Minutes Regarding Vaccine Queue Estimate
The website said the estimated wait was forty-five minutes. The actual wait was three hours and twelve minutes. Nobody had calibrated the estimate.
Minutes of the Queue Estimate Review Committee Meeting 1 of 1
Attendees
The committee convened to review the accuracy of wait time estimates displayed on the appointment scheduling portal. No quorum requirements were established because the committee had not met before.
Item 1: The Estimates
The portal displayed a wait time of forty-five minutes at the time of the undersigned's arrival. The undersigned arrived at the location and waited for three hours and twelve minutes.
This represents a discrepancy of one hundred and forty-seven minutes.
Item 2: The Source of the Estimate
The estimate was produced by a system that calculates expected wait times based on appointment volume, historical throughput, and estimated processing time per appointment. These inputs were last calibrated in a prior period.
The calibration period did not reflect current conditions.
Item 3: Analysis
The displayed wait time was calibrated against a scenario of normal throughput. The scenario of normal throughput has not existed since March 2020. The estimates have continued to be displayed without revision.
An estimate that has never been compared to actuals is not an estimate. It is a guess formatted as an estimate.
Item 4: Consequence
People planning their schedules around the displayed estimate arrived at locations expecting a forty-five minute wait. They committed to that window. They did not bring additional provisions, arrange care coverage for the subsequent four hours, or make contingency plans for the duration.
The forty-five minute estimate produced a population of people who were reasonably surprised to still be waiting at hour two.
Item 5: The Absurd Rule
The portal displays a disclaimer in small text below the estimate indicating that wait times may vary. The disclaimer has not been updated to indicate that wait times currently vary by approximately two hundred percent.
The disclaimer is technically accurate. It is not useful.
Item 6: Why This Has Not Been Fixed
Fixing the estimate requires someone to collect actual wait time data, compare it to displayed estimates, calculate a correction factor, and update the model. This is a task of moderate complexity that has not been assigned to anyone.
The estimate continues to display forty-five minutes.
Item 7: Observation
The undersigned is not critical of the people staffing the location. They were processing appointments as quickly as the process allowed. The queue was long. The process took as long as it took.
The estimate was wrong in a way that was entirely avoidable and has remained wrong in a way that continues to be entirely avoidable.
Meeting adjourned.