Framework for Drug Formulary Decision Using Multiple-Criteria Decision Analysis.

Background. Reviewing drugs to determine coverage or reimbursement level is a complex process that involves significant time and expertise. Review boards gather evidence from the submission provided, input from clinicians and patients, and results of clinical and economic reviews. This information consists of assessments on multiple criteria that often conflict with one another. Multiple-criteria decision analysis (MCDA) includes methods to address complex decision making problems with conflicting objectives and criteria. We propose an MCDA approach that infers a utility model based on reviews of previously submitted drugs. Methods. We use a recent extension of the UTilitiés Additives DIScriminantes approach, UTADISGMS. This disaggregation approach deconstructs a portfolio of elements such as a set of drugs that have been reviewed and for which a decision has been made. It derives global and marginal utility functions that are consistent with the preferences exhibited by the review boards in their recommendations. We apply the method to oncology drugs reviewed in Canada between 2011 and 2017. We also illustrate how to conduct scenario analyses and predict the coverage decisions for new drugs. Results. Applying the method yields a utility value for each submission along with a set of thresholds that partition the utility values based on the submission outcomes. Scenario analyses illustrate the predictive ability of the method. Conclusion. Preference disaggregation is an indirect way of eliciting an additive global utility value function. It requires less of a cognitive effort from the decision making bodies because it infers preferences from the data rather than relying on direct assessments of model parameters. We illustrate how it can be applied to validate existing decisions and to predict the recommendation of a new drug.

Emergency department patient flow simulations using spreadsheets.

INTRODUCTION: Patient flow computer simulations allow Emergency Department stakeholders to assess operational interventions, develop utilization and performance measures, and produce estimates for budgeting or planning purposes. Key challenges of traditional discrete-event computer simulation software are their inherent complexity for modeling, coding, or analyzing output and their significant costs and training. We propose a simulation platform that runs in spreadsheets. Because of their low cost, popularity and powerful functionality and performance, spreadsheets also allow for the development and management of simulations that efficiently output results that are just as reliable as those from traditional software. METHODS: A spreadsheet simulation is developed by modeling one row as one simulated minute (more than 20,000 rows for a 2-week period). Uncertainty in arrivals, patient type, routing, and treatment times is modeled using the "rand()" function to simulate the state of the Emergency Department at a given point in time. The patient is tracked with embedded "if()" functions and summary statistics are obtained through range functions. We use an equivalence test to determine whether the resulting average length-of-stay figures are the same as those of a traditional simulation platform. RESULTS: We find little significant difference in average length-of-stay figures between both models. CONCLUSIONS: Spreadsheet simulations are as effective as traditional simulations but easier to use, understand, and implement. Spreadsheet software is widely available, at a fraction of the cost of discrete-event simulation software. Coding spreadsheet simulations may be more challenging as it requires a different and more novel expertise than traditional computer programming. However, spreadsheets can be organized to reference existing datasets, thus minimizing the burden of copying and likelihood of transcription errors and information leakage. Output analysis can also be customized with user-specific performance statistics and charts.

The relationship between inpatient discharge timing and emergency department boarding.

BACKGROUND: Patient crowding and boarding in Emergency Departments (EDs) impair the quality of care as well as patient safety and satisfaction. Improved timing of inpatient discharges could positively affect ED boarding, and this hypothesis can be tested with computer modeling. STUDY OBJECTIVE: Modeling enables analysis of the impact of inpatient discharge timing on ED boarding. Three policies were tested: a sensitivity analysis on shifting the timing of current discharge practices earlier; discharging 75% of inpatients by 12:00 noon; and discharging all inpatients between 8:00 a.m. and 4:00 p.m. METHODS: A cross-sectional computer modeling analysis was conducted of inpatient admissions and discharges on weekdays in September 2007. A model of patient flow streams into and out of inpatient beds with an output of ED admitted patient boarding hours was created to analyze the three policies. RESULTS: A mean of 38.8 ED patients, 22.7 surgical patients, and 19.5 intensive care unit transfers were admitted to inpatient beds, and 81.1 inpatients were discharged daily on September 2007 weekdays: 70.5%, 85.6%, 82.8%, and 88.0%, respectively, occurred between noon and midnight. In the model base case, total daily admitted patient boarding hours were 77.0 per day; the sensitivity analysis showed that shifting the peak inpatient discharge time 4h earlier eliminated ED boarding, and discharging 75% of inpatients by noon and discharging all inpatients between 8:00 a.m. and 4:00 p.m. both decreased boarding hours to 3.0. CONCLUSION: Timing of inpatient discharges had an impact on the need to board admitted patients. This model demonstrates the potential to reduce or eliminate ED boarding by improving inpatient discharge timing in anticipation of the daily surge in ED demand for inpatient beds.

How process analysis could improve the implementation of spinal cord stimulation treatment for chronic pain.

SUMMARY Spinal cord stimulation has been in clinical use for the treatment of chronic pain for over four decades. Since the initial use by Norman Shealy, the indications for its use have increased steadily over the decades to include neuropathic pain owing to failed back surgery syndrome, complex regional pain syndrome and painful diabetic peripheral neuropathies. To date, the precise mechanism of action of spinal cord stimulation remains unclear, yet it is still one of the most expensive interventional treatment modalities available in pain medicine with increasing application across the world. Given the worldwide focus on cost-effective care, there is an opportunity to focus on process analysis as a mechanism for optimizing the operations within and between all specialties engaged in the provision of care in pain medicine. Here, we propose a process analysis approach to model, measure and improve the delivery of disease-based care to enhance effective treatment with a costly modality. Systems-based process analysis is not widely utilized in pain medicine, and there is a limited body of evidence for its application. The purpose of this article is to generate interest in the discipline of process analysis in pain medicine, as it has found value in other healthcare settings and industries. We mention the applicability across countries and specialties that we hope will increase the awareness of this concept and possibly generate interest in further examination by investigators that will lead to the development of highly efficient and effective healthcare delivery processes and systems across the globe.

Adding more beds to the emergency department or reducing admitted patient boarding times: which has a more significant influence on emergency department congestion?

STUDY OBJECTIVE: We evaluate a computer simulation model designed to assess the effect on emergency department (ED) length of stay of varying the number of ED beds or altering the interval of admitted patient departure from the ED. METHODS: We created a computer simulation model (Med Model) based on institutional data and augmented by expert estimates and assumptions. We evaluated simulations of increasing the number of ED beds, increasing the admitted patient departure and increasing ED census, analyzing potential effects on overall ED length of stay. Multiple sensitivity analyses tested the robustness of the results to changes in model assumptions and institutional data. RESULTS: With a constant ED departure rate at the base case and increasing ED beds, there is an increase in mean length of stay from 240 to 247 minutes (95% confidence interval [CI] 0.8 to 12.6 minutes). When keeping the number of beds constant at the base case and increasing the rate at which admitted patients depart the ED to their inpatient bed, the mean overall ED length of stay decreases from 240 to 218 minutes (95% CI 16.8 to 26.2 minutes). With a 15% increase in daily census, the trends are similar to the base case results. The sensitivity analyses reveal that despite a wide range of inputs, there are no differences from the base case. CONCLUSION: Our computer simulation modeled that improving the rate at which admitted patients depart the ED produced an improvement in overall ED length of stay, whereas increasing the number of ED beds did not.

Simple standardized patient handoff system that increases accuracy and completeness.

PURPOSE: The Joint Commission on Accreditation of Healthcare Organizations (JCAHO) defines a "handoff" as a contemporaneous, interactive process of passing patient-specific information from one caregiver to another for the purpose of ensuring the continuity and safety of patient care. The purpose of this study was to conduct a comprehensive investigation on the determinants of an effective handoff management system. Specifically, we sought to address the following null hypotheses: There is no difference before and after implementation of a new, low-cost, low-tech process for surgery patient handoffs in accuracy of information, completeness, clarity of exact time of patient transfer, and number of tasks appropriately handed off. METHODS: Baseline description of the handoff process was mapped from 3 direct observation sessions by an efficiency operations team. A focus group with residents, nurses, hospital administrators, and surgeons was held to identify concerns with the baseline process and to identify important features of a handoff system. These data were used to create an electronic survey for residents to indicate level of agreement with importance of various features and qualities of a handoff system. Longitudinal telephone surveys were performed with residents throughout and after the development period to determine the residents' perceptions of the completeness, accuracy, clarity of handoff time, and method of information transfer, as well as the frequency with which residents were expected to perform tasks that should have been performed by outgoing residents. An online survey was sent to residents before and after the new handoff system was implemented to study perceptions of information quality, process operations, clarity of responsibility, and satisfaction with the handoff process. Perceptions were rated on operationally defined scales. All instruments underwent expert review for content validity and clarity of instructions and scale definition appropriateness. A standardized, and partially automated, handoff form was then developed. After a 2-week pilot study, telephone surveys were repeated. Data were analyzed using descriptive statistics, the Student t-test, and multivariate analysis. RESULTS: Compared with baseline, residents reported increased accuracy, as measured by the perceived number of inaccuracies found on sign-out sheets (p = 0.003). Completeness of the information on sign-out sheets also was improved (p = 0.015). Clarity as to the time of transfer of care from outgoing (day team) to incoming (night float) improved (p = 0.0001). The type of rotation (intensive care unit vs non-intensive care unit) did lead to an improvement (confidence interval< 99%). Across both shifts, the perceived number of inappropriate tasks transferred decreased significantly. Experience (months of training) and type of rotation did not affect these measures. CONCLUSIONS: By simplifying and standardizing the handoff instrument, we demonstrated improvements in resident perceptions of accuracy, completeness, and number of tasks transferred. This low-cost, low-tech paradigm may be useful to others.

Analysis of factors influencing length of stay in the emergency department.

OBJECTIVES: Length of stay (LOS) is a key measure of emergency department (ED) throughput and a marker of overcrowding. Time studies that assess key ED processes will help clarify the causes of patient care delays and prolonged LOS. The objectives of this study were to identify and quantify the principal ED patient care time intervals, and to measure the impact of important service processes (laboratory testing, imaging and consultation) on LOS for patients in different triage levels. METHODS: In this retrospective review, conducted at a large urban tertiary care teaching hospital and trauma centre, investigators reviewed the records of 1047 consecutive patients treated during a continuous 7-day period in January 1999. Key data were recorded, including patient characteristics, ED process times, tests performed, consultations and overall ED LOS. Of the 1047 patient records, 153 (14.6%) were excluded from detailed analysis because of incomplete documentation. Process times were determined and stratified by triage level, using the Canadian Emergency Department Triage and Acuity Scale (CTAS). Multiple linear regression analysis was performed to determine which factors were most strongly associated with prolonged LOS. RESULTS: Patients in intermediate triage Levels III and IV generally had the longest waiting times to nurse and physician assessment, and the longest ED lengths of stay. CTAS triage levels predicted laboratory and imaging utilization as well as consultation rate. The use of diagnostic imaging and laboratory tests was associated with longer LOS, varying with the specific tests ordered. Specialty consultation was also associated with prolonged LOS, and this effect was highly variable depending on the service consulted. CONCLUSIONS: Triage level, investigations and consultations are important independent variables that influence ED LOS. Future research is necessary to determine how these and other factors can be incorporated into a model for predicting LOS. Improved information systems will facilitate similar ED time studies to assess key processes, lengths of stay and clinical efficiency.