Dilemma of Accurate Data Collection

In the article How your hospital can make you sick, Consumer Reports paints a pretty negative picture of hospital acquired infections.  The data is disturbing, but without context can lead to reaching conclusions and action plans that may or may not work.

The electronic health record contains endless amounts of information, but may or may not provide the precise data researchers seek in an easily accessible form.  Optimally, the EHR collects the data automatically, without need for provider input.  However all too frequently, data is incomplete or inappropriately classified; unless an answer to a specifically requested question is input, the data may become difficult to retrieve.

Most commonly the providers do not know the questions that are needed, and do not therefore record whether they have or not done performed some action.  A good example is not recording that the patient is a MRSA carrier who then leaves the hospital with MRSA cellulitis and dehydration.  Did the patient acquire the infection prior to entering the institution or after hospital exposure?

An EHR can be used as a tool to capture this data.  The providers must know the questions and the  organization must create buy-in to collect the data.  There are various methods.  The easiest is a checklist prior to discharge that answers the questions easily with the ability to provide context.  This context can explain a behavior that may seem inappropriate.  Moreover, it can be entered into the EHR by a non-provider at a latter time.

The bottom line is the EHR can be formatted to help the clinicians answer the tough questions.  This may help various institutions avoid the dreaded headlines in the morning paper(if anyone still reads it).

ICD-10: The Elephant in the Room

ICD-10: The Elephant in the Room

            

With ICD-10 “Go-Live” compliance date being 12am - October 1, 2015 (36 hours 45 minutes away as this is being written)¹, we’ve been experimenting with the CMS/Medicare resources², and in particular their ICD-10 code lookup “tool³”.  Our object in doing so was to figure out how to quickly determine the appropriate code for any given case.  Why?

While it is most likely that physicians will not be entering their own codes on a regular basis, some will do so, either by requirement or choice.  In other cases a coder will most often be responsible for being sure the submitted code corresponds to the documentation.  In either case, the issue will be finding the appropriate code for the case in question.  If the search for that code takes an inordinate amount of time, care will be compromised when a physician has to direct attention away from patients to search for an ICD-10 item.  If it is a coder, the efficiency of coding operations will be diminished, and in many cases, more coders will thus be needed. As the cost of physicians and coders is non-zero,  such anticipated new inefficiencies can be expected to be expensive.  Hence the cost of care is quite likely to be adversely impacted.  “Affordable,” you see, exists at present in the eyes of the beholder, but not necessarily in the eyes of the one writing the check, whether it bounces or not.

            So we were trying to see how well the CMS code-lookup tool worked, and to ascertain what insights into the mind of ICD-10’s creators, we could obtain.  We recommend that each of you try this out.  The web-link is below³, or you can Google ICD-10 code lookup. Once there, ICD-10’s fun and complexity begin to be revealed. Moreover, it is clear that CMS did not hire a Google-trained developer to create its search engine.  The basic elements of good search are missing.  This one small oversight will cost millions.  A good search engine needs to have certain basic elements. For example, if you type in either cars or car in Google, you will get a search not unlike that if you type in auto.  That is, the search engine “knows” what items are identical or nearly related. This basic component affords compatibility with users, also known as humans, also known to be imperfect.  In medicine, no less does such flexibility need to be present in search engines.  But no!  The CMS ICD-10 search engine requires extreme specificity in keywords.  We wanted to see codes for an aortic dissection.  Entering “aortic dissection” gets you nowhere.  Entering aorta dissection is required. Unlimited examples can be obtained if you try it yourself.

            Therefore, if you are the one who has to put in the right ICD-10 code—good luck. But since neither luck, nor CMS fixes, are items to be banked upon, we suggest you create a “cheat sheet” of keywords to expedite the process. Post it, and have therein the common items that are used in your care setting.  Classic provider diagnosis is totally inconsistent in the ICD-10 lexicon.

Examples (try them):

1.    Acute Myocardial Infarction (does not exist as an independent entity; lised in search results are comorbidities only).

2.    Myocardial infarction (returns subsets, but lord help you if you put in acute)

a.    NSTEMI (works)

b.    STEMI also (works - with subsets for anatomical and arterial sites, if known)

3.    Unstable angina (works – but if you specify beyond 120.0, you are into transplantation and bypass graft sequelae)

a.    If you use the keyword angina- ICD-10 gives you a list of 30. Have fun reading.

4.    Pneumonia – (oh boy…you get 100 to chose from)

a.    If you put in community acquired pneumonia – nothing

b.    Hospital acquired pneumonia - nothing

c.    Better have your billing company tell you which one to select

5.    Acute Bronchitis—(works –kinda – get a list of 11,  nine of which you need to specify the organism, and the other two are organism unspecified and neither acute nor chronic). Having fun yet.

a.    Bronchitis alone expands the list

b.    The last of these is J68.0 (Bronchitis and pneumonia due to chemicals, gases, fumes, and vapors).  And if you search, then, inhalation, you get a series that begins with J69.0 (but not J68.0). You must be having fun at this point.

6.    Pharyngitis – (sort of works –the things we see all the time, e.g. early congenital syphilitic pharyngitis come to the top of the list….while those that are comparatively rare, like strep pharyngitis or acute pharyngitis, unspecified are further on down).

7.    Finally: as a final example (among many, many possibilities) of what is going to be frustrating, if you put in pulmonary embolism you get a list of six codes, none of which apply to a case of acute pulmonary embolism. If you put in pulmonary embolus, you get two codes for saddle embolus with or without cor pulmonale.  And finally if you put in just embolism, you get a goldmine (!), but it is longer than Tolstoy’s War and Peace.  At that point, I tried lung embolism and got “There are no ICD-10 Codes that match.”

Oh brother, this will be a slow process with a great deal of frustration. If readers send in their most common, basic diagnosis list with and ICD-10 equivalent, we will try to create a spreadsheet for common usage.  Keep in mind, too, that selected codes must correspond to the correct code assignment (in the ICD-10 CM/PCS).  Otherwise appropriate reimbursement for any clinical entity billed to a payor/insurance company may likely not occur.

Footnotes:

1. Countdown & Overview:

https://www.cms.gov/Medicare/Coding/ICD10/Index.html

2. Provider Resources:

https://www.cms.gov/Medicare/Coding/ICD10/ProviderResources.html

3. Code Lookup

https://www.cms.gov/medicare-coverage-database/staticpages/icd-10-code-lookup.aspx

Business Blog

XpressTechnologies practice management, billing and Electronic Health Record are ICD-10 ready and compliant and will provide any assistance in this major transition.

           

Technical Support for EHRs

Technical Support for EHRs

A recent conversation at a highly-respected and decades-old EHR vendor brought up the issue of tech-support for EDs and clinics, and how that best can be managed.  The issues are complex, and yet important when it comes to patient safety, healthcare staff backing, and the continuance of key operations that depend upon your EHR.

Those in the healthcare EHR industry know full well that a vast majority of support calls stem from issues potentially solved locally, with little or no technical know-how needed.  That is why most all tech companies, medical or not, provide FAQ support information, commonly online or on paper, that can be used to quickly consider and resolve issues.  It’s like the manual in a new car when you need to figure out a “how-to” but want to keep the vehicle running.  And few in healthcare have the time for their vehicles (EHRs) to stop running or slow down. At times though, more than the manual is needed, either for more in-depth information, or for hand-holding through some simple problem that nevertheless induces anxiety. Still, support teams commonly consist of rotating responsibility by the program developers, they rarely have time (or patience) for hand-holding.

For example, most facilities use printers, either for patient-care instructions, prescriptions, or for other printed care documents.  And yet, printers are often a big stumbling block for end users, who may be technically unskilled, new to the system, or rotational members of a staff that services multiple care facilities. Problems with these devices, however, are usually rather easily corrected.  Nevertheless, an EHR system that is functionally brought to its knees because the printer went out, can induce panic in the facility staff.

And then there are software, hardware, or internet-based technical issues that do crop up.  Who of us has not gone partially bald waiting hours online for Microsoft or Apple to provide an English speaking helper to resolve a problem that is making smoke pour from our ears, and impairing our lives.  It happens with all devices, from “smart” phones to ipads to laptops.  Yet, in the hospital or clinic environment, delay of this sort is not only frustrating, but dangerous.

How does a tech-support crew best solve such matters?  For one, be sure that your vendor has provided you with key items of access. They are:

1.    An FAQ-problem list of the most commonly encountered glitches, and how to solve them. This list should have easily located, step-by-step instructions, understandable by even the most novice clinician doing a new shift at the facility.  It should include an instruction to shut down (all devices) and restart as a next-to-last-ditch effort, and then to move to #2 if this fails. The FAQ sheet should be placed in a secure, known, location at all sites, and available to all users of the system.

2.    Phone contact information:  this should be clearly stated on the FAQ sheet. It may be helpful to distinguish more than one number, one for emergencies that have not been resolved locally, and another for issues that can wait until the next business day for resolution. 

3.    Support management by the EHR support service should include an initial re-routing message of the sort one hears from hospitals and physician offices:  “if this is an emergency, hang up and dial 911.”

In the case of technical emergencies, it might say:  “if your system is down, press 1.”

Or it might further define a true EHR emergency and say, perhaps “if you have a technical issue with your system, and it is now impairing the care of patients, press 1.”

When the issue is not urgent, further information might route the caller to voicemail to leave a message for the staff when they are available to respond.

4.    Follow-up and analysis by the vendor:  all EHR vendors should be interested in learning how requests for help are handled and resolved  A mechanism for doing so, either by email, website contact, or direct call, should be set up. That way, all involved parties can continue to improve their relationship and their abilities in keeping every installation of the system up and running.

Be sure that you and your vendor discuss elements of support, and how it is to be managed, so that you can work together to keep your facility seeing patients as efficiently as possible.

COMPARTMENTALIZING, PRIORITIZING, ACCUMULATING, AND COORDINATING

COMPARTMENTALIZING, PRIORITIZING, ACCUMULATING, AND COORDINATING: A BIT OF HISTORY:

Guest Blog (DRK) with Oversight (RB)

As an emergency medicine resident in the (very) late 1970’s I serendipitously happened to create one of the earliest renditions of an ED tracking board. Resources being scant, I went to the local office supply, to buy a very large erasable white board, which took up an entire wall.  On it were placed strips of colored tape in vertical columns for patient names, chief-complaint, stage of care, pending issues, and anticipated disposition; each room or care-space in the department had its own horizontal row.

Why in the world did I do this? Well, it is of note that after a few years, the hospital became the urban flagship for the University of Florida EM program. I was an attending. Many would wander into the area to see this “device,” and my explanation for its presence was as an attempt to make evident to all what was going on in terms of compartmentalization and strategizing. Before then, in my experience, that cognitive process remained rather non-communal, as it was usually kept within the heads of the docs and head nurses, spoken by mouth, and maybe visible, if at all, only via clipboard rack.  This new board seemed a step forward, in that other doctors, nurses, and ancillary staff, could all see what was going on, often at a single glance.*  Communication as well as prioritization (compartmentalization) was enhanced.

Now those of you who have grown up in EM and UC medicine since may well find it hard to visualize a world without tracking boards.  You may also find it hard to imagine a world without cell phones.  But in each of those instances, the world was, indeed, long (long) without these.

Since those earlier times, when working on developing one of the first EMR programs for the ED, placing tracking board functions within an electronic system became a goal.  Decades later, even with that functionality, though, we are still, for the most part struggling.  Why? Principally because efficient EM practice requires individual, personal, compartmentalization, in addition to whatever may be going on from a global, departmental view.

IMPROVING EFFICIENCY THROUGH PERSONAL COMPARTMENTALIZATION

         The Merriam-Webster Dictionary defines the verb compartmentalization as three kinds of actions:

: to separate (something) into sections or categories

: to separate (two or more things) from each other

: to put (something) in a place that is separate from other things.

        

         At its best, and even at its not-so-best, compartmentalization allows one to deal with large quantities of input at the same time, in a variety of situations, and without getting confused, agitated, error-prone, or stressed.

         A classic example is the single provider Emergency Department or Urgent Care with 8 or more simultaneous patient encounters. The stressors on the situation are

1.   Time

2.   Patient satisfaction

3.   Making the proper diagnosis and treatment plan

4.   Interruptions

5.   Interactions: staff, family, phone calls, new information.

6.   Interruptions

7.   External forces demanding unreasonably rapid patient flow.

8.   Interruptions

9.   Finally, the hardest one: following up on each individual patient to ensure that 1-8 have been handled thoughtfully and reasonably.

10. And then: finishing the endless documentation formally known as paperwork.

A mental compartmentalization technique can be used to divide up the patients in a prioritized manner that limits potential for confusion. After each patient is seen (by you), they are put in virtual mental schematic of the department, noting their location, their problem, the key bits of information needed to make the final disposition of this patient, and the typical time frame needed to make a disposition for this complaint. Think tracking board, but written within your own mind.

An example would be a patient with RLQ pain in room 2. Your disposition will perhaps be based on the results of the CT scan for appendicitis and you expect completion in about 2-3 hours. At the same time, you might have a patient with an Acute MI in room 1. Your anticipated disposition here, is arranging for catheterization and initiating a treatment protocol. Time frame should be 30 minutes to hour. Now the patient in room 8 has bronchitis. Disposition in real time with prescriptions and follow up, might be completed in   30 minutes to hour.

All this is inescapably overlaid on a fact sometimes noted, but rarely voiced: the best of the “older” ED docs would have diagnosed with ADHD as children. Had the current version of political/diagnostic correctness been in place in prior generations; likely they would have been medicated, and real talents pharmaceutically precluded.  That is, an ADHD-like cognitive mind is a benefit in the ED. Woe to the doc who has to complete one thing before going on to another (we all know them). Attention has to shift, and rapidly, but nevertheless, to be effective, a supervening mechanism to focus attention is needed.  Compartmentalization, mental compartmentalization, is indeed such a mechanism.

If you can compartmentalize your thoughts, and prioritize your actions, you can deal with the heart attack, check for the CT on the tracking board, and discharge the low acuity patient, almost simultaneously. The result is a well compartmentalized, well prioritized, tiered tracking board; the highest tier is (mental), within you; the next tier is external (physical) a displayed tracking board. The one (mental) sits inside the other (displayed), but only the displayed version is visible. Your mental board shows the rows and columns for you to easily access.  The ancient Greeks used to use this type of mnemonic scheme to facilitate memory of huge amounts of data. They would visualize a large building with many (many) rooms, cubicles, spaces, and into each would mentally place any item they wished to remember.  To retrieve, they “went” where they knew it was.

 Using this technique can help you to function efficiently, multitask, and move patients through the department.  Notice it requires a bit of balance, and assumes some conditions that may or may not be present in your facility at all times. For example if you have three MIs and an incoming rescue unit, your compartmentalization/prioritization is going to have to achieve a different level than if you have one rash to see. But it can be very helpful, even under significant stress, to be maintaining your own internal tracking board. In fact, even if you hav three MI’s going on, you can probably also handle a rash.

All this does take practice, and a key to making it work is pre-thinking the steps needed to make disposition on the case, as early in the encounter as possible (like after the first 5 words). A partner of ours (name unmentioned) was known for making dispostions at the time the first EMS call came in. If you can train yourself to think [encounter-->disposition], you will be 10 steps ahead of the game.  Even if there are three active MI’s, you will be thinking about which ones are going when to the cath lab, which are being transferred (perhaps), and also what to do with the rash. Such direct planning is rarely, if ever, written on actual tracking boards. Developing, practicing, and “seeing” your own internal tracking board and using the material/actual tracking board as a real-time assistant will significantly enhance your ability to process patients.

XpressTechnologies Electonic Health Record with charting, tracking board, prescripton writer, and follow-up instructions has been time-tested to dramatically improve efficiency.

        

 *For those of who like to research such things from an academic/theoretical viewpoint, consider the article by Berg Computer Supported Cooperative Work 8: 373–401, 1999 Kluwer Academic Publishers. Accumulating and Coordinating: Occasions for Information Technologies in Medical Work. This pivotal work is not-uncommonly referenced by another seminal thinker Bob Wears (e.g. The Chart is Dead—Long Live the Chart Annals of Emergency Medicine Volume 52 No 4- October 2008 p. 390….and many more).