Respirology | 2019
Treating moderate–severe obstructive sleep apnoea for cardiovascular health: Is what stake the stakeholder holds important?
Abstract
As clinical and scientific researchers, we live in an age where it is becoming increasingly important to engage with stakeholder groups that include individuals who represent patients, pharmaceutical and device-based industries and professional bodies. Each has a role to play in aiding with the funding and design of clinical trial and other research protocols that will answer key clinical questions about treatments for disease. Traditionally, the impetus for testing the efficacy of a treatment for a disease comes from the equipoise surrounding the question of whether the treatment will benefit health in some way. The greater the equipoise, the greater is the need to establish whether a treatment is of benefit. Usually, it is the veracity and extent of the scientific evidence base as judged by peak professional body expert panels (made up of scientists and clinician researchers) that play an important role in determining the state of this equipoise and the subsequent need for clinical trials. However, the success of these trials is becoming increasingly reliant on engagement with both patient and industry partners because of their valuable contribution to aspects of recruitment and funding. In a recent article published in Respirology, Landry et al. explored the question of equipoise surrounding the benefits of treatment of moderate–severe obstructive sleep apnoea (OSA) for aspects of cardiovascular (CV) health by surveying 12 representatives of ‘key sleep medicine stakeholder groups’. The stakeholder groups included scientists and clinician researchers as well as industry and patient representatives. The survey was based on the provision of evidence from 39 articles, which included systematic reviews/meta-analyses of randomized controlled trials (RCT) (level 1 evidence), individual RCT (level 2 evidence) and observational cohort studies. Participants were asked whether they disagreed, were unsure or whether they agreed that treatment of OSA conferred benefits for each of the three CV health outcomes: (i) CV events and death, (ii) blood pressure in OSA patients with hypertension and (iii) left ventricular function (LVF) in OSA patients with heart failure. The three categories of response were each divided into three choices as follows: Disagree 1–3, Unsure 4–6 and Agree 7–9. A consensus of opinion required at least 80% of panel votes to fall within the same response category. The results indicated that the panel agreed that OSA treatment improves blood pressure, but they were unsure about the treatment benefits for CV events and death. The results for LVF had a somewhat skewed/bimodal distribution of opinions but with a median response falling within the ‘Agree’ category. However, respondents failed to achieve the 80% unified response required for consensus. Overall, these results indicate ongoing equipoise about OSA treatment benefits for CV events/ death and LVF, which then suggests that additional level 2 evidence is still required. This was an unusual approach to devising a consensus statement because it involved both industry and patient representatives. We are certainly not aware of active involvement of industry in American Thoracic Society or European Respiratory Society consensus statements, and indeed, the authors did not indicate whether this has previously occurred. Outside of the field of sleep medicine, we are also not aware if pharma companies actively participate in consensus statements for disorders such as hypertension, osteoporosis or cancer. Presumably, the reasoning behind the involvement of industry opinion in this study was partly because the current variety of healthcare models in Australia includes these groups. This is an important issue because some may question the value of stakeholder input to consensus statements about treatment if that input involves industry or industry-conflicted panellists who could derive financial benefit separate from that associated with untied funding of research. Should you really include such groups in consensus statements? Of course, clinicians, methodologists and even patient representatives are always asked to declare their conflicts of interest before participating in such groups, and this also occurred in this study. In the context of healthcare models, over the past 20 years, Australian sleep apnoea care has been characterized by an increasing number of industry-driven ‘vertical integration’ models providing diagnosis and then management by one modality. This, in turn, has led to increasing regulatory interest from the government and investigation of these models. Perhaps the time has come to change the focus to ‘horizontal integration’ models involving multiple treatments, greater information for patients on therapeutic alternatives and closer involvement of primary care. In this way, integration of sleep apnoea care into primary care prevention and treatment of CV disease could be instituted. There are also several limitations in this survey that should be mentioned. (i) The panel only comprised 12 members, which is a small number given the multidisciplinary nature of representation (patient, clinician/scientist and industry). (ii) The analysis did not indicate whether there was any bias in consensus