Help with Turnover Rate Calculation – Keeping Headcount Stable While Filtering

Hi Fabric Community,
I'm working on a turnover rate calculation in Power BI and running into a challenge with how filters affect my headcount denominator.
Here’s the setup:

Table A contains employee profile data (e.g., employment status, career stream, worker type).
Table B contains termination records (e.g., termination date, reason).
I calculate turnover rate as:
Turnover Rate = Terminations / Average Headcount

✅ What I Need:

The termination count should respond to filters (e.g., career stream, location, time).
The average headcount should:

Be calculated over completed months in a selected time period (monthly, quarterly, annually).
Respond to filters from Table A (e.g., career stream).
Ignore filters from Table B (to avoid inflating the turnover rate).

❌ What’s Going Wrong:

When I apply filters from Table B (like termination reason), the headcount shrinks.
When I try to remove filters using REMOVEFILTERS(Table B), the headcount becomes static and doesn’t respond to filters from Table A.

🔍 What I’ve Tried:

Using CALCULATE([Monthly Headcount], REMOVEFILTERS(Table B))
Wrapping AVERAGEX over filtered months
Creating disconnected tables for filtering

💡 What I’m Looking For:

Best practices for keeping headcount filter-responsive to employee attributes but immune to termination filters.
Whether disconnected tables or alternate modeling approaches are better.
Any DAX patterns or modeling tips that have worked for others.

Thanks in advance for your help!