Effect of Wheelchair Headrest Use During Rear Impact on Pediatric Head and Neck Injury Risk Outcomes

Susan I. Fuhrman, BSME1 , Patricia Karg MSBME1 , Gina Bertocci, PhD2
1 University of Pittsburgh, Pittsburgh, PA 2 University of Louisville, Louisville, KY


Wheelchair transportation safety research has primarily addressed frontal impact consequences. This study is a first, examining effects of headrest use on pediatric injury measures under rear impact conditions. Two sets of three identical WC19 transit-option manual pediatric wheelchairs were tested: three sled tests were conducted without a headrest and three with a headrest. All tests used a 26 km/h, 10 g rear impact test pulse, seated Hybrid III 6-year old anthropomorphic test device (ATD), and 4-point strap-type surrogate wheelchair tiedowns and 3-point occupant restraints (WTORS). Results suggest a lower risk of head and neck injury with wheelchair headrest use.


wheelchair transportation safety, rear impact, pediatric injury risk, neck injury criteria, head injury criteria


This is a photo of a side view of a 6-year old Hybrid three ATD seated in a manual pediatric wheelchair. The wheelchair is secured to a test sled with four strap-type tiedowns. The ATD is wearing a three-point occupant restraint. High contract markers are placed on the head, shoulder and knee. Figure 1: Wheelchair set-up without a headrest (Click image for larger view)

Many wheelchair users are unable to transfer to a motor vehicle seat and remain seated in their wheelchairs during transportation. A goal is to make transportation in a motor vehicle as safe for those who travel in wheelchairs as it is for those who travel in original manufacturer installed vehicle seats. Historically, most wheelchair transportation safety research has focused on frontal impact events. Yet, while most fatalities from motor vehicle accidents do occur in frontal impact, rear impact accounts for the greatest number of occupant related injuries [1, 2]. In response to this concern, vehicle manufacturers have focused research efforts on developing effective head restraints [3, 4]. Our research [5] has indicated that wheelchair headrests are prescribed for over 60% of all wheelchair users, and for 80% of all pediatric wheelchair users. However, there has been no previous effort to investigate the effects of wheelchair headrest use for pediatric wheelchair-users in rear impact.

This side view photo shows a Hybrid three, six year old ATD seated in a manual pediatric wheelchair. The wheelchair is equipped with a headrest. The wheelchair is secured to a test sled with four strap-type tiedowns. The ATD is wearing a three-point occupant restraint. There is whit arrow on the photo pointing to the joint where the anterior-posterior headrest stem meets the vertical headrest stem.Figure 2: Wheelchair set-up with headrest. Arrow indicates pin location. (Click image for larger view)

Several different measures have been used in an effort to predict likelihood of human injury based on measurable criteria. Both physical and mathematical models of the head or head and neck, have been used to establish measurable parameters to assess injury risk [6, 7]. Maximum head acceleration, head injury criteria (HIC), neck injury criteria (Nij), rotational head velocity, rotational head acceleration are all measures that have been used in an effort to predict injury likelihood [8].

The purpose of this investigational baseline study was to establish quantitatively the potential benefit or harm of using a postural headrest for a wheelchair-seated pediatric occupant during rear impact. This was evaluated by comparing outcomes to a variety of injury risk measures.


Six identical Sunrise Medical Quickie Zippie transit-option (ANSI/RESNA WC19 [9]) pediatric manual wheelchairs (17.9 kg) were tested: three without headrests, three with identical Sunrise Medical single-pad headrests (Table 1).

Table 1: Sled test matrix
Sled Test Wheelchair configuration
1 No headrest
2 No headrest
3 No headrest
4 Sunrise Medical single pad headrest
5 Sunrise Medical single pad headrest
6 Sunrise Medical single pad headrest
This side view photo shows the ATD seated in the wheelchair during the rear impact collision. The wheelchair has rotated rearward, but remains secured by the tiedowns. The wheelchair is not equipped with a headrest. The ATD’s head as rotated all the way rearward such that the back of the head is in contact with the top of the seatback.Figure 3: Maximum neck extension for Test 2 with no headrest (Click image for larger view)

All wheelchairs were tested with a seated Hybrid III 6-year old ATD using a 26 km/h, 10g rear impact crash pulse. Wheelchairs were secured using four-point strap-type surrogate wheelchair tiedowns; the ATD was restrained with a three-point occupant restraint system [9]. At the time of testing, the crash pulse conformed with the then proposed ISO test pulse for use in development of a draft voluntary industry standard for wheelchair performance evaluation in rear impact [10]. All wheelchairs were equipped with matched components and identically configured. A pin, inserted in the Sunrise Medical headrest stem joint, prevented headrest anterior-posterior slippage during rear impact. (Figure 2))

This side view photo shows the wheelchair seated ATD during the rear collision. The wheelchair has rotated rearward although it remains secured by the tiedowns. This wheelchair contains a headrest. The ATD’s head remains in an upright position during this test.Figure 4: Maximum neck extension for Test 5 with a headrest (Click image for larger view)

ATD instrumentation included a triaxial accelerometer positioned at the head CG to measure head accelerations; an upper neck load cell measured neck loads and moments. High-contrast markers, placed on the ATD head (2), shoulder and knee, indicated position throughout the test. High-speed video cameras (1000 frames/sec) recorded the test. Transducer data were recorded every 0.1 ms and filtered according to SAE J211.

Kinematic data from the tests described ATD response to rear impact. A Matlab program was used to acquire and track high-contrast marker location coordinates [11]. Kinematic data from video images were used to calculate rotational head velocity, and peak rotational head acceleration, which were compared to diffuse axonal injury (DAI) criterion, introduced by Margulies [12]. DAI is associated with maximum change in angular velocity and peak angular accelerations.

Transducer output for head acceleration, neck loads and neck moments, were used to calculate head injury criteria (HIC) values and neck injury criteria (Nij) values. HIC values were calculated using Equation 1 [13]. HIC15, HIC36 and HICun were calculated using corresponding millisecond time intervals.

HIC equals the quantity of one over the quantity t-two minus t-one times the integral from t-one to t-two of acceleration d-t, closed quantity to the two and a half power. This is multiplied by the quantity t-two minus t-one.

Nij establishes critical limits for neck axial loading and bending moments. Nij was defined by [14]:

Nij equals the quantity of the axial load divided by the critical intercept load value, closed quantity. This is added to the quantity of the bending moment divided by the critical moment intercept, closed quantity.

Fz = axial load

Fint = critical intercept load value used for normalization

My = bending moment

Mint = critical moment intercept value used for normalization


The horizontal axis shows time from zero to three hundred milliseconds. The vertical axis shows acceleration from minus fourteen to two g’s. Six curves are shown. All have very close to the same profile. The sled decelerates from zero to minus ten, holds that level for approximately seventy milliseconds, then returns to zero.Graph 1: Sled acceleration profile (Click image for larger view)

In all tests, sled acceleration plateau average levels were between -9.6 g and -10.0 g, sled acceleration peaks between -11.0 g and -11.6 g, and sled change in velocity (delta-V) between 25 km/h and 25.6 km/h. Graph 1 shows sled test acceleration profiles and sled acceleration pulse reproducibility.

The horizontal axis is time form zero to three hundred milliseconds. The vertical axis is head acceleration from zero to forty g’s. There are six curves. Three in blue represent test with headrests. They have a single peak of twenty-five to thirty-five g’s. Three in orange represent the tests without headrests. They show two peaks. The first peak is thirty-two to thirty-six g’s; the second peak is twenty-seven to thirty-seven g’s.Graph 2: Translational head acceleration (Click image for larger view)

The Zippie wheelchairs remained structurally intact and the ATD maintained an upright posture throughout all rear impact tests. Initial evaluation of sled test video images (Figures 3 and 4) suggests a neck extension reduction with headrest use.

Head acceleration results (Graph 2) show higher peak head accelerations and longer duration of these accelerations during sled tests 1-3 lacking headrests when compared against tests 4-6 with headrests. Tests 1-3 show secondary acceleration peaks occurring during head-seatback contact. When compared to a proposed head acceleration protection reference value (PRV) of 80 g, [6] a low probability of associated head injury is predicted.

The graph depicts a bar chart with the HIC results from six sled tests. In all cases the tests that did not include headrests have higher values that those that did have headrests. Non headrest results for HIC-unlimited range from one hundred and thirty-five to one hundred and seventy, with headrests results range from eighty to eighty-five. For HIC-thirty-six, no-headrest values range from one hundred and twenty-four to one hundred and thirty, with headrest values are eighty to one hundred and ten. For HIC-fifteen, non-headrest results range from seventy-three to ninety-five, headrest tests range from fifty to sixty-eight. Graph 3: Head injury criteria values (Click image for larger view)

Peak head acceleration and associated duration is used in computing HIC. HIV value results for 15 ms, 36 ms, and the complete test are indicated in Graph 3. The HIC PRV for the 6-year old ATD are: HICun is 1000, HIC36 is 1000, HIC15 is 700; these levels are associated with a 23% chance of maximum abbreviated injury scale (MAIS) injury level > 3 [6]. Tests 4-6 conducted with headrests yielded average HIC values 34% lower than tests 1-3 conducted without headrests, although all HIC values were below PRVs.

The horizontal axis is normalized neck bending moment. The vertical axis is the normalized axial load. The graph is divided into four quadrants. The upper left is tension-extension. The upper right is tension-flexion. The lower left is compression-extension. The lower right is compression-flexion. There is a reference curve for Nij equal to one. Six plots are shown. The orange curves represent the tests without headrest. The orange curves extend past the reference line in the tension-extension quadrant. The blue curves, representing tests with headrests, are al well within the reference curve.Graph 4: Neck injury criteria (Click image for larger view)

Graphs 4 and 5 highlight improved neck response with headrest use. Nij > 1 is associated with 22% serious injury probability (abbreviate injury scale > 3) [15, 16]. The mechanism of greatest concern during rear impact is the tension-extension portion of Nij. Our findings show that Nij results indicate ATD neck response in rear impact without a headrest exceeds PRV Nij=1. This occurred when the ATD neck reached maximum extension. Tests with headrests had 70% average reduction in Ntension-extension over tests without headrests.

This bar graph shows the peak values for each of the Nij components: compression-extension, compression-flexion, tension-extension, and tension-flexion. A horizontal reference value is shown at Nij equal to one. Tests without headrest are shown in orange, tests with headrest are shown in blue. Only tests without headrest exceed the reference value. The reference value is exceeded in the tension-extension component, with the maximum Nij recorded at Nij equal to one point five.Graph 5: Nij peak values (Click image for larger view)

Both peak rotational head velocity change (Graph 6) and peak rotational acceleration (Graph 7) show that headrests reduce the rotational effects on the head. The rotational head response PRVs, not firmly established but cited by Klinich [6] based on Ommaya’s work [17], are associated with AIS > 3. In both graphs, values for tests without headrests exceed PRVs; those with headrests are close to PRVs. Tests 4-6 with headrests averaged 51% reduction in change in rotational head velocity, and 36% reduction in rotational head acceleration.

This is a bar graph with six sled tests listed on the horizontal axis. Peak change in rotational head velocity is shown on the vertical axis. A horizontal line shows the protection reference value equal of thirty-three radians per second. Tests without headrests are shown in orange. They all exceed the protection reference value and range from seventy to seventy-eight radians per second. Tests with headrests are shown in blue and fall much closer to the protection reference value, ranging from twenty-eight to forty-three.Graph 6: Peak change in rotational head velocity (Click image for larger view)

As noted in the DAI graph (Graph 8), all tests were below the injury threshold, with headrest-containing tests falling lowest.


This exploratory study of pediatric ATD response during rear impact demonstrates that use of a slightly modified commercially available wheelchair headrest designed solely for postural support has the potential to provide improved head and neck protection during rear impact. Each injury outcome measured, showed peak value reductions for tests with headrests; HIC, Nij, rotational head velocity, and rotational head acceleration all showed reductions in excess of 34%, and in some cases as high as 70%. This is a significant finding of this study since three-fourths of pediatric wheelchairs are prescribed with headrests and some school systems require wheelchair users to have a headrest for transportation purposes [5, 18].

The bar graph list the six sled tests on the horizontal axis. The vertical axis shows the peak rotational head acceleration in radians per second squared. The protection reference value of twenty-one hundred radians per second squared is shown with a horizontal line on the graph. The three tests without a headrest all have values exceeding the protection reference value and range from twenty-four hundred to twenty-seven hundred. The tests that included headrests are shown in blue and all fall below the protection reference value, ranging from fifteen hundred and fifty to seventeen hundred and fifty.Graph 7: Peak rotational head acceleration (Click image for larger view)

It is important to note that there are some limitations associated with this study. Use of the ATD has limitations intrinsic with ATD biofidelity, especially with respect to ATD neck response. The Hybrid III 6-year old ATD neck has come under scrutiny for its response characteristics [19]. Many of the injury criteria PRVs used in this study were initially developed for the 50th percentile male and scaled for smaller ATDs. Many scaling parameters were based on limited testing. Furthermore, DAI injury criteria were derived from primate testing.


This graph plots peak rotational head acceleration against peak change in rotational velocity. Velocity is on the horizontal axis; acceleration is on the vertical. The are two sets of reference curves on the graph. The higher curve is the infant DAI threshold, the lower curve is the adult DAI threshold. The values for all tests fell below both thresholds.  The tests that included headrests are shown in blue and are lower than the tests without headrests shown in orange.  The lowest point on the adult threshold is approximately eight thousand radians per second squared at one hundred radians per second. The tests without a headrest had values of approximately twenty-five hundred at sevety-five radians per second. Tests with headrests were at two thousand radians per second squared at forty radians per second.Graph 8: DAI injury risk (Click image for larger view)

This study was funded by the National Institute on Disability and Rehabilitation Research (NIDRR) and the Rehabilitation Engineering Research Center (RERC) on Wheelchair Transportation, grant # H133E010302. The opinions expressed herein are those of the authors and are not necessarily reflections of NIDRR opinions. Thanks to Miriam Manary for her input to sled impact testing.

Susan I. Fuhrman
Department of Rehabilitation Science and Technology
2310 Jane Street, Suite 1300
Pittsburgh, PA 15203
(412) 586-6920, (412) 586-6910 (fax),


  1. Japan Traffic Safety Association, Traffic Greenpaper. 1997. p. 115-118.
  2. NHTSA, Preliminary Assessment of NASS CDS Data Related to Rearward Seat Collapse and Occupant Injury. 1997, NHTSA: Washington, DC.
  3. Saab. Features and Specifications: Safety & Security. 2006 [cited April 5, 2006]; Available from: http://www2.saabusa.com/95s/features.asp?start=home.
  4. Johansson, S. Volvo for life: safety firsts. 2006 [cited April 5, 2006]; Available from: http://www.volvocars.us/_Tier2/WhyVolvo/Safety/SafetyFirsts.htm.
  5. Fuhrman, S.I., M.E. Buning, and P. Karg. Prescription patterns of secondary postural support devices and concerns related to their used during vehicle transportation. in RESNA 2005. 2005. Atlanta, Georgia: RESNA.
  6. Klinich, K.D., et al., Techniques for developing child dummy protection reference values, Child Injury Protection Team, Editor. 1996, General Motors.
  7. McLean, A.J. and R.W.G. Anderson, Head Injury, ed. P. Reilly and R. Bullock. 1997, London: Chaoman & Hall. 25 - 37.
  8. Nahum, A. and J. Melvin, Accidental injury-biomechanics and prevention. 1993, NY: Springer-Verlag.
  9. ANSI/RESNA, ANSI/RESNA WC-19: Wheelchairs Used as Seats in Motor Vehicles. 2000, American National Standards Institute (ANSI)/Rehabilitation Engineering Society of North America (RESNA).
  10. Flannagan, C. and M. Manary, ISO proposed rear impact standards and crash pulse severities, S.I. Fuhrman. November 14, 2005: Ann Arbor.
  11. Fuhrman, M.G., trackTargetsManually.m. 2005: Pittsburgh.
  12. Margulies, S.S. and L.E. Thibault, A proposed tolerance criterion for diffuse axonal injury in man. Journal of Biomechanics, 1992. 25(8): p. 917-23.
  13. McHenry, B.G., Head Injury Criterion and the ATB. ATB Users' Group, 2004.
  14. Kleinberger, M., et al., Development of improved injury criteria for the assessment of advanced automotive restraint systems. 1998, National Highway Traffic Safety Administration: Washington D.C. p. 114.
  15. NHTSA, FMVSS 208, Occupant Crash Protection. Vol. 49 CFR 571.208. 1993.
  16. Eppinger, R., et al., Development of Improved Injury Criteria for the Advanced Automotive Restraint Systems - II. 1999, National Highway Traffic Safety Administration: Washington D.C.
  17. Ommaya, A.K., The biomechanics of head injury: experimental aspects, in The biomechanics of trauma, A.M. Nahum and J. Melvin, Editors. 1985, Prentice-Hall.
  18. Smith, D., Discussion on pediatric headrests, S.I. Fuhrman. November 10, 2005: Pittsburgh.
  19. Sherwood, C.P., et al., Prediction of cervical spine injury risk for the 6-year-old child in frontal crashes. Traffic Injury Prevention, 2003. 4(3): p. 206-13.


  • Source Ordered
  • No Tables
  • Very Compatible


Disney produced a television show in the mid 1990s called Gargoyles. It's a great show and I'm a big fan. A few years ago Disney started to release the show on DVD. The last release was of season 2, volume 1. That was two years ago. Volume 2 has not been released. Why? Poor sales. So if you should find yourself wanting to support my work, instead I ask you pick up a copy of season 2, volume 1. It's a great show and you might find yourself enjoying it.