Expanding Use of the ProVent Score
Article information
Although advances in intensive care have enabled more patients to survive through the acute stage of critical illness, approximately 5%–10% of mechanically ventilated patients experience difficulty in weaning and thus require prolonged mechanical ventilation (PMV), which has been defined as a period of 21 days or more1. PMV is associated with a disproportionate use of medical resources resulting in substantial health care costs as well as high mortality after intensive care2. A recent meta-analysis of 124 studies from 16 different countries worldwide reporting clinical outcomes in PMV further highlights short- and long-term survival information for patients treated with PMV: approximately 30% of the patients requiring PMV did not survive to hospital discharge and over half of the patients died within 1 year3.
For these reasons, the Prolonged Mechanical Ventilation Prognostic Model (ProVent score) was developed and validated to predict 1-year mortality of patients requiring PMV, which is derived by assessing 1 point to each of four readily ascertained variables assessed on day 21 of mechanical ventilation4: need for dialysis, need for vasopressors, age 50 years or older, and platelet count less than 150×109. It has been modified by adding another age group (65 years or older), to which 2 points are assigned, yielding a 6-point score ranging from 0 to 55. These variables are objective and easy to extract so that can be linked within electronic medical records to inform clinical care. This simplicity leads to actual clinical use rather than limited to research application only, to help inform management decisions for patients receiving PMV. To date, this tool has been validated in Korean populations with simple clinical data obtained at 14 and 21 days of mechanical ventilation678.
However, the ProVent score is not perfect for assessing prognosis for the individual patient. Based on the clinical characteristics not included in ProVent, there are clearly unique patients who can reasonably expect to be much better or worse than the model predicts8. Another current limitation of ProVent is its inability to predict outcomes beyond survival that patients and families deeply care about, such as chronic critical illness and related medical resource utilization.
In this issue of Tuberculosis and Respiratory Diseases, Roh et al.9 present additional research looking for the predicted value of ProVent score in medical expenses as well as intensive care unit and hospital mortalities of patients requiring PMV. In a retrospective cohort design among 305 patients received PMV, the authors found that three components of ProVent model, such as age less than 50 years, platelet count, and need for dialysis, were associated with high medical costs. Further extending their exploration of the model, the authors also found that the ProVent score was significantly associated with short-term outcome during hospitalization, consistent with the previous report10. One of the primary weaknesses of the study, as identified by the authors, is that the true total costs of PMV patients would be underestimated because of their common omission of post-discharge care expenses. In addition, other clinical characteristics that are not included in the ProVent could be associated with high medical costs. Nonetheless, the two major findings suggest that the use of ProVent score can be expanded to predict clinical outcomes in patients requiring PMV.
Notes
Conflicts of Interest: No potential conflict of interest relevant to this article was reported.