SYSTEMS AND METHODS FOR VIEWING BIOMETRICAL INFORMATION AND DYNAMICALLY ADAPTING SCHEDULE AND PROCESS INTERDEPENDENCIES WITH CLINICAL PROCESS DECISIONING
Certain embodiments provide a method for viewing biometrical information, determining the clinical state of the patient, and dynamically adapting schedule and process interdependencies in a clinical environment. The method includes deriving a schedule including at least one patient treatment protocol based on a mathematical model using people and asset information as input. The method also includes processing input biometrical data indicating a patient status. Additionally, the method includes adapting the at least one patient treatment protocol based on the biometrical data. Further, the method includes revising the schedule based on adapting the at least one patient treatment protocol. In certain embodiments, the method includes recalculating a critical path, schedule tasks, and resource allocation based on adapting the at least one patient treatment protocol and revising the schedule. In certain embodiments, the method includes communicating with affected process stakeholders and assets regarding the revised schedule.
This application is related to and claims the benefit of priority to Provisional Application U.S. Ser. No. 60/976,582, entitled “METHOD TO VIEW BIOMETRICAL INFORMATION AND DYNAMICALLY ADAPT SCHEDULE AND PROCESS INTERDEPENDENCIES WITH CLINICAL PROCESS DECISIONING”, filed Oct. 1, 2007, the contents of which are both herein incorporated by reference.
BACKGROUND TO THE INVENTIONThe invention relates generally to business process management systems, and more particularly to scheduling systems in the clinical setting, such as healthcare delivery institutions or hospitals.
Healthcare delivery institutions are business systems that can be designed and operated to achieve their stated missions robustly. As is the case with other business systems such as those designed to provide services and manufactured goods, there are benefits to managing variation such that the stake-holders within these business systems can focus more fully on the value added core processes that achieve the stated mission and less on activity responding to variations such as emergency procedures, regular medical interventions, delays, accelerations, backups, underutilized assets, unplanned overtime by staff and stock outs of material, equipment, people and space that are impacted in the course of delivering healthcare.
Currently clinical process decisions have historically relied on the art of understanding symptoms and diagnosing causality much in alignment with the practice of the medical diagnosis arts. In an ever-evolving environment, judgment and experientially-developed mental models are utilized by the healthcare providers to utilize the information currently at hand to make decisions. Presented with similar data, the decision made from one caregiver to another typically exhibits a variation. Presented with partial information, which is the byproduct of being organized in functional departments, specialties, roles and by the nature of having partial and/or current or dated information availability on hand—clinical process decisions vary widely and typically are locally focused for lack of a systems view upstream and downstream of the decision point.
As a hospital processes care plans on an increasing patient load, these variations in medical condition and selected treatment plans perturbs the schedules of doctors, nurses and assets such as rooms and equipment. If there is protective capacity in these schedules and staff, the providers of care can manage variation while maintaining care quality. When randomness and interdependencies exceed the ability to serve, care providers are forced to make choices amongst poor alternative options; some one or some thing is going to be bottlenecked or overextended. Delays, queues, overtime, burnout and emotional decision making characterize systems that are over-taxed or beyond their ability to perform.
Where information systems exist, they are simply informational in nature. Examples include scheduled rooms, people, materials and equipment. Recent advances in locating devices such as those utilizing radio-frequency identification (RFID) technology to report a location of a tagged asset are utilized, yet personnel are loath to be tracked by wearing a device. RFID devices are not pervasive, and the systems that monitor them are not fully integrated with the requisite adjacent systems that gather contextual information. The current art is not predictive, probabilistic nor necessarily systemic. For example, knowing the location of an asset is desirable but knowing its anticipated need and interdependencies is required to make a decision to use a located asset actionable. The information required for such a decision comes from a plurality of adjacent health information systems and must have an ability to play forward into the future.
Today, current procedure duration and room status is provided without any proactive or forward-looking capability. Schedules are produced before a day's activities commence. Process status is displayed along with trending and, often, alarm functionality should a process variable trip a threshold set point. Today, processes are planned for a given volume; when that volume is exceeded or processes have sufficient variation to overtax their capability, scheduling and recovery are reduced to manual triage and experience to sort out. Typically, queues, delay, overtime and cancellation result; there is no proactive decision support to dynamically reschedule people or physical assets or supplies.
There is therefore a need for an integrated system and method for scheduling clinical activities and procedures in real time that incorporate variation, biometrical changes in the state of patient health, changes in process necessitated by protocols which must be executed that are different than what was originally scheduled, staff and equipment preferences, interdependencies and information flow into the clinical delivery of healthcare that can “look ahead” and enable “what-if” capability for prospective decision support, given the changes that are occurring, and provide viable ways going forward that reduce overall negative system impact.
SUMMARY OF THE INVENTIONCertain embodiments provide a method for viewing biometrical information and dynamically adapting schedule and process interdependencies in a clinical environment. The method includes deriving a schedule including at least one patient treatment protocol based on a mathematical model using people and asset information as input. The method also includes processing input biometrical data indicating a patient status. Additionally, the method includes adapting the at least one patient treatment protocol based on the biometrical data. Further, the method includes revising the schedule based on adapting the at least one patient treatment protocol.
In certain embodiments, the method includes recalculating a critical path, schedule tasks, and resource allocation based on adapting the at least one patient treatment protocol and revising the schedule. In certain embodiments, the method includes communicating with affected process stakeholders and assets regarding the revised schedule.
Certain embodiments provide a scheduling system for viewing biometrical information and dynamically adapting schedule and process interdependencies in a clinical environment. The system includes a scheduling module deriving a schedule including at least one patient treatment protocol based on a mathematical model using people and asset information as input. The system also includes an input module processing input biometrical data indicating a patient status. The scheduling module adapts the at least one patient treatment protocol based on the biometrical data and revises the schedule based on adapting the at least one patient treatment protocol. Certain embodiments provide a computer-readable medium having a set of instructions for execution on a computer. The set of instructions includes a scheduling routine deriving a schedule including at least one patient treatment protocol based on a mathematical model using people and asset information as input. The set of instructions also includes an input routine processing input biometrical data indicating a patient status. The scheduling routine adapts the at least one patient treatment protocol based on the biometrical data and revises the schedule based on adapting the at least one patient treatment protocol.
These and other features, aspects and advantages of the present invention will become better understood when the following detailed description is read with reference to the accompanying drawings in which like characters represent like part throughout the drawings. The embodiments shown in the drawings are presented for purposes of illustration only. It should be understood, however, that the present invention is not limited to the arrangements and instrumentality shown in the attached drawings.
Certain embodiments provide systems and methods for systemically organizing tasks and assets of a process to more effectively achieve immediate and longer-term macro objectives. In certain embodiments, scheduled tasks are organized using, for example, a critical path method (CPM) and the tasks therein are exposed to durations which are probabilistic and are either within the endogenous variation control of the system or are exogenous factors to which the system must be robust to. Measures of duration, availability and reliability to calculate an enumeration of scenarios in the context of variation is used to determine the probabilities of meeting a selected schedule (schedule risk). The probabilistic measures of duration, availability and reliability are functions of path dependent consumption and utilization decisions that are made to determine the use of the assets of the process. Using a multi-modality simulation methodology, for example, a process transfer function of the probabilistic measures may be derived. It is these estimates of duration, both endogenous and exogenous, that are described by simulation that create the task durations and logic for interdependencies which in turn are used to calculate schedule risk in the CPM method. Certain embodiments provide decision support to help make effective decisions in real time or substantially real time while embedded in a highly variable and interdependent process. These decision support embodiments can be automated or prescriptive to the process stake-holder.
Certain embodiments of the present invention provide systems and methods to manage changes to a schedule to accommodate changes that are internally or externally induced—and to do so in a way that reduces or minimizes overall health delivery system throughput or quality degradation. Certain embodiments of the present invention provide systems and methods to review what has recently happened in a process, to view actual current process operations, and to view what is on the schedule looking forward into the near term future. Specific assets such as plant & equipment, people, physical location and information are exemplary entities being tracked and dynamically managed.
Certain embodiments are adaptable and dynamically configurable such that activities can be rescheduled based upon, among other things, state(s) of patients, providers, and assets within a procedure and/or scheduled for a procedure. Increased adaptability and dynamic configurability help clinical systems and personnel to function more reliably and to be robust to exogenous forces outside of process control such as what symptoms patients present with, time of emergency, volume of patient medical demand, and the like.
Certain embodiments of the present invention reduce process variation and thereby enable staff to focus more on the clinical delivery of healthcare. This is achieved, for example, in four ways—(1) An overarching capability to reduce internal (endogenous) variation from interdependency variation that can be anticipated and subsequently managed by techniques such as in certain embodiments of the presently disclosed technology; (2) To incorporate variation into the process planning and control as far forward into the time line such that not only more accurate averages are used for scheduling, but also the probability ranges of time duration for various activities, availability, capability ranges for staff and equipment are incorporated and planned for such as in certain embodiments of the presently disclosed technology; (3) An overarching capability to combine information flows as to the status of staff, patients, equipment and facilities with the scheduled plan such that anticipatory alerts are provided when schedule risk crosses a threshold as well as a diagnosis as to the cause of the likely or actual source of the deviation that is sufficient and actionable enough for intervention by the staff to resolve or revise a plan for it such as is provided for in part by certain embodiments of the presently disclosed technology; and, (4) An overarching capability to understand and incorporate the effects of external (exogenous) variation resulting from unforecastable events such as, for example, surges, medical reason procedure delay, equipment failure and staff sickness such as is provided for in certain embodiments of the presently disclosed technology.
Certain embodiments address multiple variations including medically induced variations and variations in treatment strategies for medical conditions.
The first example, medically induced variations, occurs in normal events of care, which result from patient changes, or in execution of protocols due to emergencies, for example. Examples include a patient heart rate increasing and a nurse responding by providing medication, as well as a stroke patient arriving in the emergency department who requires a protocol that will need a diagnostic imaging device that is already scheduled for another patient.
The second example, treatment strategy variation, is variation resulting from different interpretation and response among caregivers. Examples of the second type include the historical practice of clinical process decisions in reliance on the art of understanding symptoms and diagnosing causality, much in alignment with the practice of the medical diagnosis arts. In an ever-evolving environment, judgment and experientially developed mental models are utilized by healthcare providers to utilize information currently at hand to make decisions. Presented with similar data, the decision made from one caregiver to another typically exhibits variation. Presented with partial information—which is a byproduct of being organized in functional departments, specialties, roles and by the nature of having partial and/or current or dated information availability on hand—clinical process decisions may vary widely and typically are locally focused for lack of a systems view upstream and downstream of the decision point.
In accordance with a first embodiment of the present invention, a system for use in planning procedures is provided comprising the following components: a duration estimator module configured to characterize average duration times and variations from average duration times for a given procedure or activity; a block allocation planner module configured to schedule procedures or activities in accordance with characterized times from the duration estimator module; and a user interface module configured to permit a user to visualize variation, to visualize scheduling opportunities and constraints and to view information output for use in scheduling procedures and activities. Each component or module will be described in greater detail below. In the exemplary embodiments described below, the given schedule is related to a perioperative environment. However it is to be appreciated that the methods and systems can be extended to other clinical or non-clinical process systems.
The Duration Estimator, which may be thought of as a ‘better ruler to measure with’, measures variation so that procedure times are more accurate and so that schedule risk is reduced.
The Block Allocation Planner may be thought of as a “defragmenter” for reserving blocks of time. Much like a computer hard drive defragmenter, it re-sorts the time slots and rooms to be available for the booking of surgeries in such a way as to satisfy constraints and departmental objectives. It factors in preferences and availabilities and solves for the best departmental allocation of space and time and through the allocation, help achieve department policy objectives such as case mix, outcomes, safety and to provide incentive for desired behaviors.
The user interface to the disclosed schedule decision support may include at least one of modules configured to permit a user to visualize variation, to visualize scheduling opportunities and constraints and provide output for use in scheduling procedures and activities, or combinations thereof.
In accordance with an exemplary embodiment, the scheduling module is configured to aid in initially scheduling activities and procedures and is configured to visualize scheduling opportunities and constraints and will be referred to herein as “Day Planner”. Day Planner may be thought of as a manufacturing requisition planner, it enables the scheduling of specific activities and all interdependencies—with the added benefit of simulating forward many alternative plans and contingencies.
In accordance with an exemplary embodiment, the user interface to the schedule manager as the process is occurring is configured to indicate and display variation along with suggestions as to “do-what” and enable “what-if” and will be referred to herein as “Day View”. Day View may be thought of as a “radar” for the clinical process, it brings schedule with other location and clinical information so that the staff can know when schedule deviations are occurring, what the cause is, have a way to visualize process interdependencies, have the ability to play out or simulate alternative process decisions and ultimately get the process stakeholders constructively involved in proceeding forward in a manner that has their intellectual buy-in to the course.
A user interface module is configured to provide output for use in directing procedures and activities and will be referred to herein as “Provider/Patient Kiosk”. Provider/Patient Kiosk may be thought of as workflow, the kiosk is information gathering and output for the various stakeholders in the clinical process. It gathers information that is later used to reduce variation and it interacts with the stakeholders in mediums that are natural extensions of their native environments. The user interface may include computer display and/or reports.
It is to be appreciated that the clinical process decision support methods may comprise one or more of the capabilities described above, as well as any combination thereof.
Endogenous variation is reduced within clinical process from situations arising from concurrent scheduled use of mutually exclusive resources such as doctors, nurses, equipment and space regardless of the cause being endogenously or exogenously perturbed. Methods according to the present invention move clinical process decisioning as far forward into the care process as possible. Known interdependencies are then brought to each decision point and factored in so that interdependencies are then managed and visible immediately. For example, in the instance of day planning/scheduling done in advance of the day for which activity is scheduled, some interdependencies must be known well before block schedules are derived and patients are scheduled.
Embodiments of the present invention address the immediate, inevitable deviations from plan with diagnoses capability, in near real time, for the causality of endogenously generated variation and provides decision support for prospective process recovery to either the originally proscribed state or a new, superior one, given other operating constraints that render the original plan less advantageous. Exogenous variation—that which happens to the delivery system, despite the care and safety margin in the process scheduling, such as for example unanticipated medical complications, a surge of additional procedure(s) or an unanticipatable staffing change or equipment outage—is addressed by the ability to see the impact of change at or before it is currently possible to see it and then to provide feasible path forward decision support.
Certain embodiments dynamically offer process guidance depending upon the state of the patient and availability and priority of assets, care providers, and other patients. Once a decision support user selects a path, a process of care and affected schedule(s) are updated to effect process change(s) to accommodate new activities as well as best achieve the prior scheduled or occurring activity according to one or more operating objectives or criteria established by the hospital or other clinical facility. Certain embodiments facilitate the coupling of a real or simulated biometrical patient reading to a semantic or structured data architecture and a logical interpretive decision engine. The architecture and engine determine a state of the patient and a proposed protocol to execute, if any. State and protocol information maybe determined via a process model using discrete event-based, agent-based, and/or heuristic logic simulation that models “what could be” to a Gantt/Pert based scheduling engine with both simulated and actual process feedback.
A time constant for detecting change and making decisions is useful in designing out or reducing preventable variation in a system and in making allowances for residual normal variation. A decision making time constant is enabled via one or more of the disclosed methods and systems to fall within a time constant of unrecoverable process oscillation, for example. Where there is no anticipatory forecast possible and the physical state or care of patients, staff, and/or assets necessitates action, certain embodiments offer systems and methods to help achieve department objectives by identifying adjacent activity impacts and offering a feasible way to cover a present excursion of care activity as well as viable path(s) to recover.
Major sources of endogenous variation resulting from interdependencies amongst resources are, for an exemplary embodiment in the perioperative environment, delayed assets of the process due to unanticipated complexities of interdependent activities, equipment breakdown, incomplete preparation, staff sickness, patients not biometrically responding to procedures, patients having medical changes requiring care, operating complexities, and the like.
Certain embodiments include Duration Estimator, Block Allocation Planner, Day View, Day Planner/Day Replay and Provider/Patient Kiosk as elements for information and decisioning designed to reduce systemic variations by eliminating schedule risk up front and enabling rapid on-the-fly response to unanticipated perturbations to the clinical process, for example.
It is advantageous to not only know what is scheduled, but to know that what is scheduled is robust to unplanned events, to have an assessment of the schedule risk, to have the ability to visualize likely process scenarios before hand, to be able to determine robust decisions that can be taken in the dynamic environment that will best achieve the operating objectives, to learn from what transpired, to achieve departmental objectives and to include stakeholders in the clinical process. Each of these advantageous capabilities is advanced with certain embodiments of the present invention.
Referring to
Clinical procedures have vast variation yet the advanced scheduling typically is made at a single prescribed duration time. When procedures take less than the forecast (typically estimated in the prior art as the average for a procedure plus, perhaps, a safety margin), valuable assets are underutilized and clinical flexibility is degraded because the availability is unanticipated/unactionable and further, downstream processes are impacted from patient handling, staffing and equipment turn around. When procedures take more time than forecasted, valuable assets are unavailable for other procedures that were scheduled thus backing up upstream processes, creating staffing shortfalls and fatigue stress from the emotional responses to schedule collapse and recovery.
Estimating the duration of a procedure has a number of difficulties. One forecast difficulty is the variation on single procedure cases that result from doctor/staff combinations, patient disease state severity, patient medical complications and physical attributes such as weight/height/body mass index (BMI)/surface area, information availability, equipment and supply availability, day of week, procedure prior and etc. Another still is wrong or non-recorded information such as, for example, time durations, and multiple procedures conducted with only a subset being recorded as actually having been conducted.
The objective of understanding the variance in procedure duration is so that the schedules may be made more robust. How schedules are made robust will be discussed below with reference to a module referred to as Day Planner. It can be understood what the interactions of factors are on specific procedure combinations or alternatively, to classify combinations of factors into defined duration probability densities. Either the procedure is called in a schedule planning simulation or using the time density functions directly, for example.
Referring to
In the absence of a historical record, duration estimation is achieved via expert input (not shown separately but can be included in history 24). Experts include consensus of relevant professionals collated via industry working groups, societies and academic study, nursing, administrators, anesthesiologists and surgeons. Historical data of recorded procedures and their duration are more desirous than strictly expert opinion. A preliminary analytical step is to characterize the accuracy (mean and statistical variation) of historical procedure duration vs. actual for like cases. This is achieved with a design of experiments on a subset of procedure and/or procedure clustering and then compared to the historical record. Additionally, audits are conducted to compare scheduled vs. recorded as being conducted. Art from the methods of gauge repeatability and reliability are then used to characterize the confidence interval of forecasted duration. Regardless of the accuracy, a measure of duration classification is made and its degree of uncertainty is established.
The history of procedures is combined into a relational database or spreadsheet or analytical platform capable of data mining with data incorporated into a data repository. The objective is to regress potential leading indicators against the historical duration times in an attempt to incorporate significant predictive variables that can then be used at the time of booking to allocate time.
Where it is possible to describe variation with enough confidence, then the regression equation is utilized plus a fraction of the variation to set the duration. In calculating what is ‘enough confidence’, a fraction of the variation is included in addition to the calculated duration using the regression equation.
Enough confidence is a function of the case throughput for a period, typically defined as a day given that there is a slack time off hours for schedule catch-ups to be made day to day. The cumulative probability density function for all cases in an allocated day is set to a level of schedule risk at a given time within the day. This would be expressed as a desire to be, for example, that the caseload can be processed by 7 PM with 90% confidence. A back propagation to the discrete cases is made such that the individual error terms (variation) are incorporated to the calculated duration in fractional steps until such time as the portfolio of cases can meet the daily throughput/risk set points. This may be performed manually or with the assistance of a stochastic optimization.
It can be the case that historical data is insufficient to separate between procedure(s) with statistical significance given the available leading indicators. In this instance, the histogram of all procedures can be partitioned and procedures that cannot be statistically separated may be clustered such that each is associated with a probability density function (PDF) of time duration. The total of the PDFs form the complete enumeration of the historical procedures less holdouts of known data defects
A learning loop 23 is established that finely records attributes and procedure times. As is the case when first classifying duration from the historical record, techniques including Artificial Neural Networks, Multivariate Regression, Analysis of Variance (ANOVA) and Correlation Analysis are used to refine predictive capability and tighten the confidence bounds. Additionally new descriptive attributes may be appended in order to test and improve forecast accuracy.
Referring now to
What does not exist is a method and system to allocate block time in such a way as to simultaneously satisfy stakeholder preferences. Additionally, there is a present need to have available in the market an engine to ‘re-sort’ block allocation from time to time in an institution. Similar to the need for a defragmenter utility for a computer hard drive—healthcare blocks are allocated based upon past, evolutionary activity. Staff, rooms, procedures, patient demographics, clinical specialties and competitors all drive changing assumptions of former time and space allocations.
It is to be appreciated that methods and systems described herein may enable increased throughput, schedule risk reduction, incentive driven policy, preferences and a consistent environment. Hospitals, surgeons, patients and staff are beneficiaries.
The method proceeds as depicted in
Objective decision-making is an explicit benefit of the disclosed system. Having the ability to communicate visually helps with rapid understanding. A ‘cockpit’ describing the historical fact patterns and current operating results helps to facilitate conversation. The use of cockpits has been described in other known business systems. A further extension is depicted in
The block schedule output may take many forms. An example embodiment is
In certain embodiments of the present invention, a user interface module may be provided in order to get information to a user for use in scheduling procedures. In an exemplary embodiment, user interface comprise the ability to monitor activities for a given day, and will be referred herein as “Day View”. Day View may be thought of as a radar for the perioperative environment coupled to a logic engine. More that simply alarming a problem, Day View prospectively assesses interdependencies such as, for example, a certain procedure and a piece of equipment and staff required for the procedure. Should the location and availability not be per the schedule, an alarm is first made as a warning for action. If as the time of need approaches, the interdependent items are located via passive and active identification. If the issue can be resolved such as via a phone call or quick action—an issue was avoided or minimized.
Should the interdependency not be solved, Day View has the capability to simultaneously solve for all other interdependencies and allow a ‘what-if’ scenario testing exercise to occur. The what-if may be manual or automatic.
By having a procedure finish before the schedule, potential scheduling flexibility is lost for other issues that inevitably arise during the day such as staff availability, rooms and equipment constraints. Potential throughput capacity might also be impacted in that over the course of a day, one or more additional procedures could be inserted into the schedule.
A longer than anticipated procedure can have the effect of backing up other dependent procedures for a particular patient either in the original department or cross departmentally. Additionally, the rooms and assets that were originally allocated are no longer as available and therefore other patients and activities are negatively impacted. There may be unscheduled overtime of staff or extra costs associated with turning around apparatus. New schedules must be made that impacts reworking of staff planning as well as the potential for rework on case preparation for following or interdependent procedures. The anticipated and/or needed throughput capacity may also be impacted negatively. These changes reduce operating flexibility and increase the anxiety of the clinical process stakeholders—often leading to what is commonly known as ‘burn-out’ and attrition, as well as to the potential ultimate quality of delivered healthcare.
Referring now to
Certain embodiments provide a proactive application looking ahead at procedure scheduling and duration to avoid delays by triggering an advanced warning with sufficient time to respond in an event that scheduled procedures will start or end before or after their scheduled time. Certain embodiments provide recommendation regarding one or more specific decision(s) or action(s) can be taken to add, drop or move specific cases, task and assets, for example.
In certain embodiments, scheduling is provided that accounts for available personnel, available physical space, and available blocks of time, for example. Certain embodiments account for variations associated with fluctuating demand and asset availability.
In certain embodiments, multiple criteria may be set for a scheduling process. Additionally, various risks associated with selected criteria may be taken into account. Furthermore, rather than providing a single schedule that is designed with average or generic or judgmental time buffers to compensate for variation, certain embodiments enable a plurality of scheduling scenarios to be manually or dynamically entered or simulated automatically to explore an available solution space and ramifications on current and future activities. Certain embodiments provide suggested decisions calculated to help meet one or more static, dynamic or path-dependent configurable objectives.
In certain embodiments, a scheduler provides a capability to view a single or multiple process metric, agent, and/or asset of the process as well as a dynamic impact on interdependencies such that one or more causes of variation in that process may be explicitly communicated and understood. Certain embodiments enable a dynamic view to process risk along one or more dimensions in real time (or substantially real time with some inherent system delay) or historically, for example. Additionally, certain embodiments may prompt an alarm, action and/or warning if a certain process variable trips a threshold set point, for example.
In certain embodiments, multiple simulation modalities are employed including a critical path method coupled to discrete event, agent, Monte Carlo and/or continuous simulation. Using this coupling, one or more objectives of the process may be assessed.
In certain embodiments, Day View or other user scheduler interface provides features for historical review such as replaying a day or past several days in order to extract from and discover process dynamics, training, and knowledge capture for future use as well as for administrative activity cost(s), protocol verification and billing, for example.
Utilization and asset consumption may be viewed to help understand a state of dynamic interdependencies between scheduling processes and to help determine which a schedule is likely to be met. If a determination is made that a schedule is not likely to be met, data may be viewed and/or used to help identify what assets and interdependencies are causes of schedule variance.
A future schedule view may be provided to calculate “what-if” scenario testing to help understand schedule changes and effects of endogenous variation, such as schedule adds or drops, resource availability, unforeseen delays or failures, etc. Future schedule extrapolation helps to enable a higher process entitlement via better decisions that directly and indirectly affect variation and throughput, wait times, stocks, capability and uses of assets in procedure and resource scheduling.
Thus certain embodiments provide scheduling of processes in highly dynamic environments where knowledge workers and service providers are integral agents of the process, rather than providing singular or discrete schedules with an objective and buffers allocated by judgment or heuristics alone.
While many processes benefit from embodiments disclosed herein, an example embodiment involving a hospital perioperative clinical healthcare delivery process is used for purposes of illustration only.
In an exemplary clinical process, a schedule may be derived among available assets with well-understood interdependencies. For example, a doctor cannot be in two locations at once, and a surgery may not progress unless, concurrently, the appropriate patient, doctors, nurses, information and equipment are in an operating environment suited for the procedure. Additionally, the biometrical state of the patient should be adequate for a surgery to proceed.
As shown in the flow of
In certain embodiments, a determination of an initial view of a schedule may be prefaced by a sequence of analytical work. Referring to
Responsible scheduling is considered to include an estimate of duration and a block of time allocation within which the procedure is considered likely to finish. While under-scheduling procedure time creates delays in subsequent procedure starts, over-estimating and blocking time for available assets may create under-utilized capacity.
When average duration forecasts are used in a clinical environment, and each room is considered unique outside the context of the staff's before and after tasks, chaos often results from concurrent demands. Utilizing probability density functions 750 of time for a given duration estimation of a surgery is a foundation for calculating a schedule's risk and system level performance or optimization, for example.
Referring to
A schedule risk is calculated by integrating the PDF using statistical techniques to create a cumulative probability density function for duration CPDF 760 where a probability of duration is made available. Given sufficient time 763, there is a 100% chance of the forecast duration matching the actual procedure duration. Likewise, given no time 762 there would be no probability of completing the surgery within the time interval. The expected time to complete similar tasks are found at the 50% probability 764 and is Texpected 761, which is the time where it is 50% likely the procedure will finish sooner and 50% likely that it will take longer than this time value. The scale of time zero 762 and maximum 763 is continuous and directly related to probability zero 766 to 100% 765 according to the PDF cumulative probability density function transform.
A risk that a procedure will not finish within its allotted time is determined by a probability 760 at the time allocation scheduled for the procedure. Again referring to
A lower specification limit, LSL 745, and upper specification limit, USL 746, are similarly available. The LSL 745 is a minimum probability threshold (the most schedule risk) of completing the procedure within the allocated time. The USL 746 is a degree of what is typically referred to as “cushion” or “margin” in that it is the adjustable surplus probability over an expected duration probability 764. The LSL 745 and USL 746 are adjustable parameters and may be used as logical rules in decision support 715 available to manual or automated scheduling. The set points may be tuned in a learning feedback loop that uses observed historical process data 862, learns probability set points 742, 745, 746, 761 to optimize throughput and cost and offers optimized time allocation and probabilistic set point decision support.
Certain embodiments facilitate dynamic, intelligent schedule change based on changes in the actual stochastic and interdependent processes of care occurring in the hospital. This method requires forecast durations of procedures arranged within a schedule along with interdependencies of space, people, equipment, consumables and information (e.g., 700, 701, 702, 703). Actual process feedback is provided such as from HIS, RFID, Optical recognition, telemetry and various clinical systems. An explicit mapping of interdependencies in process assets, such as those exemplified in
Examples of events necessitating modification to the schedule 703 include staff and equipment unavailability 738, upstream or downstream processes not able to provide or receive patients 736, devices needed in the scheduled tasks not functioning 735, people and equipment not in planned location 733, 734, inputs from clinical or administrative systems not adequate 732, patient biomedical adequacy or health status not within specification 731, added procedures 721 not in the schedule 703, and dropped procedures 722 for any reason. The schedule may be modified by changing assumptions 716 in the activities 700, 701, 702, 703 used to create the schedule or dynamically managed in the Day View system 704. The changes to assumptions 716 may be manual or computer generated to exploit feasible solutions.
In an environment involving many interdependencies, variation in any interdependent factor may propagate process disturbance. Certain embodiments help facilitate understanding and proactive management of factors that, if delayed or accelerated from plan, will likely increase the probability of delay and preclude a process operating objective from being met.
Returning to
In accordance with
The dynamic Day View 800 is constructed via the execution of analytical tasks as depicted in
As illustrated in
General agreement should exist between the schedule 801 and protocol 802. Referring to
A third of four uploads of data to establish the initial schedule Day View includes a Decision Support rule set 803. The rule set 803 encapsulates the logic specific to protocol and process. An illustrative example relates to a specialized surgical tool. A sterilized probe should be available in a surgical case cart. If a probe is not available, the surgery may proceed if there are two probes on standby on the preoperative floor. Logic 803, 715 may be rule-based, example-based, evidential reasoning, fuzzy logic-based, case-based, and/or other artificial intelligence-based logic, for example.
A fourth upload of data is the duration estimation PDF portfolio. Alternatively, rather than replicate an instance of this knowledge base 804, the analytical workflow 805 may have access to the Duration Classification Engine 700 deployed in the scheduling operation.
Known additions and deletions 807 are made to the initial schedule 806. Illustrative examples include a request to add an emergency surgery to a day, a staff person calls in sick, etc. Timing of this activity 807 has maximum utility if done before the start of scheduled activities and sufficiently in advance to allow time for staff review of the upcoming tasks 808.
A review of an upcoming period's process task, schedule risks and contingency plans is beneficial to provide a contextual understanding of activities as well as to solicit opinions of staff to then modify the schedule for improvement. It may also be beneficial for individual and team preparation and training, for example. At any point, schedule tasks and assets may be simulated forward. Utility may be gained in shift change handovers, for example.
At the start of scheduled activities, variations may begin to occur. Certain embodiments of the present invention help to manage variation systemically in such a way as to meet the objective of a schedule without needless non-value-added activity, such as efforts to complete actions at a rate impacting a critical path for a schedule.
Day View manages a scheduling critical path by monitoring factors that will impact the critical path. If scheduling assets are in designated location(s) and state(s) of availability, a schedule can execute as planned. If a state of an asset in a process or an activity of the process is delayed or accelerated from its designated time, place and state, a chain of dependencies may be impacted. Day View instantaneously calculates durations and the scheduled risks 706 and alarms at probabilistic set points 811, 813, 815.
Asset states 810 are relevant degrees to which an asset is available or biometrical states of a patient. As an illustrative example, a certain piece of equipment is an asset in a schedule process, and a desired state of that asset at a beginning of a designated surgical procedure is to be calibrated and sterile. The state of the equipment is according to plan if each of the requirements are met. The state of the monitored assets 810 is dynamically assessed, for example.
Asset location 820 is monitored as well. For example, Day View determines the process implication of an asset's availability at a designated location at a designated time. Extending the example used in asset state 810, a specialized piece of equipment required for a surgery must be in the operating room. A Boolean indicator of whether an asset is or is not in location as well as a degree or rate of being in the designated location in a future time may be provided, for example.
An illustrative example is an operation beginning in one hour. Specialized equipment is required for the operation at that time. However, the equipment is currently being sterilized. Based on timing information, the equipment's estimated time to sterilization is deemed sufficient to become sterilized and transported to the designated room within an hour.
An asset's dynamic metrics 830 may also be used by the Day View system, for example. Examples of asset metrics include an oxygen saturation from pulse oximeters, status codes from equipment, consumption profiles dispensed in anesthesia delivery, etc. Day View utilizes this information for availability and anticipatory forecasting, for example. An example of anticipatory forecasting is use of biometrical data from an operating room to assess progress relative to a protocol in order to assess a rate as in the asset location 820 illustrative examples discussed above. For example, anticipatory forecasting may be used to determine if a surgery with complete within the next hour.
In another example, pulse oximeter data, captured by a semantic or structured table data infrastructure, is monitored by a reasoning engine. If the readings indicate that care is required, the disclosed system provides the relationships and interdependencies of the process at a system and granular level. Where care providers are required for the new prioritized activity, the potential schedule changes are tested with simulation or heuristics. Feasible solutions are presented and new dynamic schedules are produced and flowed down to the clinical workflow control.
A dynamic view of a process and status of interdependencies is provided 840 with a comparison of what should be happening versus what is happening, what is about to happen, and/or what is scheduled to happen, for example. Enabling assets that are missing through location, rate, and/or state are highlighted to process stakeholders in the form of “satisfactory”, “warning” (pre-emptive action is required) or “danger” (schedule will not be met and alternations to it are required), for example. While “satisfactory” “warning” and “danger” are utilized to illustrate the principle, other context appropriate designators may be used. For example, a change in biometrical state of a patient and/or a patient presents with certain symptoms needing certain care protocols may be highlighted to process stakeholders.
A plurality of modes may be used to view a scheduling process in accordance with embodiments of the present invention. Four primary viewing modes including “what-was”, “what-is”, “what-will-be”, “what-if”. “What-was” 850 is a replay of activities, interdependences and risk levels as the process actually experienced. “What-is 860 is a view into what is happening in the process at the time of being viewed along with the current interdependences, states of assets, people and process risk. “What will be” 870 is the process in the future if the trending of “what-is” remains. “What-is” may be analogous to a radar on a vessel looking sufficiently far enough out over the horizon such that corrective action may be taken. “What-if” 880 is a scenario-based view to aid decision support. Alternatively, “What-if” can be retrospective to evaluate the likely outcomes had a “what-if” decision been made instead of the actual historical one.
In each of the four views, an objective is to have better process outcomes via reduction of controllable variation. Decision support may be provided to process stakeholders through the views, for example.
An analytical roadmap or workflow to build decision support using Day View is depicted in
Key interdependencies are identified around consumption and utilization of assets (see, e.g.,
For assets most likely to become on critical path and potentially delay or cause bottlenecks in a critical path within the forecast horizon, monitoring is enabled 904 along metrics to measure availability, readiness and state, for example. An actual critical path is calculated 906 knowing the structure of interdependencies and state of the assets in the process using Gantt and CPM methods, for example.
Certain embodiments help provide a degree of robustness or tolerance of a schedule to variation. PDF's of endogenous variation from equipment availability and reliability of staffing and patients 908 are used with probabilistic duration estimates 910 and exogenous variation 920, such as adds and actual or likely drops to the schedule as well as historical volatility (e.g., factoring for seasonal variation and capacity) are used in a multi-modality simulation 940 to calculate duration probabilities of scheduled tasks being completed. The simulation engine 940 is repurposed for decision support to run “what-if” scenarios, for example.
In an exemplary embodiment, a clinical process (
Patients arrive 1101 through an emergency department 1102, or walk in to receive emergency or scheduled services 1103. Live patients are discharged after clinical services are rendered 1109 and leave the hospital 1110. Within the clinical process, many departments provide clinical services and each benefit from Day View scheduling, for example. For illustrative simplification, a clinical workflow in and supporting a hospital preoperative department 1122 is considered.
Patients having surgery are either scheduled or are presented for surgery in an emergency. Scheduling of surgical services is typically made between a surgeon's or physician's office and a hospital according a surgeon's allocated blocks of time 701. Surgical procedures are scheduled days, weeks or even months in advance. Emergency procedures are not scheduled and, if reserve capacity is now available, emergency add-ons to a schedule constitute a significant source of variation. It is not always possible to set aside reserve capacity.
Patients are prepared for surgery 1104. Sometimes, patients are biometrically prepared for surgery and have documents or information such as lab results in preparation for surgery prior to the scheduled starting time. Examples of being prepared include being in a surgical preparation area (such as patients P1 and P2) and having blood pressure, heart rates, blood chemistries, and digestive chemistries within protocol and surgical permissions. Dependencies to be satisfied before preoperative surgical services can commence may be monitored. For example, if forms 1160 must be signed at least 15 minutes prior to surgery, staff may be alerted with sufficient time to complete the task prior to the schedule surgical start.
For illustrative purposes, suppose that a specialized piece of equipment E1 1161 in operating room 1 (OR1) 1120 is needed in OR4 1137 in one hour 1165, and E1 is needed in OR1 in 15 minutes 1161, or the OR1 surgery cannot commence. Likewise, the surgery schedule in OR4 in one hour and 15 minutes also cannot start without E1 1165. Further, with respect to a protocol step using E1 for patient P1, the person scheduled for surgery in OR1 has a 65% probability of being finished within 45 minutes, an 80% probability of being done within 50 minutes and a 98% probability of finishing within 70 minutes 1162. Furthermore, surgeon S1 is scheduled to report to OR2 1130 for patient P2 in two hours 1164. If signature of form F1 1160 is delayed, E1 1161 is on critical path because OR4 1137 may start late. Should F1 not be signed until 08:15 hours 1165, OR4's schedule risk is 35%, which is derived from knowing E1 is critical path and its probability of being used and released in 45 minutes (the original one hour minus 15 minutes delay) is 65% (1162), all other factors are satisfactory. If OR4 1137 is not on critical path, then this delay is an annoyance, but not critical as yet. However, note that surgeon S1 is scheduled to start OR2 1130 at 09:45 hours (1165), and that if F1 1160 is not signed until 08:15 hours (1165), that F1 and S1 are now critical path. The Day View Logic may generate a “yellow” alert to nurse N1 to review and have signed F1 at 07:45 hours, for example. The status of F1 would transition to ‘Red’ at 08:00 hours if the forms remain unsigned because of the S1 dependency on OR2 1164. As can be appreciated, combinations between numbers of rooms, patients, staff, equipment and information follow an exponential expansion of scenarios.
Extending the illustration in
Thus, in certain embodiments, a scheduling process includes a dynamic flow of people, assets and information. In the above example embodiment, once surgical procedures are rendered 1122, the clinical workflow moves patients into a post-operative (post-op) area 1150 for recovery and observation/stabilization. Patients are then removed 1107 to floor beds 1108 or discharged 1109, for example. If post-op 1150 or beds 1108 have no capacity, then the dependent antecedent processes are impacted. Day View provides visibility to upstream and downstream processes using the method illustrated above for OR dependencies, for example.
Shared services, such as nursing, surgical instrumentation, equipment technologists and cleaning are scheduled and/or called-for tasks. Availability and capacity constraints impact dependent processes. Illustrative examples include a nurse calling in to be absent from an upcoming shift, a surgical case cart not correctly inventoried, or a cleaning technician occupied in one room as another room is requiring clean-up. A ripple effect of these variations increases process variation. This process variation increases as discrete randomness accumulates.
Day View provides not only a system for identifying causes and effects of variation, but also a mechanism to help recover from unanticipated variation in a data driven and visual way, so as to facilitate conversation, ideas and buy-in of process stakeholders.
Day View may be configured to accommodate increasing degrees of exogenous variation and process objectives. For example, the Day View system may be configured to accommodate cases added to an already full schedule, as well as accommodate concurrently scheduled procedures that finish early and/or late. Process stakeholders help determine where to place new cases, what cases to postpone, which staff to add, redirect or drop, which rooms to shuffle, how to ration equipment, supplies, etc.
Certain embodiments use simulation to explore potential feasible solutions. For example, four modalities of simulation are employed to forecast asset and resource utilization assumptions. These four exemplary modalities are agent based simulation (AB), discrete event simulation (DE), continuous or system dynamic simulation (SD), and Monte Carlo simulation (MC). A critical path method (CPM) may also be repurposed in a simulation-based mode in that CPM is used to calculate critical path, probabilities of completion, and availability and feasibility of a solution.
Referring to
The presently disclosed methods and systems are differentiated from Gantt and Program Evaluation and Review Techniques (PERTs) that replace deterministic duration assumptions in CPM with task duration probabilities. Certain embodiments not only draw from a static assumption of a probability distribution function such as PERT or Gantt but also provide duration probabilities from coupling to a simulation of the physical environment. Thus the CPM or PERT method is enhanced with a most recent actual duration that is observed within the hospital's or other clinical environment's operations from protocols that are tied to the specific care pathways for each patient and also an added precision of dynamic forecasts of duration that result from incorporation of process signals and activities on a dynamic basis into probability assumptions feeding CPM/PERT.
For illustrative purposes, Task1 1202 is, on average, 1 hour in length and completes with 95% probability in 90 minutes. The current art utilizes a single number for duration, typically, the mean or mode. Therefore, nearly half of the time an early finish is realized and half time, a delay is caused. If the delay is a critical path item, other procedures, staff, assets and patients are negatively impacted.
A schedule risk is estimated by converting historical observation of similar task and context durations 1241 into a PDF 1242 and integrating into a cumulative PDF 1243 using one or several simulation methods 1240 (e.g., Monte Carlo), for example. Returning to the CPM for Task1 1202, task and total schedule risk are calculated along a full enumeration of path independent values for time assumptions of tasks. The method allows for an assessment of a given schedule's likelihood of completing on schedule, as well as for any selected time or probability. In this way, a schedule may be adjusted to complete with “50% probability”, “80% probability”, “95% probability”, and/or any other desired likelihood.
Alternatively, a schedule can be made, and a likelihood of any task being completed at a given time can be expressed in a probability, such as OR3 will be ready for surgery at 1430 hours with 85% probability and at 1500 hours with 95% probability.
Tasks are interdependent on people and assets. Path independent assumptions, though superior to static scheduling, may not be sufficient to determine if a given person will in fact be available. An independent assumption assumes that resources required for a task will be available. Certain embodiments of the CPM method disclosed herein call assets of a process as though the assets were tasks and triggers alarms accordingly if double booking or a selected critical path risk is exceeded. However, certain embodiments employ more specific treatment of resources required for tasks through use of discrete event and agent based simulation.
Discrete event simulation 1210 is an ordered step-through of determined tasks 1212, 1213 in discrete time increments. At each time step, designated resources used for the task(s) at hand are attached 1211, 1214. Should concurrent tasks involve the same resource(s), logic 1215 is called from the simulation to determine which task has priority and which release rules apply once a particular resource is attached to a particular task. Thus, it may be possible that a single resource can serve two or more tasks and not be mutually exclusive as the way CPM is broadly configured.
Agent-based simulation 1220 assumes that the resources 1221, 1222 contain a prioritization rule set and respond to the surrounding environment. By studying how the resources (agents) respond, a structure of the system emerges. This is in contrast to Monte Carlo 1240, which is configured with a path independent assumption and linear or formulaic correlation that assumes the simulated task is much the same as its historical assumption set 1241. Monte Carlo determines a system structure from historical assumptions and from discrete events in ordered tasks and logical call of assets or resources.
Dynamic processes with interdependencies, randomness and human judgment defy any singular simulation modality. Certain embodiments utilize a multi-modality approach to asset utilization so that a scheduling structure enabled with the CPM method can be made more accurate and yet preserve the benefits of the visual logic CPM 1201 communicating in an intuitive, rapid, systemic form.
Continuous or system dynamic simulation 1227 is yet a fourth modality and is employed by the Day View scheduling system to assess “stocks”, “flows” and “feedback structure” dynamics in order to assess process tendencies and policy impact. While process objectives are often related to capacity, throughput and timeliness, there are also strategic objectives that are often met or missed—not in a single shift, day or week, but over time. Examples include staff skill, “burnout”, staff turnover, reputation, capacity to serve, and financial operating margin. This class of objectives share the property of accumulation. For example, “skill” is a stock that may increase with exposure to different and/or frequent procedures. “Skill” may also erode through memory and/or dexterity loss from infrequent exposure or training. Similarly, a clinical workflow policy that schedules finish times at an expected value (the 50th percentile) will result in near daily, if not hourly schedule misses and, resultantly, a disruptive, chaotic, stressful work day for the process stakeholders.
Occasional days of chaos are typical in all domains wherein scheduling exists. Related stress from schedule chaos, unanticipated variation, emotional decision making, unplanned overtime and the like will accumulate and, for some staff, will exceed a preference or indifference point causing the staff to seek other employment. A dynamic of new hires is that they do not have departmental knowledge, relationships and trust built due to a lack of history, context and/or requisite competencies. Existing staff are then encumbered with training the new staff. Thus there is detraction from a department's ability to execute tasks when experienced staff quit and new (ongoing) replacements enter. Process variation increases with lower capability if schedule durations remain as before. Thus, other process objectives are desirable to attain and can be attained by policies and decisions executed on an hourly, daily, and/or weekly basis via the decisions made in the clinical workflow enabled with Day View.
Referring to
Referring to
In practice, there are endogenous forces that limit perpetual growth 1320. Should increased demand 1309 exceed capability 1307, a staff may become overextended or quit 1315, capability may diminish 1307, and fewer patients avail themselves of these services until demand abates to a level the staff can serve. Likewise, an overloaded staff and/or assets reduce effectiveness 1312 of asset and staff investment 1305. For example, a specialized asset for a certain procedure is obtained 1305 with intent to increase a department's capability 1307. However, staff, responding to process chaos, do not develop skill in the new competency and thus, the depreciating asset's costs are not liquated by revenue originally intended to be garnered by the investment.
Referring to
An ability to play the day forward is an improvement to the art of service-oriented processes. Prior to the start of a shift, for example, the day's schedule is advanced through time 1414 by adjusting at 1420 the virtual time 1418, such as by a slider bar or dial, and watching the schedule unfold along with the relative risks of delays or early completions of the assets of the process. In this way, the staff can be cognizant of the sensitivities and key risks to the schedule and can be additionally alert when risks increase to the key points. Likewise, a day can be replayed for study or training of decision support algorithms. The replay can be historical and/or comparative to what was planned or even what the scheduling algorithm chose as a robust path forward but was not necessarily selected.
In certain embodiments, dynamic process control adapts to changed priorities that demand resources [people and assets] versus what may have been previously scheduled. An information chain from biometrical sensor to data infrastructure to reasoning algorithm determines what gathered data means clinically in protocol selection. The information chain also determines how clinical state meaning integrates to a what-is process flow and determines how new protocol pathways are exchanged or updated before process control.
As an example, a likely stroke patient presents to a doctor, and the doctor requests a brain magnetic resonance image (MRI) as part of the care protocol, clinical workflow, or pathway. A clinical workflow is a sequence of events that executes a protocol by itself or integrated with other protocols on the same or other patients. Suppose the MRI machine was being used and was scheduled for other patients that are now less critical than the present patient. Rather than having the hospital or clinic build up queues of displaced patients or care providers, the affected schedules are preferably reconfigured to accommodate the change in MRI machine usage. It is further desirable that the new schedule minimizes or reduces factors of operation such as overtime, delays, non-value added time, cost, medical outcomes, patient and staff satisfaction, etc.
In certain embodiments, the state of “what is”, or a current state of a patient process, may be determined using information from biometrical data following a chain of transformations. Sensors or user inputs 1540 provide raw data to a data structure 1550 which in turn supplies reasoning algorithm(s) 1560 to operate on the data. The state of the patient process is then provided for a “what is” or current state analysis, as well as to further determine what the new protocol 1570, if any, is required. New protocol 1596 can be an adaptation of what is currently being performed with the processes of care 1590 or may be an insertion.
Critical path, task and resources are then recalculated at 1590 using method(s) previously described. Affected process stakeholders and assets 1598 are informed of the revised schedule (e.g., when tasks and resources will be available). A new schedule is established at 1510 with the revised information 1597, which is then made more robust using simulation forward testing 1520 methods that establish a new “should be” state. Objective functions may be called within the simulation 1520.
In certain embodiments, information may be provided using a kiosk or workstation. Kiosk screens may be displayed with schedule, planning and/or decision support information, for example. A clinical workflow may be facilitated using input and output provided at the kiosk. Kiosk screens and information may be used to driving a clinical scheduling process.
In a perioperative suite, for example, a whiteboard, one or more computer terminals (e.g., Microsoft Windows-based and/or green screen computers), and/or one or more flat panel screens hung on the wall display activity in different rooms. Color may be used to show whether or not operations are proceeding according to schedule. In certain embodiments, a decision support window may sit or pop up on top of an application running graphics on a screen.
Thus, certain embodiments provide systems and methods for using probabilities to schedule and modify schedules and clinical processes in healthcare delivery. Certain embodiments provide systems and methods to calculate the probability that a scheduled task will start at given a time using CPM and PERT with assumptions derived from an active process feedback and probabilistic historical task durations. Certain embodiments provide systems and methods to calculate the probability that a scheduled task will end at given a time using CPM and PERT with assumptions derived from an active process feedback and probabilistic historical task durations. Certain embodiments provide systems and methods to calculate the probability that a whole or any part of a schedule will start and/or complete as scheduled and/or a given time(s) using CPM and PERT with assumptions derived from an active process feedback and probabilistic historical task durations.
Certain embodiments provide systems and methods to calculate the time that scheduled task will start given a desired probability of a start occurring using CPM and PERT in a simulation or static mode drawing upon assumptions derived from an active process feedback and probabilistic historical task durations. Certain embodiments provide systems and methods to calculate the time that a scheduled task will end given a probability of an end occurring using CPM and PERT with assumptions derived from an active process feedback and probabilistic historical task durations.
Certain embodiments provide systems and methods to calculate the probability of meeting tasks in a schedule of healthcare clinical processes using the critical path method (CPM) to identify tasks and interdependencies that will be impacted by any other task or resource at any time in the schedule.
Certain embodiments provide systems and methods of using probabilistic durations with the CPM and or PERT methods to calculate probabilities of scheduled task starts of finishes and predicent and anticent tasks or resources (people, assets and information) in healthcare clinical processes.
Certain embodiments provide systems and methods to use CPM or PERT to calculate slack time for resources in the active processes of healthcare delivery by embedding the algorithms of these methods into a computerized information and decision making system which actively takes in process and related information, calculates the algorithms and provides active process decision support or automation at one or a multiple of locations.
Certain embodiments provide systems and methods to warn process stakeholders of a deviation from schedule that will under or over utilize capacity of a process in advance of the scheduled time for that process or resource using CPM and PERT with assumptions derived from an active process feedback and probabilistic historical task durations and will feed forward capacity ramifications to process stakeholders who make decisions to add or drop surgical cases or manage beds, transportation and physical assets.
Certain embodiments provide systems and methods to incorporate dynamically updated probability distributions which benefit from the historical and current actual process durations at one or a plurality of hospitals such that the assumptions gain ever more precision where observations at one hospital may be too infrequent into the CPM/PERT based logic rather than a static probability assumptions to calculate probabilities of scheduled task starts or finishes for tasks or resources (People, assets and information) in healthcare clinical processes.
Certain embodiments provide systems and methods for illuminating various states of the schedule with visual, audible, tactile, numerical or shape (or any combination there of) context relative to the degree of impact if certain tasks are not done or resources made available.
Certain embodiments provide systems and methods to identify root cause of actual and likely schedule variation by a decision logic tree incorporating schedule risk calculated from a cumulative probability density function or closed form approximation and active process status feedback combined with Gantt, CPM or PERT algorithms in healthcare delivery processes.
Certain embodiments provide systems and methods to identify what resource(s) to focus on to change the schedule using a decision logic tree incorporating schedule risk calculated from a cumulative probability density function or closed form approximation and active process status feedback combined with Gantt, CPM or PERT algorithms in healthcare delivery processes.
Certain embodiments include systems and methods providing use in combination of discrete event, agent based, continuous or system dynamics simulation or Monte Carlo simulation integrated with PERT and CPM in healthcare related processes to simulate “what if” scenarios, calculate schedule risk, find feasible process solutions, replay events of historical record in an emulation and as a transfer function that can be called from an optimization.
Certain embodiments include systems and methods providing use of agent based simulation integrated with PERT and CPM in healthcare related processes to simulate “what if” scenarios, calculate schedule risk, find feasible process solutions, replay events of historical record in an emulation and as a transfer function that can be called from an optimization.
Certain embodiments include systems and methods providing use of continuous or System Dynamics simulation integrated with PERT and CPM in healthcare related processes to simulate “what if” scenarios, calculate schedule risk, find feasible process solutions, replay events of historical record in an emulation and as a transfer function that can be called from an optimization.
Certain embodiments include systems and methods providing use of Monte Carlo simulation integrated with PERT and CPM in healthcare related processes to simulate “what if” scenarios, calculate schedule risk, find feasible process solutions, replay events of historical record in an emulation and as a transfer function that can be called from an optimization.
Certain embodiments include systems and methods providing use of singularly and/or in combination of discrete event, agent based, continuous or system dynamics and Monte Carlo simulation integrated with PERT and CPM in inter-related healthcare delivery or support processes in and amongst departments or hospital networks in a region to simulate “what if” scenarios, calculate schedule risk, find feasible process solutions, replay events of historical record in an emulation and as a transfer function that can be called from an optimization.
Certain embodiments provide systems and methods to replay “what was” emulations of actual processes at one or a plurality of hospitals for individual or team learning with an interactive user interface or structured command.
Certain embodiments provide systems and methods utilizing “what was” emulation or simulation for activity based costing estimates, checks, validations, re-engineering or justification. Certain embodiments provide systems and methods utilizing “what was” emulation or simulation for billing. Certain embodiments provide systems and methods to replay “what was” for team, peer and operational learning. Certain embodiments provide systems and methods utilizing “what was” historical records to train decision support algorithms that include case-based reasoning, rule-based, fuzzy logic, example-based evidentiary reasoning, neural net, regressive, heuristic or other artificial intelligence based algorithms.
Certain embodiments provide systems and methods utilizing “what was” to calibrate simulation model(s). Certain embodiments provide systems and methods utilizing “what was” for process study, analysis, peer review, re-engineering, best practice sharing and benchmarking.
Certain embodiments provide systems and methods utilizing the planned schedule and comparing it live against “what-is” for process study, analysis, peer review, re-engineering, best practice sharing, decision support and benchmarking.
Certain embodiments provide systems and methods utilizing “what was” for case cart and equipment, assets and resources preferences and decision support process study, analysis, peer review, re-engineering, best practice sharing, decision support and benchmarking.
Certain embodiments provide systems and methods to provide decision support as to where best, given one or more process objectives, to add and insert task(s).
Certain embodiments provide systems and methods to provide decision support as to what tasks to drop and resources to release from specific tasks in order to best meet objective(s) of the process when the schedule changes for any reason.
Certain embodiments provide systems and methods to adjust blocks of dynamically allocated time.
Certain embodiments provide systems and methods to dynamically adjust upper and lower specification limits for various alerts.
Certain embodiments provide systems and methods to call clinical protocol into the CPM structure for schedule monitoring.
Certain embodiments provide a capability to create an exception report for schedule and protocol mismatch.
Certain embodiments provide an ability to configure protocol and schedule mismatch presentation where they are matched, schedules>protocol, protocol>schedule as an exception.
Certain embodiments provide a capability to integrate equipment, biometrical or personal information into a schedule in order to determine impact on the schedule
Certain embodiments provide an ability to use optical, RFID, and EMR dynamic data to assess/forecast the status or degree of task completeness and resources utilization.
Certain embodiments provide an ability to incorporate equipment prognostics into the schedule risk calculation(s).
Certain embodiments provide an ability to incorporate PDF frequency histograms, look-up tables & cumulative probability distributions for procedure duration, staff availability and skill, equipment reliability into the CPM and simulation method assumptions.
Certain embodiments provide an ability to simulate forward to explore feasible and more robust or flexible alternative schedules that satisfy one or more objectives.
Certain embodiments provide coupling of linear programming and/or stochastic optimization to trade off process objectives for one or more process objectives.
Certain embodiments provide systems and methods to simulate forward to explore alternative schedules using CPM and PERT task and asset relationship methods where rule based or simulations utilizing agent based, discrete event, Monte Carlo or continuous differential methods are used to forecast assets of the process behavior.
Certain embodiments provide an ability to forecast the long term relationships between investment in resources, staff, information, healthcare providers with the change in capability and the attraction of more patients with resulting revenues.
Certain embodiments provide an ability to use ‘what ifs’ to plan equipment and surgical inventory logistics.
Certain embodiments provide an ability to use analytical workflow in coordinating the analytical task in an automated and semi-automated fashion.
Certain embodiments provide an ability to summarize the state of the process in either actual time, simulated forward or the emulated historical record and provide a visual indicator such as red/yellow/green, shapes, happy faces, or other audible and visual representation.
Certain embodiments provide an ability to use probabilistic set points to summarize the state of the schedule performance or risk in either actual time, simulated forward or the emulated historical record.
Several embodiments are described above with reference to drawings. These drawings illustrate certain details of specific embodiments that implement the systems and methods and programs of the present invention. However, describing the invention with drawings should not be construed as imposing on the invention any limitations associated with features shown in the drawings. The present invention contemplates methods, systems and program products on any machine-readable media for accomplishing its operations. As noted above, the embodiments of the present invention may be implemented using an existing computer processor, or by a special purpose computer processor incorporated for this or another purpose or by a hardwired system.
As noted above, embodiments within the scope of the present invention include program products comprising machine-readable media for carrying or having machine-executable instructions or data structures stored thereon. Such machine-readable media can be any available media that can be accessed by a general purpose or special purpose computer or other machine with a processor. By way of example, such machine-readable media may comprise RAM, ROM, PROM, EPROM, EEPROM, Flash, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to carry or store desired program code in the form of machine-executable instructions or data structures and which can be accessed by a general purpose or special purpose computer or other machine with a processor. When information is transferred or provided over a network or another communications connection (either hardwired, wireless, or a combination of hardwired or wireless) to a machine, the machine properly views the connection as a machine-readable medium. Thus, any such a connection is properly termed a machine-readable medium. Combinations of the above are also included within the scope of machine-readable media. Machine-executable instructions comprise, for example, instructions and data which cause a general purpose computer, special purpose computer, or special purpose processing machines to perform a certain function or group of functions.
Embodiments of the invention are described in the general context of method steps which may be implemented in one embodiment by a program product including machine-executable instructions, such as program code, for example in the form of program modules executed by machines in networked environments. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform particular tasks or implement particular abstract data types. Machine-executable instructions, associated data structures, and program modules represent examples of program code for executing steps of the methods disclosed herein. The particular sequence of such executable instructions or associated data structures represents examples of corresponding acts for implementing the functions described in such steps.
Embodiments of the present invention may be practiced in a networked environment using logical connections to one or more remote computers having processors. Logical connections may include a local area network (LAN) and a wide area network (WAN) that are presented here by way of example and not limitation. Such networking environments are commonplace in office-wide or enterprise-wide computer networks, intranets and the Internet and may use a wide variety of different communication protocols. Those skilled in the art will appreciate that such network computing environments will typically encompass many types of computer system configurations, including personal computers, hand-held devices, multi-processor systems, microprocessor-based or programmable consumer electronics, network PCs, minicomputers, mainframe computers, and the like. Embodiments of the invention may also be practiced in distributed computing environments where tasks are performed by local and remote processing devices that are linked (either by hardwired links, wireless links, or by a combination of hardwired or wireless links) through a communications network. In a distributed computing environment, program modules may be located in both local and remote memory storage devices.
An exemplary system for implementing the overall system or portions of the invention might include a general purpose computing device in the form of a computer, including a processing unit, a system memory, and a system bus that couples various system components including the system memory to the processing unit. The system memory may include read only memory (ROM) and random access memory (RAM). The computer may also include a magnetic hard disk drive for reading from and writing to a magnetic hard disk, a magnetic disk drive for reading from or writing to a removable magnetic disk, and an optical disk drive for reading from or writing to a removable optical disk such as a CD ROM or other optical media. The drives and their associated machine-readable media provide nonvolatile storage of machine-executable instructions, data structures, program modules and other data for the computer.
The foregoing description of embodiments of the invention has been presented for purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise form disclosed, and modifications and variations are possible in light of the above teachings or may be acquired from practice of the invention. The embodiments were chosen and described in order to explain the principals of the invention and its practical application to enable one skilled in the art to utilize the invention in various embodiments and with various modifications as are suited to the particular use contemplated.
Those skilled in the art will appreciate that the embodiments disclosed herein may be applied to the formation of any clinical software feedback and dynamic scheduling/planning system. Certain features of the embodiments of the claimed subject matter have been illustrated as described herein; however, many modifications, substitutions, changes and equivalents will now occur to those skilled in the art. Additionally, while several functional blocks and relations between them have been described in detail, it is contemplated by those of skill in the art that several of the operations may be performed without the use of the others, or additional functions or relationships between functions may be established and still be in accordance with the claimed subject matter. It is, therefore, to be understood that the appended claims are intended to cover all such modifications and changes as fall within the true spirit of the embodiments of the claimed subject matter.
Claims
1. A method for viewing biometrical information and dynamically adapting schedule and process interdependencies in a clinical environment, said method comprising:
- deriving a schedule including at least one patient treatment protocol based on a mathematical model using people and asset information as input;
- determining at what stage of a protocol at least one patient is using available information to infer or logically determine the protocol stage, the available information obtained from one or more data sets in the clinical environment;
- processing input biometrical data to derive the clinical state of the patient and indicate a patient status;
- adapting said at least one patient treatment protocol based on said biometrical data; and
- revising said schedule based on adapting said at least one patient treatment protocol.
2. The method of claim 1, further comprising recalculating a critical path, schedule tasks, and resource allocation based on adapting said at least one patient treatment protocol and revising said schedule.
3. The method of claim 1, further comprising communicating with affected process stakeholders and assets regarding said revised schedule.
4. The method of claim 1, wherein said biometrical data comprises at least one of patient vital sign data, patient symptom information, and patient electronic medical record data.
5. The method of claim 1, further comprising calculating, using a critical path method and program evaluation and review techniques with assumptions derived from an active process feedback and probabilistic historical task durations, at least one of the following probabilities:
- a) a probability that a scheduled task will start at a given time;
- b) a probability that a scheduled task will end at a given time; and
- c) a probability that a whole or a part of the schedule will at least one of start and end as scheduled.
6. The method of claim 1, further comprising calculating, using a critical path method and program evaluation and review techniques with assumptions derived from an active process feedback and probabilistic historical task durations, at least one of the following times in the daily schedule:
- a) a time that a scheduled task will start given a desired probability of a start occurring; and
- b) a time that a scheduled task will end given a desired probability of an end occurring.
7. The method of claim 1, further comprising at least one of the following scenarios:
- a) providing a historical “what was” emulation of the schedule;
- b) providing a current “what is” visualization of the schedule;
- c) providing a future “what will be” simulation of the schedule; and
- d) providing a predictive “what if” scenario generation for the schedule,
- based on the interdependencies, estimated task durations, input biometrical data, and the state of personnel, equipment, and locations.
8. The method of claim 7, wherein said revising step further comprises adjusting the daily schedule based on at least one of scenarios a), b), c), and d).
9. A scheduling system for viewing biometrical information and dynamically adapting schedule and process interdependencies in a clinical environment, said system comprising:
- a scheduling module deriving a schedule including at least one patient treatment protocol based on a mathematical model using people and asset information as input; and
- an input module processing input biometrical data indicating a patient status,
- wherein said scheduling module adapts said at least one patient treatment protocol based on said biometrical data and revises said schedule based on adapting said at least one patient treatment protocol.
10. The system of claim 9, further comprising a user interface module configured to allow a user to monitor activities and track a schedule of tasks for a given day, said user interface module configured to permit the user to visualize variation of schedule task times, to visualize scheduling opportunities and constraints, and to view schedule risk information with respect to the schedule of tasks for the day,
- wherein said scheduling module is coupled to said user interface module, said scheduling module configured to load the schedule of tasks for a given day, along with decision rules and protocols and to calculate a schedule risk based on a) task interdependencies between personnel, equipment, and location, b) estimated task durations, and c) a state of personnel, equipment, and locations, said logic engine using process feedback from the schedule of tasks along with probabilistic historical data in conjunction with the schedule risk to provide an alert to the user regarding the schedule risk.
11. The system of claim 9, wherein said scheduling module calculates, using a critical path method and program evaluation and review techniques with assumptions derived from process feedback from the schedule of tasks and probabilistic historical task duration data, at least one of the following probabilities:
- a) a probability that a scheduled task will start at a given time;
- b) a probability that a scheduled task will end at a given time; and
- c) a probability that a whole or a part of the schedule will at least one of start and end as scheduled.
12. The system of claim 9, wherein said scheduling module calculates, using a critical path method and program evaluation and review techniques with assumptions derived from process feedback from the schedule of tasks and probabilistic historical task duration data, at least one of the following times in the daily schedule:
- a) a time that a scheduled task will start given a desired probability of a start occurring; and
- b) a time that a scheduled task will end given a desired probability of an end occurring.
13. The system of claim 9, wherein said scheduling module executes at least one of the following scenarios for display:
- a) providing a historical “what was” emulation of the schedule;
- b) providing a current “what is” visualization of the schedule;
- c) providing a future “what will be” simulation of the schedule; and
- d) providing a predictive “what if” scenario generation for the schedule,
- based on the interdependencies, estimated task durations, and the state of personnel, equipment, and locations.
14. The system of claim 9, wherein said scheduling module recalculates a critical path, schedule tasks, and resource allocation based on adapting said at least one patient treatment protocol and revising said schedule.
15. The system of claim 9, wherein said scheduling module communicates with affected process stakeholders and assets regarding said revised schedule.
16. The system of claim 9, wherein said biometrical data comprises at least one of patient vital sign data, patient symptom information, and patient electronic medical record data
17. A computer-readable medium having a set of instructions for execution on a computer, said set of instructions comprising:
- a scheduling routine deriving a schedule including at least one patient treatment protocol based on a mathematical model using people and asset information as input; and
- an input routine processing input biometrical data indicating a patient status,
- wherein said scheduling routine adapts said at least one patient treatment protocol based on said biometrical data and revises said schedule based on adapting said at least one patient treatment protocol.
18. The computer-readable medium of claim 17, further comprising a user interface routine for displaying a schedule of tasks for a given day and allowing a user to monitor activities and track the schedule of tasks along with visualization of timing probabilities, schedule risks, and proposed schedule modifications involving times, locations, personnel, and equipment.
19. The computer-readable medium of claim 17, wherein said scheduling routine recalculates a critical path, schedule tasks, and resource allocation based on adapting said at least one patient treatment protocol and revising said schedule.
20. The computer-readable medium of claim 17, wherein said scheduling routine communicates with affected process stakeholders and assets regarding said revised schedule.
Type: Application
Filed: Feb 29, 2008
Publication Date: Apr 30, 2009
Inventors: Christopher D. Johnson (Clifton Park, NY), Kunter Seref Akbay (Niskayuna, NY), Onur IIkin Dulgeroglu (Niskayuna, NY), David Sterling Toledano (Round Lake, NY), Paul Cuddihy (Ballston Lake, NY), Jenny Weisenberg (Niskayuna, NY)
Application Number: 12/040,668
International Classification: G06Q 50/00 (20060101);
