Case Study - Arcadia Family Practice

The confirmation system trial at Arcadia can only be described as extremely successful. Practice revenue has increased because the confirmation system has reduced the no-show rate to a low 5.45%, patient acceptance of the confirmation system is over 99% and the Arcadia office staff overwhelmingly approves of the system.

The confirmation system has already started to pay for itself. Using an 18% standard no show rate (MGMA) and an assumed patient load of 22 patients scheduled per day, Arcadia would have expected to have 79 no shows over our test period; The actual number of no shows was 23. An improvement of 56 patients over the one-month test period. At $40 per patient it is a simple matter to determine that Arcadia’s revenue increased $2,240 during the time of the test.

Without Confirmation With Confirmation Patients per day 22 22 Number of days 20 20 No-show rate 18% 5% Expected Arrivals 360 416 Value @ $40 per visit $14,400 $16,640

$2,240 in additional revenue was created in twenty days. Extrapolated over an average work year of 220 workdays the confirmation system could be expected to generate an additional $24,640 in revenue.

The following graph details the daily no-show percentages:

click for larger picture

It is important to note that the one day where the no-show rate rocketed out of its normal range the practice computer system was inoperable the day prior and the confirmation system was unable to make calls for the 23rd. Correcting that number to conform to the average number of all other days would drop the average no-show rate from 5.45% to 4.7%. It is also significant to note that the one-day the patients were not called the no-show rate jumped into the reported national average range.

A brief statistical summary of the data provides some interesting insight into the actual productivity of the confirmation system. After eliminating the 23rd from the data pool due to errors beyond our control the average number of no-shows per day is exactly 1 with a standard deviation of .75 patients. In a normal distribution 99.7% of all data should be within 3 deviations above the mean. Based on this sample, 4 no-shows in one day with the calling is inconsistent with known data. That is, we would not expect it to happen. Using a confidence test, a test designed to evaluate the spread of an event around the mean based on a desired confidence level, it can be calculated that 95% of the time the number of no shows will be between .65 and 1.35 no shows per day. There are two related but separate conclusions to be drawn from the statistical tests:

• The four no-shows of the 23rd should never happen if the confirmation system is operating properly, and …
• The data clearly show that the confirmation system will not allow a practice to experience standard, known no-show levels as experienced by practices as a whole.

Initial numerical assumptions had to be refined during the test. We started by assuming that Arcadia’s 40 patients per day was a scheduled 40 patients per day. This is inaccurate. After interviewing the staff at Arcadia the number that Joanne Link thinks is most appropriate is 22 patients scheduled per day. Thus, I used 22 patients when calculating the no-show rates and the additional revenue figure. This created a "truer" picture about the performance of the confirmation system. While using the initial 40 per day would have improved the numbers substantially it would not have been accurate. Fortunately, even with the much lower number the confirmation system excelled.

The following pie charts detail the results of the patient comment form.

Clearly the practice is performing at a high level with respect to customer service to garner this kind of response.

It is important to note that the one respondent who said that the practice of automated appointment confirmation weakened his perception of the practice is a twelve-year-old boy whose mother insisted that he fill out the card. The following is a complete transcription of all comments from the patient cards:

"I think it’s great!"
"It reminded me of the appointment when I forgot."
"Convenient."
"A helpful reminder."
"I was very impressed with the system."
"Seems less personal than an actual person making the call."
"I think Dr. Morgan does a great job."
"Well pleased with Dr. Abbot."
"Strengthens my perception because without an appointment reminder I might forget about my appointment."
"Some appointments are made far in advance and it’s good to be reminded."
"Indicates to me that someone is on top of things."
"Good reminder, very helpful to help remember the appointment"
"I think it is a neat idea"
"Good Idea"
"Always very helpful when I am called"
"With my busy schedule, I think that a reminder is great because I tend to forget"
"As a busy mom I appreciate the reminder"
"Strengthens – I think it is great!"
"Very Good Idea"

Throughout the trial we have been in contact with the practice to gauge perceptions of the system. The front office personnel (Crystal and Emily) enjoy being relieved of "phone duty" to pursue work that cannot be done by a computer. Joanne Link is extremely satisfied with the system and the technical support. It is clear that the confirmation system has improved patient arrival at the practice.