psychological intervention, Outcome 2: Sucking cessation (long term)
Comparison 7: Psycholgical+orthodontic versus psychological+alternative orthodontic, Outcome 1: Sucking cessation (short term)
| Orthodontic appliances compared with no treatment for the cessation of pacifier or digit sucking habits in children | ||||||
|---|---|---|---|---|---|---|
| Patient or population: children with pacifier or digit sucking habits Settings: home Intervention: orthodontic appliances (palatal arch/palatal crib) Comparison: no treatment | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | Number of participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| No treatment | Orthodontic appliance | |||||
| Sucking cessation (short term) | 7 per 100 | 46 per 100 | RR 6.53 (1.67 to 25.53) | 70 (2) | Lowa,b | Long‐term data also show orthodontic appliances to be beneficial when compared with no treatment |
| Occlusion (changes in overbite) | MD 4.10 (2.93 to 5.27) | 24 (1) | Very lowa,b,c | Difference in mean net change in overbite between no treatment (‐0.4 mm) and orthodontic appliance (3.7 mm) = 4.1 mm | ||
| Adverse events | Insufficient information reported | |||||
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| aDowngraded due to risk of bias. | ||||||
| Psychological interventions compared with no treatment for the cessation of pacifier or digit sucking habits in children | ||||||
|---|---|---|---|---|---|---|
| Patient or population: children with pacifier or digit sucking habits Settings: home Intervention: psychological interventions Comparison: no treatment | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | Number of participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| No treatment | Pyschological interventions | |||||
| Sucking cessation (short term) | 3 per 100 | 19 per 100 | RR 6.16 (1.18 to 32.10) | 87 (2) | Lowa,b | Long‐term data also show psychological interventions to be beneficial when compared with no treatment |
| Occlusion | Not reported | |||||
| Adverse events | Insufficient information reported | |||||
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| aDowngraded due to risk of bias. | ||||||
| Villa 1997 study (n = 24) | Control | Palatal Crib | Statistically significant |
| Change in mandibular arch length (mm) | 0.03 SD 0.19 | ‐1.2 SD 0.8 | Yes, P value < 0.01 |
| Change in maxillary arch length (mm) | 0.01 SD 0.33 | ‐1.4 SD 1.4 | Yes, P value < 0.05 |
| Net change in overjet (mm) | 0.02 | ‐0.2 | |
| SD: standard deviation | |||
| Study | Outcome | Intervention A | Intervention B | Intervention C | Intervention D | Intervention E | Control |
| Christensen 1987 (n = 30) | Habit reversal | Differential reinforcement of other behaviour |
|
|
| Waiting list (no treatment) | |
| % of children with oppositional behaviour before intervention and at follow‐up | 3.1% to 0.2% | 2.5% to 0.6% | 2.7% to 2.7% | ||||
| Haryett 1967 (n = 66) | Psychology | Palatal arch | Palatal arch & psychology | Palatal crib | Palatal crib & psychology | No treatment | |
| Number upset by treatment | 1 | 0 | 2 | 6 | 2 | 0 | |
| No. reporting eating difficulty | 0 | 0 | 0 | 3 | 6 | 0 | |
| Development of mannerisms | 6 | 1 | 1 | 4 | 0 | 1 |
| Psychological treatment | Orthodontic treatment | Both Psychological and Orthodontic treatment | Control | |
| Reported upset | 1/11 (9%) | 6/22 (27%) | 4/22 (18%) | 0/50 (0%) |
| Speech difficulty | 0/11 (0%) | 9/22 (41%) | 6/22 (27%) | 0/50 (0%) |
| Eating difficulty | 0/11 (0%) | 3/22 (14%) | 6/22 (27%) | 0/45 (0%) |
| Developing mannerisms | 6/11 (55%) | 5/22 (23%) | 1/22 (5%) | 3/50 (6%) |
| Study | Intervention A | Intervention B | Intervention C | Intervention D | Intervention E | Control | How cessation was measured | Cessation of digit sucking units | |
| Azrin 1980 (n = 30) | Habit reversal |
|
|
|
| Bitter substance | % of children with cessation of the habit | 47% HR group vs 10% control at 3 months | |
| Christensen 1987 (n = 30)
| Habit reversal | Differential reinforcement of other behaviour |
|
|
| Waiting list (no treatment) | number of children with cessation of the habit | 2/10 HR group vs 1/10 DRO group vs 0/10 in WL control group | |
| Friman 1990 (n = 34) | Aversive taste treatment and reward system |
|
|
|
| Waiting list (no treatment) | % intervals of time with observed thumb sucking before and after intervention | 44% to 4% aversive taste vs 44% to 51% control (% intervals with thumb sucking) | |
| Haryett 1967 (n = 66) | Psychology | Palatal arch | Palatal arch & psychology | Palatal crib | Palatal crib & psychology | No treatment | % of children with cessation of the habit | 9.1% A vs 9.1% B vs 27.3% C vs 100% D vs 100% E vs 10% no treatment | |
| Larsson 1972 (n = 76) | Positive reinforcement (pos) | Negative reinforcement (neg) | Palatal crib (crib) |
|
| No treatment | % of children with cessation of the habit | 26% pos, 53% neg, 42% crib, 5% control | |
| DRO: differential reinforcement of other behaviour | |||||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1.1 Sucking cessation (short term) Show forest plot | 2 | 70 | Risk Ratio (M‐H, Fixed, 95% CI) | 6.53 [1.67, 25.53] |
| 1.2 Sucking cessation (long term) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 1.3 Occlusion (short term) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 2.1 Sucking cessation (short term) Show forest plot | 2 | 78 | Risk Ratio (M‐H, Fixed, 95% CI) | 6.16 [1.18, 32.10] |
| 2.2 Sucking cessation (long term) Show forest plot | 1 | 57 | Risk Ratio (M‐H, Fixed, 95% CI) | 6.25 [1.65, 23.65] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 3.1 Sucking cessation Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 4.1 Sucking cessation (short term) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 5.1 Sucking cessation (short term) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 6.1 Sucking cessation (short term) Show forest plot | 2 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 6.1.1 HR versus DRO | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 6.1.2 Positive reinforcement versus negative reinforcement | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 6.2 Sucking cessation (long term) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 7.1 Sucking cessation (short term) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 7.1.1 Psych+palatal arch versus psych+palatal crib | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
