ASA Monitor 10 2018, Vol.82, 60-61.
A recent retrospective cohort study tallied administrative and pharmacy billing data collected during the years 2008 to 2013 on patients with septic shock admitted to hospitals. The authors analyzed the impact of the 2011 norepinephrine shortage in the U.S. According to the results of the study, which of the following is most likely true?
- (A) Overall, the percentage of patients receiving norepinephrine decreased over the five years.
- (B) Among hospitals most impacted by the norepinephrine shortage, norepinephrine usage decreased approximately 75 percent during the shortage period.
- (C) In-hospital mortality among patients treated in the most impacted hospitals increased during the shortage period.
- (D) The most common alternative vasopressor used during the shortage period was epinephrine.
Announcements of drug shortages are becoming increasingly common and have drawn the attention of not only physicians but also legislators and the public at large. However, the impact these shortages have had on patient outcomes has largely been unstudied.
The authors of a recent retrospective cohort study sought to evaluate the impact the 2011 norepinephrine shortage in the U.S. had on the care of patients with septic shock. The primary goals of the study were to determine changes in mortality rates and the most common alternative vasopressor medication used during the shortage period.
An administrative and pharmacy billing database was used to access the information of interest between July 1, 2008, and June 30, 2013. The study period was broken down into quarters (three-month periods) with the baseline study period for initial data assessment occurring from July 2008 to July 2010. Hospital-level vasopressor use for each medication type was defined as the percentage of patients receiving that particular vasopressor, either alone or in combination with other vasopressors, for the management of septic shock. Hospitals that used norepinephrine for less than 60 percent of patients with septic shock were excluded from the study. The remaining hospitals were stratified into shortage hospitals and consistent-use hospitals. Shortage hospitals were defined as those that had 1) a relative decrease of more than 20 percent in norepinephrine use from baseline in at least one quarter of 2011, 2) a return to norepinephrine use rates within 10 percent of the baseline rate by the second quarter of 2012, and 3) no more than one quarter of norepinephrine use that was more than 20 percent below baseline before or after 2011. Consistent-use hospitals were defined as those that did not have a decrease in norepinephrine use of more than 20 percent in any of the four quarters of 2011. The authors of the study used consistent-use hospitals to perform a difference-in-differences analysis with shortage hospitals.
A total of 168,304 patients from 150 hospitals were included in the initial analysis. Of these hospitals, 26 (17.3 percent) were classified as shortage hospitals, 102 (68 percent) were classified as consistent-use hospitals and the remaining 22 (14.7 percent) did not fit into either classification. Norepinephrine use increased over the five-year study period (76.7 percent in the third quarter of 2008 vs. 80 percent in the second quarter of 2013). During the baseline study period, overall norepinephrine use was 78.5 percent (95 percent CI, 78.2 percent – 78.7 percent), which decreased over the shortage period, and then recovered to pre-shortage levels in 2012. The mean baseline use of norepinephrine was 79.3 percent (95 percent CI, 78.9 percent – 79.6 percent) for consistent-use hospitals with no change throughout the 2011 shortage period. However, among shortage hospitals, baseline norepinephrine use was 77 percent (95 percent CI, 76.2 percent – 77.8 percent) and decreased to a low of 55.7 percent (95 percent CI, 52.0 percent – 58.4 percent) in the second quarter of 2011. Among shortage hospitals, mean phenylephrine use increased from a baseline rate of 36.2 percent (95 percent CI, 35.3 percent – 37.1 percent) to a maximum of 54.4 percent (95 percent CI, 51.8 percent – 57.2 percent) in the second quarter of 2011.
The adjusted odds ratio for patient mortality increased to 1.15 (95 percent CI, 1.01 – 1.30) from baseline among shortage hospitals with decreased norepinephrine use of 20 percent or more during the study quarters of 2011. This in turn corresponded to an absolute mortality difference of 3.7 percent (95 percent CI, 1.5 percent – 6.0 percent). A difference-in-differences comparison of in-hospital mortality rates between shortage hospitals and consistent-use hospitals showed a similar association, with an adjusted odds ratio of 1.17 (95 percent CI, 1.06 – 1.31) for mortality among shortage hospitals in 2011.
In summary, patients with septic shock admitted to hospitals impacted by the U.S. norepinephrine shortage in 2011 had higher odds of in-hospital mortality. The most common alternative vasopressor used during this shortage period was phenylephrine. The findings of this study suggest that the norepinephrine shortage of 2011 may have resulted in hundreds of additional deaths in the U.S.
An accompanying editorial raised concerns about these possible inferences. The diagnosis of septic shock relied on administrative data; there was no adjudication or validation of this diagnosis by the investigators. The analyses generated associations between periods of shortage and outcomes, but no proof of causality. Even if mortality was increased during the shortage, the mechanism might be multifactorial and not solely related to the absence of norepinephrine.
Vail E, Gershengorn HB, Hua M, Walkey AJ, Rubenfeld G, Wunsch H . Association between US norepinephrine shortage and mortality among patients with septic shock. JAMA. 2017;317(14):1433-1442.
Donohue JM, Angus DC . National shortages of generic sterile injectable drugs: norepinephrine as a case study of potential harm. JAMA. 2017;317(14):1415-1417.