"Welcome to our company blog. Within these blog posts, we hope to share our insights on clinical quality management, the patient-centered medical home, chronic disease management in primary care, evidence-based medicine, and the use of technology in ambulatory care settings."

- David Morin, CEO and Donald Nease Jr., MD, Chief Medical Officer

Cielo MedSolutions now provides periodic email updates on the implementation of this Act and its impact on an ambulatory care practice. Below is our first update. If you'd like to be on the email list, simply let us know at info@cielomedsolutions.com.

Of course, the usual disclaimers are in place: this is a rapidly changing landscape as no final decisions have been made on the HITECH Act and views expressed in this document are only opinions.

SNOMED-CT is the Problem and Procedure Vocabulary

From the HIT Policy Committee, August 14, 2009, Report From Clinical Operations Workgroup update slides:

“Primary vocabulary standards:
· Clinical problems and procedures: SNOMED CT
· Drugs: RxNorm
· Ingredient allergies: UNII
· Lab tests: LOINC
· Units of measure: UCUM
· Administrative terminology: CAQH CORE and HIPAA”

From the HIT Policy Committee, August 14, 2009, Clinical Quality Workgroup: Progress Report update slides:

"· Multiple versions of measures to provide options
o 2011 – ICD 9 or SNOMED CT
o 2013 – ICD10 or SNOMED CT
o 2015 – SNOMED CT
· Can use internal codes using SNOMED CT expertise to map to SNOMED CT
· EHR certification should require problem list”

Views: it looks as though problem lists will ultimately need to be built with SNOMED CT versus ICD9 or ICD10 and the open question is the year it needs to be done. This could be a significant issue for systems that do not have a clinical thesaurus that can cross-reference problem terms across different vocabularies (in other words, a system needs to be able to know how an ICD9 or ICD10 code maps to a SNOMED code). In addition, registry and EMR systems that only utilize billing files for problem documentation may struggle with using an alternative coding system (problems will still come in from billing files only in ICD9). Because Cielo uses a clinical thesaurus (the ENCODE table mapped to ICD9, ICD10 and ICPC) and one of Cielo’s Medical Advisory Board members is leading a committee to map ICPC to SNOMED, this will not be an issue for Cielo Clinic.

Registries called out as key to ARRA

From the National Committee on Vital and Health Statistics Report of Hearing on “Meaningful Use” of Health Information Technology, April 28-29, 2009:

"Testifiers reported that the ability to get data out of EHRs easily – both for reporting and for creating panels of patients, is difficult with today’s EHRs. However, it was noted that in addition to embedding registry functionality in an EHR, such functionality (and others) may better be delivered through applications and services that are not part of a single all-encompassing application, such as population or disease registries."

Views: This testimony from industry leaders reinforces the growing sentiment that registry solutions are a very viable solution for meaningful use.

Modular Approach Available through EHR-M

From the HIT Policy Committee Review of Initial Recommendations by the Certification and Adoption Workgroup, Paul Egerman and Marc Probst, Intermountain Healthcare, August 14, 2009:

“Recommendation 4 – Flexible Software Sources - provide for certification of components so EHRs can be purchased from multiple sources”

Views: a registry can be certified via EHR-M. You will be able to assemble best of breed components from a variety of vendors to meet meaningful use. You will probably find this can be done at a total price-point lower than a monolith EMR. Cielo is assembling the best-of-breed products that, together, will be pre-interfaced and meet meaningful use.

Submissions Will be Electronic

“CMS noted that only measures that "can be submitted electronically" will be allowed.”

Source: CMS Sheds Light on Meaningful Use, HDM Breaking News, August 14, 2009

Views: electronic submission is good. We predict the constructs in place for registry reporting on PQRI will be used for meaningful use submissions. Therefore, a system already doing PQRI uploads, like Cielo Clinic, should have an advantage.

Dave Morin
CEO and Co-Founder
Cielo MedSolutions LLC

A must-read by e-Patient Dave and his experience downloading his medical data from his hospital into Google Health.

http://e-patients.net/archives/2009/04/imagine-if-someone-had-been-managing-your-data-and-then-you-looked.html

And people still want to use billing data in their patient registry?

Dave Morin
CEO and Co-Founder
Cielo MedSolutions

It seems as if the interest level has risen rapidly on our argument that there are many shortcomings when using billing data and ICD9 for a quality improvement program. I've been involved in numerous discussions about this over the last month and they are, in general, in-line with our thoughts at Cielo; billing data and ICD9 are plain insufficient! Usually, the discussion is with a person that has already undertaken a quality improvement program and found this out after trying to work with their billing data.

To study this further, we are working with a few providers and comparing their billing files to their Cielo Clinic database to look at how a patient's diagnoses align in these two datasets. While I can't release our findings just yet, suffice to say the results of this work are eye-opening. Look for more to come from us on this topic.

FYI - I'd also recommend reading a very good article in the February 2009 article of Healthcare Informatics that discusses this topic.-


Dave Morin
CEO
Cielo MedSolutions LLC

We were recently asked to summarize our thoughts on the shortcomings of ICD9 and billing data when used for diagnoses in clinical quality management systems. I thought I'd share our summary….


Specificity - ICD9, for many diagnoses, does not provide the required level of specificity required for evidence-based care guidelines. An example is asthma. ICD9 cannot differentiate between persistent asthma and intermittent asthma, an important distinction.

Scope – literature has documented that ICD9 can accurately represent approximately 50% of the conditions a primary care provider will encounter. When a condition cannot be properly documented, a provider must choose the “best fit”. This can be a major problem for clinical research and also affects the use of this data for care guidelines.

Accuracy – the needs of documentation for reimbursement leads to incorrect problem documentation. A common example is the need to document a diagnosis of asthma for a patient presenting with wheezing. If the patient is ultimately not asthmatic (which is usually the case), there is no way to “go back” and change their diagnosis on the billing data record. Therefore, when that billing data file is used in a registry, it inaccurately represents the asthmatic patient population. This inaccuracy can exceed 50%.

Completeness - Billing data does not document lifestyle issues like smoking and cannot capture clinical modifiers such as family history and risk factors. These elements are important for care guidelines and can be important data elements for clinical research.

We believe that these shortcomings are solved through the use of ICPC, the International Classification of Primary Care and through the use of clinician-verified diagnoses. A few prior blog entries talk about this and we'll be talking a fair bit more about it in the months ahead.

Dave Morin
CEO and Co-Founder
Cielo MedSolutions LLC

If you've spoken to us in the past, you know we are ardent supporters of a classification system called the International Classification of Primary Care (ICPC). This system, we believe, does a much better job of documenting problems and reasons for encounter in primary care than ICD-9. It also provides an ability to create episode-of-care documentation.



Why would you use it? Well, for one, it will generate a more accurate problem list and allow you to be more effective in care delivery. Second, this accuracy will help you with pay-for-performance programs as we have found a more accurate problem list leads to better documentation and hence better pay-for-performance reporting. Third, accurate problem lists are the cornerstone of a patient-centered medical home and an ICD-9 problem list based on billing data is not an accurate problem list. Fourth, it provides a unique view into symptoms and problems - as documented here in "Characterizing Breast Symptoms in Family Practice".1



1 Eberl, Margaret M., Phillips, Robert L., Jr, Lamberts, Henk, Okkes, Inge, Mahoney, Martin C.Characterizing Breast Symptoms in Family PracticeAnn Fam Med 2008 6: 528-533


Dave Morin

CEO and Co-Founder

Cielo MedSolutions LLC

If you are evaluating or researching the technology underpinnings of a medical home, you should take a look at wonderful piece (available on the internet) called "Health IT to Support the Patient-Centered Medical Home" authored by Michael Klinkman and Robert Phillips. This slide show accompanied recent testimony they gave to the National Committee on Vital and Health Statistics.

It's probably different than many of the other presentations you've seen on this topic, but what they have to say is very powerful.

Dave Morin
CEO
Cielo MedSolutions

Disclosure: Michael Klinkman is on the Medical Advisory Board of Cielo MedSolutions.

As you know if you've spoken with us or spent much time browsing our website, we're big on the fact that Cielo Clinic contains a coded, all-problem registry. Here's a brief description of why that is so important.

Your registry is only as good as its accuracy.

Many registries are built off of billing data. Not a bad place to start you may say. Just take all those people with asthma or depression codes and dump 'em in! Not so fast... Billing data sits at one end of a long process that begins with a clinician thinking, "This patient is wheezing, but I don't have enough yet to make an asthma diagnosis." So the clinician either writes down "wheezing" which gets changed to a billable diagnosis of asthma, or they realize from the start that wheezing won't get paid for and they just write asthma. When that billing data gets dumped into the registry, the patient with wheezing is now part of the asthma registry, and becomes part of the reporting for which one is responsible. Whoops!

If you have a system that allows you to accurately represent with a coded problem what that clinician is thinking, "wheezing" doesn't become "asthma", and your quality improvement efforts are focused with precision on the patients that truly have asthma.

Cielo Clinic contains a rich set of clinical terminology that was built by having primary care physicians record problems using words that make sense to them. These terms are coded, and mapped to a classification system that allows accurate aggregation of the problems into disease categories for quality management. If your quality management system can't do this, you'll spin your wheels focusing on patients that don't have the diseases you're trying to impact.

If you build it they will come - Field of Dreams

OK, maybe it's a bit corny, but that's our reason for having an "all-problem" registry. Clinicians build it by recording the problems they are working with every day. Not just on a select set of patients, but all patients. A lot of work? Not if you are getting something in return. What Cielo Clinic gives you in return is an accurate problem list on every patient, using terms you understand, not up-coded to meet billing standards. Additionally, Cielo Clinic gives you back reminders based on those problems and the ability to "on-the-fly" respond when new guidelines emerge for conditions. Because you've been accurately recording the problems, you have already built the registry for any given clinical disease category.

You're building it, because those new guidelines will come!

Don Nease, MD
Chief Medical Officer
Cielo MedSolutions

AHRQ (Agency for Healthcare Research and Quality) recently published a wonderful paper entitled "Health Information Technology for Improving Quality of Care in Primary Care Settings". The paper looks at "the link between health information technology and quality improvement in a range of primary care settings"1.

To see the document: http://healthit.ahrq.gov/portal/server.pt/gateway/PTARGS_0_1248_661809_0_0_18/AHRQ_HIT_Primary_Care_July07.pdf.

There are many insights and recommendations documented in this piece; if you are looking into technology and best practices to support improving quality of care, it is a must-read. But, there is a specific issue highlighted in the report that "hits home" because we constantly talk about it. Specifically:

1. Page 18 - "another aspect of data structures that continues to restrict improvement activities is the billing and reimbursement coding mindset that permeates much of health care data. For example, in many health IT systems, patients with asthma do not have a diagnosis of asthma; they have a data history of billed visits with a billing diagnosis of asthma. For visits to the clinic that did not involve their asthma (and hence no billing code of asthma was issued), there is no way to relate that visit to their chronic condition of asthma. Additionally, an asthma billing code is often used for a patient who arrives wheezing (whether they have a diagnosis of asthma or not). This may not look like a data problem on the surface, but if you ask the health IT system how many asthmatics are in a panel, the numbers are far from reality."
2. Page 19 - "data structures for billing and documentation are often very different from the data structures that support improvement"
3. Page 13 - "data that are constrained to billing codes may make it very difficult to track the progress of a chronic diagnosis over time"

The paper also discusses that it may be difficult to solve this problem - but we think we have solved it through the use of clinician-verified diagnoses built from a thesaurus of coded clinical terms. We do not rely on billing diagnoses; clinicians capture patient diagnoses at the point of care based on the true conditions of a patient and Cielo Clinic stores and utilizes these diagnoses. This capture is a very simple effort and has a very small time impact on a provider.

Of course, it is quite simple for our system to take a billing file and use it to pre-populate a patient database with diagnoses; we offer to do this during every Cielo Clinic installation. But, we rarely do diagnoses pre-populations as most practices are not comfortable with these data sets for clinical quality improvement.

We feel strongly that a quality program built on billing data will not be successful. Bottom line, billing data serves a different purpose than quality of care data. Billing data was never constructed to support quality of care initiatives. A different data set is needed, period. It needs to be coded, clinician-verified and uncoupled from billing diagnoses. It isn't difficult to build this data set, especially with a tool such as Cielo Clinic, and build it to the benefit of all stakeholders; providers, patients and payors.


Dave Morin
CEO
Cielo MedSolutions

1: Langley J, Beasley C. Health Information Technology for Improving Quality of Care in Primary Care Settings. Prepared by the Institute for Healthcare Improvement for the National Opinion Research Center under contract No. 290-04-0016. AHRQ Publication No. 07-0079-EF. Rockville, MD: Agency for Healthcare Research and Quality. July 2007. Page 1.