University HealthSystem Consortium Database Research Articles

Developing Models to Predict Intensive Care Unit Admission and Mortality for Adults with Acute Myeloid Leukemia (AML)

Introduction: The survival expectations for adults with acute myeloid leukemia (AML) have improved over the last 40 years, partly due to supportive care advances enabling the delivery of increasingly intensive treatment modalities. However, with increasing therapy intensity comes potential for increased use of hospital resources. Indeed, our recent work with a large national database - the University HealthSystem Consortium (UHC)– demonstrated that intensive care unit (ICU) admissions remained frequent in AML patients over the last decade and were associated with significant mortality and steeply climbing costs. As the next step, we now used this database to develop models to predict ICU admission (“ICU Admission Risk Score”) and mortality (“ICU Mortality Risk Score”) for hospitalized AML patients.Methods: A longitudinal discharge database derived from 239 U.S. UHC participating hospitals was used to retrospectively evaluate AML patients hospitalized from 2004-2012. Clinical data from each hospital's discharge summaries were extracted by certified coders and merged to create the central UHC database. Included were patients with a diagnosis of active AML based on ICD-9 CM coding. Patients were excluded if they had undergone a hematopoietic stem cell transplant. Generalized estimating equations were used to control for patients with >1 admission in the study period . Primary outcomes were ICU admission and in-hospital death after ICU admission. Independent variables included age, gender, geographic location, AML volume of the admitting hospital based on quartiles of unique patients, number of prior ICU admissions, and types of comorbidities and infections present. We then applied multivariate logistic regression to create the models in which each covariate and the multicomponent model's predictive ability was evaluated by testing the area under the receiver operating characteristic curve (AUC). We made 2 models for each outcome: a full model using all relevant covariates tested and a simplified model for facilitating use in clinical practice. Models' internal validity was evaluated with 10-fold cross-validation.Results: 73,955 AML patient admissions were identified for inclusion. Patients had a median age of 58 (range 18-102) years, were 54% male, and were hospitalized in the Northeast (33%), Central (31%), Western (18%), and Southern (17%) U.S. One percent of admissions were at very-low AML volume hospitals, 5% at low, 21% at medium, and 74% at high. Overall, 22% of patients were admitted to the ICU during their hospitalization with a median ICU length of stay of 5 (1st-3rd quartile: 4-25) days. In-hospital mortality was 28% for those requiring ICU care. Factors with the greatest risk for ICU admission and subsequent in-hospital mortality by odds ratio in multivariate analysis included younger age, treatment at a small volume center, baseline lung disease, renal disease, vascular disease, gram negative/positive sepsis, pneumonia, and increasing number of comorbidities. The AUC for the full model for ICU admission was 0.73 (95% confidence interval [CI]: 0.728-0.732) and for the simplified model (composed of number of prior admissions, thrombosis, lung disease, renal disease, vascular disease, fungal infection, and number of comorbidities; Table 1A) was 0.70 (95% CI: 0.690-0.695). The AUC for the full model for in-hospital mortality was 0.70 (95% CI: 0.701-0.705) and for the simplified model (composed of lung disease, renal disease, and number of comorbidities; Table 1B) was 0.86 (95% CI: 0.853-0.859). The AUC was actually higher for the simplified mortality model due to a small number of risk factors with high univariate AUCs; thus including the covariates with low predictive ability decreased the AUC of the large model.Conclusions: These simplified models composed of readily available clinical variables have good performance characteristics in their ability to predict hospital death, and to a lesser extent, ICU admission, for hospitalized AML patients. The inferior performance of the ICU model may be related to heterogeneity in ICU admission criteria between hospitals. As a next step, these models will be externally validated and refined with our large institutional AML dataset to improve their predictive ability. If validated, we plan to use them to risk stratify patients upon hospital admission to allow for testing of primary prevention and intervention strategies. [Display omitted] DisclosuresWalter:Aptevo Therapeutics: Research Funding; ADC Therapeutics: Research Funding.

The Impact of Selecting Specific Cohorts for Benchmarking and Interpretation of Emergency Department Patient Satisfaction Scores.

Emergency departments (EDs) patient satisfaction metrics are highly valued by hospitals, health systems, and payers, yet these metrics are challenging to analyze and interpret. Accurate interpretation involves selection of the most appropriate peer group for benchmark comparisons. We hypothesized that the selection of different benchmark peer groups would yield different interpretations of Press Ganey (PG) patient satisfaction scores. Emergency department PG summary ratings of "doctors section" and "likelihood-to-recommend" raw scores and corresponding percentiles were derived for three benchmark peer groups from three academic years (2016, 2017, and 2018). The three benchmarks are: 1) the PG Large database; 2) the PG University HealthSystem Consortium (UHC) database; and 3) the Academy of Administrators in Academic Emergency Medicine (AAAEM) database, which is composed only of EDs from academic health centers with emergency medicine residency training programs. Raw scores were converted to percentile ranks for each distribution and then compared using Welch's ANOVA and Games-Howell pairwise comparisons. For both patient satisfaction raw scores evaluated, the AAAEM database was noted to have significantly higher percentile ranks when compared to the PG Large and PG UHC databases. These results were consistent for all three time frames assessed. Benchmarking with different peer groups provides different results, with similar patient satisfaction raw scores resulting in higher percentile ranks using the AAAEM database compared to the two PG databases. The AAAEM database should be considered the most appropriate peer group for benchmarking academic EDs.

Perioperative blood transfusions following hepatic lobectomy: A national analysis of academic medical centers in the modern era

BackgroundThe purpose of the study was to characterize the prevalence and impact of perioperative blood use for patients undergoing hepatic lobectomy at academic medical centers. MethodsThe University HealthSystem Consortium (UHC) database was queried for hepatic lobectomies performed between 2011 and 2014 (n = 6476). Patients were grouped according to transfusion requirements into high (>5 units, 7%), medium (2–5 units, 6%), low (1 unit, 8%), and none (0 units, 79%) during hospital stay for comparison of outcomes. ResultsOver 20% of patients undergoing hepatic lobectomy received blood perioperatively, of which 35% required more than 5 units. Patients with high transfusion requirements had increased severity of illness (p < 0.01). High transfusion requirements correlated with increased readmission rates (23.4% vs. 19.2% vs. 16.6% vs. 13.5%), total direct costs ($31,982 vs. $20,859 vs. $19,457 vs. $16,934), length of stay (9 days vs. 8 vs. 7 vs. 6), and in-hospital mortality (10.8% vs. 2.0% vs. 0.9% vs. 2.0%) compared to medium, low, and no transfusion amounts (all p < 0.01). Neither center nor surgeon volume were associated with transfusion use. ConclusionHigh transfusion requirements after hepatic lobectomy in the United States are associated with worse perioperative quality measures, but may not be influenced by center or surgeon volume.

Variability in blood transfusions after pancreaticoduodenectomy: A national analysis of the University HealthSystem Consortium

Variability in blood use after pancreaticoduodenectomy and the associated impact on readmission, mortality, and cost is not well understood at the national level. The University HealthSystem Consortium database was queried for all pancreaticoduodenectomies performed between the years 2011-2013 (n = 9,582). Patients were grouped according to transfusion requirements into none (0 units, 64%), low (1-2 units, 15%), medium (3-5 units, 13%), and high (>5 units, 8%). Multivariable analyses were used to determine predictors of increased transfusions, readmission, in-hospital mortality, and cost. Of the patients undergoing pancreaticoduodenectomy, 36% received blood perioperatively. Patients with high transfusion requirements were less often white, more often male, and had a higher severity of illness (all P < .01). High transfusion requirements correlated with higher readmission rates (OR 1.23, P = .03), cost (RR 1.84, P < .01), length of stay (18 vs. 13 vs. 10 vs. 8 days, P < .01), and in-hospital mortality (12.5% vs. 3.1% vs. 0.5% vs. 0.4%, P < .01). Higher-volume surgeons demonstrated lower transfusion requirements (OR 0.61, P < .01). Significant variability exists nationally in transfusion practices for patients undergoing pancreaticoduodenectomy, which may be driven most by severity of illness and surgeon volume. Efforts to reduce such variability could lead to improved outcomes and healthcare cost savings.

The Importance of Safety-Net Hospitals in Emergency General Surgery

IntroductionSafety-net hospitals provide care to an inherently underprivileged patient population. These hospitals have previously been shown to have inferior surgical outcomes after complex, elective procedures, but little is known about how hospital payer-mix correlates with outcomes after more common, emergent operations. MethodsThe University HealthSystem Consortium database was queried for all emergency general surgery procedures performed from 2009 to 2015. Emergency general surgery was defined as the seven operative procedures recently identified as contributing most to the national burden. Only urgent and emergent admissions were included (n = 653,305). Procedure-specific cohorts were created and hospitals were grouped according to safety-net burden. Multivariate analyses were done to study the effect of safety-net burden on hospital outcomes. ResultsFor all seven emergency procedures, patients at hospitals with a high safety-net burden were more likely to be young and black (p < 0.01 each). Patients at high-burden hospitals had similar severity of illness scores to those at other hospitals. Compared with lower burden hospitals, in-hospital mortality rates at high-burden hospitals were similar or lower in five of seven procedures (p = NS or < 0.01, respectively). After adjusting for patient factors, high-burden hospitals had similar or lower odds of readmission in six of seven procedures, hospital length of stay in four of seven procedures, and cost of care in three of seven procedures (p = NS or < 0.01, respectively). ConclusionSafety-net hospitals provide emergency general surgery services without compromising patient outcomes or incurring greater healthcare resources. These data may help inform the vital role these institutions play in the healthcare of vulnerable patients in the USA.

Perioperative blood transfusions following hepatic lobectomy: a national analysis of academic medical centers in the modern era

Background: We aim to characterize the prevalence and impact of perioperative blood use for patients undergoing hepatic lobectomy at academic medical centers on the national level. Methods: The University HealthSystems Consortium database was queried for hepatic lobectomies performed between 2011-2014 (n = 6476). Patients were grouped according to transfusion requirements into high (>5 units, n = 474), medium (2–5 units, n = 354), low (1 unit, n = 531), and none (0 units, n = 5117) during hospital stay. Logistic regression models were used to determine predictors of increased transfusions, readmission, mortality, and cost. Results: Approximately 21% of patients undergoing hepatic lobectomy received blood perioperatively. While a majority of transfusions were intraoperative or on day of surgery, patients with high transfusion requirements were more often transfused postoperatively (54% after day of surgery vs. 45% vs. 43%). High transfusion patients had a significantly higher severity of illness and more often had major complications or comorbidities (both p < 0.01). High transfusion requirements correlated with higher 30-day readmission (23% vs. 19% vs. 17% vs. 14%), total direct cost ($31982 vs. $20859 vs. $19457 vs. $16934), length of stay (9 days vs. 8 vs. 7 vs. 6) and in- hospital mortality (11% vs. 2% vs. 1% vs. 2%)(all p < 0.01). Neither center volume nor surgeon volume were significantly associated with transfusion use following hepatic lobectomy. Conclusion: This data suggests there is a high transfusion requirement after hepatic lobectomy in the United States. Perioperative blood transfusions significantly impacted patient outcomes following hepatic lobectomy, however, were not influenced by center or surgeon volume. Focusing efforts on appropriate patient selection and minimizing comorbidities could lead to improved resource utilization and reduced healthcare costs.

Antibiotic Timing and Outcomes in Sepsis

BackgroundWe evaluated the effect of time spent in the emergency department (ED) and process of care on mortality and length of hospital stay in patients with sepsis or septic shock. MethodsAn observational cohort study was conducted on 117 patients who came through the University of Louisville Hospital ED and subsequently were directly admitted to the intensive care unit (ICU). Variables of interest were time in the ED from triage to physical transport to the ICU, from triage to antibiotic(s) ordered, and from triage to antibiotic(s) administered. Expected mortality was calculated according to the University Health System Consortium Database. Primary and secondary outcomes were in-hospital death and hospital length of stay in days, respectively. ResultsWe found no significant association between time in the ED and mortality between survivors and nonsurvivors (5.5 versus 5.7 hours, P = 0.804). After adjusting for expected mortality, a 22% increase in mortality risk was found for each hour delay from triage to antibiotic(s) ordered; a 15% increase in mortality risk was observed for each hour from triage to antibiotic(s) given. Both time from triage to antibiotic(s) ordered (hazard ratio [HR] = 0.8, P = 0.044) and time from triage to antibiotic(s) delivery (HR = 0.79, P = 0.0092) were independently associated with an increased hospital stay (HR = 0.79, P = 0.0092). ConclusionThough no significant association between mortality and ED time was demonstrated, we observed a significant increase in mortality in septic patients with both delays in antibiotic(s) order and administration. Delay in care also resulted in increased hospital stays both overall and in the ICU.

Venous Thromboembolism Rates in Transferred Patients: A Cross-Sectional Study.

Patients undergoing a transfer during a hospitalization may be more likely to be diagnosed with a venous thromboembolism (VTE) than patients who are not transferred. To determine whether transferred patients have an increased prevalence of VTE diagnosis. This was a cross-sectional study comparing VTE diagnosis rates between transferred patients and non-transferred patients. For the years 2012-2014, the University HealthSystem Consortium database of multiple community and academic medical centers throughout the United States was parsed using ICD-9 VTE diagnosis codes and patient's point of origin. Patients were included in the analysis as transferred patients if their point of origin was a skilled nursing facility, another acute care facility or another facility. Non-transferred patients were those whose point of origin was a clinic or those with a non-facility point of origin. The primary comparison of VTE prevalence during hospitalization between transferred and non-transferred patients in the years 2012-2014. Subgroup analysis looked at level I trauma status and case mix index (CMI) to determine whether these had an effect on VTE prevalence. From 2012 to 2014, a total of 225 unique hospitals and 12,036,029 patients were analyzed, and the prevalence of VTE in transferred patients and non-transferred patients was 3.43% and 1.91% (RR 1.80; 95% CI 1.78-1.81; P<0.001), respectively. VTE prevalence in transferred versus non-transferred patients at level I trauma centers was 3.42% versus 1.88% (RR=1.82; 95% CI 1.80-1.85; P<0.001). The 3-year average CMI of transferred versus non-transferred patients was 3.53 versus 2.26 (P< 0.001). Transferred patients have a higher prevalence of VTE than non-transferred patients, regardless of level I trauma designation. Higher VTE rates in transferred versus non-transferred patients was minimally correlated with CMI.

Laparoscopic subtotal cholecystectomy compared to total cholecystectomy: a matched national analysis

BackgroundLaparoscopic subtotal cholecystectomy (LSC) is considered a safe alternative to laparoscopic cholecystectomy (LC) if biliary anatomy is obscured by inflammation. While case series studies have observed low morbidity rates with LSC, the impact of operative conversion on patient outcomes is poorly understood. MethodsA national analysis of all patients who underwent LC or LSC from 2009 to 2013 was performed using the University HealthSystem Consortium database. A 1:1 propensity score match was used to compare procedural outcomes accounting for clinical and demographic factors. Matched samples had <10% standardized differences of each baseline covariate. ResultsA total of 131,082 LC and 487 LSC were performed during the study period. Compared with LC, patients undergoing LSC were more likely to be older (56 versus 48 years), male (54.2% versus 32.3%), and have higher severity of illness scores on admission (9.2% versus 3.5% extreme severity of illness; P < 0.001 each). LSC patients had a prolonged hospital length of stay (LOS, 4 versus 3 days), greater total direct cost ($9053 versus $6398), higher readmission rates (11.9% versus 7.0%), and higher mortality rates (0.82% versus 0.28%, P < 0.05 each). After matching, the difference in total direct cost persisted ($9053 versus $7,581, P < 0.001), but there were no differences in hospital LOS, readmission rates, or overall mortality. ConclusionsLSC is an important alternative to LC for the difficult gallbladder. Conversion to LSC is associated with increased patient morbidity and resource utilization leading to perceived poor outcomes, but this is due to patient factors at initial presentation. Health care providers should consider LSC if the patient may be at risk for iatrogenic injury to the biliary tract.

Association of Centers for Medicare & Medicaid Services Overall Hospital Quality Star Rating With Outcomes in Advanced Laparoscopic Abdominal Surgery

The Centers for Medicare & Medicaid Services (CMS) recently released the Overall Hospital Quality Star Rating to help patients compare hospitals based on a 5-star scale. The star rating was designed to assess overall quality of the institution; thus, its validity toward specifically assessing surgical quality is unknown. To examine whether CMS high-star hospitals (HSHs) have improved patient outcomes and resource use in advanced laparoscopic abdominal surgery compared with low-star hospitals (LSHs). Using the University HealthSystem Consortium database (which includes academic centers and their affiliate hospitals) from January 1, 2013, through December 31, 2015, this administrative database observational study compared outcomes of 72 662 advanced laparoscopic abdominal operations between HSHs (4-5 stars) and LSHs (1-2 stars). The star rating includes 57 measures across 7 areas of quality. Patients who underwent advanced laparoscopic abdominal surgery, including bariatric surgery (sleeve gastrectomy, Roux-en-Y gastric bypass), colorectal surgery (colectomy, proctectomy), or hiatal hernia surgery (paraesophageal hernia repair, Nissen fundoplication), were included. Risk adjustment included exclusion of patients with major and extreme severity of illness. Main outcome measures included serious morbidity, in-hospital mortality, intensive care unit admissions, and cost. A total of 72 662 advanced laparoscopic abdominal operations were performed in patients at 66 HSHs (n = 38 299; mean [SD] age, 51.26 [15.25] years; 12 096 [31.5%] male and 26 203 [68.4%] female; 28 971 [75.6%] white and 9328 [24.4%] nonwhite) and 78 LSHs (n = 34 363; mean [SD] age, 49.77 [14.77] years; 9902 [28.8%] male and 24 461 [71.2%] female; 21 876 [67.6%] white and 12 487 [32.4%] nonwhite). The HSHs were observed to have fewer intensive care unit admissions (1007 [2.6%] vs 1711 [5.0%], P < .001) and lower mean cost ($7866 vs $8708, P < .001). No significant difference was found in mortality between HSHs and LSHs for any advanced laparoscopic abdominal surgery. No significant difference was found in serious morbidity between HSHs and LSHs for bariatric or hiatal hernia surgery. However, for colorectal surgery, serious morbidity was lower at HSHs compared with LSHs (258 [2.2%] vs 276 [2.9%], P = .002). This study found that HSHs treat fewer ethnic minorities and have similar outcomes as LSHs for advanced laparoscopic abdominal operations. However, HSHs may represent hospitals with improved resource use and cost.

In-hospital use of non-steroidal anti-inflammatory drugs in patients withheart failure in academic centers intheUnited States.

Non-steroidal anti-inflammatory drugs are considered potentially harmful for patients with heart failure. To determine the prevalence of in-hospital NSAID use, their type, associated diagnosis and impact in clinical outcomes among patients with a diagnosis of heart failure. The University Health System Consortium Database was used to identify all first hospitalizations with an International Classification of Diseases-9 discharge diagnosis code of systolic heart failure as the primary diagnosis between January 1, 2011, and December 31st 2014. Among 65,902 patients admitted for a primary diagnosis of SHF, 2675 (4.1%) were exposed to NSAID. The most frequent NSAID used was ibuprofen (51.63%), followed by ketorolac (29.38%) naproxen (8.07%) celecoxib (5.61%), and others. On multivariable analyses, the length of stay of patients exposed to NSAID was longer compared to non-exposed (OR: 4.67, p < 0.001, 95% CI 4.10-5.25), but differences in mortality were not statistically different (OR: 0.90, p = 0.476, 95% CI 0.69-1.19). The use of NSAID in patients admitted with a primary diagnosis of systolic heart failure was low but was associated with longer length of stay. Further studies are needed to understand the impact of NSAID use in this patient population.

Malnutrition Diagnosis during Adult Inpatient Hospitalizations: Analysis of a Multi-Institutional Collaborative Database of Academic Medical Centers

BackgroundMalnutrition is a significant problem for hospitalized patients. However, the true prevalence of reported malnutrition diagnosis in real-world clinical practice is largely unknown. Using a large collaborative multi-institutional database, the rate of malnutrition diagnosis was assessed and used to assess institutional variables associated with higher rates of malnutrition diagnosis. ObjectiveThe aim of this study was to define the prevalence of malnutrition diagnosis reported among inpatient hospitalizations. DesignThe University Health System Consortium (Vizient) database was retrospectively reviewed for reported rates of malnutrition diagnosis. Participants/settingAll adult inpatient hospitalization at 105 member institutions during fiscal years 2014 and 2015 were evaluated. Main outcome measuresMalnutrition diagnosis based on the presence of an International Classification of Diseases-Ninth Revision diagnosis code. Statistical analysisHospital volume and publicly available hospital rankings and patient satisfaction scores were obtained. Multiple regression analysis was performed to assess the association between these variables and reported rates of malnutrition. ResultsA total of 5,896,792 hospitalizations were identified from 105 institutions during the 2-year period. It was found that 292,754 patients (5.0%) had a malnutrition diagnosis during their hospital stay. By institution, median rate of malnutrition diagnosis during hospitalization was 4.0%, whereas the rate of severe malnutrition diagnosis was 0.9%. There was a statistically significant increase in malnutrition diagnosis from 4.0% to 4.9% between 2014 and 2015 (P<0.01). Institutional factors associated with increased diagnosis of malnutrition were higher hospital volume, hospital ranking, and patient satisfaction scores (P<0.01). ConclusionsMissing a malnutrition diagnosis appears to be a universal issue because the rate of malnutrition diagnosis was consistently low across academic medical centers. Institutional variables were associated with the prevalence of malnutrition diagnosis, which suggests that institutional culture influences malnutrition diagnosis. Quality improvement efforts aimed at improved structure and process appear to be needed to improve the identification of malnutrition.

Diffusion of technology: Trends in robotic-assisted colorectal surgery

Following FDA approval, robotic-assisted colorectal surgery (RACS) has increased in prevalence. We aimed to identify trends in utilization and patient characteristics of RACS in the United States using the University HealthSystem Consortium database between October 2011–September 2015. Outcome measures were number and percentage of procedures performed with robotic-assistance.7100 patients were identified. The most common procedures were low anterior resection, sigmoid colectomy, abdominoperineal resection, right colectomy, rectopexy, left colectomy, and total colectomy. There was a 158% increase in RACS procedures. As a percentage of all approaches, RACS increased from 2.6% to 6.6%. The number of centers performing RACS increased from 105 to 140. Over the study period, the complexity of patients increased, with the percentage of patients with ≥3 comorbidities rising from 18% to 24% (p = 0.03) and patients with a moderate severity of illness score increasing from 35% to 41% (p = 0.04).RACS has expanded significantly in volume, number of centers, and patient selection. Further studies evaluating outcomes and cost of RACS are required to determine whether these increases are justified by improved clinical outcomes.

Extended use of perioperative antibiotics in head and neck microvascular reconstruction

PurposeMany head and neck surgical procedures are considered clean-contaminated wounds and antibiotic prophylaxis is recommended. Despite prophylaxis, the incidence of surgical site infections remains significant – especially in the setting of free tissue transfer. The antibiotic course is often of a longer duration after free tissue transfer than the recommended 24hour post-operatively. Currently, there is no consensus on appropriate antibiotic regimen or duration at this time. This study investigates the outcomes of a 7-day perioperative antibiotic regimen after microvascular reconstruction of the head and neck at our institution. Materials and methodsA retrospective review was performed of 72 patients undergoing microvascular free tissue at our institution between 09/2011 and 03/2014. The antibiotic regimen, post-operative surgical (including surgical site infections) and medical complications were noted. Our rates of complications and adverse events were compared to all surgical patients, as well as all inpatients hospital-wide with use of the University Health System Consortium database. ResultsSeventy-two subjects met inclusion criteria for this study. The majority of subjects received cefazolin/metronidazole (69.4%). Subjects with beta-lactam allergy received clindamycin (12.5%). The remainder received an alternative regimen (18.1%). All received at least 7days of antibiotics. The rate of hospital acquired C. difficile diarrhea was 0.57% hospital-wide, 1.13% in Otolaryngology patients, and 1.4% in this study. There were no instances of a multi-drug resistant infection or any adverse reactions to the administration of antibiotics. When compared with other antibiotic regimens, clindamycin was associated with a significantly increased rate of either medical or surgical infections (OR 14.38, p=0.02) and longer hospital stay (average=18days, p<0.05). ConclusionThe use of a 7-day prophylactic antibiotic regimen is not associated with an increased risk of antibiotic-associated infections, multi-drug resistant infections, or antibiotic-associated complications. The use of clindamycin is associated with increased risk of medical and surgical infections post-operatively and should be avoided in the prophylactic perioperative phase after free tissue transfer of the head and neck.

Human Albumin Use in Adults in U.S. Academic Medical Centers

To determine rates and predictors of albumin administration, and estimated costs in hospitalized adults in the United States. Cohort study of adult patients from the University HealthSystem Consortium database from 2009 to 2013. One hundred twenty academic medical centers and 299 affiliated hospitals. A total of 12,366,264 hospitalization records. Analysis of rates and predictors of albumin administration, and estimated costs. Overall the proportion of admissions during which albumin was administered increased from 6.2% in 2009 to 7.5% in 2013; absolute difference 1.3% (95% CI, 1.30-1.40%; p < 0.0001). The increase was greater in surgical patients from 11.7% in 2009 to 15.1% in 2013; absolute difference 3.4% (95% CI, 3.26-3.46%; p < 0.0001). Albumin use varied geographically being lowest with no increase in hospitals in the North Eastern United States (4.9% in 2009 and 5.3% in 2013) and was more common in bigger (> 750 beds; 5.2% in 2009 and 7.3% in 2013) compared to smaller hospitals (< 250 beds; 4.4% in 2009 to 6.2% in 2013). Factors independently associated with albumin use were appropriate indication for albumin use (odds ratio, 65.220; 95% CI, 62.459-68.103); surgical admission (odds ratio, 7.942; 95% CI, 7.889-7.995); and high severity of illness (odds ratio, 8.933; 95% CI, 8.825-9.042). Total estimated albumin cost significantly increased from $325 million in 2009 to $468 million in 2013; (absolute increase of $233 million), p value less than 0.0001. The proportion of hospitalized adults in the United States receiving albumin has increased, with marked, and currently unexplained, geographic variability and variability by hospital size.

Case mix–adjusted cost of colectomy at low-, middle-, and high-volume academic centers

Efforts to regionalize surgery based on thresholds in procedure volume may have consequences on the cost of health care delivery. This study aims to delineate the relationship between hospital volume, case mix, and variability in the cost of operative intervention using colectomy as the model. All patients undergoing colectomy (n=90,583) at 183 academic hospitals from 2009-2012 in The University HealthSystems Consortium Database were studied. Patient and procedure details were used to generate a case mix-adjusted predictive model of total direct costs. Observed to expected costs for each center were evaluated between centers based on overall procedure volume. Patient and procedure characteristics were significantly different between volume tertiles. Observed costs at high-volume centers were less than at middle- and low-volume centers. According to our predictive model, high-volume centers cared for a less expensive case mix than middle- and low-volume centers ($12,786 vs $13,236 and $14,497, P<.01). Our predictive model accounted for 44% of the variation in costs. Overall efficiency (standardized observed to expected costs) was greatest at high-volume centers compared to middle- and low-volume tertiles (z score -0.16 vs 0.02 and -0.07, P<.01). Hospital costs and cost efficiency after an elective colectomy varies significantly between centers and may be attributed partially to the patient differences at those centers. These data demonstrate that a significant proportion of the cost variation is due to a distinct case mix at low-volume centers, which may lead to perceived poor performance at these centers.

Implementation of the surgical safety checklist at a tertiary academic center: Impact on safety culture and patient outcomes

BackgroundThe impact and efficacy of the World Health Organization Surgery Safety Checklist (SSC) is uncertain. We sought to determine if the SSC decreases complications and examined the attitudes of the surgical team members following implementation of the SSC. MethodsA 28-question survey was developed to assess perspectives of surgical team members at the University of Vermont Medical Center (UVMC). The University Health System Consortium database was examined to compare the rates of nine complications before and after SSC implementation using Chi square analysis and Fisher's exact test. ResultsThere was no significant decrease in any of the nine complications 2 years after SSC implementation. There was overall agreement that the SSC improved communication, safety, and prevented errors in the operating room. However, there was disagreement between nursing and surgeons over whether all three parts of the SSC were always completed. ConclusionsImplementation of the SSC did not result in a significant decrease in perioperative morbidity or mortality. However, it did improve the perception of safety culture by operating room staff.

Association of Risk Factors, Mortality, and Care Costs of Adults With Acute Myeloid Leukemia With Admission to the Intensive Care Unit

ImportanceAdults with acute myeloid leukemia (AML) commonly require support in the intensive care unit (ICU), but risk factors for admission to the ICU and adverse outcomes remain poorly defined.ObjectiveTo examine risk factors, mortality, length of stay, and cost associated with admission to the ICU for patients with AML.Design, Setting, and ParticipantsThis study extracted information from the University HealthSystem Consortium database on patients 18 years or older with AML who were hospitalized for any cause between January 1, 2004, and December 31, 2012. The University HealthSystem Consortium database contains demographic, clinical, and cost variables prospectively abstracted by certified coders from discharge summaries. Outcomes were analyzed using univariate and multivariable statistical techniques. Data analysis was performed from November 15, 2013, to August 15, 2016.Main Outcomes and MeasuresPrimary outcomes were admission to the ICU and inpatient mortality among patients requiring ICU care. Secondary outcomes included length of stay in the ICU, total hospitalization length of stay, and cost.ResultsOf the 43 249 patients with AML (mean [SD] age, 59.5 [16.6] years; 23 939 men and 19 310 women), 11 277 (26.1%) were admitted to the ICU. On multivariable analysis (with results reported as odds ratios [95% CIs]), independent risk factors for admission to the ICU included age younger than 80 years (1.56 [1.42-1.70]), hospitalization in the South (1.81 [1.71-1.92]), hospitalization at a low- or medium-volume hospital (1.25 [1.19-1.31]), number of comorbidities (10.64 [8.89-12.62] for 5 vs none), sepsis (4.61 [4.34-4.89]), invasive fungal infection (1.24 [1.11-1.39]), and pneumonia (1.73 [1.63-1.82]). In-hospital mortality was higher for patients requiring ICU care (4857 of 11 277 [43.1%] vs 2959 of 31 972 [9.3%]). On multivariable analysis, independent risk factors for death in patients requiring ICU care included age 60 years or older (1.16 [1.06-1.26]), nonwhite race/ethnicity (1.18 [1.07-1.30]), hospitalization on the West coast (1.19 [1.06-1.34]), number of comorbidities (18.76 [13.7-25.67] for 5 vs none), sepsis (2.94 [2.70-3.21]), invasive fungal infection (1.20 [1.02-1.42]), and pneumonia (1.13 [1.04-1.24]). Mean costs of hospitalization were higher for patients requiring ICU care ($83 354 vs $41 973) and increased with each comorbidity, from $50 543 for patients with no comorbidities to $124 820 for those with 5 or more comorbidities.Conclusions and RelevanceAdmission to the ICU is associated with high mortality and cost that increase proportionally with the comorbidity burden in adults with AML. Several demographic factors and medical characteristics identify patients at risk for admission to the ICU and mortality and provide an opportunity for testing primary prevention strategies.

Pediatric surgical readmissions: Are they truly preventable?

Background/purposeReimbursement penalties for excess hospital readmissions have begun for the pediatric population. Therefore, research determining incidence and predictors is critical. MethodsA retrospective review of University HealthSystem Consortium database (N=258 hospitals; 2,723,621 patients) for pediatric patients (age 0–17years) hospitalized from 9/2011 to 3/2015 was performed. Outcome measures were 7-, 14-, and 30-day readmission rates. Hospital and patient characteristics were evaluated to identify predictors of readmission. ResultsReadmission rates at 7, 14, and 30days were 2.1%, 3.1%, and 4.4%. For pediatric surgery patients (N=260,042), neither index hospitalization length of stay (LOS) nor presence of a complication predicted higher readmissions. Appendectomy was the most common procedure leading to readmission. Evaluating institutional data (N=5785), patients admitted for spine surgery, neurosurgery, transplant, or surgical oncology had higher readmission rates. Readmission diagnoses were most commonly infectious (37.2%) or for nausea/vomiting/dehydration (51.1%). Patients with chronic medical conditions comprised 55.8% of patients readmitted within 7days. 92.0% of patients requiring multiple rehospitalizations had comorbidities. ConclusionsReadmission rates for pediatric patients are significantly lower than adults. Risk factors for adult readmissions do not predict pediatric readmissions. Readmission may be a misnomer for the pediatric surgical population, as most are related to chronic medical conditions and other nonmodifiable risk factors. Level of EvidenceLevel IV.

Variability in Standard Outcomes of Posterior Lumbar Fusion Determined by National Databases

National databases are used with increasing frequency in spine surgery literature to evaluate patient outcomes. The differences between individual databases in relationship to outcomes of lumbar fusion are not known. We evaluated the variability in standard outcomes of posterior lumbar fusion between the University HealthSystem Consortium (UHC) database and the Healthcare Cost and Utilization Project National Inpatient Sample (NIS). NIS and UHC databases were queried for all posterior lumbar fusions (International Classification of Diseases, Ninth Revision code 81.07) performed in 2012. Patient demographics, comorbidities (including obesity), length of stay (LOS), in-hospital mortality, and complications such as urinary tract infection, deep venous thrombosis, pulmonary embolism, myocardial infarction, durotomy, and surgical site infection were collected using specific International Classification of Diseases, Ninth Revision codes. Analysis included 21,470 patients from the NIS database and 14,898 patients from the UHC database. Demographic data were not significantly different between databases. Obesity was more prevalent in UHC (P= 0.001). Mean LOS was 3.8 days in NIS and 4.55 in UHC (P < 0.0001). Complications were significantly higher in UHC, including urinary tract infection, deep venous thrombosis, pulmonary embolism, myocardial infarction, surgical site infection, and durotomy. In-hospital mortality was similar between databases. NIS and UHC databases had similar demographic patient populations undergoing posterior lumbar fusion. However, the UHC database reported significantly higher complication rate and longer LOS. This difference may reflect academic institutions treating higher-risk patients; however, a definitive reason for the variability between databases is unknown. The inability to precisely determine the basis of the variability between databases highlights the limitations of using administrative databases for spinal outcome analysis.