Even though this isn't really personal finance related I still feel like there are some misconceptions here that could be addressed.
I don't know where you got the phrase "pass-through" insurance from. What you're describing is a self-funded plan. In a self-funded arrangement an employer contracts a third-party-administrator (TPA), usually one of the big health insurance carriers, to use it's provider network, process and adjudicate claims, etc. In addition to the TPA there will be some sort of stop-loss insurance coverage on each participant. Stop-loss coverage usually provides a maximum amount of risk on a given member and on the entire population for a given month and/or year and/or lifetime. The employer's risk is in between the plan deductible and the stop loss coverage (assuming the stop-loss doesn't have a maximum).
Almost all of the claim dollars in a given plan will come from very very few people. These costs typically arise out of very unforeseen diagnoses not chronic issues. A cancer patient can easily cost $1,000,000 in a year. Someone's diabetes maintenance medicine or other chronic maintenance will cost no where near what a botched surgery will in a year.
If we take a step back there are really four categories of employer insurance.
- Small group (2-100)
- Mid-Market group (101-1,000) -- sometimes this goes to 500 depending on the market
- Large group (1,001-5,000) -- over 5,000 is typically self-funded
- Self-funded Large Group (2,000-unlimited) -- under 2,000 doesn't usually make financial sense
Small group is tightly regulated. Usually plan premiums are filed with a state authority, there is no negotiating, your group's underwriting performance has zero impact on your premiums. Employers have no way of obtaining any medical/claim information on employees.
Mid-market is a pooled arrangement. The overall pool has a total increase, and your particular group performs better or worse than the pool which may impact premiums. Employers get very minor claims data, things like the few highest claims, or number of claims over a certain threshold, but no employee specific information.
Large-group is a mostly unpooled arrangement. Generally your group receives it's own rating based on its individual underwriting performance. In general the carrier is offloading some risk to a stop-loss carrier and employer's get a fair amount of insight in to claims, though again, not with employee names.
Self-funded is obviously self-contained. The employer sets up a claims checking account. The TPA has draft authority on the account. The employee's typically have no idea the plan is self funded, their ID cards will have the carrier logo, and the carrier deals with them just as it would any other member. Generally when a company is this size it has a separate benefits committee, those few people will have some level of insight in to claims performance and stop-loss activity. This committee will have nothing to do with the hiring process.
There are some new partially self-funded arrangements, which is just a really low-threshold (and relatively expensive) stop-loss program, that's becoming somewhat popular in the mid-market group size as employers attempt to reduce medical spend.
I think when you start thinking on a micro, single employee level, you really lose sight of the big picture. Why would an employer hire this guy who has this disease/chronic problem that costs $50,000 per year? And logically you can get to the conclusion that with a self-funded plan it literally costs the company the money so the company has an incentive not to hire the person. I understand the logic of the argument, but at the self funded level the plan is typically costing north of half a million dollars each month. So a mid-level HR hiring manager 1. isn't aware of specific plan claims or costs and is not part of the benefits executive committee, 2. won't be instructed to screen for health deficiencies because it's against the law, 3. a company generally won't test the water here because $50,000 per year is less than 1% of the company's annual medical expenses, 4. $50,000 is well below the cost to litigate a discrimination law-suit.
Really the flaw in your thought process is that $50,000 in annual medical expense is a lot. A harsh child-birth can run in the $250,000 range, so these companies never hire women? Or never hire men who could add a spouse who's in child bearing years? Or never hire women who might have a female spouse who could be in child bearing years? A leukemia diagnosis will ratchet up $1,000,000 in a year. Spend a bit of time in intensive care for $25,000 per day and you're fired? A few thousand bucks on diabetes meds isn't anything relative to the annual cost of your average self-funded plan.
The second flaw is that the hiring managers get insight in to specific claims. They don't.
Third, you don't hand over medical records on your resume anyway.
I typed this out in one single draft and have no intention of editing anything. I just wanted paint a broad picture, I'm sure things can be nit-picked or focused on.