A working guide for practice leaders, coding and quality leads, and the people who own a queue: why a gap report never closes a gap, the properties a gap needs before anyone can act on it, who owns each kind, how the visit closes it, and how closure and productivity are measured without lying.
By Jonan Portal Alvarez, Chief Executive Officer, Alara Healthcare (management services) · Updated September 10, 2026 · About 9 minutes
Every population health tool I have used could tell me how many gaps we had. A risk score, a gap count, a list of members sorted by something. What none of them could do was the next thing: give the gap to the person who could close it, tell them when it was due, show them the evidence, and put it in front of the provider with the patient in the room. So the report went into a spreadsheet, the spreadsheet went into an email, and the work moved by memory.
A gap report is a snapshot of a problem. A work list is a commitment to fix it. The difference is not the software. It is a handful of properties a gap has to have before anyone can act on it, and most reports supply none of them.
| Property | A gap report | A work list |
|---|---|---|
| The unit | A count, by measure or by condition. | One opportunity: one member, one gap, one year. |
| The owner | Nobody, or a department. | One named person. |
| The due date | None, or as soon as possible. | From the branch's calendar: the sweep, the measurement year, the follow-up window. |
| The status | Open or closed, as of the plan's last file. | Assigned, in progress, done, skipped with a reason, reopened, and acknowledged by the plan. |
| The evidence | Somewhere else. | Attached: what the data says and where it came from. |
| The visit | Not connected. | The provider sees the member's open items with the patient in the room. |
| Closure | Whatever the spreadsheet says. | Confirmed against the plan's file. |
The branches count gaps in different units. Risk adjustment counts conditions awaiting re-documentation. HEDIS counts member-and-measure pairs the plan reports non-compliant. MIPS counts eligible patients who have not met a measure. Transitions count discharges without a follow-up. Those units were never meant to match, and forcing them into one report is how two true numbers get called a discrepancy.
A work list solves this by changing the unit. Everything that enters it becomes an opportunity: one member, one gap, one year, with an owner, a due date, a status, and its evidence. The branch it came from is a label on the item, not a different kind of item. A coder's queue, a coordinator's queue, and the front desk's recall list are the same shape, which is what lets a lead balance them and an administrator measure them.
| Kind of gap | Where it comes from | Who owns it | What done means |
|---|---|---|---|
| Recapture gap | A condition credited last year and not yet documented this year, from the risk file. | A coder or documentation specialist prepares it; the provider rules on it in the visit. | Documented in a qualifying visit, submitted, and accepted by the payer. Not before. |
| Suspect | Evidence in claims, labs, or the chart pointing at a condition nobody has documented. | A coder reviews the evidence and surfaces it or declines to; the provider decides. | Ruled on, either way. A suspect the provider rejects is finished work, and the record says so. |
| HEDIS or Stars gap | A member-and-measure pair the plan reports non-compliant, from the plan's gap file. | A care coordinator or quality specialist: outreach, scheduling, records retrieval. | Acknowledged by the plan as closed. Captured and submitted are the waiting states in between, and nothing is lost while an item sits there. |
| MIPS open measure gap | An eligible patient who has not met a selected measure, from your own data. | The quality specialist, with the data steward when the source cannot see the care. | The numerator met in the source that feeds the measure, before completeness is counted. |
| Scheduling gap | A member with open items and no visit on the books. | The front desk and the schedulers. | A visit kept, not booked. The recall list is population health work, whatever the org chart says. |
| Transition follow-up | A discharge or an emergency visit, from an event feed or a claim. | The practice's care coordinators and clinical staff. | The follow-up visit inside the window, not the call attempted. |
| Data gap | A file that did not arrive, did not load, or failed its checks. | The data steward. | The file read and graded, and the number that depended on it trustworthy again. |
Everything upstream of the exam room prepares one moment. Five moves make it count.
For every member on tomorrow's schedule, the open items across every branch: conditions due for re-documentation, suspects with their evidence, quality measures still open, a follow-up owed. One list, one page, not three logins.
The list belongs in the workflow the provider already uses, at the moment of the visit, with the evidence behind each item. A list that lives in a portal nobody opens during clinic is a report with extra steps.
Each item is a question: is this condition present, is this measure met, does this follow-up matter. The provider answers from what is clinically true. Accept, reject, or defer, and every answer is a recorded decision, including no.
A gap is not closed because it was discussed. It is closed when the documentation supports it, the claim or submission carries it, and the source that feeds the measure can see it. The status moves when the evidence exists.
Weeks later the payer's file either shows the closure or it does not. Reconciling against that answer, rather than against the practice's belief, is what turns a closure rate into a number an auditor and a plan would both accept.
The reason spreadsheets fail as work lists is not that they cannot hold a list. It is that they cannot measure it honestly. Five rules keep the figures true.
Closing a HEDIS gap in the chart and the plan acknowledging it are two events weeks apart. The states in between, captured and submitted, exist so the wait is visible and nobody re-sends a gap the plan is still reconciling.
Skipping an item returns it to the pile, and the queue does not shrink. A skip with a reason is credited as finished work, because a documented reason is a decision. A skip without one is not counted at all.
Opportunities completed says nothing on its own. Read it beside the acceptance rate, the share of surfaced suggestions providers accepted, and the rework rate. A coder who surfaces everything scores high on volume and low on acceptance; a discerning one does the reverse. The pair is the signal.
A reopened item can mean rework, but it also means new information arrived or a policy changed. It is a signal to read, never a verdict on a person.
A decision recorded without a work item is real work that never reaches anyone's figures. A high count of these means the list was inconvenient at that moment, and the fix is the path, not the person.
Pace per active hour excludes idle time, so an afternoon of meetings does not read as a slow coder. Pace per calendar day includes weekends and leave in the denominator, which is why short ranges are volatile and the two are read together.
The list is a month-long loop, not a quarterly project. Gaps are assigned at the start of the month by value and by owner. Coders, coordinators, and the front desk work their queues. The lead watches unassigned and overdue items, which are the two numbers a manager acts on first, and moves work between people when the load is uneven. Closed work reappears the following month as a measured rate against the plan's file, and the provider sees the effect in the visit with the patient in front of them.
This is the cadence Alara Healthcare runs its management services on. The practice keeps clinical authority over every decision in it. The management partner keeps the list honest, the queues balanced, and the closure measured.
What is assigned, what is due, what was done in the last seven days, and a load bar that says light, medium, or heavy. Their own figures, with a note on what those figures will not support as a conclusion about anyone else.
Every list at once: who has what, where the load is uneven, what is unassigned or overdue, and one-click redistribution of a colleague's pending work.
Opportunity burn-down, the closure trend with a projection against the sweep deadline, work by location, and the closure rate reconciled to the payer's answer.
The member's open items with the patient in the room, and the evidence behind each one. Nothing else on this list matters if this view is empty at the moment it is needed.
Care Programs is the service Alara Healthcare runs on this module: the monthly cadence, the queues, the outreach logistics, and the reconciliation. Care Programs services
A good team is exactly who a report fails. They build the real list in a spreadsheet, and the spreadsheet is where the audit trail ends, where two people work the same member, and where the closure rate is whatever the maintainer believes. The list does not replace the team. It gives their work an owner, a date, a status, and a number that survives a question.
Weeks, usually. You close a gap when the documentation and the claim exist. The plan acknowledges it when its next file reflects that. Captured and submitted are the states in between. Re-sending an item during the wait slows the reconciliation down. Chase the plan, not the item.
Because it is not. A skip returns the item to the pile, and the queue is unchanged. Recording a reason turns the skip into a decision, and a decision is finished work, so a skip with a reason is credited. A skip without one is noise and is not counted.
By never showing volume alone. Every pace figure sits beside a quality figure: the acceptance rate for what providers agreed with, the rework rate for what came back. Pace per active hour excludes idle time, so an afternoon in meetings does not read as slow work. And every figure states what it will not support as a conclusion about someone else.
The clinician, in the visit, from what is clinically true. The coder decides whether the evidence is worth a provider's attention. The provider decides the rest. A rejected suspect is finished work, and the platform records it as a decision rather than a failure.
Bring the spreadsheet. Each row is already an opportunity waiting for an owner and a date. The first month is loading it, assigning it, and running one cadence. The second month is the first honest closure rate, because it is measured against the plan's file instead of the spreadsheet's last column.
We will load it as a work list, assign each row to a person, and show what the closure rate looks like when it is measured against the plan's file.
