Prescribing Patterns of Hydrocortisone in Septic Shock
Prescribing Patterns of Hydrocortisone in Septic Shock
The study was performed at the Emory Center for Critical Care (ECCC), which was comprised of eight ICUs totaling 135 beds throughout the course of the study (it has subsequently expanded). The ECCC includes ICU beds from both Emory University Hospital, a 579-bed academic hospital, and Emory University Hospital Midtown, a 511-bed hybrid academic/community hospital. The ICUs examined included two medical ICUs, one surgical/transplant ICU, two cardiovascular surgery ICUs, two neuroscience ICUs, and one coronary care unit. Each ICU was staffed by a single intensivist who rotated in 1- to 2-week increments. Although some intensivists rotated in only a single ICU, many rotated in different ICUs in the same hospital or at both hospitals. Each ICU also had nurse practitioners or physician assistants (affiliate providers) in house 24 × 7 and half had residents from the departments of medicine, anesthesiology, emergency medicine, and surgery.
The study had two components. First, a retrospective chart review was performed to examine provider prescribing patterns of corticosteroids in septic shock. While the data were being analyzed, a survey was sent to all intensivists within the ECCC to determine whether self-described attitudes toward prescribing of corticosteroids in septic shock were similar to actual prescribing patterns. Emory University's institutional review board approved both components of the study and waived the requirement for obtaining written informed consent.
All patients who had the Emory severe sepsis protocol initiated (identified via electronic order of the protocol) between November 1, 2009, and April 30, 2011, were evaluated. The Emory severe sepsis protocol is a standardized approach to sepsis management. The primary elements of the protocol are based upon Early Goal-Directed Therapy and include management in the first 6 hours related to fluid resuscitation, vasopressors, transfusion, inotropes, as well as timely initiation of antibiotics. The protocol next includes elements of a management phase from 6 to 24 hours including blood glucose management, respiratory management, venous thromboembolism prophylaxis, and stress ulcer prophylaxis. The final component of the protocol is a section entitled "consider" that includes the statement "consider intravenous corticosteroids (hydrocortisone 50 mg IV q6hr) in patients with septic shock who despite adequate fluid replacement require vasopressor therapy to maintain adequate blood pressure." No electronic order is associated with this suggestion in the protocol, and providers wishing to order corticosteroids must separately order hydrocortisone in a manner similar to entering any other medication in computer order entry. All ICU patients were screened twice a day by the nursing staff for the potential presence of severe sepsis via vital signs monitoring and the suspicion of infection. The decision to initiate the protocol was made by either a physician or an affiliate provider.
Patients on whom the severe sepsis protocol was initiated were further evaluated if the patient's medical record contained: 1) an order for vasopressors (defined as norepinephrine, phenylephrine, vasopressin, epinephrine, dopamine) 1 hour before or up to 24 hours after protocol initiation; 2) an order for IV hydrocortisone within 72 hours of protocol initiation; and 3) evidence that vasopressors were actually initiated on the patient within 24 hours of protocol initiation defined by charting on the electronic medical record of vasopressor infusion. The rationale for these inclusion criteria were to evaluate only patients who were: 1) identified as being septic; 2) further identified as being in septic shock requiring vasopressors; and 3) had a new order written for corticosteroids after the diagnosis of septic shock. Patients not meeting these three criteria were not analyzed further. Patients who received systemic corticosteroids within 90 days prior to initiation of the severe sepsis protocol were also excluded from further analysis. Recent corticosteroid usage was identified by inpatient orders, documented medication history, discharge prescriptions, and clinic prescriptions if the prescription would still have been active less than 3 months prior to protocol initiation.
Because there were no established ranges in the literature to categorize vasopressor usage as low, moderate, or high dose that we were able to identify, pre-hoc definitions were established for the purpose of this study (Table 1). These dosing ranges were established on an individual agent basis, and it is not the intent of the authors that the categories be interpreted as equipotent dosing of vasopressor agents.
For the component of the study that examined weaning or discontinuation of corticosteroid therapy, the first change in hydrocortisone dosing regimen was defined as the time of the first order for a dose decrease or the discontinuation of the hydrocortisone order.
To determine whether provider attitudes toward prescribing corticosteroids for septic shock matched actual behavior, a survey was developed to ascertain provider beliefs relevant to this issue. All intensivists practicing in the ECCC received a nine-question electronic survey in April 2012. Survey questions were developed to assess adherence to recommendations from the Surviving Sepsis Campaign guidelines and on the specific types of data that were collected in the first component of the study examining prescribing behavior within the ECCC. Once the survey was completed, responses were anonymously recorded in an online database, and aggregated responses were compared to actual prescribing patterns.
Materials and Methods
Location
The study was performed at the Emory Center for Critical Care (ECCC), which was comprised of eight ICUs totaling 135 beds throughout the course of the study (it has subsequently expanded). The ECCC includes ICU beds from both Emory University Hospital, a 579-bed academic hospital, and Emory University Hospital Midtown, a 511-bed hybrid academic/community hospital. The ICUs examined included two medical ICUs, one surgical/transplant ICU, two cardiovascular surgery ICUs, two neuroscience ICUs, and one coronary care unit. Each ICU was staffed by a single intensivist who rotated in 1- to 2-week increments. Although some intensivists rotated in only a single ICU, many rotated in different ICUs in the same hospital or at both hospitals. Each ICU also had nurse practitioners or physician assistants (affiliate providers) in house 24 × 7 and half had residents from the departments of medicine, anesthesiology, emergency medicine, and surgery.
Study Design
The study had two components. First, a retrospective chart review was performed to examine provider prescribing patterns of corticosteroids in septic shock. While the data were being analyzed, a survey was sent to all intensivists within the ECCC to determine whether self-described attitudes toward prescribing of corticosteroids in septic shock were similar to actual prescribing patterns. Emory University's institutional review board approved both components of the study and waived the requirement for obtaining written informed consent.
Provider Prescribing Patterns
All patients who had the Emory severe sepsis protocol initiated (identified via electronic order of the protocol) between November 1, 2009, and April 30, 2011, were evaluated. The Emory severe sepsis protocol is a standardized approach to sepsis management. The primary elements of the protocol are based upon Early Goal-Directed Therapy and include management in the first 6 hours related to fluid resuscitation, vasopressors, transfusion, inotropes, as well as timely initiation of antibiotics. The protocol next includes elements of a management phase from 6 to 24 hours including blood glucose management, respiratory management, venous thromboembolism prophylaxis, and stress ulcer prophylaxis. The final component of the protocol is a section entitled "consider" that includes the statement "consider intravenous corticosteroids (hydrocortisone 50 mg IV q6hr) in patients with septic shock who despite adequate fluid replacement require vasopressor therapy to maintain adequate blood pressure." No electronic order is associated with this suggestion in the protocol, and providers wishing to order corticosteroids must separately order hydrocortisone in a manner similar to entering any other medication in computer order entry. All ICU patients were screened twice a day by the nursing staff for the potential presence of severe sepsis via vital signs monitoring and the suspicion of infection. The decision to initiate the protocol was made by either a physician or an affiliate provider.
Patients on whom the severe sepsis protocol was initiated were further evaluated if the patient's medical record contained: 1) an order for vasopressors (defined as norepinephrine, phenylephrine, vasopressin, epinephrine, dopamine) 1 hour before or up to 24 hours after protocol initiation; 2) an order for IV hydrocortisone within 72 hours of protocol initiation; and 3) evidence that vasopressors were actually initiated on the patient within 24 hours of protocol initiation defined by charting on the electronic medical record of vasopressor infusion. The rationale for these inclusion criteria were to evaluate only patients who were: 1) identified as being septic; 2) further identified as being in septic shock requiring vasopressors; and 3) had a new order written for corticosteroids after the diagnosis of septic shock. Patients not meeting these three criteria were not analyzed further. Patients who received systemic corticosteroids within 90 days prior to initiation of the severe sepsis protocol were also excluded from further analysis. Recent corticosteroid usage was identified by inpatient orders, documented medication history, discharge prescriptions, and clinic prescriptions if the prescription would still have been active less than 3 months prior to protocol initiation.
Because there were no established ranges in the literature to categorize vasopressor usage as low, moderate, or high dose that we were able to identify, pre-hoc definitions were established for the purpose of this study (Table 1). These dosing ranges were established on an individual agent basis, and it is not the intent of the authors that the categories be interpreted as equipotent dosing of vasopressor agents.
For the component of the study that examined weaning or discontinuation of corticosteroid therapy, the first change in hydrocortisone dosing regimen was defined as the time of the first order for a dose decrease or the discontinuation of the hydrocortisone order.
Provider Attitudes Toward Prescribing Corticosteroids for Septic Shock
To determine whether provider attitudes toward prescribing corticosteroids for septic shock matched actual behavior, a survey was developed to ascertain provider beliefs relevant to this issue. All intensivists practicing in the ECCC received a nine-question electronic survey in April 2012. Survey questions were developed to assess adherence to recommendations from the Surviving Sepsis Campaign guidelines and on the specific types of data that were collected in the first component of the study examining prescribing behavior within the ECCC. Once the survey was completed, responses were anonymously recorded in an online database, and aggregated responses were compared to actual prescribing patterns.
Source...