TABLE 3

Studies that met the inclusion criteria and were reviewed

First author (year) [ref.]Study type and characteristics of study populationPatients nEvaluation of depressionReadmissionsMortalityFollow-up
Abrams (2011) [22]Retrospective study of mostly male (97%) veterans from 2006 to 2008 hospitalised for AECOPD26 591Prevalent in 11.6% (n=3077)In a 30-day period 10.4% (n=319) with prevalent depression versus 11.6% (n=2723) with no depressionIn a 30-day period 5.3% (n=163) with prevalent depression versus 3.8% (n=881) with no depression30 days
Regvat (2011) [45]Prospective study of patients hospitalised for AECOPD
Patients with prevalent psychiatric disorders, including depression, were excluded		50Evaluation of depression through PRIME-MD questionnaire on day of discharge; 42% (n=21) indicated depressionn/an/an/a
Pumar (2014) [8]Reviewn/an/an/an/an/a
Singh (2016) [48]Retrospective cohorts of patients hospitalised for AECOPD in 2001–2011 and aged >66 years80 088 patients with a total of 135 498 hospitalisationsPrevalent in 14.24% of hospitalisations (n=19 293)In a 30-day period 24.15% (n=4659) with prevalent depression versus 16.91 (n=19 645) with no depressionn/a30 days
Ng (2014) [40]Prospective cohort study of patients hospitalised for AECOPD
Conference abstract		376HADS score ≥8 in 44.4% (n=167)n/aIn a 1-year period a hazard ratio of 1.93 (95% CI 1.04–3.58) for mortality was associated with baseline depression1 year
Cavaillès (2013) [29]Reviewn/an/an/an/an/a
Irwin (2015) [33]Prospective study of patients hospitalised for AECOPD
Patients with prevalent psychiatric disorders, including depression, were excluded		85PROMIS measurement of depression was positive in 44% (n=35)n/an/an/a
Salte (2015) [46]Reviewn/an/an/an/an/a
Aimonino (2007) [23]Prospective study of patients aged ≥75 years hospitalised for AECOPD from April 2004 to April 200516Geriatric depression scale
Mean±sd depression score of 12.8±5.4 at baseline and 12.68±n/a at discharge		n/an/an/a
Yıldırım (2013) [53]Prospective study of patients hospitalised for AECOPD
Patients with prevalent psychiatric disorders, including depression, were excluded		135HADS (Turkish translated version) score ≥8 in 85.6% (n=116)n/an/an/a
Valenza (2014) [52]Prospective randomised cohort study of male patients hospitalised for AECOPD
Intervention with a controlled breathing programme		46Baseline mean±sd HADS score of 9.62±2.1 in the intervention group and 8.85±4 in the control groupn/an/an/a
Al Aqqad (2014) [25]Prospective study from January 2016 to June 2016 in Malaysia
Patients hospitalised for AECOPD and aged >60 years
Conference abstract		3745.9% showed symptoms of depression (no further clarification of test method); 5.4% categorised as severe and 40.5% as mild to moderaten/an/an/a
Pooler (2014) [44]Reviewn/an/an/an/an/a
Laurin (2012) [39]Reviewn/an/an/an/an/a
Papaioannou (2013) [43]Prospective study of patients hospitalised for AECOPD between March 2009 and June 2010 without prevalent depression230Beck depression inventory (Greek translation) score ≥19 in 39.57% (n=91) on the first day of hospitalisation84.6% (n=77) with 1.96±1.41 readmissions for depressive patients versus 28.1% (n=39) with 0.41±0.86 readmissions for nondepressive patients within 1 year25.3% (n=23) for depressive patients versus 3.6% (n=5) for nondepressive patients within 1 year1 year
Laurin (2011) [38]Reviewn/an/an/an/an/a
Ozyemisci-Taskiran (2015) [41]Retrospective study of patients hospitalised for AECOPD
Two hospitalised control groups: one with stable COPD and one without COPD		133 (AECOPD); 34 (stable COPD); 34 (without COPD)HADS (Turkish translation) score ≥8 in 39.1% (n=52); mean score 5 (Q1=2; Q3=11) for AECOPD
HADS score ≥8 in 14.7% (n=5); mean score 4 (Q1=0.25; Q3=6) for stable COPD
HADS score ≥8 in 29.4% (n=10); mean score 4.5 (Q1=1; Q3=8.25) for non-COPD		n/an/an/a
Almagro (2002) [27]Prospective study of patients (mostly male, 96%) hospitalised for AECOPD between October 1996 and May 1997135YDS was applied, but no direct data providedn/aPatients with YDS score ≥11 had a 3.11-fold higher mortality than patients with a YDS score ≤5 within 2 years2 years
Panagioti (2014) [42]Reviewn/an/an/an/an/a
Small (1992) [49]Prospective study of patients hospitalised for AECOPD26Profile of Mood States was applied
Mean±sd score for depression of 13.35±12.38 and median score 11.5
No cut-off value for depression was provided by the test		n/an/an/a
Koenig (2006) [36]Prospective cohort study of patients aged >50 years hospitalised for AECOPD between November 1999 and December 2003No data provided on overall rates of depression of screened patientsPatients were identified as depressive using the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders (4th edition)
410 patients were classified with minor deperession and 301 with major depression using the Hamilton Depression Rating Scale		n/an/aAt an average follow-up of 11.4 weeks minor depressions showed remission of 66.6% (n=273) and at an average follow-up of 21.3 weeks for major depressions 49.2% (n=148) showed remission
Jennings (2015) [34]Randomised trial of patients admitted with AECOPD between February 2010 and April 2013 and with smoking history ≥20 pack-years172 in total; (intervention group n=93; control group n=79)HADS was applied in the intervention group, but no baseline values are providedNo significant association between depression score per 1-unit increase and Readmission for 30-day follow-up (HR 1.11, 95% CI 0.96–1.27; p=0.157) and 90-day follow-up (HR 1.057, 95% CI 0.95–1.17; p=0.300)n/a30 days and 90 days
Silver (2010) [47]Cross-sectional study of patients hospitalised for AECOPD between January 2007 and December 2007
Conference abstract		69 841Prevalent diagnosis of anxiety/depression present in 27.3%
No seperated data provided		n/aOdds ratio for in-hospital mortality 0.93 (95% CI 0.81–1.06)n/a
Torres-Sánchez (2016) [51]Randomised, single-blind clinical trial of obese patients hospitalised for AECOPD49 (intervention group n=24; control group n=25)HADS mean±sd score at baseline: intervention group 6.0±2.86; control group 4.90±2.84
p=0.218		n/an/aFollow-up evaluation at discharge
Kim (2010) [35]Retrospective study of 77 patients hospitalised for AECOPD between January 2005 and May 200877Patients with prevalent diagnosis of anxiety/depression and patients taking psychotropic medication were considered to have anxiety/depression
No seperated data of anxiety/depression were provided
11.9% (n=9) of patients had anxiety/depression		Frequent exacerbators (n=35) (>1 hospitalisation within 1 year) showed 20% (n=7) anxiety/depression versus 4.8% (n=2) of single exacerbators (n=42)n/an/a
Almagro (2006) [26]Prospective study of patients hospitalised for AECOPD between October 1996 and May 1997129YDS was applied, but no general baseline data are provided
Score ≥5 is considered as cut-off for depression		YDS mean±sd score of 5±3.4 for patients who were readmitted within 1 year versus 3.7±3.1 for patients without readmission within 1 year
58.1% were readmitted within 1 year		n/a1 year
Gudmundsson (2005) [31]Prospective study of patients hospitalised for AECOPD during 2000–2001406HADS mean±sd score 5.5±3.6
HADS score ≥8 in 28.7%		60.59% (n=246) readmitted within 1 year versus 39.41% (n=160) without readmission in 1 year
No significant difference between groups in mean HADS score (5.4±3.4 versus 5.6±3.8; p=0.63) and depression level (26.5% versus 30.2%; p=0.44)		n/a1 year
Struik (2013) [50]Cross-sectional study of patients hospitalised for AECOPD between 2008 and October 2011 with GOLD III or IV disease and after 48 h of receiving invasive or noninvasive ventilation170HADS mean±sd score 7.8±sdn/an/an/a
Aimonino Ricauda (2008) [24]Prospective study of patients aged ≥75 years hospitalised for AECOPD between April 2004 and April 2005
The authors published a similar article including partly the same population in 2007 [27]		52Geriatric Depression Scale
Mean±sd depression score 17.2±6.8 at baseline and a mean±sd change of +0.7±3.2% at 6-month follow-up		n/an/an/a
Abrams (2010) [21]Retrospective cohort study of veterans hospitalised for AECOPD between October 2016 and September 2007
Conference abstract		23 306Pre-existing depression was assessed from data collected 1 year prior to admission
9.5% (n=2.216) had pre-existing depression		n/aUnadjusted 30-day mortality was higher in patients with depression (7.2% versus 4.6%)
Unadjusted 365-day mortality was similar in patients with depression (9.8% versus 9.1%)
Adjusted odds ratio of 30-day mortality with depression relative to those without 1.52 (95% CI 1.24–1.86); adjusted odds ratio of 365-day mortality with depression relative to those without 1.2 (95% CI 1.04–1.39)		n/a
Burr (2010) [28]Retrospective study of patients hospitalised for AECOPD in 2008
Two groups where defined: frequent exacerbators (>1 AECOPD event in 2008) and single AECOPD event in 2008
Conference abstract		Frequent exacerbators n=85; nonfrequent exacerbators n=89Depression prevalent in 42% of frequent exacerbators
No data provided for nonfrequent exacerbators		n/an/an/a
Aaron (2013) [20]Retrospective study of patients hospitalised for AECOPD in 2010
Conference abstract		388n/aThe study does not distinguish between depression and anxiety
29% of patients with prevalent depression/anxiety where readmitted within 1 year versus 16% of patients without depression/anxiety		n/an/a
Laçin (2011) [37]Study of patients hospitalised for AECOPD between January 2010 and February 2011
Conference abstract		114HADS score ≥8 in 53% (n=60), mean±sd HADS score 7.0±5.7n/an/an/a
Hasan (2011) [32]Study of patients hospitalised for AECOPD
Patients with pneumonia, heart failure, on long-term oral corticosteroids, any significant comorbid condition or active malignancy were excluded
Conference abstract		120HADS was applied within 2 days of hospitalisation
HADS scores ≥11 in 28% (n=34)		n/an/an/a
Franzen (2014) [30]Study of patients hospitalised between January 2012 and December 2012 in three public hospitals in Zurich canton for AECOPD
Conference abstract		94Diagnosis of depression was present in 17%n/an/an/a

AECOPD: acute exacerbation of chronic obstructive pulmonary disease; PRIME-MD: Primary Care Evaluation of Mental Disorders; HADS: Hospital Anxiety and Depression Scale; PROMIS: Patient-reported Outcomes Measurement Information System; YDS: Yesavage Depression Scale; GOLD: Global Initiative for Chronic Obstructive Lung Disease.