J Neurosurg Pediatrics 3:173–180,

3:000–000, 2009

Navigating the gray zone: a guideline for surgical decision

making in obstetrical brachial plexus injuries
Clinical article

James R. Bain, M.D., M.Sc., F.R.C.S.C.,1,2 Carol DeMatteo, M.Sc., Dip. P.&O.T.,2–4
Deborah Gjertsen, B.H.Sc. O.T., 2 and Robert D. Hollenberg, M.D., F.R.C.S.C.2,5
Divisions of 1Plastic Surgery and 5Neurosurgery, Department of Surgery, 3School of Rehabilitation Science,
and 4CanChild Centre for Childhood Disability Research, McMaster University; and 2McMaster Children’s
Hospital, Hamilton, Ontario, Canada

Object. In the literature, the best recommendations are imprecise as to the timing and selection of infants with
obstetrical brachial plexus injury (OBPI) for surgical intervention. There is a gray zone (GZ) in which the decision as
to the benefits and risks of surgery versus no surgery is not clear. The authors propose to describe this category, and
they have developed a guideline to assist surgical decision-making within this GZ.
Methods. The authors first performed a critical review of the medical literature to determine the existence of a
GZ in other clinical publications. In those reports, 47–89% of infants with OBPI fell within such a GZ. Complete
recovery in those reported patients ranged from 9 to 59%. Using a prospective inception cohort design, all infants
referred to the OBPI Clinic at McMaster Children’s Hospital were systematically evaluated up to 3 years of age. The
Active Movement Scale scores were compared for surgical and nonsurgical groups of infants in the GZ to identify
any important trends that would guide surgical decision-making.
Results. In the authors’ population of infants with OBPI, 81% fell within the GZ, of whom 44% achieved com-
plete recovery. Mean scores differed significantly between surgery and no surgery groups in terms of total Active
Movement Scale score and shoulder abduction and flexion at 6 months. Elbow flexion and external rotation differed
at 3 months.
Conclusions. There is compelling evidence that there is a group of infants with OBPI in whom the assessment
of the risk/benefit ratio for surgical versus nonsurgical treatment is not evident. These infants reside within what the
authors have called the GZ. Based on their results, a guideline was derived to assist clinicians working with infants
with OBPI to navigate the GZ. (DOI: 10.3171/2008.12.PEDS0885)

Key Words      •      obstetrical brachial plexus injury      •      outcome      •     

prospective inception cohort      •      recovery pattern      •      surgery

T
he peripheral nerve surgeon dealing with an infant
with an OBPI is always faced with the diagnos-
tic dilemma as to whether the injury includes sig-
nificant components of axonotmesis and/or neurotmesis
(which require timely surgical attention to maximize the
possibility of regeneration) versus neurapraxis where
one expects to see spontaneous full functional recovery.
Obviously the surgeon wants to avoid surgery on the
neurapraxic injuries and operate as soon as possible on
the limb destined for hypoplasia, poor motor/sensory re-
covery, and eventual functional disability.
Although there is no Level I evidence for any of the
Abbreviations used in this paper: AMS = Active Movement
Scale; GZ = gray zone; OBPI = obstetrical brachial plexus injury.
surgical or medical care given to these infants, large ret-
rospective data sets,5,6,8,12,13 a few prospective studies,20,21
and systematic reviews of the literature15,19 provide very
useful guidelines. There are some clear areas of consen-
sus in this body of literature. Infants in Narakas17 Group
IV outlined in Table 1 (total plexus injury plus Horner
syndrome) who show no signs of recovery definitely will
require reconstructive microsurgery to repair their dam-
aged plexus and improve their outcome.7,12,15,19,21 Similar-
ly, consensus reigns that infants with a mild neurapraxic
injury who demonstrate full recovery by 1 month are best
treated conservatively.
We have prospectively studied our patient population
to identify clinical criteria that would enable us to devel-
op a valid guideline for these treatment decisions. This

J. Neurosurg.: Pediatrics / Volume 3 / March 2009 173

Brought to you by Rutgers University Libraries | Unauthenticated | Downloaded 10/21/20 03:28 PM UTC
 J. R. Bain et al.
TABLE 1: Narakas classification of obstetrical brachial plexus TABLE 2: The AMS*
palsy*
Parameter Score
Group Lesion Level Clinical Picture 2–3 Wks After Birth
movement tested
I C5–6 paralysis of shoulder & biceps   shoulder flexion
II C5–7 paralysis of shoulder, biceps, & wrist/finger ex-   shoulder abduction/adduction
  tensors   shoulder external/internal rotation
III C5–T1 complete paralysis of limb including hand, no   elbow flexion/extension
  Horner sign
  forearm supination/pronation
IV C5–T1 complete paralysis of limb including hand w/
  wrist flexion/extension
  Horner sign
  finger flexion/extension
*  Data in this table were obtained from Narakas, 1987.   thumb flexion/extension
gravity-eliminated score
  no contraction 0
has allowed us to define the category of infants whose
recovery falls between these 2 extremes of the spectrum   contraction; no motion 1
as the GZ.   motion ≤0.5 range 2
  motion >0.5 range 3

Methods
There are 2 parts to this research study: a critical re-
view of the literature and a prospective cohort of children
with OBPI.
Critical Review of Literature
A standardized search strategy and critical review
of the medical literature was undertaken to determine
the existence of a GZ in other clinical publications. Ar-
ticles were included in the review if the authors clearly
described the scenario for a GZ (albeit not in those words)
and provided data that allowed an actual calculation of
the numbers of children within the GZ. Variable follow-
up times and measures were accepted. Recovery was de-
fined as complete or incomplete according to descriptions
provided by the authors.
Prospective Inception Cohort
Children with OBPI treated at McMaster Children’s
Hospital between 2002 and 2005 were analyzed. Clas-
sification of injury, changes in movement over time as
measured by the AMS,7 recovery status as measured by
the AMS, and whether surgical intervention was under-
taken were the variables of interest for this study. In this
study full recovery was defined as a perfect AMS score
attained within the first 3 months. The AMS highlighted
in Table 2 is a 7-point scale with 15 items that are used
to grade movement and not specific muscle action. The
AMS provides a score for each of the joints in the up-
per limb including the hand. Scores from 1 to 4 reflect
movement with gravity eliminated and scores from 5 to 7
are against gravity, and a score of 7 represents full range
against gravity. The total score with full active movement
against gravity is 105. The AMS provides guidelines for
standardized use in infants with OBPI. Reliability of 0.89
has been reported using weighted kappa on the total mea-
sure.
All infants referred to the Obstetrical Brachial Plex-
us Clinic at McMaster Children’s Hospital are systemati-
cally evaluated at monthly intervals until 6 months of age
  full range 4
against gravity score
  motion ≤0.5 range 5
  motion >0.5 range 6
  full range 7
*  Modified from Table 1 in Curtis et al., 2002, Copyright American So­
ci­ety for Surgery of the Hand.
and then at 3-month intervals or as needed by a multidis-
ciplinary team including an occupational therapist and a
peripheral nerve surgeon. There are ~ 4 new referrals per
month with 95% of infants < 2 months of age at referral.
Study Procedures
The prospective database of children with OBPI
treated at McMaster Children’s Hospital was analyzed to
explore the discriminators between the children in the GZ
and the overall group. Surgical and nonsurgical interven-
tion groups of infants in the GZ were evaluated using the
AMS. The AMS scores of elbow flexion, shoulder flex-
ion, abduction, external rotation, wrist extension, and to-
tal scores were examined at 1, 3, and 6 months to identify
any important trends that would guide surgical decision-
making. These particular movement items are the most
troublesome during recovery as well as the most reliably
measured. The AMS was administered by a trained and
unblinded occupational therapist.
The analysis of these discriminators permitted the
development of a guideline for surgical decision-making,
which decreases the ambiguity of current care plans. The
guideline was developed based on the existing evidence
and our clinical decision-making process. It was then
tested using the existing data as described below.
Outcome Data
Outcome data were examined at 1, 3, 4, 5, 6, and 12
months and up to 3 years of age. Not all children had reached
3 years by the time of analysis. The following are the 5 out-
come categories, defined for the guideline, which describe

174 J. Neurosurg.: Pediatrics / Volume 3 / March 2009

Brought to you by Rutgers University Libraries | Unauthenticated | Downloaded 10/21/20 03:28 PM UTC
Navigating the gray zone
TABLE 3: Publications with data that allowed for calculation of GZ*

Full
Biceps Recovery
No. of Study Type, Recovery GZ Recovery in Total Mean
Authors & Year Patients Patients in GZ (%) Population Time Measure (%) Sample (%) FU Time
Hoeksma et al., 56 historical cohort, OBPI identified at birth >3 wks Narakas Motor 51 incomplete, 66 46 mos
  2004   66   Scales   49 complete
Bisinella & Birch, 74 prospective cohort, OBPI, mean age at refer- >1 mo Mallett, Gilbert 41 incomplete, 53 32 mos
  2003   89   ral 3.2 mos (range 1–9   & Raimondi   59 complete
  mos)
Waters, 1999 66 prospective cohort, OBPI, evaluated w/in 1 mo Narakas, Mallet 95 incomplete, 12 36 mos
  87   1–3 mos old   5 complete
Noetzel et al., 94 prospective cohort, OBPI, persistent lack of 2 mos MRC scale 28 incomplete, 66 18–24 mos
  2001   71   shoulder or elbow anti-   72 complete
  gravity movement at 2
  wks, initial center eval-
  uation before 1–2 mos
Al-Qattan, 2000 43 prospective cohort, Erb palsy, seen in clinic 2 mos AMS 70 incomplete, 60 18 mos
  47   before 1 mo   30 complete
Strombeck et al., 247 prospective cohort, OBPI, sustained symp- 3 mos no standardized 91 incomplete, 27 5 yrs
  2000   63   toms of deltoid & bi-   tool   9 complete
  ceps at 2 & 3 mos
present study 125† prospective cohort, OBPI, seen in clinic be- 1 mo AMS 56 incomplete, 49 3 yrs
  81   fore 1 mos   44 complete

*  FU = follow-up; MRC = Medical Research Council.

†  Only 75 patients had AMS scores and were included in the study.

recovery status, treatment plan, and prognosis when an in-
fant exits the GZ: A, no surgery and complete recovery
(These infants exit the GZ between 1 and 3 months with
full AMS score and no functional deficits.); B, no surgery
and delayed recovery (These infants exit the GZ between 3
and 5 months and may achieve full AMS scores. At a later
age, these children exhibit residual functional impairment
such as scapular winging, shoulder weakness, and atypical
movement patterns. As there is a lack of strong outcome
evidence, conservative management is the choice.); C, no
surgery and incomplete recovery (These children exit the
GZ after 6 months with less than full AMS scores along
with significant functional impairment. The decision not
to proceed with surgery is due to parental choice.); D, early
surgery (These children exit the GZ before 6 months be-
cause of minimal progress in motor recovery, less than full
AMS scores, and significant functional impairment.); and
E, delayed surgery (These children exit the GZ between
7 and 12 months because of lack of parental readiness for
surgery or infant health issues.).
Critical Analysis
During critical analysis of the literature, the GZ was
calculated for all authors who provided data. Before 1
month of age, infants were not categorized into the GZ.
Complete recovery in the GZ is the percentage of infants
who did not attain complete recovery by 3 months, but
did have complete recovery at the follow-up evaluation.
The following formula was used to calculate the per-
centage of infants that fell within the GZ in these publi-
cations. We applied the same formula to our data: GZ =
(number of patients with delayed recovery [> 1 month]
excluding total pan plexus Narakas Group IV)/(total num-
ber of patients). The percentage of patients in the GZ with
complete recovery = (number of patients with complete
recovery in the GZ)/(total number of patients in the GZ).
The percentage of patients in the GZ with incomplete re-
covery = (number of patients with incomplete recovery in
the GZ)/(total number of patients in GZ).
Descriptive frequencies and cross-tabulations with
Pearson chi-square calculation were carried out to exam-
ine associations and relationships between an outcome of
perfect AMS score and Narakas group, and whether sur-
gery was performed and Narakas group. These analyses
compare the total group of infants with the infants in the
GZ. The total AMS score, elbow flexion and shoulder ab-
duction, external rotation, and flexion scores at 1, 3, and 6
months were compared between surgically and nonsurgi-
cally treated infants in the GZ.
The GZ guideline was tested by examining ranges and
the mean AMS scores in surgically treated versus non-
surgically treated infants. This study was approved by the
Research Ethics Board of the Faculty of Health Sciences,
McMaster University in Hamilton, Ontario, Canada.
Results
Table 3 summarizes the publications with data that
allowed us to evaluate their patients with our GZ defini-
tion. The time to spontaneous recovery described by those

J. Neurosurg.: Pediatrics / Volume 3 / March 2009 175

Brought to you by Rutgers University Libraries | Unauthenticated | Downloaded 10/21/20 03:28 PM UTC
 J. R. Bain et al.
TABLE 4: Population descriptive frequencies* TABLE 5: Patients grouped according to Narakas classification,
perfect AMS score, and surgical status
Total w/ AMS Scores GZ
Variable (125 patients) (75 patients) (61 patients) No. of Patients (%)
Perfect AMS
sex Op Performed†
Score*
  female 65 (52)
Narakas Group No Yes No Yes Total
  male 60 (48)
affected side AMS total group score
  rt 57 (46)   (75 patients)
  lt 63 (50)    I 10 (29) 25 (71) 34 (97) 1 (3) 35
  bilat (greater on rt) 4 (3)    II 19 (66) 10 (34) 19 (66) 10 (34) 29
  bilat (greater on lt) 1 (1)    III 5 (71) 2 (29) 4 (57) 3 (42) 7
birth weight (g)    IV 4 (100) 0 (0) 1 (25) 3 (75) 4
  mean ± SD 4100 ± 752    total 38 (51) 37 (49) 58 (77) 17 (23) 75
  range 1648–5710 GZ (61 patients)
Narakas Level   I 10 (36) 18 (64) 27 (96) 1 (4) 28
  I 56 (45) 35 (47) 28 (46)   II 19 (73) 7 (27) 16 (61) 10 (38) 26
  II 40 (32) 29 (39) 26 (43)   III 5 (71) 2 (29) 4 (57) 3 (43) 7
  III 17 (14) 7 (9) 7 (12)   total 34 (56) 27 (44) 47 (77) 14 (23) 61
  IV 7 (6) 4 (5) 0 *  For Narakas group, p = 0.002. For GZ, p = 0.015.
  missing 5 (4) 0 0 †  For Narakas group, p = 0.003. For GZ, p = 0.012.
op performed 26 (21) 17 (23) 14 (23)
5). For infants in the GZ, Table 6 illustrates the range of
*  Data are presented as the number of patients with percentages in AMS total scores, elbow flexion, shoulder abduction, ex-
parentheses unless otherwise stated. ternal rotation, and flexion scores at 1, 3, and 6 months.
The mean scores differed significantly between surgery
authors varied from 3 weeks to 3 months. In the literature, and no surgery groups in total AMS at 6 months. Shoul-
the GZ ranged from 47 to 89% of infants with OBPI. Com- der abduction and flexion differed significantly at 6 months
plete recovery within the GZ ranged from 9 to 59%. What while elbow flexion and external rotation differed at 3 and
is more striking is the percentage of incomplete recovery 6 months. In the surgery group all mean scores are lower
reported in the literature within the GZ, which ranged as are the ranges.
from 28 to 95%. The analysis of our population revealed There is great variability within the group in total
that 19% of our infants with OBPI either fully recovered scores particularly at 1 and 3 months. In the nonsurgical
or clearly required surgery. In the remaining 81% (those group the high standard deviations remain at 3 months,
within our GZ), 56% demonstrated incomplete recovery whereas those in the surgerical group are narrowing. How-
and 44% had complete recovery. ever, by 6 months variability in the nonsurgical group has
The data from our cohort of 125 infants included only decreased significantly and the variability remains in the
75 with AMS scores because of implementation of the surgical group. These analyses contributed to the values
measure after the database was established. This group of AMS used in creating and testing the GZ guideline il-
of 75 children will be referred to as the AMS group. They lustrated in Fig. 1.
serve as the main focus of comparative analysis, and the Figure 2 illustrates the ages and times that infants
subgroup is called the GZ. in our sample exited from the GZ. Forty-two percent ex-
Descriptive information is shown in Table 4. The ited the GZ without undergoing surgery and attained a
infants were categorized according to the Narakas clas- complete recovery (Group A). Similarly 20% of the pa-
sification17 based on root involvement. Four infants had tients underwent surgery (Group D), and the outcomes of
global paralysis (Narakas Group IV), and early surgery 75% of these patients had been decided 3 months later.
was advised. Ten children had early complete recovery by Twenty-eight percent of infants did not undergo surgery
1 month. Neither of these groups presented as a treatment (Group B) between 3 and 5 months but only attained an
decision challenge nor were they in the GZ. incomplete recovery. Seven percent of infants exited at 6
The remaining 61 infants comprise our GZ. These in- months without having undergone surgery (Group C) and
fants had Narakas Group I–III injuries. By 12–36 months, incomplete recovery. Finally, 3% exited to late/delayed
44% of children in the GZ attained complete recovery, surgery, between 7 and 12 months (Group E).
which we define as a full AMS score. In this GZ only 1
child underwent surgery and had a complete recovery. Discussion
Twenty (43%) of the children who did not have surgery had
incomplete recovery (p = 0.003, chi-square comparison). We consider the children with OBPI, whose clinical
The Narakas group is a significant factor in determin- status is such that their optimal therapy is unclear, to oc-
ing outcomes of surgery and a perfect AMS score (Table cupy what we have called the GZ. It is a zone of uncer-

176 J. Neurosurg.: Pediatrics / Volume 3 / March 2009

Brought to you by Rutgers University Libraries | Unauthenticated | Downloaded 10/21/20 03:28 PM UTC
Navigating the gray zone
TABLE 6: Elbow and shoulder AMS scores at 1, 3, and 6 months according to surgery versus no surgery groups for 61
infants in the GZ

AMS Scores*
No Op Op Performed
Area Assessed & Time Points (in mos) Range Mean ± SD Range Mean ± SD p Value†
total
  1 13–102 72.0 ± 21.9 32–90 59.6 ± 17.7 0.264
  3 69–105 88.1 ± 18.6 37–92 69.3 ± 12.8 0.517
  6 83–105 99.3 ± 6.4 66–102 82.2 ± 12.2 0.020
elbow flexion
  1 0–7 3.0 ± 2.4 0–3 .9 ± 1.1 0.234
  3 0–7 5.3 ± 1.7 0–6 1.8 ± 1.9 0.001
  6 2–7 6.3 ± 1.3 0–7 4.0 ± 2.3 0.003
shoulder abduction
  1 0–7 2.8 ± 2.2 0–5 1.3 ± 1.7 0.313
  3 1–7 4.8 ± 1.9 0–5 2.9 ± 1.9 0.014
  6 3–7 6.1 ± 1.0 2–6 3.7 ± 1.7 0.006
shoulder flexion
  1 0–7 2.8 ± 2.0 0–5 1.6 ± 1.5 0.476
  3 1–7 4.9 ± 1.7 1–6 2.9 ± 1.4 0.067
  6 2–7 6.0 ± 1.2 2–6 4.1 ± 1.5 0.002
shoulder external rotation
  1 0–7 2.8 ± 2.1 0–5 1.1 ± 1.7 0.815
  3 0–7 4.4 ± 2.2 0–5 1.2 ± 1.8 0.002
  6 1–7 5.5 ± 1.7 0–7 2.8 ± 2.4 0.007

*  The total score is 105. A score of 7 for each item indicates full movement.
†  Pearson chi-square test.

tainty. Many other groups have referred to this subset of
infants (Table 3), but they have not labeled them as such.
Recovery times reported by these authors varied from 3
weeks to 3 months. This lack of precision supports that a
GZ exists. Clinicians require additional guidance for de-
cision making within this area until long-term outcomes
are well researched and clear, definitive surgical criteria
are established.
In our population, 81% of our patients fell within
this zone at 1 month of age. The large variability in their
AMS scores at this stage makes it difficult to decide by 1
month those infants in the GZ who should undergo sur-
gery. However, decisions become clearer as the infant is
monitored through to 3 months.
A definitive decision against surgical intervention re-
quires strict criteria of full recovery. Prior series used a
broad definition of recovery and reported 70–90% spon-
taneous recovery.16 However, a recently published system-
atic review of the natural history of OBPI illustrates that
this common perception is in fact inaccurate.19 There is
now evidence to suggest that many infants who are re-
ported to “recover spontaneously” do have residual defi-
cits in function even if their nerves are intact.2,4,20
Early surgery is only definitely required for chil-
dren classified as Narakas Group IV, complete palsy and
Horner syndrome. A very small proportion of children
are at this end of the spectrum.
The development of our GZ guideline is based on the
earlier work of Gilbert,11 Michelow,16 and Curtis7 and their
colleagues. Gilbert et al.13 concluded that assessment of
biceps function best informs the need for surgery. In this
reported series of > 1000 children, the absence of biceps
function at 3 months of age was the primary indicator for
surgery. Clarke et al.5,6 and Michelow et al.16 have exten-
sively used the AMS7 in evaluating recovery over time in
infants with OBPI. Their surgical prediction model com-
bines AMS scores of elbow flexion and extension, wrist
extension, thumb extension, and finger flexion and exten-
sion. They have concluded that using elbow flexion alone
as a predictor leads to significant false-positive findings.16
Using their combined prediction model, an inaccurate
prediction rate is reduced to 5.2%.
The GZ guideline presented in Fig. 1 is a departure
from those that use only elbow flexion as a determinant
for surgery.13 Our approach is similar to the combined
prediction model of Michelow et al.16 and Clarke et al.,6
because the AMS is an important component. The AMS
has been used as the objective measure of recovery with
full recovery defined as a perfect AMS score within the
first 3 months. The AMS score alone is not enough to
evaluate recovery outcomes in the toddler and child but
does work very well as early evidence of recovery in the
young infant. In keeping with Hoeksma et al.,14 Bisinella
and Birch,3 and Waters,21 our GZ guideline also places

J. Neurosurg.: Pediatrics / Volume 3 / March 2009 177

Brought to you by Rutgers University Libraries | Unauthenticated | Downloaded 10/21/20 03:28 PM UTC
 J. R. Bain et al.

Fig. 1.  Chart showing the GZ guideline for OBPI management. The asterisks indicate that the score accounts for ≥ 2 of the
following movements: elbow flexion (biceps) shoulder abduction, external rotation shoulder flexion, and wrist extension. See
Methods for the definition of the outcome categories (A–E).

strong emphasis on the assessment of shoulder function.
These authors have made a strong case that if recovery
of shoulder as well as biceps function is not present by 1
month of age then there is a significant risk of incomplete
recovery. The residual functional problems described by
Strombeck et al.20 at 5 years after support this perspective.
We have found that infants in the GZ may demonstrate
a pattern of recovery that is often “patchy,” unusual, or
atypical, such as recovery in the elbow but not shoulder or
vice versa. Given the analysis of our AMS scores, func-
tional outcomes, and recent evidence, we now suggest the
inclusion of functional measures along with increased at-
tention to the shoulder in both surgical decision-making
and future outcome studies.

Fig. 2.  Chart showing the GZ outcomes by category. m = month.

178 J. Neurosurg.: Pediatrics / Volume 3 / March 2009

Brought to you by Rutgers University Libraries | Unauthenticated | Downloaded 10/21/20 03:28 PM UTC
Navigating the gray zone
To determine the need for surgical intervention, our
GZ guideline uses the criteria of AMS cutoff scores at each
month in ≥ 2 of the following movements: elbow flexion
(biceps), shoulder abduction, external rotation and flexion,
and wrist extension. Also, we have recorded and calculated
the mean AMS scores for our subset of children that ben-
efited from surgical intervention. If a child’s total AMS
score is less than the mean values in the surgical group
at the corresponding month, surgical intervention is prob-
ably warranted. If an infant has AMS scores ≤ 1 in ≥ 2 of
the previously mentioned movements at 1 month, then sur-
gery is discussed but the child remains in the GZ. We have
seen that infants with variable AMS scores particularly at
1 month can still demonstrate recovery and improvement.
Table 6 illustrates these movement scores at 1, 3, and 6
months in our population.
If an infant with OBPI falls into the GZ they exit in
a number of ways within the first 6 months when surgical
management is being decided. Figure 1 illustrates these
outcomes. In our population of infants, no infant exited
the zone at 1–2 months unless he or she had an AMS
score ≥ 6 in all movements on the scale. Those in Group
A exited before 3 months by demonstrating full recovery
with a complete AMS score (or > 90 points). As some
infants improve over time, more will exit the GZ without
surgery by attaining shoulder and elbow AMS scores ≥
5 by 5 months. The patients in Group B had good early
motor recovery such that surgery was not recommended
as it was believed the outcome would be equitable with
no surgery. It is imperative to note that these children did
go on to demonstrate residual functional impairment in
the shoulder and scapula not captured by the AMS. The
patients in Group D clearly required surgery, which they
underwent before or at 6 months. The patients in Group
E required surgery, but because lack of parental readi-
ness or the child’s health, it was delayed beyond 6 months.
Those in Group C exited at 6 months with a decision for
no surgery due to parental choice. These children demon-
strated significant residual impairment and primary nerve
reconstructive surgery was recommended. However, par-
ents declined this treatment option.
In light of recent evidence about long-term out-
comes,2,4,20,21 we are concerned about the 16 children in
Group B. These infants did not meet current evidence-
based surgical criteria based on their achievement and tim-
ing of AMS scores and function. This illustrates another
characteristic of children in the GZ that includes inconsis-
tency and decreased frequency of movement, despite the
infant attaining AMS scores of ≥ 5 in key muscle groups.
Strombeck et al.20 specifically reported that children who
had poor function at 3 months but had “recovered” by 6–9
months still have deficits in shoulder external rotation,
forearm supination, decreased grip strength, and biman-
ual function. Other groups have also demonstrated similar
findings.14,19,21 It is imperative that families be informed
that incomplete recovery by 5–6 months equates with in-
complete recovery in the future. The GZ ends at 6 months,
suggesting that the decision for surgery should be clear by
then. However, the existence of our Group B suggests even
more rigorous criteria are needed in addition to this guide-
line.
J. Neurosurg.: Pediatrics / Volume 3 / March 2009
Brought to you by Rutgers University Libraries | Unauthenticated | Downloaded 10/21/20 03:28 PM UTC
Our outcomes using this guideline support the view
of Gilbert et al.11–13 that earlier surgical decision-making
is possible and optimizes physiological recovery of dener-
vated muscle. Experimental and clinical data have clearly
determined that time to muscle reinnervation is one of the
most significant factors affecting the quality and quan-
tity of muscle functional recovery.10 In most experimental
models, delays to reinnervation beyond 3 months lead to
muscle fiber atrophy largely from structural protein deg-
radation, myocyte apoptosis, and secondary fibrosis and/
or fatty replacement.9 Delaying microneural surgery to
reconstruct the plexus with grafts, transfers, and/or neu-
roysis condemns the muscle to incomplete recovery.10
We have only one opportunity in the neonatal period
to maximize the neuronal input to the extremity to op-
timize the physiological muscle and bone recovery. It is
important that this opportunity is given to all infants who
truly need it. All factors that can facilitate this decision
should be considered.
The GZ guideline presented here is a multidisci-
plinary, family-centered model that incorporates the best
existing evidence and the outcomes in the children fol-
lowed by our OBPI program.
Conclusions
There is compelling evidence that there is a group of
infants in whom the present evidence and criteria for sur-
gery do not provide a clear path for decision-making as to
the need for surgical intervention. The risks of no surgery
versus the benefit of surgery are not evident. The variation
in surgical criteria and expectation for amount and timing
of recovery described in the body of literature confounds
the issue. These infants are considered to lie within a GZ
where decisions as to surgery versus no surgery are not
black and white. The guideline presented in this paper
provides another framework for decision-making that in-
cludes other areas in addition to biceps recovery or elbow
function. Although we do not have evidence from a clini-
cal trial to support that this guideline is the best approach,
within the context of our prospective data collection, we
present this guideline to assist clinicians working with in-
fants with OBPI to navigate the GZ.
Disclaimer
The authors do not report any conflict of interest concerning
the materials or methods used in this study or the findings specified
in this paper.
References
  1.  Al-Qattan MM: The outcome of Erb’s palsy when the decision
to operate is made at 4 months of age. Plast Reconstr Surg
106:1461–1465, 2000
  2.  Bellew M, Kay SP, Webb F, Ward A: Developmental and be-
havioural outcome in obstetric brachial plexus palsy. J Hand
Surg [Br] 25:49–51, 2000
  3.  Bisinella GL, Birch R: Obstetric brachial plexus lesions: a study
of 74 children registered with the British Paediatric Surveil-
lance Unit (March 1998-March 1999). J Hand Surg [Br] 28:
40–45, 2003
  4.  Brown T, Cupido C, Scarfone H, Pape K, Galea V, McComas
179

A: Developmental apraxia arising from neonatal brachial
plexus palsy. Neurology 55:24–30, 2000
  5.  Clarke HM, Al-Qattan MM, Curtis CG, Zuker RM: Obstetrical
brachial plexus palsy: results following neurolysis of conduct-
ing neuromas-in-continuity. Plast Reconstr Surg 97:974–
982, 1996
  6.  Clarke HM, Curtis CG: An approach to obstetrical brachial
plexus injuries. Hand Clin 11:563–580, 1995
  7.  Curtis C, Stephens D, Clarke HM, Andrews D: The active
movement scale: an evaluative tool for infants with obstetrical
brachial plexus palsy. J Hand Surg [Am] 27:470–478, 2002
  8.  Duclos L, Gilbert A: Obstetrical palsy: early treatment and
secondary procedures. Ann Acad Med Singapore 24:841–
845, 1995
  9.  Fu SY, Gordon T: The cellular and molecular basis of periph-
eral nerve regeneration. Mol Neurobiol 14:67–116, 1997
10.  Fu SY, Gordon T: Contributing factors to poor functional re-
covery after delayed nerve repair: prolonged denervation. J
Neurosci 15:3886–3895, 1995
11.  Gilbert A, Brockman R, Carlioz H: Surgical treatment of bra-
chial plexus birth palsy. Clin Orthop Relat Res 264:39–47,
1991
12.  Gilbert A, Pivato G, Kheiralla T: Long-term results of pri-
mary repair of brachial plexus lesions in children. Microsur-
gery 26:334–342, 2006
13.  Gilbert A, Razaboni R, Amar-Khodja S: Indications and re-
sults of brachial plexus surgery in obstetrical palsy. Orthop
Clin North Am 19:91–105, 1988
14.  Hoeksma AF, ter Steeg AM, Nelissen RG, van Ouwerkerk WJ,
Lankhorst GJ, de Jong BA: Neurological recovery in obstet-
ric brachial plexus injuries: an historical cohort study. Dev
Med Child Neurol 46:76–83, 2004
15.  McNeely PD, Drake JM: A systematic review of brachial plex-
180
Brought to you by Rutgers University Libraries | Unauthenticated | Downloaded 10/21/20 03:28 PM UTC
J. R. Bain et al.
us surgery for birth-related brachial plexus injury. Pediatr
Neurosurg 38:57–62, 2003
16.  Michelow BJ, Clarke HM, Curtis CG, Zuker RM, Seifu Y,
Andrews DF: The natural history of obstetrical brachial plex-
us palsy. Plast Reconstr Surg 93:675–680, 1994
17.  Narakas AO: Obstetrical brachial plexus injuries, in Lamb
DW (ed): The Paralysed Hand. New York: Churchill Liv-
ingstone, 1987, pp 116–135
18.  Noetzel MJ, Park TS, Robinson S, Kaufman B: Prospective
study of recovery following neonatal brachial plexus injury. J
Child Neurol 16:488–492, 2001
19.  Pondaag W, Malessy MJ, van Dijk JG, Thomeer RT: Natural
history of obstetric brachial plexus palsy: a systematic review.
Dev Med Child Neurol 46:138–144, 2004
20.  Strombeck C, Krumlinde-Sundholm L, Forssberg H: Func-
tional outcome at 5 years in children with obstetrical brachial
plexus palsy with and without microsurgical reconstruction.
Dev Med Child Neurol 42:148–157, 2000
21.  Waters PM: Comparison of the natural history, the outcome
of microsurgical repair, and the outcome of operative recon-
struction in brachial plexus birth palsy. J Bone Joint Surg
Am 81:649–659, 1999
Manuscript submitted April 28, 2008.
Accepted December 9, 2008.
A preliminary version of this work was presented at the “XIV
International Symposium on Brachial Plexus Surgery - Club A.
Narakas”, Brescia, Italy, October 2004.
Address correspondence to: Carol DeMatteo, M.Sc., Dip. P.&O.T.,
School of Rehabilitation Science, McMaster University, 1400 Main
Street West, Hamilton, Ontario, Canada, L8S 1C7. email: dematteo@
mcmaster.ca.
J. Neurosurg.: Pediatrics / Volume 3 / March 2009