Aaron E. Carroll's Blog, page 108

Why does it take so long?

Over the weekend, when I messaged Aaron about my obstructive sleep apnea (OSA) situation, his reply was, basically, “Why can’t they get you a CPAP machine right away?”

Welcome to the U.S. health system. Here are the sources of delay, with time=0 the moment the doctor had a diagnosis and called me.

I was not available when the doctor called, though I would have picked up if the caller ID had indicated it was him. I called back within a day and got voice mail. I left a message. I never got a call back. (USA! USA! USA!) Meanwhile, I suspected I did not have OSA because (a) during my sleep study, nobody came in to fit me with a CPAP machine, as I was told would happen if they saw apnea events; (b) I assumed the doctor would express more urgency if I had that diagnosis. So, score this as a communication failure for which I am only partly responsible.
~1 week later I remembered I hadn’t connected with the doctor and called back. This time I got him and my diagnosis. I can take some blame for this delay, but it was not without some justification (see above).
Knowing little more than what I could recall from the ICER report I was involved in, I told the doctor I was interested in an oral appliance. Once those words came out of my mouth, he decided not to put in a request for a CPAP machine. That made sense to me at the time, but now I think that was bad care. He trusted the patient too much. I’ve never had this diagnosis before! However, I also take some blame for my ignorance about CPAP and bias against it. A far better process would be an informed, shared decision-making one, perhaps starting with a referral to some instructional videos. I did all this work on my own this past weekend (as I’ve documented in prior “sleep apnea” posts), which reduced my fear of CPAP tremendously. Now I’m all for it.
This false assurance followed by self-education process cost me another week or so. But, the prescription for a CPAP is now in! I also have a dental appointment for an oral appliance (intake, not to actually make the thing). I can do both, at this stage, but will probably go with CPAP as my first treatment. If that fails me for some reason, I am that much further along in the oral appliance branch.
However, both branches (CPAP and getting the intake dental appointment) take about 4 weeks. The former is because of insurance delays. The latter because the dentist I want to see is busy. I’m sure I could find another dentist, but I don’t know how to judge them. Anyway, it doesn’t matter because I want to do CPAP first.
So, what is the insurance company doing for 4 weeks? Utilization management. It wants to make sure I have all the documentation to justify CPAP. This is somewhat reasonable. On one hand, nobody would get CPAP if they didn’t have some sleep problem, so wasteful utilization is somewhat bounded. On the other hand, we would not want docs just giving out prescriptions for the thing without being sure it was warranted, and the sleep problem wasn’t actually something else (like insomnia that can be cured very cheaply with CBT). CPAPs are not cheap!
Should this take 4 weeks? That’s a cost-benefit or marketing question. My insurer could certainly hire more claims processors. That’d boost admin costs, showing up in higher premiums. By the way, I’m being generous here, I was really told it could take 4-6 weeks, but could be faster. So, I figure 4 is about the average.

So, there you have it. After an insurance-created, multi-month delay in getting my diagnosis, there’s about another ~1.5 months to treatment. And that’s just to start treatment. I expect an adjustment period. It won’t actually be working for me for several weeks after I start. So, let’s call it 2 months to treatment from diagnosis. I guess it could be worse. It sure as hell could be better, but that might cost us even more than we already pay.

That’s a long time to have crappy sleep! Now you see why I’m working so hard at interim solutions. BITWMA!

@afrakt

Healthcare Triage: Everything you missed from last week

I was on vacation last week, so I didn’t get a chance to promote Healthcare Triage. First up:

In the United States, there’s a lot of worry about babies getting TOO MUCH to eat. Overweight and obesity among children is a real concern, and eating habits established in early childhood are a factor. There’s an idea that babies can direct their own food intake, and self-direct on how often and how much they eat. It turns out, babies aren’t that great at making these kinds of decisions.

Then news: Allergies and atopic disease are on the rise. Especially food allergies in kids. A HUGE new study has looked at how changes to the microbiome can lead to allergies.

And finally, the podcast: This week on the Healthcare Triage Podcast, we’re sitting down with Dr. Jonathan Fridell, organ transplant surgeon. He and Aaron talk all about how exactly organs get moved from one person to another, and I can tell you, it is FASCINATING.

Go listen. You can find the podcast in all the usual places, like iTunes, Soundcloud, or even on YouTube.

@aaronecarroll

Breakthrough

I’m an engineer at heart (and by formal training), so I know the credo:

There’s only so long I’m going to let a problem bother me before I will try to fix it.
When I set my mind to it, I will damn well fix it.
An inelegant solution beats no solution. Let them laugh, you still solved the problem.

My obstructive sleep apnea situation provides me with generous time in the wee hours for my engineering brain to tinker with ways to remain unconscious during apnea events. (The story of why a long-term treatment is taking so long is another blog post, but it’s the usual thing: our health system is not very good at some things, and delivering me a CPAP machine quickly is one of them.)

My problem in sleeping through apnea events in the early AM is that I’m a very light sleeper then. Every. Single. One. Wakes. Me. Up. Consciously. I can count them. Yes, 9 per hour (as verified in the sleep lab) seems just about right. The first one comes just moments after I drift off. My sleep is way too light to get through it.

But, I reasoned, if I could get through that first one, I’d be able to stay unconscious a lot longer, through many subsequent ones. How to do this?

There are drugs (see my prior posts), but not ones I want to take every day. Last night I was Ambien free.

Then there are an assortment of mechanical solutions, all of which are aimed at keeping the tongue out of the airway. I have so much time to experiment with these, including last night between 2:30-6AM.

I tried biting my tongue to hold it outward. Yes, that’s dumb, but one gets desperate. It didn’t work. I unbite it as soon as I fall asleep. Plus it’s very drooly, which also wakes me.

I tried sleeping on my stomach, to maximize the gravitational pull on my tongue mouth-outward. To my surprise, I can do this. It didn’t help. I drift off, staking a claim on sleep. Mean Mr. Apnea snatches it away.

I tried thrusting out my lower jaw as is achieved by mandibular advancement devices. Nope. Can’t sustain it while entering sleep. Also drooly.

I tried all manner of pillow architecture. Didn’t do a damn thing.

After all this, at 4:30AM last night, I’m thinking, thinking, thinking. There has to be a way.

Suction. I can create a partial vacuum in my mouth, easily. I bet you can too. In part it’s how you scratch your hard pallet with your tongue or click your tongue against it. Once initiated, the tongue can be pinned to the roof of the mouth, where the suction is not that hard to sustain for many minutes. All I had to do was keep the tongue suctioned up there long enough to get through the first apnea event and, voila, I entered sufficiently deep sleep that I was unconscious from 4:30-6AM, during which time, no doubt, I experienced about 15 more.

Engineering wins.

@afrakt

The Ambien trials, part 3 (and other updates)

Intervention: 5mg of Ambien at 10PM, another 2.5mg at 2:30AM

Result: Slept from 10:30PM-6AM (7.5 hours). I’d be lying if I said it was the best sleep. I can recall several wakings — more than the (old) normal. I’m definitely groggier than normal this morning too. Still got some Ambien in me from that early morning dose, small as it was.

Conclusion: This version of Ambien is not the best for me. Either an extended release version, taken at bedtime, or a fast acting version, taken at 1-2AM or so, would be better, as has been suggested by several readers.

Anyway, I’m going to wrap up these trials for a few day and go without Ambien. I’m very concerned about getting reliant on it. I definitely will want to use it during a few, upcoming trips. I’m sure I won’t have a CPAP machine (yet) for them. (Cue grumble about the stupidly slow health system, considering what we pay for it.)

Other updates:

1. Some reading, but by no means thorough, suggests to me that OSA can have some connection to inflammation and, by logical extension, perhaps tendinitis. The pathway would be that OSA causes stress hormone spikes. I guess it’s stressful to not breathe. Go figure. I have noticed for years that during times of stress, my arm and knee tendinitis is much harder to manage. (Sometimes this is how I know I am experiencing subconscious stress — before I recognize it consciously, I feel it physically.) I’ve been experiencing management challenges of both for a handful of months now. My suspicion is that this was an early warning sign of OSA. Had I only known! Hopefully, OSA treatment will also improve my tendinitis. I thought I had two problems when I may really only have one.

2. A reader and CPAP user recommended this small CPAP machine for travel.

3. Another reader, and physician, recommended these as an alternative to any of the devices I’ve discussed so far. Super simple, but expensive (and, frustratingly, disposable so the expense just lasts forever). Maybe a good travel alternative or an interim measure while waiting for a long term solution. I’ve ordered the starer kit and will report back.

4. As a sleep aid, a reader recommended trazodone as an alternative to Ambien. I’ve heard this advice before.

5. I saw the CPAP machine of yet another friend on this gathering I’m at. His was smaller and quieter than the first two I’d seen. And, by quieter, I mean effectively silent. I was impressed. This friend had also used, and liked, a dental appliance. It was years ago and he didn’t remember much about it. However, after some discussion we jointly concluded that it could not have been a mandibular advancement type.

@afrakt

The Ambien trials, part 2 (and other updates)

Taking the full, 10mg dose of Ambien “worked” better than the 5mg dose I tried two nights ago. I say “worked” in quotes because it’s confounded with lots of other things. For one, I was already more sleep deprived. For another, my anxiety is coming down somewhat about my diagnosis and what to do about it. That alone should help me sleep.

So, the fact that I woke at around 4:45AM on 10mg instead of 2:15AM on 5mg may not have been caused by the dosage difference. Also, to be more precise, on 5mg, I slept from 9PM to 2:15AM (5.25 hours), with a few, trivial moments of sleep thereafter, which hardly count. On 10mg, I slept from 10:15PM to 4:45AM (6.5 hours), with a bunch of quite nice, multi-minute dosing until 6AM (so, let’s say I got more like 7.25 hours of sleep).

But, still, what woke me both times, and interrupted my dosing thereafter, were apnea events. I’m sure of that now. This ain’t no insomnia of yesteryear.

So, my fundamental question remains: As I wait for the gears of the health system to turn to deliver me a longer-term treatment, how can I best remain unconscious during/through those events? My trouble spot is typically 1-2AM onward.

One physician recommended a dose of sublingual zolpidem (brand: Intermezzo), taken in the middle of the night. The purpose of this variant is that it is intended to treat difficulty returning to sleep from nighttime waking. I would like to try it, but my PCP hasn’t provided a prescription. The active ingredient is the same as Ambien, so I’m tempted to try a half dose of that tonight in the wee hours, maybe chewed up to speed its delivery. Crazy? (Anticipatory concern: If I end up using Ambien for a week or two, I wonder how to come off it. Taper? Or, should I not use it regularly over that span? Take a break every few nights?)

Other updates:

1. While dosing intermittently this morning, I messed around more with a pile of pillows to try to get the tongue gravity assist I wrote about yesterday. I don’t think it helped. Maybe the gravitational force one can get to act on the tongue while sleeping on one’s side — constrained by the limits of comfortable neck contortion — is simply not enough. I forgot to try some sort of stomach sleeping, but I’m pretty sure I won’t find that easy to pull off.

2. I’m on a retreat with a bunch of buddies, three of whom have OSA. All use and like their CPAP machines. Two have given me demos of theirs. They are way quieter than I imagined. They’re also bigger than I had hoped, particularly with the humidifying attachment, which also acts as a noise dampener. (Cleverly, one fellow puts ice in his humidifying compartment in the summer.) We talked at length about all the wonderful and annoying things about the machines, masks, etc. (I’ll talk to the third friend today. He’s also used oral appliances.) One thing they suggested is that if I am going on a short trip and don’t want to be bothered bringing the machine, I could just not. My case is mild and I can get a passably decent night sleep with a dose of Ambien (or maybe even without it). Be flexible. It’s nice to have a support group.

3. A TIE reader and sleep specialist physician in Calgary reached out to me. After some email, we spoke by phone for an hour yesterday, going over all my questions. This was AMAZING! I am incredibly grateful to him for the time. It’s exactly what I needed and was not getting from the US health system, within which I feel like I’m on a conveyor belt. No clinicians I’ve seen here want to talk about all the confusing issues I’m dealing with. This is all old and boring to them and I’m just one patient of dozens they have to cram into their busy days. It sucks. This kind Canadian offered to try to help me find someone locally who would give me the attention I want, if that’s possible here.

He also sent me some reading:

Canadian Agency for Drugs and Technologies in Health’s evidence review on OSA treatments. It’s more current than the ICER or AHRQ reviews I had reference in prior posts.
An even more recent clinical review
A collection of patient resources

@afrakt

Readers’ feedback on obstructive sleep apnea

I am grateful for the feedback I’ve already received about obstructive sleep apnea. (All prior posts on this topic are appropriately tagged.) Below are a few things I’ve heard and some additional, related information I dug up. I am going to presume anyone reading this will know what the acronyms and other jargon mean (if not, Google is your friend).

Disclaimer: In no case should you assume my evidence review has been thorough. I’m still cutting my way through the thicket here. None of this is clinical advice. I have a PhD only, anyway (not that there’s anything wrong with that).

1. CPAP > MAD. The most common feedback I received was to encourage me to go right to CPAP, skipping MADs. Some suggest this because the evidence shows that CPAP has better outcomes. This is true! See the ICER report and the AHRQ report that it updates.

However, it’s not that MADs are ineffective. Here’s a recent review that has come out since either of these reports that makes that point. Here’s another one. Notice that MADs aren’t worse on all measures.

An important consideration is that, as with any treatment, clinical effectiveness (as demonstrated in trials) isn’t the only variable. According to some studies, CPAP has lower adherence, which suggests there are things about it that some patients don’t like, relative to MADs. CPAP machines are also more difficult to travel with (particularly when backpacking or to locations without power). I am especially concerned about how either device will play with my fragile sleep profile (easily awakened by noise or discomfort). That could go either way. MADs are silent, but could be uncomfortable. CPAP machines make some noise (I need to hear some to appreciate how much) and also must impact comfort somewhat.

But, MADs can break. I haven’t seen relative failure rates of them vs CPAPs, but I’m willing to believe MADs fail more often. Breakage of one is why one respondent switched to a CPAP machine, though he also said that what insurance would cover played a role in this decision too. It’s potentially meaningful that I’ve yet to hear from a single patient who prefers MADs. But, not everyone who responded has tried both. Some have. Some admitted they haven’t. I suspect CPAP is just more common, perhaps a consequence of coverage (?).

Another critique of MADs is that they can cause jaw pain. True! But at least one study shows that doing jaw exercises can help with this. Clearly MADs are going to affect one’s bite and teeth more than CPAPs. On the other hand — and I’ve done zero work chasing this down — I have seen some anecdotes about CPAPs associated with ENT issues.

Also, and again I have not chased this down, I saw something about the possibility of CPAPs potentially causing a progression from obstructive sleep apnea to central sleep apnea. I guess that might make sense if the brain starts to learn it doesn’t need to do any work in keeping the airway open, but this could also be internet BS (?).

Given these other considerations, it might be reasonable to trade the lower effectiveness of a MAD for higher comfort or convenience if it provides those (for an individual patient). Ideally, a patient would try both. Over time, many do. And I may be one of them.

2. Don’t use Ambien. One respondent said this was not an OSA solution. Clearly I know that. Its purpose, in my case, would be to address short-term, OSA-exacerbated insomnia issues until I get a longer-term (and proper!) OSA device. (If you’re just joining me, yes, I know all about CBT for insomnia.)

One respondent said it or benzodiazepines were associated with future dementia. First of all, association studies — let’s be careful. Second of all, there are studies that say otherwise. Third of all, risk vs. benefit. Forth of all, short-term use. Enough about this.

3. Sleep position matters. A friend made the point (implied it, but I still have enough wit to make the inference) that OSA is a mechanical issue and is addressed with mechanical solutions. This being a physical world, I am free to experiment with various mechanical solutions on my own, as I wait for the real deal.

For obstructive sleep apnea, the issue is that the tongue is blocking the airway. In part that is (or could be?) because the airway has lost some muscle tone or resiliency of some sort (I’ve yet to internalize all the clinical jargon). But gravity also plays a role. For most sleep positions, and especially supine (on one’s back), the tongue is pulled down into the airway by gravity. Stupid gravity.

I can still remember my physics well enough to reason that the effect of gravity on the tongue would be the opposite (away from the airway) in a prone (on the stomach) position. Maybe, maybe, it’s also possible to get this effect on one’s side (where is the fancy jargon for that?) — my preference — if one can get the mouth pointing somewhat downward.

You can probably imagine how one might do this with the right pillow architecture and neck/head tilt. here’s the best picture that sort of illustrates it that I could find on the internet, if only she’d tilt her head a bit forward and/or rotate it a bit pillow-ward (see what I mean?):

This idea led me down a rabbit hole searching for evidence of sleep position on apena events. There’s stuff out there, but it’s pretty much all about how supine sucks and prone rocks, relatively speaking. OK, but what about clever, mouth downward side sleeping?

With a pile of pillows, I did a bit of experimenting with this early this morning (more to come) and it’s clear one can get some gravity going in the right direction. The process measure here is drool. If it’s coming out of your mouth, figure your tongue is subject to the same forces, provided physical laws are constant across space. I cannot yet conclude it helps any, but I was already pretty wake from hours of apnea events and lack of effectiveness of my small Ambien dose, etc.

Naturally, one could just go full prone, but that’d take considerable adjustment for me. I searched for prone sleeping pillows. Not convinced I’ve found anything worth the money. I did find this, for side sleeping, but it’s hard to tell if it’s worth it.

4. What about OTC MADs? One respondent pointed out they’re out there, which I knew. But which are worth trying (again, as an interim step before getting something better or a CPAP machine)? No idea and I couldn’t find any sound guidance on this point (feedback welcome).

5. Nobody said this, but there are a lot of crappy studies out there about OSA and treatments. I can’t be bothered to find them now, but I chased down some work about how playing various wind/brass instruments might be associated with more/less OSA. I play trumpet, so this is relevant. I don’t think any of this stuff is conclusive and the studies are not very good.

@afrakt

The Ambien trials, part 1

What does one do with oneself between diagnosis of obstructive sleep apnea and treatment? My appointment with a dentist just to start the process of getting an oral appliance is one month away. That will be followed by an adjustment period of how long I don’t know. The oral appliance may fail and then I’d go to a CPAP machine. This could all take a while. I’m thinking it’ll be months until back to anything like normal.

So, I have a lot of time to tinker with myself.

My preference is to go unconscious for the duration, but I’d settle for 7-8 hours at night. What I mean by this is that though I can’t stop the apnea events without treatment, I might be able to keep them at a subconscious level, which is better than nothing. In fact, I suspect this is how I’ve been living for months, if not years. I very much doubt that my awareness of apnea events starting in December was the first time they occurred.

Maintaining unconsciousness is hard for a recovering insomniac whose sleep is disrupted multiple times per hour. I can usually do it for the first 3-4 hours of the night. The trouble begins around 1-2AM and can last several hours. At worst, I’m up for the day. This is to be avoided!

Small doses of Benadryl (Diphenhydramine*) can work for me, usually. I didn’t know I was treating OSA this way. I thought it was just dealing with occasional, short-term insomnia spikes. (PSA: Long-term, insomnia not associated with another problem, like OSA, is best treated with cognitive behavioral therapy.) By small doses I mean 1/4 of one pill (6.25mg). A full dose is two pills (50mg).

I told my primary care doc this yesterday and she suggested I give Ambien (Zolpidem*) a try. So, now I’m tinkering with that. But, how to use it?

My first trial was last night. Being risk averse, I tried a half a pill (5mg) at bedtime. That knocked me unconscious until 2:15AM. I suspect apnea events woke me, and they certainly kept me awake thereafter.

So, the first trial failed. What’s next?

My choices are (a) take the full, 10mg pill at bedtime or (b) take a half pill (5mg) when I wake around 1-2AM. The first of these would be the “normal” way to take the drug. It’s as directed. But would it sustain through my trouble spot?

The second of these would seem to target my problem very precisely. I bet it would work in the sense of helping me sleep in the wee hours, but maybe I’d have trouble waking in the morning. That could be a risk worth taking, since I typically don’t have to get up at any particular hour.

I have 9.5 more pills in the bottle, so I can try both of these, as well as a few other variations, if I think of any.

* Approximately nobody knows these things by generic names, so I am not going to use them exclusively, if at all. Yeah, that’s giving in to the brand manufacturer. Sue me.

@afrakt

 

Massachusetts wants to drive down Medicaid drug costs. Why is the Trump administration so nervous?

This piece is co-authored by Rachel Sachs, associate professor of law at Washington University Law School. It’s cross-posted at the Health Affairs Blog.

Although drug formularies are ubiquitous in Medicare and the private insurance market, they’re absent in Medicaid. By law, state Medicaid programs that offer prescription drug coverage (as they all do) must cover all drugs approved by the U.S. Food and Drug Administration, however expensive they are and however slim their clinical benefits may be.

Massachusetts would like to change all that. In a recent waiver proposal, Massachusetts asked the Centers for Medicare and Medicaid Services (CMS) to allow it to adopt a closed formulary in Medicaid. That would allow Massachusetts to exclude certain brand-name drugs from Medicaid, increasing its leverage in price negotiations beyond what it can achieve through existing utilization management techniques like prior authorization.

Among Medicaid advocates, the proposal is controversial. Some fear that state budgets would be balanced on the backs of Medicaid beneficiaries, who could be denied access to expensive therapies. But Massachusetts thinks there’s room to drive down drug spending without threatening access to needed medications. In any event, the state has to do something. Drug spending in Massachusetts has increased, on average, 13 percent annually since 2010, threatening to “crowd out important spending on health care and other critical programs.”

By all rights, CMS should welcome Massachusetts’s proposal. Closed drug formularies are tried-and-true, market-based approaches to fostering competition over drug prices, and the Trump administration’s Council on Economic Advisers recently released a report saying that “government policy should induce price competition” in Medicaid. If Secretary of Health and Human Services (HHS) Alex Azar means it when he says that “drug prices are too high,” letting Massachusetts try out a formulary makes a ton of sense.

But The Waiver Is In Trouble

Recent press accounts, however, report that Azar is likely to reject the Massachusetts waiver. Details are sketchy, but the agency is apparently concerned that the pharmaceutical industry would sue CMS if it approved the state’s request for a closed formulary.

As law professors, this explanation mystifies us. Yes, PhRMA might sue CMS if it approves the waiver. This is America; you can sue anyone for anything. The right question is whether PhRMA could win.

We don’t see how. CMS’s legal authority to grant the waiver is secure. Under section 1115 of the Social Security Act, CMS can “waive compliance with any of the requirements” of section 1902 of the Medicaid statute. One of the requirements of section 1902 is that state Medicaid programs must cover all medically necessary drugs. CMS is therefore free to relieve states of that obligation through an 1115 waiver.

To be more precise: Section 1927 of the Medicaid statutesays that, “[i]n order for payment to be available” from the federal government for a drug, a state has to agree to cover the drug whenever it’s medically necessary. In exchange, manufacturers must sell the drug to the state’s Medicaid program at a rebated price. But states have to adhere to section 1927 only because, under section 1902(a)(54), they must “comply with the applicable requirements of section 1927.” And section 1115 allows CMS to waive compliance with any provision of section 1902, including the part that incorporates section 1927.

Once a state’s waiver is approved, the “costs of such [waiver] which would not otherwise be included as expenditures … shall, to the extent and for the period prescribed by the Secretary, be regarded as expenditures under the State plan,” and thus eligible for matching payments from the federal government. In other words, CMS can grant waivers allowing states to receive federal Medicaid dollars for a drug that isn’t purchased in conformity with section 1927.

There are some conditions on the approval of 1115 waivers. Any waiver must be “likely to assist in promoting [Medicaid’s] objectives” and must be a genuine experiment. But both conditions appear satisfied here: Making Medicaid a smarter purchaser of drugs likely advances the program’s purposes, and Massachusetts has said it will evaluate the effects that its formulary has on price and access. In any event, CMS is in no position to be finicky about the limits of 1115 waivers—when it comes to the approval of state waivers for work requirements, it’s already pushing the legal envelope. If it’s not worried about the lawsuits over work requirements, why worry about similar lawsuits from PhRMA?

What About The Rebates?

The industry’s legal objection to the Massachusetts waiver may run deeper. There’s been some suggestion that Massachusetts can’t secure a selective waiver of section 1927 of the Medicaid statute. On this argument, maybe section 1927 is an all-or-nothing deal: In exchange for guaranteed access to state Medicaid markets, drug manufacturers promise to offer substantial rebates. If access is restricted, those rebates evaporate too.

We haven’t seen a fully developed version of this argument and we have no idea if CMS is taking it seriously. But the Medicaid statute doesn’t support it. Section 1927 may have been the product of a negotiated legislative compromise—a “‘grand bargain’ struck 28 years ago,” as The New York Timesput it—but nothing in that section purports to limit CMS’s waiver authority. That’s a telling omission. Other parts of the Medicaid statute (section 1916(f), for example) do place constraints on waivers, suggesting that Congress knows how to constrain waiver authority when it wants to.

Indeed, CMS selectively waived section 1927 just a few weeks ago when the agency granted Arkansas’s request to waive “Section 1902(a)(54) insofar as it incorporates Section 1927(d)(5).” A Massachusetts waiver would look almost identical: It would waive section 1902(a)(54) to the extent that it incorporates section 1927(d), which is the part of section 1927 that curtails states’ authority to exclude drugs from coverage. Left intact would be sections 1927(a), (b), and (c), which require drug manufacturers to offer rebates. We could be missing something, but we don’t see any legal objection to that approach.

Preemption

Even more misguided is the suggestion that federal law might somehow preempt Massachusetts’ closed formulary. State laws are subject to preemption where Congress says they’re preempted (express preemption) or where a state law conflicts with the purposes of a federal statutory regime (conflict preemption). The pharmaceutical industry might look to support a preemption theory by invoking Biotechnology Industry Association v. District of Columbia, a 2007 Federal Circuit decision invalidating a Washington, D.C. law that prohibited drug manufacturers from charging “an excessive price.” In the court’s view, Washington, D.C.’s effort to regulate prices conflicted with the federal patent system’s policy of offering lucrative financial rewards for inventors.

The court’s decision is questionable, to say the least. Right or wrong, however, the decision speaks to a state effort to impose direct price controls on drugs. It doesn’t speak to the question of whether a closed drug formulary is unlawful. And it certainly doesn’t speak to the question of whether a formulary adopted with the explicit consent of the federal government is unlawful. Here, Massachusetts wouldn’t be adopting a formulary in the teeth of a federal statute. It would be adopting a formulary pursuant to a federal statute—in particular, pursuant to the terms of a Medicaid waiver contemplated by the Medicaid statute.

So What’s Going On Here?

To our eyes, the legal objections appear to be pretext. Why, then, is CMS apparently so reluctant to grant Massachusetts’ waiver? We can envision a number of possible reasons, some benign and others less so. It could be, of course, that the reporting overstates the case against the waiver. Two weeks ago, Secretary Azar said that HHS would unveil “a whole slate of” proposals on the topic in “about a month.” Maybe Secretary Azar will use the opportunity to endorse Massachusetts’ waiver.

Perhaps more plausible is the idea that CMS wants congressional buy-in before taking a step—approving closed formularies—that would undermine the legislative bargain struck between rebates and coverage. The Trump administration’s budget proposal, for example, calls for “new statutory demonstration authority to allow up to five states more flexibility in negotiating prices with manufacturers.” Maybe that reflects a genuine desire to act through legislation instead of through a waiver. Then again, CMS knows that Congress has been unable to take real action on drug prices. Waiting for Congress is tantamount to doing nothing.

There is another possibility. CMS reportedly fears that other states may request similar waivers if Massachusetts’ is approved. (Indeed, Arizona has already made such a request.) But if other states come to think that closed formularies are a good idea, why shouldn’t they follow Massachusetts’s lead? That’s how federalism is supposed to work: When a state lights on a good idea, the idea proliferates. It’s sensible for CMS to worry about copycat states when a state waiver might lead to a big increase in federal Medicaid spending, but closed drug formularies should reduce federal spending. The more states jump on board, the better.

In any event, CMS’s reticence here is in sharp contrast to its eagerness to promote the spread of work requirements. CMS has already approved work requirements for three states, and applications are pending from at least eight other states. We do not see a principled argument for distinguishing between the two policies on this basis.

If none of these explanations hold water, a more alarming possibility comes into focus. The Trump administration may have succumbed to lobbying from the pharmaceutical industry. We hope that’s not the case. It would be unlawful for CMS to deny an otherwise-legal waiver just because it threatens industry’s bottom line.

Whatever is going on here, however, it does not bear the hallmarks of good policymaking. CMS should also keep in mind that it faces litigation risk no matter what it does. PhRMA might sue if the waiver is approved, but Massachusetts might sue if it’s denied.

Secretary Azar has only been in office a few months, and he is still working to develop the agency’s policies on drug pricing. His tenure has already marked a significant improvement over Tom Price’s in a number of ways, and we are optimistic that he will prove us wrong about what’s going on at CMS. If he does not, however, we anticipate further legal battles for the administration.

@nicholas_bagley and @RESachs

 

New adventures in sleep deprivation: obstructive sleep apnea edition

I was fascinated by sleep as a kid and wrote at least one grade school science class report about it. It struck me as both mysterious and frustratingly inefficient that we become unconscious for one-third of our lives. Perhaps some part of me knew that for many periods in my future I wouldn’t.

Even as I suffered with and overcame* insomnia a few years ago (about which I’ve written a great deal here and elsewhere), I served on an ICER committee to evaluate diagnostic and treatment devices for obstructive sleep apnea (OSA). “Oh, how much more awful it would be to have that,” I thought at the time. Maybe the saints or the fully enlightened have voided their minds of all traces of instinctive revulsion at the thought of disease or disability, but I haven’t.

I should. Experiencing the affliction is not the same as imagining it. Human nature though it is, projecting discomfort at any level on the afflicted is unkind, even unjust.

I’ve come to learn that OSA exacts a different kind of sleep deprivation than insomnia and not quite as terrible as I feared (but still no walk in the park).

“You have mild obstructive sleep apnea,” the doctor told me last week. This was just after my in-lab sleep study, which followed two home studies that ended inconclusively due to equipment failure. “You wake up nine times per hour on average, thirty times during REM sleep, of which you get too little.”

“In what sense is that mild?” was my obvious question. My other, unexpressed thought was, “I don’t think ‘wake up’ means what you think it means. Believe me, were I to wake that much, I would damn well know it.” My subjective experience is that I only wake a few times, tops — each with a loud and startling snort — and most nights not at all.

This started around Christmas. During the few months of waiting for appointments, experiencing inconclusive home tests, and waiting for insurance to cover the lab sleep study, I had seen an ENT. His hypothesis was that my symptoms might be due to intermittent and otherwise benign, mild inflammation. Good theory, but, as it turns out, incorrect.

Fewer than 15 OSA events per hour is considered “mild.” Also, my oxygen level only drops three percent, another indicator of a mild case. Mild though it may be, it does’t play nice with insomnia, should one lean that way. For a fragile sleeper like me, being awakened even a few times in succession can translate into just staying awake for hours.

Yet, it turns out, it is possible to be a relatively high functioning and productive member of society with mild OSA, at least for some time. Witness I keep cranking out Upshot posts and getting my regular job done too.

Still, it’s been getting harder. The diagnosis probably also explains why it’s become increasingly difficult to exercise as much as I used to, and I admit I am tired frequently.

“Do you feel tired,” the doctor asked.

“I’m 46. How am I supposed to feel?”

There’s also memory loss, I guess due to less REM sleep. Perhaps it’s no coincidence that several of my Upshot posts over the last few years have focused on memory and memory aids.

These are the early days, between diagnosis and treatment. The health system being what it is, it’ll be at at least a month before I start treatment, probably with an oral appliance (which sounds like a toaster in the mouth). According to anecdotes and the interwebs, these can also reduce or eliminate snoring. Awesome! Some are over-the-counter. Are those any good? (I will get a prescription one made by a dentist, nevertheless.)

I have lots of other questions. I want to write more about this topic (here and at Upshot, editors permitting). Those that pertain to my own health I will ask of my own doctors. Those that pertain to the topic more broadly, I would love to talk to clinical experts. If you are one and want to chat, let me know.

* One doesn’t really overcome insomnia. “Recovering insomniac” is more apt. The tendency is always there.

@afrakt

How much is too much? What does the US actually spend on health care administration?

Elsa Pearson is a Policy Analyst with Boston University’s School of Public Health. She tweets at @epearsonbusph.

The United States spends much more on health care each year than wealthy equals around the globe. That’s not just true for spending on direct patient care, but also for spending on health care administration. Many scholars recognize the cost containment potential in curbing administrative costs. Determining just how much the US spends on health care administration and in what ways are critical first steps.

How much of US health care spending is on administration?

Health care administration includes all activities related to coordinating health and medical services, such as scheduling, billing, and claims processing. Administrative costs’ contribution to overall health care spending is large and growing.

System-level estimates for health care administrative costs are limited and often dated. One highly cited estimate suggests that administrative costs accounted for about 30% of total health care expenditures in 1999. In 2006-2007, administrative costs outpaced growth in other health care categories, such as professional services, and matched growth rates in typically costly categories, such as prescription drug spending.

Though we may not be able to pin down a current estimate of total administrative costs, we do know it’s substantial and continues to increase. Other research focuses on administrative costs in three, large subcategories: billing and insurance-related (BIR) costs, hospital administration, and physician practice administration.

BIR Administrative Costs

BIR costs are often an easily quantifiable subcategory of administrative costs. The National Academy of Medicine calculated the US spent about $361 billion on BIR costs alone in 2009, or 14.4% of total health expenditures. That number only seems to multiply; another study suggests the same costs totaled $471 billion in 2012, or 16.8% of total health expenditures (with 80% directly related to the US’ multi-payer system).

A case study of an academic health care system by Tseng, et al. showed that BIR costs accounted for 14.5% of professional revenue collected during primary care visits (as well as 25.2% during discharged ED visits and 13.4% during ambulatory surgical visits). The authors estimated the monetary impact of this administrative burden on primary care providers equated to almost $100,000/year/provider.

Hospital Administrative Costs

Research by Himmelstein, et al. found that US hospitals spend about 25% of total hospital expenditures on administration. For-profit hospitals tend to have the highest administrative costs. (This does not equate to better care, however, as for-profit hospitals often have higher death rates as well.)

Blanchfield, et al. conducted a case study of a large teaching hospital’s physician organization to determine the impact of administrative costs on the organization’s bottom line. While understood it was already a “high performance billing organization” and, thus, already ahead of the curve in reducing wasteful spending, the study found that excessive administrative costs accounted for about 12% of the organization’s revenue in 2006, or about $45 million. Processing and billing of claims alone accounted for 12.5% of those administrative costs, or $5.6 million.

Physician Practice Administrative Costs

Sakowski, et al. found administrative costs to consume 10% of revenue in a multispecialty medical group. In order to successfully maintain administrative procedures, the organization employed two administrative staff for every three clinical providers.

Research by Casalino, et al. concluded that medical practices spend about $68,000/year/physician interacting with health insurance plans. Primary care physicians, especially those in private practice, feel the most impact. Over three quarters of participants indicated that the costs associated with health plan interactions had increased within a two year period. Similarly, Papanicolas, et al. found that over half of surveyed physicians found insurance-related administrative tasks to be a significant burden.

Worth noting, high administrative costs are not just associated with misuse of resources but rather also accrue in efficient health care systems.  Tseng, et al. found no glaring inefficiencies in billing practice, such as task duplication or poor resource management, during their analysis. Blanchfield, et al. studied a physician organization with streamlined billing practices as well.

How does US health care administration spending compare to other countries?

When compared to 11 other wealthy countries, the US leads the pack in health care administration spending.

Himmelstein, et al.’s 2010-2011 estimate of about 25% of total hospital expenditures devoted to administration exceeds the Netherlands’, the next highest spender, by about 5.5 percentage points. US hospital administrative spending exceeds Canada’s and Scotland’s, the lowest spenders, by about 13 percentage points. In fact, US administrative costs have risen over the past decade while Canada’s costs have decreased.

Similar comparisons exist at the physician level as well. Morra, et al. compared the resources spent in 2006 on provider-payer interactions by small physician practices in both the US and Canada (specifically, Ontario). Canadian practices spent only 27% of the financial resources on provider-payer interaction that US practices did. Canadian support staff—nurses and administrative staff—spent one tenth of the time on these interactions compared to US support staff. The authors estimate the US could save $27.6 billion/year on administrative costs by adopting Canadian administrative practices.

Why are administrative costs so high in the US?

Though studies provide various hypotheses on why US health care administration is so expensive, the big picture is always the same: it’s structural.

The United States utilizes a largely private, multi-payer system. The complex payment scheme may cause systemic fragmentation, leading to additional, unnecessary administrative costs. Private insurers contribute the largest share of billing and insurance-related costs in the US ($198 billion in 2012) and, because the US health care system relies on them so heavily, their significant financial footprint may contribute to high overall administrative costs.

(A new Health Affairs article from Gottlieb, et al. explores billing complexity across insurance types—both public and private—in the US. They found, in general, that higher billing complexity led to higher administrative costs. The article provides fresh insight into US administrative costs and we will unpack the details in a future TIE post.)

The Netherlands, the nation with the second highest administrative expenditures, also has a multi-payer system, and one that is moving towards a market-based approach. Canada, on the other hand, has a single payer system and—coincidently?—has the lowest national hospital expenditures.

What should we do about high administrative costs in the US?

It’s worth acknowledging that administration costs are not automatically bad. Morra, et al. suggest “administrative tasks are wasteful [only] if their costs exceed the benefits they generate or if the same benefits could be achieve at a lower cost.” Policies to reduce administrative costs should aim to reduce waste while preserving worthwhile administrative activities.

A single payer health care system is often considered the “gold standard” for reducing administrative costs. However, most would agree that adopting a single payer system in the United States is a political nonstarter, though it certainly is gaining momentum in public opinion.

There are ways reap many of the administrative benefits of single-payer without moving to it. This would rely on standardization, simplification, and automation.

Standardization & Simplification

Every US payer operates within its own unique system—unique forms, processes, and rules. A “single-payer approach,” with one set of rules and operations for all payers, could reduce administrative costs. All individual payers would utilize the same claims forms, submission methods, etc., under a universal operating framework.

The current health care financial system is complex and often burdensome to providers and administrative staff and a simplified financial system may have significant cost-savings potential, or at the very least, reduce waste. Streamlining other administrative activities, such as credentialing, quality measurements, or benefit eligibility, could also contribute cost savings in the long-term.

Minimizing waste in the current system could lead to substantial savings. Berwick, et al. emphasize reducing administrative complexity (such as, inefficient rules and procedures) and reducing fraud/waste (both actual scams and the procedural regulations put in place because of them). The authors estimate that poor performance in these two categories accounted for $189-661 billion of wasteful spending in 2011.

Automation

Health information technology (HIT) and electronic health records (EHR) were developed, in part, with the intent of streamlining administrative tasks and patient care, resulting in long term cost savings.

Effectively conducting correspondence, payments, and insurance-related activities electronically may reduce administrative costs. Lee, et al. argue that the current rush towards EHR implementation lends itself as a natural opportunity for improvement. The Health Information Technology for Economic and Clinical Health Act (HITECH Act) was created to provide financial motivation for EHR implementation. A 2015 CDC report suggests the HITECH Act may have successfully incentivized providers to adopt some sort of EHR system.

Cutler, et al. propose the HITECH Act lays the foundation for comprehensive electronic correspondence. Expanding HITECH legislation to include electronic transmittance of other pertinent administrative data, such as claims data, may be a logical way to create system-wide reductions in administrative costs. The authors speculate an annual savings of $2 billion through this proposal.

However, the current use of HIT and EHR doesn’t seem to reduce administrative costs in the US as anticipated. Tseng, et al. conducted their analysis within a health care system with a complete EHR system and a centralized billing organization. The authors report administrative costs still had a significant financial impact even in light of these technological advances.

Reducing health care administration costs in the United States could have both direct and indirect impacts on the health care system as a whole. Most obvious would be the potential monetary savings by curtailing wasteful spending accrued through superfluous administrative activities. Providers could also see a positive impact on their relationships with patients, reclaiming the three hours/week spent on administrative duties for clinical care and increasing physician productivity.

US health care prices are the highest in the world. Some of that is due to ever growing administrative costs. So far, we haven’t found ways to keep those costs from rising. Overhauling the US health care system in its entirety seems unlikely. Some restructuring of health care administration could be a plausible first step towards reducing costs without harming patient care.