The Rest of the Story
March 2017 Issue
Courtesy of Shutterstock Inc.
Just under three years ago, the O&P profession was startled by the publication of the first randomized controlled trial (RCT) on the efficacy of cranial remolding orthoses.1 It was somewhat surprising that an RCT had been conducted at all. Given the reasonable evidence base in support of cranial remolding orthoses, it seemed unethical to deprive families of this treatment modality for the sake of research. But more startling were the reported results- that neither remolding orthoses nor natural history provided meaningful correction to asymmetrical head shapes.1 Treating clinicians and prescribing physicians around the world were surprised by these published reports that ran so extremely counter to their own experiences.
And yet, the publication was released and joined the collective dialogue of published evidence in this area of study. However, while it made some declarations, it did not send out the final word on the topic, and the evidence base related to the management of deformational plagiocephaly has continued to evolve. This article reviews some of the literature that has been published since the Dutch RCT, most of which directly references that work in some form.
The Immediate Response
A number of letters to the editor were sent to the British Medical Journal (BMJ) following the publication. While insightful to review all the opinions that were received on the topic, for the purposes of this review, only the first and last response will be discussed.2,3 Speaking chronologically, the first response came from veteran cranial researchers, Kevin Kelly, PhD, and Timothy Littlefield, MSEng, authors of three of the studies referenced by the RCT. Their response begins with a summary of the RCT's findings: "van Wijk and associates found that the deformation was not corrected by either 'helmet treatment' or by the natural growth of the skull (i.e., no treatment). Seeing no difference, the authors conclude that 'the use of a helmet as a standard treatment for healthy infants' should be discouraged."2
This foundation established, the letter to the editor continued:
Of significance, the authors were able to demonstrate that the "natural history" of plagiocephaly is the skull deformity will not resolve on its own. This is a significant finding which may be overlooked in the hyperbole of their second conclusion, the overreaching dismissal of all helmet therapy. We assert that the author's second conclusion is egregiously in error, and that the lack of improvement seen in their study was the direct result of their own admittedly ill-fitting helmets.2
Returning to the first of the two summarized findings, Kelly and Littlefield remind their readers that the historical premise of the need for an RCT was not to determine whether helmets were effective, but whether they were necessary. Restating the latter position, do aberrant head shapes "round out on their own?"
Kelly and Littlefield then contend that, while it was not the stated objective of the RCT, it was ultimately successful in addressing whether active treatment of deformational plagiocephaly is indicated:
In a truly novel contribution to the medical literature, van Wijk and colleagues have documented that the natural growth did not correct the skull deformation. In other words, they have shown that untreated skull deformities persist. This finding suggests to us that the prevention and treatment of the skull deformation is justified and necessary.2
Having complimented the unintended beneficial contribution of the RCT to the current body of cranial literature, Kelly and Littlefield return to the claim that helmets do not work. Citing 15 peer-reviewed publications over a period of more than 20 years in which helmet therapy has been demonstrated as an effective treatment for skull deformation, Kelly and Littlefield question the stated position of the RCT, namely that, "Based on a single, aberrant study, van Wijk and colleagues have concluded that all helmet therapy should be discouraged."2
How is it that the RCT came to conclusions that were in such stark contrast to most of the published evidence on this topic? Kelly and Littlefield ascribe the fault to the fidelity of the treatments received as part of the RCT. Citing abnormally high rates of skin irritation, ill-fitting helmets, pain, and nonacceptance, Kelly and Littlefield question: "Was the treatment delivered in an accurate and consistent manner in accord with accepted standards? Unfortunately, the answer to that question is a resounding no."2
This first letter to the editor concludes by applauding the investigators for documenting the natural history of untreated plagiocephaly, followed by an admonishment against their hasty dismissal of all helmet therapy.
The Measured Response
Kelly and Littlefield's immediate concerns are included in the more measured response of the plastic surgery section of the American Academy of Pediatrics (AAP) Executive Committee written six months after the initial publication.3 While less pointed in their responses, the authors of this letter are no less assertive in their fundamental position that "although timely and intriguing given the prevalence of positional head shape deformity in the wake of the Back to Sleep campaign, the study suffers from important and critical weaknesses that warrant mention."3
Among these weaknesses are the low number of subjects; the exclusion of infants with severe deformities, torticollis, and developmental delays (making any conclusions poorly generalizable, given that many infants with plagiocephaly meet at least one of these criteria); a lack of objective followup assessment during the study; the unreasonably high rate of side effects; and recovery rates that suggest "overall inadequate treatment of the presenting problem" that "should be considered unacceptable."3
Asserting that "numerous studies have shown excellent results with cranial orthoses that have been fit properly and appropriately monitored by orthotic specialists," the authors of the AAP letter conclude, "In summary, we find significant weaknesses in the methodology employed by van Wijk et al. that necessarily call into question any conclusions about the lack of effectiveness of helmets."3
A Parrot's Echo
Unfortunately, no matter how poor the quality of the treatments provided, any RCT will garner attention, particularly if it is the first RCT of a treatment modality. This was the case with the Dutch RCT, exemplified in a 2015 study by Rowland and Das in The Journal of Family Practice. The article in question is described as a PURLs, which stands for Priority Updates From the Research Literature. The title is direct: "Helmets for Positional Skull Deformities: A Good Idea or Not?" The subtitle is troublesome: "Probably Not. Helmets Appear to Be No More Effective Than Waiting for Natural Skull Growth to Correct the Shape of an Infant's Head."
If the busy family practitioner doesn't have time to read the entire article, there is a simple pull-out message at the top of the front page in a blue font that can't be missed: "PRACTICE CHANGER. Do not recommend helmet therapy for positional skull deformity in infants and children. Wearing a helmet causes adverse effects but does not alter the natural course of head growth."4
The article goes on to summarize the Dutch RCT without a single mention of any concerns cited by other respected voices about its methodology or atypical findings. Instead, the second page is highlighted with a lateral pull-out summary that reads: "At the end of the study, improvement in skull shape was almost the same in the helmet therapy and control groups."4
If that family practitioner has failed to catch the authors' interpretation of the RCT's findings, the article concludes with the following warning:
CHALLENGES TO IMPLEMENTATION
Parents may find this evidence hard to accept. If another physician or physical therapist recommends helmet therapy-or if a parent requests it-explaining the findings of this study may be challenging.... Since this study shows that helmets don't help correct skull deformities, parents can be assured that a helmet is unnecessary, costly and causes adverse effects.4
So now we have The Journal of Family Practice encouraging its medical readers to help parents understand that helmets do not help correct skull deformities based on a single study that the AAP asserts "suffers from important and critical weaknesses," and suggests, "overall inadequate treatment of the presenting problem" that "necessarily call into question any conclusions about the lack of effectiveness of helmets."2 The result represents a substantial challenge to evidence-based medicine with different organizations affirming radically different interpretations and recommendations.
Fortunately, in the wake of the Dutch RCT, a number of experienced craniofacial centers published their clinical observations, supported by extremely large datasets (Table 1).
Eberle et al., 2015
In the introduction to the Eberle et al. study, the authors explicitly cite the Dutch RCT as featured in The Wall Street Journal stating that watchful waiting of deformational plagiocephaly was as effective as helmet therapy.5 Their stated aim was to "determine the effectiveness of helmet therapy for positional skull deformation in a single provider's practice."5
Against the 84 subjects in the Dutch study, Eberle et al. report upon 4,248 patients managed over a five-year period. In contrast to van Wijk et al., who observe that neither helmet therapy nor natural history improved cranial deformities, Eberle et al. describe the following: "In a conservative treatment group of 3,186 infants managed with repositioning, 75 percent achieved complete correction, defined as a cranial diagonal difference of less than 6mm. The remaining 25 percent were moved to a second treatment group managed with a remolding orthosis. Of these 1,062 infants, 95 percent achieved full correction as defined above, over an average treatment duration of five months."
The authors' conclusion is straightforward, "Conservative and helmet therapy can effectively treat positional plagiocephaly.... These results are consistent with previous studies, but conflicts [with] the BJM article. This is important, as this article can be the source of insurance denials for infants who could clearly benefit from cranial molding orthotics."5
Steinburg et al., 2015
Near the same time as the Eberle et al. study, authors from Children's Memorial Hospital in Chicago released their collective observations from a seven-year period.6 A conservative treatment group that comprised 3,381 infants was managed through prescribed repositioning, and, in the majority of cases, physical therapy. Of these, 77 percent obtained complete correction (defined by this clinic as a transcranial diagonal difference of less than 5mm). Of those who failed to attain complete correction, the majority transitioned to a helmet treatment group, with a minority failing to obtain complete correction and declining further treatment. The helmet group comprised 997 infants who were initially prescribed a helmet and the 534 infants who joined the group after conservative treatments failed, yielding a population of 1,531 infants.
Complete correction by the standards reported above was obtained by 94 percent of those infants initially managed with a remolding helmet, and by 96 percent of those infants transitioned after failed conservative management attempts. Collectively, 4,062 of the 4,378 infants in the retrospective analysis attained successful correction of their cranial deformations through a combination of active repositioning, physical therapy, and orthotic remolding. The authors' conclusions are direct, "Conservative treatment and helmet therapy were found to be effective for correcting positional cranial deformation in 92.8 percent of infants."6
Wilbrand et al., 2016
The final article lacks the high patient volumes of the prior two studies, but is remarkable in its length of follow-up as it addresses the opposing outcomes observed with treated versus untreated plagiocephaly.7 The first line of the article's introduction addresses the claims and resultant opinions associated with the Dutch RCT:
A recently published randomized controlled trial regarding helmet versus no helmet therapy for early infant cranial deformation reported no benefit of orthotic treatment compared with a control group. Based on these findings and data reported by others, pediatricians, craniofacial surgeons, physiotherapists, and other professionals engaged in cranial deformation are becoming increasingly divided into advocates and opponents of orthotic correction of significant aberrance of cranial shape in early infancy.
Opponents describe side effects, such as discomfort to the child, the high cost of helmet therapy to the healthcare system, skin irritation, and the absence of functional impairment caused by an abnormal cranial shape.... In contrast to those publications, advocates of helmet therapy cite data regarding the persistence of significant deformation, mandibular asymmetry, and neurocognitive deficiency as more of a trigger than a consequence of cranial deformity. Advocates also state that sociopsychologic impairment might justify early correction of nonsynostotic deformity....
The current manuscript presents data collected during a time frame of five years after the first assessment of children with nonsynostotic cranial deformity that were either treated with an individual molding helmet or left untreated.7
The authors retrospectively selected 41 children with substantial cranial deformations who were not treated with remolding orthoses. A second group of 41 children who received treatment was subsequently selected from a potential pool of 859 children. To the extent possible, the baseline severity of their cranial deformities was matched to that observed in the first group. Then, approximately five years later, these groups returned to the clinic to obtain current cranial measurements from clinicians blinded to whether the children had undergone orthotic cranial remolding.
Even when choosing the more severe deformities among the untreated children and the less severe cases in the children who received orthotic treatment, the median deformity was more severe in the second group, with median values for initial cranial vault asymmetry (CVA) reported at 1cm and 1.65cm, respectively. For the children in the treatment group, a second set of measurements was taken about six months after their initial assessment to track the efficacy of the orthotic intervention. The authors report that the median post-helmet CVA was 0.2cm.7
Aggregating the five-year follow-up data, the authors report that the median CVA in the untreated group was unchanged, remaining at 1cm. For the children who received orthotic management, the median CVA began at 1.65cm, reduced to 0.2cm after roughly six months of helmet use, and regressed slightly to 0.3cm at the five-year follow-up.7
Within the body of their discussion, the authors assert, "In contrast to the results of other authors, our data did not show significant spontaneous improvement in cranial deformation within a timeframe of five years."7 Further, "we assert that only children with a history of helmet therapy recovered in our study population."
In comparing their observations with those of the Dutch RCT, Wilbrand et al. include the following:
The frequency of complications associated with helmet therapy has been previously evaluated and has not perceivably changed since then. In contrast, the number of complications observed by van Wijk et al. seems very high (up to 96 percent) compared with the data of other authors. This might be an important cause of the discouraging results in the treated group and the equality of the effects between the groups.
van Wijk et al. reported that 75 percent of all children had persistent skull deformation at two years of age regardless of whether the children were treated with an individual molding helmet. Such a high degree of uncorrected cranial deformities could lead practitioners to forgo performing any therapeutic procedures for clinically significant cranial deformity and define nonsynostotic cranial deformity in early childhood as an incurable burden. The disturbance in the individual effectiveness of the molding helmets used in the van Wijk et al. study-which might be because of the high numbers of side effects-was not discussed by the authors.
Although the results of the current study might not represent the final conclusion regarding the use of helmets for infant nonsynostotic cranial deformation, we have shown that helmet therapy does have a significant effect and nonsynostotic cranial deformity is correctable.7
And so, the dialogue related to the value and efficacy of cranial remolding orthoses continues. The general consensus at this point seems to be that the Dutch RCT was a high-level study of a poorly controlled and executed treatment modality. In retrospect, it is hardly surprising that helmets that were poorly fit, causing uncharacteristically frequent side effects, were found to be ineffective. Moreover, recent studies have countered the findings of the RCT with four-digit n-values, declaring loudly and simply that helmets work. Assuming the helmets are appropriately constructed and managed, it is expected that future studies, along with these recent trials, will continue to counter the bold declarations of a single, flawed RCT.
Phil Stevens, MEd, CPO, FAAOP, is in clinical practice with Hanger Clinic, Salt Lake City. He can be contacted at firstname.lastname@example.org.
- van Wijk, R. M. et al. 2014. Helmet therapy in infants with positional skull deformation: Randomized controlled trial. British Medical Journal 348, g2741
- Kelly, K. M., and T. R. Littlefield. Response to BMJ 2014;348:g2741. Accessed at www.bmj.com/content/348/bmj.g2741/rr/697461
- Steinburg, J. P., et al. American Academy of Pediatrics Executive Committee- Plastic Surgery Section. Response to BMJ 2014;348:g2741. Accessed at www.bmj.com/content/348/bmj.g2741/rr/779173
- Rowland, K., and N. Das. 2015. PURLs: Helmets for positional skull deformities: A good idea, or not? Journal of Family Practice 64 (1):44-6. Accessed at www.ncbi.nlm.nih.gov/pmc/articles/PMC4294410/pdf/JFP-64-44.pdf
- Eberle, N. A., E. J. Stelnicki, and B. Boland. 2015. Efficacy of conservative and cranial orthotic therapy in over 4000 patients treated for positional plagiocephaly over a five-year period. Plastic and Reconstructive Surgery 136 (4S), 5-6.
- Steinberg, J. P., R. Rawlani, L. S. Humphries, V. Rawlani, and F. A. Vicari. 2015. Effectiveness of conservative therapy and helmet therapy for positional cranial deformation. Plastic and Reconstructive Surgery 135 (3):833-42.
- Wilbrand, J. F., N. Lautenbacher, J. Pons-Kuhnemann, P. Streckbein, C. Kahling, M. H. Reinges, H. P. Howaldt, and M. Wilbrand. 2016. Treated versus untreated positional head deformity. Journal of Craniofacial Surgery 27 (1):13-8.