Thursday, October 14, 2010

Weighing Costs and Benefits Part IV: Costs

Some people quit taking meds that their doctors believe will relieve their symptoms of mental illness.  Why?

Because the meds don't work, because they can't afford them, because the meds make them sick.

Manifesto:

For any of these reasons, people who quit are making intelligent decisions in their own best interests.

On The Other Hand 

Sometimes the meds do work.  Sometimes people have decent health insurance with good drug coverage.  Sometimes the side effects are not as bad as the disease.  In that case, those who quit their meds are stupid.

Let's just get that right out front.

Moving On To The Costs

Today my series on weighing costs and benefits turns to the costs.  The costs do not tell you whether you should try a medication.  They simply give you the odds.  It is up to you to decide how you want to play the odds.  I calculate the odds based on the numbers of those who quit.  Those who consume have the best information about costs, what actually happens when they put these chemicals in their own particular test tubes.

How Many Of Us Are Noncompliant?

Out of 100 prescriptions that providers write, 10 consumers never consume.  They don't show up at the pharmacy at all.

28 consumers quit within the first month.  That includes those first 10.

50 quit within 60 days.

72 are outta there at six months, 78 within the year.

That leaves 22 compliant consumers.

How Do Noncompliant Consumers Explain Their Decision?

10 out of the 78 don't.  Providers failed to close the sale.  Providers would be interested to know why these 10 are pharmacy no shows, because it might help them improve their pitch.  Their assumptions are that it was because the consumers didn't understand, or the providers didn't establish trust, or that good old back up -- stigma.  But often, consumers don't report their decision.

We could invent reasons, which might be fun, top ten list, that sort of thing.  The drinking buddy said, Buck it up.  Real men don't get depressed.  The transmission fell out of the car on the way to the drug store.  My favorite -- the primary care physician said, Are you kidding?  With your blood glucose and lipid levels?  Does this so-called doctor even own a blood pressure cuff?  However, all this speculation is just that.  These 10 do not give us information about the costs of taking the medication, because they never take it.

So now we have 68 consumers who quit after they tried the meds.  AK Ashton et. al. actually asked them why.

30 (out of the 90 who actually filled the prescription) say they quit because they could not tolerate the side effects.

30 say the medication was not effective.

That already adds up to 70 nonconsumers, counting the nonstarters and leaving eight who quit for other reasons.  I will suggest some of these other reasons, and you will have to come up with the odds yourself that any of them might put you among these 8.  (They may have reasons similar to the 10 who never started.)

And by the way, these numbers vary by how many different medications the consumer has already consumed, which primarily affects the efficacy number.  They also vary by which medication is currently being considered, primarily effecting the side effect number.

We don't have all the numbers we need.  Somebody needs to be collecting this data.  A consumer group, looking at real world data over the course of a year, not the guys with 6-8 weeks of information, seeking FDA permission and doctors' cooperation to sell pills.  But the algorithm itself will work for whatever the numbers turn out to be. 

Let's Start With Side Effects

30 of the 68 who consumed and quit say they quit because of side effects.  The clinical trials, lasting eight weeks or so, report much lower numbers.  The numbers the providers give you are from the clinical trials.

The common belief among providers is that they could improve compliance by giving consumers more information up front about side effects.  Small isolated studies sometimes confirm this over the short haul.  But this belief does not stand up to more research and more time.

Up front discussion of side effects can give the consumer strategies for dealing with insomnia, reducing nausea, preventing falls when they get out of bed.  These are the side effects we notice immediately.  Maybe they are tolerable if you have social supports to get you through the roughest first weeks.  Sometimes your body does  acclimate, and the immediate side effects become less bothersome.

But sometimes these strategies don't work.  Social supports wear out.  Mom has to go home and stop helping you with the kids.  You run out of sick leave.  The body does not adjust.  And sometimes these side effects are indications that you are taking the wrong medication!

But the major side effects appear later.  Which are the most bothersome?  The results: weight gain (31%), erectile dysfunction (25%), failure to reach orgasm (24%) and fatigue (21%).

Weight gain -- a few pounds in the first few months are not a problem.  You hardly notice.  But over the months, when you are moving from overweight to obese, you get a reality check on what this medication really costs.  Morbid obesity takes 8-10 years off your life.

Tell that to your psychiatrist when you complain and he/she says you have to weigh your costs and benefits.  Your doctor may not even know about how serious the health risks of obesity are.  Obesity even increases the risk of dementia.  But psychiatrists treat psychological problems with pharmacology.  They do not treat your heart, pancreas or liver.

Then there are the sexual side effects.  When you started the medication, you weren't getting much anyway.  That was one of the symptoms -- loss of interest in formerly pleasurable activities.  But six months later when you're not getting any, you (and your partner) recalculate your costs and benefits.

Hence, these noncompliance numbers go up over time.

Side Effects In The Algorithm

The major competition between makers of psychotropic medications has always been on this side effect issue.  It turns out, we just won't keep taking stuff that makes us feel worse.  So sometimes you can find studies that pit one against the other and get real numbers about side effects.

STAR*D found that in just 8 weeks, a combo of lithium/sertraline (Zoloft) got an intolerable rate of 45%, 2-5 times any other treatment.  Effectiveness rate -- 9%.  I wonder how many of the 91% who didn't get better would have been better off if they had taken nothing at all.

Or to put a finer point on it, did lithium/sertraline make matters worse?  They didn't test against placebo, so we don't know.

If the odds of harm are five times the odds of help, I will give it a pass.  That is like rolling the dice, looking for one particular number.  Only it's not dice; it is my body.  That is my personal decision, made after my eighth trial.  It is up to you how you play the odds.

For the sake of the algorithm, SE means the odds that you will quit taking this medication within a year because of side effects.

Efficacy -- What If It Just Doesn't Work?

We already discussed effectiveness in detail on September 2, Weighing Costs and Benefits Part II: Benefits.  Go back there for the details.  It makes more sense if you know the back story.  In summary:

Efficacy for Number of Present Trial (E#PT) means how many people got better with this med after they tried a number of others that didn't work.  Non-Spontaneous Recovery Rate (NSR) means how many people would not have gotten better if they had simply waited for the depression to go away on its own.  Efficacy for Number of Present Trial times Non-Spontaneous Recovery Rate equals Short Term Benefit (STB).  Those are the odds that it will work.

Or, E#PT X NSR = STB.

The abbreviations are there to make me look smart.  Which, as a matter of fact, I am.  Some days, I can make the smart parts of my brain connect  again and actually work smart.

Another way of looking at it: STB is a number between 1 and 100.  That many times out of a 100 are the odds that you have come up a winner.

So then the odds that the medication will not work are 100 minus Short Term Benefit.  We will call that Not Effective (NE)100 - STB = NE.  You have wasted your time, and are more discouraged than ever.  Bummer.

Now you may have noticed, the algorithm calculates the Short Term Benefit for eight to twelve weeks.  And the Short Term Cost refers to one year.  Why the difference?  Because you will likely be one of the early quitters (50%) if you don't get relief by twelve weeks.  And if you do get relief by then, you are likely to keep taking the medication for a year.  It may quit working for you eventually.  But you are probably good to go for a year.  Hence, twelve weeks for STB and twelve months for STC are probably equivalent measures.

Efficacy -- What About Those Who Quit Before They Gave The Medication An Adequate Trial?

I did not consider how many reported that they discontinued because the medication was ineffective, the 30 out of 90 that Ashton, et al, discovered in their survey.  This number is not helpful, because some of these 30 quit before the full 60 days needed to determine efficacy.

Instead, I used the efficacy numbers reported from the clinical trials.  As a result, those 8% discussed below is a larger group.  It would include the early quitters, because some of them might have gotten better if they had been more patient.

But these numbers are for illustration purposes only.  The algorithm is designed to be general, so that you can insert whatever the numbers turn out to be.

If you quit simply because the medication does not work faster than it works, and for no other reason, then you go into the stupid category.  Just to get that right out front.

Other Costs

8% (plus) quit taking the medication primarily for other reasons.  I expect that money, stigma and trust are the the big ones, with stupid in there, too, as stated above.

Money

Let's face it.  These medications cost money.  There are two costs to consider.  The first is the pills themselves.  The provider may provide you with samples, if yours is the newest wonder drug being promoted this week.  The samples likely last for two or three weeks.  This is good, if it helps you determine early on that there is no way you can tolerate the medication, even long enough for some of the side effects to become less troublesome.

On the other hand, it does not help you determine whether the medication will be effective.  That takes more time and your own money, a lot of it, if yours is the newest wonder drug being promoted this week, which you can count on, if you have failed to prove your provider a genius by getting better with his/her first or second choice.

If you have a good drug benefit, cost of medication may not be a major issue.  I now get my generics for free.  I represent a very, very small portion of the U.S. population.

I used to have insurance with a high deductible and mediocre drug benefits.  After the samples ran out, I paid $120 for a two month supply from a company that required I buy from them by mail order.  By the time the pills arrived, I had already discovered I couldn't tolerate the med.  I never even opened their bottle.  Meanwhile, I had to pay through the nose at my local pharmacy for the two weeks it took me to taper off.

In addition to the medications, you will pay for medical management, trips to the provider who will monitor your condition and tweak the chemistry experiment.

Again, these costs will vary by insurance plans and whether you have insurance at all.  With my current insurance, I pay $5/visit.  In my previous plan, I paid $40.  If I had no insurance at all, the cost would be $135.  And I see my doc every six weeks on average.

I cannot assign a number to the odds that you will quit a medication because of how much money it will cost you.  That is your call. Out of 100, what are the odds that you will quit because you cannot afford it?  We will call that $$$ in the algorithm.

Stigma

Okay, if you have made it this far through the Costs and Benefits series, you ought to be motivated enough to resist those who shame you (including yourself) for relying on a pill, for being weak, for being sick... whatever garbage they throw at you and you throw at yourself.  Please let's get over it.  I hope your stigma number is low.  But again, that is your call.  Out of 100, what are the odds you will quit for reasons of stigma? -- STG.

Trust

Next, out of 100, what are the odds that you will quit because you cannot find a provider you trust with your body, or because you think the pharmaceutical industry is corrupt? -- TR.

Stupid

Stupid is a side note.  Providers prescribe the medication because they already believe that the benefits outweigh the costs.  So they expect the stupid category is a large proportion of the noncompliers.   Only they call it confused.

Stupid is irrelevant to the algorithm, which is designed to weigh costs and benefits.  So stupid (or confused) is not in there.  Like stigma, stupid can be fixed.  But it is not a cost.  It is a prior condition.

Down And Dirty Costs

So now we simply add the odds of each of these costs together:

Side Effects plus Not Effective plus Money plus Stigma plus Trust (lack thereof) equals Short Term Costs.

SE + NE + $$$ + STG + TR = STC.

So How Do You Decide?

STC versus STB give you the odds.  Once more I repeat, they do not give you your decision.

We will look at a couple other issues and pull this all together in our next installment.  Whew.  My brain is about to explode.

Flair from Facebook
Clipart from Microsoft
Photo of die by Roland Scheichder, in the  public domain
Photo "Solution" by Salvatore Vuono
Photo "Angry Father" by Akapl616.  Permission is granted to copy
under the terms of the GNU Free Documentation License
Photo of "Tired Man" by graur codrin
Photo "Aces" by Felixco, Inc.
Photo "Loneliness" by graur razvan ionut 
Photo of Pristiq by Tom Varco.  Permission is granted to copy
under the terms of the GNU Free Documentation License
Mademoiselle Zizi Feints at Fainting, by John Sloan

1 comment:

  1. Interesting to see the compliance rate numbers.

    I spent hours doing research on the medications after I was diagnosed with Bipolar 2. I could not get much information from my psychiatrist during the short time allotted for appointments, even when I arrived with specific questions like "what is the half life of this medication?" I had to rely on what little bit I could find at my local library as this was before the days of google!

    It was a slow tedious process for 2 years, with my body as the lab rat. Fortunately I kept very detailed notes, as did my psychiatrist.

    I like the algorithm approach, although math is not my strong suit. Will be interested to see your further work on this approach.

    ReplyDelete