Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison / control (C) ³

Follow-up

Outcome measures and effect size ⁴

Comments

McMillan, 2015

Type of study:

Randomised, parallel, investigator-blinded multicentre trial

Setting: sleep centres

Country: UK

Source of funding:

National Institute for Health Research Health Technology Assessment Programme.

Inclusion criteria:

- aged ≥ 65 years

- newly diagnosed OSAS (defined as oxygen desaturation index at ≥4% desaturation threshold level for >7.5 events/hour and Epworth

Sleepiness Scale (ESS) score of ≥9)

Exclusion criteria:

- previous CPAP therapy

- arterial awake oxygen saturation <90%

- forced expiratory volume in 1 second (FEV1)//forced vital capacity (FVC) ratio < 60%

- substantial problems with sleepiness while driving (in those who are still driving)

- currently using heavy goods vehicle or professional service vehicle driving licence

- shift work

- any very severe complication of OSAS such that CPAP therapy was mandatory

- inability to give informed consent or comply with the protocol

N total at baseline:

Intervention: 140

Control: 138

Important prognostic factors²:

Age, median (25^th -75^th percentiles):

I: 69.5 (67.1-74.1)

C: 70.3 (68.0-73.8)

Sex:

I: 86% M

C: 79% M

BMI ± SD:

I: 33.9 ± 5.7

C: 33.6 ± 6.4

ESS:

I: 11.6 ± 3.4

C: 11.6 ± 4.9

Groups were comparable at baseline.

Intervention:

Autotitrating CPAP with best supportive care (BSC).

Control: best supportive care (structured advice on minimising sleepiness: improved sleep hygiene, using naps or caffeine as required and weight loss if appropriate, which was

summarised in a booklet format.)

Length of follow-up:

12 months

ESS was measured after 3 and 12 months; cost-effectiveness (including EQ-5D) after 12 months)

Loss-to-follow-up:

3 months

Intervention: N=15 (11%)

Reasons: withdrew consent (n=11), unable to be contacted (n=2), other medical condition (n=2)

Control: N=12 (9%)

Reasons: withdrew consent (n=10), unable to be contacted (n=1), other medical condition (n=1)

12 months

Intervention: N=10 (%)

Reasons: withdrew consent (n=4), unable to be contacted (n=5), died (n=1).

Control: N=6 (%)

Reasons: withdrew consent (n=2), unable to be contacted (n=3), died (n=1)

Incomplete outcome data:

3 months

Intervention

N 4 (%)

Reasons (describe)

Control:

N (%)

Reasons (describe)

Outcome measures and effect size (include 95%CI and p-value if available):

Outcome measure-1: ESS

After 3 months (mean ± SD):

I: -3.8 ± 0.4

C: -1.6 ± 0.3

Adjusted treatment effect was -2.1 (95%CI -3.0, -1.3, p<0.001), favouring CPAP.

After 12 months:

Adjusted treatment effect was -2.0 (95%CI -2.8, -1.2, p<0.001).

The effect was greater in patients with increased CPAP use or higher

baseline ESS score.

Outcome measure-2: QoL

SF-36, 3 months

Mental component scale:

1.9 (0.0 to 3.8); p=0.046

Physical component scale:

–0.2 (–2.5 to 2.1); p=0.84.

Subdomains that were significant different were: energy/vitality, and social functioning.

SF-36, 12 months

Mental component scale:

1.4 (–0.8 to 3.6); p=0.22

Physical component scale:

0.6 (–2.2 to 3.4); p=0.68.

Subdomains that were significant different were: energy/vitality, physical functioning.

Outcome measure-3: Cognitive performance

3 months (treatment effect (95%CI)

MMSE: -0.2 (–0.6 to 0.2)

DSS:  0.8 (–0.9 to 2.5)

TMT-B: 0.7 (–7.2 to 8.5)

Simple reaction time:  –12.8 (–39.9 to 14.3);

Four-choice reaction time, # correct answers:  –0.1 (–0.5 to 0.4);

Four-choice reaction time, # correct answers: 32.0 (–0.7 to 64.8); p=0.055

12 months (treatment effect (95%CI)

MMSE: 0.1 (–0.3 to 0.5)

DSS: 1.1 (–0.6 to 2.7)

TMT-B: 6.2 (–3.4 to 15.8)

Simple reaction time:  –16.4 (–39.1 to 6.2)

Four-choice reaction time, # correct answers:  0.3 (–0.2 to 0.8)

Four-choice reaction time, # correct answers: 1.8 (–33.6 to 37.2)

In older people with OSAS, CPAP reduces sleepiness and is marginally more cost-effective than BSC over 12 months.

After 3 months, 86% still used CPAP; after 12 months this was 71%.

At 3 months, CPAP usage was 1h and 52 minutes (IQR 19 min-5 h and 12 min). After 12 months, CPAP usage was 2 h and 22 min (IQR 10 min – 5 h and 9 min).

Martinez-Garcia, 2015

Type of study: open-label, randomised multicentre clinical trial with a blinded end-point

Setting: 12 clinical centers

Country: Spain

Source of funding:

Instituto de Salud Carlos III, and Linde HealthCare, Oximesa, Vitalaire and Teijin. None of these institutions played any role in the objectives, methods, data

analysis or conclusions of the study.

Inclusion criteria:

- aged ≥70 years

- AHI ≥30

Exclusion criteria:

- current use of CPAP treatment

- central sleep apnoea

- respiratory failure (defined as diurnal oxygen saturation below 90%)

- severe heart failure (New York Heart Association functional class III–IV)

- a cardiovascular event in the month prior to the inclusion in the study

- disabling hypersomnia (Epworth Sleepiness Scale (ESS) ≥18)

- previous severe impairment of cognitive performance that ruled out their inclusion in the study.

N total at baseline:

Intervention: 115

Control: 109

Important prognostic factors²:

Age ± SD

I: 75.4 ± 3.8 yrs

C: 75.6 ± 4.0 yrs

Sex:

I: 63.5% men

C: 73.4% men

BMI ± SD

I: 33.0±7.3 kg/m2

C: 32.8±5.1 kg/m2

AHI± SD

I: 53.5±15.6

C: 47.2±13.4

Groups were comparable at baseline

Describe intervention:

CPAP.

Average use of CPAP treatment was 4.9±2.5 h/night. 30% patients had <4 h per night; 70% had>4h/night.

Describe  control: No therapy

Length of follow-up:

69 months

Loss-to-follow-up:

Intervention

N=1 (1%)

Reasons: not described

Control

N=4 (4%)

Reasons: not described

Incomplete outcome data:

Intervention

N=36 (31%)

Reasons: no adherence to CPAP (≥4 h/night). Data were imputed

Control:

N=4 (4%)

Reason: lost to follow-up. Data were imputed

Outcome measures and effect size (include 95%CI and p-value if available):

Outcome measure-1: ESS

Defined as (mean±SD):

I: Baseline: 9.6±4

I: follow-up: 5.9±3.4

I: Change: -3.6±4.0; effect size: -0.91

C: Baseline 9.3±3.6

C: Follow-up: 9.2±4.0

C: Change: -0.1±2.9 (effect size -0.03

Intergroup difference:

3.4 (95%CI 2.5, 4.2); p<0.001.

Outcome measure-2: QoL

Quebec Sleep Questionnaire (QSQ) - hypersomnolence

Group diff: -0.6 (95%CI -0.9, -0.3); p<0.001

QSQ -Diurnal symptoms

Group diff: -0.8 (95%CI -1.0, -0.5) p<0.001

QSQ- nocturnal symptoms

Group diff: -1.0 (95%CI -1.2, -0.7); p<0.001

QSQ – emotions

Group diff: -0.6 (95%CI -0.9, -0.3); p<0.001

QSQ – social interaction

Group diff: -0.5 (-0.8, -0.3); p<0.001

All significant, but not clinically sign difference.

Outcome measure 3: cognitive function

Digit span

Intergroup difference:

-0.01 (95%CI -0.6, 0.4); p=0.75

Digit symbol

Intergroup difference:

-1.4 (95%CI -2.8, -0.0); p=0.047

CPAP treatment gave a significant improvement on ESS and quality of life. However, the effect on QoL did not reach clinical significance.

Compliance was similar to other studies.

The average use of CPAP treatment was 4.9±2.5 h/night, with 35 (30.4%) patients presenting <4 h per

night (69.6% with good adherence).

Dalmases, 2015

Type of study:

Randomized, evaluator-blinded, parallel-group, single-center pilot study

Setting:

Country: Spain

Source of funding:

Catalan Society of Pneumology, Spanish Ministry of Science

and Innovation.

Inclusion criteria:

- aged ≥65 years

- newly diagnosed severe OSA by PSG (≥

- ESS ≤12)

Exclusion criteria:

- Mini-mental state examination  <24

- respiratory failure

- neurologic or psychiatric disorders

- chronic heart failure

- unstable illnesses

- other sleep disorders

- contraindications for MRI

- inability to respond to questionnaires.

N total at baseline:

Intervention: 17

Control: 16

Important prognostic factors²:

Age ± SD:

I: 71.9 ± 6.0 yrs

C: 70.8 ± 5.1 yrs

Sex:

I: 65% M

C: 75% M

BMI ± SD:

I: 32.8 ± 3.9 kg/m2

C: 29.9 ± 4.2 kg/m2

AHI ± SD:

I: 61.2 ± 17.9 events/h

C: 49.5 ± 15.8 events/h

Groups were comparable at baseline; BMI and AHI were slightly different (p=0.06)

Intervention:

CPAP.

Compliance: median 6.0 (1.7) hours/night

Control: Conservative care, consisting of sleep hygiene and dietary counselling

Length of follow-up:

Three months

Loss-to-follow-up:

Intervention:

N=1 (6%)

Reasons: unable to be contacted

Control:

N=1 (6%)

Reasons: unable to be contacted

Incomplete outcome data:

Cognitive tests

Intervention

N=1 (6%)

Control

N=1 (6%)

Outcome measures and effect size (include 95%CI and p-value if available):

Outcome measure-1: ESS

Defined as (mean±SD):

I: Baseline: 7.9 ± 3.0

I: follow-up: 6.4 ± 3.8

C: Baseline: 5.8 ± 3.6

C: follow-up: 6.1 ± 2.8

LS mean difference: -0.63 (95%CI -2.93, 1.67); p=0.58

Outcome measure-2: QoL

Quebec Sleep Questionnaire (QSQ)

Defined as (mean±SD):

I: Baseline: 26.5 ± 5.5

I: follow-up: 32.8 ± 5.4

C: Baseline: 29.6 ± 2.9

C: follow-up: 30.5 ± 2.4

LS mean difference: 3.25 (95%CI 0.13, 6.38); p=0.04, favouring CPAP.

Domains that showed a significant effect: nocturnal symptoms, social interactions. 

Outcome measure 3: cognitive function

Defined as LS Mean difference (95%CI)

Digit span forward

1.06 (0.10, 2.01); p=0.03

Digit span backward

0.79 (-0.30, 1.88); p=0.15

RAVLT

7.60 (1.66, 13.55); p=0.01

Digit symbol

5.74 (1.69, 9.79); p=0.007

TMT-A

-2.68 (-15.99, 10.63); p=0.68

TMT-B

-47.64 (-81.83, -13.45); p=0.008

Semantic fluency

2.01 (-0.75, 4.76); p=0.15

Phonemic fluency

7.02 (-0.20, 14.23); p=0.056

CPAP treatment improved cognitive function in older adults. ESS did not differ between CPAP and conservative therapy. 

Crawford-Achour, 2015

Type of study:

Cohort study (PROOF study)

Setting:

Country: France

Source of funding: not industry-supported.

Inclusion criteria:

- aged ≥65 years

Exclusion criteria:

- history of myocardial infarction, stroke, heart failure, atrial fibrillation

- type 1 (insulin-dependent) diabetes mellitus;

- pacemaker implantation;

- a pathological condition limiting life expectancy to < 5 years;

- contraindication to a brain MRI;

- being institutionalized;

- or planning to move during the next 2 years.

N total at baseline:

N=126

Intervention: 33

Control: 93

Important prognostic factors²:

Age ± SD:

I: 75.0 ±1.1 years

C: 74.7 ± 1.1 years

Sex:

I: 70% M

C: 58% M

BMI ± SD:

I: 27.8 ± 3.3 kg/m2

C: 26.7 ± 3.5 kg/m2

AHI ± SD:

I: 49.0 ± 15.4 events/h

C: 40.7 ± 9.1 events/h

Baseline BMI and AHI were statistically different between treatment groups.

Intervention:

CPAP;

Self-reported mean compliance was 6.6±1.1 h/night, 7.0±0.2 nights per week. All participants estimated their compliance >4h/night.

Control:

No CPAP

Length of follow-up:

44.5 ± 26.3 months

Loss-to-follow-up:

Intervention:

N=2 (%)

Reasons; intolerance to CPAP

Incomplete outcome data:

Intervention:

N (%)

Reasons (describe)

Outcome measure 3: cognitive function

Time-treatment:

MMES: ns

VAS: ns

Delayed total recall: ns

Delayed free recall: p=0.02

Benton test: ns

TMT A: ns

TMT B: ns

Stroop (words): ns

Stroop (colors): ns

Semantic fluency: ns

Alphabetic fluency: ns

Similarities: p<0.0001.

Adjustments for sex, BMI, socio-educational level, AHI, desaturation index, did not change the results.

30% of the patients in the CPAP group reported adverse effects, none

of which led to treatment cessation.

Compliance, CPAP-use: 6 (SD 1.7) h/night

Lopez-Padilla, 2016

Type of study: historical cohort

Setting:

Country: Spain

Source of funding:

Not described. The authors declared no conflicts of interest.

Inclusion criteria:

- ≥ 80 yrs at OSA-diagnosis

Exclusion criteria:

- AHI <20

- previous treatment with

CPAP or bi-level positive pressure

- diagnosis of central sleep

apnea, obesity hypoventilation syndrome, or chronic respiratory

failure.

N total at baseline:

Intervention: 79

Control: 76

Important prognostic factors²:

Age ± SD:

I: 81.7 ± 1.7 y

C: 81.2 ± 1.7 y

Sex:

I: 53% M

C: 55% M

BMI ± SD:

I: 32.5 ± 4.4

C: 33.0 ± 5.1

AHI± SD:

I: 53.0 ± 19.2

C: 45.2 ± 18.2

Groups comparable at baseline? No; CPAP-group was significantly older and had a higher AHI and ODI.

Intervention:

CPAP (≥ 4 hours/night)

Control:

Non-CPAP (no CPAP prescription, or CPAP-use <4 hours/night)

Length of follow-up:

Follow-up measurements were done after three, six and 12 months after CPAP prescription, and every 12 months after the first year.

Median follow-up was 53 months (IQR 41-77 months).

Loss-to-follow-up:

7 patients were lost to follow-up. No reasons were described.

Incomplete outcome data:

Data from the last observation was taken into account, so no loss of outcome data.

Outcome measure 4: mortality

# deaths

I: n=35 (44%)

C: n=48 (63%)

Cox proportional analyses,  adjusted model:

HR=0.46, 95%CI 0.27, 0.78, p=0.004

Long-term CPAP treatment is associated with a higher rate of survival in very old persons with moderate to severe OSA.

Long-term adherence (CPAP-use (≥ 4 hours/night) was observed in 60% of the prescriptions.

Ou, 2015

Type of study: prospective cohort study

Setting: ambulatory

Country: China

Source of funding:

National Key Clinical Department, National Key Discipline, Guangdong Key Laboratory For Diagnosis and Treatment of Major Neurological Diseases, Guangdong Technological grant, and Guangdong Department of Science & Technology Translational Medicine Center grant.

The authors declared no conflicting interests.

Inclusion criteria:

- Elderly subjects (≥ 60 years)

- Moderate to severe sleep apnea (AHI > 20 events/h)

Exclusion criteria:

-

N total at baseline:

Intervention: 36

Control: 88

Important prognostic factors²:

Age ± SD:

I:71.3 ± 6.3 y

C: 73.4 ± 6.1 y

Sex:

I: 92% M

C: 81% M

BMI ± SD:

I: 30.8 ± 9.0 kg/m2

C: 30.9 ± 8.7 kg/m2

AHI ± SD:

I: 45.3 ± 13.05

C: 36.05 ± 13.94

ESS ± SD:

I: 8.08 ± 3.49

C: 6.19 ± 3.5

Groups comparable at baseline? CPAP-group had a significant higher ESS and AHI.

Intervention:

CPAP

Control:

Non-CPAP

Length of follow-up:

Mean follow-up time was 5 years ± 2.5 (range 1-8 years)

Loss-to-follow-up:

Not described

If patients had not visited the sleep laboratory within the last 6 months before the end of the study, and could not be reached by repeat telephone calls, the latest available dataset of the annual follow-up investigations was used for censoring of event-free survival.

Incomplete outcome data:

Not described

Outcome measure 4: mortality

I: 5.6% deaths (2/36)

C: 21.1% deaths (19/88);

P=0.035

Survival rate (Kaplan-Meier):

I: 94.4%, 95%CI: 82–99%

C: 78.4%, 95%CI: 70–87%;

log-rank χ2 = 6.3; p = 0.01

Good compliance (≥ 4 hours/night for more than 5 days per week): 66.7% (24/36); poor compliance in 33.3% (12/36).

The incidence of cardiovascular events was lower in those with good compliance.

Jennum, 2015

(we only report the data of the subpopulation of age≥60 yrs)

Type of study: historical cohort

Setting:

Country: Denmark

Source of funding: Not mentioned. The authors declared no conflicts of interests.

Inclusion criteria:

- patients with OSA (ICD-10 code G473)

- ≥ 60 yrs

Exclusion criteria:

-

Data were derived from the National Patient Registry.

N total at baseline:

Intervention: 3837

Control: 2882

Important prognostic factors²:

Not described for the subpopulation (age ≥ 60 yrs)

Groups comparable at baseline?

Intervention:

CPAP

Control:

Non-CPAP

Length of follow-up:

10 years

Loss-to-follow-up:

Not described, but probably none as it is a historical cohort

Incomplete outcome data:

Not described

Outcome measure 4: mortality

Hazard ratio=0.62 (95%CI 0.54, 0.71; p<0.0001)

Ancoli-Israel, 2008

Type of study: double-blind RCT

Setting: general clinical research center

Country: USA

Source of funding:  not industry-supported

Inclusion criteria:

- diagnosis of mild probable AD

- a CT or MRI of the brain consistent with AD done within 24 months

- stable health

- a live-in caregiver.

- English speaking patients

- a Mini Mental Status Examination (MMSE) score >17

- AHI ≥10

Exclusion criteria:

- prior diagnosis of a sleep disorder or current treatment for OSA.

- Participants with severe medical or psychiatric illnesses

N total at baseline:

Intervention: 27

Control: 25

Important prognostic factors²:

Age ± SD:

I: 78.6±6.8

C: 77.7±7.7

Sex:

I: 70% M

C: 80% M

BMI:

I: 26.1±4.2

C: 25.0±3.6

MMSE:

I: 24.3±2.8

C: 24.8±4.2

Groups comparable at baseline? Yes

Intervention:

therapeutic

CPAP (tCPAP)

Control:

Placebo (pCPAP, three weeks) followed by therapeutic CPAP (three weeks).

Control CPAP had a placebo mask with ten ¼-inch drill holes to create a large air leak and to allow for adequate air exchange with a pressure reducer placed between the whisper swivel and the CPAP tubing connected to the mask.

Length of follow-up:

6 weeks.

Placebo was used for 3 weeks.

Loss-to-follow-up:

Intervention:

N=7 (26%)

Reasons: illness, unable to tolerate CPAP, nose bleed

Control:

N=5 (20%)

Reasons: illness, unable to tolerate CPAP, unable to tolerate PSG.

Incomplete outcome data:

Intervention:

N=7 (26%)

Control:

N=5 (20%)

Outcome measure-1: AHI:

Intervention:

Baseline: 29.7±15.9

6 wks: 6.4±8.1

Control:

baseline: 26.9±15.5

After 3 wks: 34.6±22.3

Outcome measure : cognition

Between intervention and control (after 3 weeks): no difference

Before and after 3 weeks of CPAP: improvement (pre-post comparison).

Author’s conclusion: OSA may aggravate cognitive dysfunction in dementia and thus may be a reversible cause of cognitive loss in AD patients. OSA treatment seems to improve some of the cognitive functioning.

Compliance:

Intervention: 5.8±2.1 h/n for 73% of the nights.

Control: 6.4±2.5 h/n for 67% of the nights.

After switching to therapeutic CPAP, mean use was 4.9±2.4 for 62% of the nights.

Cooke, 2009

Type of study: clinical study

Setting: general clinical research center

Country: USA

Source of funding: not industry-supported

Inclusion criteria:

See above (Ancoli-Israel)

Exclusion criteria:

See above (Ancoli-Israel)

N total at baseline:

Intervention: 5

Control: 5

This is a subset of participants

from an initial 6-week RCT of Ancoli-Israel (2008)

Important prognostic factors²:

Age ± SD: 75.7±5.9 yrs

Sex: 70% men

MMSE ± SD: 22.6±4.5

BMI ± SD: 25.9±3.1 kg/m2

Groups comparable at baseline? Yes

Intervention:

CPAP

Control:

No treatment

Length of follow-up:

Mean follow-up was 13.1±5.2 months (range 6-21 months)

Loss-to-follow-up:

Incomplete outcome data:

Outcome measure- QoL (FOSQ)

No difference between groups. Effect size = -0.2

Outcome measure- cognition  (MMSE)

No difference between groups; effect size= 0.1

Author’s conclusions:

The results of this preliminary study raise the possibil­ity that sustained, long-term CPAP treatment for patients with AD and OSA may result in lasting improvements in sleep and mood as well as a slowing of cognitive deterioration.

Compliance was not measured

Aihara, 2010

Type of study: retrospective cohort

Setting:

Country: Japan

Source of funding: not industry-supported

Inclusion criteria:

- patients with OSA

- middle-aged and elderly

- CPAP-use for more than 3 years

Exclusion criteria:

N total at baseline:

Intervention:

Control:

Important prognostic factors²:

Age ± SD: 66.7 (95%CI 65-68)

Sex:  89%men

BMI ± SD: 26.5 (25.6-27.5) kg/m2

AHI: 38.5 (33.8-43.2)

ESS: 8.84 (6.9-10.8)

Intervention:

CPAP

Control:

No treatment

Length of follow-up:

Mean follow-up was 1000 days. Data was gathered at two time points as close as possible to a yearly interval (616 and 1048 days).

Loss-to-follow-up:

Not described, probably none as is was a medical chart study.

Incomplete outcome data:

Not described, probably none as is was a medical chart study.

Outcome measure-Blood pressure

Diastolic BP: baseline 82.2 (95%CI 78.4-86.1)

Baseline - 600 days: p=0.12

600-1000 days: p=0.45

Baseline-1000 days: p=0.02

Systolic BP

Baseline: 138.0 (95%CI 133.0-143.1)

Baseline - 600 days: p=0.17

600-1000 days: p=0.76

Baseline-1000 days: p=0.09

Yagihara, 2012

Type of study: Prospective case-control study

Setting: outpatient sleep clinic

Country: Brazil

Source of funding:  Associac¸a˜o Fundo de

Incentivo a Psicofarmacologia (AFIP) and Fundac¸a˜o de Amparo a` Pesquisa

do Estado de Sa˜o Paulo (FAPESP)

Inclusion criteria:

- age between 60 and

75 years

- male gender

- body mass index (BMI) <35 kg/m2.

- clinical criteria of OSAS and an AHI of ≥20 events per hour

Exclusion criteria:

- individuals who had neurological or psychiatric diseases

- sleep disorders other than OSAS

- alcohol abuse

- psychoactive drug use

- patients who had previously received treatment for OSAS

N total at baseline:

Intervention: 30

Control: 27

Important prognostic factors²:

Age ± SD: 66.4±0.7

Sex: 100% men

BMI ± SD: 27.9±0.7 kg/m2

Groups comparable at baseline? Yes, only BMI was higher in the OSA-group

Intervention:

CPAP

Control:

No treatment

Age-mathed controls

Length of follow-up:

Mean follow-up was six months

Loss-to-follow-up:

Intervention:

N (%)

Reasons (describe)

Control:

N (%)

Reasons (describe)

Incomplete outcome data:

Intervention:

N (%)

Reasons (describe)

Control:

N (%)

Reasons (describe)

Outcome measure-1: AHI:

Before: 36.6 ±3.4

Post: 4 ± 0.9, p<0.01

Outcome measure-2: ESS

Before: 10.7 (8.0-13.3)

Post: 7.4 (5.9-9.2)

Outcome measure-QoL

Measured by the SF-36

Functional:

Pre: 77.4 (95%CI 69.2-85.5)

Post:83.4 (76.9-89.9

Pain

Pre:65.4 (95%CI 52.8-78.1

Post: 77.8 (67.2-88.3

Score on the other domains did nog change

Martinez-Garcia, 2012

Type of study: prospective observational study

Setting:

Country: Spain

Source of funding: not industry-supported (public grants)

Inclusion criteria:

- 65 years or older

Exclusion criteria:

- CSA (more than 50% of apneic events)

- previous treatment with CPAP

- unwillingness to undergo a sleep study

N total at baseline:

Total included in the study: n=943

Control group (AHI<15): n=115

Mild OSA (AHI 15-29) without CPAP: n=108

Severe OSA (AHI >29) without CPAP: n=173

OSA with CPAP: n=503

Important prognostic factors²:

Age: 7001 – 71.9

Sex:  60-71.7% men

Groups comparable at baseline? Not for ESS, AHI, previous CVE

Intervention:

CPAP (OSA with CPAP)

Control:

No treatment / no OSA

- Control group (no OSA)

- mild OSA without CPAP

- severe OSA without CPAP

Length of follow-up:

Median follow-up was 69 months

Loss-to-follow-up:

N=4 (0.4%)

Incomplete outcome data:

N=4 (0,4%)

Outcome measure – Mortality

Control group: HR=1

Cardiovascular mortality

Mild OSA without CPAP: HR=1.83 (0.73-2.64)

Severe OSA without CPAP: HR=2.25 (95%CI 1.41-3.61)

OSA with CPAP: HR: 0.93 (95%CI 0.46-1.89)

All-cause

Mild OSA without CPAP: HR=1.17 (0.75-1.81)

Severe OSA without CPAP: HR=1.99 (95%CI 1.42-2.81)

OSA with CPAP: HR: 1.35 (95%CI 0.85-2.15)

Stroke

Mild OSA without CPAP: HR=4.25 (0.88-20.49)

Severe OSA without CPAP: HR=4.63 (95%CI 1.03-20.8)

OSA with CPAP: HR: 1.15 (95%CI 0.24-5.5)

Severe OSA not treated with CPAP was associated with an increased risk in cardiovascular mortality due to stroke and heart failure. CPAP treatment may decrease the risk of mortality to the level of patients without OSA.

Study reference

(first author, publication year)

Describe method of randomisation¹

Bias due to inadequate  concealment of allocation?²

(unlikely/likely/unclear)

Bias due to inadequate blinding of participants to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of care providers to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of outcome assessors to treatment allocation?³

(unlikely/likely/unclear)

Bias due to selective outcome reporting on basis of the results?⁴

(unlikely/likely/unclear)

Bias due to loss to follow-up?⁵

(unlikely/likely/unclear)

Bias due to violation of

intention to treat analysis?⁶

(unlikely/likely/unclear)

McMillan

Patients were randomised using a computerised service, on a 1 : 1 basis with a random element of 80% and stratified by disease severity (enrolment ESS score of >13 versus ≤13), functionality using the TDS score of >1 versus ≤1 and recruitment centre.

Unlikely

Likely

Unlikely

Unclear: It was not possible to blind all trial staff, although the assessments were done blind wherever possible.

Unlikely

Unlikely; dropout was 16%.

Unlikely

Martinez-Garcia, 2015

Randomisation was done with specific software. Random allocation stratified by site was used, without

any other restriction

Unlikely (The software only revealed the allocation group when an investigator provided the data of a fully eligible patient, which guaranteed the concealment of the randomisation sequence.)

Likely

Unclear

Unlikely

Unlikely

Unlikely; dropout was low.

Unclear; imputed data for 30% (because of no CPAP-adherence).

Dalmases, 2015

1:1 ratio, computer-generated list

Unclear: allocation group was only revealed after baseline data.

Likely

Unlikely

Unlikely

Unlikely

Unlikely; dropout was low

Likely; drop-outs were not analysed.

Ancoli-Israel, 2008l

Not described

Unclear

Unlikely

Unlikely

Unlikely

Unlikely

Unclear: drop-out was rather high but was comparable between groups

Likely: drop-outs were not analysed

Study reference

(first author, year of publication)

Bias due to a non-representative or ill-defined sample of patients?¹

(unlikely/likely/unclear)

Bias due to  insufficiently long, or   incomplete follow-up, or differences in follow-up between treatment groups?²

(unlikely/likely/unclear)

Bias due to ill-defined or inadequately measured outcome ?³

(unlikely/likely/unclear)

Bias due to inadequate adjustment for all  important prognostic factors?⁴

(unlikely/likely/unclear)

Crawford-Achour, 2015

Unlikely

Unlikely

Unlikely

Likely

Lopez-Padilla, 2016

Likely: historic cohort

Unclear

Unlikely (no blinding, but hard outcome measure)

Likely

Jennum, 2015

Likely: historic cohort

Unlikely

Unlikely (no blinding, but hard outcome measure)

Likely

Ou, 2015

Unlikely

Unclear (not described)

Unlikely (no blinding, but hard outcome measure)

Likely

Cooke, 2009

Likely

Likely (wide range in follow-up time)

Likely: no blinding

Unclear

Aihara, 2010

Likely: historic cohort

Unclear

Unlikely

Likely

Yagihara, 2012

Unlikely

Unlikely (not described)

Unlikely

Likely