Unattended Continuous Positive Airway Pressure ...

Unattended Continuous Positive Airway Pressure ntration Clinical Relevance and Cardiorespiratory Hazards of the Method JANOS JUHASZ, JOACHIM SCHILLEN, ANJA URBIGKEIT, THOMAS PLOCH, THOMAS PENZEL, and JORG HERMAN PETER Zentrum fOr Innere Medizin, Philipps-Universitat, Medizinische Poliklinik, Schlafmedizinisches labor, Marburg, Germany; and Koranyi National Institute for Pulmonary Medicine, Respiratory Care Unit, Budapest, Hungary

The high cost of in-laboratory sleep studies and the insuffidency of available nights in most sleep laboratories has prompted clinical trials in an attempt to find an effective and safe method for continuous positive airway pressu~e (CPAP) titration in the unsupervised home environment. Earlier publications focused on the effectiveness of this new method but did not indicate the potential hazard and the selection criteria of patients. We accomplished a prospective study to evaluate a machine-controlled CPAP titration method in an unselected group of 21 patients with obstructive sleep apnea (OSA) requiring CPAPtherapy. Uttrftended CPAPtitrations were completed in the sleep laboratory, followed by conventional CPAP titration night. The CPAP levels assessed through automatic titrations showed strong agreement with those from the control nights.fjfteen patients had no pathologic cardiorespiratory events during machine-controlled titration. Undesirable cardiorespiratory complications developed in six patients with underlying cardiokspiratory disorder. Two patients had to stop the procedure. Two patients needed a reset of the actual pressure to baseline because of .f~ntral apneas and one patient because of high-grade arrhythmia. The machine-controlled CPAP titration enabled reliable assessment of an effective pressure and appeared to be safe in selected patients with OSA. However, because the method caused significant adverse effects in some patients, we recommend that unattended CPAP titration not be attempted in patients with underlying cardiorespiratory disease. Juh'" J, Schillen J, Urblgkelt A, Ploch T, penzel T, Peter JH. Unattended continuous positive airway pressure tltr. tlon: dlnlcal relevance and cardiorespiratory hazards of the method. AM

Nasal continuous positive airway pressure (CPAP) has been demonstrated to be highly effective in the elimination of nocturnal upper airway occlusion leading to obstructive sleep apnea (OSA) (1-5). Because the positive airway pressure may induce adverse cardiorespiratory effects in high-risk patients and each patient needs an individual adjustment of the CPAP therapy, the titration of the effective CPAP should be carried out in a sleep laboratory under continuous medical supervision of cardiorespiratory functions (6-9). These requirements, combined with the high costs of polysomnography and the development of reliable ambulatory recording systems, have prompted clinical attempts to find an effective and safe method for CPAP titration in the unsupervised home environment of the patients. Miles (10) performed the first CPAP titrations under these circumstances by means of pulse oximeter safeguarded by a low O 2 alarm and using different airway pressure valves. Since 1991 they have performed unattended CPAP titrations in a selected group of patients with OSA using a new portable physiologic monitoring system (II). In a recent abstract it was shown that there was no

(Received in original form February 24, 1995 and in revised form November27, 1995)

Correspondence and requests for reprints should be addressed to Janos Juhasz, M.D., Zentrum fOr Innere Medizin, Philipps-Universitat, Medizinische Poliklinik, Schafmedizinisches Labor, Baldingerstr. 1. 35033 Marburg, Germany. Am J Respir Crlt Care Med

Vol 154. pp 359-365, 1996

J RESPI. cur CARE MED 1996;154:J59-45.

significant difference between the pressure determined by unattended titration and that assessed during CPAP titration by a technologist (12). A more recent paper reports on the diagnosis and treatment of patients with OSA in an entirely outpatient setting using a four-channel ambulatory recording device (13). Although these first published results are encouraging, the following questions still remain to be answered: (1) Is the method suitable to estimate a relevant and effective airway pressure for therapy in all groups of patients? (2) What are the underlying diseases that have to be considered as high risk or contraindication for unattended CPAP titration? (3) Do the induced cyclical changes of the airway pressure disturb the quality of sleep? To evaluate the relevance of unattended CPAP titration and to assess its potential hazard in case of underlying cardiorespiratory diseases,wecompleted a clinical study in unselected patients with obstructive sleep-related breathing disorders (SRBD). METHODS Patients Twenty-onepatients wererandomly selected from 162patients diagnosed as having OSA and judged to be candidates for nasal CPAP therapy between November 1992 and March 1993. All patients who were approached agreed when asked to participate in the evaluation of a new method of CPAP titration. All underwent a machine-controlled CPAP titration on one night in the sleep laboratory (simulating an unattended titration) and a conventional technician CPAP titration on the next night. Their age (mean ± SD) was 53.9 ± 9.1 yr (range, 37 to 72), body mass

360

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

VOL 154

TABLE 1

TABLE 2

UNDERLYING DISEASES IN THE PATIENTS WHO UNDERWENT UNATTENDED (PAP CALIBRATION

SENSOR CONFIGURATION OF THE VITALOG HMS-5000 MONITOR FOR UNATTENDED CPAP CALIBRATION Biosignals

Cardiopulmonary 14 ph

Arterial hypertension COPD Right heart failure Cardiac arrhyth mias Congestive heart failure Pulmonary arterial hypertension

7 pts 4 pts 5 pts 2 pts 1 pt Others 6 pts 6 pts 1 pt 1 pt 1 pt 1 pt

Hyperlipoproteinemia Hyperuricemia Diabetes mellitus Head/brain trauma Compensated renal failure Nasal septal deviation

Breathing

Circulation

Sleep-wake

Other Definition of obbreviation: pts

= patients.

Oronasal airflow Snoring Thoracic breathing Abdominal breathing Oxygen saturation ECG Heart rate

EOG EMC (submental) Body movements Leg movements Position Tidal volume Paradoxity Mask pressure Airflow in mask

1996

Biosensors Thermocouple Microphone RIp· RIP Pulse oximetry Chest leads (V2) RR detector Surface electrodes Surface electrodes Piezoresistive accelerometer Piezoresistive accelerometer Piezoresistjve accelerometer Derived digitally from Thoracic/abdominal breathing BiPAP sensor BiPAP sensor

• RIP = respiratory inductive plethysmography.

index,.30.67 ± 5.34 kg/rrr' (range, 23.15to 44.29). They had established OSA of body mass index, 30.67± 5.34 kg/m' (range, 23.15 to 44.29). They had established OSA of varying severity before entry into the study, with a respiratory disturbance index (RDI) of 52.1 ± 29.1 (range, 5 to 131). The demonstration of the SRBD and the need for nasal CPAP therapy was accomplished by visual scoring diagnostic studies obtained in the sleep laboratory. The presence of underlying disease was based on previous diagnoses obtained from the medical records (Table 1). Patients with underlying cardiorespiratory diseases were considered to be at high risk for complications of CPAP titration.

ysmography, pulse oximetry, laryngeal sounds by microphone, EOG, tibial electromyogram (EMG), and time. Total sleep time and time to fall asleep (characterizing sleep latency) were estimated after scoring sleep/wake periods on the Vitalog recording based on EMG, EOG, actigraphy, leg movements, heart rate, oximetry, and breathing pattern (because electroencephalogram (EEG) was not monitored).

Unattended CPAP Titration

Arrhythmia Monitoring

The machine-controlled CPAP (VITPAP) titrations werecompleted using the Vitalog HMS-5000 physiologic monitor (Vitalog Monitoring Inc., Redwood City, CA). The data (cardiorespiratory, airway pressure, and sleep) were recorded digitally in the Vitalog monitor. In addition, realtime data were recorded simultaneously on an IBM compatible computer. The sensor configuration is shown in Table 2. The vitalog HMS-5000 system is configured to use the Respironics'" BiPAP-STD equipment (Respironics Inc., Murrysville, USA) for nasal ventilation through a controlline that also receives the pressure and airflow signals from the BiPAP sensors. The BiPAP-STDventilator was set on IPAP-only mode (14).The procedure does not attempt to automatically adjust the CPAP pressure to the correct valueduring the titration but systematicallyvaries the CPAP pressure delivered following a preset algorithm. Pressure is raised and lowered from a user-selected baseline in 3D-min cycles, each of which spans 6 em H 2 0 , with steps of I em H 2 0 , each lasting 5 min. The presence of oxygen desaturations below 85% during a cycle is used to raise the baseline by one step. The absence of desaturations below 85% for 2 h causes the baseline to be lowered by one step. The maximal pressure level (15em H 2 0 ) provided in the algorithm prevents dangerously high airway pressure. A reset button on the device is provided to allow the patient to lower the airway pressure to the preselected baseline at the beginning of sleep. This reset button was also available to the technician to provide a security measure for high-risk patients during the VITPAP titration setting. The CPAP level was estimated by a physician after visually reviewing and scoring the Vitalog recording the next morning. We considered the pressure to be effective if snoring, apnea, hypopnea, and consecutive drops in Sao, were abolished (Figure I). The effective airway pressure was titrated conventionally by the laboratory staff, who were unaware of the pressure level estimated after the VITPAP titration. The pressure values chosen during the conventional titration were considered as "adequate CPAP" for long-term therapy.

Simultaneous Holter EKG monitoring (Oxford Medical Ltd., UK) was performed during the diagnostic recording and the machine-controlled titration to record cardiac arrhythmia and assess Lawn classification. Real-time electrocardiogram (ECG) recording through Vitalog monitor on computer completed the measurements during the VITPAP-titration.

Polysomnography Wecompleted simultaneous cardiorespiratory polygraphy during the VITPAP titrations and during the conventional technician-controlled CPAP titrations in the consecutive night. We recorded mask pressure and thoracic and abdominal breathing movements by respiratory inductive pleth-

Data Analysis We compared the airway pressure values estimated after machinecontrolled titration to those assessed during conventional titration. Differences of these pressure values werecalculated. The differences were evaluated using t test (STATGRAPHICS Version 5.0, Statistical Graphics Corporation, USA), and their relationship was analyzed by product-moment correlation. Clinical analysis included investigation of the patients' tolerance (Vitalog off temporarily or finally, difficulties in falling asleep), the method's safety (undesirable baseline elevations, need for reset), and the patients' risk (undesirable central apneas with falls in Sao, below 851170, secondary alveolar hypoventilation with sustained O 2 desaturations, cardiac arrhythmia) during the VITPAP titration.

RESULTS Patients' Tolerance

Fifteen out of 21 patients (68010) experienced no difficulty with the cyclical changes in the airway pressure, showing neither discomfort nor cardiac or respiratory complications during VITPAP titration. Four patients showed mild discomfort from the pressure changes, manifested by their need to reset the CPAP pressure to the baseline. In two of these there was a prolonged delay in sleep onset (36 and 92 min), necessitating temporarily discontinuing VITPAP titration. However, these patients were able to fall asleep under constant CPAP (4 and 2 em H 2 0 ), after which VITPAP titration was successfully resumed by the technician. Two other patients required final discontinuation of the VITPAP titration due to significant central apneas with O 2 desaturations below 85010, causing progressive rise in baseline pressure or sustained hypoxemia. In the 17 patients who were

361

Juhasz, Schillen, Urbigkeit, et 01.: Unattended CPAP Titration

POX SPIIlO

16 12

PR£S:~;;;~~~~;;~~~~~~~~~~~

OIl 1 EEG«

9080

l5070~IItT.~""""'~"""'rWJ~ ...i.n~_~~AMr-N1ll~""rnr~~~rtwAllinnJ:"~""'''' 3O15O

£~~"'-----"-------""'---""'''''--------------POStT IOI>...,.. -r"' ...,. ...., _ 00:35

01:32

Figure 1. The printout of a Vitalog registration presents three VITPAP titration cycles. At levels of baseline (p =4 cm H20) and/or ineffective airway pressure, pathologic breathing events (tidal volume) occur with consecutive 02 desaturations (Oxi) . At an effective airway pressure (arrows) or above, breathing becomes regular. Paradoxity (POX) represents asynchronous or paradoxic breathing pattern due to partial or complete upper airway obstruction. (Pressure trac ings [PRES] : CPAP =thick line; baseline presure = thin line.)

able to fall asleep without discontinuing VITPAP titrations, the average latency was 12 ± 9.5 min (range. I to 32). The estimated total sleep time was 373.5 ± 49.2 min (range. 259 to 438) in 19 patients who completed the VITPAP titration. Effectiveness of the Method

In all 19 patients who completed the VITPAP titration night. an effective CPAP pressure could be chosen. This pressure (8.36 ± 2.36; range. 4 to 16 cmH 20 ) was highly correlated (r = 0.90) to that chosen during the conventional techniciancontrolled titration (8.15 ± 2.38; range. 5 to 16 em H 2 0 ), with a mean difference of 0.21 ± 1.08 em H 20 (Figure 2). There were no differences in the pressure levels in 12 patients (63%). In four patients the effective CPAP pressure chosen based on the technician titration was lower by 2 em H 2 0 than that chosen from the VITPAP titration. In three patients the technician titration resulted in a pressure higher by 1. 1, and 2 em H 2 0 . These differ ences are both clinically and statistically insignificant (p = 0.4). Safety of the Method and Patients' Risk

The preselected baseline pressure (p = 4 em H 2 0 ) did not change throughout the VITPAP titration night in six patients. Altogether. 72 baseline elevations (BLE) appeared during the VITPAP titration in 15 patients. Thirty-eight elevations resulted during desaturations that were caused by obstructive apneas, and one more occurred during secondary hypoventilation with a Sa02 < 85% related to obstructive snoring. Thirty-three further baseline elevations arose during desaturations related to other pathologic conditions: 21 BLE in five patients due to cen-

tral apneas, 11 BLE in five patients because of drops in Sao, below 85010 due to movement artifacts, and one BLE related to rapid eye movement (REM)-hypoventilation with Sa0 2 < 85010. Twelve patients altogether needed 16 resets of the actual CPAP level to the baseline. Eleven resets were required in seven patients to enhance their falling asleep after waking and finding the actual pressure disturbingly high. Two resets were needed in

3 -

2

1

o ·1 •

·2

/

-3 -...- ---,-,--. -1

2

3

4

5

--.,.--6

- - r - - - - . , - -7

8

9

10

11 12

13

14

----.,.15

18 17 18

19

PATIENTS

meandilIerence = 0.21 :!:1 .00 anH20 (p=0.4n.s.) t, Pa'N' = CPAPVlT · CPAPPSG

Figure2. The figure shows the differences in pressure levels (~PCPAP) obtained from VITPAP titration versus conventional titration nights for each patient. (CPAPVIT: effective pressure assessed through VITPAP titration. CPAPPSG: effective pressure assessed through conventional titration.)

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 154 1996

Figure 3. Oximetry readings of simultaneous Mesam IV recordings in 10 patients. The plot shows the minimal Sao2 values from the diagnostic night compared with those from VITPAP titration night. The scatter of the values decreased from diagnostic night to the VITPAP titration niqht considerably.

two patients because of artifact increase in the pressure level. Two resets were necessary due to central apneas with desaturation below 850/0, and one more because of an increased number of ventricular ectopic beats. These latter three resets were required to avoid hazardous cardiorespiratory consequences, two of them in high-risk patients. All resets were initiated by the supervising staff. Ten patients needed no reset at all. Simultaneous Mesam IV recordings during the diagnostic night and the VITPAP titration showed significant (p < 0.05) increase in the minimal Sao, values from 73.88 ± 12.8% (range, 42 to 85) to 81.77 ± 3.67% (range, 77 to 85) in 10 patients (Figure 3). Lawn classification did not change due to unattended CPAP titration in 16 patients (13 with Lawn 0 and 3 with Lawn 1) compared with diagnostic studies. Lawn classification improved in 4 patients (Lown 1 to Lawn 0, Lawn 3a to Lawn 2, Lawn 4a to o and Lawn 3a). One patient presented significantly increased number of ventricular ectopic beats (18-20/min) at pressure values above 10 em H,O (Lawn 4a changed to Lawn 4b). The VITPAP titration did not provoke pathologic respiratory sequelae in 15 patients. Five patients presented central apneas with Sao~ > 85%. The total number of these central apneas was low in three patients (n < 10) and higher in two patients (n > 50). Central apneas with drops in Sao, below 85%, leading to ~aseline pressure elevation, appeared in three patients. Five panents developed secondary alveolar hypoventilation due to increasing CPAP, two patients presented Sao, values above 85%, and three patients had desaturations below 85%, leading to consecutive baseline elevation. 1\vo patients showed a combination of both central apneas and alveolar hypoventilation. Cardiorespiratory sequelae (central apneas with Sao, < 85% and/or hypoventilation) requiring intervention (reset or final discontinuation of VITPAP) developed altogether in six patients.

DISCUSSION Revie~ing the literature related to unattended CPAP titration, the Vitalog-controlled CPAP titration represents the first really unattended technique that automatically adjusts a pressure range at which the pathologic breathing events (snoring, apnea, hypopnea) and consecutive drops in Sao, below 85% are abolished (14, 15). Recently, many clinicians have used a portable recording device to follow their patients during a CPAP adjustment at home

but have not actually controlled the airway pressure titration. Coppola and Lawee (13) used a four-channel portable device to record the patients' breathing during CPAP adjustment at home. In their selected study population, the authors did not experience any adverse cardiorespiratory sequelae due to excessive CPAP levels. However, the airway pressures did not exceed 10em H,O in this study. Waldhorn and Wood (16) reported on CPAP titration at home requiring a technician to be present throughout the night to adjust the pressure level reviewing the analog registration of the same four-channel equipment. Another method providing a potential option for unattended CPAP titration has been published by Berthon-Jones (17). Other clinical trials have made efforts to deliver a self-regulating system (auto-CPAP) that fits dynamically the lowest airway pressure to the patient, dependent on their individual requirements, varying both intranightlyand internightly [18-22). All papers report on the feasibility and effectiveness of these CPAP titration methods. However, patients with chronic obstructive pulmonary disease (COPD) and/or cardiac disorders were excluded in every study. The present study is the first to examine the effects of unattended CPAP titration on cardiorespiratory function in an un selected group of-patients and highlights a potential hazard of this procedure in those with COPD and cardiac disease. Clinical Relevance

Our results, like those of Miles and coworkers (11, 15)and Guilleminault and colleagues (12) confirm that the effective CPAP identified by the unattended VITPAP algorithm does not differ significantly from the chosen during technician titrations. The same pressure was chosen in 12 patients (63% ofthecompleted VITPAP studies) and differences of only 1-2 em H,O were seen in the remaining seven patients. This suggests that VITPAP titrations are as reliable a method to assess effective CPAP as the conventional technique in selected patients. What small differences in pressure were found by the two techniques may have several explanations. The immediate effects of the CPAP on sleep architecture have been reported in the literature as a significant reduction in stage I and II non-rapid eye movement (NREM) sleep and a marked increase in slow-wave sleep (stage III and l.v NREM) as well as REM sleep (23, 24). This means that patients could have had a quite different sleep architecture during the conventional titration night than they did during the VITPAP night. Thus, the different sleep structure due to VITPAP titration could have modified the required effective CPAP during the second titration. Another explanation for lower CPAP levels during the control night may be the patients' individual adaptation to positive airway pressures yet within the VITPAP titration night. Our observation revealed that some patients required less positive pressure by the end of the night to abolish pathologic breathing events than they did during the first half of the titration. These observations lend support to the hypothesis of an adaptation mechanism that might have been caused either by hysteresis effect of the CPAP or by possible activation ~f the mechanoreceptors in the upper airways through the positive pressure. However, the possible effect of CPAP therapy through activation of the mechanoreceptors in the upper airway is still controversial (2, 25). Clinical studies with auto-CPAP have shown that the requirement for an optimal pressure fluctuates across several nights (21). This may result in different effective pressure. Last, but not least, the interpersonal subjectivity of the staff in the sleep laboratory could also explain the differences in the assessment of the effective CPAP level. Clinical Tolerance

Although EEG parameters were not measured according to international criteria in this study, we could draw some conclu-

363

Juhasz, Schillen, Urbigkeit, et al.: Unattended CPAP Titration

sions on the sleep quality based on the registrations of non-EEG parameters and video observation. With respect to the estimated total sleep times, we did not see relevant diminution related to Vitalog titration compared with our experiences during conventional titration. Some patients were considerably affected by increasing pressure and could not fall asleep within the given time-delay during VITPAP titration. This was also reflected in the relative high number of patients (n = 7) who needed a reset after waking up because the actual pressure was disturbing. It is noticeable that the reset option during unattended titration was provided for such instances. Patients having difficulties in falling asleep had an RDI (50.43 ± 25.65) similar to the other patients (52.1 ± 29.1). Thus, a distinctive severity of the OSA was not a factor for difficulties in falling asleep with positive pressure. We explained this disturbing effect as sleep onset with unfamiliar circumstances. safety of the Method: Baseline Elevations and Resets

The option of baseline elevation in the algorithm provided automatic adjustment of a pressure range without O2 desaturations. Baseline elevations related to obstructive events (apneas, hypopneas, snoring) were recognized as clinically desirable. Baseline elevations related to other pathologic events, such as central apneas or O 2 desaturations due to movement artifacts, were considered to be clinically undesirable. Krieger and associates (6) reported dangerous sustained hypoxemia associated with CPAP treatment. Hemodynamic measurements by Podszus and colleagues (26) showed a decline in cardiac output with incremental positive airway pressure in patients with OSA. This hazard may be potentiated through underlying cardiorespiratory diseases. Thus, patients with cardiorespiratory diseases are at high risk during unattended CPAP titration. That is why baseline and airway pressure elevations due to central apneas and sustained secondary O 2 desaturation as well as artifact desaturations were regarded as potentially dangerous in the course of VITPAP titration. Thirty-three out of 72 baseline elevations (46010) appeared in our study during such events. These resulted in undesirable sequelae, requiring intervention in six patients. Twentytwo baseline elevations related to central breathing events were observed in six patients. Fifteen out of these undesirable 22 elevations occurred in two patients (8 and 7), who required discontinuation of the VITPAP titration totally. The relatively high number of baseline elevations due to drops in Sao 2 below 85010 related to movement artifacts (11 in 5 patients) was especially surprising. We found it unsuccessful as well as potentially dangerous that desaturations below 85010 are considered the only clinical indicator for the control of the automatic baseline elevations. Thus, movement artifacts resulting in significant drops of Sao, may lead to undesirable baseline (airway pressure)elevations with consecutive adverse cardiorespiratory sequelae (central apneas, secondary alveolar hypoventilation, increased number of ventricular ectopic beats) in high-risk patients. It is advisable to control the baseline elevations by more parameters simultaneously, such as airflow (BiPAP sensor) and breathing efforts (paradoxity-a relevant indicator for upper airway obstruction). This would allow an appropriate recognition of and differentiation between central and/or obstructive breathing events and ensure clinically more reliable airway pressure adjustment. The deteriorated flow as a reliable indicator for obstructive events has been proved by reports on auto-CPAP (17) and by Condos and colleagues (27) recently. Effects on Cardiorespiratory Function Central apneas may occur at excessive pressure values during CPAP titration that can be abolished by decreasing the applied CPAP. The underlying pathomechanism to develop such apnea

during CPAP is not yet clear. Skatrud and Dempsey (28) reported that passive hyperventilation, which induced reduction in endtidal Paco2 , caused central apneas during NREM sleep. We suspect similar mechanisms that may explain central apneas in patients receiving CPAP. In this manner, the appearance of central apneas should be regarded as an indicator for inappropriate positive pressure adjustment that functionally burdens the patient. This may be aggravated in patients with underlying cardiorespiratory diseases. The association of central apneas with oxygen desaturations was considered as a pathologic event in our study. Central apneas without desaturations below 85010 did not appear to be dangerous during VITPAP titration because they did not result in baseline (airway pressure) elevation. One patient with congestive heart failure presented multiple central apneas during VITPAP at a baseline of 5 em H 2 0 . Central apneas were associated with Sao 2 above 85010 without baseline elevations. However, the duration of central apneas increased in relation to the increase in airway pressure. Thus, this patient was considered to be at high risk during VITPAP titration despite the lack of severe Oz desaturation. Central apneas with desaturations below 850,10 caused clinically undesirable baseline and airway pressure elevations in two patients. The central apneas leading to permanent increase in the airway pressure did not result in direct adverse cardiorespiratory sequelae in one patient without cardiac or respiratory failure. However, he woke up repeatedly due to excessivepositive pressure and could not complete the VITPAP titration. Conventional CPAP titration was achieved in the same night. Another patient with cardiorespiratory failure related to coronary heart disease presented central apneas and oxygen desaturations below 85010 with undesirable pressure elevations and developed alveolar hypoventilation due to excessive increase in airway pressure (Figure 4). He needed to discontinue the VITPAP titration finally and technician-controlled titration could be completed. Cardiac arrhythmia does not generally appear during correctly adjusted CPAP therapy. On the contrary, arrhythmia related to sleep apneas is abolished through therapy (29). However,Meurice and coworkers (30) reported atrial arrhythmia as a complication of nasal CPAP. Four patients entering our study presented highgrade ventricular arrhythmia (Lown 3a 'and 4a). Three of them showed an improvement in Lown classification during VITPAP titration night. Lown classification worsened in one patient with severe right heart failure (4a to 4b). At baseline pressure he had 8-10 ventricular ectopic beats/min. The number of ectopic beats increased with airway pressure elevations. Wefound 18-20 polytope ectopic beats/min as well as couplets and quadrigeminia at CPAP levels above 10em H 2 0 . After reset of the airway pressure to the baseline, the number of ectopic beats diminished immediately to 8-10. The patient did not present excessive hypoxemia during the period of multiple polytope ectopic beats, so it may not have been the cause of the worsening arrhythmia. The increased end-diastolic overdistension of the right ventricle through the positive intrapulmonary pressure might have contributed to this condition (31). Conclusions

Our results, in agreement with other studies, show that machinecontrolled CPAP titration represents a method for unattended CPAP titration that may provide a technical alternative to conventional CPAP titration. The method is suitable to assess a clinically relevant and effective airway pressure for therapy and has been shown to be safe for a selected group of patients with OSA and without underlying cardiorespiratory disorders. However, in patients with certain disorders in addition to OSA (e.g.,cardiac arrhythmia, congestive heart failure, and functional disturbances of respiratory control during sleep), our data demonstrate that

364

--------------- -

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 154 1996

AIR

FUlIl f'llJOJO ~IIC:!C::I!~~~::::!l:C~!:::!:::k::~l!::::~:!::=:!::!!:~=~:::c~~=::::!~~~:::!!::!:~:=:!=

!lOll U

TO

1 l

DU Art

l [

11 019 I EE8'"

H

(12

M

12T T[

0:51

Kal. 12 IIinutn/SCl"e«'l

0:58

Figure 4. The printout shows baseline and airway pressure elevations (arrows) that are triggered by central apneas and consecutive drops in Sa02 below 85% (Oxi) during Vitalog-controlled CPAPtitration. The patient developed alveolar hypoventilation (HV) due to excessive increase in airway pressure (PRES). Of note, this patient had a diagnostic study that showed exclusively obstructive events (RDI: 49), but also had significant cardiorespiratory failure felt to be due to coronary heart disease. At baseline pressure (p = 4 cm H20) obstructive events occurred (thick bars in upper row). The central respiratory events shown on this tracing (indicated by thin bars in upper row) appeared on excessive CPAP therapy. After termination of VITPAP titration in this patient, central events disappeared, and a technologist-controlled titration was successful at establishing a therapeutic pressure of CPAP.

these conditions may be exacerbated during inappropriately high CPAP pressures that my occur during unattended titration. Because of these undesirable respiratory and/or hemodynamic consequences, we feel a supervised CPAP adjustment in the sleep laboratory is mandatory for these patients. In our study all titrations, while automatic, were under medical supervision, and these complications were immediately identified. Thus, truly unattended machine-controlled titration remains controversial in general practice. However, we feel it may serve as a reasonable alternative to conventional technician-controlled titration in selected patients or in the controlled conditions of the sleep laboratory. Acknowledgment The authors thank David M. Rapoport, M.D., for his advice and for reviewingthe manuscript Theyalsothank StimotronGmbH (Germany) for putting the Vitalog HMS-5000monitor at their disposal to complete this study.

References I. Sullivan, C. E., F. G. Issa, M. Berthon-Jones, and L. Eves. 1981.Reversal of obstructive sleep apnoea by continuous positive airway pressure aplied through the nares. Lancet 1:862-865. 2. Rapoport, D. M., S. M. Garay, and R. M. Goldring. 1983. Nasal CPAP in obstructive sleep apnea: mechanisms of action. Bull. Eur. Physiopathol. Respir. 19:616-620. 3. Strohl, K. P., N. S. Cherniack, and B. Gothe. 1986. Physiologic basis of therapy for sleep apnea: state of art. Am. Rev. Respir. Dis. 134: 791-802.

4. Rajagopal K. R., L. L. Bennett, T. A. Dillard, C. J. Tellis, and M. F. Tenholder. 1986.Overnight nasal CPAP improves hypersomnolence in obstructive sleep apnea. Chest 90:172-176. 5. Sforza E., J. Krieger, E. Weitzenblum, M. Appril, E. Lampert, and J. Ratamaharo. 1990.Long-term effects of treatment with nasal continuous positive airway pressure on daytime lung function and pulmonary hemodynamics in patients with obstructive sleep apnea. Am. Rev. Respir. Dis. 141:866-870. 6. Krieger, J., E. Weitzenblum, J. P. Monassier, C. Stoeckel, and D. Kurtz. 1983. Dangerous hypoxemia during continuous positive airway pressure treatment of obstructive sleep apnea. Lancet 2:1429-1430. 7. Becker, H., J. H. Peter, and P. von Wichert. 1992. Nasal continuous positive airway pressure. Eur. Respir. Rev. 2(Suppl. 10):400-408. 8. Deutsche Gesellschaft fiir Pneumologie, Arbeitsgruppe niichtliche Atrnungs- und Kreislauf~egulationsstorungen.·1993. Empfehlungen zur niichtlichen nasalen Beatmungstherapie bei Atmungsstorungen, Pneumologle 47:333-335.

9. Smith, P. L., D. W. Hudgel, L. G. Olson, M. Partinen, D. M. Rapoport, C. L. Rosen, F. B. Skatrud, R. E. Waldhorn, P. R. Westbrook, and T. Young. 1994. Indications and standards for use of nasal continuous positive airway pressure (CPAP) in sleep apnea syndromes: ATS statement. Am. J. Respir. Cril. Care Med. 150:1738-1745. 10. Miles, L. E. 1987.Optimization of nasal-CPAP airflow pressure by use of home oximetry recordings. Sleep Res. 16:568. II. Miles, L. E., G. D. Buschek, D. P. McClintock, S. C. Miles, L. R. Narvios, and X. Y. Wang. 1992. Development and application of two automatic nasal CPAP titration procedures for use in the unsupervised home environment (abstract). J. Sleep Res. I(Suppl. 1):150. 12. Guilleminault, C., R. Stoohs, L. Miles, J. Cattano, and E. Kulikowski.

Juhasz, Schillen, Urbigkeit, et 01.: Unattended CPAP Titration

13. 14.

15.

16. 17. 18. 19. 20. 21. 22.

1992.Unattended CPAP titration: toward a smart machine (abstract). Am. Rev. Respir. Dis. 145:A725. Coppola, M. P., and M. Lawee. 1993.Management of obstructive sleep apnea syndrome in the home: the role of portable sleep apnea recording. Chest 104:19-25. Using the HMS 4000/5000 to carry-out an automatic nasal CPAP titration procedure. In Vitacore Users Manual for Vitalog HMS-4000 (Lunch-Box") and Vitalog HMS-5000 (Pocket Polygraph") Physiological Monitoring and Actuating Systems. Version 1.0h. Vitalog Monitoring, Inc., Redwood City, CA. 69-73. Miles, L. E., G. D. Bushek, D. P. McClintock, S. C. Miles, L. R. Narvios, and Y. X. Wang. 1993. Development and application of automatic nasal CPAP titration procedures for use in the unsupervised home environment. Sleep 16(Suppl. 8):1l8-119. Waldhorn, R. E., and K. Wood. 1993. Attended home titration of nasal continuous positive airway pressure therapy for obstructive sleep apnea. Chest 104:1707-1710. Berthon-Jones, M. 1993. Feasibility of a self-setting CPAP machine. Sleep 16(Suppl. 8):120-123. Burk, J. R., E. A. Lucas, J. R. Axe, K. Behbehani, and F.-C. Yen. 1992.Auto-CPAP in the treatment of obstructive sleep apnea: a new approach. Sleep Res. 21:182. Robert, D., G. Bourdon, B. Langevin, P. Leger, and A. Guez. 1993. Usefulness and efficacy of an auto adjustable CPAP machine in the treatment of OSAS (abstract). Am. Rev. Respir. Dis. 147:A680. Sadrnoori, B. 1994. Evaluation of self adjusting nasal CP AP (DPAP) in the treatment of adult obstructive sleep apnea. Sleep Res. 23:386. McDannold, M. D., M. Berthon-Jones, N. Zaretsky, G.ftux, and M. B. Scharf. 1994. Night-to-night variability in optimal CPAP pressure using auto CAP titration in a single patient. Sleep Res. 23:453. Davies, W. L., and J. R. Stradling. 1994.Auto-CPAP, comparison with constant pressure (abstract). J. Sleep Res. 3(Suppl. 1):57.

365 23. Issa, F. G., and C. E. Sullivan. 1986. The immediate effects of nasal continuous positive airway pressure treatment on sleep pattern in patients with obstructive sleep apnea syndrome. Electroenceph. Clin. Neurophysiol. 63:10-17. 24. Aldrich, M., A. Eiser, M. Lee, and J. E. Shipley. 1989.Effects of continuous positive airway pressure on phasic events of REM sleep in patients with obstructive sleep apnea. Sleep 12:413-419. 25. Strohl, K. P., and S. Redline. 1986. Nasal CPAP therapy, upper airway muscle activation, and obstructive sleep apnea. Am. Rev. Respir. Dis. 134:555-558. 26. Podszus, T., H. Becker, J. H. Peter, and P. von Wichert. 1989. Left ventricular function under nCPAP ventilation (abstract). Eur. Respir. J. 2(Suppl. 8):785s. 27. Condos, R., R. G. Norman, I. Krishnasamy, N. Peduzzi, R. M. Goldring, and D. M. Rapoport. 1994. Flow limitation as a noninvasive assessment of residualupper-airway resistanceduring continuous positive airway pressuretherapy of obstructive sleepapnea. Am. J. Respir. Crit. Care Med. 150:475-480. 28. Skatrud, B. J., and J. A. Dempsey. 1983. Interaction of sleep state and chemical stimuli in sustaining rhythmic ventilation. J. Appl. Physiol. 55:813-822. 29. Becker, H., M. Faust, U. Kohler, and J. H. Peter. 1987. Reversibility of severe cardiac arrhythmias in sleep apnea under nasal continuous positive pressuretherapy. In J. H. Peter, T. Podszus, and P. von Wichert, editors. Sleep-related Disorders and Internal Diseases. SpringerVerlag, New York. 380-387. 30. Meurice, J. C., J. Mergy, C. Rostykus, P. Dore, J. Paquereau, and F. Patte. 1992. Atrial arrhythmia as a complication of nasal CPAP. Chest 102:640-642. 31. Schulman, D. S., J. W. Biondi, R. A. Matthay, P. G. Barash, B. L. Zaret, and R. Soufer. 1988. Effect of positive end-expiratory pressure on right ventricular performance. Am. J. Med. 84:57-67.