How to Identify a Process for Improvement in Health Care
Process improvement is essential to business and in particular health care. Health care has come under great scrutiny due to increasing costs. Even hospitals, clinics and doctors that are doing well in the areas of care and cost can always improve. Lean and Six Sigma are just two tools that will guide you in finding opportunities for improvement in health care, positively impacting patient care, relationships with physicians, employees, suppliers and customers.
Identify opportunities for improvement in health care by defining goals. This should be done in a top-down order, where the goals for the overall organization are identified, then goals for each department are identified and finally goals for each process are identified. Organization-wide goals might include higher standings in such health care accreditation and reporting associations as the Joint Commission, a reduction in overall costs, increased patient throughput, decreased employee turnover and more practicing physicians. Department-wide goals might include reduced turnaround times for procedures, a decrease in duplicate paperwork and increased employee satisfaction. Process goals might include shortened patient wait times and streamlined processes. Utilize such tools as: mission statements, process maps and value stream mapping. Process maps show a procedure in its entirety, including time frame and resources. A value stream map focuses on value-added parts of a process, running horizontally across the value stream map, while non-value added parts of the process are indicated vertically.
Look for ways to measure the goals of the organization, the departments and the processes, utilizing dashboards and data sampling. Dashboards are used to visually show the status of a process in relationship to the attainment of a goal. A dashboard indicates what part of the process is being measured, such as patients per day, what the current count is in relationship to the goal and will be conditionally color-coded to have visual impact, showing whether the process is on target (green), in danger of not making a target (yellow) or not making the target without drastic measures (red). Data sampling collects information for statistical analysis and takes into consideration data points -- the turnaround time for procedures.
Collect data and analyze it to identify outliers that will pinpoint opportunities for improvement. The data collected will depend greatly on the time or goal that is being measured. For example, if you are looking to improve the number of check rewrites required, your data would include all checks that are written; the checks that are re-written would be indicated as such so that you can find a percentage to know what your starting point is. An outlier is a data point that is outside the majority of data points. Outliers are opportunities for quick improvements. An outlier might be a turnaround time on an x-ray reading that took 24 hours because the reading results were misplaced in an email inbox. A quick fix to this would be to set up an email alert when messages from the radiologist, whether outsourced or in-house, are received.
Utilize such tools as ANOVA, standard deviation, regression analysis and normality testing to analyze the data. ANOVA, analysis of variance, is a statistical calculation that identifies sources of variances. Data points may be time per patient in the MRI rooms. Sources for variance within the times may be day of the week, requesting physician, age of patient, gender of patient or number of staff members. Standard deviation doesn't look at individual variances within data points but will find the norm of the data points collected and, in doing so, will identify those data points that lie outside the norm. These are instances that need to be reviewed to determine what occurred to produce a data point outside the norm. Regression analysis creates a model for forecasting with numerous variables. Normality testing work backs through a process to determine whether a model is realistic and attains the desired results. If the goal is to have a turnaround time for imaging readings of three hours, then the following information would be inserted into a normality test: minutes from when the image is created and when it is sent to the radiologist, time for the radiologist to read it, time for the results to be sent back to the health care organization and time for the results to be sent to the ordering physician. If those times do not total less than three hours under normal circumstances, then the model will not fulfill the goals desired.
Use the goals, measurements, data and the statistical analysis to identify opportunities for improvement, utilizing such tools as: control/influence matrix, brainstorming, critical to quality, voice of customer, current reality trees and prerequisite trees. A control/influence matrix leads a project team through parts of a process, determining whether the project team has control, influence or neither control nor influence during a part of the process. Discussions need to be held with all interested parties of a process: the process owner (management), the processors (staff-level employees) and the customers (whether patients or physicians) to determine what is critical to quality. Voice of the customer focuses on the improvement based on the customers' needs and wants. Interviews, surveys or questionnaires can be used to find out what the customer is saying. Current reality trees (CRTs) show project teams a current list of problems within a process. A CRT will lead the project team through the downstream effects of problems within the process. Downstream effects of incomplete physician orders might be that a patient waits in an exam room with a paper robe on for an hour while staff members try to locate the physician for clarification on orders. A prerequisite tree is read from top to bottom, with the desired outcome listed at the top, and the prerequisites for obtaining that outcome listed below it. Prerequisites might be obstacles in a current process which aren't working, such as proper lab instructions not being communicated to a patient or a part of a process that doesn't currently exist that needs to be inserted, such as requiring that all employees use direct deposit.