Editorial <\/p>\n\n\n\n
October 9, 2024<\/p>\n\n\n\n
<\/a>The benefits of the intensive care unit (ICU) for delivering high-quality care and improving survival for critically ill patients are primarily due to the people who work there\u2014the specialized physicians, nurses, and other health care professionals who comprise the interprofessional critical care team.1<\/a><\/sup>\u00a0Yet, over the last century, the field is increasingly defined not by people, but by technology\u2014mechanical ventilators, invasive monitors, and, most recently, artificial intelligence (AI) to predict outcomes and guide therapy.2<\/a><\/sup>\u00a0Although some of these technologies undoubtedly save lives, technology in and of itself is not a guarantee of better patient outcomes, as exemplified by the pulmonary artery catheter that was found to increase costs of care without improving patient outcomes.3<\/a><\/sup><\/p>\n\n\n\n ICU telemedicine\u2014defined as the provision of care to critically ill patients by health care professionals located remotely\u2014is an intervention that uniquely sits at the intersection of people and technology. The primary function of ICU telemedicine is to expand access to intensivists in the setting of increasing demand for critical care.4<\/a><\/sup> However, ICU telemedicine also frequently incorporates other technological innovations, such as advanced predictive algorithms, to support clinician decision-making.5<\/a><\/sup> Its appeal likely stems from the way it bridges both worlds, satisfying our fondness for technology and our inherent belief in the value of human decision-making in health care. Despite its widespread implementation, there remains a paucity of high-quality evidence for or against its use.4<\/a><\/sup> Most studies are limited by pre-post study designs without concurrent controls, precluding strong conclusions about the overall effectiveness of ICU telemedicine.4<\/a><\/sup>,6<\/a><\/sup><\/a><\/p>\n\n\n\n In the TELE-critical Care vs usual Care on ICU Performance (TELESCOPE) trial published this week in JAMA<\/em>,7<\/a><\/sup>Pereira and colleagues overcome some of these prior limitations to provide important new information regarding the effect of ICU telemedicine on patient outcomes. The authors performed a parallel cluster randomized trial across 30 Brazilian ICUs. The intervention itself had 3 core components: (1) multidisciplinary weekday rounds led by remote intensivists using telemedicine technology, (2) monthly audit and feedback meetings between remote intensivists and local leadership to discuss care performance indicators, and (3) the distribution of 19 evidence-based clinical protocols to local clinicians. The primary outcome was ICU length of stay; secondary outcomes included a measure of ICU efficiency, in-hospital mortality, and several markers of care quality.<\/a><\/p>\n\n\n\n Between June 2019 and May 2021, 17\u202f024 critically ill adult patients were enrolled, with good balance between patient and ICU characteristics in both groups. Median time under intervention was 20 months. Tele-visits occurred on 68% of eligible patient-days and 74% of the medical recommendations from remote intensivists were accepted by local clinicians. Despite these results, the authors found that the intervention did not reduce ICU length of stay; there were also no differences in any secondary outcome compared with usual care.<\/a><\/p>\n\n\n\n This study has several strengths. First, it was well-designed with acceptable uptake of the intervention across ICUs. Although some might argue that 68% compliance with remote consultation is too low to have an effect, in reality, the \u201cright\u201d dose of telemedicine needed to impact outcomes is unknown. Moreover, we view 68% compliance as a remarkable achievement given the complexity of the intervention and realities of the COVID-19 pandemic, which occurred during the study period. Second, the intervention smartly emphasized multidisciplinary decision-making and implementation of evidence-based practices over passive remote monitoring. That is, the intervention favored the human side of ICU telemedicine rather than the technological side. This approach is consistent with the literature showing that the greatest benefit of intensivist involvement in the care of critically ill patients lies in their routine implementation of evidence-based practices rather than in recognizing and responding to clinical deterioration.8<\/a><\/sup><\/a><\/p>\n\n\n\n Overall, this study provides the strongest evidence to date that improvements in quality of care and outcomes from ICU telemedicine are not assured. At the same time, however, the lack of benefit shown in TELESCOPE does not invalidate the use of ICU telemedicine more broadly. In reality, the value of telemedicine is highly context-specific\u2014to have any effect, the intervention needs to be customized to individual sites based on local needs.9<\/a><\/sup>This truism also makes systematic evaluations like TELESCOPE inherently difficult. TELESCOPE\u2019s pragmatic study design necessarily assumed that each ICU was somewhat similar at baseline and would respond to the intervention in the same way. However, ICUs vary greatly when it comes to staffing models, resources, case mix, and culture.4<\/a><\/sup> For these reasons, it is difficult to interpret the study\u2019s findings in the absence of more granular, contextual data for each ICU. (What were the baseline staffing models of ICUs included in the study? To what extent were checklists and protocols incorporated into usual care preintervention?) Absent this information, it is difficult to assess for variation in effect of the intervention or to know whether this same intervention might have worked in other ICUs. It is also challenging to know how to refine the intervention to make it more effective.<\/a><\/p>\n\n\n\n More broadly, TELESCOPE is a vivid example of the difficulties in rigorously evaluating any technology in health care. It is instructive to consider the history of the pulmonary artery catheter, which is at first glance a simple monitoring technology. On further consideration, however, it is an intervention inexorably tied to the people who use it. First reported in 1970 and widely implemented in ICUs over subsequent decades, the pulmonary artery catheter enabled the real-time monitoring of right heart pressures and cardiac output of critically ill patients.10<\/a><\/sup>After a 1996 study showed increased mortality and cost among patients monitored with pulmonary artery catheters,3<\/a><\/sup> their use greatly declined. Subsequent studies revealed that physician experience and practice style were important factors influencing the impact of pulmonary artery catheters.11<\/a><\/sup> In this way, the effectiveness of even a simple technology like the pulmonary artery catheter is impossible to separate from its organizational context.<\/a><\/p>\n\n\n\n With this in mind, we believe TELESCOPE reveals 2 key lessons about evaluating technological innovations in the ICU. First, the authors demonstrated the ability to conduct a well-designed, pragmatic randomized clinical trial of a technological intervention in the ICU. As such, this study shows that it is possible to bring the same level of rigor that we typically expect elsewhere in clinical research. Given that it is possible to perform high-quality randomized trials of technology\u2014even complex technology like ICU telemedicine\u2014we should not settle for observational data to guide decision-making. Second, TELESCOPE highlights the need for studies of technological innovations to specifically account for the local organizational structures, resources, culture, and unmet needs of individual sites. Because effectiveness of an intervention depends on the local environment, randomized clinical trials cannot be agnostic to the context in which they take place. Instead, interventions should be flexible to the ways in which ICUs differ, allowing for customization based on local care patterns. Without this flexibility, we risk performing negative clinical trials that fail to yield actionable results.<\/a><\/p>\n\n\n\n Admittedly, this approach would mean sacrificing some of the scientific rigor desired by journal editors and research funders\u2014if the intervention can change by site, then what exactly is the intervention in the first place? To overcome this limitation, we advocate for an approach to studying critical care delivery in which rigorous methodologies are utilized to develop and tailor interventions specific to the unique needs of individual ICUs, yet are still testable in rigorous clinical trials.<\/a><\/p>\n\n\n\n Human-centered design is one such methodology to help address complex problems in existing practice paradigms.12<\/a><\/sup> A research framework in which the needs, behaviors, and experiences of end users drive product or service design and implementation, human-centered design uses a systematic, iterative approach to generate innovative solutions designed around the needs of the end user, taking into account the local environment, user characteristics, and specific workflows.13<\/a><\/sup> In the context of critical care delivery, which involves multidisciplinary teams across a variety of hospital settings, the framework of human-centered design will end up looking like team-centered design, in which interventionists engage not with individual stakeholders, but rather teams and\/or organizations as a collective whole. The end result will be interventions that may differ by site, but will do so in a way that is systematic, transparent, and reportable.<\/a><\/p>\n\n\n\n With the rise of transformative technologies like AI in health care, these issues are becoming ever more salient. If evaluation for AI proceeds like it did for ICU telemedicine, we will end up with a large number of before-and-after studies with mixed results, followed by negative randomized clinical trials. The alternative, more informative path is to develop context-sensitive technological interventions and then test them in randomized trials in which the intervention varies by site, but in a transparent way. This research will provide clinicians and policymakers with meaningful guidance about how to best implement these technologies in practice, enabling new technology to fulfill its promise to improve health.<\/p>\n","protected":false},"excerpt":{"rendered":" Editorial October 9, 2024 Evaluating Co […]<\/p>\n","protected":false},"author":3,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[24,23],"tags":[],"_links":{"self":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts\/26998"}],"collection":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=26998"}],"version-history":[{"count":1,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts\/26998\/revisions"}],"predecessor-version":[{"id":26999,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts\/26998\/revisions\/26999"}],"wp:attachment":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=26998"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=26998"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=26998"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}