months. The outcome of ‘‘conscious’’ per se
does not reflect level of disability. However,
the Task Force review indicated that for adults,
within the 52% of patients recovering con-
sciousness after 1 month in VS, only 24% be-
came independent by GOS criteria. This figure
dropped to 16% for VS lasting 3 months and to
only 4% if taken out to at least 6 months.
Not surprisingly, nontraumatic VS carries a
far less optimistic prognosis. Figure 9–4 shows
comparison percentages for adult and pediatric
patients with nontraumatic VS. For adult VS
patients remaining in VS at 1 month, only 15%
regained consciousness (with only 4% inde-
pendent by GOS). These percentages worsened
to 8% and 0% for patients remaining in a non-
traumatic VS for 3 and 6 months, respectively.
Based on these data, the Task Force paper
suggested that VS after 12 months following
TBI, or 3 months following an anoxic injury,
should be considered essentially permanent.
However, it is important to recognize that a small
number of patients may recover from VS be-
yond these time points.
67–69
Such late recovery
past the cutoffs for permanent VS from both
anoxic and traumatic etiologies has generally
been to levels of severe disability, including the
minimally conscious state.
66
Nevertheless, appli-
cation of these statistics to individual cases can
be risky, unless independent evidence of the
mechanism of brain injury is available, as rare
cases of late recovery continue to be reported.
The uncertainty in prognosis in such cases
highlights the need for better methods, such as
direct measurements of cerebral function, to
help identify cases where recovery is likely.
Mortality is very high within the first year;
approximately one-third of patients die.
64,65
If
patients remain alive after a year, mortality per
year is low and some patients may continue to
live for many years.
66
Plum and Schiff studied
one patient who had remained in PVS for 25
years (see Figure 9–8). Most patients in VS die
from infection of the pulmonary system or
urinary tract.
Clinical, Imaging, and
Electrodiagnostic Correlates of
Prognosis in the Vegetative State
A few clinical signs or confirmatory tests, in-
cluding those negative predictors for coma in
general (as reviewed previously), help predict
the prognosis of VS. As noted, abnormal SSEPs
reliably indicate cortical damage and a high
probability of remaining in VS following anoxic
and traumatic brain injuries. However, normal
evoked responses do not predict recovery. In a
study of 124 patients in VS or MCS following
TBI, three variables predicted recovery of abil-
ity to follow commands: (1) initial score on the
Disability Rating Scale (DRS), (2) rate of change
on the DRS measure in the first 2 weeks of
observation, and (3) the time of admission to a
rehabilitation program following injury.
70
Several structural and functional correlates
of VS have been examined. A prospective study
of MRI imaging correlates of 80 patients re-
maining in VS following TBI at 6 to 8 weeks
(with MRI and clinical follow-up at 2, 3, 6, 9,
and 12 months) found that 42 patients who
remained in VS at 1 year showed that structural
injuries within the corpus callosum and dorso-
lateral brainstem significantly predicted nonre-
covery (214-fold and sevenfold higher probabi-
lity of nonrecovery from VS, respectively, based
on adjusted odds ratios accounting for age,
GCS, pupillary dysfunction, and number of
brain lesions).
71
Overall, this model achieved a
classification rate of 87.5% for identifying pa-
tients who would not recover past VS.
Quantitative
fluorodeoxyglucose-positron
emission tomography (FDG-PET) studies mea-
suring resting cerebral metabolism have con-
sistently demonstrated that global cerebral me-
tabolism is markedly reduced to 40% to 50% of
normal metabolic rates in most VS patients (see
page 365). Unfortunately, early identification
of low metabolic activity is not a clear predictor
of outcome and some patients have recovered
consciousness despite significant remaining ab-
normalities in resting metabolic level.
72
The N-
acetylaspartate (NAA)-to-creatine (Cr) ratio on
magnetic resonance spectroscopy (MRS) of the
thalamus is low in all patients in VS, but lower
in those who do not recover
73
; however, only a
few patients have been studied.
Efforts to predict outcome or characterize
VS using EEG have been disappointing. EEG
studies may remain abnormal as patients im-
prove or, conversely, improve when patients
do not.
74
Event-related potentials (ERPs) may
hold more promise. These potentials require
cortical processing of stimuli that can be pas-
sively presented to subjects in VS. The re-
sponses are long latency with peak activation
Consciousness, Mechanisms Underlying Outcomes, and Ethical Considerations
359