Globally, chronic obstructive pulmonary disease (COPD) is associated with increased morbidity, mortality, hospitalization, impaired quality of life, and increased healthcare burden [1-3].
Acute exacerbation of COPD (AECOPD) is a common cause of hospital admission for patients with COPD, representing a major cost implication for the healthcare system [4]. AECOPD admission has been listed as one of the 10 leading causes of hospital admission in the United States [5]. A nationwide audit in the United Kingdom showed that about one-third of hospitalized COPD patients were readmitted within 90 days, with a mortality rate of 15.3% [6]. Based on projection model in 12 Asian countries, it was estimated that 56.6 million people with aged 30 years and above were suffering from moderate to severe COPD [7]. The overall prevalence of COPD in these Asian countries was 6.3%, and ranged from 3.5% in Hong Kong and Singapore to 6.7% in Vietnam. The World Health Organization (WHO) projected COPD to be the third main cause of death worldwide by the year 2030, following ischemic heart disease and cerebrovascular disease [8].
Despite increasing evidence on the impact and factors related to AECOPD on mortality and hospital readmission in Malaysia, where multiple ethnicities and different healthcare systems may provide varying insights into this issue, there is no available data. This includes data that evaluates the disease complications and prognosis among patients after exacerbation discharge. This information would be useful in providing better patient care and in planning treatment strategies and financial budgeting that are related to hospital admission of these patients. Therefore, the present study was designed to identify the rate and prognostic factors that affect poor outcomes (mortality and readmission) among patients hospitalized for AECOPD.
A historical retrospective cohort study was conducted at the biggest tertiary public hospital in Penang, in the northern region of West Malaysia. Ethical approval for this study and review of patient records was obtained from the Medical Research and Ethics Committee of the Ministry of Health, Malaysia (approval No. NMMR-12-513-12643).
The study population were identified using computerized patient lists from the records office and were confirmed using the admission books from the wards. Data on inpatient stays were obtained manually from patient medical records and outpatient clinic records. Mortality data were obtained from the National Registration Department of Malaysia until the end of November 2013.
All patients (≥40 years old) who were hospitalized with a primary diagnosis of AECOPD over a 5-year period between January 2007 and December 2011 were retrospectively included in the study. The first hospital admission with AECOPD during the study period was considered as the index hospitalization or baseline. Hospitalization for AECOPD was defined as admission with a primary diagnosis of AECOPD (International Classification of Diseases, tenth revision [ICD-10] code J44.1). AECOPD was defined as a change in baseline patient symptoms, beyond variation and requiring change in regular medications or increase in dose based on the Global Initiative for Chronic Obstructive Pulmonary Disease (GOLD) guidelines [9].
Patients, who died during index hospitalization, were transferred to other hospitals, or were untraceable in medical records from the record office or clinics, were excluded, as were patients with a primary diagnosis for hospitalization other than AECOPD.
The demographic and clinical data were collected at baseline. The Charlson comorbidity index was used to quantify comorbidities among patients [10]. Other comorbidities, which are not included in the Charlson index were documented separately. Oral corticosteroid use was defined as taking at least 5 mg prednisolone daily or equivalent within the last month before index hospital admission. Active smoking was defined as having smoked at least one cigarette per day within the last 6 months.
Data were presented as median and interquartile range (IQR) for continuous variables and as percentage for categorical variables. Time intervals to mortality or first rehospitalization for AECOPD were plotted using Kaplan–Meier estimates [11]. Time to death was censored at the end of follow-up period on 30 November 2013. The time to rehospitalization for AECOPD was censored on 31 December 2012, if not, until the time of non-COPD readmission, or time of death. To explore the prognostic factors of outcomes, univariate analysis was performed using a log–rank test. Variables that were significant in the univariate analysis and other potentially clinically important variables were used as covariates in multivariable Cox proportional hazards model to test the independent effect of each predictor for both mortality and readmission [12]. A two-sided
During the 5-year period between January 2007 and December 2011, 461 patients with a primary diagnosis of AECOPD were identified. Of these, 139 patients were excluded (
Characteristics of patients admitted with acute exacerbation of chronic obstructive pulmonary disease (COPD)
Characteristics | Frequency, n (%) |
---|---|
Male | 240 (92.7) |
Female | 19 (7.3) |
Chinese | 124 (47.9) |
Malay | 86 (33.2) |
Indian | 47 (18.1) |
Others | 2 (0.8) |
Family | 230 (88.8) |
Living alone | 23 (8.9) |
Nursing home | 6 (2.3) |
Married | 222 (85.7) |
Single | 28 (10.8) |
Divorced | 8 (3.1) |
Widowed | 1 (0.4) |
Active smoker | 123 (47.5) |
Former smoker | 136 (52.5) |
Smoking history (pack/years) [Median (interquartile range)] | 20 (10–35) |
Alcoholic | 29 (11.2) |
Previous COPD hospitalization | 40 (15.4) |
Previous non-COPD hospitalization | 58 (22.4) |
Long-term oxygen therapy at home | 9 (3.5) |
Oral corticosteroid before admission | 10 (3.9) |
The median age for the study sample was 70.1 years (IQR = 59.8–76.9 years), 92.7% were male, and the 76.1% were elderly (≥60 years). The Chinese ethnic group represented nearly half of the cases, followed by one third being Malays. Nearly half (47.5%) were active smokers.
The median length of hospital stay (LOS) was 4 days (IQR = 3–6 days). Only 10.8% needed ventilation support. Patients ≥60 years old were more likely to be ventilated than younger patients (11.7% vs 8.1%,
During hospitalization, almost all patients were treated with a combination of β2 agonist and anticholinergics (93.1%). Half (49%) were treated with a short acting β2 agonist (SABA), while inhaled corticosteroids were used in only 31% of them. Systemic corticosteroids, antibiotics and xanthine derivatives were used by 83.8%, 78% and 76.8% of the patients, respectively. Regular medications at discharge included a combination of β2 agonist and anticholinergics (81.1%), xanthine derivatives (66.7%), SABA (54.3%), inhaled corticosteroids (31%), and inhaled anticholinergics (21.3%). Oral corticosteroids were prescribed for 47.7% and antibiotics for 33.3% of the patients upon discharge.
Readmission to hospital because of AECOPD was common. The risks of readmission at 1, 3, and 5-years were 35.1%, 43.2%, and 48.6%, respectively. During the study period, 24.7% of the cohort was readmitted only once, 10% twice, 6.6% thrice, 3.5% readmitted 4 times, and 5.4% readmitted 5 times or more. The overall readmission rate was highest for the elderly (52.8%) compared with younger (35.5%) patients (
Univariate analysis performed by stratified Kaplan–Meier curves using a log–rank test showed that being elderly (age ≥60 years), being ventilated during index admission, long-term oxygen therapy (LTOT) upon discharge, and having a history of 2 or more hospital admissions because of AECOPD in the previous year of baseline, were significant factors associated with hospital readmission (
Univariate analysis and multivariate Cox proportional hazard analysis for independent prognostic factors associated with readmission
Prognostic factors | Univariate | Multivariate | |
---|---|---|---|
HR (95% CI) | |||
Age ≥60 years | 0.02 | 1.78 (1.12-2.85) | 0.02 |
LTOT at discharge | 0.001 | 3.63 (1.98-6.95) | <0.001 |
History of ≥2 COPD admission in the previous year | 0.009 | 2.92 (1.47-5.80) | 0.002 |
Ventilation during Index admission | 0.001 | 0.19 (0.007-0.47) | <0.001 |
LTOT = long-term oxygen therapy, HR = hazard ratio, CI = confidence interval
Patients who had a history of ≥2 COPD admissions and being discharged with LTOT had a nearly 3 to 3.5 times likelihood of being readmitted at any time compared with patients without these factors. The risk of readmission among the elderly was nearly two times greater than those below 60 years (
During the follow up period, 157 patients died, mainly because of respiratory failure (41.4%). This was followed by old age (22.9%), cardiovascularly-related conditions (16.6%), septicemia (10.8%), lung cancer (3.8%), and other causes (4.5%).
The overall mortality rates after 1, 3, and 5 years were 26.3%, 49.8%, and 59.5%, respectively. The median survival time in the cohort was 1,087 days (2.97 years). The estimated median survival among patients who died during the study was 467 days (IQR = 129–923.5).
Univariate analysis conducted using stratified Kaplan–Meier curves using a log–rank test identified old age (≥60 years), use of LTOT, antiplatelet agents, diuretics, and nebulizer bronchodilator use upon discharge as independent prognostic factors that affect mortality after hospital discharge. Although discharge with the use of a nebulizer for respiratory medications was significant in univariate analysis (
Prognostics factors for mortality according to univariate and multivariate Cox regression analysis
Prognostic factors | Univariate | Multivariate | |
---|---|---|---|
HR (95% CI) | |||
Age ≥60 years | <0.001 | 3.63 (1.98 -6.95) | 0.006 |
LTOT at discharge | <0.001 | 2.78 (1.54 -5.02) | 0.001 |
Charlson comorbidity score ≥4 | 0.02 | 0.72 (0.36 -1.35) | 0.31 |
Antiplatelets at discharge | 0.02 | 1.26 (0.81 -1.97) | 0.30 |
Diuretics at discharge | 0.02 | 1.36 (0.80 -2.32) | 0.26 |
LTOT = Long-term oxygen therapy, HR = hazard ratio, CI = confidence interval
The present study is one of the first to our knowledge to systematically study the poor outcome among patients admitted to hospital with AECOPD in Malaysia. It confirms data from the previous studies on the risk of mortality and readmission after hospital discharge. The risk was high at one year and continued to increase annually.
The rate of hospital readmission in this study was within the range of other published studies, between 25% and 79.9% for one year [13-19] and 44% and 56% for 5 years [18, 20]. Similar increasing trends were observed for mortality rates ranging from 21% to 49.1% for one year [14, 18, 20-26] and 38.2% to 75.9% for five years [18, 20, 25, 27]. The wide range of hospital readmission and mortality rates observed in these studies may be the result of the different
methods used (study design, population, data collection, selection criteria) and severity of the diseases among patients included in the study.
Several predictors for poor outcomes were identified. Old age (≥60 years) and the use of LTOT were identified as independent factors that increased both mortality and readmission. Frequency of COPD readmissions and ventilation use during hospitalization were found as prognostic factors for readmission alone.
Aging has been identified as a risk factor for both mortality [14, 18, 27-30] and readmission [18, 31] in previously published studies. Age-related physiological changes, particularly in the lung, might explain this outcome. Reduction in respiratory muscle strength, vital capacity, and total alveolar surface can accelerate pulmonary function deterioration among the elderly with COPD who already had loss of lung function because of the disease itself [32, 33]. Some studies [30, 31] showed advanced age as a predictor for a long hospital stay, but not readmission, while others [34] reported it as a predictor for a shorter time to the first readmission after discharge.
In line with our results, LTOT was associated with short [29] and long-term [35] mortality after discharge and readmission [13]. Another study reported that patients using LTOT had a shorter time to the first readmission after discharge (
The need for ventilation during the index of admission was found to be a protective factor for readmission in the current study. This could be the result of the relatively small number of patients who were ventilated (n = 28) and probability of being readmitted to other hospitals. Usually, the need for ventilation is considered a marker for the severity of exacerbation and life threatening episodes of acute respiratory failure [38]. Our study showed that the majority (67.9%) of patients being ventilated died during the study period, and more than half (53.6%) died without hospital readmission, which may also explain the lower rate of hospital readmission.
Although nebulizer usage was identified as a factor that affected mortality after discharge in univariate analysis, it was not retained in the multivariate model because of the small number of patients using it. A study conducted in 5 Nordic countries reported that nebulized bronchodilator usage upon discharge was associated with a higher risk of mortality in bivariate, but not in multivariate analysis [39]. Another study supported the previous result by reporting that patients using home nebulizers had experienced a higher risk of readmission (HR, 1.7; 95% CI, 1.4-1.9) [31]. The home nebulizer is usually used among patients who are suffering from shortness of breath despite getting maximum inhaler therapy [40], or who are unable to use other devices because of cognitive or physical impairments [41]. Hence, nebulizer use can be considered as a sign of poor disease or general patient condition.
Some nonrespiratory medications at discharge such as antiplatelet agents and diuretics were found to be predictors for mortality. The use of these medications may reflect the existence of other comorbidities such as hypertension, pulmonary edema, and cardiovascular diseases among patients. However, this association was not found to be statistically significant in the multivariate analysis. The use of certain drugs, such as anticholinergics has been shown to increase the risk of readmission [39], whereas inhaled corticosteroids and/or long acting β-2 agonists decreased the risk of mortality after discharge among COPD patients [28].
Our study showed patients with Chinese ethnicity had a higher rate of mortality and readmission than Malays or Indians, but the difference was not significant. This could be a consequence of the higher number of Chinese patients included in the present study. A study conducted in Singapore reported that patients with Malay ethnicity were found to be more likely to have frequent readmission [17]. However, the finding was not significant. In a large cohort study, nonwhite ethnicity was identified as a factor that was associated with a decrease in the risk of mortality and readmission [18]. It should be noted that studies investigating the association of poor outcomes among COPD patients and ethnicity are very limited.
We found that a high score on the Charlson index (≥4) may increase the risk of mortality only in univariate analysis. Similarly, comorbidities or the Charlson index score were not significant factors for readmission [13, 14, 17], admission [42], or mortality [14, 24, 25] in several studies. In some, the Charlson score was a risk factor for mortality [22, 29], or readmission [29].
The present study has several limitations such as its retrospective nature, lack of information regarding COPD and severity of dyspnea, and some missing data on spirometry, blood gases, and body mass index, these therefore being excluded in the final analysis. The small number of patients being ventilated was another limitation. Moreover, readmission rates might have been underestimated if patients had been readmitted to smaller peripheral hospitals. Therefore, an extensive clinic record review was performed to capture and ensure accuracy of findings.
The present study demonstrated a high rate of mortality and readmission among patients discharged after AECOPD. Elderly patients and those using LTOT had a shorter time for mortality and readmission at any time after hospital discharge. This study has identified frequent hospital admissions because of AECOPD as a prognostic factor that increases the risk of further hospital readmission. It may provide new insight to clinicians regarding the status of patients admitted because of AECOPD. This will help them to stratify patients based on their risk of developing poor outcomes. Once they are stratified as having a high risk for poor outcome, the decision on regular monitoring through frequent clinic follow-ups can be planned. This study may lead to a clearer understanding of the future work that might now be needed, such as designing an interventional study as a follow-up to this work.