Health Maintenance Working Group

The original intent of Health Maintenance tool was to account for both established clinical guidelines and local community practice. Since its initial posting, guidelines and practice have changed such that there is a need to re-evaluate the current design of the tool. I am interested in pulling together a group of interested docs to review the current design and determine the contents of a revised version. I’d like to kick off the project with an informal meeting within the next month. Given the nature of the work, I think we can accomplish much of the task thereafter via email, blogging or other electronic means.

The goals of this project include:

• Selecting and reviewing the latest clinical practice guidelines
• Reviewing the literature for some of the more controversial screening measures
• Determining the appropriate HM elements
• Determining appropriate time intervals
• Reviewing and editing verbiage of the alerts

Secondary goals include:

• Piloting various electronic means of communication: email, blog, wiki, twitter, web-conference
• Previewing developments in HM too

I expect that, if we are successful with the electronic communications, there may only be a need for an introductory meeting and possibly a final meeting for sign-off. Please let me know whether you would be interested. At minimum, I’d like a group of 4-6 docs to take part in this project.

I’m looking forward to a lively and fruitful discussion.

Health Maintenance...again

So I've think the IS team has struck paydirt. After hours, days and weeks of techno-struggle, it seems the Health Maintenance performance dilemma has been solved. In fact, it may finally be working as designed (see 4/22 post).

I give you fair warning that it can take up to 45 seconds or a minute if you are loading a patient record with a great deal of information for the first time since the fix. Go back to it in the future and it should load quickly, finally relying on the HealthFacts tool I've alluded to in previous posts.

Now that the Health Mainteance tool appears to be functioning again, I will be looking for folks to join in a working group to consider a revamp, taking into account the various practice guidelines and latest evidence. Anyone interested knows how to find me... or I'll come find you.

Health Maintenance trial fix

I mentioned a week or two ago that we are working on improving the loading time of Health Maintenance. Today we ran a "proof of concept" trial to see whether our strategy would work. While I haven't heard from more than just a couple of folks, it seems that the HM tool is loading more quickly.

It was quickly discerned by these docs that the creatinine expectation loads as though never satisfied. As this was just a test run, the creatinine deficiency will not be part of the fix. Once Josh is able to write new code to shift the Health Maintenance tool back into gear, you will find that the creatinine measure will accurately load re: expected vs. satisfied. However, not all that is accurate is precise! The creatinine will show as expected if it had not been done in >365 days, but whether displayed as satisfied or expected, will show the current date and not the date it was last performed. This might have to be the compromise we choose to live with until a real fix comes with future Cerner upgrades.

The bottom line is that you will be able have this health maintenance measure flagged and be logged accurately (and precisely) in the diabetes registry; you will just have to make sure that it has the correct date when running your screen. For the most part, I would assume that virtually all HbA1c measures have an associated creatinine, so that would be a place to start.

Health Maintenance Performance

As many of you know and have reported, the performance, i.e. speed, of the Health Maintenance tool has been absolutely lousy for the past few months. While we thought we had improved function a couple of months ago with the introduction of a new tool meant to speed up the loading of data, we soon found this had no effect. The idea was that this tool, called Health Facts, would cache individual patient data so that each time you went to HM, the only data that would need to load would be those that had not previously hit the page. Well, whether that tool works or not, Health Maintenance remains slow and dysfunctional.

We've been told by Cerner that Health Maintenance "works as designed". So we now know that the design is faulty. What we have found out is that the tool is designed to find every clinical event that satisfies the health maintenance measure. (A clinical event is any result that we are counting, e.g. LDL.) So, while this isn't such a big deal for an event that occurs every 10 years (tetanus) or even every year (mammogram), a serum creatinine can occur dozens, even hundreds, of times in the lifetime of the patient (data exists for the past 5 years). We have found that there are patients who have had hundreds of these events and that the tool records each of these events to a so-called table. What would be considerably more reasonable would be that the last posted value be used to satisfy the HM expectation and so avoid slowing down the process of displaying the HM elements.

So the bad news is that, and we (meaning, Josh Wherry of IS fame and uber-talent) have haunted Cerner as much as is feasible. The system designers are aware of our dissatisfaction, but, because of evolving software code, are unable to make the fix for us.

But the good news is that we have a clear understanding of the bad news and that Josh has written a work-around piece of code. This will enable the HM tool to work effectively and efficiently until we are able to take on updated code from Cerner later this year. The fix, though, will involve a single hitch for it to work.

Without getting into any more detail than I already have, once this work-around is tested and then brought to a computer near you, one test will display in somewhat funky fashion: the creatinine. The test will show as an above-the-line expectation if not resulted within the past 365 days, but will not display the last date that it was actually performed. Conversely, if the creatinine had been done in the last 365 days, such that the expectation has been satisfied, the date noted below the line will read incorrectly.

The better news is that the correct date of the creatinine, and presumably the performance of HM in general, will present "as designed" by the end of this year without the need for Josh's work-around.

Thanks Josh!

A brief return to Health Maintenance

I've been working on a reasonable review of the new Problem and Diagnosis List functions that will be found in CIS Enhanced View. While I'm quite confident the upgraded version will solve a host of problems we've all encountered with the current version, there are some key changes that I want to make sure I explain so that I can refer you back to this site in the event of any difficulties.

So... I see from the poll and I've heard in conversation that there remains some concern about the Health Maintenance tool. I suggest folks go back to December's and the first half of January's posts to check what I've had to say thus far. I've also added a few major links that take on this topic.

Here are a few issues that I've heard about:

• The addition of the Zostavax vaccination- most of us in adult primary care are aware not only of the "availability" of the vaccine, but of the guideline that recommends a single dose for all adults 60 years and older. While the efficacy data are not exactly compelling and the cost of the vaccine is prohibitive, particularly when payers want no part of it, I included this HM element because it seems reasonable to track and is part of the MHQP preventive health set. Take a look at this slide for one way of handling the RED mark:
  

Uploaded on authorSTREAM by nrk99

• Pneumococcal vaccine does not have a Cancel Permanently selection. I will have this added as soon as technologically feasible.

• Consider changing the due date time frame for certain expectations, e.g. Td/Tdap, to "turn red" earlier. So for Td, the expectation could come due at 9 years rather than 10 years, thereby avoiding that booster being delayed in the event that the patient visits during the 10th year only to return many months after the 10 year expectation. As I've mentioned in an earlier post, the next big upgrade of CIS will include an updated and customizable version of HM that will allow the physician to make these changes based on patient needs rather than set code. For now, I'm glad to entertain this question and would appreciate feedback. I believe it would be practical only for those expectations that have long intervals since this could be a set up for over-testing.

• Prostate Cancer screening - get rid of it... This is a not uncommon suggestion and, as I approach 50, I would love to host a lively if not definitive debate here. For now, the PSA and DRE will remain on the HM tool and, like the Zostavax, can be canceled if you so choose.

Stay tuned for Problem and Diagnosis Lists - Enhanced View style.

Upcoming Changes to Health Maintenance

I've mentioned that the Health Maintenance tool will become more user-friendly with a major upgrade later this year. In the meantime, I will be working with the CIS team to enhance the effectiveness of the tool as it now exists. The next task at hand is the building of more disease-specific expectations.

During the past week, I've submitted cardiovascular disease diagnoses that will trigger yearly lipid profile testing. To give you an idea of why it can be something of a technical chore to make this "simple" health maintenance expectation possible, I searched using multiple terms and arrived at 341 distinct diagnoses. Each of these must then be associated with each component of the lipid profile. While hundreds of CVD diagnoses have been identified, I'm sure some of you have found others in the SNOMED code system that should qualify a patient for the FLP requirement.

We will be working to add triggers for all forms of cardio-, cerebro-, and peripheral vascular disease as well as hyperlipidemia diagnoses in the coming weeks.

Let me know how we're doing and any other ideas you have.

Health Maintenance and the Diabetes Registry

You may be aware that quarterly diabetes registry reports were recently sent to all BMP adult primary practice managers and medical directors. Admittedly, we are still working out some of the core and composite calculation quirks that are used for global and group assessments. Assuredly, the demographic and lab details are accurate and clearly defined in the tables. What is lacking, however, are many of the features that are used to assess quality and, ultimately, pay-for-performance. These include:

• date of last diabetic retinal exam
• date of last comprehensive diabetic foot exam
• date of pneumococcal and influenza vaccines

The first two of these need to be manually entered into the health maintenance tool. Depending on the work design of a particular practice setting, this might be done by the MA upon patient's arrival to the office, by practice associate (PA) when a letter comes from the ophthalmologist/optometrist, or the physician upon review during chart review or during an office visit. My personal preference has been to enter the data during the patient visit as I spend time reviewing the health maintenance profile. Depending on where the retinal exam letter is buried, the entry of these two data items takes considerably less than a minute.

The vaccinations are, or at least should be, fed into the health maintenance tool as a consequence of the procedure being documented by the MA/RN. In other words, if you see a red mark in the HM tool, then the patient likely needs the vaccine.

It may be the case, however, that a patient received a vaccine at another site, e.g. VA, CVS, senior center. If another member of the staff has not entered that information, this can be done by entering into the immunization profile and attributing the procedure to "Historical Data" with an approximate date. If this profile is unfamiliar territory, I'd recommend asking your staff to enter this data when vital signs and other such information is entered.

The bottom line is that your engagement with the Health Maintenance tool will help make the registry reports, and thus quality & P4P measures, accurate and meaningful. I'd be glad to give more detailed instruction for those who are not comfortable with manual data entry.

More on Health Maintenance

Another sometimes problematic health maintenance issue is the time frame for each of the preventive measures. I am in complete agreement that the current HM tool is painfully inflexible in that one cannot change the time frame in the event of an (often common) exception. For example, if a patient has a history of colon adenomas, the timing of the next exam may be anywhere between 1 and 5 years, depending on pathology and physician judgment. While the colonoscopy can be documented in the Procedure Profile, the date of the next expected colonoscopy cannot be altered in the Health Maintenance tool.

The following is the list of time frame defaults for preventive measures. They are based on the guidelines I alluded to in my first post last month. I don't doubt that a couple of these are debatable, and not just for special populations. For example, bone density tests are often performed every 2 years depending on age and clinical judgment.

Mammogram: every 365 days
Cervical cancer screening: every 365 days
Bone density testing: every 5 years (1825 days)
Prostate DRE: every 365 days
PSA: every 365 days
Colorectal screening: if postponed, then 365 days; colonoscopy - every 10 years; flex sig - every 5 years; stool occult blood - every 365 days
Chlamydia screening: every 365 days
Tobacco exposure: every 365 days
Influenza vaccine: every 365 days
Pneumococcal vaccine: once
Tetanus booster: every 10 years
Zostavax: once

In the same way that a red mark can drive us crazy, so too can the inflexibility of the time frame. Keep in mind that the HM tool was built both to meet the minimal standards of care for all patients and to maximize opportunities for quality measures performance. In this way, the tool is really an aid; it can help facilitate best practice. Ultimately, though, it is the physician, and not the guideline, that provides the exemplary care our patients expect of us.

I do have some good news: by the end of 2009, CIS will undergo another upgrade (that is, in addition to the Enhanced View upgrade to be rolled out across Baystate Health on March 24). In the later upgrade, there will be a number of significant improvements to already extant tools as well as the introduction of new and quite excellent tools (more on that later). One major improvement that I have seen demo'd is of the Health Maintenance tool. This new design will include the customizability we are all wanting for...

Health Maintenance Age Ranges

While I've got a file that contains all the minutiae of the Health Maintenance tool, from satisfier categories to time frames, I'm going to start the piecemeal review by outlining the various expectations and their associated age parameters. Let me know what you think.

Mammogram: 40-100 years
Cervical cancer screening: 21- 65
Bone density testing: 60-100
Prostate DRE: 50-100
PSA: 50-100
Colorectal screening: 50- 80
Chlamydia screening: 16- 26
Tobacco exposure: 0-100
Influenza vaccine: 50-
Pneumococcal vaccine: 65 once
Tetanus booster: every 10 years
Zostavax: 60 once

Note that these expectations are set for ALL patients and do not include disease-management build that is currently in place for diabetes, asthma and heart failure.

Another thing to keep in mind is that, in the event that you do not agree with the screening measure, e.g. mammogram 40-49 years, the so-called expectation can be satisfied by selecting deferred or another relevant rationale. Alternatively, you can leave that expectation as is, though many of us have a hard time tolerating RED.

Happy New Year!

CIS Health Maintenance Tool

Thanks for finding me here. I am starting this blog to provide a forum for sharing ideas and ideals about the electronic medical record and to help facilitate the successful adoption of CIS throughout the Baystate Medical Practices. I see this blog as one of many tools needed for peer-to-peer exchange for the sharing of ideas, critiques and criticisms of the current state of the EMR and its hoped-for future state.

The first topic that I'd like to examine is the Health Maintenance tool. As many of you know, this has been online for over a year and was developed as a means of accounting for both general and disease-specific patient care measures. While it has a number of imperfect elements, I know it has been widely adopted based on internal reporting and on the comments I receive via email. The most recent issue is the response to the new expectation for the Zostavax, or shingles vaccine, for all patients 60 years and older.

Let's begin this conversation with a review of the various guidelines that have been referred to in the development of the current state. Feel free to compare and contrast and leave your comments here, especially as they pertain to some of the more questionable elements I included during the initial build, e.g. PSA/DRE, clinical breast exam.

Collected Guidelines from AHRQ (Agency for Healthcare Research and Quality)
Prostate Cancer Screening ACS
Breast Cancer Screening ACS
MHQP Guidelines


Let's use this set of links as a starting point for our conversation. If you are interested, I can share with you some of the technical details involved in the build of the HM tool as well as a preview of the tool as it will exist in a year from now.

I'd also welcome thoughts on the best ways to communicate with the BMP community. While I know that email is usual method for institution-based communication, the blogosphere may add to the exchange by adding a new dimension of web appeal.