Group Education Sessions and Compliance With Nasal CPAP Therapy* Laura L. Likar, MD; Toni M. Panciera, RNP, RRT; Allan D. Erickson, MD; and Sharon Rounds, MD
Study objectives: To determine an effective means of improving compliance with nasal continuous positive airway pressure (CPAP) for obstructive sleep apnea (OSA). Design: Retrospective chart review. Setting: An outpatient clinic at a Veterans Affairs Medical Center. Patients: Seventy-three patients with OSA. Interventions: Hour meters on CPAP machines provided documentation of nightly machine use. A 2-h group CPAP clinic, scheduled every 6 months, provided education, support, symptom treatment, and equipment monitoring for all CPAP patients. Results: Twenty-five patients had hour meter readings taken at their first CPAP clinic. In these patients, nightly CPAP use increased from 5.2±0.6 to 6.3±0.6 h per night after attendance at one CPAP clinic (p<0.05). CPAP use increased from 5.2±0.5 before CPAP clinic to 6.3±0.6 h per night after attendance at all subsequent CPAP clinics for 34 patients (p<0.05), an improvement that was sustained over 605±34 days. Twenty-nine percent of patients increased nightly CPAP use by at least 2 h, while only 6% decreased by ;:::2 h (p<0.025). Patients receiving supplemental oxygen had higher CPAP use prior to CPAP clinic compared to patients not receiving oxygen (p<0.05). Conclusions: Attendance in a group clinic designed to encourage patient compliance with CPAP therapy provided a simple and effective means of improving treatment of OSA. (CHEST 1997; 111:1273-77) Key words: compliance; CPAP (continuous positive airway pressure); obstructive sleep apnea Abbreviations: AHI =apnea-hypopnea index; CPAP= continuous positive airway pressure; OSA = obstructive sleep apnea
Nasal continuous positive airway pressure (CPAP) is a safe and effective the rapy for clinically significant obstructive sleep apnea (OSA).l First described in 1981, nasal CPAP has become the most commonly used nonsurgical treatment of OSA. 2 Early data on patient compliance with CPAP therapy were retrospective and required self-reporting. 3 -5 Patients reported wearing the CPAP device from 50 to 85% of sleep time. Later investigators prospec*From the Department of M edicine, Brown U niversity School of Medicine, and the Pulmonary and Critical Care Division, Providence Vete rans Affairs Medical Center, Providence, RI. Partially supported by a Depa rtmen t of Veterans Affairs Merit Review Grant, by NHLBI Grant HL34009, and by GlaxoWellcome, Inc. Some data presented at the A meri can Thoracic Society annual meeting, May 1995, and published in abstract form in the American Journal of Respiratortj and Critical Care Medicine (1995; 15l:A681). Manuscript received April 29, 1996; revision accepted J anuary 15, 1997. Reprint requests: Sharon Rounds, MD, Pulmonary and Critical Care Division, Department of Veterans Affairs Medical Center, 830 Chalkstone Ave, Providence, RI 02908
tively gathered data from monitors built into CP AP devices, re gistering time of machine operation 6 -8 and/or the time when mask pressure was elevated (indicating that the mask was being worn ). 9 - 11 These objective data demonstrated consistent patient overestimation of CPAP use when compared with selfreporting by patients. Since CPAP is a safe and effective treatment of OSA, but with poor patient acceptance, it is important to improve regular CPAP use. In this study, we investigated the effect of patient education sessions on compliance with CPAP therapy. Our objective was to identify a simple and effective means of improving nightly CPAP use. Patients for whom CPAP therapy had been prescribed for OSA were seen in a group clinic setting. Patient education and support were emphasized. CPAP use by each patient was assessed by a meter that recorded hours of machine operation. We retrospectively analyzed the use of CPAP devices b ypatients to determine if this clinical intervention improved CPAP use. CHEST I 111 I 5 I MAY, 1997
MATERIALS AND METHODS Seventy-three patients were identified as using CPAP via nasal mask or pillows in April 1994 at the Veterans Affairs Medical Center in Providence, RI. All patients had been diagnosed as having OSA, either by attended afternoon nap polysomnography study or hy portable polysomnography (Edentec; Eden Prairie, Minn ) overnight study. Nap studies monitored heart rate, oxygen saturation, two-lead EEG, genioglossus electromyogram, extraocular eye movements, chest wall excursion by impedance, and nasal airflow by thermistor. Portable nocturnal polysomnography studies monitored heart rate, nasal airflow, chest wall excursion, and oxygen saturation. The diagnosis of OSA was based on an apnea-hypopnea index (AHI ) of > 10 events per hour or > 10 desaturations per hour to :s;SO%. An event was defined as a ~4% desaturation andlor arousal on EEG accompanying a transient decrease in airflow. AHI was calculated from nap studies based on the number of events per hour of sleep time and calculated from portable studies using the number of events per hour of study time. It should be noted that this calculation of AHI may underestimate th e severitv of OSA since the total studv tim e is not necessarily equal to sleep time. ' Patients with desaturation developing slowly over 30 to 60 min (suggestive of COPD-associated hypoventilation ) or with Cheyne-Stukes or other periodic breathing patterns were excluded. All patients were tested during a second night of nasal CPAP to document resolution of sleep abnormalities, including nadir oxygen saturation > 90% and AHI <4 or resolution of clinical symptoms. CPAP was applied with a CPAP machine (Remstar Choice; Respironics; Munysv:ille, Pa) with or without supplemental oxygen (added directly to the CPAP circuit), as determined by sleep studies. Nasal CPAP was applied at a press ure of 5 to 10 em H 2 0 and increased by 5 em H 2 0 increments to a level that res ulted in the above-noted improvement in saturation andlor symptoms. Supplemental oxygen was added if the patient continued to demonstrate desaturation to <85%, despite th e maximum tolerable level of nasal CPAP. Most of the CPAP machines were equipped with meters th at registered the number of hours of machine operation, but did not necessarily record the time during which the mask was worn. Patients were not told about the presence of hour meters. RespiratOJy technicians employed by the home care vendor visited most patients to check equipment and recorded meter readings. All patients were examined initially by a pulmonary physician and were seen every 6 months by that physician. At the time of diagnosis, patients were instructed by a pulmonary nurse practitioner regarding the diagnosis of OSA and CP AP machine use and cleaning. Patients viewed a videotape detailing assembly, cleaning, adjustments, and accessories of the CP AP machine. Patients were instructed to contact the home care vendor andlor the pulmonary nurse practitioner for questions or concerns. All patients were mailed notices of appointments in the CPAP clinic at th e tim e of diagnosis or in the fall of 1993, whichever came first. The CPAP clinic was managed by a pulmonary nurse practitioner with assistance from a respiratory therapist employed by the home care vendor and a registered nurse. Eight to 10 patients attended each afternoon 2-h session. In the CPAP clinic, all patients were asked to estimate the duration of CPAP use each night. Clinical symptoms were assessed and a treatment plan was initiated on the same day. Equipment function was checked and meter readings were noted. Any equipment problems were resolved during the same CPAP clinic. Disposable supplies, such as masks and tubing, were replaced if necessary. Patients were encouraged to ask questions and to compare experiences. The session usually included a lecture on a sleep disorder topic. All patients, regardless of attendance in CPAP clinic, were resched1274
uled for CP AP clinic every 6 months and were mailed a reminder prior to each CPAP clinic session. Data were collected through individual chart review. Patients were excluded from analysis if their machine had no functional meter (n =16), if no meter reading had been obtained prior to their first CPAP clinic (n= 11 ), if no follow-up meter reading was taken after CPAP clinic attendance (n = 6), or if the CPAP machine was never used (n=3). Three patients never attended CPAP clinic, but had some mete r readings during home visits. However, there were not enough patients in this category to form a nonclinic control group. Therefore, we compared hours of CPAP use per night before and after the first and subsequent CPAP clinics. Two meter readings befo re or at the first CPAP clinic attendance were necessmy to determine use prior to CPAP clinic. Twenty-five patients who had meter readings at their first CPAP clinic had CPAP use calculated after first CPAP clinic attendance from the next available meter reading. Nine patients who did not have meter readings recorded at the flrst CPAP clinic visit could not be included in the data for the first CPAP clinic, but using a later meter reading, they had CPAP use after all CPAP clinics calculated. We calculated hourly CPAP use after all subsequent CPAP clinics for all 34 patients. We arbitrarily defined compliance as 5 h of CPAP use per night. Using this definition, we separated patients into three groups (increased, decreased, and unchanged ). Patients were considered to have increased compliance if they initially used CP AP <5 h per night prior to clinic and then increased to >5 h per night after CP AP clinic. Patients with decreased compliance used CPAP >5 h per night prior to the first CPAP clinic visit, and subsequ ently became noncompliant. Unchanged patients stayed compliant or stayed noncompliant. Data were analyzed using x2 analysis, paired two-tailed Student's t test, analysis of variance followed by Fisher's Exact Test, and least squares regression analysis (Statview; Abacus Concepts, Inc; Berkeley, CaliD. Hours per night of CPAP use were calculated by dividing the difference in meter reading by the number of nights elapsed. Data were expressed as means±SE. Diflerences between means were considered statistically significant if p :s;0.05.
Thirty-four patients attended CPAP clinic and had sufficient meter readings to calculate hours per night of CPAP use before and after their first clinic visit. The characteristics of these patients are listed in Table l. All patients were male and of a wide range of ages . Mean AHI was 34 prior to treatment. The total time reviewed for the duration of the study averaged >3 years. Mean number of CPAP clinics attended was four. Figure I shows the hours per night of CPAP machine operation. The mean duration of CPAP use prior to the first CPAP clinic was 463±69 days (range, 32 to 1,532 days). Twenty-five patients brought their machine to the first clinic and had an hour meter reading recorded. The next reading was performed either at a later CPAP clinic visit or at home by the vendor. The number of days between the first CP AP clinic visit and next reading was 212±22 days (range, 54 to 483 days) . Mean CPAP use in hours per night increased by 21% after the Clinical Investigations
Table !- Patient Characteristics* Characteristic No. of patients Gender Age, yr FEV1 , L BMI, kglm2 AHI before CPAP, events per hour CPAP level, e m H 2 0 Sa0 2 nadir, %, no CPAP No. of patients using nocturn al 0 2 No. of patients using nasal pillows/mask No. of clinics attended Total duration ofobservation, d
34 All male 58:!:2 (27-75) 2.46:!:0.16 (1.24-4.57) 37:!: 1 (22-49) 34 (10-91 ) 12:!: 1 (7.5-18) 68:!: 2 (40-90) 5 9/25 4:!: 1 (1-6) 1,151:!: 75 (513-2,162)
*Data a re expressed as mean:!: SE (range). BMI = body mass index; Sa0 2 =arterial oxygen s aturation.
first CPAP clinic visit, from 5.2 to 6.3 h p ernight. Nine patients did not have meter readings recorded during their first CPAP clinic but did have hours of operation noted on subsequent CPAP clinic visits and were therefore included in the data for all CPAP clinic visits. The effects of all CPAP clinic visits are shown in Figure 1. The mean duration of CP AP use between the first CP AP clinic attended and the last recorded meter reading was 605:±:34 days (range, 62 to 1,165 days ). The results show a statistically significant improvement in CPAP use that was sustained for as long as 3 years. Each patient's estimate of subjective hourly CPAP use was obtained by questionnaire during CP AP clinic sessions and was 6.6:±:0.3 h per night. Patients overestimated CPAP use, as noted b y previous authors ,6-8·11 but the values w ere not statistically different from objective use.
We compared hours of nightly CPAP use among groups with increased, decreased, and unchanged compliance, defining compliance as 5 h of CPAP use per night (Fig 2). Ten patients increased CPAP nightly use by ::::::2 h after all CPAP clinics while two night (p < 0.05, x2 patients decreased use by 2 h p er analysis ). Thus, a significant number of patients (29%) dramatically improved CPAP nightly use, while only 6% decreased CPAP use by at least 2 h despite CPAP clinic attendance. Five patients were treated with CPAP and supplemental o),:ygen (l to 4 Umin). Patients using CPAP plus oxygen had higher initial nightly use prior to CPAP clinic attendance (8.4:±:0.8 h per night) compared to patients not using supplemental oxygen (4.6:±:0.4 h per night; p < 0.05, Student's t test). Regression analysis comparing preclinic and postclinic CPAP use showed no correlation between CPAP pressure, nadir oxygen saturation on sleep study, use of nasal pillows, or patients' subjective assessment of drowsiness. There was no correlation between hourly use before and after CPAP clinic attendance. Thu~ , having a low nightly use did not predict improvement. Comparing the patients who became compliant with the patients who became noncompliant, there were no differences in age, the number of visits to their pulmonary physicians, visits to their medical physicians, visits to their CPAP clinic, or all physician interactions combined. DISCUSSION
In this retrospective chart review, we assessed CPAP nightly use by patients who attended a CPAP Change in Compliance with CPAP Therapy
CPAP Use Before and After Clinic 8
.E., ·;: U;
a. < a.
Before Clinic After Clinic
FIGURE l. CPAP use be fore and after CPAP clinic. The 25 patients with first CPAP clinic readings recorded h ad CPAP use compared b e fore and after their first CPAP clinic attendance. Asterisk indicates p < 0.05 compared to before first clinic. CPAP use is also compared b efore and after all subsequent CPAP clinics attended f or 34 patients. Two asterisks indi cate p < 0.05 compared to before CPAP clinics (Student's t tes t).
Before Clinic After Clinics
FI GURE 2. C hange in compli ance during CPAP therapy. ComCPAP use per night. Patients with pliance was defined a s 5 h of increased co mpliance had < 5 h ofCPAP use per night before CPAP clinic attendance and increased t o ~5 h p er night after at least one CPAP clinic. Patients with decreased compliance used CPAP ~ 5 h p er night before CPAP clinic and subseque ntly becam e noncompliant. Patients with no changes stayed compliant or noncompliant before and after CPAP clinic attendance. Asterisk indicates p < 0.05 compared t o dec reased p atients; two asterisks, p < 0.05 co mpared to unchanged p atients (analysis o f vari ance and Fisher's Exact Test).
CHEST I 11 1 I 5 I MAY, 1997
clinic designed to educate and encourage patient compliance with prescribed CPAP therapy. The results demonstrated that nightly CPAP use improved and that this improvement was sustained over years in patients who attended at least one CPAP clinic. Significantly more patients had a >2 h increase in nightly CP AP use than had a nightly decrease of at least 2 h. Patients receiving supplemental oxygen with CPAP had higher initial CPAP use than patients not receiving oxygen. Other parameters, such as CP AP pressure, use of nasal pillows, nadir saturation on sleep study, or patients' subjective drowsiness did not correlate "vith CPAP use either before or after CPAP clinic attendance. Low initial CPAP use did not accurately predict improvement. Because of the retrospective nature of this study and lack of a concurrent control group, we cannot exclude other factors that may have influenced our results, such as changes in equipment (eg, masks and humidifiers). In addition, we cannot exclude the possibility that the home care vendor may have inadvertently provided the patients information about the presence of the hourly use meters. Variability among patients in the time between initiation of CP AP therapy and attendance at the first CPAP clinic may have obscured important effects of the CPAP clinic attendance. In this study conducted at a Veterans Affairs Medical Center, the subjects were all male and had a diagnosis of OSA confirmed by sleep studies. Their CPAP machines were operational 5 to 6 h per night, which is similar to published reports of compliance of 4 to 6 h per night. 4 ·6 -ll Previous literature has identified several possible characteristics of patients with poor compliance, including nasal discomfort and lack of perceived benefit. 10 Daytime drowsiness has been associated with CPAP compliance, but not in all studies. 5 - 7 •9 · 11 In this study, the only factor associated with higher initial CPAP use was concomitant oxygen therapy. This subset of patients could represent a population with more significant disease. It is also possible that the patients perceived oxygen to have greater "life-saving" value than CPAP alone. However, patients receiving oxygen were also followed up_in an oxygen clinic prior to the CPAP clinic. Oxygen clinic had a similar format to CPAP clinic and may explain the higher initial CP AP use for patients receiving supplemental oxygen. That is, both clinics were designed to reinforce patient education and understanding of their disease. Thus, two different clinics may have had additive effects on patient compliance. Few studies have addressed interventions that may improve CPAP compliance. A recent study by Fletcher and Luckett 12 evaluated the effectiveness of 1276
telephone reinforcement on CPAP use. In their prospective randomized crossover design, positive reinforcement by telephone conversation did not affect compliance. Our data suggest that intermittent group sessions improve compliance and that the effect is sustained over years. The patients included in the present study were selected because of attendance to at least one CPAP clinic, and thus may represent a more receptive subpopulation. Without matched control subjects who did not attend CPAP clinic, it is difficult to determine whether this effect was due to nonclinic learning or improved tolerance over time. Our study design measured time that the machine was operational, but did not record mask use. However, previous literature indicates that the time of machine operation correlates with time the mask is wornY-11 Therefore, it is likely that CPAP machines were operational when masks were worn and that meter readings reflected actual use. Previous studies have defined compliance as anywhere from 3 to 7 h of CPAP use per night. 6 8 ,11 Others have defined compliance as the hours per night of use, not determining a level that separated compliance from noncompliance.9 · 12 vVe arbitrarily defined compliance as 5 h of CPAP use per night, which is a rough median of previous compliance definitions. Nasal CPAP is an effective treatment for OSA and has a low associated mortality.1 Compliance is often the limiting factor for CPAP use. Hour meters provide an effective and simple method of determining use of the CPAP machine . This study shows that intermittent educational and supportive sessions resulted in sustained improvement in regular CPAP use. Noncompliant patients appeared to receive the most benefit from CPAP clinic intervention. The CPAP clinic staff both directly assessed equipment problems and immediately treated patient symptoms. Problems requiring a physician's attention were referred from CP AP clinic. The personal contact and teaching, along with interaction with other OSA patients, provided an atmosphere of elkouragement and support. Further data, including matched, controlled, prospective studies, are necessary to determine the minimal duration and frequency of CPAP clinic sessions necessary to maintain compliance with therapy. In addition, controlled studies are needed to assess the cost effectiveness of patient education and support groups as a means of improving compliance with CPAP therapy. ACKNOWLEDGMENTS: The authors thank Richard Millman, MD, for helpful discussions and William Silvia, RN , and Javne Matoian, RRT, Vanguard Home Medical Care, for excelfent assistance with patient education. Clinical Investigations
2 3 4 5 6
REFERENCES ATS Assembly on Respirat01y eurobiology and Sleep. Indications and standards for use of nasal continuous positive airway pressure in sleep apnea syn dromes. Am JRespir Crit Care Med 1994; 150:1738-45 Sullivan CE , Berthon-Jones M , Issa FQ, et al. Reversal of obstructive sleep apnoea by continuous airway press ure applied through the nares. Lancet 1981; 1:862-65 Sanders MH, Gruencll CA, Rogers RM. Patient compliance with nasal CPAP therapy for sleep apnea. Chest 1986; 90:330-33 I-Ioffstein V, Viner S, Mateika S, et al. Treatm ent of obstructive sleep apnea with nasal continuous positive ailway pressure. Am Rev Respir Dis 1992; 145:841-45 Rauscher H , Popp W, Wanke T, et a!. Acceptance of CPAP th erapy for sleep apnea. Chest 1991; 100:1019-23 Meurice JC, Dore P, Paquereau J, et a!. Predictive factors of long-term compliance with nasal continuous positive airway pressure treatment in sleep apnea syndrom e. Chest 1994; 105:429-33
7 Rauscher H , Formanek D , Popp W, et al. Self-reported vs measured compliance with nasal CP AP for obstructive sleep apnea. Chest 1993; 103:1675-80 8 Krieger J, Kurtz D. Objective measurement of compliance with nasal CPAP treatment for obstructive sleep apnoea syndrome. Eur Respir J1988; 1:436-38 9 Reeves-Hoche MK, Meek R, Zwillich C. Nasal CPAP: an objective evaluation of patient compliance . R espir Crit Care Med 1994; 149:149-54 10 Engleman HM , Martin SE, Douglas NJ. Compliance \V:ith CPAP therapy in patients with the sleep apnoea!hypopnoea syndrome. Thorax 1994; 49:263-66 11 Kribbs NB, Pack AI, Kline LR, et al. Objective measurement of nasal CPAP use by patients with obstructive sleep apnea. Am Rev Respir Dis 1993; 147:887-95 12 Fletcher EC, Luckett RA. The effect of positive reinforcement on hourly compliance in nasal continuous positive aitway pressure users with obstructive sleep apnea. Am Rev H.espir Dis 1991; 143:936-41