Many articles have discussed the factors affecting the successful weaning rate of PMV patients [2, 4–11]. The successfully weaned rate ranges from 32.3–68.1%. In Damuth reports, only 50% were successfully liberated from mechanical ventilation [12]. The many predictors of successful weaning include cause of leading patient to need for PMV, blood urea nitrogen levels, APACHE II score, albumin level, and the number of comorbidities. Many articles also discuss the factors related to 1-year survival in PMV patients. The 1-year survival rate of PMV patients also ranges widely, from about 24.3–61% [2, 5, 13–20]. The many factors related to 1-year survival include successful weaning from the ventilator, age, comorbidity, APACHE II score, thrombocytopenia, and the need for vasopressors. The literature demonstrates that being weaned from mechanical ventilation is a key factor in the long term outcomes of PMV patients. Although several studies of the PMV successful weaning rate and 1-year survival rate have been published, they vary in their definition of PMV and in their definition of the discharge status of successfully weaned PMV patients. Besides, PMV patients are admitted to weaning centers in western countries, and those successfully weaned PMV patients are immediately discharged; thus, none of these patients die in the ward. Our RCC is a weaning unit in an acute care hospital. Many PMV patients were successfully liberated from the ventilator in the RCC, then transferred to the ward, only to die before being discharged. To our knowledge, no studies have yet included successfully weaned PMV patients who died in the ward, nor did we find any article in the international literature that addressed long term survival in successfully weaned PMV patients. Therefore, in the discussion, we can analyze only our research results, and cannot compare these with results from other institutions.
Factors affecting survival in successfully weaned PMV patients who died in the ward
The outcome of patients who died in the ward was the worst in the three groups of successfully weaned PMV patients. The most common cause of death in patients who died in the ward was pneumonia (44 patients), followed by respiratory failure. When patients suffer from pneumonia or impending respiratory failure, they may need ventilator support again. Because the patients or their surrogates signed DNR/ DNI orders, these patients received no additional critical care or ventilator support, and they expired before they were discharged from the hospital.
The multivariate analysis of clinical variables and end-of-life decisions between patients who died in the ward and those who survived < 1 year showed that only signing DNR/ DNI orders was significantly different between the two groups. Signing DNR/ DNI orders is the key factor in the worse survival of patients who died in the ward compared to those who survived < 1 year.
The multivariate analysis of clinical variables between patients who died in the ward and those who survived ≥ 1 year revealed the poorer survival of patients who died in the ward was due to a higher percentage of ESRD comorbidity (13.9%) and a higher percentage of malignant comorbidity (18.6%) and higher percentage of ⩾ Four comorbidities (18.6%). As above, another factor in poorer survival was the higher percentage of signed DNR/ DNI orders in patients who died in the ward than in those who survived ≥ 1 year.
Factors affecting survival in successfully weaned PMV patients who survival < 1 year
The most common cause of death in patients who survived less than one year was pneumonia (39 patients), followed by OHCA at home. We could not trace the cause of OHCA but considered that sudden respiratory failure might have been a common cause.
The multivariate analysis of clinical variables of poor survival in survived less than one-year patients that were associated with two factors. One was age; patients who survived < 1 year were older than those who survived longer. Of those who survived < 1 year, 66.7% were older than 75 years, compared to only 35.4% of those who survived ≥ 1 year. Another factor is end-of-life decision (signed DNR/ DNI orders) was significantly different between successfully weaned PMV patients who survived < 1 year and those who survived ≥ 1 year.
End-of-life decisions in successfully weaned PMV patients
In our study, 177 patients (72.8%) signed the DNR order, including 81 patients (33.3%) who signed DRN and DNI orders. This result means that, at most, 66.7% of the survivors were willing to receive mechanical ventilation again. In Jubran’s series, 84.7% of survivors indicated a willingness to undergo mechanical ventilation again [21]. Our patients were less willing to receive second mechanical ventilation than were Jubran's patients.
Do we need to improve the survival outcome of successfully weaned PMV patients who died in the ward?
Patients in our series displayed an excellent successful weaning rate (68.1%) but an abysmal 1-year survival rate (24.3%), and we inferred this abysmal 1-year survival rate resulted from our high ward mortality. Only 305 successfully weaned PMV patients were discharged. Fully 86 (22.0%) patients died before they were discharged from the hospital. Being successfully weaned did not substantially help these patients. Fully 69.8% of the ward mortality patients signed DNR and DNI orders. Why did patients' families sign DNR/ DNI orders for these patients? Despite being successfully weaned from the mechanical ventilator, patients experienced suffering in their critical care with PMV support. This experience is a very shocking and painful one for patients. The patients are in poor health status or unconscious in the ward after successful weaning and cannot recover in the short term. Patients' families do not want their loved ones to go through that experience again. They believed that palliative therapy was a greater benefit for these patients. We have set up a shared decision-making program to address this situation (appendix). Through a medical decision-making program to help families understand the clinical course of PMV patients, the quality of medical treatment will improve, and long term medical care will be promoted. This is our goal for PMV patients.
Limitations of our study
We did not collect patient's laboratory data, respiratory parameters, APACHE II score, Glasgow Coma Scale, or other similar relevant variables. We were, therefore, unable to determine which of these measures, if any, may be related to the long term survival of successfully weaned PMV patients. The literature contains no discussion of successfully weaned PMV patients who died in the ward after successful weaning, nor studies of long term survival in successfully weaned PMV patients. Our conclusions on the long term outcomes of successfully weaned PMV patients must be interpreted with care since they derive only from our retrospective single-unit study. We have no proper strategy to improve the survival rate of ward mortality patients. We expect that the addition of other clinical experiences and strategies to improve long term survival in successfully weaned PMV patients will yield additional insights.