A retrospective clinical observation of long-term survival in patients after prolonged mechanical ventilation

The 1-year survival rate of patients on prolonged mechanical ventilation was 29-42% before 2000. By 2012, Carson reported the 1-year survival rate of prolonged mechanical ventilation patients was 52%, an unsatisfactory improvement in survival time. We explored the long-term survival rate of prolonged mechanical ventilation patients at our hospital, 2012-2017. We analyzed retrospectively data from our respiratory care center patients to investigate the discharge status, long-term survival rate, and cause of death. We also compared the 5-year survival rates of prolonged mechanical ventilation patients and our hospital lung cancer patients. The follow-up data of The The that The survival and Continuous variables were expressed as mean ± standard deviation or median (range), whereas categorical variables were expressed as frequencies and percentage. Differences in baseline characteristics, causes of respiratory failure, and number of comorbidities were evaluated using the Student's t-test for continuous variables and Pearson chi-square tests for categorical variables. Univariate analysis was used to analyze the association of each variable with 1-year survival in all PMV patients, discharged PMV patients, and RCW patients. Multivariate stepwise logistic regression models were used to assess the factors related to 1-year survival in all PMV patients, discharged PMV patients, and RCW patients. The Kaplan-Meier method used to estimate the cumulative probability of survival as a function of the number of months in all PMV patients, discharged PMV patients, RCW patients, lung cancer patients, and stage IV lung cancer patients. The log-rank test was used to compare survival rates between groups. Cox proportional hazards models were used to determine the relationship of various factors to the survival rates of the five groups of patients. All statistical analyses were conducted using the statistical package SPSS for Windows, Version 17.0 (SPSS Inc., Chicago, IL) and a P value < 0.05 was considered to show statistical significance.


Introduction:
"The Trial Plan for National Public Health Insurance Ventilator Dependent Patients Comprehensive Care System" has been promoted in Taiwan since July 2000 [1]. The program covers mechanical ventilator care in the following settings: intensive care unit, respiratory care center (RCC), respiratory care ward (RCW), and home care service. We retrospectively studied 574 RCC patients and reported the comprehensive clinical experience of patients receiving prolonged mechanical ventilation (PMV) in 2019 [2]. Carson's study shows that patients requiring PMV have poor long-term outcomes which have not improved significantly since 2006 [3]. Damuth and colleagues reported a 1-year survival rate in all weaned units in acute care hospitals of 45.2% [4]. We sought to explore the long-term survival rate of PMV patients in one RCC over six years.

Methods:
We retrospectively studied the medical records of 574 patients admitted to the RCC from January 2012 to December 2017 and collected their age, gender, comorbidities, cause of respiratory failure, RCC discharge status, weaning status, and long-term survival rate. For those who died of pneumonia, the relevant pathogens were also recorded.
Ethics approval and consent to participate: Our study was a retrospective review and analysis of medical records. Information was de-identified before use and the research risk was identified as minimal. The project was approved by the Buddhist Dalin Tzu Chi General Hospital Research Ethics Committee (Approved IRB No.: B10802009), which exempted the study from requiring informed consent from participants.

Definitions:
A PMV patient was defined as one who used a mechanical ventilator for at least six hours daily for 21 or more consecutive days. A successfully-weaned patient was defined as one who was independent from the ventilator for five consecutive days and nights. These patients were transferred to the ward.
The unsuccessfully-weaned ventilator patients included those who died in the RCC, hospice patients in the RCC, and ventilator-dependent patients. Ventilator-dependence was defined as those who could not be weaned from the ventilator and were transferred to the RCW for long-term ventilator support.
The hospital mortality rate was determined by the number of patients who died in the hospital (including those who died in the RCC, successfully-weaned patients who died in the ward before discharge, and hospice patients) divided by the total number of patients admitted to the RCC.
This 5-year survival rate of PMV patients was also compared with that of our hospital lung cancer patients during the same period.

Cause of death in PMV patients:
We collected the cause of death of all PMV patients. If the cause of death was pneumonia, we also collect the pathogens of pneumonia.

Statistical analysis:
Continuous variables were expressed as mean ± standard deviation or median (range), whereas categorical variables were expressed as frequencies and percentage. Differences in baseline characteristics, causes of respiratory failure, and number of comorbidities were evaluated using the Student's t-test for continuous variables and Pearson chi-square tests for categorical variables.
Univariate analysis was used to analyze the association of each variable with 1-year survival in all PMV patients, discharged PMV patients, and RCW patients. Multivariate stepwise logistic regression models were used to assess the factors related to 1-year survival in all PMV patients, discharged PMV patients, and RCW patients. The Kaplan-Meier method used to estimate the cumulative probability of survival as a function of the number of months in all PMV patients, discharged PMV patients, RCW patients, lung cancer patients, and stage IV lung cancer patients. The log-rank test was used to compare survival rates between groups. Cox proportional hazards models were used to determine the relationship of various factors to the survival rates of the five groups of patients. All statistical analyses were conducted using the statistical package SPSS for Windows, Version 17.0 (SPSS Inc., Chicago, IL) and a P value < 0.05 was considered to show statistical significance.

Results:
A total of 574 patients were admitted to the RCC during the six years of the study. Of these, 357 patients (62.2%) were men and 217 (37.8%) were women. The mean age was 72.4 years. Three hundred and ninety-one patients (68.1%) were successfully weaned from the ventilator, and 83 (14.5%) unsuccessfully-weaned with ventilatordependence patients were transferred to the RCW. In all, 186 patients died in the hospital, including 95 RCC patients, 5 hospice patients, and 86 successfully-weaned patients who expired before hospital discharge. The hospital mortality rate was 32.4%.
The long-term follow-up data of 403 PMV patients included 186 who died in the hospital, 157 discharged PMV patients, and 60 RCW patients followed until October 30, 2018. The 1-year survival rate of all PMV patients was 24.3%. The mean age of all PMV patients who survived for more than one year was significantly different from that of those who survived less than one year (67.4 years vs. 75.0 years, P < 0.001). In all PMV patients, those with no comorbidity (P = 0.002, odds ratio [OR] = 3.645) and successfully weaned patients (P < 0.001, OR = 3.492) had better 1-year survival rates. Patients older than 75 years (P = 0.005, OR = 0.464), those with ESRD (P = 0.040, OR = 0.275), and those with four or more comorbidities (P = 0.021, OR = 0.180) had poorer 1-year survival rates [2]. The 1-year survival rate of discharged PMV patients was 50.3%. The mean age of discharged PMV patients who survived for more than one year was significantly different from that of those who survived less than one year (66.69 years vs. 78.04 years, P < 0.001). Discharged PMV patients older than 75 years had a poorer 1-year survival rate (P = 0.001, OR = 0.308), as did those with four or more comorbidities (P = 0.031, OR = 0.094). The 1year survival rate of RCW patients was 31.7%. The mean age of RCW patients who survived for more than one year was no significantly different from that of those who survived less than one year (72.36 years vs 69.78 years, P = 0.513). Multivariate analysis showed no significant difference in the clinical variables of RCW patients who did or did not survive one year or more.
Pneumonia was the leading cause of death in PMV patients. In all, 159 PMV patients (39.5%) died due to pneumonia, including 106 patients with detected pneumonia pathogen. Sixty percent of detected pneumonia pathogens were polypathogenic. The most-frequent pneumonia pathogens in PMV patients were, in order, Acinetobacter baumannii (29.7%), Pseudomonas aeruginosa ( 17.7%) and Klebsiella pneumonia ( 12.0%).
Kaplan-Meier analysis of all PMV patients, discharged PMV patients, and RCW patients showed that discharged PMV patients had the best 5-year survival rate. The 5-year survival rate did not significantly differ between all PMV patients and RCW patients. Compared to our hospital lung cancer patients, discharged PMV patients had a better 5-year survival rate than all lung cancer patients and stage IV lung cancer patients. All PMV patients had a worse 5-year survival rate than all lung cancer patients. The 5-year survival rate did not significantly differ between all PMV patients and stage IV lung cancer patients. RCW patients had a worse 5-year survival rate than all lung cancer patients. The 5-year survival rate did not significantly differ between RCW patients and stage IV lung cancer patients. (Fig. 2 and Fig. 3).

Discussion:
Overall 1-year survival rate of all PMV patients: Dettmer identified six factors (increasing age, vasopressor use, thrombocytopenia, acute kidney injury, inability to wean from the ventilator, and preexisting kidney injury) that had a strong association with mortality in PMV patients [5]. The literature reports that the 1-year survival rate of PMV patients ranges 37.2-61% (Table 1) [ [6][7][8][9][10][11][12][13][14]. Older age, failure to wean, and a higher Acute Physiology, Age, Chronic Health Evaluation II (APACHE II) score were associated with a poor 1-year survival rate. In Carson 's study [11], patients older than 65 years had a poorer 1-year survival rate than younger patients. In contrast, we found that being older than 75 years was associated with a poor 1-year survival rate. Carson and colleagues reported that thrombocytopenia, requiring vasopressors, and ESRD were three other factors related to a poor 1-year survival rate [11]. Huang's study revealed that congestive heart failure, ESRD, malignancy, and liver cirrhosis were factors related to a poor 1-year survival rate, but having a tracheostomy was related to an excellent 1-year survival rate [14]. Lin and colleagues found that young age and absence of liver cirrhosis were associated with a good 1-year survival rate [13]. In our series, having ESRD or four or more comorbidities was associated with a poor 1-year survival rate. Patients with no comorbidities and those successfully weaned from ventilation had an excellent 1-year survival rate. Despite our high rate of successful weaning (68.1%), our PMV patients had an abysmal 1-year survival rate (24.3%), which was due to our high hospital mortality rate. The factors influencing the 1-year survival rate of PMV patients agreed with those reported in the literature.   [18]. Clearly, weaning from the ventilator is important for long-term survival in PMV patients.

Causes of death in PMV patients:
Pneumonia was the leading cause of death in our PMV patients. Carbapenems-resistant Gram-negative bacilli make up 35.4% of detected pneumonia pathogens, especially CRAB. Chien showed that the rise of multidrugresistant A. baumannii was significantly correlated with previous exposure to piperacillin/tazobactam and imipenem. The multidrug-resistant microorganisms that cause pneumonia have been associated with increased 6-month mortality in PMV patients [19]. Another Taiwan's study showed that the prior use of imipenem, meropenem, piperacillin/tazobactam, or fourth-generation cephalosporins was an the independent risk factor for extensively drug-resistant A. baumannii hospital-acquired infections [20]. Having a carbapenems-resistant Gramnegative bacillus infection was fatal in our PMV patients. We must pursue the prudent use of broad-spectrum antibiotics, especially the strict use of carbapenems in our hospital.
Declining mortality rate in PMV patients: The declining mortality rate differed among all PMV patients, discharged PMV patients, and RCW patients. All PMV patients showed a steep decline in mortality within the first three months, a slower decline in the second year, and a subsequent plateau in mortality after two years. Discharged PMV patients showed a steady decline in mortality rate within the first year, a slower decline in the second year, and a subsequent plateau after two years.
RCW patients had a very steep decline in mortality within the first six months, a slower decline in the second year, and a subsequent plateau in decline after two years. RCW patients displayed a faster decline in mortality rate than discharged PMV patients, with the same 2-year mortality rate as for all PMV patients. This finding means that, despite the long-term invasive mechanical ventilator support they received, RCW patients did not enjoy a better long-term survival rate. Again, successful weaning from the mechanical ventilator is a key factor in the long-term outcome of PMV patients. In Stoller's study, the mortality rate of patients discharged from the weaning unit steeply declined in the first two years (68%), with a slower decline after that (5-year mortality rate, 81%) [6]. The mortality rates and trends in decline of our PMV patients were similar to those reported in the series by Stollers. Aboussouan and colleagues reported that 55% of patients (both weaned and not weaned) died within the first year after discharge, with a subsequent plateau in mortality such that 40% remained alive by the second and third years [21]. In our series of weaned and not weaned patients, the 1-year, 2-year, 3-year, and 5- year survival rates were 45,2%, 37.3%, 35.0%, and 32.3%, respectively, slightly less than the survival rates reported by Aboussouan. Conclusion: The 1-year survival rates of all PMV patients, discharged PMV patients, and RCW patients were 24.3%, 50.3%, and 31.7%, respectively. The 5-year survival rates of all PMV patients, discharged PMV patients, and RCW patients were 18.1%, 40.05%, and 11.7%, respectively. The 5-year survival rate of discharged PMV patients was better than that of lung cancer patients. However, the 5-year survival rates of all PMV patients and RCW patients were comparable to stage IV lung cancer patients. Successful weaning from mechanical ventilation and pneumonia prevention for PMV patients are key factors in improving long-term outcomes in PMV patients. Figure 1 The five years mortality rate of all PMV patients, RCW patients and Discharged PMV patient The Kaplan-Meier curve of all PMV patients, RCW patients and discharged PMV patient and lung cancer patients Figure 3 The Kaplan-Meier curve of all PMV patients, RCW patients and discharged PMV patient and lung cancer stage IV patients