1-Year Survival of Subjects Discharged From a Long-Term Chronic Ventilator Unit

Discussion

We found that subjects discharged from our long-term chronic ventilator unit were most likely to be alive after 1 y if they had shorter stays in the ICU and were sent home versus to a SNF. Neither the subjects' initial severity of illness upon presentation to an ICU nor the length of stay at our long-term chronic ventilator unit were significant predictors of mortality after long-term chronic ventilator unit discharge.

The survival of an acute critical illness is often the beginning of a long medical journey for patients. For example, Lone and Walsh discussed the association of multi-organ failure on mortality 5 y after a critical illness.²⁰ They evaluated the status of organ systems according to the SOFA definition in 745 subjects during their ICU stay. Using a multi-variable analysis adjusting for each organ failure and confounders, they found that cardiovascular, hepatic, and respiratory failure carried the greatest risk of death within 5 y after discharge from the ICU (odds ratio 2.1–2.5). However, all individual organ failures were associated with a high 5-y mortality (P < .001), suggesting persistent negative effects from the acute illness on long-term health. Lone and Walsh did not describe where the ICU survivors received their post-ICU care or the many resources likely allocated to allow these subjects the chance to ultimately survive their acute illness.²⁰ Because the critical illness frequently extends beyond the initial ICU stay, it is logical to assume that the post-ICU care may have a significant impact on the final outcome of the original critical illness. As the management of chronic critically ill patients is fairly new, best practices for these patients are unclear; however, it appears from our study that if patients can be ultimately discharged home (with home discharge serving as a proxy for medical stability with or without significant recovery of physiological function), they will have an associated survival benefit.

The syndrome of chronic critical illness is complex. Most patients are older adults with underlying comorbid conditions who develop an acute illness requiring critical care (eg, sepsis, ARDS).^8,21 The hallmark of chronic critical illness is prolonged ventilation, but other characteristic clinical features include neuromuscular weakness, brain dysfunction, malnutrition, and complex wounds.^8,21 The prevalence of this complex syndrome is increasing. For example, in a retrospective cohort conducted between 2004 and 2009 from Massachusetts, North Carolina, Nebraska, New York, and Washington, the prevalence rate of chronic critical illness was 34.4 per 100,000 persons, with a steady increase in patients < 75 y (peaking at 82.1 per 100,000 persons for the 75–79 y age group).²² Extrapolating these data to the United States as a whole, the 20% increase of chronic critical illness prevalence in the aforementioned time frame increased in-hospital costs from $15.6 billion in 2004 to $26.0 billion in 2009.²² Novel ways to manage these patients in a cost-effective way must be seen as a priority for both physicians and policy makers.

The concept of the LTACH was developed to optimally manage patients with chronic critical illness. LTACHs are defined by the United States Centers for Medicare and Medicaid Services as acute care hospitals with an average stay ≥ 25 d.¹⁴ As previously mentioned, LTACHs were established in the early 1980s when a shift toward prospective payment for United States hospitals began; ie, hospitals were paid a set amount for each patient rather than have payments determined by cost.²³ Over time, LTACHs evolved from purely respiratory-focused hospitals (such as a long-term chronic ventilator unit) to hospitals that provide care for patients with all types of chronic critical illnesses.^8,19 Moreover, the number of LTACHs has been on the rise, with a 33% increase from 2003 to 2007 associated with a doubling of Medicare spending toward these facilities from $2.7 billion to $5.2 billion. LTACHs were shown to utilize less health care cost while achieving similar survival in the management of chronic critically ill patients versus ongoing ICU management.²² However, not all LTACHs are the same.^19,24 One type of LTACH is the SNF, which is seen as more cost-effective versus other LTACHs in the sense that less intensive care is offered, presumably for an ideal patient who needs less intensive care. Yet how to determine the ideal patient has not been well demonstrated in practice and warrants further evaluation in future trials.

In our cohort of subjects, the stay in our long-term chronic ventilator unit did not differ between survivors and non-survivors. However, the stay in our long-term chronic ventilator unit was likely influenced not by subjects demands and physiological parameters, but rather by the NHT202 policy in which emphasis was placed on sending stable patients to more cost-effective areas. As mentioned before, how to best define the ideal patient suitable for downgrade to other LTACHs is challenging because there are no ideal guidelines to suggest which patients are more prepared than others, or what step-down units need to specifically offer to help patients end their long medical journey at home. Furthermore, identifying patients who are clinically ready to be sent home may be more obvious than those who need more time with a higher level of care in LTACHs, which could thus impact survival if those patients are transferred prematurely. Therefore, while our study cannot speak to the safety of transfer from one LTACH to another, we strongly advocate that more attention be placed on patients requiring longer time in chronic care, specifically around resource allocation and best practices that ultimately have our ICU survivors return home.

Defining a longitudinal care model for post-ICU patients with ongoing intensive care needs must become a priority. Because these patients impact many aspects of medical care, family life, and health budget, and because they have a future that remains uncertain with regard to mortality and morbidity, a health model that in of itself is uncertain and unexplored will only add to anxiety, confusion, and cost to all parties involved. Furthermore, as noted previously, LTACHs are not all the same and have varying resources that may or may not be appropriate for certain patients. Prior publications that discuss the benefit of LTACHs are confounded in that the type of services they provide is not often clear (eg, ratio of physical therapists to patient; respiratory therapist to patient, etc).¹⁷ However, as the field continues to build data with regard to the best care that LTACHs can provide, health care providers should be aware of current policies that may influence the management of LTACH patients. For instance, in 2006, the state of Maryland passed the NHT202 with the intention to facilitate access to ventilator care while guaranteeing that patients receive services that are most cost-effective.¹⁶ With the NHT202, a utilization control agent performs periodic continued-stay reviews to determine whether chronic hospital services remain necessary and appropriate.¹⁶ In other words, if a patient could receive a lower level of care, then the control agent would initiate the process for a transfer. While we advocate for more research to be allocated to these chronic critically ill patients managed in LTACHs, we wish to emphasize that physicians and health care providers become aware of policies that may affect the care of these patients. When the science is able to address the management of these patients in a more robust manner, such policies should be reassessed to assure they are in accordance with the data.

The creation of an effective longitudinal care model for critically ill survivors will take time. However, implementing certain interventions now may help create the background data necessary to better understand these patients' pathophysiology, needs, and prognosis. First, more research is necessary to better characterize these patients. The biochemical composition of these patients is complex, though may parallel other existing physiological profiles. For instance, the frailty phenotype encompasses malnutrition, deconditioning, and cardiovascular issues,²⁵ all of which are present in those who have chronic critical illness. Thus, collaborative efforts may expedite our understanding of post-ICU patients, which may lead to interventions. Second, education and awareness of these patients is variable. Specifically, critical care medicine is learning more and more of morbidity outcomes in its survivors, yet few interventions are known and well implemented that will positively affect patients beyond the ICU. Third, ownership of the care of the chronic critically ill is hard to define and likely does not constitute one profession. The difficulty in establishing one profession to be the champions of these patients is likely hindered by frequent transfers of care and the varying backgrounds of health providers at those facilities. Therefore, training to care for these patients must be a priority. Finally, there should be more open and honest discussions with the families of these patients, specifically during the ICU stay. The families often have poor insight into the journey they are about to embark on, especially because as their loved one is downgraded in care, the care will begin to appear less synchronized and organized. One variable that contributes to the poor insight of the families is likely the lack of knowledge from the critical care team and/or hospital team that cares for the patients before they leave the acute hospital setting. Thus, education and awareness of health providers who initially care for these patients is needed so that family meetings can be as informative as possible to discuss the medical expedition ahead.

Our study has several limitations. First, subjects who left our long-term chronic ventilator unit and were re-admitted to the hospital were not further followed. This represents an important subgroup of patients who require repeat hospitalizations while still receiving intensive care therapy; future studies should explore these subjects further. Second, we identified diagnoses based on International Classification of Diseases, 9th Revision, which is subject to coding errors. Third, we only explored one long-term chronic ventilator unit and discharges within one state. We did this so we could address the health care policies of Maryland and its impact on post-ICU patient care. Thus, this investigation may serve to help initiate change in policy, in Maryland and in other states. Fourth, we only addressed mortality after discharge from long-term chronic ventilator unit and did not explore other morbidities (eg, mental health) that are known to be prevalent on post-ICU discharges and how they were addressed at home or at SNFs. Fifth, exploring family concerns for these patients should be a priority for future studies. For instance, the transition from one facility to another may have emotional and health consequences on family members that may negatively affect them; these were not explored. Families may also have reservations in allowing patients to come home (eg, economic barriers); these were also not explored, but both should be a focus of future studies. Sixth, there is no clinical score to evaluate the candidacy of a patient to be transferred from one LTACH to another (while there are scores to assess whether a patient needs an LTACH²³); we did note whether subjects left the long-term chronic ventilator unit with a tracheostomy in place, but we did not have a formal way to assess the level of complex care patients needed in a manner similar to how a SOFA score implies level of severity for patients needing critical care. A formal scoring system should be explored for patients in an LTACH, especially if transfer of care is considered. Seventh, we did not review deaths in the long-term chronic ventilator unit; however, this should be investigated and better understood with regard to subjects who do not survive this initial transfer. Finally, we did not explore the medical resources available at the SNF (eg, how many respiratory therapists were there, nursing to patient ratio, frequency of physician visits). Addressing these in the future may provide better guidance for policy and institutional changes.