Abstract
BACKGROUND: Home oxygen therapy is commonly prescribed for patients who remain hypoxemic at hospital discharge, although evidence supporting this practice is lacking. This study aimed to evaluate oxygen prescription and follow-up for patients who were prescribed post-discharge short-term oxygen therapy (STOT) and to assess their long-term outcome.
METHODS: A retrospective audit was undertaken of subjects prescribed STOT following hospitalization at a single site in Melbourne, Australia, between January 2011 and December 2015. During the study period, a designated clinical pathway for STOT prescription and follow-up after hospital discharge was in place. Chart review was performed to collect subject demographics and comorbidities, results of oxygen assessment (arterial blood gas and 6-min walk tests) and prescription, and results at follow-up re-assessment and mortality.
RESULTS: Over five 5 years, 205 subjects were prescribed STOT upon hospital discharge. Common indications for oxygen treatment were chronic lung disease (54%) and dyspnea palliation (26%). Of the 152 subjects who were discharged with non-palliative oxygen therapy, 28% did not fulfil the recommended prescribing criteria or did not have recommended assessments. Among the 118 subjects who returned for re-assessment 4 weeks after initial oxygen provision, 47 (40%) did not fulfill criteria for long-term oxygen therapy. The 1-y cumulative survival rate for the study population was 56%.
CONCLUSIONS: A significant proportion of subjects who were prescribed post-discharge STOT did not fulfill the recommended prescribing criteria. The long-term prognosis for subjects who were prescribed post-discharge STOT was poor.
Introduction
Hypoxemia is common in patients hospitalized with a range of acute illnesses, but it usually resolves within days to weeks of treatment of the underlying disorder. Short-term oxygen therapy (STOT) is often prescribed for patients who remain hypoxemic at hospital discharge, although there are no studies to support this practice. As a consequence, criteria for assessing eligibility for home oxygen therapy during hospitalization are not standardized and are often adapted directly from those used for determining eligibility for long-term oxygen therapy. Although the benefits of long-term oxygen therapy have been established in patients with stable COPD and resting hypoxemia, its clinical benefits in patients with hypoxemia following hospitalization are uncertain. In addition to its cost, oxygen therapy may cause significant adverse effects including oxygen toxicity, physical limitations with decreased mobility and increased falls risk, social stigma, nasal irritation, and fire hazards.
The Fifth Oxygen Consensus Conference recommended that patients discharged with STOT following hospitalization be assessed for continued eligibility for long-term oxygen therapy within 90 d of discharge.1 Current guidelines from the Thoracic Society of Australia and New Zealand and the British Thoracic Society recommend reassessment within a similar timeframe.2,3 Previous studies reported that up to 58% of subjects who were prescribed STOT at hospital discharge did not fulfill criteria for long-term oxygen therapy during follow-up assessment.4–7 It is also noted that relatively few subjects (35–65%) underwent appropriate reassessments for eligibility for long-term oxygen therapy following hospital discharge.4,8,9 Post-discharge STOT has been identified by the American Thoracic Society and American College of Chest Physicians as one of the top 5 areas in adult pulmonary medicine requiring improved delivery of care.10
A clinical pathway is used at our institution for the assessment and management of patients referred for post-discharge STOT. We performed this audit to evaluate the use of the clinical pathway in the assessment, management, and follow-up of subjects discharged with STOT following hospital admission. In addition, we examined predictors for the ongoing requirement of long-term oxygen therapy and the prognosis for subjects who were prescribed post-discharge STOT.
QUICK LOOK
Current knowledge
Short-term oxygen therapy is commonly prescribed for patients who remain hypoxemic at hospital discharge, although there is a lack of evidence supporting this practice. Previous studies reported a significant proportion of subjects who no longer met eligibility criteria for long-term oxygen therapy at reassessments, and 35-65% failed to attend follow-up appointments.
What this paper contributes to our knowledge
Despite the use of a clinical pathway, 30% of subjects discharged with short-term oxygen therapy did not have appropriately detailed oxygen assessments or did not meet recommended prescribing criteria. Optimal medical care was not offered to subjects with COPD. The long-term prognosis for subjects who were discharged with short-term oxygen therapy was poor.
Methods
A retrospective audit of subjects prescribed STOT following hospitalization at Austin Health, a tertiary Victorian hospital, between January 2011 and December 2015 was performed. This study was approved by the Austin Health Human Research Ethics Committee (LNR/17/Austin/352).
At Austin Health, a standardized referral form to the oxygen therapy clinic is used to prescribe STOT at hospital discharge. A routine follow-up appointment at the oxygen therapy clinic after 4 weeks post-discharge is scheduled once the referral has been processed. Prescribing criteria for STOT at hospital discharge are based on the guidelines from the Thoracic Society of Australia and New Zealand for long-term oxygen therapy (Table 1).3 Similar prescribing criteria are used for prescription of ongoing long-term oxygen therapy and exertional oxygen therapy at the follow-up assessments. Education regarding oxygen therapy is provided by a nurse educator to all patients prior to discharge, and oxygen equipment is delivered to the discharge destination by the designated oxygen service provider.
Prescribing Criteria for Home Oxygen Therapy
Data collected from the hospital-scanned medical record included patient demographics, referring medical units, diagnosis, types of oxygen therapy prescribed (ie, continuous, exertional, nocturnal, or palliative oxygen therapy), major comorbidities, details of initial and follow-up oxygen assessments (ie, 6-min walk test, arterial blood gases, sleep study), lung function tests (ie, spirometry, gas transfer), length of hospital stay, and mortality. For subjects with COPD, additional information on medical therapies and participation in pulmonary rehabilitation were collected.
Statistical analyses were performed using Graphpad Prism (v5, Graphpad Software, San Diego, California) and SPSS (v23, IBM, Armonk, New York). Categorical variables were expressed as absolute number (frequencies). The Fisher exact test was used to compare frequencies between groups. Data distributions were tested for normality using the Kolmogorov-Smirnov test. Parametric distributions were analyzed with t tests for comparisons of the 2 groups. For nonparametric data, the Mann-Whitney test was used for 2-group comparisons. Statistical significance was accepted at P < .05. Potential predictors for requirement of long-term oxygen therapy at follow-up assessments were evaluated using univariate logistic regression analyses. Selection of variables was based on factors considered to predict persistent hypoxemia and long-term requirement for oxygen therapy, including age, gender, co-existing cardiorespiratory comorbidities, smoking history, lung function, baseline oxygen assessment results (ie, 6-min walk distance and nadir saturation, arterial blood gas), evidence of chronic hypoxemia, and number of hospitalizations in the preceding 12 months. Kaplan-Meier analyses were performed to estimate subject survival.
Results
A total of 205 subjects were included in the analysis. Baseline characteristics of the subjects are shown in Table 2. The study population was predominantly male (58%) with a median age of 75 y with an interquartile range (IQR) of 68–83. The majority of subjects were admitted under respiratory medicine (29%), general medicine (20%), and medical oncology (20%). The median length of hospital stay was 10 d (IQR 7–18). Thirty-four (17%) subjects were re-admitted within 6 weeks following discharge from the initial hospitalization. Of the 205 subjects, 181 subjects were prescribed continuous oxygen therapy (of whom 135 [75%] were also prescribed exertional oxygen therapy), 22 subjects were prescribed exertional oxygen therapy only, and 2 subjects were prescribed nocturnal and exertional oxygen therapy.
Baseline Subject Characteristics
Indications for Oxygen Therapy
The most common indication for STOT at hospital discharge was chronic lung disease (111 subjects: COPD = 70 subjects; interstitial lung disease = 33 subjects; dual lung pathologies = 5 subjects; bronchiectasis = 3 subjects). Post-discharge oxygen therapy was requested for 53 (26%) subjects for palliation (malignancy = 43 subjects; congestive cardiac failure = 5 subjects; COPD = 4 subjects; interstitial lung disease = 1 subject). Other indications for STOT included congestive cardiac failure (16 subjects), chronic hypoventilation syndrome (10 subjects), co-existing COPD and congestive cardiac failure (5 subjects), acute respiratory illnesses (pulmonary embolism = 3 subjects; pneumonia = 2 subjects), hepatopulmonary syndrome (3 subjects), and pulmonary arterial hypertension (2 subjects).
In-Patient Assessments
Of the 152 subjects who were discharged with nonpalliative oxygen therapy, 127 (84%) had appropriate oxygen assessments during hospitalization (Table 3). Both subjects who were prescribed nocturnal oxygen therapy underwent sleep studies during hospitalization for acute on chronic respiratory failure after medical stabilization. Table 4 shows the proportions of subjects who did not meet recommended prescribing criteria for different types of oxygen therapy. Overall, 43 (29%) subjects who were prescribed STOT did not fulfill recommended prescribing criteria (18 [12%] had no significant resting hypoxemia) or did not have recommended assessments (23 [15%] had no arterial blood gas tests, and 2 [1%] did not perform a 6-min walk test).
Proportion of Subjects Prescribed Home Oxygen Therapy at Hospital Discharge Who Had Appropriate In-Patient Assessment
Proportion of Subjects Prescribed Home Oxygen Therapy at Hospital Discharge Who Did Not Fulfill the Recommended Prescribing Criteria
Follow-up Assessments
Of the 190 subjects who survived > 30 d after hospital discharge, 118 (62%) appeared for follow-up assessment. The median time between hospital discharge and follow-up was 45 (IQR 38–60) d. Of the subjects who were prescribed post-discharge STOT for dyspnea palliation and survived > 30 d (42 subjects), only 6 (14%) had re-assessment. There were no significant differences in the attendance rates for follow-up assessments among subjects with various cardiopulmonary diseases. Compared to subjects who did not have follow-up assessments, those who underwent review were more likely to have baseline evidence of sequelae of chronic hypoxemia, including polycythemia, pulmonary hypertension, and right heart failure (55% vs 23%, P < .001), and they had a lower gas transfer factor (diffusing capacity of the lung for carbon monoxide 42% predicted [IQR 36–48] vs 48% predicted [IQR 39.5–55.5], P = .009).
Among the 118 subjects who attended the follow-up assessments, 47 (40%) no longer fulfilled criteria for long-term oxygen therapy; 51 (43%) subjects continued to fulfill criteria for continuous oxygen therapy, and 20 (17%) fulfilled the criteria for exertional oxygen therapy only. In the univariate logistic regression analyses, there was a significant association between lower PaO2 values at hospital discharge and ongoing requirements for long-term oxygen therapy (Table 5).
Univariate Logistic Analysis of Potential Baseline Predictors for Fulfilling Criteria for Long-Term Oxygen Therapy at Follow-up Assessments
Management of COPD
COPD (n = 75) was the most common primary diagnosis in this study population. Among the subjects with a primary diagnosis of COPD, 61 (81%) were treated with inhaled corticosteroids and dual bronchodilators, and 4 were discharged without any inhaler therapy. The remaining subjects were treated with a combination therapy of inhaled corticosteroids and long-acting beta agonists (n = 6) or long-acting muscarinic agonists only (n = 4). Only 29 subjects (39%) were referred for pulmonary rehabilitation.
Mortality
Fifteen subjects died within 30 d after hospital discharge. Figure 1 shows the overall cumulative survival rate for the study population. The overall 1-y cumulative survival rate for subjects discharged with STOT was 56%. For subjects who were prescribed nonpalliative oxygen therapy, the 1-y cumulative survival rate was 61%. The median post-discharge survival duration for subjects who were prescribed palliative and nonpalliative oxygen therapy was 49 d and 984 d, respectively.
Kaplan-Meier analysis for survival of the study population over time.
Discussion
This study highlights issues inherent with STOT prescription at hospital discharge. Detailed oxygen assessments were often incomplete, and adherence to recommended prescribing criteria was suboptimal. A significant proportion of subjects failed to attend their review appointments for reassessment for long-term oxygen therapy eligibility. Among those who attended the follow-up review, 40% no longer fulfilled criteria for long-term oxygen therapy. Apart from underlying chronic lung diseases, palliation of dyspnea (predominantly relating to malignancy) was a common indication for prescribing STOT following hospitalization. The majority of subjects with COPD who were discharged with STOT were not referred for pulmonary rehabilitation. The long-term prognosis for subjects who were prescribed STOT at hospital discharge was poor.
There is a lack of evidence concerning the need for STOT at hospital discharge in the presence of continuing hypoxemia following an acute illness; however, it is a common practice due to clinical concerns about the potential risks associated with ongoing hypoxemia. There is accumulating evidence that hyperoxia due to oxygen therapy may lead to worse clinical outcomes in acute illnesses where physiological intuition might have suggested a beneficial response. Hyperoxia has been shown to be associated with increased mortality in patients with cardiac arrest,11,12 stroke,13 acute myocardial infarction,14 and septic shock.15 Breathing 30–60 min of supplementary oxygen at 28% has been shown to lead to elevated systemic and airway markers of oxidative stress and inflammatory cytokines in both healthy volunteers and patients with COPD.16,17 In addition, hyperoxia has been shown to be associated with acute lung injury and bacterial dissemination in animal models.18–21 Further research to assess the effects of home oxygen therapy in patients with recent acute illnesses is needed to guide our practice.
Despite the presence of a designated pathway for assessment and follow-up of subjects discharged with STOT following hospitalization, in-hospital assessments lacked essential data or did not adhere to recommended prescribing criteria in a significant proportion of subjects. Approximately 17% of subjects did not undergo a baseline arterial blood gas measurement prior to discharge on continuous oxygen therapy, and 13% did not exhibit sufficient hypoxemia to warrant such therapy. This suggests that more attention should be paid to educating in-patient medical teams about the need for adequate assessment and prescription to avoid inappropriate use of STOT. Given that a significant proportion of subjects did not meet eligibility criteria for long-term oxygen therapy at reassessment, patient education should also be provided prior to hospital discharge to emphasize the likely short-term nature of home oxygen therapy post-discharge and the importance of reassessment. It is known that patients can develop psychological dependence on oxygen therapy.22 Realistic expectations of the effects of STOT and the likely potential withdrawal of the therapy after reassessment should be discussed.
Although hospital readmission rates in patients with exacerbations of COPD have been reported to be reduced with the use of oxygen therapy as part of a post-hospitalization multicomponent care program,23 there are conflicting results from retrospective studies on the impact of long-term oxygen therapy on hospitalization rates.24–27 Nevertheless, some health care professionals perceive that oxygen therapy may prevent potential hospital readmissions by improving recovery or managing relapses at home. It is possible that the use of home oxygen therapy may result in increased health care utilization for support or due to falls or oxygen toxicity as a result of its use. It is important to consider other management methods to optimize patient care following hospitalization. Pulmonary rehabilitation following exacerbation of COPD has been shown to improve patients' health-related quality of life and exercise capacity, with potential to reduce hospital readmissions.28 However, only 39% of subjects with COPD in this study were referred for pulmonary rehabilitation.
Post-discharge STOT was commonly prescribed for palliation of symptoms. This finding is in agreement with results from previous qualitative studies on physicians' perspectives on oxygen therapy.29 Although oxygen therapy has been shown to relieve breathlessness during laboratory-based exercise tests in patients with COPD who were mildly or non-hypoxemic, it has not been shown to improve symptoms or quality of life in the home setting.30 A previous randomized controlled trial found that palliative oxygen was not superior to room air for relieving dyspnea in subjects with life-limiting illness and refractory dyspnea without significant hypoxemia.31 More importantly, oxygen therapy poses significant physical and psychological burdens on both the patients with life-limiting illnesses and their caregivers.32,33 A detailed discussion of the potential effects of oxygen therapy should be provided to patients and their carers prior to prescribing. The use of alternative pharmacologic and nonpharmacologic strategies should be considered for managing refractory breathlessness in this group of patients.
The prognosis for subjects who were discharged with STOT following hospitalization was poor. The overall 1-y survival rate of 61% for the nonpalliative subjects was comparable with previously reported data for subjects hospitalized for heart failure and COPD, which ranged from 56% to 67%34–37 and from 55% to 78%,38–42 respectively. A cohort study of a sample of Medicare beneficiaries > 65 y old from the United States reported 3-y mortality rate of 39.5% for those who were admitted to the ICU during hospitalization.43 Our results suggest that patients prescribed with STOT at hospital discharge are at high risk of poor outcomes. This might be due to a high rate of multimorbidity, as noted in our study.
This study has some limitations. This study was conducted in a single institution. Nonetheless, all subjects managed by different medical units within the institution were included for analysis. Due to the retrospective nature of the study, there were missing data due to incomplete assessments. This highlighted the deficits in the current management pathway for this group of patients. Given the lack of standardized criteria for prescribing STOT at hospital discharge, our findings may not be generalizable to other hospitals. However, the prescribing criteria we use for STOT following hospitalization are commonly used at other institutions. In addition, rates of subjects in this study who both failed to attend for reassessment and no longer fulfilled prescribing criteria for long-term oxygen therapy at reassessment are comparable to those of previous studies. This suggests that our study population may be representative.
Conclusions
This study highlighted significant issues with the prescription and management of STOT for patients with hypoxemia at hospital discharge. Adherence to the clinical pathway for assessment and prescription of STOT following hospitalization was poor. There is a clear need to prospectively evaluate the effects of STOT in patients with hypoxemia following hospitalization to guide clinical practice. Strategies for educating clinicians and patients regarding the roles of home oxygen therapy following hospitalization should be implemented to improve both patient care and health care delivery.
Footnotes
- Correspondence: Yet Hong Khor MD, Department of Respiratory and Sleep Medicine, Austin Health, 145 Studley Road, Heidelberg 3084 VIC, Australia. E-mail: yethong.khor{at}austin.org.au.
Dr Khor has disclosed relationships with Air Liquide, Boehringer Ingelheim, and the National Health and Medical Research Council. Dr McDonald has disclosed relationships with Boehringer Ingelheim, Novartis, GSK, Pfizer, Menarini, and Air Liquide. Dr Wong has disclosed no conflicts of interest.
- Copyright © 2019 by Daedalus Enterprises
