[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
Purchase Options:
[Skip to Content Landing]
Figure 1.
Recruitment, Randomization, and Retention of Participants
Recruitment, Randomization, and Retention of Participants

aPatients may have more than 1 reason for not being eligible for the study.

bData on hospitalizations, emergency department visits, and deaths were collected via medical record and vital statistics record review, and were analyzed for all study participants except for 1 participant in the intervention group who withdrew from the study shortly after hospital discharge due to moving out of state.

Figure 2.
Time to First Chronic Obstructive Pulmonary Disease–Related Acute Care Event (Hospitalization or Emergency Department Visit) or Death
Time to First Chronic Obstructive Pulmonary Disease–Related Acute Care Event (Hospitalization or Emergency Department Visit) or Death

The median time to first event for the intervention group is 122 days (95% CI, 78-180) compared with greater than 180 days for the usual care group (exact value is not computed because it is beyond the 6-month observation period). Cox proportional hazards model adjusted for hospital unit, age, oxygen use, and prior hospitalization.

Figure 3.
Cumulative Number of Chronic Obstructive Pulmonary Disease (COPD)–Related and All-Cause Events per Participant by Event Type and Study Group at 1, 3, and 6 Months After Discharge
Cumulative Number of Chronic Obstructive Pulmonary Disease (COPD)–Related and All-Cause Events per Participant by Event Type and Study Group at 1, 3, and 6 Months After Discharge

The boxes in the graphs show the median and interquartile range (IQR) of the data, with the bottom and top indicating the 25th and 75th percentiles, respectively; the upper whisker extends from the top of the box to the largest value no further than 1.5 times the IQR. The bottom whiskers extend from the bottom of the boxes to the smallest value no further than 1.5 times the IQR; outliers outside the whiskers range are also presented (dots). The circles and triangles indicate the mean number of events for usual care and the intervention, respectively. The black lines across the boxes indicate the median. Boxplots at each time point are staggered to avoid superimposition. Boxplots do not show when the 75th percentile of all data is zero. Whiskers do not show when all data points except for outliers are at zero.

aEmergency department visits that led to a hospitalization are not included in the emergency department visit counts.

Table 1.  
Study Participants’ Characteristics
Study Participants’ Characteristics
Table 2.  
Mean Change in Health-Related Quality of Life, as Measured by St George’s Respiratory Questionnaire, at 6 Months After Discharge by Study Groupa
Mean Change in Health-Related Quality of Life, as Measured by St George’s Respiratory Questionnaire, at 6 Months After Discharge by Study Groupa
1.
Kochanek  KD, Murphy  SL, Xu  JQ, Arias  E.  Mortality in the United States, 2016: NCHS Data Brief, No. 293. Hyattsville, MD: National Center for Health Statistics; 2017.
2.
Ford  ES, Croft  JB, Mannino  DM, Wheaton  AG, Zhang  X, Giles  WH.  COPD surveillance: United States, 1999-2011.  Chest. 2013;144(1):284-305. doi:10.1378/chest.13-0809PubMedGoogle ScholarCrossref
3.
Brault  MW, Hootman  J, Helmick  CG, Theis  KA, Armour  BS; Centers for Disease Control and Prevention (CDC).  Prevalence and most common causes of disability among adults: United States, 2005.  MMWR Morb Mortal Wkly Rep. 2009;58(16):421-426.PubMedGoogle Scholar
4.
Kumbhare  SD, Beiko  T, Wilcox  SR, Strange  C.  Characteristics of COPD patients using United States emergency care or hospitalization.  Chronic Obstr Pulm Dis. 2016;3(2):539-548. doi:10.15326/jcopdf.3.2.2015.0155PubMedGoogle Scholar
5.
Ospina  MB, Mrklas  K, Deuchar  L,  et al.  A systematic review of the effectiveness of discharge care bundles for patients with COPD.  Thorax. 2017;72(1):31-39. doi:10.1136/thoraxjnl-2016-208820PubMedGoogle ScholarCrossref
6.
Hansen  LO, Young  RS, Hinami  K, Leung  A, Williams  MV.  Interventions to reduce 30-day rehospitalization: a systematic review.  Ann Intern Med. 2011;155(8):520-528. doi:10.7326/0003-4819-155-8-201110180-00008PubMedGoogle ScholarCrossref
7.
Prieto-Centurion  V, Markos  MA, Ramey  NI,  et al.  Interventions to reduce rehospitalizations after chronic obstructive pulmonary disease exacerbations: a systematic review.  Ann Am Thorac Soc. 2014;11(3):417-424. doi:10.1513/AnnalsATS.201308-254OCPubMedGoogle ScholarCrossref
8.
Shah  T, Press  VG, Huisingh-Scheetz  M, White  SR.  COPD readmissions: addressing COPD in the era of value-based health care.  Chest. 2016;150(4):916-926. doi:10.1016/j.chest.2016.05.002PubMedGoogle ScholarCrossref
9.
Pedersen  PU, Ersgard  KB, Soerensen  TB, Larsen  P.  Effectiveness of structured planned post discharge support to patients with chronic obstructive pulmonary disease for reducing readmission rates: a systematic review.  JBI Database System Rev Implement Rep. 2017;15(8):2060-2086. doi:10.11124/JBISRIR-2016-003045PubMedGoogle ScholarCrossref
10.
Naylor  MD, Aiken  LH, Kurtzman  ET, Olds  DM, Hirschman  KB.  The care span: the importance of transitional care in achieving health reform.  Health Aff (Millwood). 2011;30(4):746-754. doi:10.1377/hlthaff.2011.0041PubMedGoogle ScholarCrossref
11.
Zwerink  M, Brusse-Keizer  M, van der Valk  PD,  et al.  Self management for patients with chronic obstructive pulmonary disease.  Cochrane Database Syst Rev. 2014;(3):CD002990. doi:10.1002/14651858.CD002990.pub3PubMedGoogle Scholar
12.
Newham  JJ, Presseau  J, Heslop-Marshall  K,  et al.  Features of self-management interventions for people with COPD associated with improved health-related quality of life and reduced emergency department visits: a systematic review and meta-analysis.  Int J Chron Obstruct Pulmon Dis. 2017;12:1705-1720. doi:10.2147/COPD.S133317PubMedGoogle ScholarCrossref
13.
Rice  KL, Dewan  N, Bloomfield  HE,  et al.  Disease management program for chronic obstructive pulmonary disease: a randomized controlled trial.  Am J Respir Crit Care Med. 2010;182(7):890-896. doi:10.1164/rccm.200910-1579OCPubMedGoogle ScholarCrossref
14.
Bourbeau  J, Julien  M, Maltais  F,  et al; Chronic Obstructive Pulmonary Disease axis of the Respiratory Network Fonds de la Recherche en Santé du Québec.  Reduction of hospital utilization in patients with chronic obstructive pulmonary disease: a disease-specific self-management intervention.  Arch Intern Med. 2003;163(5):585-591. doi:10.1001/archinte.163.5.585PubMedGoogle ScholarCrossref
15.
Aboumatar  H, Naqibuddin  M, Chung  S,  et al; BREATHE Study Patient Family Partners Group.  Better Respiratory Education and Treatment Help Empower (BREATHE) study: methodology and baseline characteristics of a randomized controlled trial testing a transitional care program to improve patient-centered care delivery among chronic obstructive pulmonary disease patients.  Contemp Clin Trials. 2017;62:159-167. doi:10.1016/j.cct.2017.08.018PubMedGoogle ScholarCrossref
16.
Hibbard  JH, Mahoney  ER, Stockard  J, Tusler  M.  Development and testing of a short form of the Patient Activation Measure.  Health Serv Res. 2005;40(6, pt 1):1918-1930. doi:10.1111/j.1475-6773.2005.00438.xGoogle ScholarCrossref
17.
Jones  PW, Quirk  FH, Baveystock  CM.  The St George’s Respiratory Questionnaire.  Respir Med. 1991;85(suppl B):25-31. doi:10.1016/S0954-6111(06)80166-6PubMedGoogle ScholarCrossref
18.
Jones  PW.  Interpreting thresholds for a clinically significant change in health status in asthma and COPD.  Eur Respir J. 2002;19(3):398-404. doi:10.1183/09031936.02.00063702PubMedGoogle ScholarCrossref
19.
Chapman  KR, Bergeron  C, Bhutani  M,  et al.  Do we know the minimal clinically important difference (MCID) for COPD exacerbations?  COPD. 2013;10(2):243-249. doi:10.3109/15412555.2012.733463PubMedGoogle ScholarCrossref
20.
Calverley  PM.  Minimal clinically important difference: exacerbations of COPD.  COPD. 2005;2(1):143-148. doi:10.1081/COPD-200050647PubMedGoogle ScholarCrossref
21.
Jones  PW, Beeh  KM, Chapman  KR, Decramer  M, Mahler  DA, Wedzicha  JA.  Minimal clinically important differences in pharmacological trials.  Am J Respir Crit Care Med. 2014;189(3):250-255. doi:10.1164/rccm.201310-1863PPPubMedGoogle ScholarCrossref
22.
McGhan  R, Radcliff  T, Fish  R, Sutherland  ER, Welsh  C, Make  B.  Predictors of rehospitalization and death after a severe exacerbation of COPD.  Chest. 2007;132(6):1748-1755. doi:10.1378/chest.06-3018PubMedGoogle ScholarCrossref
23.
Hurst  JR, Vestbo  J, Anzueto  A,  et al; Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints (ECLIPSE) Investigators.  Susceptibility to exacerbation in chronic obstructive pulmonary disease.  N Engl J Med. 2010;363(12):1128-1138. doi:10.1056/NEJMoa0909883PubMedGoogle ScholarCrossref
24.
Bahadori  K, FitzGerald  JM.  Risk factors of hospitalization and readmission of patients with COPD exacerbation: systematic review.  Int J Chron Obstruct Pulmon Dis. 2007;2(3):241-251.PubMedGoogle Scholar
25.
Varadhan  R, Wang  SJ.  Standardization for subgroup analysis in randomized controlled trials.  J Biopharm Stat. 2014;24(1):154-167. doi:10.1080/10543406.2013.856023Google ScholarCrossref
26.
Fan  VS, Gaziano  JM, Lew  R,  et al.  A comprehensive care management program to prevent chronic obstructive pulmonary disease hospitalizations: a randomized, controlled trial.  Ann Intern Med. 2012;156(10):673-683. doi:10.7326/0003-4819-156-10-201205150-00003PubMedGoogle ScholarCrossref
27.
Vegda  K, Nie  JX, Wang  L, Tracy  CS, Moineddin  R, Upshur  RE.  Trends in health services utilization, medication use, and health conditions among older adults: a 2-year retrospective chart review in a primary care practice.  BMC Health Serv Res. 2009;9:217.Google ScholarCrossref
28.
Pinkhasov  RM, Wong  J, Kashanian  J, et al.  Are men shortchanged on health? perspective on health care utilization and health risk behavior in men and women in the United States.  Int J Clin Pract. 2010;64(4):475-487.Google ScholarCrossref
Views 7,032
Citations 0
Original Investigation
October 8, 2019

Effect of a Hospital-Initiated Program Combining Transitional Care and Long-term Self-management Support on Outcomes of Patients Hospitalized With Chronic Obstructive Pulmonary Disease: A Randomized Clinical Trial

Author Affiliations
  • 1Armstrong Institute for Patient Safety and Quality, Johns Hopkins School of Medicine, Baltimore, Maryland
  • 2Division of General Internal Medicine, Department of Medicine, Johns Hopkins School of Medicine, Baltimore, Maryland
  • 3Department of Health, Behavior, and Society, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland
  • 4Welch Center for Prevention, Epidemiology, and Clinical Research, Johns Hopkins University, Baltimore, Maryland
  • 5Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland
  • 6Department of Anesthesiology and Critical Care Medicine, Johns Hopkins School of Medicine, Baltimore, Maryland
  • 7Division of Health Sciences Informatics, Johns Hopkins School of Medicine, Baltimore, Maryland
  • 8Malone Center for Engineering in Healthcare, Whiting School of Engineering, Johns Hopkins University, Baltimore, Maryland
  • 9Pulmonary and Critical Care Medicine, Johns Hopkins School of Medicine, Baltimore, Maryland
  • 10Johns Hopkins Bayview Medical Center, Baltimore, Maryland
  • 11Johns Hopkins Biostatistics Center, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland
  • 12Health Policy Research Group, University of Oregon, Eugene
JAMA. 2019;322(14):1371-1380. doi:10.1001/jama.2019.11982
Key Points

Question  Can a hospital-initiated program result in reduced acute care use and better quality of life for patients hospitalized for chronic obstructive pulmonary disease (COPD)?

Findings  In this single-site randomized clinical trial that included 240 patients with COPD, a 3-month program that combined transition and long-term management support, compared with usual care, resulted in a greater number of COPD-related hospitalizations and emergency department visits (1.40 vs 0.72 per participant); this comparison was statistically significant. There was no significant change in health-related quality of life (2.81 vs −2.69 in the 100-point St George’s Respiratory Questionnaire) at 6 months.

Meaning  This type of program may result in more acute care use among patients with COPD, but requires further research to determine the reason for this unexpected finding.

Abstract

Importance  Patients hospitalized for chronic obstructive pulmonary disease (COPD) exacerbations have high rehospitalization rates and reduced quality of life.

Objective  To evaluate whether a hospital-initiated program that combined transition and long-term self-management support for patients hospitalized due to COPD and their family caregivers can improve outcomes.

Design, Setting, and Participants  Single-site randomized clinical trial conducted in Baltimore, Maryland, with 240 participants. Participants were patients hospitalized due to COPD, randomized to intervention or usual care, and followed up for 6 months after hospital discharge. Enrollment occurred from March 2015 to May 2016; follow-up ended in December 2016.

Interventions  The intervention (n = 120) involved a comprehensive 3-month program to help patients and their family caregivers with long-term self-management of COPD. It was delivered by nurses with special training on supporting patients with COPD using standardized tools. Usual care (n = 120) included transition support for 30 days after discharge to ensure adherence to discharge plan and connection to outpatient care.

Main Outcomes and Measures  The primary outcome was number of COPD-related acute care events (hospitalizations and emergency department visits) per participant at 6 months. The co-primary outcome was change in participants’ health-related quality of life measured by the St George’s Respiratory Questionnaire (SGRQ) at 6 months after discharge (score, 0 [best] to 100 [worst]; 4-point difference is clinically meaningful).

Results  Among 240 patients who were randomized (mean [SD] age, 64.9 [9.8] years; 61.7% women), 203 (85%) completed the study. The mean (SD) baseline SGRQ score was 62.3 (18.8) in the intervention group and 63.6 (17.4) in the usual care group. The mean number of COPD-related acute care events per participant at 6 months was 1.40 (95% CI, 1.01-1.79) in the intervention group vs 0.72 (95% CI, 0.45-0.97) in the usual care group (difference, 0.68 [95% CI, 0.22-1.15]; P = .004). The mean change in participants’ SGRQ total score at 6 months was 2.81 in the intervention group and −2.69 in the usual care group (adjusted difference, 5.18 [95% CI, −2.15 to 12.51]; P = .11). During the study period, there were 15 deaths (intervention: 8; usual care: 7) and 339 hospitalizations (intervention: 202; usual care: 137).

Conclusions and Relevance  In a single-site randomized clinical trial of patients hospitalized due to COPD, a 3-month program that combined transition and long-term self-management support resulted in significantly greater COPD-related hospitalizations and emergency department visits, without improvement in quality of life. Further research is needed to determine reasons for this unanticipated finding.

Trial Registration  ClinicalTrials.gov Identifier: NCT02036294

Introduction

Quiz Ref IDChronic obstructive pulmonary disease (COPD) was the fourth leading cause of death in the United States in 2016, and has been a leading cause of morbidity and disability.1-3 Patients with COPD receiving emergency department (ED) and hospital care are more likely to have lower education and income and comorbidities.4 A review of discharge bundle interventions to improve outcomes of hospitalized patients with COPD showed a modest effect on reducing hospitalizations, and no effects on mortality or quality of life.5 Transitional care studies of patients with COPD are few, often limited to addressing postacute care needs in the 30-day postdischarge period, and did not focus on long-term chronic disease self-management skills.6-10 However, this may be insufficient for improving patient outcomes and reducing future acute care use.

Studies on COPD self-management support, mostly conducted in outpatient settings, have demonstrated improvements in health-related quality of life (HRQOL) and reductions in COPD-related acute care events.11,12 These studies have included ongoing education, action plans, and long-term self-management support.11-14 To our knowledge, no similar interventions have been tested among hospitalized patients with COPD, although hospitalization may offer a unique opportunity to engage patients and family caregivers in self-management of this condition.

In this study, a patient-centered, hospital-initiated, 3-month program that combines transition support and chronic disease self-management (the BREATHE Program) was developed and evaluated. The program aimed to improve quality of life and reduce acute care use among patients with COPD.15 The study’s primary hypothesis was that compared with participants who received usual transitional care, participants randomized to receive this program would have a lower number of COPD-related acute care events and better HRQOL at 6 months after hospital discharge.15

Methods

The Johns Hopkins Institutional Review Board approved this study. Written consent was obtained from all participants. Detailed study methods are described elsewhere.15 The trial protocol and statistical analysis plans are available in Supplement 1 and Supplement 2, respectively.

Study Design and Setting

This single-blinded randomized clinical trial had 2 groups (intervention and usual transitional care). The study took place at Johns Hopkins Bayview Medical Center, a 447-bed academic community hospital in Baltimore, Maryland, that serves the largest number of patients with COPD within the Johns Hopkins Health System (JHHS). Most hospitalized patients with COPD are treated at 1 of 4 medical units at the hospital. All patients admitted to these units starting in March 2015 were screened for eligibility.15 Enrollment ended in May 2016 and follow-up ended in December 2016.

Patients were eligible if they were either admitted to the Johns Hopkins Bayview Medical Center with a diagnosis of acute COPD exacerbation or had a previous COPD diagnosis and were receiving additional treatment to control COPD symptoms in the current hospitalization.15 Additional eligibility criteria included being aged older than 40 years, having a smoking history of more than 10 pack-years, understanding the English language, having no terminal illness (<6-month life expectancy) other than COPD, having no severe cognitive dysfunction (able to follow simple instructions), not being homeless, and expecting discharge to home.15 Eligibility was confirmed via medical record review and clinician confirmation. No participants were excluded based on discharge diagnosis.

Participants were randomized in a 1:1 ratio to either study group based on a pregenerated sequence of assignments. Randomization was stratified by hospital unit, and a computer algorithm was used to perform a blocked randomization assignment within strata with randomly selected block sizes of 2, 4, or 6. Data collectors and outcomes assessors were blinded to group allocation; however, due to the nature of the intervention, participants and clinicians were not blinded.

Study Group Descriptions
Intervention Group

The study intervention was co-developed with patients who have COPD, caregivers, and other stakeholders, with a focus on improving hospitalized patients’ HRQOL and reducing their future need for seeking emergency care.15,16 It included 3 components deemed necessary and complementary to achieving study goals15:

  1. Transition support to try to ensure that patients and caregivers were prepared for discharge and understood the postdischarge plan of care.

  2. Individualized COPD self-management support to help patients take medications correctly, recognize exacerbations signs and follow action plan (eAppendix 1 in Supplement 3), practice breathing exercises and energy conservation techniques, maintain an active lifestyle, seek help as needed, and stop smoking.

  3. Facilitated access to community programs and treatment services.

Quiz Ref IDThe intervention was delivered by COPD nurses (ie, nurses with special training on supporting patients with COPD using standardized tools). The nurses met with the patient (and caregiver whenever possible) during the hospital stay and for 3 months after discharge. They provided self-management support and addressed barriers to care. The program followed a patient-centered partnership approach and was delivered during a series of sessions held at the hospital and after discharge via home visit or telephone.15

Comparison Group

Participants in the comparison group received the usual transitional care provided at the study site. This included assigning a general transition coach to follow up the patient for 30 days after discharge, focusing on adherence to the discharge plan, and connecting to outpatient care. eTable 1 in Supplement 3 compares the intervention and usual care groups.

Data Collection

Patient consent, baseline assessment, and randomization were completed during hospital stay. The assessment included patient report on education, income, patient activation measure (assesses an individual’s knowledge, skill, and confidence for managing their health),16 comorbidities, and race/ethnicity (collected to report on minorities’ representation in randomized clinical trials, using 2 separate questions with specified response categories). Data collection telephone calls were conducted at 1 week and 1, 3, and 6 months after discharge. Acute care visits were assessed via medical record review for all visits within JHHS. Participants were asked at each telephone call if they had visited any non-JHHS hospital or ED and, if so, their records were requested and reviewed. For participants who could not be reached at 6 months, visits within the JHHS and any outside visits reported previously were reviewed. Records of each visit were independently reviewed by 2 physicians to determine whether the visit was COPD-related using predefined criteria.15 A third physician adjudicated any unresolved conflicts. Data on deaths were collected via medical record and death certificate reviews.15

Outcomes

The study’s prespecified primary outcome was the number of COPD-related acute care events, defined as hospitalizations and ED visits, per participant over the 6 months after discharge. A co-primary outcome was the change in participants’ HRQOL as measured by the St George’s Respiratory Questionnaire (SGRQ) score at 6 months after discharge. The COPD-related events outcome was the primary design variable to power the study, and we would not have considered this study positive without inferring a benefit on this outcome. The SGRQ outcome was chosen as a key supportive outcome given its patient-centeredness and importance to interpreting intervention effects. The SGRQ is a valid instrument for measuring HRQOL in patients with respiratory disease, with a total score and scores for symptom, physical activity, and impact domains (score range, 0 [best]-100 [worst]).17 Total score’s minimum clinically important difference is 4 points.17,18

Prespecified secondary outcomes were (1) 6-month mortality rate and (2) time to death or first COPD-related hospitalization or ED visit. As part of post hoc supplementary analyses, the following outcomes were compared: (1) mean number of COPD-related and all-cause acute care events per participant and the individual components (hospital and ED visits separately) at each time point; (2) percentage of participants who had at least 1 COPD-related acute care event; (3) change in participants’ SGRQ domain scores; and (4) percentage of participants whose HRQOL improved, stayed the same, or deteriorated within each group.

Intervention implementation was tracked and adverse events, including hospitalizations, deaths, and falls resulting in an acute care visit, were monitored.

Statistical Analyses

Sample size was calculated to detect a difference of 0.25 in the mean cumulative number of COPD-related visits per participant between study groups, with 80% power and type I error of .05 (2-sided). The 0.25 difference, which was based on results of an earlier COPD self-management trial, is considered clinically meaningful.13,19-21 The estimated sample size was 120 per group.15

Unadjusted analyses of the treatment effect under intention to treat were performed using negative binomial regression for the cumulative count of events, and linear regression for change in SGRQ total and domain scores. Additional analyses were conducted, adjusting for baseline SGRQ and unit for change in SGRQ scores and predictors of hospitalization (age, home oxygen use, and hospitalization within past year) and unit for acute care event counts.22-24

All analyses included robust estimates of variance and accounted for within-unit correlation. Normality and homoskedasticity of residuals were evaluated for linear regression models. SGRQ scores for patients who died were substituted with 100. The effect of missingness on primary outcomes was evaluated by comparing the patient characteristics of those with and without the missing outcome. Survival analyses and Cox proportional hazard models were used to compare the probability of not dying or having a COPD-related acute care event, by study group, adjusting for predictors of hospitalization. The proportional hazard assumption was evaluated with a test of a zero slope for the log-hazard ratio.

Post hoc analyses included logistic regression to compare the odds of having at least 1 COPD-related event, and responder analysis of participants whose HRQOL improved, stayed the same, or deteriorated (categorized change in SGRQ score as improved if ≤−4, and deteriorated if ≥4).17,18,21 A sensitivity analysis was performed to evaluate the effect of missingness on HRQOL using a mixed-effects generalized linear model with robust variance estimates, clustering within hospital units, with patient as random effect, and using all available SGRQ total scores (baseline, 3 months, and 6 months).

Post hoc exploratory subgroup analyses to assess heterogeneity of treatment effects were performed to evaluate the difference in treatment effect in subgroups of patients whose characteristics might have affected the primary outcome. The analyses were performed for the subgrouping variables of age, sex, marital status, home oxygen use, heart failure diagnosis, and baseline patient activation level. The treatment effects were estimated by the inclusion of an interaction term in a negative binomial regression model, standardizing with inverse probability weighting to address the correlation between the multiple subgrouping variables.25

Analysis was performed in Stata versions 14 and 15 (StataCorp). Main analyses were prespecified.15 Statistical significance was considered for P < .05 (2-sided). No adjustment for multiple comparisons was performed, and therefore all analyses of secondary and other supplementary outcomes should be considered exploratory.

Results

Figure 1 depicts participant recruitment, enrollment, and follow-up. Of 802 patients screened for participation, 417 met eligibility criteria and 240 provided consent. Baseline characteristics of the participants were similar between the study groups, except for a higher percentage of smoking, home oxygen use, heart failure diagnosis, and use of combined β-agonist/anticholinergic and short-acting β-agonists in the intervention group (Table 1).

No data were missing for baseline covariates. The mean (SD) participant age was 64.9 (9.8) years and most participants were white (82.5%) and female (61.7%). eTable 2 in Supplement 3 summarizes study participants’ comorbidities. The cumulative number of events at 1, 3, and 6 months after discharge were counted based on medical record review for participants living at those time points. Sixteen participants (14%) in the usual care group and 15 participants (13%) in the intervention group were lost to follow-up. No significant differences were found in baseline characteristics of these participants compared with the overall study sample.

Primary Outcomes
Effect on COPD-Related Acute Care Events

The total number of COPD-related acute care events in this study was 238 (196 COPD-related hospitalizations and 42 COPD-related ED visits). The study primary outcome was the number of COPD-related acute care events per patient over 6 months after discharge. This outcome was assessed for all participants living at 6 months (intervention: n = 112; usual care: n = 113). The mean number of COPD-related events per participant at 6 months was 1.40 (95% CI, 1.01-1.79) in the intervention group and 0.72 (95% CI, 0.45-0.97) in the usual care group (difference, 0.68 [95% CI, 0.22-1.15]; P = .004).

Effect on HRQOL

Mean (SD) baseline SGRQ scores were 62.3 (18.8) in the intervention group and 63.6 (17.4) in the usual care group (Table 2). Data were available to calculate the 6-month change in SGRQ scores for 88 and 91 participants in the intervention and usual care groups, respectively (80% of living participants). Baseline characteristics for age, hospital unit, home oxygen use, hospitalization in past year, and forced expiratory volume among patients with missing change in SGRQ scores were compared with other study participants. No significant differences were found except for more missing scores among patients older than 60 years (eTable 3 in Supplement 3). At 6 months after discharge, the mean change in SGRQ total score was 2.81 and −2.69 for the intervention and usual care groups, respectively (difference, 5.50 [95% CI, −2.57 to 13.57]; P = .12). The difference was similar after adjustment for hospital unit and baseline SGRQ score (adjusted difference, 5.18 [95% CI, −2.15 to 12.51]; P = .11) (Table 2). Differences in SGRQ scores had wide confidence intervals, increasing uncertainty in estimating intervention effects on HRQOL.

Secondary Outcomes

There were 8 deaths (6.7%) in the intervention group and 7 deaths (5.8%) in the usual care group (P > .99).

Figure 2 depicts Kaplan-Meier plots for time to first COPD-related acute care event or death. The 6-month event-free survival probability for no death or COPD-related acute care event was 0.45 in the intervention group and 0.58 in the usual care group (log rank P = .02). The adjusted hazard ratio for an event was 1.46 (95% CI, 1.02-2.09; P = .04) (test of proportional hazard assumption met; P = .93). (eFigures 1 and 2 in Supplement 3 depict Kaplan-Meier plots by type of acute care event).

Supplementary Post Hoc Analyses

Figure 3 depicts the cumulative number of COPD-related and all-cause events per participant by event type and study group at 1, 3, and 6 months after discharge. The incidence rate ratio of COPD-related events at 6 months for the intervention compared with usual care group was 1.96 (95% CI, 1.57- 2.43; P < .001) before adjustment and 1.59 (95% CI, 1.32-1.92; P < .001) after adjustment for age, home oxygen use, discharge unit, and hospitalization in the prior year. The incidence rates and incidence rate ratios of COPD-related and all-cause events by event type and study group before and after adjustment are depicted in eTable 4 in Supplement 3. The treatment effect was similar after adjustment for age, home oxygen use, discharge unit, and hospitalization in prior year (eTable 4 in Supplement 3).

The percentage of participants who experienced at least 1 COPD-related acute care event was 52% in the intervention group and 38% in the usual care group (odds ratio, 1.74 [95% CI, 1.03-2.98]; P = .04). Of those participants, 22%, 10%, 5%, and 14% had 1, 2, 3, and 4 or more COPD-related acute care events in the intervention group compared with 20%, 12%, 3%, and 3% in the usual care group, respectively (Fisher exact P = .01).

Table 2 depicts differences in SGRQ domain scores between study groups. There were no significant differences in domain scores. The responder analysis showed that at 6 months, the HRQOL improved, stayed the same, or deteriorated among 38%, 19%, and 43% of patients in the intervention group compared with 51%, 25%, and 24% of patients in the usual care group, respectively (Pearson χ2P = .03). eFigure 3 in Supplement 3 depicts a parallel line plot of change in SGRQ scores by study group. A sensitivity analysis, performed to assess the effect of missing values on HRQOL using mixed-effects generalized linear model, allowed for examination of 586 SGRQ total scores across 231 patients and showed findings that were consistent with the primary analysis. (Estimated differences in SGRQ at 6 months compared with baseline for the intervention and usual care groups were 2.51 [95% CI, −1.61 to 6.63] and −2.43 [95 % CI, −6.37 to 1.50], respectively; see the eAppendix 2 and eTable 5 in Supplement 3 for more details.)

The exploratory subgroup analyses assessing the heterogeneity of treatment effects found significant intervention interactions with sex and with patient activation level (eFigure 4 in Supplement 3). Further analysis looking at the 4 subgroups of men and women with high and low activation levels revealed a significant interaction (overall test of interaction terms P <.001). Both men and women with high activation levels had a higher number of acute care events in the intervention group compared with the control group, although this effect was more pronounced among men (incidence rate ratio for men, 8.08 [95% CI, 3.72 to 17.53), and for women, 2.17 [95% CI, 1.02 to 4.63]). This effect persisted after adjustment and weighting for variables (eTable 6 in Supplement 3).

Intervention Implementation

The COPD nurse visited 103 of 120 participants at least once in the hospital. Intervention group participants had a mean of 6.1 sessions (interquartile range, 4-8) (eTable 7 in Supplement 3).

Adverse Events

Adverse events were reviewed at 3- to 6-month intervals. There were 339 hospitalizations and 3 falls resulting in acute care visits. No adverse events were attributed to the study intervention.

Discussion

Quiz Ref IDIn a single-site randomized clinical trial of patients hospitalized due to COPD, a 3-month program that combined transition and long-term self-management support resulted in significantly greater COPD-related hospitalizations and ED visits. There was no significant difference in HRQOL.

COPD self-management interventions, mostly implemented in outpatient settings, have shown reductions in COPD-related hospitalizations but most excluded participants based on comorbidities and provided medical interventions (eg, steroid or antibiotics prescription for use as needed).11 One study using a COPD action plan with self-initiated treatment was stopped prematurely for concerns of increased mortality.26 In the current study, patients with comorbidities were not excluded and the intervention offered no medical treatment services. The intervention, which was co-developed with patients, caregivers, and others (including clinicians of various disciplines and health care leaders), focused on engaging both patients and caregivers in COPD self-management. It included an action plan that focused on early recognition of signs and symptoms of acute exacerbations and contacting a physician early for those.

The increase in COPD-related acute care use in the intervention compared with usual care group found in this study was in the opposite direction from what had been hypothesized. This inconsistency necessarily raises a question about the validity of the study findings and means that study interpretation needs to be cautious and provisional.

Quiz Ref IDIf the findings represent a true effect, the post hoc subgroup analyses raise possible mechanisms, although these need to be considered speculative. The increase in acute care events was limited to patients with high activation status at baseline and was more common among men than women. Patients with greater activation may have responded to the intervention by being more vigilant about detecting signs of early exacerbations and taking action to get medical attention within 24 hours, as recommended by their action plan. Inability to access their regular physician within that timeframe, due to various reasons, could have led them to seek acute care services. In addition, increased communications with clinicians about exacerbation signs might have led to increased referrals to the emergency department (and subsequent hospitalizations). More acute care use by men than women may relate to less experience with seeking ambulatory care and lack of established connections with the health care system for routine care services. The latter has been demonstrated in other studies.27,28

Limitations

This study has several limitations. First, the study was conducted at a single site, and therefore, the results may not be generalizable to all patients with COPD. Second, the study population included a high proportion of low-income and less-educated participants, and these participants may have greater challenges in accessing the health care system for urgent visits outside of coming to the emergency department. Third, there were small differences in the study groups at baseline (eg, higher percentages of current smokers and continuous home oxygen therapy use) that might have led to greater health care utilization in the intervention group. Fourth, for participants who could not be contacted at 6 months after discharge, it was not possible to measure their QOL and verify acute events treated outside the JHHS. However, these participants were similar between study groups and their baseline characteristics were not different from the rest of study participants. Fifth, spirometry evidence of airflow obstruction was not required for enrollment into the study and it is possible that some participants may have been incorrectly diagnosed as having COPD. Quiz Ref IDSixth, because the direction of the results were opposite of what was anticipated, the findings should be considered hypothesis-generating rather than definitive.

Conclusions

In a single-site randomized clinical trial of patients hospitalized due to COPD, a 3-month program that combined transition and long-term self-management support resulted in significantly greater COPD-related hospitalizations and emergency department visits, without improvement in quality of life. Further research is needed to determine reasons for this unanticipated finding.

Back to top
Article Information

Corresponding Author: Hanan Aboumatar, MD, MPH, Armstrong Institute for Patient Safety and Quality, Johns Hopkins School of Medicine, 750 E Pratt St, 15th Floor, Baltimore, MD 21202 (habouma1@jhmi.edu).

Accepted for Publication: July 23, 2019.

Author Contributions: Dr Aboumatar had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Aboumatar, Naqibuddin, Chung, Chaudhry, Bone, Gurses, Knowlton, Pronovost, Rand, Roter, Sylvester, Wolff, Hibbard, Wise.

Acquisition, analysis, or interpretation of data: Aboumatar, Naqibuddin, Chung, Kim, Saunders, Pronovost, Putcha, Rand, Thompson, Wise.

Drafting of the manuscript: Aboumatar, Naqibuddin, Chung, Chaudhry, Thompson, Hibbard, Wise.

Critical revision of the manuscript for important intellectual content: Aboumatar, Naqibuddin, Chung, Chaudhry, Kim, Saunders, Bone, Gurses, Knowlton, Pronovost, Putcha, Rand, Roter, Sylvester, Thompson, Wolff, Wise.

Statistical analysis: Aboumatar, Chung, Thompson.

Obtained funding: Aboumatar, Gurses, Rand.

Administrative, technical, or material support: Aboumatar, Naqibuddin, Chung, Chaudhry, Kim, Saunders, Bone, Gurses, Roter, Sylvester, Wise.

Supervision: Aboumatar, Naqibuddin, Pronovost, Hibbard.

Conflict of Interest Disclosures: Drs Aboumatar, Saunders, Gurses, and Wolff; Ms Chung; Ms Chaudhry; and Mr Kim reported receiving grants from the Patient-Centered Outcomes Research Institute (PCORI). Dr Putcha reported receiving grants from the National Heart, Lung, and Blood Institute. Dr Hibbard reported receiving personal fees from Insignia Health and having equity ownership in Insignia Health, as well as receiving royalties on the Patient Activation Measure paid by Insignia Health through the University of Oregon. Dr Wise reported receiving grants and personal fees from AstraZeneca/Medimmune, Boehringer Ingelheim, and GlaxoSmithKline; personal fees from AbbVie, Contrafect, Kiniksa, Novartis, Pulmonx, Roche, Spiration, Sunovion, Syneos, Merck, Circassia, Pneuma, Verona, Bonti, Denali, and Aradigm; grants from Pearl Therapeutics and Sanofi; and nonfinancial support from Propeller Health. No other disclosures were reported.

Funding/Support: Funding was provided by PCORI (Award IHS 1304-7118).

Role of the Funder/Sponsor: PCORI had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The opinions in this publication are solely the responsibility of the authors and do not necessarily represent the views of PCORI, its board of governors, or methodology committee.

Meeting Presentation: This study was presented at the American Public Health 2018 Annual Meeting; November 12, 2018; San Diego, California.

Data Sharing Statement: See Supplement 4.

Additional Contributions: We acknowledge the important contributions of Edna Shattuck, RN, RRT, a study co-investigator and patient partner who died before publication of study results, and all the members of the BREATHE Patient Family Partners Group. They were compensated for their role in the study. We are also grateful for the support from the Johns Hopkins Homecare Group (Hajara Adebowale, RN, BSN; Donna Kurtz, RN; Leslie Piet, RN, MA, CCM, CHPN; Laura Syron, RN, MSN, MPH; and Mary Myers, MS, RN) and the Johns Hopkins Biostatistics Center. Ms Adebowale, Ms Kurtz, and Ms Piet performed the role of the chronic obstructive pulmonary disease nurse on study and were paid for this work. Ms Syron and Ms Myers were not compensated. The Johns Hopkins Biostatistics Center received compensation. We thank the members of the BREATHE Study data safety and monitoring board: David Levine, MD (Johns Hopkins University); Albert Wu, MD (Johns Hopkins University); Sanjay V. Desai, MD (Johns Hopkins University); and John Linnell (patient advocate on the board). Drs Levine, Wu, and Desai received no compensation for their role in the study; Mr Linnell received an honorarium. We also acknowledge the support and contributions of Kai Shea, MSW, LCSW-C (Johns Hopkins Bayview Social Work Department); Jamie Sullivan, MPH (COPD Foundation); Eric E. Howell, MD (Johns Hopkins Hospitalist service); Emmanuel Garcia-Morales, PhD (Johns Hopkins Armstrong Institute); and Joseph Neiman, MD (Johns Hopkins Armstrong Institute). Ms Shea was not compensated. Ms Sullivan and Drs Howell, Garcia-Morales, and Neiman were compensated for their role in the study.

Additional Information: This article is a corrected version of an article that has been retracted (Aboumatar H, Naqibuddin M, Chung S, et al. Effect of a program combining transitional care and long-term self-management support on outcomes of hospitalized patients with chronic obstructive pulmonary disease: a randomized clinical trial. JAMA. 2018;320(22):2335-2343 doi:10.1001/jama.2018.17933).

References
1.
Kochanek  KD, Murphy  SL, Xu  JQ, Arias  E.  Mortality in the United States, 2016: NCHS Data Brief, No. 293. Hyattsville, MD: National Center for Health Statistics; 2017.
2.
Ford  ES, Croft  JB, Mannino  DM, Wheaton  AG, Zhang  X, Giles  WH.  COPD surveillance: United States, 1999-2011.  Chest. 2013;144(1):284-305. doi:10.1378/chest.13-0809PubMedGoogle ScholarCrossref
3.
Brault  MW, Hootman  J, Helmick  CG, Theis  KA, Armour  BS; Centers for Disease Control and Prevention (CDC).  Prevalence and most common causes of disability among adults: United States, 2005.  MMWR Morb Mortal Wkly Rep. 2009;58(16):421-426.PubMedGoogle Scholar
4.
Kumbhare  SD, Beiko  T, Wilcox  SR, Strange  C.  Characteristics of COPD patients using United States emergency care or hospitalization.  Chronic Obstr Pulm Dis. 2016;3(2):539-548. doi:10.15326/jcopdf.3.2.2015.0155PubMedGoogle Scholar
5.
Ospina  MB, Mrklas  K, Deuchar  L,  et al.  A systematic review of the effectiveness of discharge care bundles for patients with COPD.  Thorax. 2017;72(1):31-39. doi:10.1136/thoraxjnl-2016-208820PubMedGoogle ScholarCrossref
6.
Hansen  LO, Young  RS, Hinami  K, Leung  A, Williams  MV.  Interventions to reduce 30-day rehospitalization: a systematic review.  Ann Intern Med. 2011;155(8):520-528. doi:10.7326/0003-4819-155-8-201110180-00008PubMedGoogle ScholarCrossref
7.
Prieto-Centurion  V, Markos  MA, Ramey  NI,  et al.  Interventions to reduce rehospitalizations after chronic obstructive pulmonary disease exacerbations: a systematic review.  Ann Am Thorac Soc. 2014;11(3):417-424. doi:10.1513/AnnalsATS.201308-254OCPubMedGoogle ScholarCrossref
8.
Shah  T, Press  VG, Huisingh-Scheetz  M, White  SR.  COPD readmissions: addressing COPD in the era of value-based health care.  Chest. 2016;150(4):916-926. doi:10.1016/j.chest.2016.05.002PubMedGoogle ScholarCrossref
9.
Pedersen  PU, Ersgard  KB, Soerensen  TB, Larsen  P.  Effectiveness of structured planned post discharge support to patients with chronic obstructive pulmonary disease for reducing readmission rates: a systematic review.  JBI Database System Rev Implement Rep. 2017;15(8):2060-2086. doi:10.11124/JBISRIR-2016-003045PubMedGoogle ScholarCrossref
10.
Naylor  MD, Aiken  LH, Kurtzman  ET, Olds  DM, Hirschman  KB.  The care span: the importance of transitional care in achieving health reform.  Health Aff (Millwood). 2011;30(4):746-754. doi:10.1377/hlthaff.2011.0041PubMedGoogle ScholarCrossref
11.
Zwerink  M, Brusse-Keizer  M, van der Valk  PD,  et al.  Self management for patients with chronic obstructive pulmonary disease.  Cochrane Database Syst Rev. 2014;(3):CD002990. doi:10.1002/14651858.CD002990.pub3PubMedGoogle Scholar
12.
Newham  JJ, Presseau  J, Heslop-Marshall  K,  et al.  Features of self-management interventions for people with COPD associated with improved health-related quality of life and reduced emergency department visits: a systematic review and meta-analysis.  Int J Chron Obstruct Pulmon Dis. 2017;12:1705-1720. doi:10.2147/COPD.S133317PubMedGoogle ScholarCrossref
13.
Rice  KL, Dewan  N, Bloomfield  HE,  et al.  Disease management program for chronic obstructive pulmonary disease: a randomized controlled trial.  Am J Respir Crit Care Med. 2010;182(7):890-896. doi:10.1164/rccm.200910-1579OCPubMedGoogle ScholarCrossref
14.
Bourbeau  J, Julien  M, Maltais  F,  et al; Chronic Obstructive Pulmonary Disease axis of the Respiratory Network Fonds de la Recherche en Santé du Québec.  Reduction of hospital utilization in patients with chronic obstructive pulmonary disease: a disease-specific self-management intervention.  Arch Intern Med. 2003;163(5):585-591. doi:10.1001/archinte.163.5.585PubMedGoogle ScholarCrossref
15.
Aboumatar  H, Naqibuddin  M, Chung  S,  et al; BREATHE Study Patient Family Partners Group.  Better Respiratory Education and Treatment Help Empower (BREATHE) study: methodology and baseline characteristics of a randomized controlled trial testing a transitional care program to improve patient-centered care delivery among chronic obstructive pulmonary disease patients.  Contemp Clin Trials. 2017;62:159-167. doi:10.1016/j.cct.2017.08.018PubMedGoogle ScholarCrossref
16.
Hibbard  JH, Mahoney  ER, Stockard  J, Tusler  M.  Development and testing of a short form of the Patient Activation Measure.  Health Serv Res. 2005;40(6, pt 1):1918-1930. doi:10.1111/j.1475-6773.2005.00438.xGoogle ScholarCrossref
17.
Jones  PW, Quirk  FH, Baveystock  CM.  The St George’s Respiratory Questionnaire.  Respir Med. 1991;85(suppl B):25-31. doi:10.1016/S0954-6111(06)80166-6PubMedGoogle ScholarCrossref
18.
Jones  PW.  Interpreting thresholds for a clinically significant change in health status in asthma and COPD.  Eur Respir J. 2002;19(3):398-404. doi:10.1183/09031936.02.00063702PubMedGoogle ScholarCrossref
19.
Chapman  KR, Bergeron  C, Bhutani  M,  et al.  Do we know the minimal clinically important difference (MCID) for COPD exacerbations?  COPD. 2013;10(2):243-249. doi:10.3109/15412555.2012.733463PubMedGoogle ScholarCrossref
20.
Calverley  PM.  Minimal clinically important difference: exacerbations of COPD.  COPD. 2005;2(1):143-148. doi:10.1081/COPD-200050647PubMedGoogle ScholarCrossref
21.
Jones  PW, Beeh  KM, Chapman  KR, Decramer  M, Mahler  DA, Wedzicha  JA.  Minimal clinically important differences in pharmacological trials.  Am J Respir Crit Care Med. 2014;189(3):250-255. doi:10.1164/rccm.201310-1863PPPubMedGoogle ScholarCrossref
22.
McGhan  R, Radcliff  T, Fish  R, Sutherland  ER, Welsh  C, Make  B.  Predictors of rehospitalization and death after a severe exacerbation of COPD.  Chest. 2007;132(6):1748-1755. doi:10.1378/chest.06-3018PubMedGoogle ScholarCrossref
23.
Hurst  JR, Vestbo  J, Anzueto  A,  et al; Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints (ECLIPSE) Investigators.  Susceptibility to exacerbation in chronic obstructive pulmonary disease.  N Engl J Med. 2010;363(12):1128-1138. doi:10.1056/NEJMoa0909883PubMedGoogle ScholarCrossref
24.
Bahadori  K, FitzGerald  JM.  Risk factors of hospitalization and readmission of patients with COPD exacerbation: systematic review.  Int J Chron Obstruct Pulmon Dis. 2007;2(3):241-251.PubMedGoogle Scholar
25.
Varadhan  R, Wang  SJ.  Standardization for subgroup analysis in randomized controlled trials.  J Biopharm Stat. 2014;24(1):154-167. doi:10.1080/10543406.2013.856023Google ScholarCrossref
26.
Fan  VS, Gaziano  JM, Lew  R,  et al.  A comprehensive care management program to prevent chronic obstructive pulmonary disease hospitalizations: a randomized, controlled trial.  Ann Intern Med. 2012;156(10):673-683. doi:10.7326/0003-4819-156-10-201205150-00003PubMedGoogle ScholarCrossref
27.
Vegda  K, Nie  JX, Wang  L, Tracy  CS, Moineddin  R, Upshur  RE.  Trends in health services utilization, medication use, and health conditions among older adults: a 2-year retrospective chart review in a primary care practice.  BMC Health Serv Res. 2009;9:217.Google ScholarCrossref
28.
Pinkhasov  RM, Wong  J, Kashanian  J, et al.  Are men shortchanged on health? perspective on health care utilization and health risk behavior in men and women in the United States.  Int J Clin Pract. 2010;64(4):475-487.Google ScholarCrossref
×