Six One-Page Worksheets Turn Each Method Into Team Work

Each worksheet is a one-page tool that turns a chapter’s method into something a team fills in together. They work on paper, in a spreadsheet, or as the starting point for an AI prompt. Keep the completed sheets: a quarter’s worth of them is the practice’s own record of what it learned.

Fill In One Row per Rung for One Program and One Month#

Capture Ladder worksheet.

For one program and one month, fill in one row per rung.

Circle the largest drop. That is the rung to work on first, no matter which rung is easiest to fix.

Log Care Minutes and Admin Minutes for Two Weeks#

Admin-minutes time study.

For two weeks, every person who touches the program keeps a simple log. Each line records the patient-month it belongs to (a study ID, never a name), the task, the minutes, and one of two labels: care (the patient would recognize it as care) or admin (finding, chasing, counting, reconciling, re-keying, validating). At the end, divide total admin minutes by the number of enrolled patient-months in the period, billed or not. Then sort the admin tasks by total minutes. The top three are your first automation targets.

Sort Every Monthly Task Into Machine, Assisted, or Human#

Human-or-machine sort.

List every task in the program’s monthly cycle, one per line. For each, mark one column: machine does it (remember, count, route, check), machine assists (summarize, draft, classify), or human decides (clinical judgment, consent conversation, escalation). Any task marked “human decides” that a person does the same way every time belongs in a different column. Any task marked “machine does it” that could hurt a patient if it went wrong needs a human check.

Model the Program at Low, Best-Estimate, and High Values#

Practice Math three-scenario worksheet.

Model the program three times, once with low values for every input, once with your best estimate, and once with high values, and compare the bottom row.

Fill in every input three times. If the program only works in the high column, it doesn’t work.

Name Where Each of Nine Components Lives and Who Would Notice#

Care-Delivery Operating System map.

For each of the nine components (eligibility logic, enrollment and consent, care plans, work queues, escalation pathways, documentation, billing validation, audit evidence, quality and outcome review), write down where it lives today: software, a spreadsheet, or a person’s memory. Then write the name of the one person who would notice if it stopped working. Every component with no name next to it is a failure waiting for a date.

Test Ten Random Patient-Months Against Seven Tests#

Honest Standard monthly check.

Pull ten enrolled patient-months at random. For each, answer yes or no to the seven tests: medical necessity, real service, patient understanding, true documentation, right supervision, measured outcome, and willingness to stop. Any “no” gets a named owner and a fix before next month’s claims go out. Two months in a row with the same “no” means the process, not the person, needs changing.