What the denominator is in a completion rate

Any completion rate can be raised without doing more work: shrink the denominator. That is why the first question an experienced buyer asks is what the measurement runs against, and why the answer has to be written down before starting.

A defensible denominator is defined by a mechanical rule: every order written in a period, across a declared set of services, with no filter for whether the patient has a phone, no filter for whether they answered, and no filter for whether anyone got round to calling.

How does a rate get inflated without lying?

• By excluding patients without a valid phone number, who are precisely the hardest ones to book.
• By counting only the orders the team managed to work within the month.
• By grouping several billable services from one order into a single unit when that helps, and splitting them when the opposite helps.
• By stretching the window until the late bookings fall inside it.
• By removing patients from the period once they were seen at another provider.

Those exclusions have a formal name in US quality measurement: a denominator exclusion is a case the measured entity removes from the denominator before the numerator criteria are ever evaluated. None of those five decisions is dishonest on its own, and all five raise the figure. The defence is publishing them: how many cases each exclusion removed and what the rate would have been without it. That, plus a simultaneous control group, is what makes two months comparable. The full methodology is in how we measure.