Changes in Interprofessional Practice During the COVID-19 Pandemic and the Effect on Patient Outcomes: A Retrospective Observational Study.

BACKGROUND: Changes in healthcare delivery were required during the first year of the COVID-19 pandemic. OBJECTIVE: The purpose of this study was to determine the impact of the approach to care of the COVID-19 patient on nursing sensitive indicators and nutrition therapy and the utilization of rehabilitation services during the first year of the pandemic in the acute care setting. METHOD: A retrospective study of 894 patients admitted with a COVID-19 diagnosis was conducted between March 2020 and February 2021 in 3-month cohorts. All charts were reviewed for general demographics and hospital data, nursing quality indicators, and nutritional and rehabilitation services for the first 30 days of admission. RESULTS: Differences in patient characteristics were noted among the cohorts. Variations were observed between time points in hospital-acquired pressure injury occurrence, with mechanical ventilation and proning being independent predictors of hospital-acquired pressure injuries. There were differences noted in the percentage of patients with a central line-associated bloodstream infection among the time points (P < .001), but there were no differences noted in catheter-associated urinary tract infections (P = .20). Overall, 15.5% had a malnutrition diagnosis, with most patients receiving 50% of prescribed calorie and protein needs. Rehabilitation services increased over time with these services being initiated earlier in the later cohorts (P < .001). DISCUSSION: The results of this study demonstrated the impact of the pandemic on outcomes in the areas of nursing, nutrition, and rehabilitation, which varied across quarterly cohorts as we learned and developed new practices and adapted to a novel pandemic.

Original Research: Patient Perception of Fall Risk in the Acute Care Setting.

PURPOSE: Hospitals are implementing a variety of fall prevention programs to reduce the fall rates of hospitalized patients. But if patients don't perceive themselves to be at risk for falling and don't adhere to fall prevention strategies, such programs are likely to be less effective. The purpose of this study was to describe the perceptions of fall risk among hospitalized patients across four acute care specialty services. METHODS: One hundred patients who had been admitted to the study hospital and who had a Morse Fall Scale score over 45 were asked to complete the Patient Perception Questionnaire, a tool designed to explore a patient's confidence regarding their fall risk, fear of falling, and intention to engage in fall prevention activities. Morse Fall Scale scores were collected via retrospective chart review. Data were analyzed using descriptive statistics, Pearson correlation coefficients, and independent sample t tests. RESULTS: Participants' mean age was 65 years; 52% were male, 48% female. Although all 100 participants were deemed at risk for falls per their Morse Fall Scale scores, only 55% considered themselves to be at such risk. As patients' confidence in their ability to perform mobility tasks increased, their intention to ask for help and fear of falling significantly decreased. Patients who had been admitted as the result of a fall demonstrated significantly lower confidence scores and higher fear scores. CONCLUSIONS: Patients who score high on fall risk assessments often don't perceive themselves to be at high risk for falling, and thus might not engage in fall prevention activities. Developing a fall risk assessment method that incorporates both a patient's physiological condition and their perception of their fall risk could help reduce fall rates in the acute care setting.

One Year Post-Discharge Outcomes After Implementation of an ICU Early Mobility Protocol.

BACKGROUND: Early mobility benefits include improved strength, decreased length of stay (LOS), and delirium. The impact of an early mobility protocol on return to activities of daily living (ADL) is less studied. OBJECTIVE: The aim of this study was to examine 1-year outcomes including ADL performance after the institution of an ICU early mobility protocol. METHODS: One year after the initiation of an early mobility protocol in 7 intensive care units (ICUs) at an academic medical center, patients with an ICU stay of 7 days or more were enrolled in a 1-year follow-up phone call study. Baseline demographic data included the following: average ICU mobility and highest ICU mobility level achieved (4 levels), highest ICU mobility score (10 levels) at ICU admission, ICU discharge (DC), hospital DC, LOS, and delirium positive days. At 4 time points after DC (1, 3, 6, 12 months), patients were contacted regarding current residence, employment, readmissions, and current level of ADL from the Katz ADL (scored 0-6) and Lawton instrumental ADL scales (scored 0-8). RESULTS: A convenience sample of 106 patients was enrolled with a mean age of 58 ± 15.4 years, ICU LOS of 18 ± 11.5 days, and hospital LOS of 37.5 ± 31 days; 58 (55%) were male; 4 expired before DC. Mobility results included mean mobility level of 1.6 ± 0.8, mean highest mobility level 3.3 ± 0.9; ICU mobility score was 5.9 ± 2.4 at time of ICU DC and 7.3 ± 2.5 at hospital DC. Katz ADL scores improved from 4.8 at 1 month to 5.6 at 12 months (P = .002), and Lawton IADL scores improved from 4.2 to 6.6 (P < .001). Mobility scores were predictors of 1 month Katz (P = .004) and Lawton (P < .001) scores. None of the mobility levels or scores were predictive for readmissions. Most patients were not working before admission, and not all returned to work. Days positive for delirium were predictive of 1 month Katz and Lawton (P = .014, .002) scores. Impact of delirium was gone by 1 year. DISCUSSION: In this critically ill patient population followed for 1 year, ICU mobility positively impacted return to ADLs and improved ADLs over time but not readmissions. Delirium positive days decreased ADL scores, but the effect diminished over time.

Implementation of an Interdisciplinary AACN Early Mobility Protocol.

BACKGROUND: Increasing mobility in the intensive care unit is an important part of the ABCDEF bundle. Objective To examine the impact of an interdisciplinary mobility protocol in 7 specialty intensive care units that previously implemented other bundle components. METHODS: A staggered quality improvement project using the American Association of Critical-Care Nurses mobility protocol was conducted. In phase 1, data were collected on patients with intensive care unit stays of 24 hours or more for 2 months before and 2 months after protocol implementation. In phase 2, data were collected on a random sample of 20% of patients with an intensive care unit stay of 3 days or more for 2 months before and 12 months after protocol implementation. RESULTS: The study population consisted of 1266 patients before and 1420 patients after implementation in phase 1 and 258 patients before and 1681 patients after implementation in phase 2. In phase 1, the mean (SD) mobility level increased in all intensive care units, from 1.45 (1.03) before to 1.64 (1.03) after implementation (P < .001). Mean (SD) ICU Mobility Scale scores increased on initial evaluation from 4.4 (2.8) to 5.0 (2.8) (P = .01) and at intensive care unit discharge from 6.4 (2.5) to 6.8 (2.3) (P = .04). Complications occurred in 0.2% of patients mobilized. In phase 2, 84% of patients had out-of-bed activity after implementation. The time to achieve mobility levels 2 to 4 decreased (P = .05). Intensive care unit length of stay decreased significantly in both phases. CONCLUSIONS: Implementing the American Association of Critical-Care early mobility protocol in intensive care units with ABCDEF components in place can increase mobility levels, decrease length of stay, and decrease delirium with minimal complications.

The Use of Physical Therapy ICU Assessments to Predict Discharge Home.

OBJECTIVES: To establish cutoff values for making recommendations for discharge to the home setting using standardized physical therapy assessments. DESIGN: Retrospective study. SETTING: Five ICUs at a large academic medical center. PATIENTS: 1,203 ICU patients. INTERVENTION: None. MEASUREMENTS AND MAIN RESULTS: The Functional Status Score for the ICU and the ICU Mobility Scale were collected during the initial physical therapy assessment, at ICU discharge, and prior to hospital discharge. The Activity Measure for Post-Acute Care-Inpatient Mobility Short Form "6 clicks" was only collected during the initial physical therapy assessment. Receiver Operating Characteristic curves were used to determine a potential cutoff value for discharge home. The Receiver Operating Characteristic was adjusted for ICU and hospital length of stay along with mobility status prior to hospital admission. Cutoff values were then determined by using Youden's Index. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were calculated based on these cut off values. The Functional Status Score for the ICU at ICU discharge was the best predictor of a discharge to the home setting in patients who had an ICU admission. The area under the curve for the Functional Status Score for the ICU at ICU discharge was 0.80. A Functional Status Score for the ICU score at ICU discharge of 19 or higher predicted discharge to home with a sensitivity of 82.9% and specificity of 73.6% CONCLUSIONS:: The Functional Status Score for the ICU at ICU discharge provided the best accuracy for making a timely recommendation for discharge home in patients who had an ICU admission.

Implementation of Fall Preventions Over the Past 15 Years: Impact on Inpatient Injury and Insights for the Future.

BACKGROUND: Hospital fall rates have changed minimally with preventive measures; however, the effect on injury rate is unclear. PURPOSE: The purpose was to determine whether fall-related injuries have changed over time. METHODS: A retrospective comparison was done of 1134 adult inpatient falls in 2017 to 1235 falls in 2001-2002 for injury and fall circumstances. Separate comparisons were made of patient characteristics by service line for 2017. RESULTS: Severe fall injuries declined from 6% to 2.4%. Elimination issues remained the most common circumstance (38.9% and 42%). In 2017, malnutrition (31.6%), low function (61.4%), fall history (26.3%), and use of high-risk medications (83.2%) were common in patients who fell. Predictors of falls with injury by patient population were as follows: surgery-male gender (P = .01), low function (P = .006), elimination issues (P = .04); oncology-low function (P = .04); and neurology-low function (P = .02). CONCLUSIONS: Severe fall-related injuries have decreased in the past 15 years. The most common circumstance for falls remains elimination issues.

Evaluation of Sensor Technology to Detect Fall Risk and Prevent Falls in Acute Care.

BACKGROUND: Sensor technology that dynamically identifies hospitalized patients' fall risk and detects and alerts nurses of high-risk patients' early exits out of bed has potential for reducing fall rates and preventing patient harm. During Phase 1 (August 2014-January 2015) of a previously reported performance improvement project, an innovative depth sensor was evaluated on two inpatient medical units to study fall characteristics. In Phase 2 (April 2015-January 2016), a combined depth and bed sensor system designed to assign patient fall probability, detect patient bed exits, and subsequently prevent falls was evaluated. METHODS: Fall detection depth sensors remained in place on two medicine units; bed sensors used to detect patient bed exits were added on only one of the medicine units. Fall rates and fall with injury rates were evaluated on both units. RESULTS: During Phase 2, the designated evaluation unit had 14 falls, for a fall rate of 2.22 per 1,000 patient-days-a 54.1% reduction compared with the Phase 1 fall rate. The difference in rates from Phase 1 to Phase 2 was statistically significant (z = 2.20; p = 0.0297). The comparison medicine unit had 30 falls-a fall rate of 4.69 per 1,000 patient-days, representing a 57.9% increase as compared with Phase 1. CONCLUSION: A fall detection sensor system affords a level of surveillance that standard fall alert systems do not have. Fall prevention remains a complex issue, but sensor technology is a viable fall prevention option.

Anatomy of Inpatient Falls: Examining Fall Events Captured by Depth-Sensor Technology.

BACKGROUND: Sensor technology offers a new way to identify patient movement, detect falls, and automatically alert health care staff when falls occur. The information gained from analyzing actual fall events can be beneficial in developing individualized fall prevention strategies, informing nursing staff about the nature of falls, and identifying opportunities to make the patient care environment safer. METHODS: A six-month performance improvement pilot was conducted at Barnes-Jewish Hospital (St. Louis) to assess the ability of a depth-sensor system to capture inpatient fall events within patient hospital rooms. Depth sensors were installed on two inpatient medicine units with a history of high fall rates. The depth sensors captured actual fall events on video. Video clips were reviewed and analyzed to identify the characteristics of patient falls, staff response times, and environmental conditions contributing to falls. RESULTS: A total of 16 falls involving 13 patients were recorded by depth sensors. Six of the 13 patients who fell were classified as high risk on the basis of the hospital's fall rating tool. Common contributing factors included difficulty rising from their bed, weakened lower extremities, and unsteady or slow gait. Eleven of the falls involved patients reaching for objects in their path in an effort to achieve stability. Nurses had less than two minutes from the time a patient began to exit a bed to the time a fall occurred. Patients expressed few complaints with depth sensors installed in rooms. CONCLUSION: Fall-detection sensor systems offer valuable data for analyzing the nature of patient falls, with the potential promise of prescribing specific fall interventions for patients and to identify staff development opportunities. Hospitals should understand these devices' benefits and limitations and how they affect nursing practice.

Subclinical Peroneal Neuropathy: A Common, Unrecognized, and Preventable Finding Associated With a Recent History of Falling in Hospitalized Patients.

PURPOSE: Identification of modifiable risk factors for falling is paramount in reducing the incidence and morbidity of falling. Peroneal neuropathy with an overt foot drop is a known risk factor for falling, but research into subclinical peroneal neuropathy (SCPN) resulting from compression at the fibular head is lacking. The purpose of our study was to determine the prevalence of SCPN in hospitalized patients and establish whether it is associated with a recent history of falling. METHODS: We conducted a cross-sectional study of 100 medical inpatients at a large academic tertiary care hospital in St Louis, Missouri. General medical inpatients deemed at moderate to high risk for falling were enrolled in the summer of 2013. Patients were examined for findings that suggest peroneal neuropathy, fall risk, and a history of falling. Multivariate logistic regression was used to correlate SCPN with fall risk and a history of falls in the past year. RESULTS: The mean patient age was 53 years (SD = 13 years), and 59 patients (59%) were female. Thirty-one patients had examination findings consistent with SCPN. After accounting for various confounding variables within a multivariate logistic regression model, patients with SCPN were 4.7 times (95% CI, 1.4-15.9) more likely to report having fallen 1 or more times in the past year. CONCLUSIONS: Subclinical peroneal neuropathy is common in medical inpatients and is associated with a recent history of falling. Preventing or identifying SCPN in hospitalized patients provides an opportunity to modify activity and therapy, potentially reducing risk.

The occurrence of antibodies to heparin-platelet factor 4 in cardiac and thoracic surgical patients receiving desirudin or heparin for postoperative venous thrombosis prophylaxis.

INTRODUCTION: This randomized, exploratory study compared the incidence of heparin-dependent antibodies associated with subcutaneous (SC) desirudin or heparin given for deep-vein thrombosis prophylaxis following cardiac and thoracic surgery. MATERIALS AND METHODS: Adult patients scheduled for elective cardiac or thoracic surgery received desirudin 15 mg SC twice daily or unfractionated heparin 5000 units SC thrice daily. Duration of thrombosis prophylaxis was determined by the treating physician. Primary outcome measure was the incidence of new antibody formation directed against platelet factor 4 (PF4)/heparin complex. Secondary outcomes included bleeding and thrombotic complications. Blood was tested for anti-PF4/heparin antibodies at baseline, after surgery prior to study drug administration, postdrug day (PDD) 2, PDD 7, and at 1 month. Doppler studies were done before discharge. RESULTS: Of 120 patients, 61 received desirudin, 59 received heparin. New PF4/heparin antibodies occurred in 10.2% and 13.6% of desirudin- and heparin-treated patients, respectively. Among desirudin patients with no heparin exposure, none (0/36) developed PF4/heparin antibodies versus 17.1% with heparin exposure. Incidence of deep venous thrombosis was 4.9% and 3.4% in the desirudin and heparin groups, respectively. Two heparin-group patients developed pulmonary embolism. Two patients per group had bleeding events; no patients required re-exploration for bleeding complications. Median chest tube output was similar with desirudin (900 mL) and heparin (692 mL) as was blood transfusion requirements of more than 2 units (5/61, desirudin; 2/59 heparin). CONCLUSIONS: The incidence of thrombotic events was low in both groups. There were no safety concerns, and desirudin was not associated with anti-PF4/heparin antibodies.

Prevalence of undiagnosed obstructive sleep apnea among adult surgical patients in an academic medical center.

BACKGROUND: Obstructive sleep apnea (OSA) affects approximately 20% of US adults, of whom about 90% are undiagnosed. While OSA may increase risk of perioperative complications, its prevalence among surgical patients is unknown. We tested the feasibility of screening surgical patients for OSA and determined the prevalence of undiagnosed OSA. METHODS: In a prospective, observational study adult surgical patients were screened for OSA in an academic hospital. Patients without an OSA diagnosis who screened high-risk were offered a home sleep study to determine if they had OSA. The results were compared with polysomnography (PSG) when available. Charts of high-risk patients were examined for postoperative complications. High-risk patients received targeted interventions as part of a hospital safety initiative. RESULTS: There were 2877 patients screened; 661 (23.7%) screened high-risk for OSA, of whom 534 (81%) did not have diagnosed OSA. The portable sleep study detected OSA in 170/207 (82%) high-risk patients without diagnosed OSA. Twenty-six PSGs confirmed OSA in 19 of these patients. Postoperatively there were no respiratory arrests, two unanticipated ICU admissions, and five documented respiratory complications. CONCLUSION: Undiagnosed OSA is prevalent in adult surgical patients. Implementing universal screening is feasible and can identify undiagnosed OSA in many surgical patients. Further investigation is needed into perioperative complications and their prevention for patients with undiagnosed OSA.

A randomized study in diabetic patients undergoing cardiac surgery comparing computer-guided glucose management with a standard sliding scale protocol.

OBJECTIVE: The aim of this study was to compare a standard insulin protocol with a computer-guided glucose management system to determine which method achieves tighter glucose control. DESIGN: A prospective, randomized trial. SETTING: A cardiothoracic intensive care unit (ICU) in a large academic medical center. PARTICIPANTS: Forty patients with diabetes mellitus who were scheduled for cardiac surgery. INTERVENTIONS: After induction of anesthesia and for the first 9 hours in the ICU, each subject received a standardized infusion of a 10% glucose solution at a rate of 1.0 mL/kg/h (ideal body weight). The subjects were then randomized to have their glucose controlled by either a paper-based insulin protocol or by a computer-guided glucose management system (CG). The desired range for blood glucose was set between 90 and 150 mg/dL. MEASUREMENTS AND MAIN RESULTS: There were no differences between groups in baseline characteristics. Patients in the CG group spent more time in the desired range during both the intraoperative phase (49% v 27%, p = 0.001) and the ICU phase (84% v 60%, p < 0.0001). There were no statistical differences between groups in the number of hypoglycemia episodes. CONCLUSIONS: The computer-guided glucose management system achieved tighter blood glucose control than a standard paper-based protocol in diabetic patients undergoing cardiac surgery. However, the low proportion of blood glucose recordings within the desired range in both groups during the intraoperative period reflects the challenges associated with achieving normoglycemia during cardiac surgery.

Mild hypercapnia after uncomplicated heart surgery is not associated with hemodynamic compromise.

OBJECTIVE: The aim of this study was to examine the effects of small changes in PaCO(2) on hemodynamic parameters after uncomplicated heart surgery with cardiopulmonary bypass. DESIGN: This was a prospective, randomized crossover study. SETTING: A large academic medical center. PARTICIPANTS: Twenty-four subjects who were scheduled for elective cardiac surgery were enrolled in this study. INTERVENTIONS: Each subject underwent the normal procedures that are associated with cardiac surgery. General anesthesia, including muscle relaxation, were continued in the immediate postoperative period. Measured tidal volumes and minute ventilation were kept constant for the duration of the study. Target PaCO(2) concentrations of 30, 40, and 50 mmHg were achieved by adding varying amounts of exogenous CO(2) gas to the inhaled oxygen. Various measurements were made at each target PaCO(2), including cardiac index, mixed venous oxygen saturation, blood pressure, heart rate, and pulmonary artery pressure. MEASUREMENTS AND MAIN RESULTS: Twenty-four patients were enrolled. Seven were withdrawn before commencement of the study. The cardiac index increased when the PaCO(2) was increased from 30 to 40 mmHg (p < 0.001) and remained unchanged between 40 and 50 mmHg. Mixed venous oxygen saturation increased (p < 0.001) with elevations in PaCO(2) up to 50 mmHg and decreased again when the PaCO(2) was returned to 30 mmHg. The blood pressure decreased (p < 0.001) with increasing PaCO(2). The pulmonary pressure increased (p < 0.001) with elevations in PaCO(2). No patient became hemodynamically unstable or had any arrhythmias. CONCLUSION: The findings of this study suggest that unless there is a specific contraindication to mild hypercapnia, such as pulmonary hypertension or hemodynamic instability, concerns about mild respiratory acidosis should not prevent weaning of sedation and mechanical ventilation after uncomplicated heart surgery.

Clostridium difficile in cardiac surgery: risk factors and impact on postoperative outcome.

BACKGROUND: Clostridium difficile-associated diarrhea (CDAD) is a potentially preventable and often troublesome gastrointestinal complication after cardiac surgery. METHODS: A retrospective study was performed of 8,405 cardiac surgery patients at two institutions between January 1997 and August 2004. Preoperative cardiac risk factors, perioperative factors including blood product transfusion, antibiotic utilization, and postoperative morbidity and mortality were recorded. Univariate and multivariate analyses were performed comparing C. difficile patients with a control group matched by date of surgery and institution. RESULTS: Sixty-six of the 8,405 patients identified with toxin-positive CDAD produced an overall incidence of 0.79% (0.70% at institution A and 1.09% at institution B), with a peak overall incidence of 5.45% in June 2003. Independent prognostic factors for CDAD by multivariate analysis included advancing age (odds ratio [OR] 1.028, 95% confidence interval [CI]: 1.001 to 1.056; p = 0.034), female sex (OR 2.026, 95% CI: 1.102 to 3.722; p = 0.022), blood product transfusion (OR 3.277, 95% CI: 1.292 to 8.311; p = 0.006), and increasing cumulative days of antibiotic administration (OR 1.046, 95% CI: 1.014 to 1.080; p = 0.004). There were no differences in the proportion of fluoroquinolones, cephalosporins, or penicillin derivatives administered between groups. The diagnosis of CDAD was associated with a greater median length of mechanical ventilation (25 hours versus 12 hours, p < 0.001), longer intensive care unit stay (5 days versus 2 days, p < 0.001), and extended hospital stay (21 days versus 7 days, p < 0.001), with no difference in 30-day mortality (7.6% versus 9.5%, p = 0.80). CONCLUSIONS: Although the overall incidence of CDAD was low, alteration in transfusion practices and antibiotic utilization may impact the development of CDAD among cardiac surgical patients.

Use of the Patient State Index (PSI) to assist in the diagnosis of perioperative neurological injury and brain death.

We present a series of three postoperative cases that were admitted to a cardiothoracic intensive care unit (ICU) after major surgery. Due to the possible presence of residual postoperative neuromuscular blockade after surgery, a processed electroencephalograph (EEG) was applied prior to starting sedation. This was markedly abnormal in all three cases, and not in keeping with the residual anesthesia. The patients were immediately transported for a CT scan. In all three cases there was severe neurological injury incompatible with survival and end of life decisions were made. Although the utility of quantitative EEG technology, like the Bispectral index (BIS) or Patient State Analyzer (PSA), is becoming better defined in the operating room, the role in the ICU is less clear. We propose that the ICU use of the PSA 4000 may have affected our decision weighing the risk versus benefit of transporting a fresh postoperative case to the radiology suite, expedited the neurological diagnosis, and may have reduced overall ICU resource utilization.

Low-dose intrathecal morphine does not delay early extubation after cardiac surgery.

PURPOSE: This study was designed to examine the efficacy of low-dose intrathecal morphine (ITM) on extubation times and pain control after cardiac surgery. METHODS: 43 patients undergoing elective cardiac surgery were enrolled in this prospective, randomized, double-blind placebo controlled trial. Patients were given a pre-induction dose of ITM (6 microg x kg(-1) per ideal body weight in 5 mL normal saline, group ITM) or 5 mL of intrathecal normal saline (group ITS). Anesthesia was induced with thiopental (3 mg x kg(-1)), sufentanil, midazolam and rocuronium. The total allowable doses of sufentanil and midazolam for the entire case were limited to 0.5 microg x kg(-1) and 0.045 mg x kg(-1) respectively. Anesthesia was maintained with isoflurane before and during cardiopulmonary bypass (CPB), and with propofol after CPB. In the postanesthesia care unit, patients received nurse-administered morphine followed by patient-controlled analgesia morphine. Serial visual analogue scale pain scores, morphine use, mini-mental state examinations and pulmonary function tests were measured for 48 hr. Patient satisfaction questionnaires were completed at the time of discharge. RESULTS: Mean times to extubation from the application of dressings were short and did not differ between groups (ITM = 41.4 +/- 33.0 min, ITS = 39.2 +/- 37.1 min). During the first 24 hr postoperatively, the ITM group had improved pain control and a lower iv morphine requirement than the control group, both at rest and during deep breathing. Both forced expiratory volume in one second and forced vital capacity were improved in the ITM group. There were no differences in spinal-related side effects or in the overall complication rates. Patient satisfaction was high in both groups. CONCLUSION: Low-dose ITM for cardiac surgery did not delay early extubation, but it improved postoperative analgesia and pulmonary function.

Inhaled prostacyclin is safe, effective, and affordable in patients with pulmonary hypertension, right heart dysfunction, and refractory hypoxemia after cardiothoracic surgery.

BACKGROUND: The purpose of this study was to describe our institutional experience in using inhaled prostacyclin as a selective pulmonary vasodilator in patients with pulmonary hypertension, refractory hypoxemia, and right heart dysfunction after cardiothoracic surgery. METHODS: Between February 2001 and March 2003, cardiothoracic surgical patients with pulmonary hypertension (mean pulmonary artery pressure >30 mm Hg or systolic pulmonary artery pressure >40 mm Hg), hypoxemia (PaO(2)/fraction of inspired oxygen <150 mm Hg), or right heart dysfunction (central venous pressure >16 mm Hg and cardiac index <2.2 L.min(-1).m(-2)) were prospectively administered inhaled prostacyclin at an initial concentration of 20,000 ng/mL and then weaned per protocol. Hemodynamic variables were measured before the initiation of inhaled prostacyclin, 30 to 60 minutes after initiation, and again 4 to 6 hours later. RESULTS: One hundred twenty-six patients were enrolled during the study period. At both time points, inhaled prostacyclin significantly decreased the mean pulmonary artery pressure without altering the mean arterial pressure. The average length of time on inhaled prostacyclin was 45.6 hours. There were no adverse events attributable to inhaled prostacyclin. The average cost for inhaled prostacyclin was 150 US dollars per day. Compared with nitric oxide, which costs 3000 US dollars per day, the potential cost savings over this period were 681,686 US dollars. CONCLUSIONS: Inhaled prostacyclin seems to be a safe and effective pulmonary vasodilator for cardiothoracic surgical patients with pulmonary hypertension, refractory hypoxemia, or right heart dysfunction. Overall, inhaled prostacyclin significantly decreases mean pulmonary artery pressures without altering the mean arterial pressure. Compared with nitric oxide, there is no special equipment required for administration or toxicity monitoring, and the cost savings are substantial.