[Read in black and white]

0.

So there was this radio segment, back in May, for KQED's California Report, called "Sorry, The Therapist Can't See You — Not Now, Not Anytime Soon" by April Dembosky. It's about the difficulties people in California have in getting access to psychotherapy. The whole thing is a half hour, and you can hear it at Soundcloud.

In the segment "Frustrated You Can’t Find a Therapist? They’re Frustrated, Too", which you can read a transcript of at that link, Dembosky explores the fact that the problem's not that there's few therapists...

Insurance companies say there’s a shortage of therapists.

But it’s not that simple. Especially in urban areas, there are lots of therapists. They just don’t want to work with the insurance companies.

[...] Nearly half of therapists in California don’t take insurance, according to a recent survey from the California Association of Marriage and Family Therapists. The same is true of psychiatrists. There are two reasons why, Klein says.

“One, because the reimbursement rates don’t provide a living wage,” Klein says. “You can’t own a home and drive a car and survive on what in-network providers pay you.”

Most insurance companies pay therapists in their networks between $60 and $80 per session. In the Bay Area and Los Angeles, therapists say the market rate for therapy is more like $150 or $200 a session.

That's the moment at which I think Dembosky misses the turn.

I think something rhetorically bad and subtle happened, and Dembosky wasn't hip enough to the insider baseball of mental health insurance coverage to recognize it.

In reality, the one we all live in, the actual market rate is completely irrelevant to the discussion of how much insurance companies do and should pay therapists for therapy. But that's something both therapists and insurance companies know. Invoking "the market rate" is a rhetorical maneuver, something both insurance companies and therapists do to justify their arguments about appropriate compensation. But both insurance companies and therapists know that this rhetorical "market rate" they argue about is a fiction. Since insurance company compensation of psychotherapy is often set as a function of the "market rate" – actually the insider term is "usual, customary and reasonable rate" (UCR rates) and the function is usually a percentage – insurance companies who pay therapists make up numbers that are low so they can pay less, and therapists who take insurance make up numbers – basically "sticker prices" none of their patients actually pay, since they all have insurance paying – that are high so they'll be paid more. Insurers and therapists locked in this rhetorical battle know "the market rate" whereof they speak – and over which they vie – is nonsense, but it's the nonsense the terms of the fight is expressed in.

But as I said, the market rate, whether the actual one or the fictional, rhetorical one, actually has nothing to do what what Dembosky is trying to reveal.

The issue is not that "most insurance companies pay therapists in their networks between $60 and $80 per session" while there's some (according to therapists) market rate of "more like $150 or $200 a session".

The problem is that $60 and $80 per session. Dembosky accepts – no doubt it's what the therapists she talked to said – that the problem with "$60 to $80 per session" is that it is so much less than the market rate. But the core problem isn't comparative.

It's arithmetic.

1.

When I asked, in my poll earlier today, what people thought psychotherapist compensation was like, several people attempted to start from what they guessed was the per-session rate and multiply out by something like 40 hours per week. Good try! (Wrong – but reasonable!)

But not one person who did that guessed a per-session rate less than $100 per hour.

That "between $60 and $80 per session" Dembosky hears about in San Francisco is what I'm seeing in Boston, too. I work for a clinic that pays me a cut of the take, and because my compensation is a percentage of what they're paid by the insurance company for my work, I get to see what they're getting paid. Now, I have few patients and they take very many insurance products at that clinic; I only know for the insurers of the patients I've treated, which is a small subset of the larger whole. But of the insurers that have paid for my work, the highest has been Massachusetts Behavioral Health Partnership (MBHP), a MassHealth (Massachusetts Medicaid) plan: it pays a hair under $86. I understand from some hints dropped that MBHP is the highest paying insurance in the state. The lowest has been Neighborhood Health Plan (NHP), which is paying a hair under $56. So $56 to $86 per session: the range is just a bit wider, but essentially the same.

By the way, we're discussing 90834. That's the Current Procedural Terminology(tm) (CPT) code for "Psytx pt&/family 45 minutes", translation: "individual psychotherapy for patient with or without family present, 45 minutes". (Used to be 50 minutes, until Jan 1, 2013.) Initial (diagnostic/assessment) sessions get paid under a different code with a higher rate, because they take more time; group therapy and family therapy are other codes, with differing rates. But 90834 is the vast bulk of psychotherapists' billing to insurance companies, so that's what is always implicit in these discussions. Nothing nefarious here, but just making it explicit. So now, you can go do your own research.

So psychotherapists are being paid $60 to $80 per (90834) session by insurance companies, not the $100 to $300 per session that some commenters guessed therapists were being paid, generally. In fairness, I didn't ask you what you thought insurance companies paid, I asked you what you thought therapists were paid, and for those of you who have paid cash on the barrel-head yourselves, it never crossed your minds that insurance companies might pay way, way less than a therapist charges patients directly.

2.

Now, most people, when they hear "$60 to $80 per (90834) session", think, "Well, that's not bad. Sure, it's not $200/hr, but it's nothing to sneeze at."

libraryhungry cannily caught the problem there:

I'm surprised by how many people are conflating "hours of work per week" with "hours of paid sessions per week."

Therapists who take insurance aren't working for $60 to $80 per hour, they're working for $60 to $80 per session.

The obvious question to then ask is "how many sessions per week can a therapist actually do?" It's an interesting question, but in many ways the wrong question. People get fascinated by the question of how much time does it take to actually do one psychotherapy session. It's not unreasonable to wonder: the amount of time a therapist has to spend on things outside the session – paperwork, collateral contacts (i.e. conferring with the patient's PCP, psychiatrist, endocrinologist, Methadone clinic, etc.), phone calls, arguing the insurer into paying for services rendered, waiting on hold to argue the insurer into paying for services rendered, etc. – can be huge, and none of it is compensated separately. It is explicitly, contractually included in the per-session fee. When you accept an insurer's terms, you agree that part of what you owe them for that money is all that stuff. They pay you by the session, but they're not just paying for the session. So it makes sense that people would fixate on how much time all that stuff actually takes.

But – and I say this as someone who has stayed at a clinic doing paperwork to midnight – that's a very secondary or tertiary concern.

The actually relevant question is subtly different. It's "how many sessions per week does a therapist get to do?"

There's three dimensions to the answer: schedule preferences, patient density, and non-attendence.

3.

In another part of the same radio segment, "Single Mom’s Search for Therapist Foiled by Insurance Companies", Dembosky relates a little investigation:

To find out, we decided to conduct our own survey and called all the psychologists — 100 in total — that were listed on the Blue Shield website for Natalie’s plan in San Francisco.

Here’s what happened:

[Original cute graphic here.] We called 100 psychologists in San Francisco.

Of those, 22 never called back and 2 had wrong numbers. That left 76.

48 said they weren't taking insurance or accepting new patients. That left 28.

And only 8 had appointments after regular work hours.

The end result: 28 psychologists actually had appointments. And only eight of them had slots available outside regular work hours. Eight out of 100.

/recordscratch

I have a whole separate post in the pipeline about the use of her phrase "regular work hours". But let's take it as given, and observe how Dembosky considers any psychotherapy appointment availability during "regular work hours" as effectively non-existant.

She's not wrong to – a psychotherapy appointment you can't go because your livelihood won't permit it is not actually a psychotherapy appointment for you – but let's look at the logical consequences of her assumptions. How many outside-of-regular-work-hours appointment times are there in a week?

When people ask about availability of appointments "outside of work hours" they usually mean weekday evenings. While Dembosky says "outside regular work hours" at the end, she describes it above as "after regular work hours". When people think of psychotherapy sessions when they're not at work, they usually think of them as being after work.

They do not, however, mean "Wednesday at 11pm", right? No, they mean something conveniently after work, for a workday assumed to end at 5pm, and have a bit of a commute to the therapy appointment. In other words, the latent assumption of when therapy should happen – the "when" had in mind by prospective clients calling therapists, looking for appointments – is 6pm to 8pm, Monday through Thursday. Since the actual therapy session takes a little under an hour, most therapists schedule them on the hour (and use the in-between time for paperwork, bathroom breaks, and checking phone messages). So, one appointment at 6pm and one appointment at 7pm, on each of four days of the week.

There are eight possible patient sessions between 6pm and 8pm, Monday through Thursday. Eight.

Let's call those the premium slots. Obviously, nobody is making a living on eight sessions a week; therapists have to try to book patients into other, less desirable, time slots.

Well, there's other times that are "outside of work hours". Maybe the therapist can convince patients to take appointments on Friday nights (another two slots), or at 8pm weekdays (another five slots), or at 5pm weekdays (another five slots). That gets us up to a total of 20 slots, half high demand, half lower demand, but still, technically "outside of work hours".

When else? Well, some therapists offer appointments before work hours. That generally means 7am weekdays. I gather those five slots are about as in-demand as 8pm weekdays. You could offer 6am and 8am weekdays (10 more slots), but they are in even lower demand. So weekday mornings yet you 5 second-tier slots, and 10 third-tier slots.

Note that working to 9pm (when an 8pm session ends) and working at 6am is kind of interesting. Few people try to do both.

What other outside-of-work times are available? Weekends. Some therapists offer Saturdays. I understand these are second- or third-tier desirability; there's basically 8 on Saturday and 8 on Sunday (9am to 5pm).

Finally, you could offer sessions after 9pm on weeknights and after 5pm on weekends. Those are fourth-tier desirability.

So, "outside of work hours" actually means: 20 slots after 5pm, 15 slots before 9am, and 16 slots on weekends, for a total of 51 hypothetical "outside of work hours" slots.

And look what working that schedule would be like: at work by 6am, having an unprofitable 8 hour gap, then working another four hours to 9pm. It is not possible working that schedule to get eight hours sleep per night. If you have any sort of commute to your workplace, either you chill your heels at your office or clinic for eight hours, or you double your commute. No days off at all – it's a seven days a week schedule. It means never seeing any of your own family who keep conventional hours themselves.

In short, it's impractical. So most therapists combine some "outside work hours" slots with slots during work hours.

This is, of course, incredibly frustrating to patients who want to see a therapist "after work hours". Those premium slots – 6pm to 8pm, M-Th – go like hotcakes, and then all the therapists have left to offer are less convenient, less desirable "outside work hours" slots, or slots during work hours.

I'm frankly astonished that of the twenty eight therapists they found taking patients, more than a quarter had evening slots. That's amazing.

Dembosky's team had to call 100 therapists to find eight who had evening hours? A little worse than one in ten? A couple of years ago, my 4pm Tuesday slot opened up, and I sat down with the clinic's waiting list. I called down ten names before I got to someone who still wanted therapy, and could meet at 4pm on Tuesdays.

Thank goodness I worked at a clinic, and so there was a waiting list. Imagining that in private practice – self employment – and having to wait through nine prospective patient contacts before getting one that could take my one schedule opening. Imagine being self employed and losing out on nine leads because nobody wants the non-premium slot you have available.

The world is full of therapists – at least the US is – and they'd mostly all be delighted to only ever see patients between 6pm and 8pm, M-Th. But you do realize, to do that, we would have to charge one eighth of our necessary weekly revenues for each of those sessions, yes? That is, figure out what the therapist needs to earn per week, and divide that by eight.

Here, I'll help: Fanny Mae used to have a guideline based on actuarial studies of consumer financial stability: one week's pay (gross) is one month's housing. Their studies showed that if you don't earn your housing costs in one week – that is, if you are putting more than a quarter of your income to housing – you're not a good risk for a mortgage that big, because you are living too close to the edge, and can't really afford one.

Right now, in the urban Boston area, studio (one-room) apartments are going for about $1,500/mo rent. So to meet the Fanny Mae guideline to be able to "afford" a minimal apartment for one's self – no dependents – a therapist would have to have earn $1,500/week.

Fifteen hundred divided by eight is $187.50 per session.

(But that's actually bogus. I just conflated revenues with income. It's actually much, much worse.)

It most definitely is not $60 to $80 per session.

And here we come to one of the ugly bits of the industry nobody talks about.

Insurance companies want to keep the number of therapists they allow to "take" their insurance products to a minimum. The obvious reason is that if there's no therapists available to treat you, they don't have to pay for your treatment, and it saves them money. The non-obvious reason is to pressure patients into taking those non-premium schedule slots.

If insurance companies took every therapist, then every therapist would be able to rely on filling exactly eight slots – anything beyond that is dicey. Therapist doesn't have a premium slot available? Go to another one. By constricting the consumer's choices to a limited roster of therapists, those therapists are in the position to say, "Look, I have 4pm Tuesday open. Take it or leave it."

This is brutal on the prospective patients, but the underlying problem isn't going away. Therapists have limited high-desirability time slots because that's all there are. The only ways for therapists to make a living are either to be paid enough for high-desirability slots that they don't need to see patients at other lower-desirability slots, or therapists have to see patients at times less convenient, or rather quite hard, for the patients.

You can compromise to some extent, but the therapist who books patients 5pm to 8pm (ending at 9pm) still has only 20 slots available per week, so now we're dividing necessary weekly revenues by twenty. So does that make $60 to $80 per session work?

This brings us to the next problem.

4.

Imagine you are a prospective patient, and you have talked to two therapists who both tell you they're taking new patients, and who both work the same 20 evening hours – all the rest of their scheduled work time is during work hours, so as far as you are concerned, that part of their schedule doesn't exist.

One of these two imaginary therapists just started, and has no patients in those twenty time slots. "Pick any one you like," he says. The other of these two imaginary therapists is booked up to his eyeteeth. He has one of those twenty slots open.

What do you think the likelihood is that you'll be able to make the schedule work out with the therapist with the one slot? What do you think the odds are that you'll go with the therapist with twenty open slots over the therapist with one?

Or put another way, from the therapist's perspective: if your only slot left is 8pm on Friday, what are the odds that you will be able to fill it?

Schedule flexibility increases the likelihood that a therapist can meet your schedule needs. But the flip side of that is the more a therapist's schedule is booked with patients, the less schedule flexibility the therapist has; and the less schedule flexibility a therapist has, the harder it is for them to book more patients.

So there becomes this really interesting question of whether a therapist working the 20 evening sessions actually can book that twenty-slot schedule tight.

There is this asymptotic scheduling problem where the more patients you have the harder it gets to book your remaining slots.

So a therapist can decide "I'm going to work N session-slots per week", but in doing so, they're not magically going to book N sessions per week. They may, if all else is equal, asymptotically approach N sessions per week, but booking fully to N is hard-to-impossible.

(As an aside? I may have hit N on my Tuesday schedule. I am veddy, veddy close to being booked solid, and have a patient maybe coming back after an absence, and I think she can take the one slot I should have coming open in two weeks. This is me bragging. But being realistic here: two parts 133t schedulemancy, three parts dumb luck.)

A therapist's ability to book as close as possible to N, for whatever N they pick, has a lot to do with the demand for their services in their locale.

There's a lot of things, aside from desirability of schedule slots, which mediate potential clients' demand for a therapist's services. The obvious one is "do you take my insurance?" and slightly less obvious, alluded to above, is "how many other therapists around here take my insurance?"

But then there's also just straight up population density. If you practice in a big city, there are simply more people around who might want your services. This is why rural areas are starved for therapists. Way out in thinly settled areas, there often simply aren't enough human beings around to keep a therapist busy.

If you're a therapist and all you have is 10am Monday free, your odds of filling that slot are way better if you work in a city of a million than a town of ten-thousand (to say nothing of a village of 500). People here get confused by the concept of "per capita". The per capita number of therapists may actually be higher in a big city than a small rural town – that is there may be more patients per therapist in the small town – but that's largely irrelevant. If there are 10 people who want a therapist in a town of 10k, and 1000 people who want a therapist in a city of 1M, the per-capital demand is identical, but my odds of there being somebody who wants 10am Monday just went up by a factor of 100.

What all this means is that however many patient sessions a week a therapist is open to booking, they aren't going to get that many. They're going to have some percentage of schedule slots standing open, because getting patients into the corners of your schedule is hard. Therapists can optimize for this by sticking to the urban cores, and avoiding low population density areas – which we all generally do.

So I just answered another part of "why you can't find a therapist where you live" if you don't live in a big city.

To recap: there's a much more limited schedule of desirable slots for offering therapy than perhaps you realized, and then when a therapist tries to populate a schedule there's things that make it hard to fill.

To put it another way: Therapists can offer fewer (and less desirable) sessions a week than you might have realized. Second, of however many session slots they offer (i.e. their work schedule), they will book even fewer patients that that.

But wait, it gets worse.

A lot worse.

5.

So let us say that, despite all the forgoing, somehow a therapist manages to book a forty-hour work week full of patients. Yes, actually has booked every one of forty slots in a week.

Therapists who accept insurance don't get paid for patients they don't see. Doesn't sound controversial, no? Allow me to explicate: all(?) insurances contractually forbid the therapist for billing the insurance company for missed appointments. Most insurance companies contractually forbid the therapist from charging the patient a missed appointment fee.

And that's just missed appointments – which, btw, are conventionally in the biz not just missed appointments, but appointments canceled with less than 24 hours notice – and those aren't the only appointments patients don't attend.

So far I have followed the conventions of how discussions of therapist compensation go: discussing patient bookings as if we were discussing the sale of magazines at a newstand.

But we are not selling individual issues – though that is exactly how insurance companies pay – we're providing ongoing services.

If I book a patient onto Thursdays, come late November, either that patient reschedules to another day – assuming I have an empty slot that works for their schedule – or I'm not seeing them one week. Or any other patient on Thursdays.

Therapists don't get paid holidays. Or paid vacation, or paid sick leave, or paid personal days. You see patients, or you don't get paid: that's true whether the patient pays cash or you're being paid by insurance companies.

But that's just therapist time off: the patients aren't usually too keen about keeping appointments on Christmas or the Fourth of July, either.

And they do things like go on vacations. They take individual weeks off therapy to deal with other things in their lives. Nobody expects patients to attend therapy every week (or on every day of their regular schedule); it would be absurd and unreasonable to.

But when your patient takes four weeks to go hiking the Appalachian Trail, their schedule slot is empty for those four weeks. It's not like they're doing you some sort of wrong. But what are you supposed to do with a slot that's just four weeks long? You can't intake another patient for that slot (unless somehow you know the patient only needs four weeks of treatment), unless you have some other provision in your schedule for the ongoing care of that patient.

So in other words, not only do therapists necessarily book fewer sessions than they offer, therapists necessarily see fewer patients than they book.

How many fewer?

That depends somewhat on your patients' insurance. There have been studies, though many have problems. I gather patients who pay out of pocket have the highest rate of attendance and patients with insurance they get through an employer have a slightly lower rate of attendance. I'm unclear exactly what that rate is.

But the one thing that has been studied the most, and is best known to practitioners, is the rate of non-attendance of people who get their insurance through the state, usually for being poor. Medicaid patients.

Medicaid patients attend, on average, about 67% of their psychotherapy sessions. Yes, studies consistently show that Medicaid patients do not attend one in three of their psychotherapy sessions.

This is not because poor people are flaky or bad at being patients or don't understand the concept of the appointment – though all those things can be true, too.

There are three big classes of reason. First off, the poorer you are, the less control you have over your schedule. If you have a low status job, you don't get to tell your boss, "No, I can't work late tonight", if you don't have disposable income, you can't just get a sitter for the kids. Second, not having disposable income means your ability to keep commitments is thwarted by how brittle the conditions of your life are to mishap: not having money to spend on car maintenance means your car is more likely to break down, and you can't just take a cab if the bus never comes. Third, sometimes – often, in mental health care – the reason someone is poor is that they have a disability, which makes it hard to attend appointments, either directly, by the symptoms impairing their ability to make the trip, or because they have so many medical appointments schedule management gets hard.

(So I've just answered the question "Why can't I find a therapist who takes Medicaid?")

This level of non-attendance is so well known in the mental health industry, that the assumption is simply that one third of sessions won't be attended. At clinics where therapists are hired "full-time", the terms of employment includes a minimum quota of patient sessions seen (billed) per week. The industry standard quota for a full time therapist is twenty-six sessions per week – because 26 is a third less than 40.

Yes, the industry assumption is that a "full-time" therapist is booking 40 patients a week, and seeing 26.

At this point, I hope you understand that the question of how much time it takes at therapist to do all the attendant work for a patient session is basically immaterial. If you're dedicating a 40 hour work week and only getting to bill for 26 sessions, it doesn't matter that it takes you one hour of every two to just do paperwork. That's okay: you have the time.

6.

To see 26 patients a week, a therapist must book 40 patients a week. But to book 40 patients a week, a therapist must offer more than 40 slots a week. How much more? I'm not sure – I've never tried to do it.

But I know of a few therapists who simply worked 60 hour work weeks. Ten hours a day, six days a week. And yes, they attempted to book all 60 hours, though of course they could only asymptotically approach it.

As an aside, to spell out what is implied by a 40 patient sessions booked per week schedule: most insurances only allow one session per patient per week. (There are exceptions.) That means booking 40+ patient sessions per week means having 40+ patients.

And, actually, many insurances will only pay for every-other-week sessions unless the patient is very ill. To book 40 every-other-week patients per week, you need a caseload of eighty patients.

I talked to a colleague who had a 60 hour per week schedule, and he had something like 90 open cases. Ninety patients. Ninety life stories, ninety family trees, ninety evidence based treatment plans, ninety courses of treatment to keep track of. How did he do it?

Apparently – according to ex-patients of his whom I have treated – he didn't. He had no idea of whom he was talking to session to session, or what their lives were like.

Also, he burnt out and no longer works as a therapist.

---

This post brought to you by the 106 readers who funded my writing it – thank you all so much! You can see who they are at my Patreon page. If you're not one of them, and would be willing to chip in so I can write more things like this, please do so there.

Please leave comments on the Comment Catcher comment, instead of the main body of the post – unless you are commenting to get a copy of the post sent to you in email through the notification system, then go ahead and comment on it directly. Thanks!