Table 2.
Data extraction table of all articles included in the systematic review
| Study (author/date) | Design | Sample size | Study setting | Length of follow-up | Participant characteristics | Outcomes/findings |
|---|---|---|---|---|---|---|
| Cohort studies | ||||||
| Alcazar et al33 | Cross-sectional observational (cohort) ×2 groups | n=127 | Spain hospital outpatients | 12 months | COPD, had/had not had admission in previous 12 months, >40 years of age | Univariate analysis; anxiety and depression significant predictors/factors for exacerbation/admission, but did not show in multivariate analysis; also related to lower BODE score; HADS used to measure anxiety and depression |
| Almagro et al26 | Prospective cohort study | n=135 | Spain, acute inpatient hospital | 3 years | All patients admitted with AECOPD October 1996–May 1997 | Those admitted >1 per year had increased depression and also increased mortality; depression associated with increased mortality (3× higher) as well as more frequent readmission; Yesavage scale26 used to measure depression |
| Chen and Narsavage51 | Longitudinal cohort study | n=145 | Six rural hospitals in Taiwan | 90 days (3 months) | Patients admitted for AECOPD; excluded those with psychiatric problems and included different cultural subgroups | Only significant factors were age and level of daily functioning; depression not significant; rural and cultural subgroups may have affected patients’ perceptions of their condition and depression, also QOL |
| Coventry et al37 | Prospective cohort study (frequent exacerbators versus infrequent exacerbators) | n=79 | Secondary care trusts in Manchester which had SED teams | 12 months | Participants were in SED follow-up system so had social support; excluded if had mental illness (self-selected group); investigator not blinded to their baseline psychological status | Depression significant factor leading to readmission; anxiety, depression, QOL deteriorated over 12-month follow-up period; low socioeconomic status also leads to increased readmissions; frequent exacerbators are more depressed; HADS used to measure anxiety and depression |
| Dalal et al60 | Retrospective cohort study using US claims data; COPD + anxiety/depression and COPD alone | n=31,483 | US | 2 years | Claims databases used to identify those who had been admitted and also had diagnosis of anxiety and/or depression and then followed | Higher ER and GP visits in those who had anxiety and depression, as well as COPD. Higher cumulative all cause costs and higher COPD-related costs |
| Eisner et al38 | Prospective cohort study (control group used) | n=1,202 | US, Kaiser Permanente (KP) Medical Care program | 1 year | Already on KP program in the US (care program) | Study concentrates on anxiety, not depression; increased disease severity linked to increased anxiety and increased exacerbations (exposure/response) relationship; dyspnea showed strongest association to anxiety; anxiety associated with younger age, females, nonwhite race, lower socioeconomic class, current smokers; anxiety also leads to increased perception of symptoms; (HADS, BODE, MRC SF12 used) |
| Fan et al47 | Prospective cohort study | n=610 | US (NETT trial; patients on medical strand) | 3 years; used clinical trial records to compare all-cause mortality | Had severe COPD; already on medical strand of NETT trial | 26.9% of participants were hospitalized; had ER visits for exacerbations; no association was found with depression and 1-year mortality, but it was for 3-year mortality; female sex associated with increased risk of hospitalization, but no significant relationship between female and depression; high prevalence of depression in severe patients, but only 37% receiving medication; neither depression nor anxiety associated with exacerbations or hospitalization; BDI, state trait anxiety inventory used 60.6% readmitted in 1 year; more likely to readmit if had decreased QOL, decreased lung function, on LTOT, had previous admissions; there was no significant difference between the group with anxiety and depression and the group without. Significant correlation between QOL and depression in relation to admission and anxiety; HADS; SGRQ |
| Gudmundsson et al19 | Prospective cohort study | n=406 | Nordic countries | 12 months (contacted 1 year after discharge re number of readmissions, also checked records) | Admitted >24 hours with AECOPD; moderate-to-severe COPD | |
| Jennings et al4 | Retrospective then prospective for 1 year after completing PR program | n=194 | US tertiary care referral center where PR took place | 12 months | All patients completed PR program with moderate-to-severe COPD | Patients with depression were more likely to have exacerbations and readmit (independent predictor); they were 2.8× more likely to have an exacerbation; females with COPD more likely to be depressed; (BDI, SF36, Charlson Comorbidity Index) |
| Ng et al45 | Prospective cohort | n=376 | Singapore ×2 secondary care hospitals | 12 months | FEV1 % predicted <70%; 20 pack-years; (self-report hospital admissions) | 44.4% depressed in total sample; more deaths after discharge in depressed group (21.5% v10.5%); increased LOS in depressed group; decrease SGRQ scores in depressed group; (HADS, SGRQ cut-off point for depression 8 on HADS used) |
| Quint et al42 | Prospective cohort study; (frequent versus infrequent exacerbators) | n=169 | London secondary care | 12 months | Taken from London COPD study. Had no exacerbations in the previous 12 months | Frequent exacerbators had higher baseline depression scores; females more depressed than men; depression significantly increased from baseline to exacerbation and admission. (CES-D, SGRQ, MRC) |
| Regvat et al43 | Prospective cohort study | n=50 | Slovinia Hospital based | 3 months (June–August, summer); interviewed on day of discharge | Successive patients admitted to a hospital with COPD exacerbation between two dates; physician diagnosed | Hospitalized patients showed a higher prevalence of anxiety and/or depression (50%); showed less improvement in their dyspnea; increased PaO2; decreased PaCO2; increased pH on admissions; ? hypersensitivity of receptors in respiratory center so sense of exacerbation felt much earlier and leads to panic (PRIME-MD: but did not do the questionnaire, just the interview section) |
| Xu et al44 | Multicenter prospective cohort study | n=491 | People’s Republic of China | 12 months monitored by telephone after discharge | >30 years of age, COPD diagnosis, no deterioration or change in meds or symptoms 4 weeks prior to start of the study | Depressed patients had a higher proportion of concurrent anxiety and more severe dyspnea; had more exacerbations and hospital admissions and lower levels of self-efficacy, social support, and QOL; depression identified in patients with stable COPD was significantly associated with a higher risk of exacerbations and hospital admissions (overall worse health profile) HADS used |
| Case control/case series studies | ||||||
| Almagro et al34 | Prospective observational (case series) no comparison group | n=129 | Spanish hospitals in Barcelona | 7 months | All those admitted with AECOPD between October 1996–May 1997 | Significant factors associated with readmission were increased PaCO2 and decreased QOL; depression a factor in bivariate analysis only not multivariate analysis (Yesavage scale used to measure depression) |
| Carneiro et al36 | Case series study | n=45 | Portuguese hospital | 66 weeks (16 months) | All those admitted with AECOPD between September 2006–January 2008 | Lower QOL related to depression, also lower FEV1; those who were depressed had more readmissions for AECOPD and longer LOS; 56% of study population had depression |
| Kim et al46 | Retrospective case series study | n=77 | Korea University hospital | 12 months | All those admitted to respiratory wards in the hospital over a 3-year period; notes examined for 12 months of data | Anxiety/depression not shown to be significant factors, but there may have been underdiagnoses due to study being retrospective therefore true numbers of cases not picked up; showed that underweight and hypercapnia on discharge were significant factors for frequent readmission with exacerbation of COPD |
| Laurin et al40 | Prospective follow up study (case series) | n=110 | Canada outpatients at two hospitals | 2 years | All patients on self- management program; all had an exacerbation previously | Patients with psychiatric disorder had significantly higher rates of exacerbation (outpatient); no difference between inpatient exacerbations; had exacerbations sooner on discharge, particularly outpatient ones; (psychiatric interview/anxiety disorders interview schedule, MRC) |
| Soler et al50 | Case control study (control group COPD but no admissions) | n=64 | Valencia, Spain secondary care hospitals | 12 months | FEV1 % predicted <50%, admission in previous 12 months; all male and all severe COPD; taken from a larger study looking at hospital care impact on COPD | QOL significantly poorer in those admitted to hospital; depression not mentioned, only anxiety; STAI-S/T- anxiety; inhaled salmetrol and cardiac arrhythmias seen as significant factors but they were all severe patients |
| Survey design | ||||||
| Abrams et al32 | Survey (review of electronic databases and records) | n=26,591 | US, Veterans administration hospitals | N/A; no face-to-face contact | Excluded patients not having any follow up after discharge. Mostly men, >50 years of age, numerous comorbidities as from Veteran association | Depression and anxiety significant predictors of both mortality and readmission (30 days). Also more likely to smoke if depressed so affect COPD severity (no tool to assess anxiety and depression identified) |
| Cao et al35 | Cross-sectional survey | n=186; decreased to 146 postdischarge | General hospitals in Singapore | 12 months | 50 years of age and only COPD; excluded those with the most severe COPD and also those with psychiatric disorders | Significant factors were long duration of having COPD >5 years; severity of COPD (FEV1 % <50%);consumption of antipsychotic drugs; depression; poor family support; no PR; HADS used to measure anxiety and depression |
| Qualitative design | ||||||
| Bailey48 | Focused ethnography and narrative analysis | Ten patients, ten nurses, 15 family caregivers | Two hospitals in mining community of Northern Ontario, Canada | 4 months; ×1 interview with each while in hospital | Extreme dyspnea > two previous admissions | Suggests that anxiety might provide an important indication of actual illness severity and assist nurses in determining the support and care that these patients require; essential that anxiety is recognized as an emic sign and not necessarily the cause of dyspnea for patients with COPD in acute exacerbations |
| Nicolson and Anderson49 | Focus groups | n=20 | UK ×3 GP surgeries in Sheffield | n/a | Diagnosed with chronic bronchitis only (excludes emphysema) | SGRQ; sense of loss, psychological distress, disruption; coping with limitations imposed on them; less autonomy which increased dependence; loss of control which leads to increased depression/decreased QOL |
| Maurer et al41 | Workshop of chest physicians (report) | n/a | US | n/a | Chest physicians | Depression undertreated; depression linked to decreased QOL which leads to decreased compliance; this leads to increased exacerbations, admissions, and also higher mortality risk on discharge following exacerbation |
| Mixed methods | ||||||
| Gruffydd-Jones et al39 | Prospective observational study and focus groups using semistructured interview | n=25; n=6 for focus groups with two groups of three | SW England; DGH with no specialist respiratory support post discharge (generic teams) | n/a | Patients admitted to DGH with AECOPD; purposive sampling for focus groups to reflect heterogeneity of severity of disease and geographical spread; mean age 76 years | HAD used; reluctance to seek medical help in an exacerbation; fear and anxiety associated with acute breathlessness; educational needs; follow-up after discharge/lack of planning; remaining depressed postdischarge and nothing done about it |
Abbreviations: AECOPD, acute exacerbations of chronic obstructive pulmonary disease; QOL, quality of life; NETT, National Emphysema Treatment Trial; HADS, Hospital Anxiety and Depression Scale; BODE, body-mass index, airflow obstruction, dyspnea, and dxercise; ER, emergency room; GP, general practitioner; LTOT, long-term oxygen therapy; SGRQ, St George’s Respiratory Questionnaire; FEV1, forced expiratory volume in 1 second; SED, supported early discharge; COPD, chronic obstructive pulmonary disease; LOS, length of stay; PaCO2, partial pressure of carbon dioxide; PaO2, partial pressure of oxygen; STAI, State Trait Anxiety Inventory; S/T, state/trait; MRC, medical research council; PR, pulmonary rehabilitation; BDI, Becks Depression Inventory; CES-D, Centre for Epidemiologic Studies-Depression scale; PRIME-MD, Primary Care Evaluation of Mental Disorders; DGH, district general hospital.