88 / 280
hospital charges to identify total charges for the groups
(
Figure 3
). As expected, the addition of home health care
resulted in a greater disparity between patients receiving a
tracheostomy versus those receiving MD only. We observed
that the Trach
1
MD group incurred fewer charges than the
Trach group in all 3 years, although this difference was not
statistically significant (Year 1
P
= .27, Year 2
P
= .30,
Year 3
P
= .29). This trend may be attributed to a higher
rate of decannulation among Trach
1
MD (5 of 9, 56%)
versus Trach (2 of 12, 17%) over the 3-year period.
To examine the contribution of syndromic status, all
patients were grouped based on diagnosis of isolated PRS
(34 patients, 72%) or syndromic (including unknown syn-
dromes) PRS (13 patients, 28%). As shown in
Table 1
,
although the MD group had a higher percentage of nonsyn-
dromic patients, this difference was not significant (
P
=
.10).
Figure 4
compares the charges between these 2
groups. Syndromic patients had higher associated charged
for all 3 years; however, this was only significant during
year 2 (
P
= .03).
Discussion
Tracheostomy effectively bypasses tongue-based obstruction
and remains the gold standard for severe obstruction that
may occur with PRS. However, tracheostomy has greater
associated morbidity including negative long-term speech
effects, difficulties with feeding, psychosocial delays, fre-
quent hospital admissions for tracheitis and pneumonia,
buildup of airway granulation tissue, and occasional need for
complicated revision surgery, including laryngotracheoplasty
and cricotracheal resection.
11-14
Further, patients with a tra-
cheostomy typically require multiple ML&B procedures to
investigate these morbidities and to prepare for decannula-
tion. Of greatest significance, tracheostomy is associated with
a small but real chance of mortality (1%-5%).
15
Mandibular distraction differs in that it directly addresses
the primary problem, micrognathia. Using MD to lengthen
the mandible provides greater room for the tongue and oral
soft tissues and indirectly pulls them forward by their attach-
ments to the mandible, correcting glossoptosis and improving
airway obstruction. A growing body of studies indicate MD
helps PRS patients treated with tracheostomy achieve decan-
nulation sooner or avoid tracheostomy altogether.
9,11,14,16-24
Complications associated with MD include hardware mal-
function, infection, damage to tooth buds, and nerve injury
and pain, although the actual incidence varies depending on
surgeon experience and technique.
25,26
Two groups have performed cost analyses to compare
tracheostomy to MD for PRS. Kohan et al
27
examined 149
neonates with PRS treated with either internal MD (n = 43)
or tracheostomy (n = 73). They reported a 2-fold higher cost
for the Trach group ($382,246) compared to MD group
($193,128) over a 4-year follow-up period. The cost differ-
ence was due to an increased length of ICU stay in patients
receiving tracheostomy. Hong et al
28
examined 52 patients
with PRS: 21 received MD, and 31 had a tracheostomy.
With 1 year of follow-up data, the Trach group had a 1.6-
fold increase in cost compared to the MD group ($92,164 vs
$57,649, Canadian dollars). This cost difference was attrib-
uted to increased hospital stay for tracheostomy patients, as
their health system mandates 90 days in house for home tra-
cheostomy care arrangement. Both studies used averaged
operative and ICU per diem fees rather than individual
patients’ billed charges, as done in our study. Also, neither
0
50000
100000
150000
200000
250000
300000
MD
Trach
Trach+MD
Year 1
Year 2
Year 3
Figure 3.
Estimated total charges ($USD) of groups inclusive of
home health care costs. Total charges over 3 years following initial
surgical intervention for patients with Pierre Robin sequence
receiving mandibular distraction (MD), tracheostomy (Trach), or
tracheostomy with subsequent MD (Trach
1
MD), including home
tracheostomy-care costs (eg, supplies, equipment rental, and home
nursing fees). Median values (horizontal line) are presented with
twenty-fifth through seventy-fifth percentile ranges. Statistical analy-
ses by year: years 1-3:
P
\
.01 (Kruskal-Wallis comparison for non-
parametric data). A Wilcoxon rank sum test was used to compare
Trach vs Trach
1
MD groups at each time point: year 1:
P
= .27;
year 2:
P
= .30; year 3:
P
= .29.
0
20000
40000
60000
80000
100000
120000
Non-syndromic
Syndromic
Year 1 Year 2 Year 3
Figure 4.
Annual charges ($USD) of syndromic versus non-syn-
dromic patients. Cost comparison (not including home care
charges) of patients with non-syndromic versus syndromic Pierre
Robin sequence over 3-year period. Median values (horizontal line)
presented with twenty-fifth through seventy-fifth percentile ranges.
Wilcoxon rank sum tests used to compare the 2 groups at each
time point: Year 1:
P
= .26; Year 2:
P
= .03; Year 3:
P
= .13.
Runyan et al
66