Frenotomy for tongue‐tie in newborn infants
Abstract
Background
Tongue‐tie, or ankyloglossia, is a condition whereby the lingual frenulum attaches near the tip of the tongue and may be short, tight and thick. Tongue‐tie is present in 4% to 11% of newborns. Tongue‐tie has been cited as a cause of poor breastfeeding and maternal nipple pain. Frenotomy, which is commonly performed, may correct the restriction to tongue movement and allow more effective breastfeeding with less maternal nipple pain.
Objectives
To determine whether frenotomy is safe and effective in improving ability to feed orally among infants younger than three months of age with tongue‐tie (and problems feeding).
Also, to perform subgroup analysis to determine the following.
• Severity of tongue‐tie before frenotomy as measured by a validated tool (e.g. Hazelbaker Assessment Tool for Lingual Frenulum Function (ATLFF) scores < 11; scores ≥ 11) (Hazelbaker 1993).
• Gestational age at birth (< 37 weeks' gestation; 37 weeks' gestation and above).
• Method of feeding (breast or bottle).
• Age at frenotomy (≤ 10 days of age; > 10 days to three months of age).
• Severity of feeding difficulty (infants with feeding difficulty affecting weight gain (as assessed by infant's not regaining birth weight by day 14 or falling off centiles); infants with symptomatic feeding difficulty but thriving (greater than birth weight by day 14 and tracking centiles).
Search methods
We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, Embase and CINAHL up to January 2017, as well as previous reviews including cross‐references, expert informants and journal handsearching. We searched clinical trials databases for ongoing and recently completed trials. We applied no language restrictions.
Selection criteria
Randomised, quasi‐randomised controlled trials or cluster‐randomised trials that compared frenotomy versus no frenotomy or frenotomy versus sham procedure in newborn infants.
Data collection and analysis
Review authors extracted from the reports of clinical trials data regarding clinical outcomes including infant feeding, maternal nipple pain, duration of breastfeeding, cessation of breastfeeding, infant pain, excessive bleeding, infection at the site of frenotomy, ulceration at the site of frenotomy, damage to the tongue and/or submandibular ducts and recurrence of tongue‐tie. We used the GRADE approach to assess the quality of evidence.
Main results
Five randomised trials met our inclusion criteria (n = 302). Three studies objectively measured infant breastfeeding using standardised assessment tools. Pooled analysis of two studies (n = 155) showed no change on a 10‐point feeding scale following frenotomy (mean difference (MD) ‐0.1, 95% confidence interval (CI) ‐0.6 to 0.5 units on a 10‐point feeding scale). A third study (n = 58) showed objective improvement on a 12‐point feeding scale (MD 3.5, 95% CI 3.1 to 4.0 units of a 12‐point feeding scale). Four studies objectively assessed maternal pain. Pooled analysis of three studies (n = 212) based on a 10‐point pain scale showed a reduction in maternal pain scores following frenotomy (MD ‐0.7, 95% CI ‐1.4 to ‐0.1 units on a 10‐point pain scale). A fourth study (n = 58) also showed a reduction in pain scores on a 50‐point pain scale (MD ‐8.6, 95% CI ‐9.4 to ‐7.8 units on a 50‐point pain scale). All studies reported no adverse effects following frenotomy. These studies had serious methodological shortcomings. They included small sample sizes, and only two studies blinded both mothers and assessors; one did not attempt blinding for mothers nor for assessors. All studies offered frenotomy to controls, and most controls underwent the procedure, suggesting lack of equipoise. No study was able to report whether frenotomy led to long‐term successful breastfeeding.
Authors' conclusions
Frenotomy reduced breastfeeding mothers’ nipple pain in the short term. Investigators did not find a consistent positive effect on infant breastfeeding. Researchers reported no serious complications, but the total number of infants studied was small. The small number of trials along with methodological shortcomings limits the certainty of these findings. Further randomised controlled trials of high methodological quality are necessary to determine the effects of frenotomy.
PICOs
Plain language summary
Surgical release of tongue‐tie for the treatment of tongue‐tie in young babies
Review question: Tongue‐tie is a potentially treatable cause of breastfeeding problems ‐ if a baby is tongue‐tied and is having feeding difficulties, does releasing the tongue‐tie help?
Background: Tongue‐tie is a condition whereby the membrane between the tongue and the floor of the mouth is too tight or too short. This may cause feeding problems for the baby and/or nipple pain for a breastfeeding mother.
Study characteristics: Five randomised controlled trials enrolling 302 infants met the inclusion criteria.
Key results: In an infant with tongue‐tie and feeding difficulties, surgical release of the tongue‐tie does not consistently improve infant feeding but is likely to improve maternal nipple pain. Further research is needed to clarify and confirm this effect.
Quality of evidence: The quality of the evidence is very low to moderate because overall only a small number of studies have looked at this condition, the total number of babies included in these studies was low and some studies could have been better designed.
Authors' conclusions
Summary of findings
| Undefined | ||||||
| Patient or population: tongue‐tie in newborn infants | ||||||
| Outcomes | Illustrative comparative risks (mean and SD) | Relative effect | Number of participants | Quality of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with no frenotomy or sham procedure | Risk with frenotomy | |||||
| Infant breastfeeding assessed by validated scale ‐ IBFAT scores following procedure | Mean IBFAT scores following procedure in the control group was 8.1 (SD 0.9) | Mean IBFAT scores following procedure in the frenotomy group was 11.6 (SD 0.8) | Mean difference is 3.50 (3.06 to 3.94) | 58 | ⊕⊕⊝⊝ | IBFAT score is based on a 12‐point scale |
| Infant breastfeeding assessed by validated scale ‐ LATCH scores following procedure | Mean LATCH scores following procedure in the control group was 6.8 to 8.5 (SD < 1.9) | Mean LATCH scores following procedure in the frenotomy group was 6.8 to 8.4 (SD < 2) | Mean difference is ‐0.07 (‐0.63 to 0.48) | 155 | ⊕⊕⊝⊝ | LATCH score is based on a 10‐point scale |
| Maternal nipple pain assessed by a validated pain scale ‐ visual analogue pain scale | Mean visual analogue pain scale scores following procedure in the control group was 2.9 to 5.5 (SD < 2.6) | Mean IBFAT scores following procedure in the frenotomy group was 1.6 to 5.3 (SD < 2.4) | Mean difference is ‐0.74 (‐1.35 to ‐0.13) | 183 | ⊕⊕⊝⊝ | Visual analogue pain scale score is based on a 10‐point scale |
| Maternal nipple pain assessed by a validated pain scale ‐ SF‐MPQ pain scale | Mean SF‐MPQ scores following procedure in the control group was 13.5 (SD 1.5) | Mean IBFAT scores following procedure in the frenotomy group was 4.9 (SD 1.5) | Mean difference is ‐8.60 (‐9.37 to ‐7.83) | 58 | ⊕⊕⊝⊝ | SF‐MPQ score is based on a 50‐point scale |
| *The risk in the intervention group (and its 95% confidence interval) is based on assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI) | ||||||
| GRADE Working Group grades of evidence | ||||||
| aImprecision (small total participant population) bRisk of bias (incomplete blinding) | ||||||
Background
Description of the condition
The lingual frenulum is a fold of mucous membrane that extends from the floor of the mouth to the midline of the underside of the tongue. It helps to stabilise the base of the tongue and does not normally interfere with tongue tip movement (Marchesan 2005). Tongue‐tie (ankyloglossia) is a condition in which the lingual frenulum has an anterior attachment near the tip of the tongue and may be unusually short, tight and thick (Jackson 2012). This causes virtual adhesion of the tongue tip to the floor of the mouth and can result in restricted tongue tip movement (Marchesan 2005). The exact cause of 'tongue‐tie' (ankyloglossia) is not known. Genetics may play a role, as the condition tends to run in some families (Coryllos 2004). Prevalence is about 4% to 11% among newborns (Hogan 2005; Messner 2000a; Messner 2000b; Ricke 2005).
Tongue‐tie has been cited as a cause of poor breastfeeding because the infant is unable to attach or stay latched on, and because maternal nipple pain may result (Coryllos 2004Hogan 2005). In older children and adults, tongue‐tie has been implicated as a cause of speech delay, abnormal dentition, poor oral hygiene and inability to play wind instruments (Krol 2007). As an infant breastfeeds, the tongue moves with peristalsis over maternal lactiferous sinuses and extracts milk. When the infant's tongue movement is restricted, as is the case with severe tongue‐tie, reduced movement may affect milk extraction, and friction may be present between the tongue or gums and the nipple, causing damage to the nipple and maternal pain (Coryllos 2004Hogan 2005). References to tongue‐tie causing speech problems date back to Aristotle in the third century BC (Obladen 2010). The association between breastfeeding difficulty and tongue‐tie has been recognised for at least 500 years (Obladen 2010). In recent years, with recognition and encouragement of exclusive breastfeeding as the optimal primary mode of infant feeding, the justification for frenotomy has shifted from improving speech problems to improving breastfeeding (Obladen 2010), re‐igniting the historical debate as to the role of tongue‐tie in breastfeeding difficulties (Kumar 2012).
The diagnosis of tongue‐tie depends on an assessment of the structure and function of the lingual frenulum. Diagnostic classification systems vary from simple visual inspection and/or palpation of the frenulum to a more complex multi‐scale classification system such as the Hazelbaker Assessment Tool for Lingual Frenulum Function (ATLFF) (Hazelbaker 1993). The ATLFF is a highly reliable screening tool (Amir 2006) that was designed to be used in assessment of infants younger than three months of age. It assesses the function and appearance of the frenulum. A score of 14 indicates normal function, between 11 and 14 is acceptable and less than 11 indicates significant tongue‐tie that requires frenotomy. An appearance scale includes values to 10, and lower scores indicate tongue‐tie. A score lower than eight is suggestive of tongue‐tie, but surgery is not recommended unless a functional problem is noted.
Description of the intervention
Frenotomy and frenuloplasty are the two main surgical procedures used in the treatment of infants with tongue‐tie (Lalakea 2002). Frenotomy, or clipping of the frenulum, is the procedure of choice in infants because it is relatively quick and easy to perform. The infant is swaddled and is placed supine on the examining table, while an assistant supports the head and neck. The clinician/surgeon elevates the tongue and exposes the frenulum, which is then incised with sharp, straight, blunt‐ended scissors (Berry 2012; Hogan 2005). Some operators describe crushing the frenulum before incision. Direct pressure is then applied to the frenulum with a piece of gauze. Bleeding is reportedly scant and is controlled by the pressure (Lalakea 2002). The incision usually is not sutured, and the infant most often recovers quickly from the procedure and is able to feed directly afterwards. In infants, frenotomy is usually performed without analgesia or anaesthetic (Lalakea 2002). The use of a laser to perform the frenotomy is becoming more frequent (Kotlow 2011).
Frenuloplasty, an operation that lengthens the frenulum, is the preferred procedure for patients over one year of age (Lalakea 2002). This intervention will not be included in this review.
How the intervention might work
Surgical release of the tongue‐tie through frenotomy may correct the restriction to infant tongue movement during feeding to allow more effective breastfeeding and less maternal nipple pain resulting from decreased friction between the infant's lower gum/tongue and the nipple (Kumar 2012).
Why it is important to do this review
Diagnosis and management of tongue‐tie remain controversial.
It is uncertain whether ankyloglossia is a congenital oral anomaly requiring treatment or a normal variant. One survey (Messner 2000b) found that most lactation consultants believe tongue‐tie to be a frequent cause of infant breastfeeding difficulties that could be solved by frenotomy. In marked contrast, 90% of paediatricians and 70% of otolaryngologists believe that tongue‐tie never, or rarely, causes a feeding problem (Messner 2000a). However, medical organisations such as the American Academy of Pediatrics (Coryllos 2004) and the National Institute for Health and Care Excellence (NICE 2005) now acknowledge that tongue‐tie, or ankyloglossia, is a significant clinical entity that should be treated as early as possible to minimise breastfeeding problems. Given that breastfeeding benefits both infants and mothers, it is important for the clinician to address any condition that may impair breastfeeding (Edmunds 2011).
Objectives
To determine whether frenotomy is safe and effective in improving ability to feed orally among infants younger than three months of age with tongue‐tie (and problems feeding).
Subgroup analysis
-
Severity of tongue‐tie before frenotomy as measured by a validated tool (e.g. ATLFF (scores < 11; scores ≥ 11)) (Hazelbaker 1993).
-
Gestational age at birth (< 37 weeks' gestation; 37 weeks' gestation and above).
-
Method of feeding (breast or bottle).
-
Age at frenotomy (≤ 10 days of age; > 10 days to three months of age).
-
Severity of feeding difficulty (infants with feeding difficulty affecting weight gain (as assessed by infant's not regaining birth weight by day 14 or falling off centiles); infants with symptomatic feeding difficulty but thriving (greater than birth weight by day 14 and tracking centiles).
Methods
Criteria for considering studies for this review
Types of studies
Randomised or quasi‐randomised controlled trials or cluster‐randomised trials.
Types of participants
Infants three months of age or younger with a diagnosis of tongue‐tie who are orally feeding. To be included, another problem must be present that could be related to the tongue‐tie, specifically, infant feeding problems or maternal nipple pain in a breastfeeding mother. We planned to exclude patients with other coexisting oral pathology that might affect oral feeding, for example, cleft palate.
Types of interventions
-
Frenotomy versus no frenotomy. Lactation consultant interventions were accepted if provided to both groups.
-
Frenotomy versus sham procedure. Lactation consultant interventions were accepted if provided to both groups.
Types of outcome measures
Primary outcomes
-
Infant feeding assessed within 48 hours, within two to seven days and after seven days following the procedure with the use of a validated scale such as the LATCH score (Jensen 1994) or the Infant Breastfeeding Assessment Tool (IBFAT) (Mathews 1988). The LATCH is a scoring system that assesses latch, swallowing, maternal nipple, maternal comfort and assistance the mother needs to position the infant. Each area gets a score between 0 and 2 with a total possible score of 10. The IBFAT assesses readiness to feed, rooting, fixing (latching on) and sucking. Each item is scored between 0 and 3, and a score of 10 to 12 represents successful breastfeeding. Bottle‐feeding will be assessed within 48 hours, within two to seven days and after seven days following the procedure, subjectively, by reports of more efficient sucking and less drooling
Secondary outcomes
-
Maternal nipple pain assessed within 48 hours, within two to seven days and after seven days following frenotomy by a validated pain scale (e.g. Short‐Form McGill Pain Questionnaire (SF‐MPQ)) (Melzack 1975)
-
Qualitative assessment of infant feeding by parental survey performed within 48 hours of the procedure
-
Duration of breastfeeding (days)
-
Cessation of breastfeeding as assessed by maternal report within four weeks of the procedure
-
Infant pain as assessed by a validated pain scale (e.g. Modified Behavioral Pain Scale (MBPS) (Taddio 1995), Neonatal Infant Pain Scale (NIPS) (Lawrence 1993), CRIES pain scale (cries, requires oxygen, shows increased vital signs and expression and is sleepless) (Krechel 1995)) before, during and up to one hour post frenotomy
-
Excessive bleeding at the time or within 24 hours of frenotomy (as determined by study investigators)
-
Infection at the site of frenotomy requiring treatment with antibiotics within seven days of the procedure
-
Damage to the tongue or submandibular ducts noted within seven days of the procedure (as determined by study investigators)
Search methods for identification of studies
Electronic searches
Two review authors (JO'S, JF) independently performed electronic database searches, including electronic searches of the Cochrane Central Register of Controlled Trials (CENTRAL; 2016, Issue 1), MEDLINE (1966 to January 2016), Embase (1980 to January 2016) and the Cumulative Index to Nursing and Allied Health Literature (CINAHL; 1982 to January 2016), and searched previous reviews including cross‐references, expert informants and journal handsearching. We searched MEDLINE, Embase and CINAHL for relevant articles using the following search terms: Infant AND Tongue Tie, Infant OR Newborn OR neonate (explode) [MeSH heading] AND Tongue Tie (explode) [MeSH heading] OR ankyloglossia. and Frenotomy [MeSH heading] OR Frenulotomy OR Frenuloplasy [MeSH heading]. We applied no language restrictions. We also searched clinical trial registries for current and recently completed trials (Australia and New Zealand Clinical Trials Register (ANZCTR); clinicaltrials.gov; controlled-trials.com; who.int/ictrp; and Oxford Database of Perinatal Trials).
Searching other resources
The search strategy included communication with expert informants and searches of bibliographies of reviews and trials for references to other trials, as well as searches of previous reviews including cross‐references, abstracts and conferences and symposia proceedings of the Perinatal Society of Australia and New Zealand and the Pediatric Academic Societies (American Pediatric Society, Society for Pediatric Research and European Society for Pediatric Research) from 1990 to 2015. We planned to contact the corresponding investigator of any unpublished trials to request information. We considered unpublished studies and studies reported only as abstracts as eligible for review if final trial data were available. We intended to contact the corresponding authors of identified randomised controlled trials (RCTs) to ask for additional information about their studies if we required further data.
Data collection and analysis
We used the standard methods of Cochrane as documented in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011) and the methods of the Cochrane Neonatal Review Group (CNRG).
Selection of studies
Review authors independently assessed for inclusion all potential studies identified by the search strategy. We resolved disagreements through discussion.
Specifically, we:
-
merged search results by using reference management software and removed duplicate records of the same report;
-
examined titles and abstracts to remove irrelevant reports;
-
retrieved full text of potentially relevant reports;
-
linked together multiple reports of the same study;
-
examined full‐text reports to assess for compliance of studies meeting eligibility criteria;
-
corresponded with investigators, when appropriate, to clarify study eligibility;
-
at all stages noted reasons for inclusion and exclusion of articles; and
-
made final decisions on study inclusion and proceeded to data collection.
Data extraction and management
Review authors independently extracted data from full‐text articles using a specifically designed spreadsheet to manage the information. We resolved discrepancies through discussion; if required, we intended to consult a review arbiter. We entered data into Review Manager software 5.3 (RevMan 2014) and checked them for accuracy.
Assessment of risk of bias in included studies
We used the standardised review methods of the CNRG (http://neonatal.cochrane.org/en/index.html) to assess the methodological quality of included studies. Review authors independently assessed study quality and risk of bias using the following criteria, as documented in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).
-
Random sequence generation: Was the allocation sequence adequately generated?
-
Allocation concealment: Was allocation adequately concealed?
-
Blinding of participants and personnel for each main outcome or class of outcomes: Was knowledge of the allocated intervention adequately prevented during the study?
-
Blinding of outcome assessors: Were the outcome assessors blinded?
-
Incomplete outcome data for each main outcome or class of outcomes: Were incomplete data adequately addressed?
-
Selective outcome reporting: Are reports of the study free of the suggestion of selective outcome reporting? We tried to locate protocols to assess outcome reporting bias.
-
Other sources of bias: Was the study apparently free of other problems that could put it at high risk of bias? We gave particular attention to completeness of follow‐up of all randomly assigned infants and to the length of follow‐up studies to identify whether any benefits claimed were robust.
We intended to request additional information and clarification of published data from the authors of individual trials if required. We assessed each trial for risk of bias based on the criteria listed above and marked each as:
-
'low' risk of bias;
-
'unclear' risk of bias; or
-
'high' risk of bias.
We judged each criterion as being at 'low risk' of bias, 'high risk' of bias or 'unclear' risk of bias (for lack of information or uncertainty over the potential for bias).
Measures of treatment effect
We analysed the results of included studies using the statistical package Review Manager software 5.3 (RevMan 2014).
We used the standard method of the CNRG, applying a fixed‐effect model for meta‐analysis. In assessing treatment effects for dichotomous data, we reported the risk ratio (RR) and the risk difference (RD), along with 95% confidence intervals, for categorical outcomes. If the RD was statistically significant, we calculated the number needed to treat for an additional beneficial outcome (NNTB) and the number needed to treat for an additional harmful outcome (NNTH) (1/RD). For outcomes measured on a continuous scale, we used the weighted mean difference, along with 95% confidence intervals.
Included studies reported outcomes of change in infant feeding ability and in maternal pain using different validated scales. For infant feeding, researchers presented LATCH scores (a 10‐point scale) and IBFAT scores (a 12‐point scale). Included studies reported maternal pain assessed by the 10‐point visual analogue pain scale and the 50‐point SF‐MPQ. As these different scales rely on very different units of reporting and show subtle differences, we did not combine scores for analysis, and we presented results in subgroups according to the different scales used.
Unit of analysis issues
We combined randomised trials in a single meta‐analysis using the generic inverse variance method.
Dealing with missing data
We planned to contact the authors of all published studies if we required clarification or additional information. We planned to describe the number of participants with missing data in the Results section and in the Characteristics of included studies table. We presented results only for available participants. We intended to discuss the implications of missing data in the Discussion section of the review.
Assessment of heterogeneity
We used RevMan 5.3 (RevMan 2014) to assess the heterogeneity of treatment effects between trials. We used the two formal statistical approaches described below.
-
Chi2 test for homogeneity: We calculated whether statistical heterogeneity is present by using the Chi2 test for homogeneity (P < 0.1). Because this test has low power when the number of studies included in the meta‐analysis is small, we set the probability at the 10% level of significance (Higgins 2011).
-
I2 statistic, to ensure that pooling of data was valid: We quantified the impact of statistical heterogeneity by using I2 statistics available in RevMan 2014, which describes the percentage of total variation across studies due to heterogeneity rather than to sampling error. We graded the degree of heterogeneity as follows: 0% to 30%: potentially trivial (not important) heterogeneity; 31% to 50%: low heterogeneity; 51% to 75%: moderate heterogeneity; and 76% to 100%: high heterogeneity. Had we found evidence of apparent or statistical heterogeneity, we planned to assess the source of heterogeneity by performing sensitivity and post hoc subgroup analyses to look for sources of bias or methodological differences between heterogeneous trials (e.g. differences in study quality, participants, intervention regimens, outcome assessments).
Assessment of reporting biases
We tried to obtain the study protocols of all included studies to compare outcomes reported in the protocol versus those reported in the findings for each of the included studies. We intended to investigate reporting and publication bias by examining the degree of asymmetry of a funnel plot if we identified 10 or more trials. When we suspected reporting bias (see selective reporting bias above), we intended to contact study authors to ask them to provide missing outcome data. When this was not possible and we suspected that missing data might introduce serious bias, we intended to explore the impact of including such studies in the overall assessment of results by performing a sensitivity analysis.
Data synthesis
We performed statistical analyses according to the recommendations of the CNRG (http://neonatal.cochrane.org/en/index.html). We analysed all randomly assigned infants on an intention‐to‐treat (ITT) basis. We analysed treatment effects in individual trials and used a fixed‐effect model for meta‐analysis in the first instance to combine the data. When substantial heterogeneity existed, we examined the potential cause of heterogeneity by performing subgroup and sensitivity analyses. When we judged the meta‐analysis to be inappropriate, we analysed and interpreted individual trials separately. For estimates of typical risk ratio and risk difference, we used the Mantel‐Haenszel method. For measured quantities, we used the inverse variance method.
Quality of evidence
We used the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach, as outlined in the GRADE Handbook (Schünemann 2013), to assess the quality of evidence for infant breastfeeding when assessed by a validated scale and maternal nipple pain.
Two review authors independently assessed the quality of the evidence for each of the outcomes above. We considered evidence from RCTs as high quality but downgraded the evidence one level for serious (or two levels for very serious) limitations on the basis of the following: design (risk of bias), consistency across studies, directness of the evidence, precision of estimates and presence of publication bias. We used the GRADEpro Guideline Development Tool to create summary of findings Table 1 to report the quality of the evidence.
The GRADE approach results in an assessment of the quality of a body of evidence according to one of four grades.
-
High: We are very confident that the true effect lies close to that the estimate of effect.
-
Moderate: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of effect but may be substantially different.
-
Low: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of effect.
-
Very low: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
Subgroup analysis and investigation of heterogeneity
We planned to carry out the following subgroup analyses.
-
Severity of tongue‐tie as measured by a validated tool (e.g. ATLFF (scores < 11; scores ≥ 11)) (Hazelbaker 1993).
-
Gestational age at birth (< 37 weeks' gestation; 37 weeks' gestation and above).
-
Method of feeding (breast or bottle).
-
Age at frenotomy (≤ 10 days of age; > 10 days to three months of age).
-
Severity of feeding difficulty (infants with feeding difficulty affecting weight gain (assessed by infant's not regaining birth weight by day 14 or dropping off centiles by three months); infants with feeding difficulty but normal weight gain).
Sensitivity analysis
We intended to explore methodological heterogeneity through the use of sensitivity analysis. We classified studies as having low risk of bias if they had adequate sequence generation and allocation concealment and reported losses less than 10% on ITT analysis.
Results
Description of studies
See Characteristics of included studies and Characteristics of excluded studies.
Results of the search
We present a summary of our search in Figure 1. We encountered no disagreement between assessors (JO'S, JF, DT) regarding inclusion or exclusion of studies, quality assessment or data extraction. We pooled available data and analysed them as listed below.
Study flow diagram.
Included studies
We included five small trials in this review (N = 302).
-
Berry 2012 is a single‐centre study performed in the United Kingdom.
-
-
Objective: to investigate if an immediate and sustained improvement in breastfeeding occurs following frenotomy.
-
Population: infants younger than four months of age with tongue‐tie and breastfeeding difficulties.
-
Intervention: infants randomised to frenotomy or no frenotomy.
-
Outcomes: subjective and objective feeding assessment and maternal pain scores.
-
-
Buryk 2011 is a single‐centre study performed in America.
-
-
Objective: to investigate if frenotomy decreased maternal pain, improved breastfeeding scores and led to breastfeeding over a longer period.
-
Population: infants younger than 30 days of age with breastfeeding difficulties whose mothers experienced nipple pain.
-
Intervention: infants randomised to frenotomy or sham procedure.
-
Outcomes: change in maternal pain scores and infant feeding scores.
-
-
Dollberg 2006 is a single‐centre study performed in Israel.
-
-
Objective: to investigate if frenotomy in infants with ankyloglossia improves nipple pain and/or latching on during attempts to breastfeed.
-
Population: breastfeeding infants younger than 21 days of age whose mothers were experiencing nipple pain.
-
Intervention: infants randomised to frenotomy followed by a sham procedure or a sham procedure followed by frenotomy.
-
Outcomes: change in maternal pain scores and infant latch scores.
-
-
Emond 2013 is a single‐centre study performed in the United Kingdom.
-
-
Objective: to investigate if immediate frenotomy was superior to breastfeeding support.
-
Population: breastfeeding infants younger than two weeks of age with mild to moderate tongue‐tie and breastfeeding difficulties.
-
Intervention: infants randomised to frenotomy or standard care.
-
Outcomes: change in breastfeeding scores and maternal pain scores.
-
-
Hogan 2005 is a single‐centre study performed in the United Kingdom.
-
-
Objective: to investigate if frenotomy was superior to standard care in infants with tongue‐tie and feeding difficulties.
-
Population: bottle‐feeding and breastfeeding infants with tongue‐tie and feeding difficulty.
-
Intervention: infants randomised to intensive lactation support or frenotomy.
-
Outcomes: subjective improvement in feeding.
-
Excluded studies
See Characteristics of excluded studies. We excluded two studies (Ngerncham 2013; Yousefi 2015) ‐ the first because it was a cross‐sectional study with no control group, the second because it compared frenotomy versus frenuloplasty in children up to 12 years of age.
Risk of bias in included studies
We included in the analysis randomised controlled trials that compared frenotomy versus no frenotomy or frenotomy versus sham procedure in newborn infants. Overall, the five included trials had small study populations and a high incidence of methodological shortcomings. We discuss below specific methodological issues regarding these studies. See the risk of bias graph in Figure 2 and summary in Figure 3.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Allocation
Random sequence generation
All five studies reported the method of randomisation. Berry 2012, Buryk 2011, Dollberg 2006 and Hogan 2005 used computer‐generated randomisation, and Emond 2013 used independent telephone‐based randomisation.
Allocation concealment
Four studies described the methods used to conceal allocation. Berry 2012 and Hogan 2005 used sealed envelopes created by a person not involved in the trial, and Dollberg 2006 used sealed envelopes and specified that the envelopes were opaque but did not state who created them. Emond 2013 used a randomisation code that was kept by an independent body. Buryk 2011 did not state how investigators concealed allocation.
Blinding
Blinding of participants and research or clinical personnel
All five studies reported this. Hogan 2005 did not blind participants nor personnel. Buryk 2011 performed single‐blinding (mothers only, not treating personnel), Emond 2013 also used single‐blinding (researchers blinded, mothers not blinded) and Berry 2012 and Dollberg 2006 blinded both participants and personnel.
Blinding of outcome assessment
All five studies reported this. Berry 2012, Buryk 2011 and Dollberg 2006 blinded the assessor. Emond 2013 blinded the assessor of the primary outcome and assessors of some secondary outcomes but did not blind the mother and did not blind the assessors of other secondary outcomes. Hogan 2005 was an unblinded study.
Incomplete outcome data
Exclusions after randomisation
Berry 2012 had three postrandomisation exclusions (5% of randomised participants), and Dollberg 2006 had one postrandomisation exclusion (4% of randomised participants). Both studies reported that exclusions were due to failure of blinding. Emond 2013 had one postrandomisation exclusion (< 1% of randomised participants) resulting from loss to follow‐up. Buryk 2011 and Hogan 2005 had no postrandomisation exclusions.
Selective reporting
Two studies selectively reported outcomes. Berry 2012 reported that infant latch onto the breast objectively improved but did not provide data; Berry 2012 also reported on breastfeeding success at three months but did not report results by study group. Buryk 2011 reported no differences between groups in longer‐term breastfeeding rates but did not present results of the two groups separately. Dollberg 2006 reported changes in pain scores before and after the intervention in the intervention arm only and on request produced scores for the control arm.
Other potential sources of bias
All five studies offered and provided frenotomy to control infants, suggesting lack of equipoise. Two studies ‐ Berry 2012 and Dollberg 2006 ‐ performed frenotomy on all participants as part of the study protocol. Buryk 2011, Emond 2013 and Hogan 2005 offered frenotomy to controls and had uptakes of 77% to 97%.
Effects of interventions
Frenotomy versus no frenotomy or sham procedure
Five studies compared frenotomy versus no frenotomy (Berry 2012; Emond 2013; Hogan 2005) or frenotomy versus sham procedure (Buryk 2011; Dollberg 2006). We summarise the main findings in summary of findings Table 1.
Primary outcomes
Infant feeding assessed with a validated tool (Analysis 1.1)
Four trials provided outcome data for this comparison (Berry 2012; Buryk 2011; Dollberg 2006; Emond 2013). Three studies reported infant breastfeeding assessed with a validated scale (Buryk 2011; Dollberg 2006; Emond 2013) based on two different measurement scales (Analysis 1.1; Figure 4).
Forest plot of comparison: 1 Frenotomy versus no frenotomy or sham procedure, outcome: 1.1 Infant breastfeeding assessed by a validated scale.
-
IBFAT score (Buryk 2011): MD 3.50 on a 12‐point scale, 95% CI 3.06 to 3.94; 58 participants (Analysis 1.1.1).
-
LATCH score (Dollberg 2006; Emond 2013): MD ‐0.07 on a 10‐point scale, 95% CI ‐0.63 to 0.48; 155 participants (heterogeneity: Chi² = 0.02, df = 1, P = 0.88; I² = 0%) (Analysis 1.1.2).
A fourth trial (Berry 2012) also reported that infants in the intervention group had higher LATCH scores after the intervention but provided no data.
We did not combine LATCH and IBFAT scores for meta‐analysis, as studies had a high degree of heterogeneity and scales used different units of measure.
Subgroup analyses
Severity of tongue‐tie
One trial included only infants with severe tongue‐tie (Buryk 2011) and reported improvement in infant breastfeeding (increase in IBFAT scores) in the frenotomy group compared with the control group: MD 3.50 on a 12‐point scale, 95% CI 3.06 to 3.94; 58 participants.
One trial included only infants with moderate tongue‐tie (Emond 2013) and found no objective improvement in feeding scores: MD ‐0.10 on a 10‐point scale, 95% CI ‐0.75 to 0.55; 105 participants.
Timing of feeding assessment
Two trials reported on infant breastfeeding within 48 hours of the intervention (Buryk 2011; Dollberg 2006).
-
IBFAT score (Buryk 2011): MD 3.50 on a 12‐point scale, 95% CI 3.06 to 3.94; 58 participants.
-
LATCH score (Dollberg 2006): MD 0.0 on a 10‐point scale, 95% CI ‐1.08 to 1.08; 50 participants.
One trial (Emond 2013) reported no difference in LATCH scores assessed between two and seven days after the intervention: MD ‐0.10 on a 10‐point scale, 95% CI ‐0.75 to 0.55; 105 participants (Analysis 1.2).
Subgroup analysis by gestational age or by severity of feeding difficulty or method of feeding was not possible owing to lack of data.
Secondary outcomes
Effect on maternal nipple pain (Analysis 1.3)
Four trials reported this outcome (Berry 2012; Buryk 2011; Dollberg 2006; Emond 2013) using two different scales.
-
Berry 2012, Dollberg 2006 and Emond 2013 assessed this outcome after the first breastfeed using a 10‐point visual analogue pain scale. Meta‐analysis showed a significant reduction in maternal nipple pain in the frenotomy group compared with the control group: MD ‐0.74, 95% CI ‐1.35 to ‐0.13 (heterogeneity: Chi² = 1.85, df = 2, P = 0.40; I² = 0%) (Analysis 1.3.1).
-
Buryk 2011 used the 50‐point SF‐MPQ Pain Assessment Tool after five days of feeding and found a significant reduction in the frenotomy group compared with the control group: MD ‐8.60, 95% CI ‐9.37 to ‐7.83 (Analysis 1.3.2).
We did not combine visual analogue pain scale and SF‐MPQ scores for meta‐analysis because studies had a high degree of heterogeneity and scales used different units of measure.
Qualitative assessment of infant feeding by parental survey performed within 48 hours of the procedure (Analysis 1.4)
Two trials (Berry 2012; Hogan 2005) reported high rates of improvement in breastfeeding following frenotomy compared with control (typical risk ratio (RR) 3.48, 95% CI 2.18 to 5.56). The meta‐analysis contained high levels of heterogeneity (Chi² = 15.63, df = 1, P < 0.0001; I² = 94%) (Figure 5).
Forest plot of comparison: 1 Frenotomy versus no frenotomy or sham procedure, outcome: 1.6 Qualitative assessment of infant feeding by parental survey performed within 48 hours of the procedure.
Hogan 2005 reported that all intervention infants (n = 8) versus no control infants (n = 9) had improved bottle‐feeding.
Emond 2013 obtained qualitative data from a subgroup of the study population and found that most parents reported improvement in infant feeding directly after the intervention and in the days that followed, but investigators did not state when they performed the survey.
Duration of breastfeeding or cessation of breastfeeding as assessed by maternal report within four weeks of the procedure
No trial reported the duration of breastfeeding or cessation of breastfeeding as assessed by maternal report within four weeks. However, four studies reported on rates of breastfeeding at specific time points. Berry 2012 reported that 58/59 were breastfeeding at three months but did not report this by treatment group. Buryk 2011 reported that 36/58 were breastfeeding at two months, 23/58 at six months and 14/58 at 12 months. Study authors also did not report by treatment group. Emond 2013 reported that at eight weeks, 43/52 in the intervention group and 40/50 in the control group were at least partially breastfed. Hogan 2005 reported that 12/20 in the intervention group and 10/20 in the control group were breastfed for at least four months.
Infant pain assessed by a validated pain scale before, during and up to one hour post frenotomy
None of the included studies reported this outcome.
Excessive bleeding at the time or within 24 hours of frenotomy (Analysis 1.5)
No infants in any of the five trials experienced excessive bleeding following the intervention (RR not estimable; RD 0.00, 95% CI ‐0.03 to 0.03; 302 participants).
Infection at the site of frenotomy requiring treatment with antibiotics within seven days of the procedure (Analysis 1.6)
No infants in any of the five trials experienced infection at the intervention site requiring antibiotics (RR not estimable; RD 0.00, 95% CI ‐0.03 to 0.03; 302 participants).
Damage to the tongue or submandibular ducts noted within seven days of the procedure (Analysis 1.7)
No infants in any of the five trials experienced damage to the tongue or submandibular ducts (RR not estimable; RD 0.00, 95% CI ‐0.03 to 0.03; 302 participants).
Discussion
Summary of main results
We included in this review five trials with a total sample size of 302 infants. Only one small trial reported that frenotomy objectively improved infant breastfeeding (Buryk 2011). Pooled analysis of two other trials (N = 155) showed no objective change following frenotomy (Dollberg 2006; Emond 2013) but showed consistently reduced maternal nipple pain.
Studies to date have not answered the clinically relevant question of whether frenotomy in tongue‐tied infants with feeding difficulties results in longer‐term breastfeeding success and resolution of maternal pain. No studies reported duration of breastfeeding by group, and all had very high contamination rates in the control group, making longer‐term outcomes meaningless.
Overall completeness and applicability of evidence
All studies were performed at maternity hospitals in the United Kingdom, America and Israel. All centres that carried out the studies had lactation support and frenotomy readily available. The total number of infants studied to date is 302. This number is too small to confirm the efficacy and safety of frenotomy, even though no study reported any significant adverse effects. Populations studied to date were healthy term infants with moderate or severe tongue‐tie. All primary outcomes were short‐term, many assessed after just one feed following intervention. Some studies followed up infants for several months after the intervention, but as the contamination rate within the control group was very high, longer‐term results are meaningless. Studies to date have not answered the clinically relevant question of whether frenotomy in tongue‐tied infants results in longer‐term breastfeeding success and resolution of maternal pain.
The potential to improve infant breastfeeding while reducing maternal nipple pain may be greatest when an infant is given the diagnosis of severe tongue‐tie. However, only one trial has studied frenotomy in this population. Again, only one study has examined frenotomy in infants given the diagnosis of moderate tongue‐tie and found that frenotomy did not have an objective effect on infant feeding nor on maternal nipple pain but was subjectively effective.
Quality of the evidence
We assessed the quality of the evidence using the GRADE method and classified evidence for major outcomes as low quality on the basis of small sample sizes (both in individual studies and in combined studies), inconsistent blinding and high risk of bias. The included studies had low rates of participant drop‐out. Most study authors responded when we requested further information. We noted high degrees of heterogeneity between studies and observed that investigators used different diagnostic tools to diagnose the condition, reported differing degrees of tongue‐tie severity and used different scales to assess outcomes. The greatest weakness among studies to date is that investigators offered frenotomy to all controls and provided the procedure for a large majority of them, which strongly implies lack of equipoise about effectiveness of the intervention during the study and reduces the quality of results.
Potential biases in the review process
We used the standard methods of Cochrane Neonatal in conducting this systematic review. Our inclusive search strategy would have included all relevant studies.
Agreements and disagreements with other studies or reviews
Several other published reviews have examined effects of frenotomy for tongue‐tie in young infants (Algar 2009; Bowley 2014; CADTH 2016; Cho 2010; Edmunds 2011; Francis 2015; Hall 2005; Hong 2013; Ito 2014; Lalakea 2003; Power 2015; Segal 2007; Suter 2009). All included observational studies as well as randomised trials. All reviews recognised a role for frenotomy when evidence indicates feeding difficulties or maternal nipple pain. However, review authors provided different interpretations of the strength of available evidence supporting frenotomy. Algar 2009Bowley 2014Cho 2010Edmunds 2011Hong 2013Ito 2014Lalakea 2003 and Segal 2007 concluded that evidence is sufficient to recommend frenotomy in an infant with breastfeeding problems and tongue‐tie. CADTH 2016Hall 2005Power 2015Suter 2009 and Francis 2015 concluded that frenotomy may be helpful and is safe but that definitive evidence is lacking. The most recent review (CADTH 2016) concluded that frenotomy is a safe procedure that leads to maternally perceived benefit in short‐term breastfeeding outcomes, concurring with this review that objective improvement in symptoms and long‐term benefit is less certain. CADTH 2016 concluded that it remains to be proved whether frenotomy provides meaningful improvement in breastfeeding difficulties, especially over the long term. Another recent review (Francis 2015) included the same five randomised trials included in this review, together with a retrospective cohort and 23 case series. Those review authors concluded that the strength of the evidence supporting frenotomy for improved breastfeeding outcomes is low to insufficient and acknowledged that maternal reported effect is greater. They identified gaps in the evidence including longer‐term outcomes, applicability to infants born at non‐tertiary centres, lack of consistency in diagnosis, effectiveness of non‐surgical interventions and optimal age of intervention. Two reviews acknowledged that the optimal time of intervention is unknown (Bowley 2014; Power 2015). Bowley 2014 recognised that frenotomy can improve breastfeeding problems in a tongue‐tied child but advised waiting at least two weeks before performing the procedure. All reviews concluded that the procedure appears to have a low complication rate when performed by trained operators. Many reviews reported that available studies have methodological flaws and risk of bias.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Forest plot of comparison: 1 Frenotomy versus no frenotomy or sham procedure, outcome: 1.1 Infant breastfeeding assessed by a validated scale.
Forest plot of comparison: 1 Frenotomy versus no frenotomy or sham procedure, outcome: 1.6 Qualitative assessment of infant feeding by parental survey performed within 48 hours of the procedure.
Comparison 1: Frenotomy versus no frenotomy or sham procedure, Outcome 1: Infant breastfeeding assessed by a validated scale
Comparison 1: Frenotomy versus no frenotomy or sham procedure, Outcome 2: Infant breastfeeding assessed by a validated scale 2 to 7 days following procedure
Comparison 1: Frenotomy versus no frenotomy or sham procedure, Outcome 3: Maternal nipple pain assessed by a validated pain scale
Comparison 1: Frenotomy versus no frenotomy or sham procedure, Outcome 4: Qualitative assessment of infant feeding by parental survey performed within 48 hours of procedure
Comparison 1: Frenotomy versus no frenotomy or sham procedure, Outcome 5: Excessive bleeding at the time or within 24 hours of frenotomy (as determined by study investigators)
Comparison 1: Frenotomy versus no frenotomy or sham procedure, Outcome 6: Infection at the site of frenotomy requiring treatment with antibiotics within 7 days of procedure
Comparison 1: Frenotomy versus no frenotomy or sham procedure, Outcome 7: Damage to the tongue and/or submandibular ducts noted within 7 days of procedure (as determined by study investigators)
| Undefined | ||||||
| Patient or population: tongue‐tie in newborn infants | ||||||
| Outcomes | Illustrative comparative risks (mean and SD) | Relative effect | Number of participants | Quality of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with no frenotomy or sham procedure | Risk with frenotomy | |||||
| Infant breastfeeding assessed by validated scale ‐ IBFAT scores following procedure | Mean IBFAT scores following procedure in the control group was 8.1 (SD 0.9) | Mean IBFAT scores following procedure in the frenotomy group was 11.6 (SD 0.8) | Mean difference is 3.50 (3.06 to 3.94) | 58 | ⊕⊕⊝⊝ | IBFAT score is based on a 12‐point scale |
| Infant breastfeeding assessed by validated scale ‐ LATCH scores following procedure | Mean LATCH scores following procedure in the control group was 6.8 to 8.5 (SD < 1.9) | Mean LATCH scores following procedure in the frenotomy group was 6.8 to 8.4 (SD < 2) | Mean difference is ‐0.07 (‐0.63 to 0.48) | 155 | ⊕⊕⊝⊝ | LATCH score is based on a 10‐point scale |
| Maternal nipple pain assessed by a validated pain scale ‐ visual analogue pain scale | Mean visual analogue pain scale scores following procedure in the control group was 2.9 to 5.5 (SD < 2.6) | Mean IBFAT scores following procedure in the frenotomy group was 1.6 to 5.3 (SD < 2.4) | Mean difference is ‐0.74 (‐1.35 to ‐0.13) | 183 | ⊕⊕⊝⊝ | Visual analogue pain scale score is based on a 10‐point scale |
| Maternal nipple pain assessed by a validated pain scale ‐ SF‐MPQ pain scale | Mean SF‐MPQ scores following procedure in the control group was 13.5 (SD 1.5) | Mean IBFAT scores following procedure in the frenotomy group was 4.9 (SD 1.5) | Mean difference is ‐8.60 (‐9.37 to ‐7.83) | 58 | ⊕⊕⊝⊝ | SF‐MPQ score is based on a 50‐point scale |
| *The risk in the intervention group (and its 95% confidence interval) is based on assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI) | ||||||
| GRADE Working Group grades of evidence | ||||||
| aImprecision (small total participant population) bRisk of bias (incomplete blinding) | ||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1.1 Infant breastfeeding assessed by a validated scale Show forest plot | 3 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 1.1.1 IBFAT scores following procedure | 1 | 58 | Mean Difference (IV, Fixed, 95% CI) | 3.50 [3.06, 3.94] |
| 1.1.2 LATCH scores following procedure | 2 | 155 | Mean Difference (IV, Fixed, 95% CI) | ‐0.07 [‐0.63, 0.48] |
| 1.2 Infant breastfeeding assessed by a validated scale 2 to 7 days following procedure Show forest plot | 1 | 105 | Mean Difference (IV, Fixed, 95% CI) | ‐0.10 [‐0.75, 0.55] |
| 1.3 Maternal nipple pain assessed by a validated pain scale Show forest plot | 4 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 1.3.1 Visual analogue pain scale | 3 | 183 | Mean Difference (IV, Fixed, 95% CI) | ‐0.74 [‐1.35, ‐0.13] |
| 1.3.2 SF‐MPQ pain scale | 1 | 58 | Mean Difference (IV, Fixed, 95% CI) | ‐8.60 [‐9.37, ‐7.83] |
| 1.4 Qualitative assessment of infant feeding by parental survey performed within 48 hours of procedure Show forest plot | 2 | 114 | Risk Ratio (M‐H, Fixed, 95% CI) | 3.48 [2.18, 5.56] |
| 1.5 Excessive bleeding at the time or within 24 hours of frenotomy (as determined by study investigators) Show forest plot | 5 | 302 | Risk Difference (M‐H, Fixed, 95% CI) | 0.00 [‐0.03, 0.03] |
| 1.6 Infection at the site of frenotomy requiring treatment with antibiotics within 7 days of procedure Show forest plot | 5 | 302 | Risk Ratio (M‐H, Fixed, 95% CI) | Not estimable |
| 1.7 Damage to the tongue and/or submandibular ducts noted within 7 days of procedure (as determined by study investigators) Show forest plot | 5 | 302 | Risk Ratio (M‐H, Fixed, 95% CI) | Not estimable |
