More Fixes for the OBRA Regulatory Process

Part Five of a Series on the Survey Process

by Steven Levenson, MD, CMD

This month's column continues the discussion of specific changes in survey instructions that could improve the OBRA regulatory process. These include more consistent survey methods and better ways of identifying negative outcomes and determining "avoidable" results.

Clarify & Standardize Survey Methods

Surveyors should use a single, consistent method for all reviews. As with providing patient care, the survey process involves fact finding, problem recognition, and cause-and-effect identification. But we know that surveyors, like most people, may find it difficult to set aside their personal preferences and predispositions in order to interpret evidence objectively. And once they draw a conclusion, they may then tend to accept only evidence that supports that conclusion, rather than appreciating contradictory evidence; also, they are more likely to interpret specific information as supporting their predetermined conclusions.

To circumvent these propensities, the Centers for Medicare and Medicaid Services should reorganize the State Operations Manual to make its requirements clearer. In particular, the general instructions on how to gather and analyze information (survey tasks 5 and 6) should be consolidated. After all, why shouldn't the survey process reflect the same techniques of information analysis and problem solving that apply to care processes and quality-improvement initiatives?

Surveyors should be given explicit steps to follow for their reviews. Similarly, the SOM should organize process indicators based on valid evidence for all areas under review (see "Simple Fixes for the OBRA Regulatory Process," in the December 2001 Caring, p. 32) into specific steps. Clinically based investigation should be based on consistent clinical guidelines. For example, reviewing for proper management of falls and fall risk would include the following steps, among others.

Determine whether the facility staff:

1. Identified individuals at risk for falling
2. Reviewed and described actual falls
3. Identified causes of falling, based on factors specific to that individual
4. Instituted cause-specific interventions to prevent or minimize resident fall risk, falls, and complications from falls, or addressed the expectation of a continued risk when cause-specific interventions or adjustments cannot be made.

The SOM should also group investigative criteria and questions ("probes") under these steps. For example, step 3 above should include such questions as: "Did facility staff evaluate an individual's gait and balance?" and "Did facility staff review an individual's medication regimen for a possible association with falling?"

The rules of evidence in the survey process should be consistent, not situational. Surveyor guidelines should tell surveyors to collect enough information to answer such questions, and clarify how to decide when a given observation or collection of facts warrant a particular conclusion. Instructions should not encourage surveyors to draw premature conclusions or let them selectively bypass the aforementioned processes to draw conclusions.

True, one or several facts may indicate that a problem exists, and sometimes, facts can be interpreted in only one way. But in everyday life, initial suspicions are not always verified, and there may be several possible explanations for identical observations.

Isolated findings or situations rarely prove poor care in and of themselves, without requiring further investigation. Unless there is general agreement that certain information invariably means only certain things, surveyors should be required to consider all relevant information and say why alternate explanations are not plausible or relevant.

When facts could reflect several possible explanations, depending on the evidence, it is inappropriate to jump to conclusions without first considering alternative explanations and other information that might provide a different context.

Finally, surveyors need to understand that bad outcomes do not justify investigative shortcuts. The SOM should state that the same evidentiary methods be used regardless of the severity of a result.

Survey instructions should not be biased towards the notion that serious outcomes are more likely to be someone's fault and therefore less evidence is needed to prove the case. That notion contradicts a basic principle of our legal system: that more serious accusations carrying potentially more serious consequences require more--not less--proof. A serious result may not be anyone's fault, whereas a less serious result may reflect grossly inadequate care. Determining accountability for a result should not be confused with deciding whether performance was adequate.

Change the Approach to Identifying Negative Outcomes

The SOM and Interpretive Guidelines speak repeatedly about "negative outcomes." Negative outcomes are defined as situations in which an individual experiences either a decline in condition or function or fails to improve or to achieve his or her highest practicable outcome. Again, the surveyor guidance offers specific examples of negative outcomes but does not provide a general procedure for identifying them.

The fact is, there should be limits on defining negative outcomes. The regulatory process must recognize in practice--not just on paper--that negative outcomes often are not preventable, and should concentrate primarily on important situations where outcomes realistically could be improved.

Surveyor guidance should also limit the scope of negative outcomes that the review process scrutinizes. Certain bad outcomes--abuse, neglect, or causing preventable functional decline by using problematic medications inappropriately--are unacceptable. But currently, survey guidelines seem in places to encourage surveyors to cast a very broad net, rather than focusing attention on negative outcomes of genuine concern. In fact, they let surveyors use limited or circumstantial evidence to decide that some situations constitute negative outcomes, especially with respect to quality-of-life issues.

The realities of human existence apply in nursing homes, as elsewhere. We all experience such "negative outcomes" as falling, conflict, anxiety, embarrassment, pain, illness, and so on. Similarly, individuals with multiple physical, functional, and psychiatric impairments who live in institutions inevitably have some unpleasant or disruptive experiences.

A resident's individual experience should not by itself guide a search for fault finding, but serve only as a clue to investigate further for possibly significant systemic deficits.

The bottom line is that surveys should concentrate on important, potentially preventable negative outcomes that occur due to correctable process problems, rather than on usual and customary disturbances that most people experience in similar circumstances in everyday life.

Revamp the Notion of "Potential" Negative Outcomes

A whole level of compliance determination is based on conclusions about "potential for negative outcomes"--that is, results that did not occur but were allegedly more likely because of a faulty or deficient practice.

Certainly, some identifiable interventions may effectively mitigate, if not prevent, undesired outcomes in frail older and other chronically ill individuals. But the ability to link current actions to future results in specific cases requires a high level of knowledge and experience. Risk-management publications tell us that every year, there are hundreds of thousands of opportunities for things to go wrong in nursing homes---but in most cases, things go right. The SOM contains little practical guidance in this area of "potential" negative outcomes--leaving it wide open to speculation and opinion, and fraught with problems.

Improving the judgment and skills of surveyors in determining when process problems might produce potentially negative outcomes is a daunting task. Instead of embarking on such an undertaking, it makes more sense to use appropriate process indicators to replace the whole confusing concept.

The survey guidelines should provide relevant process indicators and set a threshold for process compliance. Then, it would be much easier to identify situations in which processes are potentially inadequate, and focus attention on correcting inadequate processes that significantly increase the risk for subsequent negative outcomes.

Change How Surveyors Link Processes & Outcomes

Consequences of the survey process (primarily, penalties related to finding deficiencies and required plans of correction) depend heavily on surveyor determinations of whether facility processes and practices caused or contributed to negative outcomes.

Generally, the conclusion that faulty facility processes or practices caused or contributed to negative outcomes ("failure to improve or maintain function") results in more serious sanctions and consequences than "medically unavoidable" outcomes (resulting from a resident's underlying conditions, problems, or risks). Thus, an appropriate determination of "avoidability" depends on correctly identifying both negative outcomes and faulty processes or practices.

The Guidance to Surveyors should incorporate--not create anew--approved care processes and practices, and guide surveyors more clearly about identifying a link between inadequate processes and negative outcomes. The current instructions lack clear, orderly process expectations. They gloss over the issue of avoidable vs. unavoidable outcomes when they refer to "accepted standards of care" as though everyone automatically knows what those are.

In fact, disagreements over what is "expected" or "acceptable" practice are at the center of most disputes between states and providers involving quality-of-care issues.

For example, failure to complete a portion of the MDS or order a lab test are unlikely by themselves to cause significant unplanned weight loss, although surveyors may make such a connection based on investigative protocols. On the other hand, failing to identify medications and treatable conditions associated with anorexia may well contribute to weight loss, but surveyors may fail to make such a link because they don't look or don't recognize the significance of such findings.

To differentiate "avoidable" from "unavoidable" outcomes, surveyors need clear instructions--not just simple criteria or case examples--for evaluating specific situations. The conclusion that something was "avoidable" should be supported by evidence that:

• Staff and practitioners did not do what they should have done, did what they should not have done, or did not do something correctly.
• As a direct result of the above, something happened to the individual that otherwise would not likely have happened.
• Staff and practitioners could reasonably have been expected to act differently, and that acting differently would most likely produce a different result.

We need to do away with the insidious notion that all deficiencies are equally wrong and that the only place for differentiating them is in determining their scope and severity. Instead, surveyor instructions in this area should reinforce the following points:

• Negative outcomes are not necessarily anyone's fault; the fact that a negative outcome is serious does not automatically make it more likely that a faulty practice caused it.
• Finding a process error or omission and a negative outcome does not alone prove that the practice caused the outcome.
• Minor or intermittent errors or omissions are much less likely to cause negative outcomes than recurrent or major process problems; finding that staff omitted many important steps of a desired process is very different from finding that one or several steps were omitted or done incorrectly on isolated occasions.
• An error that is detected in time to prevent a cascade of subsequent complications is very different from a series of errors or omissions in a single process that fail to lead to timely adjustments.

Change the Approach to Surveyor Training & Oversight

Nursing facility staff and practitioners carry the major responsibility for improving long-term care practices. But the survey process-- and surveyor actions and conclusions--have a major influence on that performance. Both care providers and care reviewers such as surveyors must use similar investigative and problem-solving methods and recognize the same process expectations.

Like facility staff and practitioners, surveyors need clear, consistent training, detailed procedural guidance and oversight, and feedback about their performance. The variability in surveyor and state-agency performance--like the variability in care itself--reflects uncontrolled differences in the implementation of a complex investigative process that involves gathering and interpreting a great deal of technical information.

Steven Levenson, MD, CMD, is a Multi-Facility Medical Director in Baltimore, MD, and Chair of Caring's Editorial Board.

back to top