Child's Nervous System

, Volume 21, Issue 11, pp 986–990

Head circumference and chronic positive pressure ventilation in children: a pilot study

Authors

    • Pediatric Pulmonary RehabilitationFranciscan Hospital for Children
    • Division of Respiratory Disease, Children’s Hospital BostonHarvard Medical School
  • Ning Tat Hamilton Hui
    • United Christian Hospital
  • Helene M. Dumas
    • Research Center for Children with Special Health Care NeedsFranciscan Hospital for Children
  • Stephen M. Haley
    • Research Center for Children with Special Health Care NeedsFranciscan Hospital for Children
    • Health and Disability Research InstituteBoston University
  • Linda Specht
    • Medical and Rehabilitative ServicesFranciscan Hospital for Children
  • Andrew A. Colin
    • Division of Respiratory Disease, Children’s Hospital BostonHarvard Medical School
Original Paper

DOI: 10.1007/s00381-004-1126-9

Cite this article as:
Kharasch, V.S., Hui, N.T.H., Dumas, H.M. et al. Childs Nerv Syst (2005) 21: 986. doi:10.1007/s00381-004-1126-9

Abstract

Methods

We reviewed the cases of 11 children <12 years of age with complex medical conditions and respiratory failure requiring chronic positive pressure ventilation (CPPV). We conducted a within-group comparison of average head circumference (HC) percentiles with each child’s age-expected 50th percentile value and a between-groups comparison with children with no history of ventilation. We examined the relationship between HC, peak levels of positive inspiratory pressure (PIP), and length of time on CPPV.

Results

We found that children on CPPV had an average HC value at the 71st percentile, significantly greater (p=0.009) than the age-expected 50th percentile; that children on CPPV had a greater (p=0.003) average HC percentile compared with children with complex medical conditions not on CPPV; and that peak levels of PIP had a moderately strong correlation to HC (r=0.689, p=0.019).

Conclusion

We conclude that children on CPPV have larger than expected HC and HC appears related to the peak level of PIP. Research to further investigate the relationship between HC and CPPV in children appears to be warranted.

Keywords

ChildrenMechanical ventilationPositive pressureCraniumMacrocephaly

Introduction

While macrocephaly or an enlarged occipitofrontal cranial circumference has not been reported in children on chronic positive pressure ventilation (CPPV), DeLemos and Tomasovic [3] have suggested that continuous levels of positive airways pressure may increase cranial venous blood volume and pressure by obstructing venous blood return to the right heart. This study was triggered by our observation that children with complex medical conditions and attendant respiratory failure (without a systemic or metabolic disorder) who were treated in a chronic care pulmonary rehabilitation program had large head circumferences. We therefore embarked on a study to systematically address the question whether mechanical ventilation is associated with an increase in head size. The purpose of this descriptive pilot study was to determine:
  1. 1.

    If the average head circumference (HC) percentile of children who are receiving CPPV is different from their age-expected 50th percentile values

     
  2. 2.

    If an age-matched comparison group of children with complex medical conditions but not on CPPV shows a difference in average HC from children on CPPV

     
  3. 3.

    Whether a relationship between HC, peak positive inspiratory pressure (PIP), and duration of CPPV could be found

     

Materials and methods

Patients

A convenience sample, a nonprobability, non-randomized sampling procedure, was used to select the most available subjects for this study. Every child less than 12 years of age seen in consultation at the Franciscan Hospital for Children (FHC), Boston, MA who was dependent on CPPV via tracheostomy for at least 1 month were enrolled sequentially over a 2-month period from November to December 2002. We included children who were either admitted to or followed as outpatients at a pediatric rehabilitation facility by a pediatric pulmonary group and who met the above criteria.

All children on CPPV (n=11) with complex medical conditions including congenital anomalies (n=5), prematurity (25–34 weeks’ gestation) (n=3), or acquired neurological injuries (traumatic brain or spinal cord injuries; n=3) were included. None of the children had an underlying systemic or metabolic disorder that could account for a large head size. There were 6 boys and 5 girls in the study group. The mean age of the study group was 44 months (SD 46, range 2–135 months). The children had been ventilated for an average of 27 months (SD 42, range 2–132 months). Ventilation pressures were conventional with PIPs ranging from 15 to 20 cm H2O (mean 16.7 cm H2O, SD 1.9; Table 1).
Table 1

Characteristics of the study sample. PIP positive inspiratory pressure

Case number

Sex

Age (months)

PIP at time of study

Diagnosis

1

Girl

2

16

Moebius syndrome, central hypoventilation

2

Boy

3

20

Premature birth (25 weeks), bilateral grade 3 intraventricular hemorrhage, chronic lung disease

3

Boy

8

16

Premature birth (28 weeks), severe chronic lung disease left grade 3 intraventricular hemorrhage

4

Boy

9

17

Arthrogryposis, right diaphragmatic dyskinesia

5

Girl

9

16

Congenital myopathy

6

Girl

15

20

Premature birth (34 weeks), congenital myopathy, chronic lung disease

7

Boy

48

15

Cervical cord injury, right diaphragmatic paralysis

8

Boy

84

15

Teratology of Fallot, tracheomalacia, bronchomalacia

9

Girl

84

18

C2 cervical cord injury, quadriplegia

10

Girl

88

16

Mitochondrial disorder, mental retardation, seizures, scoliosis

11

Boy

135

15

Birth trauma with spinal cord injury/quadraplegia

A comparison group (n=9) of children from the same practice, but not dependent on CPPV was also studied. This comparison group included 7 boys and 2 girls with a mean age of 18 months (SD 13, range 4–36 months). These children also had complex medical problems and included 4 children with tracheostomy tubes. None of the children had microcephaly and children with hydrocephalus were excluded from both groups (Table 2).
Table 2

Characteristics of the age-matched controls

Case number

Sex

Age at time of study (months)

Diagnosis

1

Boy

4

Spastic quadriplegia, seizure disorder, gastroesophageal reflux

2

Girl

4

Nocturnal apnea, hypotonia, tracheostomy

3

Boy

5

Right facial vascular malformation, tracheostomy

4

Boy

9

Developmental delay, atrial septal defect

5

Boy

18

Seizure disorder, recurrent aspiration pneumonia

6

Boy

24

Developmental delay, asthma

7

Boy

26

Chromosomal abnormalities, tracheostomy

8

Girl

35

Pulmonary atresia post Glenn Shunt, asthma

9

Boy

36

Wolff–Hirschorn syndrome, asthma, tracheostomy

Methods

Medical records were reviewed for diagnosis, age, HC, CPPV, duration of CPPV, and PIP level. Although HC was recorded for pre- and post-CPPV levels, we used only the post-CPPV levels in these analyses, as all other data were cross-sectional. The determination of the HC percentile was based on a standard pediatric head circumference chart [9] and the percentile was recorded for each child.

To compare the mean head circumference of the study population with published reference data, we used a one-sample t test to compare the mean HC percentile of the study group (children with CPPV, n=11) with the 50th percentile. To compare our study subjects with the comparison group, we conducted two analyses. First, we compared the average HC percentile of all of the children with CPPV to the percentiles of the 9 children in the comparison group using an independent t test. Since 4 of the children with CPPV were considerably older than any of the children in the comparison group, we also compared the HC percentiles of the 7 youngest children with CPPV with the HC percentiles of the comparison group. This second analysis allowed us to assess only children from the two groups who were under the age of 4 years. Lastly, we used bivariate Pearson product moment correlations to determine the relationship between HC, PIP, and duration of CPPV. A p value of less than 0.05 was considered significant. Data were analyzed using the Statistical Programs for the Social Sciences (SSPS, Chicago, IL, USA).

Observations and results

All children with CPPV had a head circumference at or above the 50th HC percentile (range 50th HC percentile to the 97th HC percentile). The mean HC percentile of children with CPPV was the 71st percentile for HC and was significantly greater than the expected 50th percentile value (p<0.009, CI=6.5–35.4). When comparison with the group not on CPPV was made, significant differences in HC percentile were also noted both for the whole group of children with CPPV (n=11; HC=71st percentile, p=0.003) and for the CPPV group under 4 years of age (n=7, HC=67th percentile, p=0.02) (Fig. 1). Four children had head circumference data available before CPPV was initiated. In case 2, HC increased from the 10th to the 50th percentile; in case 3, HC increased from the <3rd to the 50th percentile; in case 4, HC increased from the 50th to the >97th percentile; and in case 5, HC increased from the 10th to the 50th percentile, further demonstrating increased head circumference after initiation of CPPV. Table 3 depicts the moderately strong relationship between HC, PIP, and duration of ventilation.
Fig. 1

Average head circumference of children on chronic positive pressure ventilation (CPPV) and comparison group of children not on CPPV

Table 3

Pearson product moment correlation coefficient matrix. CPPV chronic positive pressure ventilation

 

 

Head circumference

PIP

Time on CPPV

Head circumference

Pearson correlation, sig (two-tailed)

1.00

0.689

−0.077

 

0.019*

0.822

N

11

11

11

PIP

Pearson correlation, sig (two-tailed)

0.689

1.00

−0.149

0.019*

 

0.662

N

11

11

11

Time on CPPV

Pearson correlation, sig (two-tailed)

−0.077

−0.149

1.00

0.822

0.662

 

N

11

11

11

*Correlation significant at p<0.05

Discussion

We found that children with complex medical conditions on CPPV had head circumferences that are larger than normal based on standard head circumference charts and also larger in comparison to children with complex medical conditions who do not require CPPV. The intracranial volume and pressure effect of CPPV in children with chronic respiratory failure has not been previously described. In 1987, Schreiner et al. [11] defined chronic respiratory failure as a condition requiring mechanical ventilation greater than 28 days. Children with chronic respiratory failure and long-term positive pressure ventilation in this study had an average of 21% greater HC (>2 SD above the norm) than children in the normal population, supporting our clinical observation.

There is no known relationship between HC and mechanical ventilation and therefore no mechanism to explain it is to be found. Blood flow from any cerebral vein to the right atrium is determined by the pressure gradient between two points. Right atrial pressure is influenced by atrial volume and chamber compliance, and the continually subatmospheric intrathoracic pressure (ITP). Elevation of ITP narrows the cerebral-cardiac gradient and decreases venous return to the right heart. This was shown to lead to a substantial elevation of ICP in adults similar to the effect of superior vena cava occlusion [2, 12]. In children, compliant skull bones may result in intracranial volume expansion as the means of dissipation of the increased pressure resulting in macrocephaly. This effect may be similar to hydrocephalus in which an enlarged head circumference is the manifestation of increased pressure in children whose fontanels are still open.

In newborn lambs, Ratjen found that rapid thoracoabdominal compressions (RTC), a technique for physiological measurements in infants, resulted in proportional increases in ICP [10]. RTC led to linear increases in transpulmonary pressure and increased ICP parallel to thoracoabdominal pressure with ICP returning to baseline levels with the release of the chest compression [10]. More direct evidence of the relationship was demonstrated by Milligan [8]. Transient increases in HC in mechanically ventilated newborns were demonstrated by a proportional relationship between the skull volume expansion and the applied airway pressure. Cranial volume increased appreciably when more than 20% of the applied airway pressure, ranging from 13 to 31 cm H2O, was transmitted to the pleural space, or when the absolute pleural pressure was maintained at 4 cm H2O or higher than atmospheric pressure.

The effects of positive end-expiratory pressure (PEEP) on ICP is well documented in the literature [6, 13]. PEEP at 10 and 15 cm H2O produced significant increases in ICP while PEEP at 5 cm H2O had no such effects. It has been observed that while widespread use of PEEP has improved outcomes in hyaline membrane disease, many infants develop neurological sequelae from intracranial hemorrhage [7, 14]. Increased intrathoracic pressures may cause engorgement of vertebral venous networks and displace CSF from the spinal canal to the cranium causing increased ICP and hydrocephalus [4]. Other mechanisms that may support a relationship between positive pressure ventilation and increased cranial volume include impaired control of the autoregulation that allows central pressure changes to be transmitted to the cerebral vasculature [5, 7].

Interestingly, it appears that noncompliant lungs may be protective against the increase in ICP. Milligan proposed that infants with the most severe lung disease are at least risk from transmission of airway pressure to the cerebral circulation [8]. Non-compliant lungs had poor transmission of applied pressure and thus minimal change in cranial volume with each application of positive pressure. In normally compliant lungs, changes in airway pressure were associated with appreciable changes in cranial volume [8]. Similarly, decreased intracranial pressure effects of positive pressure were noted in oleic acid-injured cat lungs where compliance was poor [1].

Compared with children with complex medical conditions who were not on CPPV, patients who were on CPPV showed larger HC percentiles. It is important to note that our study population included 4 patients with decreased lung compliance secondary to chronic parenchymal disease and 7 patients with normally compliant lungs requiring positive pressure ventilation for bellows failure secondary to muscle weakness. Given the previously cited observation, it appears that a larger number of our patients were at higher risk of developing intracranial effects of positive intrathoracic pressure. Data collected for only 4 of our 11 cases further supports our observations. For these four cases we documented HC prior to the institution of positive pressure ventilation and found a substantial increase in HC after initiation of mechanical ventilation, again suggesting the intracranial effect of PIP.

Conclusions

Despite our small sample size, there was a strong relationship between level of PIP and HC percentile. In this initial study, we were not able to determine whether the level of PIP directly influences cranial volumes and pressures. These data do suggest, however, that this relationship needs further study. The clinical consequences of increased HC secondary to venous engorgement on the developing brain are unknown. Longitudinal, ideally also controlled studies are needed to further explore this association. Ultimately, the question will be whether clinical practice of managing ventilator settings, such as lowering PIP, may have a neurological benefit for children on mechanical ventilation.

Future research in this area should include studies of larger groups of children over longer periods. At this time, however, we would recommend that for all children on prolonged ventilation, HC percentiles before and after the initiation of mechanical ventilation should be followed, and also, that PIP be carefully monitored as well as minute ventilation. It is possible that sustaining the same minute ventilation by increasing the mandatory respiratory rate and decreasing peak pressure may be effective in preventing the putative effect on head circumference that we report here.

Copyright information

© Springer-Verlag 2005