Matthew G. Johnson

Intensive Intraoperative Insulin Therapy versus Conventional Glucose Management during Cardiac Surgery: A Randomized Trial

Context Intensive insulin therapy used to maintain normoglycemia during intensive care after cardiac surgery improves perioperative outcomes. Its effect during cardiac surgery is unknown. Contributions The authors randomly assigned 400 cardiac surgical patients to tight glycemic control (blood glucose level, 4.4 to 5.6 mmol/L [80 to 100 mg/dL]) during surgery or usual intraoperative care. All patients received tight glycemic control in the cardiac intensive care unit. The groups had the same risk for perioperative adverse events (risk ratio, 1.0 [95% CI, 0.8 to 1.2]). The intensive treatment group had more strokes (8 vs. 1) and more deaths (4 vs. 0) than the conventional treatment group. Caution The authors performed the study at a single center. Implications Maintaining normoglycemia during cardiac surgery does not improve outcomes and might worsen them. The Editors Hyperglycemia occurs frequently in patients with and without diabetes during cardiac surgery, especially during cardiopulmonary bypass surgery (1, 2). In a study by Van den Berghe and colleagues (3), intensive insulin therapy after surgery reduced morbidity and death in critically ill patients, most of whom underwent cardiac surgery. As a result, professional organizations have recommended rigorous glycemic control in hospitalized patients (4) and strict glycemic control is now routine practice during the postoperative period in cardiac surgical patients. However, no consensus exists on the optimal management of intraoperative hyperglycemia in cardiac surgical patients because of the lack of evidence from randomized trials. Researchers are increasingly extrapolating evidence from studies that assess the role of strict postoperative glycemic control in critically ill patients to advocate for intravenous insulin therapy for patients in the operating room (3, 57). Evidence, strictly from observational studies, suggests that tight intraoperative glycemic control may reduce postoperative complications (810). We recently reported, in a retrospective, observational study of 409 cardiac surgical patients, that intraoperative hyperglycemia was an independent risk factor for perioperative complications, including death, after adjustment for postoperative glucose concentrations. Each 1.1-mmol/L (20 mg/dL) increase in glucose concentration greater than 5.6 mmol/L (>100 mg/dL) during surgery was associated with a 34% increase in the likelihood of postoperative complications (8). An association between intraoperative hyperglycemia and adverse outcomes based on observational studies does not prove causality. Because hyperglycemia can adversely affect immunity, wound healing, and vascular function, the concept that normoglycemia be maintained during the relatively brief duration of cardiac surgery seems plausible (1116). On the other hand, the degree of intraoperative hyperglycemia may merely reflect the severity of underlying stress. If so, prevention of hyperglycemia might not reduce perioperative complications, and the risks and costs of intensive intraoperative glycemic management may outweigh the benefits. Simple, safe, and effective insulin infusion algorithms that achieve rigorous intraoperative glycemic control are lacking. To address these questions, we conducted a randomized, controlled trial at 1 center to determine whether maintenance of near normoglycemia during cardiac surgery by using intraoperative intravenous insulin infusion reduced perioperative death and morbidity when added to rigorous postoperative glycemic control. Methods Design Overview This was a randomized, open-label, controlled trial with blinded assessment. We randomly assigned patients to receive intensive insulin therapy to maintain intraoperative glucose levels between 4.4 (80 mg/dL) and 5.6 mmol/L (100 mg/dL) or conventional treatment. By design, both groups were postoperatively treated with strict glycemic control to ensure that the observed difference in outcome could be attributed to the effects of intraoperative glycemic control. Setting We performed the study at St. Marys Hospital, Rochester, Minnesota, which is a tertiary care teaching hospital with 1157 beds and an average of more than 41000 admissions per year. Participants Adults undergoing elective cardiac surgery between July 2004 and April 2005 were eligible for enrollment in our study. We excluded patients who had off-pump cardiopulmonary bypass procedures. The Mayo Foundation Institutional Review Board, Rochester, Minnesota, approved the protocol. Randomization and Interventions Before we enrolled patients in our randomized trial, we enrolled 20 patients in a 2-week pilot trial to ensure that the anesthesiologists in the operating room and the nursing staff in the intensive care units (ICUs) had adequate experience with the study insulin infusion algorithm. The 20 patients received intensive insulin therapy during surgery and for 24 hours after surgery. The pilot period data allowed us to modify the graded insulin infusion to achieve desired glucose concentration goals. We built safety features into our infusion protocol to minimize hypoglycemia. We discontinued the infusion when glucose levels were less than 4.4 mmol/L (<80 mg/dL) and initiated dextrose infusion. When glucose levels decreased to less than 3.3 mmol/L (<60 mg/dL), we treated hypoglycemia according to a standardized hypoglycemia protocol. Per protocol, patients treated in the pilot phase were not included in the analyzed cohort. Study coordinators obtained written informed consent from all patients who met eligibility criteria. We randomly assigned patients to receive intensive or conventional intraoperative insulin therapy. Randomization was computer-generated with permuted blocks of 4, with stratification according to surgeon, surgical procedure (coronary artery bypass grafting [CABG] with or without other procedures and no CABG), and diabetes. The randomization assignments were concealed in opaque, sealed, tamper-proof envelopes that were opened sequentially by study personnel after participants signed the patient consent form. We could not possibly know, before obtaining consent, the few patients who would not have intraoperative hyperglycemia (glucose concentration of 5.6 mmol/L or more [100 mg/dL]). Therefore, per protocol, patients who gave consent were randomly assigned, and those whose glucose levels were less than 5.6 mmol/L (<100 mg/dL) during surgery were not included in the final analyses. Intraoperative Period Intensive Treatment Patients in the intensive treatment group received a continuous intravenous insulin infusion, 250 units of NovoLin R (Novo Nordisk, Princeton, New Jersey) in 250 mL of 0.45% sodium chloride, when their blood glucose levels exceeded 5.6 mmol/L (>100 mg/dL). We adjusted the infusions to maintain blood glucose levels between 4.4 (80 mg/dL) and 5.6 mmol/L (100 mg/dL). We adjusted the dose according to a standardized algorithm used by anesthesiologists (Appendix Table 1). Appendix Table 1. Insulin Infusion Protocol* Conventional Treatment Patients in the conventional treatment group did not receive insulin during surgery unless their glucose levels exceeded 11.1 mmol/L (200 mg/dL). If glucose concentration was between 11.1 (200 mg/dL) and 13.9 mmol/L (250 mg/dL), patients received an intravenous bolus of 4 units insulin every hour until the glucose concentration was less than 11.1 mmol/L (<200 mg/dL). If the intraoperative glucose concentration was greater than 13.9 mmol/L (>250 mg/dL), patients received an intravenous infusion of insulin that was continued until the glucose level was less than 8.3 mmol/L (<150 mg/dL). In both study groups, we measured arterial plasma glucose concentration every 30 minutes, starting just before anesthetic induction by using hexokinase method on a Double P Modular System (Roche Diagnostics, Indianapolis, Indiana). Intraoperative procedures, including cardiopulmonary bypass, monitoring, laboratory testing, and treatment, were left to the discretion of anesthesiologists and cardiac surgeons. There was no standard protocol for monitoring and managing intraoperative potassium levels. Postoperative Period Intravenous insulin infusion was started in patients in the conventional treatment group on their arrival in the ICU. Thereafter, both study groups were treated identically, with the intravenous insulin infusion rates adjusted by a nursing staff that was not involved with the study according to a standard protocol. The target blood glucose range was 4.4 (80 mg/dL) to 5.6 mmol/L (100 mg/dL) (Appendix Table 1). Arterial blood glucose levels were measured every 1 to 2 hours by using the Accu-Check Inform blood glucose monitoring system (glucometer) (Roche Diagnostics). During the first 24 hours after surgery, patients were given only clear liquids by mouth; we did not administer subcutaneous insulin or oral diabetic medications during this time. Thereafter, the hospital diabetes consulting service saw all patients and provided individualized recommendations for ongoing care. Outcomes and Measurements The primary outcome variable was a composite of death, sternal wound infections, prolonged pulmonary ventilation, cardiac arrhythmias (new-onset atrial fibrillation, heart block requiring permanent pacemaker, or cardiac arrest), stroke, and acute renal failure within 30 days after surgery. Secondary outcome measures were length of stay in the ICU and hospital. Trained study personnel identified the occurrence of a complication through chart abstraction by using confirmable, objective criteria in accordance with standardized definitions from the Society of Thoracic Surgeons (STS) database committee (17). Personnel who assessed outcomes were not aware of patient treatment assignment or of the study hypothesis. Follow-up Procedures We contacted patients by telephone and used a standardized telephone survey at 30 days after surgery to assess outcomes that occurred after discharge. We considered pat

Effect of Hospital Follow-up Appointment on Clinical Event Outcomes and Mortality

BACKGROUND Decreasing hospital readmission and patient mortality after hospital dismissal is important when providing quality health care. Interventions recently proposed by the Centers for Medicare and Medicaid Services to reduce avoidable hospital readmissions include providing patients with clear discharge instructions and appointments for timely follow-up visits. Although research has demonstrated a correlation between follow-up arrangements and reduced hospital readmission in specific patient populations, the effect of hospital follow-up in general medicine patients has not been assessed. METHODS For this study, we reviewed hospital dismissal instructions for general medicine patients dismissed in 2006 from Mayo Clinic hospitals in Rochester, Minnesota (n = 4989), and determined whether specific appointment details for follow-up were documented. Survival analysis and propensity score-adjusted proportional hazards regression models were developed to investigate the association of follow-up appointment arrangements with hospital readmission, emergency department visits, and mortality at 30 and 180 days after discharge. RESULTS Of the 4989 dismissal summaries, 3037 (60.9%) contained instructions for a follow-up appointment. No difference was found between those with a documented follow-up appointment vs those without regarding hospital readmission, emergency department visits, or mortality 30 days after dismissal. However, those with a documented follow-up appointment were slightly more likely to have an adverse event (hospital readmission, emergency department visit, or death) within 180 days after dismissal. CONCLUSIONS Improved discharge processes, including arrangement of hospital follow-up appointments, do not appear to improve readmission rates or survival in general medicine patients. Therefore, national efforts to ensure follow-up for all patients after hospital dismissal may not be beneficial or cost-effective.

Association of heparin-dependent antibodies and adverse outcomes in hemodialysis patients: a population-based study.

OBJECTIVE To determine whether some adverse outcomes of hemodialysis could be explained by subclinical heparin-induced thrombocytopenia (HIT). PATIENTS AND METHODS Platelet factor 4 (PF4)-heparin antibodies were measured by enzyme-linked immunosorbent assay In a population-based cohort of hemodlalysis patients. Participants were then followed up prospectively for thromboembollc events, cardiovascular events, or death. RESULTS Of the 59 hemodialysis patients residing In Olmsted County, Minnesota, 57 (97%) agreed to study participation. The mean +/- SD age of the patients was 64 +/- 17 years (median hemodialysis duration, 23 months), and 27 (47%) were women. The enzyme-linked Immunosorbent assay was positive for PF4-heparin antibodies in 2 patients (3.5%). The PF4-heparin antibody content varied over a 10-fold range and was not associated with the duration of hemodialysis (P = .99). During a median follow-up of 798 days, 16 thrombotic events, 37 cardiovascular events, and 23 deaths (Including 13 cardiovascular deaths) occurred. After adjusting for the Framingham risk score, the all-cause mortality rate was significantly higher for patients with the highest tertile of PF4-heparln antibody content compared with patients in the lower tertilles (hazard ratio, 2.47; P = .03). Furthermore, 8 (73%) of deaths in this tertile were due to cardiovascular causes (hazard ratio, 4.14; P = .02). CONCLUSIONS Despite repetitive heparin exposure, the prevalence of HIT In patients undergoing maintenance hemodialysis is no greater than that anticipated for other patient populations. However, to our knowledge, this is the first study to show an association between elevated PF4-heparin antibodies and Increased mortality rates in hemodlalysis patients.

Economic implications of nighttime attending intensivist coverage in a medical intensive care unit.

Objective: Our objective was to assess the cost implications of changing the intensive care unit staffing model from on-demand presence to mandatory 24-hr in-house critical care specialist presence. Design: A pre-post comparison was undertaken among the prospectively assessed cohorts of patients admitted to our medical intensive care unit 1 yr before and 1 yr after the change. Our data were stratified by Acute Physiology and Chronic Health Evaluation III quartile and whether a patient was admitted during the day or at night. Costs were modeled using a generalized linear model with log-link and &ggr;-distributed errors. Setting: A large academic center in the Midwest. Patients: All patients admitted to the adult medical intensive care unit on or after January 1, 2005 and discharged on or before December 31, 2006. Patients receiving care under both staffing models were excluded. Intervention: Changing the intensive care unit staffing model from on-demand presence to mandatory 24-hr in-house critical care specialist presence. Measurements and Main Results: Total cost estimates of hospitalization were calculated for each patient starting from the day of intensive care unit admission to the day of hospital discharge. Adjusted mean total cost estimates were 61% lower in the post period relative to the pre period for patients admitted during night hours (7 pm to 7 am) who were in the highest Acute Physiology and Chronic Health Evaluation III quartile. No significant differences were seen at other severity levels. The unadjusted intensive care unit length of stay fell in the post period relative to the pre period (3.5 vs. 4.8) with no change in non-intensive care unit length of stay. Conclusions: We find that 24-hr intensive care unit intensivist staffing reduces lengths of stay and cost estimates for the sickest patients admitted at night. The costs of introducing such a staffing model need to be weighed against the potential total savings generated for such patients in smaller intensive care units, especially ones that predominantly care for lower-acuity patients.

Widely used track and trigger scores: Are they ready for automation in practice?

INTRODUCTION Early Warning Scores (EWS) are widely used for early recognition of patient deterioration. Automated alarm/alerts have been recommended as a desirable characteristic for detection systems of patient deterioration. We undertook a comparative analysis of performance characteristics of common EWS methods to assess how they would function if automated. METHODS We evaluated the most widely used EWS systems (MEWS, SEWS, GMEWS, Worthing, ViEWS and NEWS) and the Rapid Response Team (RRT) activation criteria in use in our institution. We compared their ability to predict the composite outcome of Resuscitation call, RRS activation or unplanned transfer to the ICU, in a time-dependent manner (3, 8, 12, 24 and 36 h after the observation) by determining the sensitivity, specificity and positive predictive values (PPV). We used a large vital signs database (6,948,689 unique time points) from 34,898 unique consecutive hospitalized patients. RESULTS PPVs ranged from less than 0.01 (Worthing, 3 h) to 0.21 (GMEWS, 36 h). Sensitivity ranged from 0.07 (GMEWS, 3 h) to 0.75 (ViEWS, 36 h). Used in an automated fashion, these would correspond to 1040-215,020 false positive alerts per year. CONCLUSIONS When the evaluation is performed in a time-sensitive manner, the most widely used weighted track-and-trigger scores do not offer good predictive capabilities for use as criteria for an automated alarm system. For the implementation of an automated alarm system, better criteria need to be developed and validated before implementation.

Association between value-based purchasing score and hospital characteristics

BackgroundMedicare hospital Value-based purchasing (VBP) program that links Medicare payments to quality of care will become effective from 2013. It is unclear whether specific hospital characteristics are associated with a hospital’s VBP score, and consequently incentive payments.The objective of the study was to assess the association of hospital characteristics with (i) the mean VBP score, and (ii) specific percentiles of the VBP score distribution. The secondary objective was to quantify the associations of hospital characteristics with the VBP score components: clinical process of care (CPC) score and patient satisfaction score.MethodsObservational analysis that used data from three sources: Medicare Hospital Compare Database, American Hospital Association 2010 Annual Survey and Medicare Impact File. The final study sample included 2,491 U.S. acute care hospitals eligible for the VBP program. The association of hospital characteristics with the mean VBP score and specific VBP score percentiles were assessed by ordinary least square (OLS) regression and quantile regression (QR), respectively.ResultsVBP score had substantial variations, with mean score of 30 and 60 in the first and fourth quartiles of the VBP score distribution. For-profit status (vs. non-profit), smaller bed size (vs. 100–199 beds), East South Central region (vs. New England region) and the report of specific CPC measures (discharge instructions, timely provision of antibiotics and beta blockers, and serum glucose controls in cardiac surgery patients) were positively associated with mean VBP scores (p<0.01 in all). Total number of CPC measures reported, bed size of 400–499 (vs. 100–199 beds), a few geographic regions (Mid-Atlantic, West North Central, Mountain and Pacific) compared to the New England region were negatively associated with mean VBP score (p<0.01 in all). Disproportionate share index, proportion of Medicare and Medicaid days to total inpatient days had significant (p<0.01) but small effects. QR results indicate evidence of differential effects of some of the hospital characteristics across low-, medium- and high-quality providers.ConclusionsAlthough hospitals serving the poor and the elderly are more likely to score lower under the VBP program, the correlation appears small. Profit status, geographic regions, number and type of CPC measures reported explain the most variation among scores.

Do pre-existing complications affect the failure to rescue quality measures?

Background A project sponsored by the University Health System Consortium has addressed the inaccuracy and high variability across institutions concerning the use of the failure to rescue (FTR) quality indicator defined by the Agency for Healthcare Research and Quality (AHRQ). Results indicated that of the complications identified by the quality indicator, 29.5% were pre-existing upon hospital admission. Objective The purpose of our study was to investigate the possible bias to FTR measures by including cases of complications that were pre-existing at admission. Methods Hospital discharges between 1 January 1996 and 30 September 2007 were retrospectively gathered from administrative databases. Using definitions outlined by the AHRQ and the National Quality Forum (NQF), FTR rates were calculated. Using present on admission coding, FTR rates were recalculated to differentiate between the rates of pre-existing and that of acquired cases. Results Using the AHRQ definition, the overall FTR rate was 11.60%. The FTR rate for patients with pre-existing complications was 8.85%, whereas patients with complications acquired during hospitalisation had an FTR rate of 18.46% (p<0.001). The NQF FTR rate was 9.93%. Pre-existing and acquired FTR rates using the NQF measure were 9.42% and 12.77%, respectively (p<0.001). Conclusions Current definitions of FTR measures meant to identify inhospital complications appear biased by the inclusion of problems at admission. Furthermore, many patients with these complications are excluded from the algorithms. When taking into account the timing of the “complications”, these measures can be useful for internal quality control. However, it should be stressed that the usefulness of the measures to compare institutions will be dependent on coding practices of institutions. Validation using chart review may be required.

Evaluating implementation of a rapid response team: Considering alternative outcome measures

OBJECTIVE Determine the prolonged effect of rapid response team (RRT) implementation on failure to rescue (FTR). DESIGN Longitudinal study of institutional performance with control charts and Bayesian change point (BCP) analysis. SETTING Two academic hospitals in Midwest, USA. PARTICIPANTS All inpatients discharged between 1 September 2005 and 31 December 2010. INTERVENTION Implementation of an RRT serving the Mayo Clinic Rochester system was phased in for all inpatient services beginning in September 2006 and was completed in February 2008. MAIN OUTCOME MEASURE Modified version of the AHRQ FTR measure, which identifies hospital mortalities among medical and surgical patients with specified in-hospital complications. RESULTS A decrease in FTR, as well as an increase in the unplanned ICU transfer rate, occurred in the second-year post-RRT implementation coinciding with an increase in RRT calls per month. No significant decreases were observed pre- and post-implementation for cardiopulmonary resuscitation events or overall mortality. A significant decrease in mortality among non-ICU discharges was identified by control charts, although this finding was not detected by BCP or pre- vs. post-analyses. CONCLUSIONS Reduction in the FTR rate was associated with a substantial increase in the number of RRT calls. Effects of RRT may not be seen until RRT calls reach a sufficient threshold. FTR rate may be better at capturing the effect of RRT implementation than the rate of cardiac arrests. These results support prior reports that short-term studies may underestimate the impact of RRT systems, and support the need for ongoing monitoring and assessment of outcomes to facilitate best resource utilization.

Automated detection of follow-up appointments using text mining of discharge records

OBJECTIVE To determine whether text mining can accurately detect specific follow-up appointment criteria in free-text hospital discharge records. DESIGN Cross-sectional study. SETTING Mayo Clinic Rochester hospitals. PARTICIPANTS Inpatients discharged from general medicine services in 2006 (n = 6481). INTERVENTIONS Textual hospital dismissal summaries were manually reviewed to determine whether the records contained specific follow-up appointment arrangement elements: date, time and either physician or location for an appointment. The data set was evaluated for the same criteria using SAS Text Miner software. The two assessments were compared to determine the accuracy of text mining for detecting records containing follow-up appointment arrangements. MAIN OUTCOME MEASURES Agreement of text-mined appointment findings with gold standard (manual abstraction) including sensitivity, specificity, positive predictive and negative predictive values (PPV and NPV). RESULTS About 55.2% (3576) of discharge records contained all criteria for follow-up appointment arrangements according to the manual review, 3.2% (113) of which were missed through text mining. Text mining incorrectly identified 3.7% (107) follow-up appointments that were not considered valid through manual review. Therefore, the text mining analysis concurred with the manual review in 96.6% of the appointment findings. Overall sensitivity and specificity were 96.8 and 96.3%, respectively; and PPV and NPV were 97.0 and 96.1%, respectively. ANALYSIS of individual appointment criteria resulted in accuracy rates of 93.5% for date, 97.4% for time, 97.5% for physician and 82.9% for location. CONCLUSION Text mining of unstructured hospital dismissal summaries can accurately detect documentation of follow-up appointment arrangement elements, thus saving considerable resources for performance assessment and quality-related research.

Incorporating the Last Four Digits of Social Security Numbers Substantially Improves Linking Patient Data from De‐identified Hospital Claims Databases

OBJECTIVE Assess algorithms for linking patients across de-identified databases without compromising confidentiality. DATA SOURCES/STUDY SETTING Hospital discharges from 11 Mayo Clinic hospitals during January 2008-September 2012 (assessment and validation data). Minnesota death certificates and hospital discharges from 2009 to 2012 for entire state (application data). STUDY DESIGN Cross-sectional assessment of sensitivity and positive predictive value (PPV) for four linking algorithms tested by identifying readmissions and posthospital mortality on the assessment data with application to statewide data. DATA COLLECTION/EXTRACTION METHODS De-identified claims included patient gender, birthdate, and zip code. Assessment records were matched with institutional sources containing unique identifiers and the last four digits of Social Security number (SSNL4). PRINCIPAL FINDINGS Gender, birthdate, and five-digit zip code identified readmissions with a sensitivity of 98.0 percent and a PPV of 97.7 percent and identified postdischarge mortality with 84.4 percent sensitivity and 98.9 percent PPV. Inclusion of SSNL4 produced nearly perfect identification of readmissions and deaths. When applied statewide, regions bordering states with unavailable hospital discharge data had lower rates. CONCLUSION Addition of SSNL4 to administrative data, accompanied by appropriate data use and data release policies, can enable trusted repositories to link data with nearly perfect accuracy without compromising patient confidentiality. States maintaining centralized de-identified databases should add SSNL4 to data specifications.