Rethinking the CPR: Is Perfect the Enemy of the Good? - Industry Trend or Event

Health Management Technology, May, 1999 by Francine R. Gaillour

Aiming for the "less than perfect" Computerized Patient Record may achieve something more useful for the organization.

Remember the original claims and designs for the Computerized Patient Record (CPR)? Full workflow automation, expert-assisted physician order entry, pop-up alerts, reminders, point-and-click structured note entry, a pristine clinical data repository with abundant data outcomes, and so on and so forth, toward the perfect model?

If the vision seems fuzzy at times, even more so are the reasons why the perfect CPR system has not been widely implemented: the technology is too complex and too expensive, vendors haven't built "it" yet, doctors won't use computers, standards don't exist, and the list goes on. After looking closer at this whole conundrum, a modest suggestion is made that we aim for a "less than perfect" CPR in order to achieve something useful for the organization.

Let's first look at the factors that may be interrupting the journey to the high road and then discuss a possible approach for a smoother path along a middle road.

Potholes Along the Road to CPR Perfection

In gathering the stories of what organizations have attempted, where their efforts stopped, or why they tried something different, several factors appear to have dampened the collective ambitions for the CPR:

1 Factor 1: Some CPR components are easier to roll out than others (see Table 1). While this seems painfully obvious to those close to these issues, this fact appears to get lost when one compares the typical RFP wish list with the realities of what most organizations have been able to successfully implement. For example, lab results and transcription review are generally widespread.

Table 1.

Implementation of CPR Components

Easier

Transcription capture
Scanned documents
Patient schedule
Lab results repository and view(*)
Registration and charges (ICD-9 capture)(*)
   E-mail
Ambulatory prescription writer(*)
Population health screening reports(*)
Web access
Automated abstraction of minimum
standard data set(*)
Manual abstraction of minimum
clinical data set(*)

Harder

Structured progress note entry via point and click(*)
Patient flow tracking(*)
Fully populated clinical repository(*)
Computerized problem list maintenance(*)
In-box routing of results and alerts
Full physician order entry(*)
Automated provider alert(*)
Cross-continuum multi-system database integration
Automated clinical pathways and medication charting(*)

(*) High value data for performance outcomes, clinical process improvement

But organizations have a difficult time with broad implementation (100 percent buy-in of clinicians) of other CPR components, including point of care nursing documentation, physician order entry, or physician structured progress note entry.

This is evident in the words of one physician who exclaimed during his organization's IS strategy meeting, "Real doctors don't do data entry." Whether they door don't, one truth seems to have emerged and that is almost all physicians will do some, but only a few will do a lot. So when looked upon as a pure "workflow automation" tool, the CPR is having a tough time. Instead, it does better when used as a tool for passive information review.

Hence, the CPR components that have the widest acceptance by physicians are those that a physician primarily views passively: a physician desktop with links to lab results review, daily appointment list, transcription review, a list of patient visits and ICD-9 diagnoses--with the bulk of this information captured by transaction systems and HL7 interfaces or through direct entry by lower paid clerical staff.

So while every RFP contains excruciating details of what the "hard" modules should do, and every vendor dutifully designs these features, the fact is that most vendors and customers are not focusing on what works for the masses. That is unfortunate because the "easy" components, besides providing very high value in terms of information access, offer high value coded data that can be used for operational and clinical process improvement.

2 Factor 2: The different healthcare market constituents value CPR components differently. While a physician group practice of fewer than 10 physicians may be able to achieve a paperless environment with a simple EMR system, the emerging IDN has higher expectations of the CPR. Enterprise capability requires that hundreds of physicians and nurses participate in the IS rollout in order to string together the disparate organizational departments. In some respects, this drives a "lowest common denominator" approach in terms of what the IDN can expect from the masses.

However, having 100 percent of clinicians accessing results and transcription reviews (the easy modules) is infinitely more valuable than having only a handful of clinicians doing direct data entry (the harder modules). Broad implementation of even the basic CPR components creates some measure of IDN workflow standardization, in addition to the creation of a rich data pool for back-end outcomes integration.