Printer Friendly

Systematic approach reduced patient waiting times.

At Northwest Family Physicians, Ltd., a medical group practice in Glendale, Ariz., process mapping, a graphic step-by-step depiction, was used to study the flow and traffic patterns of the entire patient encounter (figure 2, page 13). A common source document, the super bill, was used to record appointment time, arrival time, chart up time, doctor sees patient time, and sign out time. Deming's Statistical Process Control (SPC) was used to analyze the data and monitor and control the patient flow process. If any process is in control, it will vary around a mean. Statistical techniques are used to determine the upper and lower ranges. Twice daily, monitoring provides feedback. Physician feedback is essential for efficiency.

The SPC method of analysis requires the use of run charts to obtain feedback on the process before and after any changes. (A run chart is a graph plotting the average wait time on the y axis and the day of the month on the x axis.) Ishikawa diagrams were used to analyze the causes of any unwanted variation of a process. An Ishikawa diagram is a fishbone shaped outline that lists people, technology, materials, and methods as they apply to a particular output problem. Histograms were used to illustrate results of data comparison. A histogram is a bar graph displaying an attribute from the greatest quantity to the least.

Finally, to succeed in gathering consistent and accurate data, everyone in the office was educated on the importance of documentation and the steadfast commitment to running on time.

Analysis and Results

Initially, the simple act of measuring every patient's waiting time improved the average waiting time from 29 minutes to 16 minutes by December 1993. We feel this was due to increased awareness (the Hawthorne Effect). Through process mapping, we found two main causes of variation: external and internal.

External Causes of Variation

External causes of variation include factors that can prolong waiting time (process out of control) and that are not in full control of the physician: patient timeliness, not showing, unique problems, and emergencies. Daily run charts revealed that some patients arrived early and some late. If patients are seen in the order of arrival rather than at appointment times, all subsequent patients are seen late unless there is a gap in the schedule. We initiated a strict policy of seeing patients only at their designated appointment times. Starting in March 1994, patients who were more than 10 minutes late were rescheduled or waited until the next available opening. By October 1994, the timeliness of patients improved markedly. The percentage of late patients during the month of March was approximately 21 percent. Seven months later, the percentage of late patients dropped to 1.9 percent.

Internal Causes of Variation

Internal causes of variation include situations that can prolong waiting times and that are controlled by the physician or staff. This encompasses the time interval from when the physician sees the patient until checkout, the core of the patient encounter process. Physician inefficiencies show here.

Analysis of the core process determined that there are three types of patients in the encounter process:

* Type A. An established patient who is seen by the physician without procedures.

* Type B. An established patient who is seen by the physician and receives procedures.

* Type C. New patients who are seen by the physician and require physicals and procedures.

Data were collected to evaluate the physician's efficiency when treating Type-A patients. Control charts of Type-A patients seen during March 1994 showed that the physician was inefficient. This conclusion was inferred from control charts (x-bar and range charts). Two-hundred-twenty-three Type-A patients seen during March 1994 were rated by the control chart.

The x-bar chart (figure 3, page 14) measures the mean (average) waiting time of each day's sample of 10 Type A patients. The abnormal distribution of daily averages indicates that the process is out of control. This infers that the physician is not seeing Type-A patients in a timely manner. The points are erratic; two points are significantly above the upper control limits, and a few points are on or below the lower control limits. This is due to special causes of variation (human error) and should be investigated.

Range charts measure the range of values for each sample. They reveal any tendency of the process to behave more or less randomly. The data in figure 4, page 14, reveals an abnormal distribution of the amount of time it takes the physician to see 10 Type-A patients per day during a one-month period. This mirrors the findings from the x-bar chart. One point is above the upper control limit and few points are near the control limits. The process is out of control and is not within a normal range of distribution. This is due to special causes (human error) and should be investigated.

Core Process Investigation

A Pareto chart was constructed by using histograms to show the frequency 6f the attributes of the special causes of variation. A Pareto analysis is performed by arraying all the causes of an effect in sequential order from the highest to lowest. The underlying principle is that 80 percent of the effects are from 20 percent of the causes.

The Pareto chart in figure 5, page 14, shows that the most common attribute associated with prolonged patient waiting was a phenomenon that we labeled Out of Order (000). A patient with the earlier of two adjacent appointments is late, and the person with the later appointment is very early. The person arriving early is seen first even though his or her appointment is not till later. Therefore, charts are being processed in order of patient arrival, not the time of appointment. This increases appointment delays more than would seeing patients in strict order.

The next common cause of patient waiting refers to a situation that we call the Herbie Effect (HERB), from the book The Goal. The "Herbie Effect" is an example of a series of sequential events that are delayed at a bottleneck. Each subsequent event is late, despite 100 percent efficiency elsewhere.

Double Booking (DBLBK) is the next common phenomenon. Two patients are scheduled for the same time, both are told they have appointments, and both are given appointment reminder cards, but only one is written in the book. The problem is discovered only on the arrival of both patients.

Complex Problem (CMPLX) is the next common attribute associated with prolonging patient waiting time. "Complex problem" refers to a situation where a person presents with an illness that needs quick, accurate diagnosis and the diagnosis is uncertain. It requires more time in examination, history taking, laboratory work, x-rays etc. The event is completely unpredictable.

Late refers to the patient's arriving late for his or her appointment. It results in some isolated instances of people having to wait. One third of patients arrive late. Only 10 percent are seen beyond the upper control limit. The "Late" phenomenon becomes more critical as scheduling density increases.

Start refers to the process of starting late in the morning or in the afternoon. The first patient in the morning may not see the doctor for three to seven minutes after the appointment time. If one starts late, then one may stay late.

Results

As a result of the Pareto analysis, we decided to attack patient timeliness at the same time as physician and staff inefficiencies. The strict discipline of not seeing patients until their designated appointment and of rescheduling patients arriving more than 10 minutes late for their appointments helped to significantly improve attributes associated with patient timeliness, such as the "Out of Order" phenomenon, the "Herbie Effect," and "Late."

Daily graphs showed a marked decrease in the frequency of the "Out of Order" phenomenon, accompanied by a decrease in the "Herbie Effect." The percentage of patients arriving late decreased 19 percent between March and October. Only 2 percent of patients now arrive late each month.

The new appointment discipline did not improve key factors associated with internal causes of variation (physician and staff inefficiencies), such as Double Booking and St Strict discipline initiated by the physician to start on time, have all the materials, data, and personnel necessary at the appropriate time; and not to handle more than one problem at a time, helped to significantly improve the Start phenomenon. The Double Booking event is still a significant problem.

Figure 6, above, illustrates how changes initiated by the physician in March 1994 to improve patient waiting have improved overall efficiency. On average, the amount of time the physician spends with patients is in control, which means that the physician has been able to more evenly divide the amount of time spent with Type-A patients. However, according to figure 7, above, the physician is still out of control, which means he is still taking longer than his allotted time of 10 minutes to examine certain Type-A patients. Further analysis is under way to solve the problem.

Conclusion

Overall, this study of patient waiting time has led to a conclusion that the patient process is really a system and is subject to systems malfunction. Structure determines behavior, and structure is composed of key factors and their interrelationships. Key factors are correct scheduling, patient density, physician efficiency and effectiveness, human complexity, patient timeliness, preparation for the encounter (materials, data, personnel), lost charts, process measurement, and steadfast intention to run on time. The changes initiated at Northwest Family Physicians Ltd. significantly improved the physician's patient encounter process. Patients are extremely satisfied with the changes.

The long-term plan is to continue to improve the quality of the patient encounter process at Northwest Family Physicians Ltd. by focusing on the key factors. Presently, study is under way to improve the efficiency and effectiveness of support personnel. This will focus on the scheduling of patients. We are considering grouping patients into categories and scheduling them accordingly. This arrangement is much like batch processing and should decrease waiting time and maintain the individuality of patients' care. Lost charts are another problem that wastes the time of clerical staff. However, the real challenge is how to handle complex problems, which take a bit longer to solve, and still be a caring, nonmechanistic physician.

[Figures 1 to 7 ILLUSTRATION OMITTED]

Rose Buckle, RN, MHSA is Outpatient Clinical Coordinator, Veterans Affairs Hospital, Phoenix, Ariz., and Ted J. Stuart Jr., MD, is CEO of Northwest Family Physicians, Ltd., Glendale, Ariz. The authors may reached at 5422 W. Thunderbird, Rd., Suite 20, Glendale, Ariz. 85306, 602/942-1151, FAX 602/843-2388.
COPYRIGHT 1996 American College of Physician Executives
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1996, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Author:Stuart, Ted J., Jr.
Publication:Physician Executive
Date:Apr 1, 1996
Words:1753
Previous Article:Outpatient critical pathways: five advantages for physicians who act now.
Next Article:March to national health reform continues, but slowly.
Topics:


Related Articles
Helping physicians manage challenging patient encounters.
RBRVS: old concepts, bad press.
How hospital wayfinding is important to healthcare.
Challenges in patient safety and quality: replacing discouragement with hope.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |