Customize your JAMA Network experience by selecting one or more topics from the list below.
Science and sound clinical care both rely on honest reporting and, when necessary, self-correction. In this issue of JAMA, the Notice of Retraction1 and republication of the article by Aboumatar et al titled, “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,”2 represent a commitment by the authors to ensure an accurate scientific record. Following discovery of an error in the analysis of data from the initial report,3 the reanalysis by the authors now shows that the study’s original conclusion changed remarkably from showing a strong benefit of the intervention to showing harm. The results of this study, as in all high-quality research, are important regardless of outcome. The authors have acknowledged and addressed the error in an open and transparent way. The integrity of science is built on the principle that scientists are forthright in their research, and these authors adopted an earnest approach to amend their error.
The initial study by Aboumatar et al3 was a randomized clinical trial of a 3-month intensive chronic obstructive pulmonary disease (COPD) intervention that combined transitional care support and chronic disease self-management. While results of the retracted publication suggested that patients in the intervention group had fewer COPD-related acute care events (hospitalizations and emergency department visits), the reanalysis demonstrated that the intervention substantially increased the risk of COPD-related acute care events.2 The mean number of COPD-related acute care events (hospitalizations and emergency department visits) per participant at 6 months was 1.40 (95% CI, 1.01-1.79) in the intervention group vs 0.72 (95% CI, 0.46-0.97) in the usual care group (difference, 0.68 [95% CI, 0.22-1.15]; P = .004).2 Although there was no significant effect on health-related quality of life between the 2 groups, the direction of the results was consistent with the primary findings.
The increased risk of harm associated with the study intervention was both unexpected and important. As the authors indicated, these results should be interpreted with caution, and additional research is needed to understand why this intervention was associated with increased acute events. At the same time, this work is consistent with other literature. Numerous clinical trials on strategies to prevent COPD hospitalizations have produced conflicting results, with all recent studies failing to demonstrate benefit and some reporting harm.4 The increased risk of acute care events found in the current study has parallels with the study by Fan et al5 that found higher mortality among patients with COPD who participated in a comprehensive care management program.
With this study further contributing to inconsistent COPD trial outcomes, an important question is: “Why do well-designed and well-intentioned care management interventions for individuals with COPD often fail to improve outcomes?” These behavioral interventions are complex and the mechanism of harm when it occurs is rarely clear. Although the reasons for increased COPD acute care events in this study are not apparent, several characteristics of the intervention, study population, and outcome of interest may provide insight into this unanticipated finding.
It is possible that inherent characteristics of the intervention caused a clinical deterioration among study participants. In their post hoc analysis, the authors found that the increased risk of acute care events was isolated to patients with high levels of activation, based on a measure of self-perceived knowledge, skills, and confidence in managing their health. Self-management programs could cause these patients to become overconfident in their abilities and delay seeking care at times that may have mitigated adverse events. Alternatively, the intervention may have led these individuals to be more aware of their respiratory symptoms and seek higher levels of care that, in turn, contributed to significantly increased acute care events.
The intensity of transitional care in the intervention differentiates it from previous studies and may account for some of the differences noted between the intervention and control groups. Researchers engaged patients during hospitalization and provided frequent patient outreach for 3 months following hospital discharge. On average, intervention patients had 6 sessions with a COPD nurse, each lasting approximately 20 minutes. Every one of these sessions was a touch point that could identify patient concerns and lead to an escalation in care. For some patients, frequent interactions and focused attention on their disease may produce greater anxiety and lead to unnecessary clinical care.
The design of the COPD self-management program incorporated into the intervention could also contribute to the increased risk of acute care events. Previous studies of self-management interventions for patients with COPD have demonstrated improvements in health-related quality of life and reduction in respiratory-related hospital admissions, although these studies have largely focused on outpatient settings, where patients tend to have less-severe disease and may respond better to simple interventions that address an increase in respiratory symptoms.6 The intervention in the current study also did not include provision of antibiotics or steroids for self-medication of exacerbations, a critical component of some previously effective self-management interventions.7 Rather, patients were instructed to contact clinicians when they developed respiratory symptoms. Clinicians responding to patients who relay more frequent respiratory symptoms may have a lower threshold for referring them for emergency department evaluation and hospital admission.
The focus of the intervention on COPD-specific management may have inadequately addressed comorbidities. Inclusion criteria for the study was based on COPD diagnostic codes and in-hospital inhaler therapies regardless of the indication for hospitalization. Misdiagnosis of COPD is common,8 and patients may have undiagnosed or unaddressed conditions influencing their trajectory. Appropriately, the authors did not exclude patients based on comorbidities, which are common among patients with COPD. Most rehospitalizations after COPD exacerbations are due to conditions other than COPD.9 Even though there was no significant difference between intervention and control patients in secondary analysis of all-cause readmissions, focusing the intervention primarily on COPD management could have missed an opportunity to address other causes of acute care events and provide more comprehensive care.
This study was a single-site intervention with a unique patient population, and the results may differ in other populations. Study participants tended to have socioeconomic disadvantages, including lower levels of education and income, and most patients indicated that they did not have anyone helping them with their health care. The study intervention may have opened channels for more regular care for these patients, including acute care referrals that they would not have otherwise pursued. A similar finding was noted among states that expanded Medicaid under the Affordable Care Act, with increased access to care resulting in more primary care visits and overnight hospital stays.10
Results of this study and others underscore the challenges with focusing policies on COPD acute care events. Some features of the intervention and patient population may result in more acute care events, although the close follow-up and patient education included in this study may be considered components of high-quality COPD care, regardless of their effect on subsequent hospitalization or emergency department visits. Furthermore, some patients with worsening respiratory symptoms may, in fact, prefer treatment in the hospital, and it is not clear that these acute care events are necessarily bad outcomes.
The Hospital Readmission Reduction Program of the Centers for Medicare & Medicaid Services financially penalizes hospitals with high 30-day risk-adjusted readmission rates for select conditions, including COPD.11 The study by Aboumatar et al did not focus on 30-day readmission rates as a primary outcome and there was no difference in this measure between the intervention and control groups. Yet this study is a reminder that we lack strong evidence on how to prevent COPD hospitalizations, and there are concerns that efforts to reduce readmissions may yield unintended consequences.12
To provide the highest-quality care for patients with COPD, clinicians should rely on established evidence-based practices that improve patient-centered outcomes. Current guidelines offer recommendations that achieve this goal by promoting smoking cessation, treating with guideline-concordant medications, initiating oxygen for patients with severe resting hypoxia, referring to pulmonary rehabilitation, and addressing other medical, psychiatric, and socioeconomic conditions affecting their health.13
In summary, the study by Aboumatar et al offers a cautionary tale of the challenges in designing interventions to reduce COPD acute care events. The scientific literature is replete with good intentions gone awry. It is important to report the problems, pitfalls, and unintended consequences of these efforts, and even call into question policies that are built on these measures. Clinicians, administrators, and health systems searching for ways to reduce COPD hospitalizations are still left with limited guidance. Therefore, clinicians should continue to provide the best possible COPD care based on existing evidence. Intensive interventions beyond these basic measures warrant additional research and must consider potential harms.
Corresponding Author: Seppo Rinne, MD, PhD, Center for Healthcare Organization & Implementation Research, Veterans Affairs, 200 Springs Rd, Bedford, MA 01730 (email@example.com).
Conflict of Interest Disclosures: Dr Lindenauer reported receiving salary support from the Yale Center for Outcomes Research and Evaluation under a contract from the Centers for Medicare & Medicaid Service to develop and maintain hospital outcome measures for chronic obstructive pulmonary disease. Dr Au reported receiving grants from the Department of Veterans Affairs; National Heart, Lung, and Blood Institute; and American Lung Association and personal fees from the American Thoracic Society, American Board of Internal Medicine, and Novartis. No other disclosures were reported.
Rinne ST, Lindenauer PK, Au DH. Unexpected Harm From an Intensive COPD Intervention. JAMA. 2019;322(14):1357–1359. doi:10.1001/jama.2019.12976
Coronavirus Resource Center