Inotropes and vasopressors are frequently used in the treatment of circulatory failure, and the dose of vasopressor and inotrope medication required to establish adequate hemodynamic status is a predictor of mortality in intensive care unit (ICU) patients (1–4). The cardiovascular Sequential Organ Failure Assessment (SOFA) subscore on the first ICU day has good discrimination for hospital mortality in cardiac ICU (CICU) patients (5). Although a Cochrane meta-analysis failed to identify significant differences in mortality between different vasopressors in unselected patients with shock, there were potential differences in adverse event profiles (6). Vasoactive drugs can produce serious adverse effects such as arrhythmias and tissue ischemia, and patients with underlying cardiac disease might be more susceptible to these effects (2, 7). Patients with cardiogenic shock receiving norepinephrine may be at decreased risk of arrhythmias and mortality compared with patients receiving dopamine or epinephrine (8–11).
Despite recent trends demonstrating an increased similarity with the medical ICU population, the significant burden of cardiovascular comorbidities in CICU patients may influence the safety profile of vasopressors and inotropes for treatment of circulatory failure and shock in this population (12–15). Therefore, this study was designed to determine the trends and outcomes associated with the use of vasopressors and inotropes, specifically norepinephrine, in a contemporary CICU population.
MATERIALS AND METHODS
Study design, setting, and participants
We performed a historical cohort analysis using a previously constructed institutional database of patients admitted to the CICU at the Mayo Clinic Hospital, St. Mary's Campus (5). This CICU is a closed 16-bed ICU which cares for critically ill patients with primary cardiac diseases or significant cardiac comorbidities, but does not admit postoperative cardiac surgery patients or care for patients receiving extracorporeal membrane oxygenator or durable left ventricular assist device support. Unique adult patients at least 18 years old admitted to the CICU from January 1, 2007 to December 31, 2015, were included; only data from the first CICU admission were used for patients with subsequent readmissions (16). Patients less than 18 years of age, patients admitted before January 1, 2007, or still undergoing hospitalization on December 31, 2015, and patients who did not provide Minnesota Research Authorization under Minnesota state law statute 144.295 were excluded.
Data sources and definitions
We collected data on demographic and laboratory variables, invasive and noninvasive ventilation use, inpatient cardiac procedures, use of continuous renal replacement therapy (CRRT) and CICU and hospital length of stay (LOS), as previously described (5). Acute Physiology and Chronic Health Evaluation (APACHE)-III scores were calculated automatically using data from the first 24 h of CICU admission (17). SOFA scores and individual SOFA organ failure subscores were automatically generated daily, as were the maximum daily SOFA score for the first week in the CICU and peak cardiovascular SOFA subscore during the CICU stay (Supplemental Table 1a, https://links.lww.com/SHK/A897) (5, 18). The Charlson Comorbidity Index (CCI) was calculated from electronic health record data and used to identify prior medical conditions (19). Discharge ICD-9 diagnostic codes from the index hospitalization were reviewed for relevant diagnoses.
Vasoactive drugs were defined as intravenous vasopressors and inotropes administered via continuous infusion, including dobutamine (DOB), dopamine (DOPA), epinephrine (EPI), norepinephrine (NE), phenylephrine (PHEN), vasopressin (VASO), and milrinone (MIL). Low-dose (LD) DOPA was defined as a peak dose at most 5 mcg/kg/min, and high-dose (HD) DOPA was defined as a peak dose more than 5 mcg/kg/min. The peak Vasoactive-Inotropic Score (VIS) was calculated using peak vasoactive drug doses during the CICU stay (in mcg/kg/min): VIS = DOB + DOPA + (10 × PHEN + MIL) + (100 × [EPI + NE]) + (10,000 × units/kg/min VASO); one VIS unit is considered equivalent to 1 mcg/kg/min of DOB or DOPA or 0.01 mcg/kg/min of EPI or NE (20). Peak doses of NE, EPI, DOPA, PHEN, and VASO were used to calculate the peak norepinephrine-equivalent (NEE) dose (Supplemental Table 1b, https://links.lww.com/SHK/A897) and peak Cumulative Vasopressor Index (CVI) (Supplemental Table 1c, https://links.lww.com/SHK/A897) (3, 21).
The primary endpoint was all-cause hospital mortality. Secondary endpoints included all-cause CICU and 30-day mortality, as well as the incidence of severe acute kidney injury (AKI) during hospitalization, defined as doubling of serum creatinine, increase in serum creatinine to at least 4.0 mg/dL or new dialysis initiation during hospitalization among patients not previously on dialysis. Mortality data were extracted from Mayo Clinic electronic databases, state of Minnesota electronic death certificates and the Rochester Epidemiology Project database, as previously described (22). Patients receiving one or more vasoactive drugs were compared with patients not receiving vasoactive drugs. Categorical variables are reported as number (%), and the Pearson χ2 test was used to compare groups. Continuous variables are reported as mean (± standard deviation, SD); Student t test was used to compare two groups and analysis of variance was used to compare more than 2 groups. The Cochran–Armitage trend test was used to assess temporal changes in categorical variables. Univariate analysis was performed using continuous variables as predictors of hospital mortality to generate area under the receiver-operator characteristic curve (AUROC) values for each relevant variable; the optimal cutoff for predicting mortality was defined by the highest value of Youden's J index (sensitivity + specificity – 1). Stepwise backward multivariable regression was performed (P < 0.25 to enter and P < 0.1 to leave the model) to select predictors of hospital mortality for multivariable logistic regression, including demographics, comorbidities, mean arterial pressure (MAP) during CICU stay, CICU therapies, discharge diagnoses, and mean week 1 SOFA score as candidate variables. P values <0.05 were considered statistically significant. Statistical analyses were performed using JMP version 13.0 Pro (SAS Institute, Cary, NC).
We screened 12,904 adult admissions to the CICU during the study period and excluded 2,900 patients (1,877 readmissions, 755 patients with no Minnesota Research Authorization, and 268 patients admitted outside of the study period), yielding 10,004 unique patients as previously reported (Supplemental Figure 1, https://links.lww.com/SHK/A897) (5). The mean age of the population was 67 ± 15 years, and 3,746 (37%) patients were female (Table 1). A total of 2,468 (24.7%) patients received vasoactive drugs during the CICU stay, including 378 (15.3%) who received only inotropes (i.e., DOB or MIL), 1,540 (62.4%%) who received only vasopressors (i.e., DOPA, EPI, NE, PHEN, or VASO), and 550 (22.3%) who received both inotropes and vasopressors. Among patients receiving vasoactive drugs at any time during the CICU course, 2,211 (89.6%) received these drugs within the first 24 h; only 257 (3.3%) patients who did not receive vasoactive drugs within the first 24 h subsequently received these drugs. Patients receiving one or more vasoactive drugs had higher illness severity scores and more frequent AKI than those receiving none (Table 1). These differences were most marked among patients receiving more than one vasoactive drug (Table 1).
There was an increase in the overall use of vasoactive drugs during the study period (Fig. 1A), from 22.6% of patients in 2007 to 28.4% in 2015 (P < 0.001). As shown in Supplemental Table 2, https://links.lww.com/SHK/A897, DOPA was the most commonly used vasoactive drug (being used in 49% of patients receiving vasoactive drugs), followed by NE (being used in 29% of patients receiving vasoactive drugs). Among patients receiving vasoactive drugs, the use of MIL increased (from 18.0% in 2007 to 25.4% in 2015, P < 0.001; Fig. 1A) and the use of NE increased (from 19.7% in 2007 to 56.2% in 2015, P < 0.001; Fig. 1B) during the study period. The overall use of DOPA decreased (from 65.7% in 2007 to 27.6% in 2015, P < 0.001; Fig. 1B) over time, mainly driven by a decrease in the use of HD DOPA (from 48.0% in 2007 to 8.1% in 2015, P < 0.001; Fig. 1B) without a change in the use of LD DOPA (Fig. 1A). The use of other vasoactive drugs did not change significantly over time (Fig. 1A and Fig. 1B). Among patients receiving DOPA, the average maximum DOPA dose decreased over time during the study period (from 10.3 mcg/kg/min in 2007 to 5.3 mcg/kg/min in 2015). Baseline characteristics, illness severity scores, and vasoactive drug usage and dosing during the time periods of 2007 to 2009, 2010 to 2012, and 2013 to 2015 are shown in Supplemental Table 3, https://links.lww.com/SHK/A897; there were no significant differences in overall severity of illness (as measured by the APACHE-III score or maximum week 1 SOFA score) or vasoactive drug dosage (based on VIS or NEE) between time periods.
All-cause mortality in CICU, hospital, and at 30-days was 5.7%, 9.1%, and 11.4%, respectively. Patients receiving vasoactive drugs were at increased risk of hospital mortality (24.3% vs. 4.1%, unadjusted odds ratio [OR] 7.54, 95% confidence interval [CI], 6.51–8.73, P < 0.001). Hospital mortality was higher (all P < 0.001) in patients receiving one vasoactive drug (unadjusted OR vs. no vasoactive drugs 3.82, 95% CI, 3.14–4.64) or more than one vasoactive drug (unadjusted OR vs. no vasoactive drugs 12.93, 95% CI, 10.97–15.25; unadjusted OR vs. one vasoactive drug 3.39, 95% CI, 2.78–4.13). There was a stepwise increase in short-term mortality as a function of the number of vasoactive drugs used (Fig. 2).
Compared with patients not receiving any vasoactive drugs, patients receiving each of the vasoactive drugs had higher unadjusted hospital mortality (all P < 0.05). However, patients receiving inotropes only (DOB or MIL) without vasopressors have similar hospital mortality (5.8% vs. 4.1%, P = 0.10) compared with patients not receiving any vasoactive drugs. Peak number of vasoactive drugs, cardiovascular SOFA subscore, CVI, NEE, and VIS were all significant (all P < 0.001) univariate predictors of hospital mortality (Supplemental Table 4, https://links.lww.com/SHK/A897). Peak cardiovascular SOFA had the highest AUROC value for hospital mortality across the entire population. The relationship between peak cardiovascular SOFA and short-term mortality is shown in Supplemental Figure 2, https://links.lww.com/SHK/A897; patients with a peak cardiovascular score of 4 had hospital mortality exceeding 50%. As demonstrated in Table 2, multivariable analysis identified the peak VIS (OR 1.013 per 1 unit increase in VIS, 95% CI, 1.009–1.017, P < 0.001) as a significant predictor of hospital mortality in the entire population after adjustment for multiple markers of illness severity, CICU therapies, and discharge diagnoses.
Among patients receiving vasoactive drugs, those receiving DOB, HD DOPA, EPI, NE, VASO, or more than one vasoactive drug had higher unadjusted short-term mortality and those receiving LD DOPA, or MIL had lower unadjusted short-term mortality (all P < 0.001). Unadjusted hospital mortality for patients receiving each vasoactive drug is shown in Figure 3, as a function of whether the patient received one or more than one vasoactive drug. Characteristics of patients receiving each vasoactive drug are shown in Table 3; although these groups are not mutually exclusive, significant differences in vasopressor dosage and illness severity are noted. Patients receiving NE tended to have greater illness severity, higher overall vasoactive drug requirements, and more frequent diagnoses of sepsis, cardiac arrest, and cardiogenic shock than patients receiving other vasoactive drugs (Table 3); on the contrary, patients receiving MIL tended to have lower illness severity and overall vasoactive drug requirements. In the subgroup of patients receiving vasoactive drugs, peak cardiovascular SOFA subscore, CVI, NEE, and VIS were all significant (all P < 0.001) univariate predictors of hospital mortality (Supplemental Table 4, https://links.lww.com/SHK/A897); peak VIS and peak NEE had the highest AUROC values for hospital mortality among patients receiving vasoactive drugs. Among patients receiving vasoactive drugs, multivariable analysis again identified use of NE (OR 0.66, 95% CI, 0.49–0.90, P = 0.008), peak VIS (OR 1.018 per 1 unit increase in VIS, 95% CI, 1.013–1.022, P < 0.001), and use of IABP (OR 0.61, 95% CI, 0.43–0.67, P = 0.006) as predictors of hospital mortality; use of DOB (OR 1.36, 95% CI, 0.98–1.89, P = 0.068) was not a significant predictor of hospital mortality (Table 2). We performed a post hoc analysis including a statistical interaction term between NE usage and peak VIS in the logistic regression model among patients receiving vasoactive drugs, finding that this interaction was nominally significant (P = 0.049); use of NE remained associated with hospital mortality even after accounting for this interaction (OR 0.66, 95% CI, 0.48–0.89, P = 0.007).
Peak VIS was selected over peak CVI or peak NEE by all multivariable models as a predictor of hospital mortality. Figure 4 demonstrates the relationship between peak VIS and short-term mortality. Among patients receiving vasoactive drugs (Supplemental Table 2, https://links.lww.com/SHK/A897), the median peak VIS was 9 (interquartile range 4, 21.66). There was a stepwise increase in unadjusted short-term mortality as a function of VIS quartile (Supplemental Figure 3, https://links.lww.com/SHK/A897, P < 0.01 pairwise between groups); patients in the top VIS quartile had hospital mortality exceeding 50%. Patients in the top VIS quartile accounted for more than half of hospital deaths among patients receiving vasoactive drugs and 36% of all hospital deaths overall.
This is the largest study evaluating patterns and outcomes of vasopressor and inotrope drug usage in a contemporary CICU population. Overall use of vasoactive drugs increased during the study period, with an increase in NE and MIL usage and a decrease in HD DOPA usage over time. As expected, patients receiving vasoactive drugs were at increased risk of short-term mortality and AKI, particularly patients requiring multiple vasoactive drugs or higher vasoactive drug doses. A dose-dependent effect was demonstrated between vasoactive drug dose and hospital mortality using the peak VIS. Patients in the highest peak VIS quartile had hospital mortality exceeding 50%, as did patients with a peak cardiovascular SOFA subscore equal to 4. After adjusting for overall vasopressor and inotrope requirements and illness severity using multivariable logistic regression with a statistical interaction term, patients receiving norepinephrine for vasopressor support had a lower risk of hospital mortality, although the findings of this observational study should be considered exploratory.
Few prior studies have examined vasopressor and inotrope usage in CICU patients independent of admission diagnosis. A recent study from our institution by Thongprayoon et al. compared the use of different vasopressors in ICU patients (23). Similar to the present study, use of NE increased over time from 2007 to 2013 in CICU patients. Unlike the present study, a significant change in the overall use of DOPA was not observed in CICU patients in the study by Thongprayoon et al (23). These differences may be explained by the different methodologies used: Thongprayoon et al. used a time-sensitive vasopressor utilization index to quantify per-patient duration of vasopressor usage rather than the number of patients on each drug during the ICU stay, as in the present study (23). Time-dependent measures of vasopressor usage may have been influenced by prolonged use of LD DOPA for patients awaiting heart transplantation at our institution, many of whom were managed in the CICU during the study time frame. The frequent use of DOPA in this CICU population contrasts substantially with the patterns of vasopressor usage recently reported for patients with sepsis (24). The present study also identified an increase in the use of MIL over time, which was not examined in the prior study; reasons for this change may relate to an increase in the prevalence of heart failure over time.
Vasopressor requirements are included in multiple ICU severity of illness scores, such as the SOFA score (1, 2). In our prior study, the cardiovascular SOFA subscore on the first CICU day was a good predictor of short-term mortality in CICU patients, as was the peak cardiovascular SOFA subscore in this study (5). Other methods of quantifying vasopressor requirements (i.e., CVI, VIS, NEE) have not been previously compared for the purpose of predicting mortality in CICU populations. Peak VIS was selected by our multivariable logistic regression models over peak CVI or peak NEE as a predictor of hospital mortality, potentially because VIS includes the inodilators DOB and MIL which are relevant in CICU patients (3, 20, 21). Peak VIS provided independent prognostic value when added to ICU severity of illness scores, emphasizing the importance of vasopressor load as a predictor of outcomes in CICU patients. Our results expand on the recent study by Na et al., which demonstrated an association between VIS and adjusted mortality among patients with cardiogenic shock (4). Unlike our study, Na et al. focused on patients with cardiogenic shock rather than unselected CICU patients, explaining why higher VIS cutoffs were associated with mortality in that study compared with our study (4). Although high doses of vasopressor drugs are a marker of greater illness severity, the highest doses of vasoactive drugs were likely used as salvage therapy for patients who were dying of shock. These patients had very high mortality, emphasizing the poor outcomes seen in patients with severe shock requiring high vasoactive drug doses (3). Notably, the relatively modest vasopressor doses that define a cardiovascular SOFA subscore equal to 4 identified a population with more than 50% hospital mortality.
Cardiogenic shock is an important CICU diagnosis typically requiring vasopressor support, and vasopressor requirements are an important predictor of mortality in cardiogenic shock (1, 25–27). Only one in four patients requiring vasoactive drugs in this study carried a diagnosis of cardiogenic shock, reflecting the increasing prevalence of sepsis and undifferentiated or mixed shock states in contemporary CICU populations (12–14, 27). Notably, we identified important limitations in the use of ICD-9 codes in this population, as only 42% of patients receiving vasopressors carried a discharge diagnosis of sepsis or shock. As seen in prior studies of CICU patients, sepsis and cardiogenic shock were associated with higher hospital mortality on multivariable analysis in our CICU population as a whole, although this was not seen among patients receiving vasoactive drugs (13). This emphasizes the potential importance of sepsis as a driver of shock and mortality in CICU patients.
We observed substantial variation in hospital mortality as a function of the number and type of vasoactive drugs used, although this generally mirrored standard measures of illness severity and overall vasoactive drug dosage. Patients receiving inotropes alone (particularly milrinone) were at low risk of death, reflecting the use of these drugs for heart failure rather than shock. Despite higher crude hospital mortality reflecting greater vasoactive drug requirements and illness severity among patients receiving NE, the use of NE was associated with lower mortality than other vasoactive drugs in our population after adjustment for vasoactive drug requirements and illness severity. Randomized clinical trials and meta-analyses have demonstrated that NE is associated with fewer arrhythmias and lower mortality than DOPA in both septic shock and cardiogenic shock (8, 10, 28). Our observed trends in DOPA and NE usage over time likely reflect a switch from HD DOPA to NE for vasopressor support based on these clinical trials, with the timing of this practice change coinciding with the publication of the SOAP-II trial and subsequent meta-analyses (8, 28). Recent observational studies have demonstrated that use of EPI in cardiogenic shock was associated with higher mortality and more extensive cardiac and renal injury (9, 11). A recent randomized trial likewise demonstrated a higher rate of treatment failure among patients receiving EPI compared with NE for cardiogenic shock (29). Therefore, the lower observed mortality with NE in our study may reflect relatively higher mortality in patients receiving EPI or DOPA, which have a propensity to cause harmful excessive beta-adrenergic stimulation and arrhythmias; notably, we did not observe a direct association between use of either DOPA or EPI and mortality in our population (2, 7, 8, 10, 28, 29). Importantly, the point estimate for mortality reduction by NE in this study is likely overestimated when compared with the more modest 10% to 12% relative risk reductions seen in randomized studies (8, 28). The presence of a statistical interaction between use of NE and peak VIS for prediction of hospital mortality may imply that the VIS might overestimate the effect of NE on mortality. Our findings, along with prior studies, support the use of NE as the first-line vasopressor in CICU patients, as in other critically ill patients with shock (2, 3, 6, 8, 10, 11, 28). We suspect that the trend toward higher mortality observed with the use of DOB among patients receiving vasoactive drugs may be due to residual confounding, although use of DOB has been associated with higher mortality among patients with heart failure (30).
This exploratory study has a number of important limitations inherent to retrospective cohort studies, particularly the potential for residual confounding by unmeasured variables and an inability to determine the factors that may have influenced vasoactive drug utilization. This single-center referral population likely differs from CICU populations at other centers, and local practice patterns regarding vasopressor and inotrope usage could have influenced the associations between different drugs and mortality leading to confounding by indication. However, the observed prevalence of vasoactive drug use and hospital mortality in this study mirrors another recent study of CICU patients (15). Peak vasoactive drug doses were used to calculate the CVI, VIS, and NEE, rather than maximum simultaneous doses as in prior studies; the lack of a time stamp for the peak vasoactive drug dose prevents statistical adjustment for physiological variables from that time point and prevents analysis based on the initial vasoactive drug used (3, 20, 21). The fact that many patients received more than one vasoactive drug during their CICU stay may limit the ability to determine the individual effects of each drug on clinical outcomes. Despite the importance of fluid resuscitation in the management of patients with sepsis and septic shock, data on fluid administration were not available. The observed association between NE and mortality should be interpreted with caution and could reflect inadequate statistical adjustment by the VIS, which was derived for use in cardiac surgery patients who likely differ substantially from our CICU population (20). The observed statistical interaction between use of NE and peak VIS supports this notion. Notably, the vasopressor equivalency ratios used in the CVI, VIS, and NEE differ, reflecting disagreement on equivalent vasopressor doses in the published literature (2, 3, 20, 21). Changes in NE utilization over time could have influenced the reported effects of NE on mortality, although no temporal trends in hospital mortality were observed and year of hospital admission was not a significant predictor of mortality. Data regarding admission diagnoses were not available and could have influenced vasoactive drug usage and observed outcomes. As a surrogate for admission diagnoses, we used discharge ICD-9 diagnosis codes which likely reflect major problems and comorbidities during the hospital stay. Although we adjusted for relevant ICD-9 diagnoses in our multivariable models, this did not allow for direct evaluation of drug–disease interactions that could have influenced the effects of individual vasoactive drugs on outcomes. Treatment-emergent adverse effects such as arrhythmias or ischemic complications that might be attributed to vasoactive drugs were not routinely available, preventing determination of why patients receiving NE had lower adjusted mortality.
Vasopressor and inotrope usage is common in CICU patients, and the need for vasoactive drugs is associated with high short-term mortality. Higher vasoactive drug doses, as quantified by the VIS, predict hospital death when added to established ICU severity of illness scores. In this CICU population, use of NE was independently associated with lower mortality independent of overall vasoactive drug requirements and illness severity, validating the observed trend toward increased use of NE over time. Among patients receiving vasoactive drugs, NE was still associated with lower mortality after adjustment for illness severity and vasoactive drug requirements; use of EPI and DOPA did not appear to be associated with the risk of mortality in this population. Further studies will be needed to help guide the choice of vasopressor and inotrope medications in patients with acute cardiac disease, who may display different adverse event profiles from other critically ill patients. Until then, our results are hypothesis-generating but support the use of NE as the vasopressor of choice in CICU patients with shock.
The authors thank the dedicated physicians and nurses who care for our CICU patients every day.
1. Yadav H, Harrison AM, Hanson AC, Gajic O, Kor DJ, Cartin-Ceba R. Improving the accuracy of cardiovascular component of the sequential organ failure assessment score. Crit Care Med
43 (7):1449–1457, 2015.
2. Jentzer JC, Coons JC, Link CB, Schmidhofer M. Pharmacotherapy update on the use of vasopressors and inotropes in the intensive care unit. J Cardiovasc Pharmacol Ther
20 (3):249–260, 2015.
3. Jentzer JC, Vallabhajosyula S, Khanna AK, Chawla LS, Busse LW, Kashani KB. Management of refractory vasodilatory shock. Chest
154 (2):416–426, 2018.
4. Na SJ, Chung CR, Cho YH, Jeon K, Suh GY, Ahn JH, Carriere KC, Park TK, Lee GY, Lee JM, et al. Vasoactive inotropic score as a predictor of mortality in adult patients with cardiogenic shock: medical therapy versus ECMO. Rev Esp Cardiol (Engl Ed)
72 (1):40–47, 2019.
5. Jentzer JC, Bennett C, Wiley BM, Murphree DH, Keegan MT, Gajic O, Wright RS, Barsness GW. Predictive value of the sequential organ failure assessment score for mortality in a contemporary cardiac intensive care unit population. J Am Heart Assoc
7 (6):e008169, 2018.
6. Gamper G, Havel C, Arrich J, Losert H, Pace NL, Mullner M, Herkner H. Vasopressors for hypotensive shock. Cochrane Database Syst Rev
7. Schmittinger CA, Torgersen C, Luckner G, Schroder DC, Lorenz I, Dunser MW. Adverse cardiac events during catecholamine vasopressor therapy: a prospective observational study. Intensive Care Med
38 (6):950–958, 2012.
8. De Backer D, Biston P, Devriendt J, Madl C, Chochrad D, Aldecoa C, Brasseur A, Defrance P, Gottignies P, Vincent JL. Comparison of dopamine and norepinephrine in the treatment of shock. N Engl J Med
362 (9):779–789, 2010.
9. Tarvasmaki T, Lassus J, Varpula M, Sionis A, Sund R, Kober L, Spinar J, Parissis J, Banaszewski M, Silva Cardoso J, et al. Current real-life use of vasopressors and inotropes in cardiogenic shock—adrenaline use is associated with excess organ injury and mortality. Crit Care
20 (1):208, 2016.
10. Rui Q, Jiang Y, Chen M, Zhang N, Yang H, Zhou Y. Dopamine versus norepinephrine in the treatment of cardiogenic shock: a PRISMA-compliant meta-analysis. Medicine (Baltimore)
96 (43):e8402, 2017.
11. Leopold V, Gayat E, Pirracchio R, Spinar J, Parenica J, Tarvasmaki T, Lassus J, Harjola VP, Champion S, Zannad F, et al. Epinephrine and short-term survival in cardiogenic shock: an individual data meta-analysis of 2583 patients. Intensive Care Med
44 (6):847–856, 2018.
12. Katz JN, Minder M, Olenchock B, Price S, Goldfarb M, Washam JB, Barnett CF, Newby LK, van Diepen S. The genesis, maturation, and future of critical care cardiology. J Am Coll Cardiol
68 (1):67–79, 2016.
13. Holland EM, Moss TJ. Acute noncardiovascular illness in the cardiac intensive care unit. J Am Coll Cardiol
69 (16):1999–2007, 2017.
14. Sinha SS, Sjoding MW, Sukul D, Prescott HC, Iwashyna TJ, Gurm HS, Cooke CR, Nallamothu BK. Changes in primary noncardiac diagnoses over time among elderly cardiac intensive care unit patients in the United States. Circ Cardiovasc Qual Outcomes
10 (8):e003616, 2017.
15. Goldfarb M, van Diepen S, Liszkowski M, Jentzer JC, Pedraza I, Cercek B. Noncardiovascular disease and critical care delivery in a contemporary cardiac and medical intensive care unit. J Intensive Care Med
34 (7):537–543, 2019.
16. Herasevich V, Pickering BW, Dong Y, Peters SG, Gajic O. Informatics infrastructure for syndrome surveillance, decision support, reporting, and modeling of critical illness. Mayo Clin Proc
85 (3):247–254, 2010.
17. Chandra S, Kashyap R, Trillo-Alvarez CA, Tsapenko M, Yilmaz M, Hanson AC, Pickering BW, Gajic O, Herasevich V. Mapping physicians’ admission diagnoses to structured concepts towards fully automatic calculation of acute physiology and chronic health evaluation score. BMJ Open
1 (2):e000216, 2011.
18. Aakre C, Franco PM, Ferreyra M, Kitson J, Li M, Herasevich V. Prospective validation of a near real-time EHR-integrated automated SOFA score calculator. Int J Med Inform
19. Singh B, Singh A, Ahmed A, Wilson GA, Pickering BW, Herasevich V, Gajic O, Li G. Derivation and validation of automated electronic search strategies to extract Charlson comorbidities from electronic medical records. Mayo Clin Proc
87 (9):817–824, 2012.
20. Nguyen HV, Havalad V, Aponte-Patel L, Murata AY, Wang DY, Rusanov A, Cheng B, Cabreriza SE, Spotnitz HM. Temporary biventricular pacing decreases the vasoactive-inotropic score after cardiac surgery: a substudy of a randomized clinical trial. J Thorac Cardiovasc Surg
146 (2):296–301, 2013.
21. Trzeciak S, McCoy JV, Dellinger RP, Arnold RC, Rizzuto M, Abate NL, Shapiro NI, Parillo JE, Hollenberg SM. Early increases in microcirculatory perfusion during protocol-directed resuscitation are associated with reduced multi-organ failure at 24 h in patients with sepsis. Intensive Care Med
34 (12):2210–2217, 2008.
22. Rocca WA, Yawn BP, St Sauver JL, Grossardt BR, Melton LJ 3rd. History of the Rochester Epidemiology Project: half a century of medical records linkage in a US population. Mayo Clin Proc
87 (12):1202–1213, 2012.
23. Thongprayoon C, Cheungpasitporn W, Harrison AM, Carrera P, Srivali N, Kittamongkolchai W, Erdogan A, Kashani KB. Temporal trends in the utilization of vasopressors in intensive care units: an epidemiologic study. BMC Pharmacol Toxicol
17 (1):19, 2016.
24. Scheeren TWL, Bakker J, De Backer D, Annane D, Asfar P, Boerma ED, Cecconi M, Dubin A, Dunser MW, Duranteau J, et al. Current use of vasopressors in septic shock. Ann Intensive Care
9 (1):20, 2019.
25. Katz JN, Stebbins AL, Alexander JH, Reynolds HR, Pieper KS, Ruzyllo W, Werdan K, Geppert A, Dzavik V, Van de Werf F, et al. Predictors of 30-day mortality in patients with refractory cardiogenic shock following acute myocardial infarction despite a patent infarct artery. Am Heart J
158 (4):680–687, 2009.
26. Geppert A, Dorninger A, Delle-Karth G, Zorn G, Heinz G, Huber K. Plasma concentrations of interleukin-6, organ failure, vasopressor support, and successful coronary revascularization in predicting 30-day mortality of patients with cardiogenic shock complicating acute myocardial infarction. Crit Care Med
34 (8):2035–2042, 2006.
27. van Diepen S, Katz JN, Albert NM, Henry TD, Jacobs AK, Kapur NK, Kilic A, Menon V, Ohman ME, Sweitzer NK, et al. Contemporary management of cardiogenic shock: a scientific statement from the American Heart Association. Circulation
136 (16):e232–e268, 2017.
28. De Backer D, Aldecoa C, Njimi H, Vincent JL. Dopamine versus norepinephrine in the treatment of septic shock: a meta-analysis∗. Crit Care Med
40 (3):725–730, 2012.
29. Levy B, Clere-Jehl R, Legras A, Morichau-Beauchant T, Leone M, Frederique G, Quenot JP, Kimmoun A, Cariou A, Lassus J, et al. Epinephrine versus norepinephrine for cardiogenic shock after acute myocardial infarction. J Am Coll Cardiol
72 (2):173–182, 2018.
30. Tacon CL, McCaffrey J, Delaney A. Dobutamine for patients with severe heart failure: a systematic review and meta-analysis of randomised controlled trials. Intensive Care Med
38 (3):359–367, 2012.