National Guideline Clearinghouse Goes Kaput

Although the practice of medicine has existed for thousands of years, it substantially improved with the implementation of the scientific method. Experiments and research studies improved diagnosis and treatment. Now so much information is published that no person can read everything unless, possibly, it’s limited to an extremely narrow subspecialty.  In addition, different studies can come up with opposing results, and it can be difficult to make sense of all the available information.

To remedy that, various groups have published guidelines to help clinicians decide what to do. For example, new guidelines for high blood pressure were recently published. The American Diabetes Association just updated their guidelines for Standards of Medical Care in Diabetes.

So how does one find out about existing guidelines, other than doing a web search or coming across it in a journal? Well in 1998 the National Guideline Clearinghouse was created. It formed a collection of guidelines that met minimum quality criteria. By June 2018 there were more than 2000 guidelines listed that could be searched by specialty. In July of 2018 all of that information became unavailable on the website because of federal government budget cuts.

The website was originally created by the Agency for Healthcare Research and Quality (AHRQ), in partnership with the American Medical Association (AMA) and the American Association of Health Plans (now America Health Insurance Plans).

In the last year of operation, the National Guideline Clearinghouse’s budget was about $1.2 million dollars. This is only about 1% of the money spent globally on developing guidelines, and an even much lower percentage of the cost of medical care. The guidelines can improve care and save money, but only if people can find them. Both my company’s electronic health record and my county medical society’s website have the National Clearinghouse Guidelines integrated to reach them with a click. I’m sure we’re not the only ones who routinely used it.

Perhaps a better repository can and will be built, but in the meantime I think the government should fund the National Guideline Clearinghouse and bring it back online.  This was not a case of trimming fat from the national budget, but a self-inflicted stroke where the government cut off the blood flow (money) to a portion of our collective brain. We’re the worse for it.

ICD-10 Keeps Getting More Painful

As I previously discussed,  a year ago we transitioned from the disease classification ICD-9 to ICD-10. That has been painful, but they keep making tweaks that require more work.

I guess the powers that be decided that more than 155,000 diagnoses were not enough when they recently changed many diabetes diagnoses (a day or two ago, at least, my organization implemented the latest edition). Now it’s no longer sufficient to say that someone has Type 2 Diabetes Mellitus with Diabetic Neuropathy [E11.40], for example, but I now have to specify in addition whether it’s with or without long term insulin use, or if it’s unspecified. That means all my carefully constructed Problem Lists on my patients no longer work. Every diabetic medication I reorder will have to be changed as they are associated with a diagnosis.

Across all my patients I’d estimate that’s close to 1000 changes I will need to make. Assuming it takes me 30 seconds each time (I’m probably a lot faster than most of my colleagues) that’s over 8 hours, so a full work day. Multiply that across all the primary care doctors and that’s a lot of time – about 1000 people working years! We have a shortage of primary care physicians and I think there are many better ways to spend our time.

I typed “type 2 diabetes mellitus” into my electronic medical record. I eventually scrolled to the bottom to see a message that there were 3158 diagnoses loaded, but that the results had been limited due to it being a common phrase! Many of these were synonyms, and one can save favorites, but I think it’s ludicrous that we have so many codes for just one disease. Those who promulgated moving to ICD-10 claimed the higher specificity would lead to all kind of advantages by being more precise, but in reality physicians can’t spend all day just to pick a diagnoses and they are going to pick something close that will satisfy the billing system. For many diagnoses you can’t even get precise agreement. There are various codes for uncontrolled diabetes, for example, but if you ask different doctors what that means, you’ll get different answers.

Patients with diabetes have to suffer from complications of their disease, increased medical costs, and being stuck more often for blood or injections. It’s too bad their physicians have to suffer more as well.

Asinine Associations

As I previously wrote, when physicians place orders, they have to associate diagnoses. This is becoming even more painful as we move towards ICD-10, of which I’ll have more to say later.

I’m sure this was an attempt by the government to save money, but in the vast majority of cases the ordering physician has no secondary gain, and they order the test because they think it’s the right thing to do. I can understand it for some expensive tests or procedures, but many are just plain obvious.

I think lawmakers should have a taste of their own medicine. When they need office supplies, they should have to give a reason. Here, I’ll help them out with a few items to help them understand how it works:

StaplesTo attach separate pieces of paper.

NotepadTo write down information.

PenTo apply in conjunction with a notepad to convey information.

ChairTo help counteract gravity to prevent leg and back pain and fatigue.

Laser Printer TonerTo print out things using a laser printer.

Light BulbTo counteract darkness.

Epic SmartPhrases

If you do not use the electronic medical record Epic Hyperspace, this article is probably of no interest to you.

I previously wrote that I have done a lot of customizations to Epic. In this post I’ll explain how I use SmartPhrases. These are text, ranging from one word, to multiple pages of material, generated by typing the name of the SmartPhrase, preceded by a period.

My approach is to be modular in creating SmartPhrases, as I’ll demonstrate below. I also don’t like to pull in information into my note, such as past history, labs, etc. as the information is already in Epic and it just clutters up the note (I do bring in much of the information for physicals because I think that’s the one time it’s useful to have everything in one note. If acting as a specialist and doing a consultation I might do the same.) Unless I’m doing a physical, which has its own scripts, my baseline script is .soap which looks like this: Continue reading “Epic SmartPhrases”

Medication Small Print

WhenSolu-Medrol_crop I give a cortisone injection, I have to document it in our electronic medical records. I’ve always included the dose, how administered (intramuscular), and the lot number. This week my company added the requirement that we include the NDC number, as insurance companies wanted the information.

It’s just one more administrative requirement, but what really makes it bad is trying to read the number off the bottle. As you can see from the photo, the font is very small! I suggested the policy was age discrimination, but that didn’t get far.

Dangers of a Transparent Medical Record

Over the years there has been a push to allow more patient access to their records. Patients in our system that have signed up for MyChart with our electronic medical record Epic, can see most of their test results shortly after they are back, even if not yet reviewed by the physician who ordered them.

To avoid patients misinterpreting or reacting badly to the results, certain tests, such as HIV and pathology, are not automatically released. Although releasing the results automatically acts as a safeguard for abnormal results that may have been missed by the physician, it also opens the possibility of patients inappropriately acting on the results.

I’ve had patients make changes in their medications after seeing their results, and without even discussing it with me until their next visit. One patient who saw that his sleep study showed he had sleep apnea, borrowed a friends extra CPAP machine, adjusted the settings based on his research on the internet, then tried it out for 2 weeks!

As the saying goes, a little knowledge can be dangerous. Patients should use the information in their charts to inform them, and help ask educated questions, but not to replace their physician. There is often a lot more to making a diagnosis and deciding on a treatment than just looking at test results.

Epic Customizations

Epic is the electronic medical record (EMR) I use at work (actually my company calls it MultiCare Connect). There are a number of customizations one can do to increase efficiency, and I’ve done a lot. Keep reading to learn just how much.

There are SmartPhrases. These are shortcuts to write out text. Instead of writing “past medical history” one can write “.pmh” and it will automatically enter the three words. A SmartPhrase can contain other SmartPhrases enabling one to generate the shell of a note with a few key strokes, then just fill in the portions that cannot be automatically added.

There are SmartLists. This allows one to basically use a drop down box to select one or more options. Although I have 55 of them, each one consists of multiple entries. For example, if I want to enter the specialists a patient sees, I may select the cardiology and gastroenterology SmartLists, then select the physician the patient sees from each list.

There are a number of different Preference Lists. This is typically used to order various things and save them in the way you want, to save time later. In my Medication Preference List, for example, for an antibiotic it may say to take the medication twice a day until finished and include the proper amount of pills. In the description I write that it’s an antibiotic so my patients know what the medicine is for, and to lessen the chances the pharmacist may misread it. I also include an end date so the medication won’t show up on the patient’s medication list after they have finished taking them. It’s a bit of work, but once saved, it’s very quick to use in the future.

SmartText is kind of a SmartPhrase tied to certain situations. I’ve probably written more than one, but there is no easy way for me to look up which ones I’ve created.

SmartSets allows one to set up templates to do such things as place multiple orders and associate them with diagnoses and notes, and basically speed up various paperwork we have to do. Unfortunately a few years ago Epic made it much more difficult to write or edit SmartSets, so I’ve pretty much stopped working on them.

Letter Templates are just like they sound. I have one to tell women their PAP smear was normal, one to ask their employer to excuse them from work, one asking for a patient to be excused from jury duty due to their medical condition, etc. Unfortunately, as with the SmartSets, it’s no longer easy to write new ones or modify my existing ones.

For the medication dictionary, not only can I add words, but I can set it to auto correct words. If I type “referal”, for example, it will automatically change it to “referral”.

I use Dragon NaturallySpeaking to dictate parts of my notes, but I go beyond with custom scripts. For example, if I say, “order anemia panel” it will enter the proper codes for a CBC, iron/TIBC, ferritin and vitamin B12.

I’m constantly updating, but as of last week, these are the customizations I’ve done in each category.

Tool Number
SmartPhrase 1203
SmartText 1
SmartSet 28
Letter Template 16
Dictionary 5861
Preference Lists
Charges 4
EKG 3
Office Visits 1
Education 3
Immunizations/Injections 13
Labs 335
Imaging 170
Medications 3594
Orders 3
Procedures 41
Referrals 278
Supplies 2
Dragon Scripts 302
Total 11914

Each of the 11,914 items is a customization. It may be as simple as a word added to the dictionary, or represents paragraphs of text, a list of hundreds of items, or dozens of lines of computer code. Having been on Epic since 1998, that means I’ve averaged about a 1000 customizations a year.

All these customizations makes Epic very powerful, but unfortunately it was not designed well to share. Many of the items, such as Preference Lists, can be shared, but only by individuals importing someone’s list. If someone imports my Medication Preference List, it goes out of date as soon as I make a change. My list is so long it may take 5 minutes or so to import the list, and even if faster, most people are not going to remember to import the list regularly. It’s like backing up one’s computer. If not set to do so automatically, most people won’t do it. In addition when one imports someone’s list, it doesn’t show where it came from. I think it would be far better if people could subscribe to preference lists similarly to how one follows people on Facebook or Twitter. My medication preference list was designed for internists seeing adult patients. A family practitioner should be able to subscribe my list, to use on their adult patients, and another list to cover their pediatric patients. It is difficult to share customizations within my own company, and far harder still to share with people in other medical groups. Consequently thousands of people have to reinvent the Epic wheel.

Viewing Doctor’s Notes

Should patients be allowed to see doctor’s notes? Legally they can, but that doesn’t necessarily mean it’s a good idea. Patients would like to, but physicians are not so sure. People make strong arguments, but I think it’s really nuanced.

Physicians often write down the differential diagnoses. Say you’ve lost a little weight without trying. It could be a lot of things such as stress, cancer, an overactive thyroid, an ulcer, HIV AIDS, tuberculosis or a thousand other things. After asking a number of questions and doing an exam, I may decide that it’s unlikely that there’s any serious medical problem going on and prescribe a medication for depression. When I see you back in a month I’ll order additional tests if you have not responded as expected, and particularly if you’re still losing weight. But in my first note, I would have likely at least mentioned some of the diagnostic possibilities, and probably using medical terminology such as malignancy. I write these for a number of reasons. Mostly it makes for good care. Just in case it turns out to not be depression, when I look back at my prior note it will remind me of some of the concerns I had. It also provides a road map of what I was thinking if the patient has to see another physician, whether it’s because I’m on vacation, they have to go to the emergency room or see another physician. The note is also necessary due to malpractice concerns. Doctors are usually not expected to know the future, but the legal assumption is that if you didn’t write it down, it didn’t happen. If you don’t show that you considered the possibility of a serious condition, the presumption is it didn’t cross your mind.

Psychiatrists are allowed to protect their notes. Is that because their patients are too unstable to see their notes? Is it because the psychiatrist needs to record things that a patient may misinterpret? During the course of treatment they might have some insight about a patient’s problems, but not know whether their guess is right. By recording their thoughts they can later go back and review them, improve their diagnosis and treatment, and better help the patient. Well the majority of psychiatric care in the United States is actually provided by primary care physicians. There are not enough psychiatrists to treat all the cases of depression and anxiety. But primary care physicians notes are not similarly protected.

One measure of the benefit of a treatment is the number needed to treat. For example, one may need to treat 20 patients with a cholesterol medication for a year for every heart attack prevented. Conversely is the number needed to harm. Depending on age, it’s estimated that for about every 1500 abdominal CT scans, one person will get cancer as a result of the radiation. A good clinician will be correct the majority of time. How many patients will be harmed by reading chart notes (needless worry, additional tests that have their own risks and costs ordered because of that fear, physicians not recording important information for fear of it being read by a patient) for every patient that benefits?

I’m not embarrassed by what I write in a patient’s chart, but patients might be if they read it. Imagine a man asks a family member to review his medical records to see if they think he has been getting good care, given his recent heart attack. He probably forgot that a few years ago he spoke with me about sexual problems he was having.

The system I suggest would be a juried one. Patients could request their records, and in most cases the physicians would grant access to most or all of the record. If there was parts they did not want to show, they could explain why to the patient. If the patient did not accept the answer, they could appeal to a third party health advocate who would then decide whether it should be released or not. This would only apply to patients who are not bringing legal action. I think this approach would make physicians a little more comfortable, and lead to better patient care.

Medication Errors

Not infrequently Express Scripts, Medco, or other similar companies send a fax to alert me that my patient is taking two similar medications. Occasionally it’s intentional, but most of the time it means something went wrong.

Sometimes I change a patients’ medication to something similar to achieve better efficacy, to minimize side effects, or due to cost. Although I always put the changes in writing for the patient, telling them what to start and what to stop, this doesn’t always work. Patients may get an automatic refill of the original medication from the pharmacy or call it in when they notice a pill bottle is almost empty. Sometimes they go by a medication list they’ve generated, but not updated, rather than the printout I give them.

Sometimes patients end up on two similar medications after getting one from a specialist who doesn’t realize a patient is taking something, because the patient didn’t bring the list I gave them, and they don’t remember everything they take. For example I might have the patient on lisinopril for hypertension, and their cardiologist prescribes the similar benazepril.

A similar medication error happens when we tell patients to stop a medication and they don’t for similar reasons as above.

So the faxes are helpful when these things are caught, but it would be better if it occurred at the the time the prescription is sent to the pharmacy.  Ideally the pharmacy computer would automatically connect to the physician’s electronic medical record (EMR), particularly the primary care doctor, and compare medication lists. If they had medications to refill that didn’t match the EMR record, they would call to double check if the patient could not give them a good reason for the discrepancy. In addition, the pharmacy computer could keep track of all the chronic medications a patient has filled. If the patient doesn’t get the prescription refilled in a timely manner, their computer would query the physician computer to make sure it was still an active medication. If so they would call the patient (and maybe in the future talk to the patient’s medication list carried on their computer/mobile device) and remind them to refill their medication, assuming someone hadn’t stopped it, the patient was taking samples, or some other good reason.

If you use a program such as Quicken, you can download credit card and other transactions and reconcile them with entries you’ve entered. Comparing medications would be a similar process.

There are certainly barriers to such a solution. Electronic health records would need to have medication fields standardized, and there would need to be protocols to exchange the information. I’m not sure, but I think some of this already exists. Of course there are legal issues such as HIPAA.

As John Lennon said, “You may say I’m a dreamer, but I’m not the only one. I hope someday you’ll join us, and the world will be as one.”