PSG and Portable Monitoring

Report
Polysomnography and Portable
Monitoring
BY
AHMAD YOUNES
PROFESSOR OF THORACIC MEDICINE
Mansoura faculty of medicine
POLYSOMNOGRAPHY
• POLYSOMNOGRAPHY is the standard test for
1-Diagnosis of suspected sleep-related breathing
disorders
2- Positive airway pressure titration
3- Document the efficacy of prior surgical treatment
and determine the efficacy of oral appliance
treatment.
4- Evaluation of parasomnias (under certain
conditions).
5- Narcolepsy (when combined with a multiple sleep
latency test)
Portable Monitoring
•Attended cardiorespiratory pulmonary studies (airflow
+ two effort channels, oximetry, and heart rate) are an
acceptable alternative to PSG for diagnosis of
obstructive sleep apnea (OSA) in patients with high
pretest probability of OSA or for determining the
adequacy of prior surgery for OSA or effectiveness of
current OA treatment.
•Unattended PM monitoring is also acceptable for
assessing the efficacy of non-PAP treatment of OSA.
•The Centers for Medicare and Medicaid Services
(CMS) now recognize a diagnosis of OSA by
unattended PM as acceptable for reimbursement of
CPAP treatment
POLYSOMNOGRAPHY
• PSG is the standard diagnostic study for evaluation
of a suspected SRBD.
• A diagnostic study may be repeated if the initial
study was negative for sleep apnea and there is
a high clinical index of suspicion for this
disorder.
• PSG should be performed preoperatively for
planned surgery to treat snoring or suspected
OSA.
• A PSG is indicated after surgery for OSA (after
surgical healing) to document effectiveness.
POLYSOMNOGRAPHY
• For patients with prior effective surgical
treatment (documented by PSG), the PSG may
be repeated at a later time if symptoms of sleep
apnea return.
• For a patient using an OA as treatment for OSA,
the PSG may also be repeated while the patient
wears the OA if the patient’s symptoms return.
• If a patient on CPAP loses more than 10% of
body weight, a diagnostic PSG is indicated to
determine whether CPAP is still needed.
5
Mechanism of CSA
POLYSOMNOGRAPHY
• Number of circumstances in which SRBDs are very common
(patients with systolic or diastolic heart failure, recent or
past stroke or transient ischemic attack (TIA), coronary
artery disease, and tachyarrhythmias or bradyarrhythmias)
PSG is NOT routinely indicated in those circumstances
unless a clinical evaluation reveals a reasonable suspicion
for SRBD.
• Most clinicians would also place resistant hypertension or
pulmonary hypertension of unknown etiology in this
category as well.
• The clinician should recognize that many patients with
significant OSA do not complain of daytime sleepiness.
• A history of snoring and gasping would suggest a PSG is
indicated.
• It should also be noted that some patients with OSA
complain of insomnia.
• The practice parameters do list neuromuscular diseases as a
group of disorders in which PSG is indicated for evaluation
of SDB.
POLYSOMNOGRAPHY
• Routine evaluation of chronic lung disease is not an
indication for PSG unless coexistent OSA is suspected.
• Nocturnal oximetry is a useful tool for determining
whether nocturnal oxygen desaturation is occurring in a
patient with COPD.
• A saw-tooth pattern is suggestive of sleep apnea.
• A PSG for PAP titration is the standard procedure to
select a level of pressure for treatment.
• The titration can be performed on a separate night after
a diagnostic PSG or during the second part of the night
during a split (partial night) study.
POLYSOMNOGRAPHY
• A split sleep study is recommended when
(1) the diagnostic portion shows an AHI > 40 /hr with at least 2 hours
of monitoring,
(2) there is an AHI of 20 to 40 with special clinical circumstances such
as severe desaturation or arrhythmia thought due to OSA
(3) at least 3 hours remain for the PSG titration.
If the PSG titration does not last at least 3 hours or is not adequate, a
repeat PSG titration is indicated.
• If the patient is being treated on CPAP and is NOT doing well, a
repeat PSG study on CPAP is indicated. However, before this
expensive procedure ,it is essential to document adequate
objective adherence and to optimize treatment and the mask
interface.
• PSG is also indicated if a patient on CPAP gains more than 10% of
body weight to determine whether the pressure is adequate.
However, a repeat PSG titration may not be clinically indicated.
Hypercapnic CSA
• Impaired Central Drive ("Won’t Breathe"):
1-Tumors or trauma-induced lesions to brainstem
2-congenital central hypoventilation syndrome (Ondine curse)
3-opioid-induced CSA
4- OHS
Impaired Respiratory Motor Control ("Can’t Breathe")
1-myasthenia gravis
2-amyotrophic lateral sclerosis
3-post-polio syndrome
4-myopathies
5-Chest wall syndromes such as kyphoscoliosis
PSG Indications: Nonrespiratory Disorders
• A PSG preceding an MSLT is indicated for evaluation of
suspected narcolepsy or to help differentiate narcolepsy
from idiopathic hypersomnia.
• PSG is indicated for evaluation of suspected periodic
limb movement disorder but NOT the restless legs
syndrome (RLS; a clinical diagnosis).
• A PSG is indicated to evaluate (1) nocturnal behavior
possibly due to seizures, (2)atypical parasomnia
behavior (frequent episodes each night ,stereotypic
behavior, or behavior unusual for age), (3) nocturnal
behavior/parasomnia that has resulted in injury to the
patient or others (4) presumed parasomnia or nocturnal
seizure disorder that does not respond to conventional
treatment, or (5) if there are legal/forensic implications
of nocturnal behavior.
PSG NOT Indicated CSA
• PSG is not indicated for evaluation of insomnia (unless OSA is
suspected) or unless insomnia does not respond to usual
treatment.
• PSG is not indicated for diagnosis of depression or insomnia with
depression.
• The practice parameters for use of PSG in evaluation of insomnia
state “PSG is indicated when initial diagnosis is uncertain,
treatment fails (behavioral or pharmacologic), or precipitous
arousals occur with violent or injurious behavior (Guideline).
• Although the PSG finding of a short REM latency is common in
depression, this finding is not sufficiently sensitive or specific to
warrant a PSG in evaluating patients with suspected depression.
• A PSG is NOT indicated for (1) typical uncomplicated and
noninjurious parasomnias when the diagnosis is clearly delineated
or (2) patients with known seizure disorders who have no nocturnal
complaints.
Approach to the Sleepy Patient on PAP Treatment
• If a patient with OSA continues to have daytime sleepiness on
adequate PAP treatment (documented adequate adherence and
adequate control of respiratory events), there are two options., an
alerting agent such as modafinil may be added. If there is a
suspicion of narcolepsy in addition to OSA, a nocturnal PSG on
PAP treatment (documenting adequate treatment) followed by an
MSLT on PAP can be performed
• If unambiguous cataplexy is present, the patient likely has
narcolepsy and the MSLT is simply confirmatory.
• It is essential that adequate PAP treatment for OSA be confirmed
before expensive testing. This includes documentation of adequate
objective adherence.
• Many PAP devices also have the ability to record residual AHI.
• A high residual AHI would be an indication for an adjustment in
pressure (empirical increase) or a PSG PAP titration.
• The PAP device estimate of residual AHI is not always accurate, but
a high value has a reasonable positive predictive value that the
residual AHI is indeed elevated.
• A surprising number of patients (up to 15%) on chronic PAP
treatment are not adequately treated (AHI >10/hr).
Approach to Reading the PSG
• Before the PSG is read, a review of the clinical history with special
attention to symptoms of sleep apnea, narcolepsy, RLS, and
medications is very useful.
• The presence of underlying lung disease may help explain a low
awake arterial oxygen saturation (SaO2) or low baseline sleeping
SaO2.
• A clinical history of pacemaker insertion or known atrial fibrillation is
also very helpful in providing a useful interpretation of ECG findings.
• If a PAP titration is planned for a patient currently using CPAP, the
current treatment pressure level should be noted.
• All digital PSG systems have a view that shows graphical summary
information of the entire night. It is often useful to look at the big
picture before going through the data in smaller time windows.
• The biocalibrations are often helpful in noting the appearance of eyesopen wake in a given patient and whether the patient produces an
alpha rhythm with eye closure .
Summary view shows three periods of significant arterial
oxygen desaturations (A, B, and C) during rapid eye movement
(REM) sleep. In addition, the summary view shows that no
supine REM sleep was recorded. SpO2 = pulse oximetry.
PORTABLE MONITORING (HOME SLEEP TESTING, OUT
OF CENTER SLEEP TESTING)
• The tests are not always performed in the home ;hence,
the terms HST or PM are not ideal but are used in much
of the literature on this subject.
• In the past, PM has been used to diagnose OSA in
settings in which access to PSG is limited or delayed.
• The original classification used “level I, II, III, and IV” to
refer to different classes of monitoring but currently the
terminology is “type 1, 2, 3, and 4.”
• The Centers for Medicare and Medicaid Services (CMS)
has a different classification for monitoring The CMS
terminology defines the respiratory disturbance index
(RDI) as the total number of apneas and hypopneas per
hour of monitoring time. Therefore, the index
determined by PM (no EEG) would be an RDI using the
CMS definition. CMS also refers to PM as HST.
Accuracy of PMs
• There are a number of reasons why the AHI from PSG and PM might
differ. Even if the same type of sensors are used and the same
number of events are identified, the two devices would still give
different AHI values as monitoring time (PM) would exceed total sleep
time (PSG).
• It is also possible that the oximeters used in different systems could
differ in their ability to detect desaturations.
• Comparing PM and PSG AHI values during simultaneous recording
reduces the effects of night-to-night variability.
• The AHI values could differ simply due to night-to-night variability in
the AHI. In general for milder OSA, the amounts of REM and supine
sleep are the major determinants of the AHI. These patients have
elevated AHI mainly in the supine position or during REM sleep. In
milder OSA patients, night-to-night differences in the proportions of
REM and supine sleep are a significant source of night-to-night
variability.
Accuracy of PMs
• Differences in the AHI during simultaneous
monitoring were likely due to the effect of using
monitoring time versus TST to compute the AHI.
• During PM monitoring at home, less supine sleep
also contributed to differences in the AHI.
• To compensate for night-to-night variability, many
PM devices now have the ability to monitor multiple
nights.
• One study suggested that night-to-night variability in
the home might be less than in the sleep center.
Clinical Use of PM
•
•
•
•
•
•
•
Indications for Attended PM is rarely used in the attended setting.
In 1994 and 1997 practice parameters,attended type 2 or 3 PMs were stated to
be acceptable alternatives when treatment of OSA is urgent and PSG is
delayed, the patient is unable to be studied in the sleep center (safety or
immobility), or the diagnosis of OSA is already established and the purpose is
to evaluate the response to treatment.
The 2003 practice parameters for the use of portable monitoring added that
certain PMs may be used in the attended setting to rule in or to rule out OSA.
The 2005 practice parameters for PSG stated that attended cardiorespiratory
studies (type 3 PM) are an acceptable alternative to PSG for diagnosis of OSA
in patients with a high probability of OSA as long as a negative PM study was
followed by PSG.
Attended cardiorespiratory studies were also said to be acceptable for
preoperative evaluation for planned surgery for snoring or OSA. After surgery,
attended cardiorespiratory studies were acceptable to document surgical
effectiveness in patients with moderate to severe OSA.
A specific statement concerning the use of attended cardiorespiratory studies
in patients being considered for an OA treatment of snoring or suspected
sleep apnea was not made in any of the practice parameters.
In the attended setting, cardiorespiratory studies are acceptable to document
the effectiveness of an OA (after adequate adjustment) or in OSA patients
treated with surgery or an OA when symptoms return.
Indications for Unattended PM
• PM is most often used in the unattended setting.
• The recent AASM PMGs state that unattended PM “may” be indicated
for diagnosis of OSA in patients with a high pretest probability or for
documenting the efficacy of non-PAP treatments for OSA—IF guidelines
for patient selection and procedures for PM performance and
interpretation are followed.
• In brief, PM must be combined with a comprehensive sleep evaluation by
a qualified physician, patients must have a high probability of moderate
to severe OSA, there must be an absence of medical co-morbid
conditions that degrade the accuracy of PM (severe lung disease,
neuromuscular disease, and congestive heart failure). Patients with comorbid sleep disorders (e.g., narcolepsy) requiring PSG should also be
excluded. A minimum of airflow, respiratory effort, and oxygen saturation
must be measured (type 3 PM). If a PM study is negative or technically
inadequate, a PSG should be ordered to avoid a false-negative PM
result.
Patient Selection for PM
• Ideally, each patient should be seen by a sleep physician before PM testing.
• If this is not possible, an evaluation can be performed with questionnaires
before or at the time of testing.
• Review of the medical record to exclude patients with co-morbidities that
degrade PM accuracy is also important. these include patients with severe
pulmonary disease, neuromuscular disease, or congestive heart failure
(CHF).
• The rationale is that such patients may exhibit hypoventilation without
discrete respiratory events or Cheyne-Stokes breathing (common in severe
systolic CHF).
• However, one can argue that if PM devices use the same sensors that are
used for PSG (the recommendation), then PM and PSG should have similar
ability to detect central apnea, Cheyne-Stokes breathing, or hypoventilation
(manifested by a low SaO2 without discrete events).
• The counter argument is that such patients will likely need a PSG PAP
titration. In this case, a split-sleep study may be more cost effective than PM
followed by a PSG titration. PM also would miss arrhythmias because pulse
rate obtained from the oximeter rather than an ECG tracing is usually
recorded.
Recommended PM Methodology
•
•
•
•
•
•
•
•
•
•
The PMGs recommend monitoring at a minimum airflow ,respiratory effort, and
SaO2 .
The recommended sensors for PM are the same as those recommended for
PSG in the AASM scoring manual.
Note that CMS uses the metric RDI rather than AHI for PM [(apneas +
hypopneas)/monitoring time]. RDI usually means AHI + RERA index.
Adequately trained personnel should either place the monitoring equipment on
the patient or train them on the application of the sensors. This is essential to
avoid a high percentage of technically inadequate studies.
The PM data must be viewed in the raw form, and if automated scoring is used, it
must be edited for accuracy.
A physician must look at the raw data as well as the data summary before
making an interpretation.
It was recommended that PSG be interpreted by a sleep physician or a
physician associated with an accredited sleep center.
For quality assurance, standard operating procedures for the PM process must
exist.
To verify adequate scoring, interrater reliability on scoring of PM studies should
be performed on a routine basis and documented.
If PM is inadequate technically or if the study results are negative in a patient
with a high pretest probability of having OSA, an attended PSG should be
performed.
Recommended PM Methodology
• Numerous devices are available for PM. Devices having
more sensors provide more information but are more
difficult for patients to place.
• It is always a trade off between information and complexity
of sensor application.
• Respiratory effort is monitored by chest and abdominal
respiratory inductance plethysmography (RIP) bands.
• An oximetry channel and derived heart rate are also
recorded as well as body position., actigraphy to enable
elimination of periods of wake/artifact from the final index
time (the monitoring time used for AHI calculation) is
available in some devices.
PM devices. A, Embletta (Embla). B, WatchPAT
100 (Itamar). C, Stardust II (Philips-Respironics). D,
ARES (Advanced Brain Monitoring).
A 60-second tracing shows an obstructive hypopnea recorded with a type
3 PM device (PDX by Philips-Respironics). An oronasal thermal device
(NO flow) and nasal pressure (N Press), chest and abdominal respiratory
inductance plethysmography (RIP) bands, snoring, oximetry, pulse rate
(from the oximeter), and body position (S = supine) are recorded. SpO2 =
pulse oximetry.
Tracing from a type 3 PM device (Embletta ) shows nasal pressure,
oronasal thermal flow, chest and abdominal respiratory inductance
plethysmography (RIP) channels, oximetry (SpO2), and pulse. The
illustrated event is a central apnea of the Cheyne-Stokes type. This was an
unexpected finding. The patient had no history of known congestive heart
failure but did report nocturnal dyspnea.
In this tracing from an Embletta, the nasal flow channel is inadequate
(likely nasal cannula dislodged or occluded) but the derived XFlow
(derivative of thorax and abdomen RIP bands) provides a reasonable
estimate of flow. SpO2 = pulse oximetry.
Peripheral Arterial Tonometry
•
•
•
•
•
•
•
Unique PM devices that detect respiratory events by recording changes
in sympathetic tone (rather than airflow) using peripheral arterial
tonometry (PAT) are also available.
Devices using this technology (Watch PAT 100, 200, Itamar Medical) are
worn on the wrist.
The devices have two probes—a PAT probe and an oximetry probe—
worn on separate digits.
The PAT signal is a measure of the blood volume in the digit.
Increases in sympathetic tone stimulate alpha receptors on the digital
blood vessels, causing constriction.
A reduction in blood flow to the digits decreases the finger tip volume and
the PAT signal.
Because surges in sympathetic tone follow respiratory event termination,
the combination of a decrease in PAT signal, a fall in pulse oximetry
(SpO2) followed by an increase, and an increase in heart rate allows
determination of respiratory events.
Watch PAT 100
Watch PAT 100
• The device has a built-in actigraphy to help with estimation of an
appropriate index time (used to compute an event index).
• Recently, the combination of actigraphy and the PAT signal has
been used to determine estimates of wake, non–rapid eye
movement (NREM) sleep, and REM sleep because the sympathetic
tone characteristics of these sleep stages differ.
• Newer models also have a body position sensor and a snore
sensor.
• The device has been validated with several studies.
• The device cannot be used in patients on alpha blockers (e.g.,
terazosin) and with patients in atrial fibrillation.
• Another downside is that the PAT probes are relatively expensive.
• The national carrier determination for HST recognizes “3 channels
of monitoring including PAT,oximetry, and actigraphy” as a valid PM
method.
Practical Considerations in PM
• Devices with more sensors provide more information but are more difficult to
apply.
• The software should provide accurate autoscoring to minimize the amount of
event editing required.
• If the software is similar to that used for PSG, this may be an advantage to
reduce training costs.
• The durability of the device and cost of expendables (e.g., nasal cannula)
should be considered.
• PM devices can be placed in the sleep center or in the home by a
technologist. Alternatively, the patient can be trained on the device and apply
the device himself or herself at home.
• PM devices are typically either returned to the sleep center the following day
or mailed if patients live a distance from the center.
• Device loss can be a major expense.
• The application of the PM device by a technologist in private homes is
expensive and has safety issues. Therefore, having patients come to the
sleep center is recommended if possible.
• The more complex the PM device, the less likely patients can apply it
themselves.
.
Practical Considerations in PM
• Adequate training is essential..
• Common problems include dislodgment of either the nasal
cannula or the oximeter probe as well as pulling leads out of
the PM device during body movement
• It is also useful to have patients complete a brief sleep diary
to record their estimate of how long they slept and if the
night of sleep was fairly typical.
• An occasional patient will sleep very poorly with the device
attached. If minimal sleep is recorded, a false-negative
study is likely.
• Devices that can record more than one night provide another
monitoring night opportunity and may also reduce the
influence of night-to-night variability.
Integration of PM into the Overall Patient Care Algorithm
• Diagnosis of OSA using PM is only the first part of the process if
the study is positive.
• It is expected that in populations with a high probability of OSA, a
high percentage of PM tests will be positive.
• If PAP is chosen for treatment, there are several alternative
pathways to proceed .
• The standard approach would be to perform a PSG PAP titration
and subsequent PAP treatment.
• Patients could also use an auto-PAP (autotitrating, positive airway
pressure [APAP]) device at home for 3 days or more and
information obtained could be used to select a pressure for chronic
CPAP treatment.
• Commonly, the 90th or 95th percentile pressure is chosen for
treatment with a fixed pressure (CPAP).
• A third possible approach is starting CPAP at a pressure derived
using prediction equations with subsequent adjustment based on
oximetry, symptoms, or machine estimates of residual AHI.
• Finally, simply treating the patient with an auto-PAP (autoadjusting)
eliminates the need for titration.
Overall Approach to Using PSG and PM
• A clinical evaluation determines whether there is a high
probability of OSA, if other sleep disorders are present, or
whether complicating issues are present that will likely
require a PSG titration.
• Patients with a high probability of OSA undergo PM, and if
OSA is diagnosed, they can have APAP treatment or APAP
titration followed by CPAP treatment.
• If the PM is negative, a PSG can be performed.
• If other sleep disorders are suspected or there are
complicating factors such as CHF (Cheyne-Stokes possible),
narcotics, obesity hypoventilation possible (supplemental
oxygen or massive obesity), a PSG with split study if
indicated is performed.
• Although PM can often diagnose Cheyne-Stokes breathing,
a PSG titration will likely be needed.
ACTIGRAPHY
• Actigraphy utilizes a portable device (the actigraph) usually worn on the
wrist that records movement over an extended period of time .
• Sleep-wake patterns are estimated from the pattern of movement.
• Software is available to estimate TST and wake time from the data.
• The sleep/wake pattern of actigraph data is extremely valuable in
documenting patterns of sleep and wake.
• Actigraphy is still not reimbursed by most insurance plans or Medicare.
• Studies of actigraphy have usually compared the results with that of PSG
or sleep logs. However, actigraphy does not measure sleep (EEG) or the
subjective experience of sleep (sleep logs).
Actigraphy. The dark black represents the amount of activity (when
present = wake). The light blue represents periods of low activity
identified as sleep. The light gray represents ambient light detection.
Indications
• Actigraphy provides a fairly accurate estimate of sleep
patterns in normal, healthy adult populations and in patients
suspected of certain sleep disorders.
• Actigraphy is indicated to assist in the evaluation of patients
with advanced sleep phase syndrome (ASPS), delayed
sleep phase syndrome (DSPS), shift work disorder, in the
evaluation of patients suspected of jet lag disorder and non–
24-hour sleep-wake syndrome (including that associated
with blindness).
• Actigraphy is also useful for documenting the response to
treatment in circadian disorders.
• When PSG is not available, actigraphy is indicated to
estimate TST in patients with OSA.
Actigraphy in OSA
• If the number of events detected by PM was divided by a better estimate of
TST than the monitoring time, this might improve the estimate of the AHI.
• Patients with OSA by the very nature of the disorder have periodic
movements at the end of respiratory events.
• A special actigraphic system optimized for use in OSA patients was
compared with PSG on an epoch-by-epoch basis. The overall sensitivity
and specificity to identify sleep were 89% and 69%, respectively. The
agreement ranged from 86% in the normal subjects to 86%, 84%, and 80%
in the patients with mild, moderate, and severe OSA, respectively.
• There was a tight agreement between the mean values of actigraphy and
PSG in determining mean sleep efficiency, TST, and sleep latency.
Whereas for most individuals, the difference between the PSG and the
actigraphy was relatively small, for some, there was a substantial
disagreement.
• Another study did not find good agreement between actigraphic and PSG
estimates of TST.
• Actigraphy combined with respiratory monitoring are a better estimate of
the AHI than use of time in bed. The utility of actigraphy with PM remains to
be determined. However, it certainly allows elimination of large portions of
the night in which wake is obvious.
Patients with Hypersomnia
• Practice parameters on use of PSG/MSLT to
diagnose narcolepsy stated that sleep logs may be
obtained for 1 week before to assess sleep-wake
schedules.
• Not all patients can keep accurate sleep logs.
Therefore, actigraphy could be potentially useful to
document sleep patterns before MSLT .
• Current practice parameters for actigraphy do not
list evaluation of sleep before MSLT as an
indication.
Actigraphy in Insomnia
• The practice parameters for actigraphy did not recommend testing as a
routine evaluation of patients with insomnia.
• However, actigraphy was said to be “useful” in documenting sleep-wake
patterns.
• Therefore, documenting sleep patterns rather than the absolute amount
of sleep and wake is likely the best use of the actigraph.
• In insomnia patients, periods of low activity in which patients lay quietly
in bed but are awake may be scored as sleep by actigraphy software.
• When performing actigraphy, it is essential to require patients to
complete a sleep log (lights off, lights on, out of bed, actigraph off for
shower, and so on, estimated TST; and sleep latency).
• This information enables a correct interpretation of actigraphic tracings.
For example, total absence of movement usually indicates that the
actigraph has been removed (shower, swimming), but here patient logs
are essential to verify device removal.
Actigraphy in Insomnia
• Many modern wrist actigraph devices also provide the capability of
recording the amount of ambient light.
• This is useful in patients with circadian rhythm sleep disorders in whom
light exposure may either exacerbate or improve the underlying disorder.
• Some actigraphy devices also have event markers that the patient can
press when getting in bed, trying to sleep, or after awakening in the
morning.
• Sleep logs and actigraphy provides complementary information.
• Both actigraphy and sleep logs underestimated TST and overestimated
wake time.
• Actigraphy underestimated the sleep latency whereas sleep logs
overestimated the sleep latency.
• Actigraphy was found to be more sensitive at detecting treatment effects
than sleep logs. In contrast, different results was found in a group of
older adults treated for primary insomnia (expected to have more wake
time). Actigraphy slightly overestimated TST and underestimated wake.

similar documents