Abstract
Context:
Term neonates treated with therapeutic hypothermia (TH) may experience delays in acquiring skills for oral feeding, thus prolonging hospital admission.
Objective:
To determine whether osteopathic manipulative treatment (OMT) can decrease the overall length of stay (LOS) for term neonates by accelerating the transition to full oral feeds.
Methods:
A pilot study was conducted to compare LOS in term neonates treated with OMT to matched historical controls. Eligibility criteria were gestational age greater than or equal to 37 weeks, mild to moderate encephalopathy, absent seizure activity, and no brain injury on magnetic resonance imaging. Treated neonates had OMT twice between day of life four and seven, then twice per week on nonconsecutive days until hospital discharge. Secondary outcomes were to compare the number of days with nasogastric tube in both OMT and historically matched control groups and to characterize somatic dysfunction patterns in the primary affected areas of the craniosacral mechanism.
Results:
Twelve of 28 neonates treated with TH between October 2017 and August 2018 met eligibility criteria for the study and were matched 3:1 to 36 historical controls. On average, mothers of neonates who received OMT were older than the historical control mothers (31.3 years [SD, ±6.0] vs. 27.4 [SD, ±4.5]; p=0.02) Nineteen of the 36 historical control neonates, but no neonates in the OMT group, were intubated for a median of two days (interquartile range 1, 4). The mean LOS in the OMT group was 9.1 vs. 11.6 days for historical controls (p=0.048); however, in a sensitivity analysis excluding intubated neonates from the historical controls, the difference was 9.1 vs. 10.1 days (p=0.21). All neonates were orally feeding at discharge. Neonates given OMT had between two and four treatments lasting 8–20 minutes. By the final treatment, the craniosacral mechanism’s intraosseous and cranial base compressions and physiologic motion were improved. There were no adverse effects noted during OMT sessions.
Conclusion:
Our findings suggest that OMT performed on term neonates treated with TH may decrease the amount of time needed to acquire the necessary skills for oral feeding. Although this result was not statistically significant, it may be clinically significant. A larger prospective clinical trial may have the power needed to detect a statistically significant reduction in LOS and number of days to full oral feeds in this patient population.
Roughly two to five per 1,000 births are affected by neonatal encephalopathy (NE); this term describes abnormal neurologic examination features in a neonate in the first days of life related to hypoxia and ischemia around the time of birth.
1 The current standard of care for neonates with NE is to initiate a neuroprotective therapy, therapeutic hypothermia (TH), within the first six hours of life to increase chances of survival and preserve neurodevelopment.
2, 3, 4, 5 TH is achieved by cooling the neonate to 33–34 °C for 72 hours, followed by a gradual period of rewarming over more than 12 hours. During this time, neonates are not fed enterally because of concern for potential decreased gut perfusion and associated complications.
6 As such, a barrier to hospital discharge for these neonates is a successful transition from nasogastric (NG) feeds to either breast or bottle feeding. In neonates without brain injury on magnetic resonance imaging (MRI), hospitalization is often briefly extended to allow the neonate to develop the skills necessary to achieve full oral feeding before discharge. For neonates with brain injury, depending on the severity, these skills may take much longer to develop, or may never be acquired, resulting in need for percutaneous endoscopic gastrostomy tube placement.
Osteopathic manipulative treatment (OMT) helps to effectively stabilize and regulate the autonomic nervous system
7 and the cranial nerves that are important in sucking and latching reflexes, which, in turn, facilitates the acquisition of skills necessary for oral feeding.
8, 9 Additionally, the neonatal cranium is strikingly malleable at this age, with the occipital bone comprising four parts and the sphenoid and temporal bones each comprising three parts. This increase in mobility results in increased susceptibility to developing intraosseous compressions when subjected to the forces of a normal labor and delivery process.
8 Traumatic deliveries that require interventions such as vacuum extraction, forceps, or maneuvers for shoulder dystocia can exacerbate or add additional strain to the forces of normal labor, which potentially increases the risk of these intraosseous compressions.
8 Previous studies
10, 11 have demonstrated that OMT can improve oral feeding skills and decrease risk for gastrointestinal dysfunction in premature neonates, leading to a mean reduction of six days in the hospital length of stay (LOS). A relationship between craniosacral strain patterns and the associated symptoms (e.g., poor latch, poor suck, vomiting, hyperactive peristalsis, tremor, hypertonicity, and irritability) has been shown to affect the nervous, circulatory, and respiratory systems, as evidenced by difficult or irregular respirations, excessive mucus, and marked cyanosis.
8 In this study, we hypothesized that term neonates treated with therapeutic hypothermia (TH) would be able to accelerate the transition to full oral feeds through the application of OMT, thus decreasing their overall LOS. We also hypothesized that these neonates would demonstrate improved somatic dysfunction patterns through decreased prevalence of nonphysiologic strain patterns within the craniosacral mechanism.
The primary outcome measure was a comparison of total hospital LOS in OMT managed neonates vs. historical controls. LOS was determined based on the date of birth (same as the date of admission) and date of discharge and was counted in full days.
Two secondary outcomes were investigated. One secondary outcome measure compared the number of days to full oral feeds between study groups and was assessed by analyzing the number of days a neonate required a nasogastric (NG) tube for feeding. This data was extracted from the intake and output documentation in the electronic health record (EHR) of each patient’s chart. Full oral feeding volume was defined as a 24-hour period in which the NG tube was not used; this time frame was selected to avoid confounding from smaller increments that could be attributable to nursing care or other factors not related to OMT.
The other secondary outcome measure was the identification of somatic dysfunction patterns through documented findings on the osteopathic structural examination (OSE) performed before and after OMT. A qualitative analysis of patterns specific to the craniosacral mechanism was completed using these data.
Patients were identified for enrollment by the principal investigator (A.C.), a neonatal neurologist who was actively involved in the clinical care of each of these patients. The PI was responsible for determining eligibility and obtaining informed consent during the period of TH, after confirmation of a normal brain MRI. TH was carried out per the current standard of care according to existing clinical guidelines and was not part of this investigation. OMT was provided after TH was complete at no charge to the patient, insurance, or any other third-party payer. OMT was performed by one of three selected providers (K.B., A.B., B.B.). Patients were not compensated for participation.
Two OMT sessions were performed between fourth and seventh day of life. For neonates who were hospitalized longer, OMT was continued twice per week on nonconsecutive days until the neonate was discharged. Each neonate was evaluated with an OSE before each treatment, which included examination of the head, cervical, thoracic, lumbar, sacral, pelvic, and rib cage regions for underlying somatic dysfunctions. Subjective comments were noted from parents or nursing staff. The OSE findings and subjective comments were then documented at each OMT encounter (Appendix). Specific OMT techniques were used at each provider’s discretion and included a combination of myofascial release, balanced ligamentous tension, balanced membranous tension, and osteopathic cranial manipulative medicine. Per current standard protocol, neonates ingested nothing by mouth throughout TH and started on NG tube feeds after completion of TH.
Baseline differences between the OMT and historical controls were compared using χ2 tests or Fisher exact tests, as appropriate, for categorical variables and t tests or their nonparametric equivalents for continuous variables. Nonparametric variables were presented as medians and interquartile ranges. The primary analysis compared LOS for neonates in the OMT group to the historical controls. In the secondary analysis, the number of days for the neonates to achieve full oral feeding was compared. Lastly, an additional secondary analysis compared the somatic dysfunction patterns within the craniosacral mechanism of neonates before and after receiving OMT.
A power calculation demonstrated that a two-group t test with p value of 0.05 would have 80% power to detect the difference between the OMT group mean of 7.3 days and the historical controls mean of 11.5 days. The difference in means was 4.2 days, assuming that the expected standard deviation is 4.4 when the sample sizes in the OMT group or the historical controls are 12 and 36, respectively.
This pilot study assessed the effect of OMT on hospital LOS and achievement of full oral feedings in term neonates following treatment with TH. The difference in LOS was statistically significant and showed a decrease of 2.5 days for neonates treated with OMT. However, statistical significance was lost in a sensitivity analysis that excluded intubated neonates from the historical control group. These neonates were excluded from this analysis due to the known association between intubation and delayed oral feeding and increased LOS. Although not statistically significant, the OMT group did have a one-day shorter LOS compared with nonintubated historical controls. The TH cohort demonstrated a decrease of one day within the context of an admission that lasted only nine to 12 days. As such, TH neonates had fewer opportunities to receive OMT; therefore, both a larger sample size and a prospectively recruited control group would be required to definitively assess the effect on LOS. Given the potential psychosocial and even financial benefits of shorter hospital LOS, the results of this pilot study suggest a randomized control trial is worth pursuing.
We analyzed two secondary outcomes: review of days to full oral feeding, measured as full days with NG tube use, and qualitative analysis of craniosacral strain patterns. Analysis of days to full oral feeds showed no statistically significant difference between OMT treated neonates and historical controls. The way this data was collected may have limited our ability to detect a significant difference. Historical control data was collected and documented in 24-hour increments of time, which made it difficult to appreciate improvement in the volume consumed orally from a single feeding to the next. Subjectively, both nurses and parents reported an improvement in oral feeding, and there was an objective improvement appreciated in the craniosacral strain patterns in treated neonates. Transient neuropathies can develop in the nerves responsible for effective latch, suck and swallowing reflexes, such as the vagus, glossopharyngeal or hypoglossal nerves that reside within foramina at the cranial base.
9 These neuropathies commonly are induced into the craniosacral mechanism from the trauma of the labor and delivery process resulting in nonphysiologic strain patterns.
8, 9
Qualitative analysis of strain patterns was exploratory and performed only on the small sample of the 12 patients who received OMT. Our findings were consistent with those described in previous osteopathic studies conducted in both neonatal patients and healthy adults.
10, 11, 13 One study
8 of 1,250 healthy neonates demonstrated a higher prevalence of physiologic craniosacral strain patterns following a normal and uncomplicated labor and delivery – specifically, torsions and sidebending rotation strains. This study
8 suggested that the strains that are nonphysiologic (compressions, vertical, lateral strains) would be more prevalent following trauma to the system due to disruptions specifically at the cranial base (i.e., following difficult birth). Our pretreatment OSE findings in the 12 neonates in our study support this pattern of increased prevalence of nonphysiologic strain patterns. This finding is consistent with the traumatic birth process these neonates experienced.
Traumatic births appear to increase the risk of developing nonphysiologic strain patterns, thus inducing transient neuropathies at the cranial base and creating dysfunctions in a neonate’s latch, suck, or swallowing reflexes. The anatomy of the neonatal cranium makes it particularly susceptible to the compressive forces of birth. Specifically, the occipital bone comprises four parts, and the temporal and sphenoid bones comprise three parts. This anatomic fact demonstrates why certain areas have an increased vulnerability for compression. Areas of vulnerability include the hypoglossal canal and the jugular foramen. The hypoglossal canal is formed between the basilar and condylar parts of the occipital bone, which houses the hypoglossal nerve. The jugular foramen is formed where the occiput and temporal bones articulate, which houses the glossopharyngeal, vagus, and accessory nerves. These nervous structures are directly responsible for the functions required for a neonate to feed successfully.
8, 9
Only a small percentage (one; 8.3%) of the neonates in the OMT group in this study still demonstrated evidence of nonphysiologic strain at the time of discharge, suggesting a beneficial effect from the OMT intervention. Future studies could investigate the differences in underlying pathophysiology that contribute to prolonged neuropathies and, in turn, difficulties or delays in progression to full oral feeds.
The main limitation of this study was the unintended confounding variables resulting from historically matched controls - specifically, that more than half of the control patients were intubated. While this significant issue was addressed through a sensitivity analysis, statistical significance was lost in part due to reduced sample size. Additionally, OMT was not performed in a manner that permitted nurses and parents to be blinded and may have contributed to overly positive subjective reports of feeding improvement. For this small, unfunded pilot study, there were insufficient resources for an optimal randomized control trial in which parents and nurses could be blinded to treatment status. Finally, there are limitations in the extrapolation of the osteopathic data because multiple osteopathic physicians diagnosed and treated the neonates, which may have created interobserver variation in the reported findings. However, multiple providers found similar craniosacral strain patterns within each patient, suggesting that these observations may in fact be a true representation, albeit one limited by a small sample size.
The current study highlighted the second and third osteopathic tenets by demonstrating the body’s inherent ability for self-healing and self-regulating and the importance of the interconnectedness of structure and function, respectively. Our OSE findings suggest that increased stress and strain introduced into the craniosacral mechanism during birth in neonates subsequently treated with TH increases the prevalence of nonphysiologic strain patterns and reduction in the primary respiratory mechanism, specifically at the cranial base. This intraosseous compression alters the normal physiologic strain pattern, thus transiently inhibiting the nerves responsible for effective latch, suck, and swallowing reflexes. OMT directed at decompression of the cranial base can effectively target these strains and improved primary respiratory motion and restoration of function.
This form was used by the authors to document findings and subjective comments at each OMT encounter and is reprinted without edits.
Infant Subject Number: ___________________
Date of Treatment: ______________________
Day of Life of Infant: _____________________
Time Treatment Started: __________________
Finished: _________________
Name of Treating Physician: _________________________
Admission Date: ___________ Discharge Date: __________
Osteopathic structural exam: Craniosacral strain patterns: -
Patient Position:
-
Head:
-
Cervical:
-
Thoracic:
-
Lumbar:
-
Sacrum:
-
Pelvis:
-
Rib cage:
- ❑ Occipital
-
❑ Right compression
-
❑ Left compression
-
❑ Bilateral compression
-
❑ Free
- ❑ Temporal
-
❑ Right restriction
-
❑ Left restriction
-
❑ Bilateral restriction
-
❑ Free
- ❑ Sphenobasilar symphysis (SBS)
Adverse Events: -
❑ No
-
❑ Yes (please describe):
(Simple check box for the qualitative collected data on the somatic dysfunction patterns seen in the craniosacral mechanism that would have direct effects on the neurologic, cardiac and respiratory function of these infants.)
Osteopathic structural exam: Craniosacral strain patterns: -
Patient Position:
-
Head:
-
Cervical:
-
Thoracic:
-
Lumbar:
-
Sacrum:
-
Pelvis:
-
Rib cage:
- ❑ Occipital
-
❑ Right compression
-
❑ Left compression
-
❑ Bilateral compression
-
❑ Free
- ❑ Temporal
-
❑ Right restriction
-
❑ Left restriction
-
❑ Bilateral restriction
-
❑ Free
- ❑ Sphenobasilar symphysis (SBS)
Adverse Events: -
❑ No
-
❑ Yes (please describe):
(Simple check box for the qualitative collected data on the somatic dysfunction patterns seen in the craniosacral mechanism that would have direct effects on the neurologic, cardiac and respiratory function of these infants.)
×